* etc/publicsuffix.txt: Update from source.
[bpt/emacs.git] / lisp / progmodes / cc-engine.el
CommitLineData
c38e0c97 1;;; cc-engine.el --- core syntax guessing engine for CC mode -*- coding: utf-8 -*-
785eecbb 2
ba318903 3;; Copyright (C) 1985, 1987, 1992-2014 Free Software Foundation, Inc.
785eecbb 4
e309f66c
AM
5;; Authors: 2001- Alan Mackenzie
6;; 1998- Martin Stjernholm
d9e94c22 7;; 1992-1999 Barry A. Warsaw
5858f68c
GM
8;; 1987 Dave Detlefs
9;; 1987 Stewart Clamen
785eecbb 10;; 1985 Richard M. Stallman
0ec8351b 11;; Maintainer: bug-cc-mode@gnu.org
785eecbb 12;; Created: 22-Apr-1997 (split from cc-mode.el)
bd78fa1d
CY
13;; Keywords: c languages
14;; Package: cc-mode
785eecbb
RS
15
16;; This file is part of GNU Emacs.
17
b1fc2b50 18;; GNU Emacs is free software: you can redistribute it and/or modify
785eecbb 19;; it under the terms of the GNU General Public License as published by
b1fc2b50
GM
20;; the Free Software Foundation, either version 3 of the License, or
21;; (at your option) any later version.
785eecbb
RS
22
23;; GNU Emacs is distributed in the hope that it will be useful,
24;; but WITHOUT ANY WARRANTY; without even the implied warranty of
25;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
26;; GNU General Public License for more details.
27
28;; You should have received a copy of the GNU General Public License
b1fc2b50 29;; along with GNU Emacs. If not, see <http://www.gnu.org/licenses/>.
785eecbb 30
3afbc435
PJ
31;;; Commentary:
32
a66cd3ee
MS
33;; The functions which have docstring documentation can be considered
34;; part of an API which other packages can use in CC Mode buffers.
35;; Otoh, undocumented functions and functions with the documentation
36;; in comments are considered purely internal and can change semantics
37;; or even disappear in the future.
38;;
39;; (This policy applies to CC Mode as a whole, not just this file. It
40;; probably also applies to many other Emacs packages, but here it's
41;; clearly spelled out.)
42
d9e94c22
MS
43;; Hidden buffer changes
44;;
45;; Various functions in CC Mode use text properties for caching and
46;; syntactic markup purposes, and those of them that might modify such
0386b551
AM
47;; properties but still don't modify the buffer in a visible way are
48;; said to do "hidden buffer changes". They should be used within
49;; `c-save-buffer-state' or a similar function that saves and restores
50;; buffer modifiedness, disables buffer change hooks, etc.
d9e94c22 51;;
0386b551
AM
52;; Interactive functions are assumed to not do hidden buffer changes,
53;; except in the specific parts of them that do real changes.
d9e94c22 54;;
0386b551
AM
55;; Lineup functions are assumed to do hidden buffer changes. They
56;; must not do real changes, though.
d9e94c22 57;;
0386b551
AM
58;; All other functions that do hidden buffer changes have that noted
59;; in their doc string or comment.
60;;
61;; The intention with this system is to avoid wrapping every leaf
62;; function that do hidden buffer changes inside
63;; `c-save-buffer-state'. It should be used as near the top of the
64;; interactive functions as possible.
65;;
66;; Functions called during font locking are allowed to do hidden
67;; buffer changes since the font-lock package run them in a context
68;; similar to `c-save-buffer-state' (in fact, that function is heavily
69;; inspired by `save-buffer-state' in the font-lock package).
d9e94c22
MS
70
71;; Use of text properties
72;;
73;; CC Mode uses several text properties internally to mark up various
74;; positions, e.g. to improve speed and to eliminate glitches in
75;; interactive refontification.
76;;
7bfc3fdb
MS
77;; Note: This doc is for internal use only. Other packages should not
78;; assume that these text properties are used as described here.
79;;
0ec1d2c5
AM
80;; 'category
81;; Used for "indirection". With its help, some other property can
82;; be cheaply and easily switched on or off everywhere it occurs.
83;;
d9e94c22 84;; 'syntax-table
c8018ede
AM
85;; Used to modify the syntax of some characters. It is used to
86;; mark the "<" and ">" of angle bracket parens with paren syntax, and
87;; to "hide" obtrusive characters in preprocessor lines.
d9e94c22
MS
88;;
89;; This property is used on single characters and is therefore
90;; always treated as front and rear nonsticky (or start and end open
91;; in XEmacs vocabulary). It's therefore installed on
92;; `text-property-default-nonsticky' if that variable exists (Emacs
93;; >= 21).
94;;
95;; 'c-is-sws and 'c-in-sws
96;; Used by `c-forward-syntactic-ws' and `c-backward-syntactic-ws' to
97;; speed them up. See the comment blurb before `c-put-is-sws'
98;; below for further details.
99;;
100;; 'c-type
101;; This property is used on single characters to mark positions with
0386b551
AM
102;; special syntactic relevance of various sorts. Its primary use is
103;; to avoid glitches when multiline constructs are refontified
d9e94c22
MS
104;; interactively (on font lock decoration level 3). It's cleared in
105;; a region before it's fontified and is then put on relevant chars
106;; in that region as they are encountered during the fontification.
107;; The value specifies the kind of position:
108;;
109;; 'c-decl-arg-start
110;; Put on the last char of the token preceding each declaration
111;; inside a declaration style arglist (typically in a function
112;; prototype).
113;;
114;; 'c-decl-end
115;; Put on the last char of the token preceding a declaration.
116;; This is used in cases where declaration boundaries can't be
117;; recognized simply by looking for a token like ";" or "}".
118;; `c-type-decl-end-used' must be set if this is used (see also
119;; `c-find-decl-spots').
120;;
121;; 'c-<>-arg-sep
122;; Put on the commas that separate arguments in angle bracket
123;; arglists like C++ template arglists.
124;;
125;; 'c-decl-id-start and 'c-decl-type-start
126;; Put on the last char of the token preceding each declarator
127;; in the declarator list of a declaration. They are also used
128;; between the identifiers cases like enum declarations.
129;; 'c-decl-type-start is used when the declarators are types,
130;; 'c-decl-id-start otherwise.
131;;
132;; 'c-awk-NL-prop
133;; Used in AWK mode to mark the various kinds of newlines. See
134;; cc-awk.el.
135
3afbc435
PJ
136;;; Code:
137
0ec8351b 138(eval-when-compile
51f606de 139 (let ((load-path
130c507e
GM
140 (if (and (boundp 'byte-compile-dest-file)
141 (stringp byte-compile-dest-file))
142 (cons (file-name-directory byte-compile-dest-file) load-path)
51f606de 143 load-path)))
d9e94c22 144 (load "cc-bytecomp" nil t)))
130c507e
GM
145
146(cc-require 'cc-defs)
d9e94c22 147(cc-require-when-compile 'cc-langs)
130c507e 148(cc-require 'cc-vars)
d9e94c22 149
51f606de 150\f
d9e94c22
MS
151;; Make declarations for all the `c-lang-defvar' variables in cc-langs.
152
153(defmacro c-declare-lang-variables ()
154 `(progn
485fe977
RS
155 ,@(apply 'nconc
156 (mapcar (lambda (init)
157 `(,(if (elt init 2)
158 `(defvar ,(car init) nil ,(elt init 2))
159 `(defvar ,(car init) nil))
160 (make-variable-buffer-local ',(car init))))
161 (cdr c-lang-variable-inits)))))
d9e94c22
MS
162(c-declare-lang-variables)
163
164\f
165;;; Internal state variables.
166
167;; Internal state of hungry delete key feature
168(defvar c-hungry-delete-key nil)
169(make-variable-buffer-local 'c-hungry-delete-key)
170
0386b551
AM
171;; The electric flag (toggled by `c-toggle-electric-state').
172;; If t, electric actions (like automatic reindentation, and (if
173;; c-auto-newline is also set) auto newlining) will happen when an electric
174;; key like `{' is pressed (or an electric keyword like `else').
175(defvar c-electric-flag t)
176(make-variable-buffer-local 'c-electric-flag)
177
d9e94c22
MS
178;; Internal state of auto newline feature.
179(defvar c-auto-newline nil)
180(make-variable-buffer-local 'c-auto-newline)
181
0386b551 182;; Included in the mode line to indicate the active submodes.
cb694ab7
AM
183;; (defvar c-submode-indicators nil)
184;; (make-variable-buffer-local 'c-submode-indicators)
d9e94c22 185
a66cd3ee
MS
186(defun c-calculate-state (arg prevstate)
187 ;; Calculate the new state of PREVSTATE, t or nil, based on arg. If
188 ;; arg is nil or zero, toggle the state. If arg is negative, turn
189 ;; the state off, and if arg is positive, turn the state on
190 (if (or (not arg)
191 (zerop (setq arg (prefix-numeric-value arg))))
192 (not prevstate)
193 (> arg 0)))
194
d9e94c22 195\f
037558bf
MS
196;; Basic handling of preprocessor directives.
197
198;; This is a dynamically bound cache used together with
199;; `c-query-macro-start' and `c-query-and-set-macro-start'. It only
200;; works as long as point doesn't cross a macro boundary.
201(defvar c-macro-start 'unknown)
202
203(defsubst c-query-and-set-macro-start ()
037558bf
MS
204 (if (symbolp c-macro-start)
205 (setq c-macro-start (save-excursion
0386b551
AM
206 (c-save-buffer-state ()
207 (and (c-beginning-of-macro)
208 (point)))))
037558bf
MS
209 c-macro-start))
210
211(defsubst c-query-macro-start ()
037558bf
MS
212 (if (symbolp c-macro-start)
213 (save-excursion
0386b551
AM
214 (c-save-buffer-state ()
215 (and (c-beginning-of-macro)
216 (point))))
037558bf
MS
217 c-macro-start))
218
56d093a9
AM
219;; One element macro cache to cope with continual movement within very large
220;; CPP macros.
221(defvar c-macro-cache nil)
222(make-variable-buffer-local 'c-macro-cache)
223;; Nil or cons of the bounds of the most recent CPP form probed by
224;; `c-beginning-of-macro', `c-end-of-macro' or `c-syntactic-end-of-macro'.
225;; The cdr will be nil if we know only the start of the CPP form.
226(defvar c-macro-cache-start-pos nil)
227(make-variable-buffer-local 'c-macro-cache-start-pos)
228;; The starting position from where we determined `c-macro-cache'.
229(defvar c-macro-cache-syntactic nil)
230(make-variable-buffer-local 'c-macro-cache-syntactic)
231;; non-nil iff `c-macro-cache' has both elements set AND the cdr is at a
232;; syntactic end of macro, not merely an apparent one.
233
234(defun c-invalidate-macro-cache (beg end)
235 ;; Called from a before-change function. If the change region is before or
f003f294 236 ;; in the macro characterized by `c-macro-cache' etc., nullify it
56d093a9
AM
237 ;; appropriately. BEG and END are the standard before-change-functions
238 ;; parameters. END isn't used.
239 (cond
240 ((null c-macro-cache))
241 ((< beg (car c-macro-cache))
242 (setq c-macro-cache nil
243 c-macro-cache-start-pos nil
244 c-macro-cache-syntactic nil))
245 ((and (cdr c-macro-cache)
246 (< beg (cdr c-macro-cache)))
247 (setcdr c-macro-cache nil)
248 (setq c-macro-cache-start-pos beg
249 c-macro-cache-syntactic nil))))
250
037558bf
MS
251(defun c-beginning-of-macro (&optional lim)
252 "Go to the beginning of a preprocessor directive.
253Leave point at the beginning of the directive and return t if in one,
254otherwise return nil and leave point unchanged.
255
0386b551
AM
256Note that this function might do hidden buffer changes. See the
257comment at the start of cc-engine.el for more info."
56d093a9
AM
258 (let ((here (point)))
259 (when c-opt-cpp-prefix
260 (if (and (car c-macro-cache)
261 (>= (point) (car c-macro-cache))
262 (or (and (cdr c-macro-cache)
263 (<= (point) (cdr c-macro-cache)))
264 (<= (point) c-macro-cache-start-pos)))
265 (unless (< (car c-macro-cache) (or lim (point-min)))
266 (progn (goto-char (max (or lim (point-min)) (car c-macro-cache)))
267 (setq c-macro-cache-start-pos
268 (max c-macro-cache-start-pos here))
269 t))
270 (setq c-macro-cache nil
271 c-macro-cache-start-pos nil
272 c-macro-cache-syntactic nil)
273
274 (save-restriction
275 (if lim (narrow-to-region lim (point-max)))
276 (beginning-of-line)
277 (while (eq (char-before (1- (point))) ?\\)
278 (forward-line -1))
279 (back-to-indentation)
280 (if (and (<= (point) here)
281 (looking-at c-opt-cpp-start))
282 (progn
283 (setq c-macro-cache (cons (point) nil)
284 c-macro-cache-start-pos here)
285 t)
286 (goto-char here)
287 nil))))))
037558bf
MS
288
289(defun c-end-of-macro ()
290 "Go to the end of a preprocessor directive.
0386b551
AM
291More accurately, move the point to the end of the closest following
292line that doesn't end with a line continuation backslash - no check is
293done that the point is inside a cpp directive to begin with.
037558bf 294
0386b551
AM
295Note that this function might do hidden buffer changes. See the
296comment at the start of cc-engine.el for more info."
56d093a9
AM
297 (if (and (cdr c-macro-cache)
298 (<= (point) (cdr c-macro-cache))
299 (>= (point) (car c-macro-cache)))
300 (goto-char (cdr c-macro-cache))
301 (unless (and (car c-macro-cache)
302 (<= (point) c-macro-cache-start-pos)
303 (>= (point) (car c-macro-cache)))
304 (setq c-macro-cache nil
305 c-macro-cache-start-pos nil
306 c-macro-cache-syntactic nil))
307 (while (progn
308 (end-of-line)
309 (when (and (eq (char-before) ?\\)
310 (not (eobp)))
311 (forward-char)
312 t)))
313 (when (car c-macro-cache)
314 (setcdr c-macro-cache (point)))))
037558bf 315
0ec1d2c5
AM
316(defun c-syntactic-end-of-macro ()
317 ;; Go to the end of a CPP directive, or a "safe" pos just before.
318 ;;
319 ;; This is normally the end of the next non-escaped line. A "safe"
320 ;; position is one not within a string or comment. (The EOL on a line
321 ;; comment is NOT "safe").
322 ;;
323 ;; This function must only be called from the beginning of a CPP construct.
324 ;;
325 ;; Note that this function might do hidden buffer changes. See the comment
326 ;; at the start of cc-engine.el for more info.
327 (let* ((here (point))
328 (there (progn (c-end-of-macro) (point)))
56d093a9
AM
329 s)
330 (unless c-macro-cache-syntactic
331 (setq s (parse-partial-sexp here there))
332 (while (and (or (nth 3 s) ; in a string
333 (nth 4 s)) ; in a comment (maybe at end of line comment)
334 (> there here)) ; No infinite loops, please.
335 (setq there (1- (nth 8 s)))
336 (setq s (parse-partial-sexp here there)))
337 (setq c-macro-cache-syntactic (car c-macro-cache)))
0ec1d2c5
AM
338 (point)))
339
51c9af45
AM
340(defun c-forward-over-cpp-define-id ()
341 ;; Assuming point is at the "#" that introduces a preprocessor
342 ;; directive, it's moved forward to the end of the identifier which is
343 ;; "#define"d (or whatever c-opt-cpp-macro-define specifies). Non-nil
344 ;; is returned in this case, in all other cases nil is returned and
345 ;; point isn't moved.
346 ;;
347 ;; This function might do hidden buffer changes.
348 (when (and c-opt-cpp-macro-define-id
349 (looking-at c-opt-cpp-macro-define-id))
350 (goto-char (match-end 0))))
351
037558bf
MS
352(defun c-forward-to-cpp-define-body ()
353 ;; Assuming point is at the "#" that introduces a preprocessor
354 ;; directive, it's moved forward to the start of the definition body
0386b551
AM
355 ;; if it's a "#define" (or whatever c-opt-cpp-macro-define
356 ;; specifies). Non-nil is returned in this case, in all other cases
357 ;; nil is returned and point isn't moved.
358 ;;
359 ;; This function might do hidden buffer changes.
360 (when (and c-opt-cpp-macro-define-start
361 (looking-at c-opt-cpp-macro-define-start)
037558bf
MS
362 (not (= (match-end 0) (c-point 'eol))))
363 (goto-char (match-end 0))))
364
365\f
d9e94c22
MS
366;;; Basic utility functions.
367
0386b551 368(defun c-syntactic-content (from to paren-level)
d9e94c22
MS
369 ;; Return the given region as a string where all syntactic
370 ;; whitespace is removed or, where necessary, replaced with a single
0386b551
AM
371 ;; space. If PAREN-LEVEL is given then all parens in the region are
372 ;; collapsed to "()", "[]" etc.
373 ;;
374 ;; This function might do hidden buffer changes.
375
d9e94c22 376 (save-excursion
0386b551
AM
377 (save-restriction
378 (narrow-to-region from to)
379 (goto-char from)
380 (let* ((parts (list nil)) (tail parts) pos in-paren)
381
382 (while (re-search-forward c-syntactic-ws-start to t)
383 (goto-char (setq pos (match-beginning 0)))
384 (c-forward-syntactic-ws)
385 (if (= (point) pos)
386 (forward-char)
387
388 (when paren-level
389 (save-excursion
390 (setq in-paren (= (car (parse-partial-sexp from pos 1)) 1)
391 pos (point))))
392
393 (if (and (> pos from)
394 (< (point) to)
395 (looking-at "\\w\\|\\s_")
396 (save-excursion
397 (goto-char (1- pos))
398 (looking-at "\\w\\|\\s_")))
399 (progn
400 (setcdr tail (list (buffer-substring-no-properties from pos)
401 " "))
402 (setq tail (cddr tail)))
403 (setcdr tail (list (buffer-substring-no-properties from pos)))
404 (setq tail (cdr tail)))
405
406 (when in-paren
407 (when (= (car (parse-partial-sexp pos to -1)) -1)
408 (setcdr tail (list (buffer-substring-no-properties
409 (1- (point)) (point))))
410 (setq tail (cdr tail))))
411
412 (setq from (point))))
413
414 (setcdr tail (list (buffer-substring-no-properties from to)))
415 (apply 'concat (cdr parts))))))
416
417(defun c-shift-line-indentation (shift-amt)
418 ;; Shift the indentation of the current line with the specified
419 ;; amount (positive inwards). The buffer is modified only if
420 ;; SHIFT-AMT isn't equal to zero.
421 (let ((pos (- (point-max) (point)))
422 (c-macro-start c-macro-start)
423 tmp-char-inserted)
424 (if (zerop shift-amt)
425 nil
426 ;; If we're on an empty line inside a macro, we take the point
427 ;; to be at the current indentation and shift it to the
428 ;; appropriate column. This way we don't treat the extra
429 ;; whitespace out to the line continuation as indentation.
430 (when (and (c-query-and-set-macro-start)
431 (looking-at "[ \t]*\\\\$")
432 (save-excursion
433 (skip-chars-backward " \t")
434 (bolp)))
435 (insert ?x)
436 (backward-char)
437 (setq tmp-char-inserted t))
438 (unwind-protect
439 (let ((col (current-indentation)))
440 (delete-region (c-point 'bol) (c-point 'boi))
441 (beginning-of-line)
442 (indent-to (+ col shift-amt)))
443 (when tmp-char-inserted
444 (delete-char 1))))
445 ;; If initial point was within line's indentation and we're not on
446 ;; a line with a line continuation in a macro, position after the
447 ;; indentation. Else stay at same point in text.
448 (if (and (< (point) (c-point 'boi))
449 (not tmp-char-inserted))
450 (back-to-indentation)
451 (if (> (- (point-max) pos) (point))
452 (goto-char (- (point-max) pos))))))
d9e94c22
MS
453
454(defsubst c-keyword-sym (keyword)
455 ;; Return non-nil if the string KEYWORD is a known keyword. More
456 ;; precisely, the value is the symbol for the keyword in
457 ;; `c-keywords-obarray'.
458 (intern-soft keyword c-keywords-obarray))
459
460(defsubst c-keyword-member (keyword-sym lang-constant)
461 ;; Return non-nil if the symbol KEYWORD-SYM, as returned by
462 ;; `c-keyword-sym', is a member of LANG-CONSTANT, which is the name
463 ;; of a language constant that ends with "-kwds". If KEYWORD-SYM is
464 ;; nil then the result is nil.
465 (get keyword-sym lang-constant))
466
467;; String syntax chars, suitable for skip-syntax-(forward|backward).
468(defconst c-string-syntax (if (memq 'gen-string-delim c-emacs-features)
469 "\"|"
470 "\""))
471
0386b551 472;; Regexp matching string limit syntax.
d9e94c22
MS
473(defconst c-string-limit-regexp (if (memq 'gen-string-delim c-emacs-features)
474 "\\s\"\\|\\s|"
475 "\\s\""))
476
0386b551
AM
477;; Regexp matching WS followed by string limit syntax.
478(defconst c-ws*-string-limit-regexp
479 (concat "[ \t]*\\(" c-string-limit-regexp "\\)"))
480
d9e94c22
MS
481;; Holds formatted error strings for the few cases where parse errors
482;; are reported.
a66cd3ee 483(defvar c-parsing-error nil)
d9e94c22
MS
484(make-variable-buffer-local 'c-parsing-error)
485
486(defun c-echo-parsing-error (&optional quiet)
d9e94c22
MS
487 (when (and c-report-syntactic-errors c-parsing-error (not quiet))
488 (c-benign-error "%s" c-parsing-error))
489 c-parsing-error)
490
491;; Faces given to comments and string literals. This is used in some
492;; situations to speed up recognition; it isn't mandatory that font
493;; locking is in use. This variable is extended with the face in
494;; `c-doc-face-name' when fontification is activated in cc-fonts.el.
1e330469 495(defvar c-literal-faces
0386b551
AM
496 (append '(font-lock-comment-face font-lock-string-face)
497 (when (facep 'font-lock-comment-delimiter-face)
498 ;; New in Emacs 22.
499 '(font-lock-comment-delimiter-face))))
500
501(defsubst c-put-c-type-property (pos value)
502 ;; Put a c-type property with the given value at POS.
503 (c-put-char-property pos 'c-type value))
504
505(defun c-clear-c-type-property (from to value)
5a89f0a7 506 ;; Remove all occurrences of the c-type property that has the given
0386b551
AM
507 ;; value in the region between FROM and TO. VALUE is assumed to not
508 ;; be nil.
509 ;;
510 ;; Note: This assumes that c-type is put on single chars only; it's
511 ;; very inefficient if matching properties cover large regions.
512 (save-excursion
513 (goto-char from)
514 (while (progn
515 (when (eq (get-text-property (point) 'c-type) value)
516 (c-clear-char-property (point) 'c-type))
517 (goto-char (next-single-property-change (point) 'c-type nil to))
518 (< (point) to)))))
037558bf 519
d9e94c22
MS
520\f
521;; Some debug tools to visualize various special positions. This
522;; debug code isn't as portable as the rest of CC Mode.
523
524(cc-bytecomp-defun overlays-in)
525(cc-bytecomp-defun overlay-get)
526(cc-bytecomp-defun overlay-start)
527(cc-bytecomp-defun overlay-end)
528(cc-bytecomp-defun delete-overlay)
529(cc-bytecomp-defun overlay-put)
530(cc-bytecomp-defun make-overlay)
531
532(defun c-debug-add-face (beg end face)
533 (c-save-buffer-state ((overlays (overlays-in beg end)) overlay)
534 (while overlays
535 (setq overlay (car overlays)
536 overlays (cdr overlays))
537 (when (eq (overlay-get overlay 'face) face)
538 (setq beg (min beg (overlay-start overlay))
539 end (max end (overlay-end overlay)))
540 (delete-overlay overlay)))
541 (overlay-put (make-overlay beg end) 'face face)))
542
543(defun c-debug-remove-face (beg end face)
544 (c-save-buffer-state ((overlays (overlays-in beg end)) overlay
545 (ol-beg beg) (ol-end end))
546 (while overlays
547 (setq overlay (car overlays)
548 overlays (cdr overlays))
549 (when (eq (overlay-get overlay 'face) face)
550 (setq ol-beg (min ol-beg (overlay-start overlay))
551 ol-end (max ol-end (overlay-end overlay)))
552 (delete-overlay overlay)))
553 (when (< ol-beg beg)
554 (overlay-put (make-overlay ol-beg beg) 'face face))
555 (when (> ol-end end)
556 (overlay-put (make-overlay end ol-end) 'face face))))
557
558\f
559;; `c-beginning-of-statement-1' and accompanying stuff.
130c507e 560
64001211
RS
561;; KLUDGE ALERT: c-maybe-labelp is used to pass information between
562;; c-crosses-statement-barrier-p and c-beginning-of-statement-1. A
563;; better way should be implemented, but this will at least shut up
564;; the byte compiler.
0386b551 565(defvar c-maybe-labelp)
64001211 566
d9e94c22
MS
567;; New awk-compatible version of c-beginning-of-statement-1, ACM 2002/6/22
568
a66cd3ee
MS
569;; Macros used internally in c-beginning-of-statement-1 for the
570;; automaton actions.
571(defmacro c-bos-push-state ()
572 '(setq stack (cons (cons state saved-pos)
573 stack)))
574(defmacro c-bos-pop-state (&optional do-if-done)
575 `(if (setq state (car (car stack))
576 saved-pos (cdr (car stack))
577 stack (cdr stack))
578 t
579 ,do-if-done
580 (throw 'loop nil)))
581(defmacro c-bos-pop-state-and-retry ()
582 '(throw 'loop (setq state (car (car stack))
583 saved-pos (cdr (car stack))
584 ;; Throw nil if stack is empty, else throw non-nil.
585 stack (cdr stack))))
586(defmacro c-bos-save-pos ()
587 '(setq saved-pos (vector pos tok ptok pptok)))
588(defmacro c-bos-restore-pos ()
589 '(unless (eq (elt saved-pos 0) start)
590 (setq pos (elt saved-pos 0)
591 tok (elt saved-pos 1)
592 ptok (elt saved-pos 2)
593 pptok (elt saved-pos 3))
594 (goto-char pos)
595 (setq sym nil)))
596(defmacro c-bos-save-error-info (missing got)
597 `(setq saved-pos (vector pos ,missing ,got)))
598(defmacro c-bos-report-error ()
599 '(unless noerror
600 (setq c-parsing-error
601 (format "No matching `%s' found for `%s' on line %d"
602 (elt saved-pos 1)
603 (elt saved-pos 2)
604 (1+ (count-lines (point-min)
605 (c-point 'bol (elt saved-pos 0))))))))
606
607(defun c-beginning-of-statement-1 (&optional lim ignore-labels
608 noerror comma-delim)
609 "Move to the start of the current statement or declaration, or to
610the previous one if already at the beginning of one. Only
611statements/declarations on the same level are considered, i.e. don't
612move into or out of sexps (not even normal expression parentheses).
613
5a89f0a7 614If point is already at the earliest statement within braces or parens,
a85fd6da
AM
615this function doesn't move back into any whitespace preceding it; it
616returns 'same in this case.
617
d9e94c22
MS
618Stop at statement continuation tokens like \"else\", \"catch\",
619\"finally\" and the \"while\" in \"do ... while\" if the start point
620is within the continuation. If starting at such a token, move to the
621corresponding statement start. If at the beginning of a statement,
622move to the closest containing statement if there is any. This might
623also stop at a continuation clause.
a66cd3ee 624
0386b551
AM
625Labels are treated as part of the following statements if
626IGNORE-LABELS is non-nil. (FIXME: Doesn't work if we stop at a known
cb694ab7
AM
627statement start keyword.) Otherwise, each label is treated as a
628separate statement.
a66cd3ee 629
cb694ab7
AM
630Macros are ignored \(i.e. skipped over) unless point is within one, in
631which case the content of the macro is treated as normal code. Aside
632from any normal statement starts found in it, stop at the first token
633of the content in the macro, i.e. the expression of an \"#if\" or the
634start of the definition in a \"#define\". Also stop at start of
635macros before leaving them.
a66cd3ee 636
a85fd6da 637Return:
d28e7f28 638'label if stopped at a label or \"case...:\" or \"default:\";
a85fd6da
AM
639'same if stopped at the beginning of the current statement;
640'up if stepped to a containing statement;
641'previous if stepped to a preceding statement;
642'beginning if stepped from a statement continuation clause to
643 its start clause; or
644'macro if stepped to a macro start.
645Note that 'same and not 'label is returned if stopped at the same
646label without crossing the colon character.
a66cd3ee
MS
647
648LIM may be given to limit the search. If the search hits the limit,
649point will be left at the closest following token, or at the start
650position if that is less ('same is returned in this case).
651
652NOERROR turns off error logging to `c-parsing-error'.
653
cb694ab7
AM
654Normally only ';' and virtual semicolons are considered to delimit
655statements, but if COMMA-DELIM is non-nil then ',' is treated
656as a delimiter too.
0386b551
AM
657
658Note that this function might do hidden buffer changes. See the
659comment at the start of cc-engine.el for more info."
a66cd3ee 660
d9e94c22
MS
661 ;; The bulk of this function is a pushdown automaton that looks at statement
662 ;; boundaries and the tokens (such as "while") in c-opt-block-stmt-key. Its
663 ;; purpose is to keep track of nested statements, ensuring that such
5a89f0a7 664 ;; statements are skipped over in their entirety (somewhat akin to what C-M-p
d9e94c22 665 ;; does with nested braces/brackets/parentheses).
a66cd3ee
MS
666 ;;
667 ;; Note: The position of a boundary is the following token.
668 ;;
d9e94c22
MS
669 ;; Beginning with the current token (the one following point), move back one
670 ;; sexp at a time (where a sexp is, more or less, either a token or the
671 ;; entire contents of a brace/bracket/paren pair). Each time a statement
672 ;; boundary is crossed or a "while"-like token is found, update the state of
673 ;; the PDA. Stop at the beginning of a statement when the stack (holding
674 ;; nested statement info) is empty and the position has been moved.
675 ;;
676 ;; The following variables constitute the PDA:
677 ;;
678 ;; sym: This is either the "while"-like token (e.g. 'for) we've just
679 ;; scanned back over, 'boundary if we've just gone back over a
680 ;; statement boundary, or nil otherwise.
681 ;; state: takes one of the values (nil else else-boundary while
682 ;; while-boundary catch catch-boundary).
683 ;; nil means "no "while"-like token yet scanned".
684 ;; 'else, for example, means "just gone back over an else".
685 ;; 'else-boundary means "just gone back over a statement boundary
686 ;; immediately after having gone back over an else".
687 ;; saved-pos: A vector of either saved positions (tok ptok pptok, etc.) or
688 ;; of error reporting information.
689 ;; stack: The stack onto which the PDA pushes its state. Each entry
690 ;; consists of a saved value of state and saved-pos. An entry is
691 ;; pushed when we move back over a "continuation" token (e.g. else)
692 ;; and popped when we encounter the corresponding opening token
693 ;; (e.g. if).
694 ;;
695 ;;
b414f371 696 ;; The following diagram briefly outlines the PDA.
a66cd3ee
MS
697 ;;
698 ;; Common state:
d9e94c22
MS
699 ;; "else": Push state, goto state `else'.
700 ;; "while": Push state, goto state `while'.
701 ;; "catch" or "finally": Push state, goto state `catch'.
702 ;; boundary: Pop state.
a66cd3ee
MS
703 ;; other: Do nothing special.
704 ;;
d9e94c22
MS
705 ;; State `else':
706 ;; boundary: Goto state `else-boundary'.
707 ;; other: Error, pop state, retry token.
708 ;;
709 ;; State `else-boundary':
710 ;; "if": Pop state.
711 ;; boundary: Error, pop state.
712 ;; other: See common state.
713 ;;
714 ;; State `while':
715 ;; boundary: Save position, goto state `while-boundary'.
716 ;; other: Pop state, retry token.
717 ;;
718 ;; State `while-boundary':
719 ;; "do": Pop state.
720 ;; boundary: Restore position if it's not at start, pop state. [*see below]
721 ;; other: See common state.
722 ;;
723 ;; State `catch':
724 ;; boundary: Goto state `catch-boundary'.
725 ;; other: Error, pop state, retry token.
726 ;;
727 ;; State `catch-boundary':
728 ;; "try": Pop state.
729 ;; "catch": Goto state `catch'.
730 ;; boundary: Error, pop state.
731 ;; other: See common state.
732 ;;
733 ;; [*] In the `while-boundary' state, we had pushed a 'while state, and were
734 ;; searching for a "do" which would have opened a do-while. If we didn't
735 ;; find it, we discard the analysis done since the "while", go back to this
736 ;; token in the buffer and restart the scanning there, this time WITHOUT
737 ;; pushing the 'while state onto the stack.
738 ;;
a66cd3ee
MS
739 ;; In addition to the above there is some special handling of labels
740 ;; and macros.
741
742 (let ((case-fold-search nil)
743 (start (point))
744 macro-start
745 (delims (if comma-delim '(?\; ?,) '(?\;)))
746 (c-stmt-delim-chars (if comma-delim
747 c-stmt-delim-chars-with-comma
748 c-stmt-delim-chars))
d28e7f28 749 c-in-literal-cache c-maybe-labelp after-case:-pos saved
0386b551
AM
750 ;; Current position.
751 pos
752 ;; Position of last stmt boundary character (e.g. ;).
753 boundary-pos
754 ;; The position of the last sexp or bound that follows the
755 ;; first found colon, i.e. the start of the nonlabel part of
756 ;; the statement. It's `start' if a colon is found just after
757 ;; the start.
758 after-labels-pos
759 ;; Like `after-labels-pos', but the first such position inside
760 ;; a label, i.e. the start of the last label before the start
761 ;; of the nonlabel part of the statement.
762 last-label-pos
763 ;; The last position where a label is possible provided the
764 ;; statement started there. It's nil as long as no invalid
765 ;; label content has been found (according to
020716e1 766 ;; `c-nonlabel-token-key'). It's `start' if no valid label
0386b551
AM
767 ;; content was found in the label. Note that we might still
768 ;; regard it a label if it starts with `c-label-kwds'.
769 label-good-pos
b248a85d
AM
770 ;; Putative positions of the components of a bitfield declaration,
771 ;; e.g. "int foo : NUM_FOO_BITS ;"
772 bitfield-type-pos bitfield-id-pos bitfield-size-pos
0386b551
AM
773 ;; Symbol just scanned back over (e.g. 'while or 'boundary).
774 ;; See above.
775 sym
776 ;; Current state in the automaton. See above.
777 state
778 ;; Current saved positions. See above.
779 saved-pos
780 ;; Stack of conses (state . saved-pos).
781 stack
782 ;; Regexp which matches "for", "if", etc.
783 (cond-key (or c-opt-block-stmt-key
a66cd3ee 784 "\\<\\>")) ; Matches nothing.
0386b551
AM
785 ;; Return value.
786 (ret 'same)
787 ;; Positions of the last three sexps or bounds we've stopped at.
788 tok ptok pptok)
a66cd3ee
MS
789
790 (save-restriction
791 (if lim (narrow-to-region lim (point-max)))
792
793 (if (save-excursion
794 (and (c-beginning-of-macro)
795 (/= (point) start)))
796 (setq macro-start (point)))
797
d9e94c22 798 ;; Try to skip back over unary operator characters, to register
a66cd3ee
MS
799 ;; that we've moved.
800 (while (progn
801 (setq pos (point))
0386b551
AM
802 (c-backward-syntactic-ws)
803 ;; Protect post-++/-- operators just before a virtual semicolon.
804 (and (not (c-at-vsemi-p))
805 (/= (skip-chars-backward "-+!*&~@`#") 0))))
d9e94c22
MS
806
807 ;; Skip back over any semicolon here. If it was a bare semicolon, we're
0386b551 808 ;; done. Later on we ignore the boundaries for statements that don't
d9e94c22
MS
809 ;; contain any sexp. The only thing that is affected is that the error
810 ;; checking is a little less strict, and we really don't bother.
a66cd3ee
MS
811 (if (and (memq (char-before) delims)
812 (progn (forward-char -1)
813 (setq saved (point))
0386b551 814 (c-backward-syntactic-ws)
a66cd3ee
MS
815 (or (memq (char-before) delims)
816 (memq (char-before) '(?: nil))
d9e94c22 817 (eq (char-syntax (char-before)) ?\()
0386b551 818 (c-at-vsemi-p))))
a66cd3ee
MS
819 (setq ret 'previous
820 pos saved)
821
822 ;; Begin at start and not pos to detect macros if we stand
823 ;; directly after the #.
824 (goto-char start)
825 (if (looking-at "\\<\\|\\W")
826 ;; Record this as the first token if not starting inside it.
827 (setq tok start))
828
b248a85d
AM
829
830 ;; The following while loop goes back one sexp (balanced parens,
831 ;; etc. with contents, or symbol or suchlike) each iteration. This
832 ;; movement is accomplished with a call to c-backward-sexp approx 170
833 ;; lines below.
834 ;;
835 ;; The loop is exited only by throwing nil to the (catch 'loop ...):
836 ;; 1. On reaching the start of a macro;
837 ;; 2. On having passed a stmt boundary with the PDA stack empty;
838 ;; 3. On reaching the start of an Objective C method def;
839 ;; 4. From macro `c-bos-pop-state'; when the stack is empty;
840 ;; 5. From macro `c-bos-pop-state-and-retry' when the stack is empty.
a66cd3ee
MS
841 (while
842 (catch 'loop ;; Throw nil to break, non-nil to continue.
843 (cond
b248a85d 844 ;; Are we in a macro, just after the opening #?
a66cd3ee 845 ((save-excursion
0386b551 846 (and macro-start ; Always NIL for AWK.
a66cd3ee
MS
847 (progn (skip-chars-backward " \t")
848 (eq (char-before) ?#))
849 (progn (setq saved (1- (point)))
850 (beginning-of-line)
851 (not (eq (char-before (1- (point))) ?\\)))
d9e94c22 852 (looking-at c-opt-cpp-start)
a66cd3ee
MS
853 (progn (skip-chars-forward " \t")
854 (eq (point) saved))))
855 (goto-char saved)
856 (if (and (c-forward-to-cpp-define-body)
857 (progn (c-forward-syntactic-ws start)
858 (< (point) start)))
859 ;; Stop at the first token in the content of the macro.
860 (setq pos (point)
861 ignore-labels t) ; Avoid the label check on exit.
862 (setq pos saved
863 ret 'macro
864 ignore-labels t))
b248a85d 865 (throw 'loop nil)) ; 1. Start of macro.
a66cd3ee 866
d9e94c22
MS
867 ;; Do a round through the automaton if we've just passed a
868 ;; statement boundary or passed a "while"-like token.
a66cd3ee
MS
869 ((or sym
870 (and (looking-at cond-key)
871 (setq sym (intern (match-string 1)))))
872
873 (when (and (< pos start) (null stack))
b248a85d 874 (throw 'loop nil)) ; 2. Statement boundary.
a66cd3ee 875
d9e94c22
MS
876 ;; The PDA state handling.
877 ;;
037558bf 878 ;; Refer to the description of the PDA in the opening
d9e94c22
MS
879 ;; comments. In the following OR form, the first leaf
880 ;; attempts to handles one of the specific actions detailed
881 ;; (e.g., finding token "if" whilst in state `else-boundary').
882 ;; We drop through to the second leaf (which handles common
883 ;; state) if no specific handler is found in the first cond.
884 ;; If a parsing error is detected (e.g. an "else" with no
885 ;; preceding "if"), we throw to the enclosing catch.
886 ;;
887 ;; Note that the (eq state 'else) means
888 ;; "we've just passed an else", NOT "we're looking for an
889 ;; else".
a66cd3ee
MS
890 (or (cond
891 ((eq state 'else)
892 (if (eq sym 'boundary)
893 (setq state 'else-boundary)
894 (c-bos-report-error)
895 (c-bos-pop-state-and-retry)))
896
897 ((eq state 'else-boundary)
898 (cond ((eq sym 'if)
899 (c-bos-pop-state (setq ret 'beginning)))
900 ((eq sym 'boundary)
901 (c-bos-report-error)
902 (c-bos-pop-state))))
903
904 ((eq state 'while)
905 (if (and (eq sym 'boundary)
906 ;; Since this can cause backtracking we do a
907 ;; little more careful analysis to avoid it:
908 ;; If there's a label in front of the while
909 ;; it can't be part of a do-while.
910 (not after-labels-pos))
911 (progn (c-bos-save-pos)
912 (setq state 'while-boundary))
d9e94c22 913 (c-bos-pop-state-and-retry))) ; Can't be a do-while
a66cd3ee
MS
914
915 ((eq state 'while-boundary)
916 (cond ((eq sym 'do)
917 (c-bos-pop-state (setq ret 'beginning)))
d9e94c22
MS
918 ((eq sym 'boundary) ; isn't a do-while
919 (c-bos-restore-pos) ; the position of the while
920 (c-bos-pop-state)))) ; no longer searching for do.
a66cd3ee
MS
921
922 ((eq state 'catch)
923 (if (eq sym 'boundary)
924 (setq state 'catch-boundary)
925 (c-bos-report-error)
926 (c-bos-pop-state-and-retry)))
927
928 ((eq state 'catch-boundary)
929 (cond
930 ((eq sym 'try)
931 (c-bos-pop-state (setq ret 'beginning)))
932 ((eq sym 'catch)
933 (setq state 'catch))
934 ((eq sym 'boundary)
935 (c-bos-report-error)
936 (c-bos-pop-state)))))
937
d9e94c22
MS
938 ;; This is state common. We get here when the previous
939 ;; cond statement found no particular state handler.
a66cd3ee 940 (cond ((eq sym 'boundary)
d9e94c22
MS
941 ;; If we have a boundary at the start
942 ;; position we push a frame to go to the
943 ;; previous statement.
944 (if (>= pos start)
945 (c-bos-push-state)
946 (c-bos-pop-state)))
a66cd3ee
MS
947 ((eq sym 'else)
948 (c-bos-push-state)
949 (c-bos-save-error-info 'if 'else)
950 (setq state 'else))
951 ((eq sym 'while)
0386b551
AM
952 ;; Is this a real while, or a do-while?
953 ;; The next `when' triggers unless we are SURE that
40ba43b4 954 ;; the `while' is not the tail end of a `do-while'.
a66cd3ee 955 (when (or (not pptok)
d9e94c22 956 (memq (char-after pptok) delims)
0386b551
AM
957 ;; The following kludge is to prevent
958 ;; infinite recursion when called from
959 ;; c-awk-after-if-for-while-condition-p,
960 ;; or the like.
961 (and (eq (point) start)
962 (c-vsemi-status-unknown-p))
963 (c-at-vsemi-p pptok))
a66cd3ee
MS
964 ;; Since this can cause backtracking we do a
965 ;; little more careful analysis to avoid it: If
0386b551
AM
966 ;; the while isn't followed by a (possibly
967 ;; virtual) semicolon it can't be a do-while.
a66cd3ee
MS
968 (c-bos-push-state)
969 (setq state 'while)))
970 ((memq sym '(catch finally))
971 (c-bos-push-state)
972 (c-bos-save-error-info 'try sym)
973 (setq state 'catch))))
974
975 (when c-maybe-labelp
976 ;; We're either past a statement boundary or at the
977 ;; start of a statement, so throw away any label data
978 ;; for the previous one.
979 (setq after-labels-pos nil
980 last-label-pos nil
981 c-maybe-labelp nil))))
982
d9e94c22
MS
983 ;; Step to the previous sexp, but not if we crossed a
984 ;; boundary, since that doesn't consume an sexp.
a66cd3ee
MS
985 (if (eq sym 'boundary)
986 (setq ret 'previous)
d9e94c22
MS
987
988 ;; HERE IS THE SINGLE PLACE INSIDE THE PDA LOOP WHERE WE MOVE
0386b551
AM
989 ;; BACKWARDS THROUGH THE SOURCE.
990
0386b551 991 (c-backward-syntactic-ws)
0386b551 992 (let ((before-sws-pos (point))
0386b551
AM
993 ;; The end position of the area to search for statement
994 ;; barriers in this round.
b248a85d 995 (maybe-after-boundary-pos pos))
0386b551 996
b248a85d
AM
997 ;; Go back over exactly one logical sexp, taking proper
998 ;; account of macros and escaped EOLs.
0386b551
AM
999 (while
1000 (progn
1001 (unless (c-safe (c-backward-sexp) t)
1002 ;; Give up if we hit an unbalanced block. Since the
1003 ;; stack won't be empty the code below will report a
1004 ;; suitable error.
a66cd3ee 1005 (throw 'loop nil))
b248a85d
AM
1006 (cond
1007 ;; Have we moved into a macro?
1008 ((and (not macro-start)
1009 (c-beginning-of-macro))
1010 ;; Have we crossed a statement boundary? If not,
1011 ;; keep going back until we find one or a "real" sexp.
1012 (and
1013 (save-excursion
1014 (c-end-of-macro)
1015 (not (c-crosses-statement-barrier-p
1016 (point) maybe-after-boundary-pos)))
1017 (setq maybe-after-boundary-pos (point))))
1018 ;; Have we just gone back over an escaped NL? This
1019 ;; doesn't count as a sexp.
1020 ((looking-at "\\\\$")))))
1021
1022 ;; Have we crossed a statement boundary?
1023 (setq boundary-pos
1024 (cond
1025 ;; Are we at a macro beginning?
1026 ((and (not macro-start)
1027 c-opt-cpp-prefix
1028 (looking-at c-opt-cpp-prefix))
1029 (save-excursion
1030 (c-end-of-macro)
1031 (c-crosses-statement-barrier-p
1032 (point) maybe-after-boundary-pos)))
1033 ;; Just gone back over a brace block?
1034 ((and
1035 (eq (char-after) ?{)
1036 (not (c-looking-at-inexpr-block lim nil t)))
1037 (save-excursion
1038 (c-forward-sexp) (point)))
1039 ;; Just gone back over some paren block?
1040 ((looking-at "\\s\(")
1041 (save-excursion
1042 (goto-char (1+ (c-down-list-backward
1043 before-sws-pos)))
1044 (c-crosses-statement-barrier-p
1045 (point) maybe-after-boundary-pos)))
1046 ;; Just gone back over an ordinary symbol of some sort?
1047 (t (c-crosses-statement-barrier-p
1048 (point) maybe-after-boundary-pos))))
91af3942 1049
b248a85d
AM
1050 (when boundary-pos
1051 (setq pptok ptok
1052 ptok tok
1053 tok boundary-pos
1054 sym 'boundary)
1055 ;; Like a C "continue". Analyze the next sexp.
1056 (throw 'loop t))))
a66cd3ee
MS
1057
1058 ;; ObjC method def?
1059 (when (and c-opt-method-key
1060 (setq saved (c-in-method-def-p)))
1061 (setq pos saved
1062 ignore-labels t) ; Avoid the label check on exit.
b248a85d
AM
1063 (throw 'loop nil)) ; 3. ObjC method def.
1064
1065 ;; Might we have a bitfield declaration, "<type> <id> : <size>"?
1066 (if c-has-bitfields
1067 (cond
1068 ;; The : <size> and <id> fields?
1069 ((and (numberp c-maybe-labelp)
1070 (not bitfield-size-pos)
1071 (save-excursion
1072 (goto-char (or tok start))
1073 (not (looking-at c-keywords-regexp)))
1074 (not (looking-at c-keywords-regexp))
1075 (not (c-punctuation-in (point) c-maybe-labelp)))
1076 (setq bitfield-size-pos (or tok start)
1077 bitfield-id-pos (point)))
1078 ;; The <type> field?
1079 ((and bitfield-id-pos
1080 (not bitfield-type-pos))
1081 (if (and (looking-at c-symbol-key) ; Can only be an integer type. :-)
1082 (not (looking-at c-not-primitive-type-keywords-regexp))
1083 (not (c-punctuation-in (point) tok)))
1084 (setq bitfield-type-pos (point))
1085 (setq bitfield-size-pos nil
1086 bitfield-id-pos nil)))))
a66cd3ee 1087
0386b551
AM
1088 ;; Handle labels.
1089 (unless (eq ignore-labels t)
1090 (when (numberp c-maybe-labelp)
cb694ab7
AM
1091 ;; `c-crosses-statement-barrier-p' has found a colon, so we
1092 ;; might be in a label now. Have we got a real label
1093 ;; (including a case label) or something like C++'s "public:"?
d28e7f28
AM
1094 ;; A case label might use an expression rather than a token.
1095 (setq after-case:-pos (or tok start))
020716e1
AM
1096 (if (or (looking-at c-nonlabel-token-key) ; e.g. "while" or "'a'"
1097 ;; Catch C++'s inheritance construct "class foo : bar".
1098 (save-excursion
1099 (and
1100 (c-safe (c-backward-sexp) t)
1101 (looking-at c-nonlabel-token-2-key))))
d28e7f28
AM
1102 (setq c-maybe-labelp nil)
1103 (if after-labels-pos ; Have we already encountered a label?
1104 (if (not last-label-pos)
1105 (setq last-label-pos (or tok start)))
1106 (setq after-labels-pos (or tok start)))
1107 (setq c-maybe-labelp t
1108 label-good-pos nil))) ; bogus "label"
cb694ab7
AM
1109
1110 (when (and (not label-good-pos) ; i.e. no invalid "label"'s yet
1111 ; been found.
1112 (looking-at c-nonlabel-token-key)) ; e.g. "while :"
0386b551
AM
1113 ;; We're in a potential label and it's the first
1114 ;; time we've found something that isn't allowed in
1115 ;; one.
1116 (setq label-good-pos (or tok start))))
1117
1118 ;; We've moved back by a sexp, so update the token positions.
a66cd3ee
MS
1119 (setq sym nil
1120 pptok ptok
1121 ptok tok
1122 tok (point)
b248a85d
AM
1123 pos tok) ; always non-nil
1124 ) ; end of (catch loop ....)
1125 ) ; end of sexp-at-a-time (while ....)
91af3942 1126
a66cd3ee
MS
1127 ;; If the stack isn't empty there might be errors to report.
1128 (while stack
1129 (if (and (vectorp saved-pos) (eq (length saved-pos) 3))
1130 (c-bos-report-error))
1131 (setq saved-pos (cdr (car stack))
1132 stack (cdr stack)))
1133
1134 (when (and (eq ret 'same)
1135 (not (memq sym '(boundary ignore nil))))
1136 ;; Need to investigate closer whether we've crossed
1137 ;; between a substatement and its containing statement.
5ab78d3d
AM
1138 (if (setq saved
1139 (cond ((and (looking-at c-block-stmt-1-2-key)
1140 (eq (char-after ptok) ?\())
1141 pptok)
1142 ((looking-at c-block-stmt-1-key)
1143 ptok)
1144 (t pptok)))
a66cd3ee
MS
1145 (cond ((> start saved) (setq pos saved))
1146 ((= start saved) (setq ret 'up)))))
1147
0386b551
AM
1148 (when (and (not ignore-labels)
1149 (eq c-maybe-labelp t)
d9e94c22 1150 (not (eq ret 'beginning))
0386b551 1151 after-labels-pos
b248a85d 1152 (not bitfield-type-pos) ; Bitfields take precedence over labels.
0386b551
AM
1153 (or (not label-good-pos)
1154 (<= label-good-pos pos)
1155 (progn
1156 (goto-char (if (and last-label-pos
1157 (< last-label-pos start))
1158 last-label-pos
1159 pos))
1160 (looking-at c-label-kwds-regexp))))
a66cd3ee
MS
1161 ;; We're in a label. Maybe we should step to the statement
1162 ;; after it.
1163 (if (< after-labels-pos start)
1164 (setq pos after-labels-pos)
1165 (setq ret 'label)
0386b551
AM
1166 (if (and last-label-pos (< last-label-pos start))
1167 ;; Might have jumped over several labels. Go to the last one.
a66cd3ee
MS
1168 (setq pos last-label-pos)))))
1169
d28e7f28 1170 ;; Have we got "case <expression>:"?
a66cd3ee 1171 (goto-char pos)
d28e7f28
AM
1172 (when (and after-case:-pos
1173 (not (eq ret 'beginning))
1174 (looking-at c-case-kwds-regexp))
1175 (if (< after-case:-pos start)
0000ee90
AM
1176 (setq pos after-case:-pos))
1177 (if (eq ret 'same)
1178 (setq ret 'label)))
d28e7f28
AM
1179
1180 ;; Skip over the unary operators that can start the statement.
a66cd3ee 1181 (while (progn
0386b551
AM
1182 (c-backward-syntactic-ws)
1183 ;; protect AWK post-inc/decrement operators, etc.
1184 (and (not (c-at-vsemi-p (point)))
1185 (/= (skip-chars-backward "-+!*&~@`#") 0)))
a66cd3ee
MS
1186 (setq pos (point)))
1187 (goto-char pos)
1188 ret)))
785eecbb 1189
b248a85d
AM
1190(defun c-punctuation-in (from to)
1191 "Return non-nil if there is a non-comment non-macro punctuation character
1192between FROM and TO. FROM must not be in a string or comment. The returned
1193value is the position of the first such character."
1194 (save-excursion
1195 (goto-char from)
1196 (let ((pos (point)))
1197 (while (progn (skip-chars-forward c-symbol-chars to)
1198 (c-forward-syntactic-ws to)
1199 (> (point) pos))
1200 (setq pos (point))))
1201 (and (< (point) to) (point))))
1202
785eecbb 1203(defun c-crosses-statement-barrier-p (from to)
a66cd3ee
MS
1204 "Return non-nil if buffer positions FROM to TO cross one or more
1205statement or declaration boundaries. The returned value is actually
d9e94c22
MS
1206the position of the earliest boundary char. FROM must not be within
1207a string or comment.
a66cd3ee
MS
1208
1209The variable `c-maybe-labelp' is set to the position of the first `:' that
1210might start a label (i.e. not part of `::' and not preceded by `?'). If a
0386b551
AM
1211single `?' is found, then `c-maybe-labelp' is cleared.
1212
1213For AWK, a statement which is terminated by an EOL (not a \; or a }) is
1214regarded as having a \"virtual semicolon\" immediately after the last token on
b414f371 1215the line. If this virtual semicolon is _at_ from, the function recognizes it.
0386b551
AM
1216
1217Note that this function might do hidden buffer changes. See the
1218comment at the start of cc-engine.el for more info."
536610a4
AM
1219 (let* ((skip-chars
1220 ;; If the current language has CPP macros, insert # into skip-chars.
1221 (if c-opt-cpp-symbol
1222 (concat (substring c-stmt-delim-chars 0 1) ; "^"
1223 c-opt-cpp-symbol ; usually "#"
1224 (substring c-stmt-delim-chars 1)) ; e.g. ";{}?:"
1225 c-stmt-delim-chars))
1226 (non-skip-list
1227 (append (substring skip-chars 1) nil)) ; e.g. (?# ?\; ?{ ?} ?? ?:)
1228 lit-range vsemi-pos)
1229 (save-restriction
1230 (widen)
1231 (save-excursion
1232 (catch 'done
1233 (goto-char from)
1234 (while (progn (skip-chars-forward
1235 skip-chars
1236 (min to (c-point 'bonl)))
1237 (< (point) to))
1238 (cond
1239 ;; Virtual semicolon?
1240 ((and (bolp)
1241 (save-excursion
1242 (progn
1243 (if (setq lit-range (c-literal-limits from)) ; Have we landed in a string/comment?
1244 (goto-char (car lit-range)))
1245 (c-backward-syntactic-ws) ; ? put a limit here, maybe?
1246 (setq vsemi-pos (point))
1247 (c-at-vsemi-p))))
1248 (throw 'done vsemi-pos))
1249 ;; In a string/comment?
2e492df3 1250 ((setq lit-range (c-literal-limits from))
536610a4
AM
1251 (goto-char (cdr lit-range)))
1252 ((eq (char-after) ?:)
1253 (forward-char)
1254 (if (and (eq (char-after) ?:)
1255 (< (point) to))
1256 ;; Ignore scope operators.
1257 (forward-char)
1258 (setq c-maybe-labelp (1- (point)))))
1259 ((eq (char-after) ??)
1260 ;; A question mark. Can't be a label, so stop
1261 ;; looking for more : and ?.
1262 (setq c-maybe-labelp nil
1263 skip-chars (substring c-stmt-delim-chars 0 -2)))
b1e8e010
AM
1264 ;; At a CPP construct or a "#" or "##" operator?
1265 ((and c-opt-cpp-symbol (looking-at c-opt-cpp-symbol))
1266 (if (save-excursion
1267 (skip-chars-backward " \t")
1268 (and (bolp)
1269 (or (bobp)
1270 (not (eq (char-before (1- (point))) ?\\)))))
1271 (c-end-of-macro)
1272 (skip-chars-forward c-opt-cpp-symbol)))
536610a4
AM
1273 ((memq (char-after) non-skip-list)
1274 (throw 'done (point)))))
1275 ;; In trailing space after an as yet undetected virtual semicolon?
1276 (c-backward-syntactic-ws from)
b5eb9035
AM
1277 (when (and (bolp) (not (bobp))) ; Can happen in AWK Mode with an
1278 ; unterminated string/regexp.
1279 (backward-char))
536610a4
AM
1280 (if (and (< (point) to)
1281 (c-at-vsemi-p))
1282 (point)
1283 nil))))))
0386b551
AM
1284
1285(defun c-at-statement-start-p ()
1286 "Return non-nil if the point is at the first token in a statement
1287or somewhere in the syntactic whitespace before it.
1288
1289A \"statement\" here is not restricted to those inside code blocks.
1290Any kind of declaration-like construct that occur outside function
1291bodies is also considered a \"statement\".
1292
1293Note that this function might do hidden buffer changes. See the
1294comment at the start of cc-engine.el for more info."
1295
1296 (save-excursion
1297 (let ((end (point))
1298 c-maybe-labelp)
1299 (c-syntactic-skip-backward (substring c-stmt-delim-chars 1) nil t)
1300 (or (bobp)
1301 (eq (char-before) ?})
1302 (and (eq (char-before) ?{)
1303 (not (and c-special-brace-lists
1304 (progn (backward-char)
1305 (c-looking-at-special-brace-list)))))
1306 (c-crosses-statement-barrier-p (point) end)))))
1307
1308(defun c-at-expression-start-p ()
1309 "Return non-nil if the point is at the first token in an expression or
1310statement, or somewhere in the syntactic whitespace before it.
1311
1312An \"expression\" here is a bit different from the normal language
1313grammar sense: It's any sequence of expression tokens except commas,
1314unless they are enclosed inside parentheses of some kind. Also, an
1315expression never continues past an enclosing parenthesis, but it might
1316contain parenthesis pairs of any sort except braces.
1317
1318Since expressions never cross statement boundaries, this function also
1319recognizes statement beginnings, just like `c-at-statement-start-p'.
1320
1321Note that this function might do hidden buffer changes. See the
1322comment at the start of cc-engine.el for more info."
1323
1324 (save-excursion
1325 (let ((end (point))
1326 (c-stmt-delim-chars c-stmt-delim-chars-with-comma)
1327 c-maybe-labelp)
1328 (c-syntactic-skip-backward (substring c-stmt-delim-chars 1) nil t)
1329 (or (bobp)
1330 (memq (char-before) '(?{ ?}))
1331 (save-excursion (backward-char)
1332 (looking-at "\\s("))
1333 (c-crosses-statement-barrier-p (point) end)))))
785eecbb
RS
1334
1335\f
d9e94c22
MS
1336;; A set of functions that covers various idiosyncrasies in
1337;; implementations of `forward-comment'.
1338
1339;; Note: Some emacsen considers incorrectly that any line comment
1340;; ending with a backslash continues to the next line. I can't think
1341;; of any way to work around that in a reliable way without changing
1342;; the buffer, though. Suggestions welcome. ;) (No, temporarily
1343;; changing the syntax for backslash doesn't work since we must treat
1344;; escapes in string literals correctly.)
1345
1346(defun c-forward-single-comment ()
1347 "Move forward past whitespace and the closest following comment, if any.
1348Return t if a comment was found, nil otherwise. In either case, the
1349point is moved past the following whitespace. Line continuations,
1350i.e. a backslashes followed by line breaks, are treated as whitespace.
1351The line breaks that end line comments are considered to be the
1352comment enders, so the point will be put on the beginning of the next
1353line if it moved past a line comment.
1354
1355This function does not do any hidden buffer changes."
1356
1357 (let ((start (point)))
1358 (when (looking-at "\\([ \t\n\r\f\v]\\|\\\\[\n\r]\\)+")
1359 (goto-char (match-end 0)))
1360
1361 (when (forward-comment 1)
1362 (if (eobp)
1363 ;; Some emacsen (e.g. XEmacs 21) return t when moving
1364 ;; forwards at eob.
1365 nil
1366
1367 ;; Emacs includes the ending newline in a b-style (c++)
1368 ;; comment, but XEmacs doesn't. We depend on the Emacs
1369 ;; behavior (which also is symmetric).
1370 (if (and (eolp) (elt (parse-partial-sexp start (point)) 7))
1371 (condition-case nil (forward-char 1)))
1372
1373 t))))
1374
1375(defsubst c-forward-comments ()
1376 "Move forward past all following whitespace and comments.
1377Line continuations, i.e. a backslashes followed by line breaks, are
1378treated as whitespace.
1379
0386b551
AM
1380Note that this function might do hidden buffer changes. See the
1381comment at the start of cc-engine.el for more info."
d9e94c22
MS
1382
1383 (while (or
1384 ;; If forward-comment in at least XEmacs 21 is given a large
1385 ;; positive value, it'll loop all the way through if it hits
1386 ;; eob.
1387 (and (forward-comment 5)
1388 ;; Some emacsen (e.g. XEmacs 21) return t when moving
1389 ;; forwards at eob.
1390 (not (eobp)))
1391
1392 (when (looking-at "\\\\[\n\r]")
1393 (forward-char 2)
1394 t))))
1395
1396(defun c-backward-single-comment ()
1397 "Move backward past whitespace and the closest preceding comment, if any.
1398Return t if a comment was found, nil otherwise. In either case, the
1399point is moved past the preceding whitespace. Line continuations,
1400i.e. a backslashes followed by line breaks, are treated as whitespace.
1401The line breaks that end line comments are considered to be the
1402comment enders, so the point cannot be at the end of the same line to
1403move over a line comment.
1404
1405This function does not do any hidden buffer changes."
1406
1407 (let ((start (point)))
1408 ;; When we got newline terminated comments, forward-comment in all
1409 ;; supported emacsen so far will stop at eol of each line not
1410 ;; ending with a comment when moving backwards. This corrects for
1411 ;; that, and at the same time handles line continuations.
1412 (while (progn
1413 (skip-chars-backward " \t\n\r\f\v")
1414 (and (looking-at "[\n\r]")
0386b551 1415 (eq (char-before) ?\\)))
d9e94c22
MS
1416 (backward-char))
1417
1418 (if (bobp)
1419 ;; Some emacsen (e.g. Emacs 19.34) return t when moving
1420 ;; backwards at bob.
1421 nil
1422
1423 ;; Leave point after the closest following newline if we've
1424 ;; backed up over any above, since forward-comment won't move
1425 ;; backward over a line comment if point is at the end of the
1426 ;; same line.
1427 (re-search-forward "\\=\\s *[\n\r]" start t)
1428
bf2c1571 1429 (if (if (let (open-paren-in-column-0-is-defun-start) (forward-comment -1))
d9e94c22
MS
1430 (if (eolp)
1431 ;; If forward-comment above succeeded and we're at eol
1432 ;; then the newline we moved over above didn't end a
1433 ;; line comment, so we give it another go.
bf2c1571
AM
1434 (let (open-paren-in-column-0-is-defun-start)
1435 (forward-comment -1))
d9e94c22
MS
1436 t))
1437
1438 ;; Emacs <= 20 and XEmacs move back over the closer of a
1439 ;; block comment that lacks an opener.
1440 (if (looking-at "\\*/")
1441 (progn (forward-char 2) nil)
1442 t)))))
1443
1444(defsubst c-backward-comments ()
1445 "Move backward past all preceding whitespace and comments.
1446Line continuations, i.e. a backslashes followed by line breaks, are
1447treated as whitespace. The line breaks that end line comments are
1448considered to be the comment enders, so the point cannot be at the end
0386b551
AM
1449of the same line to move over a line comment. Unlike
1450c-backward-syntactic-ws, this function doesn't move back over
1451preprocessor directives.
d9e94c22 1452
0386b551
AM
1453Note that this function might do hidden buffer changes. See the
1454comment at the start of cc-engine.el for more info."
d9e94c22
MS
1455
1456 (let ((start (point)))
1457 (while (and
0386b551 1458 ;; `forward-comment' in some emacsen (e.g. XEmacs 21.4)
d9e94c22
MS
1459 ;; return t when moving backwards at bob.
1460 (not (bobp))
1461
14b8c3d9
AM
1462 (if (let (open-paren-in-column-0-is-defun-start moved-comment)
1463 (while
1464 (and (not (setq moved-comment (forward-comment -1)))
1465 ;; Cope specifically with ^M^J here -
1466 ;; forward-comment sometimes gets stuck after ^Ms,
1467 ;; sometimes after ^M^J.
1468 (or
1469 (when (eq (char-before) ?\r)
1470 (backward-char)
1471 t)
1472 (when (and (eq (char-before) ?\n)
1473 (eq (char-before (1- (point))) ?\r))
1474 (backward-char 2)
1475 t))))
1476 moved-comment)
d9e94c22
MS
1477 (if (looking-at "\\*/")
1478 ;; Emacs <= 20 and XEmacs move back over the
1479 ;; closer of a block comment that lacks an opener.
1480 (progn (forward-char 2) nil)
1481 t)
1482
1483 ;; XEmacs treats line continuations as whitespace but
1484 ;; only in the backward direction, which seems a bit
1485 ;; odd. Anyway, this is necessary for Emacs.
1486 (when (and (looking-at "[\n\r]")
1487 (eq (char-before) ?\\)
1488 (< (point) start))
1489 (backward-char)
1490 t))))))
1491
1492\f
d9e94c22 1493;; Tools for skipping over syntactic whitespace.
a66cd3ee 1494
d9e94c22
MS
1495;; The following functions use text properties to cache searches over
1496;; large regions of syntactic whitespace. It works as follows:
1497;;
1498;; o If a syntactic whitespace region contains anything but simple
1499;; whitespace (i.e. space, tab and line breaks), the text property
1500;; `c-in-sws' is put over it. At places where we have stopped
1501;; within that region there's also a `c-is-sws' text property.
1502;; That since there typically are nested whitespace inside that
1503;; must be handled separately, e.g. whitespace inside a comment or
1504;; cpp directive. Thus, from one point with `c-is-sws' it's safe
1505;; to jump to another point with that property within the same
1506;; `c-in-sws' region. It can be likened to a ladder where
1507;; `c-in-sws' marks the bars and `c-is-sws' the rungs.
1508;;
1509;; o The `c-is-sws' property is put on the simple whitespace chars at
1510;; a "rung position" and also maybe on the first following char.
1511;; As many characters as can be conveniently found in this range
1512;; are marked, but no assumption can be made that the whole range
1513;; is marked (it could be clobbered by later changes, for
1514;; instance).
1515;;
1516;; Note that some part of the beginning of a sequence of simple
1517;; whitespace might be part of the end of a preceding line comment
1518;; or cpp directive and must not be considered part of the "rung".
1519;; Such whitespace is some amount of horizontal whitespace followed
1520;; by a newline. In the case of cpp directives it could also be
1521;; two newlines with horizontal whitespace between them.
1522;;
1523;; The reason to include the first following char is to cope with
1524;; "rung positions" that doesn't have any ordinary whitespace. If
1525;; `c-is-sws' is put on a token character it does not have
1526;; `c-in-sws' set simultaneously. That's the only case when that
1527;; can occur, and the reason for not extending the `c-in-sws'
1528;; region to cover it is that the `c-in-sws' region could then be
1529;; accidentally merged with a following one if the token is only
1530;; one character long.
1531;;
1532;; o On buffer changes the `c-in-sws' and `c-is-sws' properties are
1533;; removed in the changed region. If the change was inside
1534;; syntactic whitespace that means that the "ladder" is broken, but
1535;; a later call to `c-forward-sws' or `c-backward-sws' will use the
1536;; parts on either side and use an ordinary search only to "repair"
1537;; the gap.
1538;;
1539;; Special care needs to be taken if a region is removed: If there
1540;; are `c-in-sws' on both sides of it which do not connect inside
1541;; the region then they can't be joined. If e.g. a marked macro is
1542;; broken, syntactic whitespace inside the new text might be
1543;; marked. If those marks would become connected with the old
1544;; `c-in-sws' range around the macro then we could get a ladder
1545;; with one end outside the macro and the other at some whitespace
1546;; within it.
1547;;
1548;; The main motivation for this system is to increase the speed in
1549;; skipping over the large whitespace regions that can occur at the
1550;; top level in e.g. header files that contain a lot of comments and
1551;; cpp directives. For small comments inside code it's probably
1552;; slower than using `forward-comment' straightforwardly, but speed is
1553;; not a significant factor there anyway.
1554
1555; (defface c-debug-is-sws-face
1556; '((t (:background "GreenYellow")))
1557; "Debug face to mark the `c-is-sws' property.")
1558; (defface c-debug-in-sws-face
1559; '((t (:underline t)))
1560; "Debug face to mark the `c-in-sws' property.")
1561
1562; (defun c-debug-put-sws-faces ()
1563; ;; Put the sws debug faces on all the `c-is-sws' and `c-in-sws'
1564; ;; properties in the buffer.
1565; (interactive)
1566; (save-excursion
0386b551 1567; (c-save-buffer-state (in-face)
d9e94c22
MS
1568; (goto-char (point-min))
1569; (setq in-face (if (get-text-property (point) 'c-is-sws)
1570; (point)))
1571; (while (progn
1572; (goto-char (next-single-property-change
1573; (point) 'c-is-sws nil (point-max)))
1574; (if in-face
1575; (progn
1576; (c-debug-add-face in-face (point) 'c-debug-is-sws-face)
1577; (setq in-face nil))
1578; (setq in-face (point)))
1579; (not (eobp))))
1580; (goto-char (point-min))
1581; (setq in-face (if (get-text-property (point) 'c-in-sws)
1582; (point)))
1583; (while (progn
1584; (goto-char (next-single-property-change
1585; (point) 'c-in-sws nil (point-max)))
1586; (if in-face
1587; (progn
1588; (c-debug-add-face in-face (point) 'c-debug-in-sws-face)
1589; (setq in-face nil))
1590; (setq in-face (point)))
1591; (not (eobp)))))))
1592
1593(defmacro c-debug-sws-msg (&rest args)
1594 ;;`(message ,@args)
1595 )
1596
1597(defmacro c-put-is-sws (beg end)
0386b551 1598 ;; This macro does a hidden buffer change.
d9e94c22
MS
1599 `(let ((beg ,beg) (end ,end))
1600 (put-text-property beg end 'c-is-sws t)
1601 ,@(when (facep 'c-debug-is-sws-face)
1602 `((c-debug-add-face beg end 'c-debug-is-sws-face)))))
1603
1604(defmacro c-put-in-sws (beg end)
0386b551 1605 ;; This macro does a hidden buffer change.
d9e94c22
MS
1606 `(let ((beg ,beg) (end ,end))
1607 (put-text-property beg end 'c-in-sws t)
1608 ,@(when (facep 'c-debug-is-sws-face)
1609 `((c-debug-add-face beg end 'c-debug-in-sws-face)))))
1610
1611(defmacro c-remove-is-sws (beg end)
0386b551 1612 ;; This macro does a hidden buffer change.
d9e94c22
MS
1613 `(let ((beg ,beg) (end ,end))
1614 (remove-text-properties beg end '(c-is-sws nil))
1615 ,@(when (facep 'c-debug-is-sws-face)
1616 `((c-debug-remove-face beg end 'c-debug-is-sws-face)))))
1617
1618(defmacro c-remove-in-sws (beg end)
0386b551 1619 ;; This macro does a hidden buffer change.
d9e94c22
MS
1620 `(let ((beg ,beg) (end ,end))
1621 (remove-text-properties beg end '(c-in-sws nil))
1622 ,@(when (facep 'c-debug-is-sws-face)
1623 `((c-debug-remove-face beg end 'c-debug-in-sws-face)))))
1624
1625(defmacro c-remove-is-and-in-sws (beg end)
0386b551 1626 ;; This macro does a hidden buffer change.
d9e94c22
MS
1627 `(let ((beg ,beg) (end ,end))
1628 (remove-text-properties beg end '(c-is-sws nil c-in-sws nil))
1629 ,@(when (facep 'c-debug-is-sws-face)
1630 `((c-debug-remove-face beg end 'c-debug-is-sws-face)
1631 (c-debug-remove-face beg end 'c-debug-in-sws-face)))))
1632
1633(defsubst c-invalidate-sws-region-after (beg end)
1634 ;; Called from `after-change-functions'. Note that if
1635 ;; `c-forward-sws' or `c-backward-sws' are used outside
1636 ;; `c-save-buffer-state' or similar then this will remove the cache
1637 ;; properties right after they're added.
0386b551
AM
1638 ;;
1639 ;; This function does hidden buffer changes.
d9e94c22
MS
1640
1641 (save-excursion
1642 ;; Adjust the end to remove the properties in any following simple
1643 ;; ws up to and including the next line break, if there is any
1644 ;; after the changed region. This is necessary e.g. when a rung
1645 ;; marked empty line is converted to a line comment by inserting
1646 ;; "//" before the line break. In that case the line break would
1647 ;; keep the rung mark which could make a later `c-backward-sws'
1648 ;; move into the line comment instead of over it.
1649 (goto-char end)
1650 (skip-chars-forward " \t\f\v")
1651 (when (and (eolp) (not (eobp)))
1652 (setq end (1+ (point)))))
1653
1654 (when (and (= beg end)
1655 (get-text-property beg 'c-in-sws)
2a15eb73 1656 (> beg (point-min))
d9e94c22
MS
1657 (get-text-property (1- beg) 'c-in-sws))
1658 ;; Ensure that an `c-in-sws' range gets broken. Note that it isn't
1659 ;; safe to keep a range that was continuous before the change. E.g:
1660 ;;
1661 ;; #define foo
1662 ;; \
1663 ;; bar
1664 ;;
1665 ;; There can be a "ladder" between "#" and "b". Now, if the newline
1666 ;; after "foo" is removed then "bar" will become part of the cpp
1667 ;; directive instead of a syntactically relevant token. In that
1668 ;; case there's no longer syntactic ws from "#" to "b".
1669 (setq beg (1- beg)))
1670
1671 (c-debug-sws-msg "c-invalidate-sws-region-after [%s..%s]" beg end)
1672 (c-remove-is-and-in-sws beg end))
1673
1674(defun c-forward-sws ()
1675 ;; Used by `c-forward-syntactic-ws' to implement the unbounded search.
0386b551
AM
1676 ;;
1677 ;; This function might do hidden buffer changes.
d9e94c22
MS
1678
1679 (let (;; `rung-pos' is set to a position as early as possible in the
1680 ;; unmarked part of the simple ws region.
1681 (rung-pos (point)) next-rung-pos rung-end-pos last-put-in-sws-pos
1682 rung-is-marked next-rung-is-marked simple-ws-end
1683 ;; `safe-start' is set when it's safe to cache the start position.
1684 ;; It's not set if we've initially skipped over comments and line
1685 ;; continuations since we might have gone out through the end of a
1686 ;; macro then. This provision makes `c-forward-sws' not populate the
1687 ;; cache in the majority of cases, but otoh is `c-backward-sws' by far
1688 ;; more common.
1689 safe-start)
1690
1691 ;; Skip simple ws and do a quick check on the following character to see
1692 ;; if it's anything that can't start syntactic ws, so we can bail out
1693 ;; early in the majority of cases when there just are a few ws chars.
1694 (skip-chars-forward " \t\n\r\f\v")
1695 (when (looking-at c-syntactic-ws-start)
1696
1697 (setq rung-end-pos (min (1+ (point)) (point-max)))
1698 (if (setq rung-is-marked (text-property-any rung-pos rung-end-pos
1699 'c-is-sws t))
1700 ;; Find the last rung position to avoid setting properties in all
1701 ;; the cases when the marked rung is complete.
1702 ;; (`next-single-property-change' is certain to move at least one
1703 ;; step forward.)
1704 (setq rung-pos (1- (next-single-property-change
1705 rung-is-marked 'c-is-sws nil rung-end-pos)))
1706 ;; Got no marked rung here. Since the simple ws might have started
1707 ;; inside a line comment or cpp directive we must set `rung-pos' as
1708 ;; high as possible.
1709 (setq rung-pos (point)))
1710
2143fa32 1711 (with-silent-modifications
d9e94c22
MS
1712 (while
1713 (progn
1714 (while
1715 (when (and rung-is-marked
1716 (get-text-property (point) 'c-in-sws))
1717
1718 ;; The following search is the main reason that `c-in-sws'
1719 ;; and `c-is-sws' aren't combined to one property.
1720 (goto-char (next-single-property-change
1721 (point) 'c-in-sws nil (point-max)))
1722 (unless (get-text-property (point) 'c-is-sws)
1723 ;; If the `c-in-sws' region extended past the last
1724 ;; `c-is-sws' char we have to go back a bit.
1725 (or (get-text-property (1- (point)) 'c-is-sws)
1726 (goto-char (previous-single-property-change
1727 (point) 'c-is-sws)))
1728 (backward-char))
1729
1730 (c-debug-sws-msg
1731 "c-forward-sws cached move %s -> %s (max %s)"
1732 rung-pos (point) (point-max))
1733
1734 (setq rung-pos (point))
1735 (and (> (skip-chars-forward " \t\n\r\f\v") 0)
1736 (not (eobp))))
1737
1738 ;; We'll loop here if there is simple ws after the last rung.
1739 ;; That means that there's been some change in it and it's
1740 ;; possible that we've stepped into another ladder, so extend
1741 ;; the previous one to join with it if there is one, and try to
1742 ;; use the cache again.
1743 (c-debug-sws-msg
1744 "c-forward-sws extending rung with [%s..%s] (max %s)"
1745 (1+ rung-pos) (1+ (point)) (point-max))
1746 (unless (get-text-property (point) 'c-is-sws)
1747 ;; Remove any `c-in-sws' property from the last char of
1748 ;; the rung before we mark it with `c-is-sws', so that we
1749 ;; won't connect with the remains of a broken "ladder".
1750 (c-remove-in-sws (point) (1+ (point))))
1751 (c-put-is-sws (1+ rung-pos)
1752 (1+ (point)))
1753 (c-put-in-sws rung-pos
1754 (setq rung-pos (point)
1755 last-put-in-sws-pos rung-pos)))
1756
1757 (setq simple-ws-end (point))
1758 (c-forward-comments)
1759
1760 (cond
1761 ((/= (point) simple-ws-end)
1762 ;; Skipped over comments. Don't cache at eob in case the buffer
1763 ;; is narrowed.
1764 (not (eobp)))
1765
1766 ((save-excursion
1767 (and c-opt-cpp-prefix
1768 (looking-at c-opt-cpp-start)
1769 (progn (skip-chars-backward " \t")
1770 (bolp))
1771 (or (bobp)
1772 (progn (backward-char)
1773 (not (eq (char-before) ?\\))))))
1774 ;; Skip a preprocessor directive.
1775 (end-of-line)
1776 (while (and (eq (char-before) ?\\)
1777 (= (forward-line 1) 0))
1778 (end-of-line))
1779 (forward-line 1)
1780 (setq safe-start t)
1781 ;; Don't cache at eob in case the buffer is narrowed.
1782 (not (eobp)))))
1783
1784 ;; We've searched over a piece of non-white syntactic ws. See if this
1785 ;; can be cached.
1786 (setq next-rung-pos (point))
1787 (skip-chars-forward " \t\n\r\f\v")
1788 (setq rung-end-pos (min (1+ (point)) (point-max)))
1789
1790 (if (or
1791 ;; Cache if we haven't skipped comments only, and if we started
1792 ;; either from a marked rung or from a completely uncached
1793 ;; position.
1794 (and safe-start
1795 (or rung-is-marked
1796 (not (get-text-property simple-ws-end 'c-in-sws))))
1797
1798 ;; See if there's a marked rung in the encountered simple ws. If
1799 ;; so then we can cache, unless `safe-start' is nil. Even then
1800 ;; we need to do this to check if the cache can be used for the
1801 ;; next step.
1802 (and (setq next-rung-is-marked
1803 (text-property-any next-rung-pos rung-end-pos
1804 'c-is-sws t))
1805 safe-start))
b2acd789 1806
0ec8351b 1807 (progn
d9e94c22
MS
1808 (c-debug-sws-msg
1809 "c-forward-sws caching [%s..%s] - [%s..%s] (max %s)"
1810 rung-pos (1+ simple-ws-end) next-rung-pos rung-end-pos
1811 (point-max))
1812
1813 ;; Remove the properties for any nested ws that might be cached.
1814 ;; Only necessary for `c-is-sws' since `c-in-sws' will be set
1815 ;; anyway.
1816 (c-remove-is-sws (1+ simple-ws-end) next-rung-pos)
1817 (unless (and rung-is-marked (= rung-pos simple-ws-end))
1818 (c-put-is-sws rung-pos
1819 (1+ simple-ws-end))
1820 (setq rung-is-marked t))
1821 (c-put-in-sws rung-pos
1822 (setq rung-pos (point)
1823 last-put-in-sws-pos rung-pos))
1824 (unless (get-text-property (1- rung-end-pos) 'c-is-sws)
1825 ;; Remove any `c-in-sws' property from the last char of
1826 ;; the rung before we mark it with `c-is-sws', so that we
1827 ;; won't connect with the remains of a broken "ladder".
1828 (c-remove-in-sws (1- rung-end-pos) rung-end-pos))
1829 (c-put-is-sws next-rung-pos
1830 rung-end-pos))
1831
1832 (c-debug-sws-msg
1833 "c-forward-sws not caching [%s..%s] - [%s..%s] (max %s)"
1834 rung-pos (1+ simple-ws-end) next-rung-pos rung-end-pos
1835 (point-max))
1836
1837 ;; Set `rung-pos' for the next rung. It's the same thing here as
1838 ;; initially, except that the rung position is set as early as
1839 ;; possible since we can't be in the ending ws of a line comment or
1840 ;; cpp directive now.
1841 (if (setq rung-is-marked next-rung-is-marked)
1842 (setq rung-pos (1- (next-single-property-change
1843 rung-is-marked 'c-is-sws nil rung-end-pos)))
1844 (setq rung-pos next-rung-pos))
1845 (setq safe-start t)))
1846
1847 ;; Make sure that the newly marked `c-in-sws' region doesn't connect to
1848 ;; another one after the point (which might occur when editing inside a
1849 ;; comment or macro).
1850 (when (eq last-put-in-sws-pos (point))
1851 (cond ((< last-put-in-sws-pos (point-max))
1852 (c-debug-sws-msg
1853 "c-forward-sws clearing at %s for cache separation"
1854 last-put-in-sws-pos)
1855 (c-remove-in-sws last-put-in-sws-pos
1856 (1+ last-put-in-sws-pos)))
1857 (t
1858 ;; If at eob we have to clear the last character before the end
1859 ;; instead since the buffer might be narrowed and there might
1860 ;; be a `c-in-sws' after (point-max). In this case it's
1861 ;; necessary to clear both properties.
1862 (c-debug-sws-msg
1863 "c-forward-sws clearing thoroughly at %s for cache separation"
1864 (1- last-put-in-sws-pos))
1865 (c-remove-is-and-in-sws (1- last-put-in-sws-pos)
1866 last-put-in-sws-pos))))
2143fa32 1867 ))))
b2acd789 1868
d9e94c22
MS
1869(defun c-backward-sws ()
1870 ;; Used by `c-backward-syntactic-ws' to implement the unbounded search.
0386b551
AM
1871 ;;
1872 ;; This function might do hidden buffer changes.
d9e94c22
MS
1873
1874 (let (;; `rung-pos' is set to a position as late as possible in the unmarked
1875 ;; part of the simple ws region.
1876 (rung-pos (point)) next-rung-pos last-put-in-sws-pos
1877 rung-is-marked simple-ws-beg cmt-skip-pos)
1878
1879 ;; Skip simple horizontal ws and do a quick check on the preceding
91af3942 1880 ;; character to see if it's anything that can't end syntactic ws, so we can
d9e94c22
MS
1881 ;; bail out early in the majority of cases when there just are a few ws
1882 ;; chars. Newlines are complicated in the backward direction, so we can't
1883 ;; skip over them.
1884 (skip-chars-backward " \t\f")
1885 (when (and (not (bobp))
1886 (save-excursion
1887 (backward-char)
1888 (looking-at c-syntactic-ws-end)))
1889
1890 ;; Try to find a rung position in the simple ws preceding point, so that
1891 ;; we can get a cache hit even if the last bit of the simple ws has
1892 ;; changed recently.
1893 (setq simple-ws-beg (point))
1894 (skip-chars-backward " \t\n\r\f\v")
1895 (if (setq rung-is-marked (text-property-any
1896 (point) (min (1+ rung-pos) (point-max))
1897 'c-is-sws t))
1898 ;; `rung-pos' will be the earliest marked position, which means that
1899 ;; there might be later unmarked parts in the simple ws region.
1900 ;; It's not worth the effort to fix that; the last part of the
1901 ;; simple ws is also typically edited often, so it could be wasted.
1902 (goto-char (setq rung-pos rung-is-marked))
1903 (goto-char simple-ws-beg))
1904
2143fa32 1905 (with-silent-modifications
d9e94c22
MS
1906 (while
1907 (progn
1908 (while
1909 (when (and rung-is-marked
1910 (not (bobp))
1911 (get-text-property (1- (point)) 'c-in-sws))
1912
1913 ;; The following search is the main reason that `c-in-sws'
1914 ;; and `c-is-sws' aren't combined to one property.
1915 (goto-char (previous-single-property-change
1916 (point) 'c-in-sws nil (point-min)))
1917 (unless (get-text-property (point) 'c-is-sws)
1918 ;; If the `c-in-sws' region extended past the first
1919 ;; `c-is-sws' char we have to go forward a bit.
1920 (goto-char (next-single-property-change
1921 (point) 'c-is-sws)))
1922
1923 (c-debug-sws-msg
1924 "c-backward-sws cached move %s <- %s (min %s)"
1925 (point) rung-pos (point-min))
1926
1927 (setq rung-pos (point))
1928 (if (and (< (min (skip-chars-backward " \t\f\v")
1929 (progn
1930 (setq simple-ws-beg (point))
1931 (skip-chars-backward " \t\n\r\f\v")))
1932 0)
1933 (setq rung-is-marked
1934 (text-property-any (point) rung-pos
1935 'c-is-sws t)))
1936 t
1937 (goto-char simple-ws-beg)
1938 nil))
1939
1940 ;; We'll loop here if there is simple ws before the first rung.
1941 ;; That means that there's been some change in it and it's
1942 ;; possible that we've stepped into another ladder, so extend
1943 ;; the previous one to join with it if there is one, and try to
1944 ;; use the cache again.
1945 (c-debug-sws-msg
1946 "c-backward-sws extending rung with [%s..%s] (min %s)"
1947 rung-is-marked rung-pos (point-min))
1948 (unless (get-text-property (1- rung-pos) 'c-is-sws)
1949 ;; Remove any `c-in-sws' property from the last char of
1950 ;; the rung before we mark it with `c-is-sws', so that we
1951 ;; won't connect with the remains of a broken "ladder".
1952 (c-remove-in-sws (1- rung-pos) rung-pos))
1953 (c-put-is-sws rung-is-marked
1954 rung-pos)
1955 (c-put-in-sws rung-is-marked
1956 (1- rung-pos))
1957 (setq rung-pos rung-is-marked
1958 last-put-in-sws-pos rung-pos))
1959
1960 (c-backward-comments)
1961 (setq cmt-skip-pos (point))
a66cd3ee 1962
d9e94c22
MS
1963 (cond
1964 ((and c-opt-cpp-prefix
1965 (/= cmt-skip-pos simple-ws-beg)
1966 (c-beginning-of-macro))
1967 ;; Inside a cpp directive. See if it should be skipped over.
1968 (let ((cpp-beg (point)))
1969
1970 ;; Move back over all line continuations in the region skipped
1971 ;; over by `c-backward-comments'. If we go past it then we
1972 ;; started inside the cpp directive.
1973 (goto-char simple-ws-beg)
1974 (beginning-of-line)
1975 (while (and (> (point) cmt-skip-pos)
1976 (progn (backward-char)
1977 (eq (char-before) ?\\)))
1978 (beginning-of-line))
1979
1980 (if (< (point) cmt-skip-pos)
1981 ;; Don't move past the cpp directive if we began inside
1982 ;; it. Note that the position at the end of the last line
1983 ;; of the macro is also considered to be within it.
1984 (progn (goto-char cmt-skip-pos)
1985 nil)
1986
1987 ;; It's worthwhile to spend a little bit of effort on finding
1988 ;; the end of the macro, to get a good `simple-ws-beg'
1989 ;; position for the cache. Note that `c-backward-comments'
1990 ;; could have stepped over some comments before going into
1991 ;; the macro, and then `simple-ws-beg' must be kept on the
1992 ;; same side of those comments.
1993 (goto-char simple-ws-beg)
1994 (skip-chars-backward " \t\n\r\f\v")
1995 (if (eq (char-before) ?\\)
1996 (forward-char))
1997 (forward-line 1)
1998 (if (< (point) simple-ws-beg)
1999 ;; Might happen if comments after the macro were skipped
2000 ;; over.
2001 (setq simple-ws-beg (point)))
2002
2003 (goto-char cpp-beg)
2004 t)))
2005
2006 ((/= (save-excursion
2007 (skip-chars-forward " \t\n\r\f\v" simple-ws-beg)
2008 (setq next-rung-pos (point)))
2009 simple-ws-beg)
2010 ;; Skipped over comments. Must put point at the end of
2011 ;; the simple ws at point since we might be after a line
2012 ;; comment or cpp directive that's been partially
2013 ;; narrowed out, and we can't risk marking the simple ws
2014 ;; at the end of it.
2015 (goto-char next-rung-pos)
2016 t)))
2017
2018 ;; We've searched over a piece of non-white syntactic ws. See if this
2019 ;; can be cached.
2020 (setq next-rung-pos (point))
2021 (skip-chars-backward " \t\f\v")
2022
2023 (if (or
2024 ;; Cache if we started either from a marked rung or from a
2025 ;; completely uncached position.
2026 rung-is-marked
2027 (not (get-text-property (1- simple-ws-beg) 'c-in-sws))
2028
2029 ;; Cache if there's a marked rung in the encountered simple ws.
2030 (save-excursion
2031 (skip-chars-backward " \t\n\r\f\v")
2032 (text-property-any (point) (min (1+ next-rung-pos) (point-max))
2033 'c-is-sws t)))
a66cd3ee 2034
d9e94c22
MS
2035 (progn
2036 (c-debug-sws-msg
2037 "c-backward-sws caching [%s..%s] - [%s..%s] (min %s)"
2038 (point) (1+ next-rung-pos)
2039 simple-ws-beg (min (1+ rung-pos) (point-max))
2040 (point-min))
2041
2042 ;; Remove the properties for any nested ws that might be cached.
2043 ;; Only necessary for `c-is-sws' since `c-in-sws' will be set
2044 ;; anyway.
2045 (c-remove-is-sws (1+ next-rung-pos) simple-ws-beg)
2046 (unless (and rung-is-marked (= simple-ws-beg rung-pos))
2047 (let ((rung-end-pos (min (1+ rung-pos) (point-max))))
2048 (unless (get-text-property (1- rung-end-pos) 'c-is-sws)
2049 ;; Remove any `c-in-sws' property from the last char of
2050 ;; the rung before we mark it with `c-is-sws', so that we
2051 ;; won't connect with the remains of a broken "ladder".
2052 (c-remove-in-sws (1- rung-end-pos) rung-end-pos))
2053 (c-put-is-sws simple-ws-beg
2054 rung-end-pos)
2055 (setq rung-is-marked t)))
2056 (c-put-in-sws (setq simple-ws-beg (point)
2057 last-put-in-sws-pos simple-ws-beg)
2058 rung-pos)
2059 (c-put-is-sws (setq rung-pos simple-ws-beg)
2060 (1+ next-rung-pos)))
2061
2062 (c-debug-sws-msg
2063 "c-backward-sws not caching [%s..%s] - [%s..%s] (min %s)"
2064 (point) (1+ next-rung-pos)
2065 simple-ws-beg (min (1+ rung-pos) (point-max))
2066 (point-min))
2067 (setq rung-pos next-rung-pos
2068 simple-ws-beg (point))
2069 ))
2070
2071 ;; Make sure that the newly marked `c-in-sws' region doesn't connect to
2072 ;; another one before the point (which might occur when editing inside a
2073 ;; comment or macro).
2074 (when (eq last-put-in-sws-pos (point))
2075 (cond ((< (point-min) last-put-in-sws-pos)
2076 (c-debug-sws-msg
2077 "c-backward-sws clearing at %s for cache separation"
2078 (1- last-put-in-sws-pos))
2079 (c-remove-in-sws (1- last-put-in-sws-pos)
2080 last-put-in-sws-pos))
2081 ((> (point-min) 1)
2082 ;; If at bob and the buffer is narrowed, we have to clear the
2083 ;; character we're standing on instead since there might be a
2084 ;; `c-in-sws' before (point-min). In this case it's necessary
2085 ;; to clear both properties.
2086 (c-debug-sws-msg
2087 "c-backward-sws clearing thoroughly at %s for cache separation"
2088 last-put-in-sws-pos)
2089 (c-remove-is-and-in-sws last-put-in-sws-pos
2090 (1+ last-put-in-sws-pos)))))
2143fa32 2091 ))))
785eecbb 2092
d9e94c22 2093\f
580fba94
AM
2094;; Other whitespace tools
2095(defun c-partial-ws-p (beg end)
2096 ;; Is the region (beg end) WS, and is there WS (or BOB/EOB) next to the
2097 ;; region? This is a "heuristic" function. .....
b414f371 2098 ;;
c421028a
AM
2099 ;; The motivation for the second bit is to check whether removing this
2100 ;; region would coalesce two symbols.
580fba94
AM
2101 ;;
2102 ;; FIXME!!! This function doesn't check virtual semicolons in any way. Be
2103 ;; careful about using this function for, e.g. AWK. (2007/3/7)
2104 (save-excursion
2105 (let ((end+1 (min (1+ end) (point-max))))
2106 (or (progn (goto-char (max (point-min) (1- beg)))
2107 (c-skip-ws-forward end)
2108 (eq (point) end))
2109 (progn (goto-char beg)
2110 (c-skip-ws-forward end+1)
2111 (eq (point) end+1))))))
2112\f
0386b551 2113;; A system for finding noteworthy parens before the point.
e1c458ae 2114
0ec1d2c5
AM
2115(defconst c-state-cache-too-far 5000)
2116;; A maximum comfortable scanning distance, e.g. between
2117;; `c-state-cache-good-pos' and "HERE" (where we call c-parse-state). When
0d26e0b6 2118;; this distance is exceeded, we take "emergency measures", e.g. by clearing
0ec1d2c5
AM
2119;; the cache and starting again from point-min or a beginning of defun. This
2120;; value can be tuned for efficiency or set to a lower value for testing.
2121
d9e94c22
MS
2122(defvar c-state-cache nil)
2123(make-variable-buffer-local 'c-state-cache)
2124;; The state cache used by `c-parse-state' to cut down the amount of
0ec1d2c5
AM
2125;; searching. It's the result from some earlier `c-parse-state' call. See
2126;; `c-parse-state''s doc string for details of its structure.
0386b551 2127;;
d9e94c22
MS
2128;; The use of the cached info is more effective if the next
2129;; `c-parse-state' call is on a line close by the one the cached state
2130;; was made at; the cache can actually slow down a little if the
2131;; cached state was made very far back in the buffer. The cache is
2132;; most effective if `c-parse-state' is used on each line while moving
2133;; forward.
e1c458ae 2134
0386b551
AM
2135(defvar c-state-cache-good-pos 1)
2136(make-variable-buffer-local 'c-state-cache-good-pos)
0ec1d2c5
AM
2137;; This is a position where `c-state-cache' is known to be correct, or
2138;; nil (see below). It's a position inside one of the recorded unclosed
2139;; parens or the top level, but not further nested inside any literal or
2140;; subparen that is closed before the last recorded position.
0386b551
AM
2141;;
2142;; The exact position is chosen to try to be close to yet earlier than
2143;; the position where `c-state-cache' will be called next. Right now
2144;; the heuristic is to set it to the position after the last found
2145;; closing paren (of any type) before the line on which
2146;; `c-parse-state' was called. That is chosen primarily to work well
2147;; with refontification of the current line.
0ec1d2c5
AM
2148;;
2149;; 2009-07-28: When `c-state-point-min' and the last position where
2150;; `c-parse-state' or for which `c-invalidate-state-cache' was called, are
2151;; both in the same literal, there is no such "good position", and
2152;; c-state-cache-good-pos is then nil. This is the ONLY circumstance in which
2153;; it can be nil. In this case, `c-state-point-min-literal' will be non-nil.
2154;;
2155;; 2009-06-12: In a brace desert, c-state-cache-good-pos may also be in
2156;; the middle of the desert, as long as it is not within a brace pair
2157;; recorded in `c-state-cache' or a paren/bracket pair.
2158
2159
2160;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2161;; We maintain a simple cache of positions which aren't in a literal, so as to
2162;; speed up testing for non-literality.
9657183b 2163(defconst c-state-nonlit-pos-interval 3000)
0ec1d2c5
AM
2164;; The approximate interval between entries in `c-state-nonlit-pos-cache'.
2165
2166(defvar c-state-nonlit-pos-cache nil)
2167(make-variable-buffer-local 'c-state-nonlit-pos-cache)
e84efb70
AM
2168;; A list of buffer positions which are known not to be in a literal or a cpp
2169;; construct. This is ordered with higher positions at the front of the list.
2170;; Only those which are less than `c-state-nonlit-pos-cache-limit' are valid.
0ec1d2c5
AM
2171
2172(defvar c-state-nonlit-pos-cache-limit 1)
2173(make-variable-buffer-local 'c-state-nonlit-pos-cache-limit)
2174;; An upper limit on valid entries in `c-state-nonlit-pos-cache'. This is
2175;; reduced by buffer changes, and increased by invocations of
2176;; `c-state-literal-at'.
2177
56d093a9
AM
2178(defvar c-state-semi-nonlit-pos-cache nil)
2179(make-variable-buffer-local 'c-state-semi-nonlit-pos-cache)
2180;; A list of buffer positions which are known not to be in a literal. This is
2181;; ordered with higher positions at the front of the list. Only those which
2182;; are less than `c-state-semi-nonlit-pos-cache-limit' are valid.
2183
2184(defvar c-state-semi-nonlit-pos-cache-limit 1)
2185(make-variable-buffer-local 'c-state-semi-nonlit-pos-cache-limit)
2186;; An upper limit on valid entries in `c-state-semi-nonlit-pos-cache'. This is
2187;; reduced by buffer changes, and increased by invocations of
f003f294 2188;; `c-state-literal-at'. FIXME!!!
56d093a9 2189
691e26ae 2190(defsubst c-state-pp-to-literal (from to &optional not-in-delimiter)
13d49cbb
AM
2191 ;; Do a parse-partial-sexp from FROM to TO, returning either
2192 ;; (STATE TYPE (BEG . END)) if TO is in a literal; or
2193 ;; (STATE) otherwise,
2194 ;; where STATE is the parsing state at TO, TYPE is the type of the literal
2195 ;; (one of 'c, 'c++, 'string) and (BEG . END) is the boundaries of the literal.
2196 ;;
691e26ae
AM
2197 ;; Unless NOT-IN-DELIMITER is non-nil, when TO is inside a two-character
2198 ;; comment opener, this is recognized as being in a comment literal.
2199 ;;
13d49cbb
AM
2200 ;; Only elements 3 (in a string), 4 (in a comment), 5 (following a quote),
2201 ;; 7 (comment type) and 8 (start of comment/string) (and possibly 9) of
2202 ;; STATE are valid.
0ec1d2c5 2203 (save-excursion
13d49cbb 2204 (let ((s (parse-partial-sexp from to))
691e26ae
AM
2205 ty co-st)
2206 (cond
2207 ((or (nth 3 s) (nth 4 s)) ; in a string or comment
13d49cbb
AM
2208 (setq ty (cond
2209 ((nth 3 s) 'string)
691e26ae 2210 ((nth 7 s) 'c++)
13d49cbb 2211 (t 'c)))
0ec1d2c5 2212 (parse-partial-sexp (point) (point-max)
691e26ae
AM
2213 nil ; TARGETDEPTH
2214 nil ; STOPBEFORE
2215 s ; OLDSTATE
2216 'syntax-table) ; stop at end of literal
2217 `(,s ,ty (,(nth 8 s) . ,(point))))
2218
2219 ((and (not not-in-delimiter) ; inside a comment starter
2220 (not (bobp))
2221 (progn (backward-char)
99f7b0a6
AM
2222 (and (not (looking-at "\\s!"))
2223 (looking-at c-comment-start-regexp))))
691e26ae
AM
2224 (setq ty (if (looking-at c-block-comment-start-regexp) 'c 'c++)
2225 co-st (point))
2226 (forward-comment 1)
2227 `(,s ,ty (,co-st . ,(point))))
2228
2229 (t `(,s))))))
13d49cbb
AM
2230
2231(defun c-state-safe-place (here)
2232 ;; Return a buffer position before HERE which is "safe", i.e. outside any
2233 ;; string, comment, or macro.
e84efb70
AM
2234 ;;
2235 ;; NOTE: This function manipulates `c-state-nonlit-pos-cache'. This cache
2236 ;; MAY NOT contain any positions within macros, since macros are frequently
2237 ;; turned into comments by use of the `c-cpp-delimiter' category properties.
2238 ;; We cannot rely on this mechanism whilst determining a cache pos since
2239 ;; this function is also called from outwith `c-parse-state'.
0ec1d2c5
AM
2240 (save-restriction
2241 (widen)
2242 (save-excursion
2243 (let ((c c-state-nonlit-pos-cache)
56d093a9 2244 pos npos high-pos lit macro-beg macro-end)
0ec1d2c5
AM
2245 ;; Trim the cache to take account of buffer changes.
2246 (while (and c (> (car c) c-state-nonlit-pos-cache-limit))
2247 (setq c (cdr c)))
2248 (setq c-state-nonlit-pos-cache c)
2249
2250 (while (and c (> (car c) here))
56d093a9 2251 (setq high-pos (car c))
0ec1d2c5
AM
2252 (setq c (cdr c)))
2253 (setq pos (or (car c) (point-min)))
2254
56d093a9
AM
2255 (unless high-pos
2256 (while
2257 ;; Add an element to `c-state-nonlit-pos-cache' each iteration.
2258 (and
2259 (<= (setq npos (+ pos c-state-nonlit-pos-interval)) here)
2260
2261 ;; Test for being in a literal. If so, go to after it.
2262 (progn
2263 (setq lit (car (cddr (c-state-pp-to-literal pos npos))))
2264 (or (null lit)
2265 (prog1 (<= (cdr lit) here)
2266 (setq npos (cdr lit)))))
2267
2268 ;; Test for being in a macro. If so, go to after it.
2269 (progn
2270 (goto-char npos)
2271 (setq macro-beg
2272 (and (c-beginning-of-macro) (/= (point) npos) (point)))
2273 (when macro-beg
2274 (c-syntactic-end-of-macro)
2275 (or (eobp) (forward-char))
2276 (setq macro-end (point)))
2277 (or (null macro-beg)
2278 (prog1 (<= macro-end here)
2279 (setq npos macro-end)))))
2280
2281 (setq pos npos)
2282 (setq c-state-nonlit-pos-cache (cons pos c-state-nonlit-pos-cache)))
2283 ;; Add one extra element above HERE so as to to avoid the previous
2284 ;; expensive calculation when the next call is close to the current
2285 ;; one. This is especially useful when inside a large macro.
2286 (setq c-state-nonlit-pos-cache (cons npos c-state-nonlit-pos-cache)))
0ec1d2c5
AM
2287
2288 (if (> pos c-state-nonlit-pos-cache-limit)
2289 (setq c-state-nonlit-pos-cache-limit pos))
13d49cbb 2290 pos))))
9858f6c3 2291
56d093a9
AM
2292(defun c-state-semi-safe-place (here)
2293 ;; Return a buffer position before HERE which is "safe", i.e. outside any
2294 ;; string or comment. It may be in a macro.
2295 (save-restriction
2296 (widen)
2297 (save-excursion
2298 (let ((c c-state-semi-nonlit-pos-cache)
2299 pos npos high-pos lit macro-beg macro-end)
2300 ;; Trim the cache to take account of buffer changes.
2301 (while (and c (> (car c) c-state-semi-nonlit-pos-cache-limit))
2302 (setq c (cdr c)))
2303 (setq c-state-semi-nonlit-pos-cache c)
c38e0c97 2304
56d093a9
AM
2305 (while (and c (> (car c) here))
2306 (setq high-pos (car c))
2307 (setq c (cdr c)))
2308 (setq pos (or (car c) (point-min)))
c38e0c97 2309
56d093a9
AM
2310 (unless high-pos
2311 (while
2312 ;; Add an element to `c-state-semi-nonlit-pos-cache' each iteration.
2313 (and
2314 (<= (setq npos (+ pos c-state-nonlit-pos-interval)) here)
c38e0c97 2315
56d093a9
AM
2316 ;; Test for being in a literal. If so, go to after it.
2317 (progn
2318 (setq lit (car (cddr (c-state-pp-to-literal pos npos))))
2319 (or (null lit)
2320 (prog1 (<= (cdr lit) here)
2321 (setq npos (cdr lit))))))
c38e0c97 2322
56d093a9
AM
2323 (setq pos npos)
2324 (setq c-state-semi-nonlit-pos-cache
2325 (cons pos c-state-semi-nonlit-pos-cache))))
2326
2327 (if (> pos c-state-semi-nonlit-pos-cache-limit)
2328 (setq c-state-semi-nonlit-pos-cache-limit pos))
2329 pos))))
2330
13d49cbb
AM
2331(defun c-state-literal-at (here)
2332 ;; If position HERE is inside a literal, return (START . END), the
2333 ;; boundaries of the literal (which may be outside the accessible bit of the
2334 ;; buffer). Otherwise, return nil.
2335 ;;
2336 ;; This function is almost the same as `c-literal-limits'. Previously, it
9858f6c3 2337 ;; differed in that it was a lower level function, and that it rigorously
13d49cbb
AM
2338 ;; followed the syntax from BOB. `c-literal-limits' is now (2011-12)
2339 ;; virtually identical to this function.
2340 (save-restriction
2341 (widen)
2342 (save-excursion
2343 (let ((pos (c-state-safe-place here)))
2344 (car (cddr (c-state-pp-to-literal pos here)))))))
0ec1d2c5
AM
2345
2346(defsubst c-state-lit-beg (pos)
2347 ;; Return the start of the literal containing POS, or POS itself.
2348 (or (car (c-state-literal-at pos))
2349 pos))
2350
2351(defsubst c-state-cache-non-literal-place (pos state)
9c4aeabf 2352 ;; Return a position outside of a string/comment/macro at or before POS.
0ec1d2c5 2353 ;; STATE is the parse-partial-sexp state at POS.
9c4aeabf
AM
2354 (let ((res (if (or (nth 3 state) ; in a string?
2355 (nth 4 state)) ; in a comment?
2356 (nth 8 state)
2357 pos)))
2358 (save-excursion
2359 (goto-char res)
2360 (if (c-beginning-of-macro)
2361 (point)
2362 res))))
0ec1d2c5
AM
2363
2364;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2365;; Stuff to do with point-min, and coping with any literal there.
2366(defvar c-state-point-min 1)
2367(make-variable-buffer-local 'c-state-point-min)
2368;; This is (point-min) when `c-state-cache' was last calculated. A change of
2369;; narrowing is likely to affect the parens that are visible before the point.
2370
2371(defvar c-state-point-min-lit-type nil)
2372(make-variable-buffer-local 'c-state-point-min-lit-type)
2373(defvar c-state-point-min-lit-start nil)
2374(make-variable-buffer-local 'c-state-point-min-lit-start)
2375;; These two variables define the literal, if any, containing point-min.
2376;; Their values are, respectively, 'string, c, or c++, and the start of the
2377;; literal. If there's no literal there, they're both nil.
2378
2379(defvar c-state-min-scan-pos 1)
2380(make-variable-buffer-local 'c-state-min-scan-pos)
2381;; This is the earliest buffer-pos from which scanning can be done. It is
2382;; either the end of the literal containing point-min, or point-min itself.
2383;; It becomes nil if the buffer is changed earlier than this point.
2384(defun c-state-get-min-scan-pos ()
2385 ;; Return the lowest valid scanning pos. This will be the end of the
2386 ;; literal enclosing point-min, or point-min itself.
2387 (or c-state-min-scan-pos
2388 (save-restriction
2389 (save-excursion
2390 (widen)
2391 (goto-char c-state-point-min-lit-start)
2392 (if (eq c-state-point-min-lit-type 'string)
2393 (forward-sexp)
2394 (forward-comment 1))
2395 (setq c-state-min-scan-pos (point))))))
2396
2397(defun c-state-mark-point-min-literal ()
2398 ;; Determine the properties of any literal containing POINT-MIN, setting the
2399 ;; variables `c-state-point-min-lit-type', `c-state-point-min-lit-start',
2400 ;; and `c-state-min-scan-pos' accordingly. The return value is meaningless.
2401 (let ((p-min (point-min))
2402 lit)
2403 (save-restriction
2404 (widen)
2405 (setq lit (c-state-literal-at p-min))
2406 (if lit
2407 (setq c-state-point-min-lit-type
2408 (save-excursion
2409 (goto-char (car lit))
2410 (cond
2411 ((looking-at c-block-comment-start-regexp) 'c)
2412 ((looking-at c-line-comment-starter) 'c++)
2413 (t 'string)))
2414 c-state-point-min-lit-start (car lit)
2415 c-state-min-scan-pos (cdr lit))
2416 (setq c-state-point-min-lit-type nil
2417 c-state-point-min-lit-start nil
2418 c-state-min-scan-pos p-min)))))
2419
2420
2421;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2422;; A variable which signals a brace dessert - helpful for reducing the number
2423;; of fruitless backward scans.
2424(defvar c-state-brace-pair-desert nil)
2425(make-variable-buffer-local 'c-state-brace-pair-desert)
e84efb70 2426;; Used only in `c-append-lower-brace-pair-to-state-cache'. It is set when
0ec1d2c5
AM
2427;; that defun has searched backwards for a brace pair and not found one. Its
2428;; value is either nil or a cons (PA . FROM), where PA is the position of the
2429;; enclosing opening paren/brace/bracket which bounds the backwards search (or
2430;; nil when at top level) and FROM is where the backward search started. It
2431;; is reset to nil in `c-invalidate-state-cache'.
2432
2433
2434;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2435;; Lowish level functions/macros which work directly on `c-state-cache', or a
2436;; list of like structure.
2437(defmacro c-state-cache-top-lparen (&optional cache)
2438 ;; Return the address of the top left brace/bracket/paren recorded in CACHE
2439 ;; (default `c-state-cache') (or nil).
2440 (let ((cash (or cache 'c-state-cache)))
2441 `(if (consp (car ,cash))
2442 (caar ,cash)
2443 (car ,cash))))
2444
2445(defmacro c-state-cache-top-paren (&optional cache)
2446 ;; Return the address of the latest brace/bracket/paren (whether left or
2447 ;; right) recorded in CACHE (default `c-state-cache') or nil.
2448 (let ((cash (or cache 'c-state-cache)))
2449 `(if (consp (car ,cash))
2450 (cdar ,cash)
2451 (car ,cash))))
2452
2453(defmacro c-state-cache-after-top-paren (&optional cache)
2454 ;; Return the position just after the latest brace/bracket/paren (whether
2455 ;; left or right) recorded in CACHE (default `c-state-cache') or nil.
2456 (let ((cash (or cache 'c-state-cache)))
2457 `(if (consp (car ,cash))
2458 (cdar ,cash)
2459 (and (car ,cash)
2460 (1+ (car ,cash))))))
2461
2462(defun c-get-cache-scan-pos (here)
2463 ;; From the state-cache, determine the buffer position from which we might
2464 ;; scan forward to HERE to update this cache. This position will be just
2465 ;; after a paren/brace/bracket recorded in the cache, if possible, otherwise
2466 ;; return the earliest position in the accessible region which isn't within
2467 ;; a literal. If the visible portion of the buffer is entirely within a
2468 ;; literal, return NIL.
2469 (let ((c c-state-cache) elt)
2470 ;(while (>= (or (c-state-cache-top-lparen c) 1) here)
2471 (while (and c
2472 (>= (c-state-cache-top-lparen c) here))
2473 (setq c (cdr c)))
2474
2475 (setq elt (car c))
2476 (cond
2477 ((consp elt)
2478 (if (> (cdr elt) here)
2479 (1+ (car elt))
2480 (cdr elt)))
2481 (elt (1+ elt))
2482 ((<= (c-state-get-min-scan-pos) here)
2483 (c-state-get-min-scan-pos))
2484 (t nil))))
2485
2486;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2487;; Variables which keep track of preprocessor constructs.
8cff5feb
AM
2488(defvar c-state-old-cpp-beg-marker nil)
2489(make-variable-buffer-local 'c-state-old-cpp-beg-marker)
0ec1d2c5
AM
2490(defvar c-state-old-cpp-beg nil)
2491(make-variable-buffer-local 'c-state-old-cpp-beg)
8cff5feb
AM
2492(defvar c-state-old-cpp-end-marker nil)
2493(make-variable-buffer-local 'c-state-old-cpp-end-marker)
0ec1d2c5
AM
2494(defvar c-state-old-cpp-end nil)
2495(make-variable-buffer-local 'c-state-old-cpp-end)
2496;; These are the limits of the macro containing point at the previous call of
2497;; `c-parse-state', or nil.
2498
2499;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
c7015153 2500;; Defuns which analyze the buffer, yet don't change `c-state-cache'.
c0209c2c
AM
2501(defun c-state-balance-parens-backwards (here- here+ top)
2502 ;; Return the position of the opening paren/brace/bracket before HERE- which
2503 ;; matches the outermost close p/b/b between HERE+ and TOP. Except when
2504 ;; there's a macro, HERE- and HERE+ are the same. Like this:
0ec1d2c5 2505 ;;
c0209c2c
AM
2506 ;; ............................................
2507 ;; | |
2508 ;; ( [ ( .........#macro.. ) ( ) ] )
2509 ;; ^ ^ ^ ^
2510 ;; | | | |
2511 ;; return HERE- HERE+ TOP
0ec1d2c5
AM
2512 ;;
2513 ;; If there aren't enough opening paren/brace/brackets, return the position
c0209c2c 2514 ;; of the outermost one found, or HERE- if there are none. If there are no
da6062e6 2515 ;; closing p/b/bs between HERE+ and TOP, return HERE-. HERE-/+ and TOP
c0209c2c
AM
2516 ;; must not be inside literals. Only the accessible portion of the buffer
2517 ;; will be scanned.
2518
2519 ;; PART 1: scan from `here+' up to `top', accumulating ")"s which enclose
2520 ;; `here'. Go round the next loop each time we pass over such a ")". These
2521 ;; probably match "("s before `here-'.
0ec1d2c5
AM
2522 (let (pos pa ren+1 lonely-rens)
2523 (save-excursion
2524 (save-restriction
2525 (narrow-to-region (point-min) top) ; This can move point, sometimes.
c0209c2c 2526 (setq pos here+)
0ec1d2c5
AM
2527 (c-safe
2528 (while
2529 (setq ren+1 (scan-lists pos 1 1)) ; might signal
2530 (setq lonely-rens (cons ren+1 lonely-rens)
2531 pos ren+1)))))
2532
c0209c2c 2533 ;; PART 2: Scan back before `here-' searching for the "("s
0ec1d2c5
AM
2534 ;; matching/mismatching the ")"s found above. We only need to direct the
2535 ;; caller to scan when we've encountered unmatched right parens.
c0209c2c
AM
2536 (setq pos here-)
2537 (when lonely-rens
2538 (c-safe
2539 (while
2540 (and lonely-rens ; actual values aren't used.
2541 (setq pa (scan-lists pos -1 1)))
2542 (setq pos pa)
2543 (setq lonely-rens (cdr lonely-rens)))))
2544 pos))
0ec1d2c5
AM
2545
2546(defun c-parse-state-get-strategy (here good-pos)
2547 ;; Determine the scanning strategy for adjusting `c-parse-state', attempting
53964682 2548 ;; to minimize the amount of scanning. HERE is the pertinent position in
0ec1d2c5
AM
2549 ;; the buffer, GOOD-POS is a position where `c-state-cache' (possibly with
2550 ;; its head trimmed) is known to be good, or nil if there is no such
2551 ;; position.
2552 ;;
2553 ;; The return value is a list, one of the following:
2554 ;;
267eacb5 2555 ;; o - ('forward START-POINT) - scan forward from START-POINT,
99f7b0a6
AM
2556 ;; which is not less than the highest position in `c-state-cache' below HERE,
2557 ;; which is after GOOD-POS.
267eacb5 2558 ;; o - ('backward nil) - scan backwards (from HERE).
99f7b0a6
AM
2559 ;; o - ('back-and-forward START-POINT) - like 'forward, but when HERE is earlier
2560 ;; than GOOD-POS.
267eacb5 2561 ;; o - ('IN-LIT nil) - point is inside the literal containing point-min.
0ec1d2c5 2562 (let ((cache-pos (c-get-cache-scan-pos here)) ; highest position below HERE in cache (or 1)
d23b0804
AM
2563 strategy ; 'forward, 'backward, or 'IN-LIT.
2564 start-point)
0ec1d2c5
AM
2565 (setq good-pos (or good-pos (c-state-get-min-scan-pos)))
2566 (cond
2567 ((< here (c-state-get-min-scan-pos))
d23b0804 2568 (setq strategy 'IN-LIT))
0ec1d2c5
AM
2569 ((<= good-pos here)
2570 (setq strategy 'forward
d23b0804 2571 start-point (max good-pos cache-pos)))
0ec1d2c5 2572 ((< (- good-pos here) (- here cache-pos)) ; FIXME!!! ; apply some sort of weighting.
d23b0804 2573 (setq strategy 'backward))
0ec1d2c5 2574 (t
99f7b0a6 2575 (setq strategy 'back-and-forward
d23b0804 2576 start-point cache-pos)))
99f7b0a6 2577 (list strategy start-point)))
0ec1d2c5
AM
2578
2579
2580;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2581;; Routines which change `c-state-cache' and associated values.
2582(defun c-renarrow-state-cache ()
2583 ;; The region (more precisely, point-min) has changed since we
2584 ;; calculated `c-state-cache'. Amend `c-state-cache' accordingly.
2585 (if (< (point-min) c-state-point-min)
2586 ;; If point-min has MOVED BACKWARDS then we drop the state completely.
2587 ;; It would be possible to do a better job here and recalculate the top
2588 ;; only.
2589 (progn
2590 (c-state-mark-point-min-literal)
2591 (setq c-state-cache nil
2592 c-state-cache-good-pos c-state-min-scan-pos
2593 c-state-brace-pair-desert nil))
2594
2595 ;; point-min has MOVED FORWARD.
2596
2597 ;; Is the new point-min inside a (different) literal?
2598 (unless (and c-state-point-min-lit-start ; at prev. point-min
2599 (< (point-min) (c-state-get-min-scan-pos)))
2600 (c-state-mark-point-min-literal))
2601
2602 ;; Cut off a bit of the tail from `c-state-cache'.
2603 (let ((ptr (cons nil c-state-cache))
2604 pa)
2605 (while (and (setq pa (c-state-cache-top-lparen (cdr ptr)))
2606 (>= pa (point-min)))
2607 (setq ptr (cdr ptr)))
2608
2609 (when (consp ptr)
2610 (if (eq (cdr ptr) c-state-cache)
2611 (setq c-state-cache nil
2612 c-state-cache-good-pos c-state-min-scan-pos)
2613 (setcdr ptr nil)
2614 (setq c-state-cache-good-pos (1+ (c-state-cache-top-lparen))))
2615 )))
2616
2617 (setq c-state-point-min (point-min)))
2618
d23b0804 2619(defun c-append-lower-brace-pair-to-state-cache (from here &optional upper-lim)
583e23bd
AM
2620 ;; If there is a brace pair preceding FROM in the buffer, at the same level
2621 ;; of nesting (not necessarily immediately preceding), push a cons onto
2622 ;; `c-state-cache' to represent it. FROM must not be inside a literal. If
2623 ;; UPPER-LIM is non-nil, we append the highest brace pair whose "}" is below
2624 ;; UPPER-LIM.
0ec1d2c5
AM
2625 ;;
2626 ;; Return non-nil when this has been done.
2627 ;;
583e23bd
AM
2628 ;; The situation it copes with is this transformation:
2629 ;;
2630 ;; OLD: { (.) {...........}
2631 ;; ^ ^
2632 ;; FROM HERE
c38e0c97 2633 ;;
583e23bd
AM
2634 ;; NEW: { {....} (.) {.........
2635 ;; ^ ^ ^
2636 ;; LOWER BRACE PAIR HERE or HERE
c38e0c97 2637 ;;
0ec1d2c5
AM
2638 ;; This routine should be fast. Since it can get called a LOT, we maintain
2639 ;; `c-state-brace-pair-desert', a small cache of "failures", such that we
2640 ;; reduce the time wasted in repeated fruitless searches in brace deserts.
2641 (save-excursion
2642 (save-restriction
d23b0804 2643 (let* (new-cons
8cff5feb
AM
2644 (cache-pos (c-state-cache-top-lparen)) ; might be nil.
2645 (macro-start-or-from
2646 (progn (goto-char from)
2647 (c-beginning-of-macro)
2648 (point)))
2649 (bra ; Position of "{".
2650 ;; Don't start scanning in the middle of a CPP construct unless
2651 ;; it contains HERE - these constructs, in Emacs, are "commented
2652 ;; out" with category properties.
2653 (if (eq (c-get-char-property macro-start-or-from 'category)
2654 'c-cpp-delimiter)
2655 macro-start-or-from
2656 from))
2657 ce) ; Position of "}"
0ec1d2c5
AM
2658 (or upper-lim (setq upper-lim from))
2659
2660 ;; If we're essentially repeating a fruitless search, just give up.
2661 (unless (and c-state-brace-pair-desert
2662 (eq cache-pos (car c-state-brace-pair-desert))
b03aabda
AM
2663 (or (null (car c-state-brace-pair-desert))
2664 (> from (car c-state-brace-pair-desert)))
0ec1d2c5 2665 (<= from (cdr c-state-brace-pair-desert)))
b03aabda 2666 ;; DESERT-LIM. Avoid repeated searching through the cached desert.
583e23bd
AM
2667 (let ((desert-lim
2668 (and c-state-brace-pair-desert
2669 (eq cache-pos (car c-state-brace-pair-desert))
b03aabda 2670 (>= from (cdr c-state-brace-pair-desert))
583e23bd
AM
2671 (cdr c-state-brace-pair-desert)))
2672 ;; CACHE-LIM. This limit will be necessary when an opening
2673 ;; paren at `cache-pos' has just had its matching close paren
b03aabda
AM
2674 ;; inserted into the buffer. `cache-pos' continues to be a
2675 ;; search bound, even though the algorithm below would skip
2676 ;; over the new paren pair.
583e23bd
AM
2677 (cache-lim (and cache-pos (< cache-pos from) cache-pos)))
2678 (narrow-to-region
2679 (cond
2680 ((and desert-lim cache-lim)
2681 (max desert-lim cache-lim))
2682 (desert-lim)
2683 (cache-lim)
2684 ((point-min)))
d23b0804
AM
2685 ;; The top limit is EOB to ensure that `bra' is inside the
2686 ;; accessible part of the buffer at the next scan operation.
2687 (1+ (buffer-size))))
0ec1d2c5
AM
2688
2689 ;; In the next pair of nested loops, the inner one moves back past a
2690 ;; pair of (mis-)matching parens or brackets; the outer one moves
2691 ;; back over a sequence of unmatched close brace/paren/bracket each
2692 ;; time round.
2693 (while
2694 (progn
2695 (c-safe
2696 (while
2697 (and (setq ce (scan-lists bra -1 -1)) ; back past )/]/}; might signal
2698 (setq bra (scan-lists ce -1 1)) ; back past (/[/{; might signal
267eacb5
AM
2699 (or (> bra here) ;(> ce here)
2700 (and
2701 (< ce here)
2702 (or (not (eq (char-after bra) ?\{))
2703 (and (goto-char bra)
2704 (c-beginning-of-macro)
2705 (< (point) macro-start-or-from))))))))
0ec1d2c5
AM
2706 (and ce (< ce bra)))
2707 (setq bra ce)) ; If we just backed over an unbalanced closing
2708 ; brace, ignore it.
2709
267eacb5 2710 (if (and ce (< ce here) (< bra ce) (eq (char-after bra) ?\{))
0ec1d2c5
AM
2711 ;; We've found the desired brace-pair.
2712 (progn
2713 (setq new-cons (cons bra (1+ ce)))
2714 (cond
2715 ((consp (car c-state-cache))
2716 (setcar c-state-cache new-cons))
2717 ((and (numberp (car c-state-cache)) ; probably never happens
2718 (< ce (car c-state-cache)))
2719 (setcdr c-state-cache
2720 (cons new-cons (cdr c-state-cache))))
2721 (t (setq c-state-cache (cons new-cons c-state-cache)))))
2722
583e23bd 2723 ;; We haven't found a brace pair. Record this in the cache.
267eacb5
AM
2724 (setq c-state-brace-pair-desert
2725 (cons (if (and ce (< bra ce) (> ce here)) ; {..} straddling HERE?
2726 bra
2727 (point-min))
2728 (min here from)))))))))
0ec1d2c5
AM
2729
2730(defsubst c-state-push-any-brace-pair (bra+1 macro-start-or-here)
2731 ;; If BRA+1 is nil, do nothing. Otherwise, BRA+1 is the buffer position
2732 ;; following a {, and that brace has a (mis-)matching } (or ]), and we
2733 ;; "push" "a" brace pair onto `c-state-cache'.
2734 ;;
2735 ;; Here "push" means overwrite the top element if it's itself a brace-pair,
2736 ;; otherwise push it normally.
2737 ;;
2738 ;; The brace pair we push is normally the one surrounding BRA+1, but if the
2739 ;; latter is inside a macro, not being a macro containing
2740 ;; MACRO-START-OR-HERE, we scan backwards through the buffer for a non-macro
2741 ;; base pair. This latter case is assumed to be rare.
2742 ;;
2743 ;; Note: POINT is not preserved in this routine.
2744 (if bra+1
2745 (if (or (> bra+1 macro-start-or-here)
2746 (progn (goto-char bra+1)
2747 (not (c-beginning-of-macro))))
2748 (setq c-state-cache
2749 (cons (cons (1- bra+1)
2750 (scan-lists bra+1 1 1))
2751 (if (consp (car c-state-cache))
2752 (cdr c-state-cache)
2753 c-state-cache)))
d23b0804 2754 ;; N.B. This defsubst codes one method for the simple, normal case,
0ec1d2c5 2755 ;; and a more sophisticated, slower way for the general case. Don't
e1dbe924 2756 ;; eliminate this defsubst - it's a speed optimization.
d23b0804 2757 (c-append-lower-brace-pair-to-state-cache (1- bra+1) (point-max)))))
0ec1d2c5 2758
d23b0804
AM
2759(defun c-append-to-state-cache (from here)
2760 ;; Scan the buffer from FROM to HERE, adding elements into `c-state-cache'
2761 ;; for braces etc. Return a candidate for `c-state-cache-good-pos'.
0ec1d2c5
AM
2762 ;;
2763 ;; FROM must be after the latest brace/paren/bracket in `c-state-cache', if
2764 ;; any. Typically, it is immediately after it. It must not be inside a
2765 ;; literal.
d23b0804 2766 (let ((here-bol (c-point 'bol here))
0ec1d2c5 2767 (macro-start-or-here
d23b0804 2768 (save-excursion (goto-char here)
0ec1d2c5
AM
2769 (if (c-beginning-of-macro)
2770 (point)
d23b0804 2771 here)))
0ec1d2c5
AM
2772 pa+1 ; pos just after an opening PAren (or brace).
2773 (ren+1 from) ; usually a pos just after an closing paREN etc.
2774 ; Is actually the pos. to scan for a (/{/[ from,
2775 ; which sometimes is after a silly )/}/].
2776 paren+1 ; Pos after some opening or closing paren.
2777 paren+1s ; A list of `paren+1's; used to determine a
2778 ; good-pos.
2779 bra+1 ce+1 ; just after L/R bra-ces.
2780 bra+1s ; list of OLD values of bra+1.
2781 mstart) ; start of a macro.
2782
2783 (save-excursion
d23b0804
AM
2784 (save-restriction
2785 (narrow-to-region (point-min) here)
2786 ;; Each time round the following loop, we enter a successively deeper
2787 ;; level of brace/paren nesting. (Except sometimes we "continue at
2788 ;; the existing level".) `pa+1' is a pos inside an opening
2789 ;; brace/paren/bracket, usually just after it.
2790 (while
2791 (progn
2792 ;; Each time round the next loop moves forward over an opening then
2793 ;; a closing brace/bracket/paren. This loop is white hot, so it
2794 ;; plays ugly tricks to go fast. DON'T PUT ANYTHING INTO THIS
2795 ;; LOOP WHICH ISN'T ABSOLUTELY NECESSARY!!! It terminates when a
2796 ;; call of `scan-lists' signals an error, which happens when there
2797 ;; are no more b/b/p's to scan.
2798 (c-safe
2799 (while t
2800 (setq pa+1 (scan-lists ren+1 1 -1) ; Into (/{/[; might signal
2801 paren+1s (cons pa+1 paren+1s))
2802 (setq ren+1 (scan-lists pa+1 1 1)) ; Out of )/}/]; might signal
2803 (if (and (eq (char-before pa+1) ?{)) ; Check for a macro later.
2804 (setq bra+1 pa+1))
2805 (setcar paren+1s ren+1)))
2806
2807 (if (and pa+1 (> pa+1 ren+1))
2808 ;; We've just entered a deeper nesting level.
2809 (progn
2810 ;; Insert the brace pair (if present) and the single open
2811 ;; paren/brace/bracket into `c-state-cache' It cannot be
2812 ;; inside a macro, except one around point, because of what
2813 ;; `c-neutralize-syntax-in-CPP' has done.
2814 (c-state-push-any-brace-pair bra+1 macro-start-or-here)
2815 ;; Insert the opening brace/bracket/paren position.
2816 (setq c-state-cache (cons (1- pa+1) c-state-cache))
2817 ;; Clear admin stuff for the next more nested part of the scan.
2818 (setq ren+1 pa+1 pa+1 nil bra+1 nil bra+1s nil)
2819 t) ; Carry on the loop
2820
2821 ;; All open p/b/b's at this nesting level, if any, have probably
2822 ;; been closed by matching/mismatching ones. We're probably
2823 ;; finished - we just need to check for having found an
2824 ;; unmatched )/}/], which we ignore. Such a )/}/] can't be in a
2825 ;; macro, due the action of `c-neutralize-syntax-in-CPP'.
2826 (c-safe (setq ren+1 (scan-lists ren+1 1 1)))))) ; acts as loop control.
2827
2828 ;; Record the final, innermost, brace-pair if there is one.
2829 (c-state-push-any-brace-pair bra+1 macro-start-or-here)
2830
2831 ;; Determine a good pos
2832 (while (and (setq paren+1 (car paren+1s))
2833 (> (if (> paren+1 macro-start-or-here)
2834 paren+1
2835 (goto-char paren+1)
2836 (setq mstart (and (c-beginning-of-macro)
2837 (point)))
2838 (or mstart paren+1))
2839 here-bol))
2840 (setq paren+1s (cdr paren+1s)))
2841 (cond
2842 ((and paren+1 mstart)
2843 (min paren+1 mstart))
2844 (paren+1)
2845 (t from))))))
2846
2847(defun c-remove-stale-state-cache (start-point here pps-point)
0ec1d2c5 2848 ;; Remove stale entries from the `c-cache-state', i.e. those which will
99f7b0a6
AM
2849 ;; not be in it when it is amended for position HERE. This may involve
2850 ;; replacing a CONS element for a brace pair containing HERE with its car.
2851 ;; Additionally, the "outermost" open-brace entry before HERE will be
2852 ;; converted to a cons if the matching close-brace is below HERE.
0ec1d2c5 2853 ;;
267eacb5 2854 ;; START-POINT is a "maximal" "safe position" - there must be no open
d23b0804 2855 ;; parens/braces/brackets between START-POINT and HERE.
0ec1d2c5
AM
2856 ;;
2857 ;; As a second thing, calculate the result of parse-partial-sexp at
267eacb5 2858 ;; PPS-POINT, w.r.t. START-POINT. The motivation here is that
0ec1d2c5 2859 ;; `c-state-cache-good-pos' may become PPS-POINT, but the caller may need to
267eacb5 2860 ;; adjust it to get outside a string/comment. (Sorry about this! The code
0ec1d2c5
AM
2861 ;; needs to be FAST).
2862 ;;
99f7b0a6 2863 ;; Return a list (GOOD-POS SCAN-BACK-POS CONS-SEPARATED PPS-STATE), where
0ec1d2c5
AM
2864 ;; o - GOOD-POS is a position where the new value `c-state-cache' is known
2865 ;; to be good (we aim for this to be as high as possible);
2866 ;; o - SCAN-BACK-POS, if not nil, indicates there may be a brace pair
2867 ;; preceding POS which needs to be recorded in `c-state-cache'. It is a
267eacb5
AM
2868 ;; position to scan backwards from. It is the position of the "{" of the
2869 ;; last element to be removed from `c-state-cache', when that elt is a
2870 ;; cons, otherwise nil.
99f7b0a6
AM
2871 ;; o - CONS-SEPARATED is t when a cons element in `c-state-cache' has been
2872 ;; replaced by its car because HERE lies inside the brace pair represented
2873 ;; by the cons.
0ec1d2c5 2874 ;; o - PPS-STATE is the parse-partial-sexp state at PPS-POINT.
d23b0804
AM
2875 (save-excursion
2876 (save-restriction
2877 (narrow-to-region 1 (point-max))
2878 (let* ((in-macro-start ; start of macro containing HERE or nil.
0ec1d2c5 2879 (save-excursion
d23b0804 2880 (goto-char here)
0ec1d2c5
AM
2881 (and (c-beginning-of-macro)
2882 (point))))
267eacb5
AM
2883 (start-point-actual-macro-start ; Start of macro containing
2884 ; start-point or nil
d23b0804 2885 (and (< start-point here)
0ec1d2c5 2886 (save-excursion
267eacb5 2887 (goto-char start-point)
0ec1d2c5
AM
2888 (and (c-beginning-of-macro)
2889 (point)))))
267eacb5 2890 (start-point-actual-macro-end ; End of this macro, (maybe
d23b0804 2891 ; HERE), or nil.
267eacb5 2892 (and start-point-actual-macro-start
0ec1d2c5 2893 (save-excursion
267eacb5 2894 (goto-char start-point-actual-macro-start)
0ec1d2c5
AM
2895 (c-end-of-macro)
2896 (point))))
267eacb5 2897 pps-state ; Will be 9 or 10 elements long.
0ec1d2c5
AM
2898 pos
2899 upper-lim ; ,beyond which `c-state-cache' entries are removed
2900 scan-back-pos
99f7b0a6 2901 cons-separated
0ec1d2c5
AM
2902 pair-beg pps-point-state target-depth)
2903
d23b0804
AM
2904 ;; Remove entries beyond HERE. Also remove any entries inside
2905 ;; a macro, unless HERE is in the same macro.
0ec1d2c5
AM
2906 (setq upper-lim
2907 (if (or (null c-state-old-cpp-beg)
d23b0804
AM
2908 (and (> here c-state-old-cpp-beg)
2909 (< here c-state-old-cpp-end)))
2910 here
2911 (min here c-state-old-cpp-beg)))
657071fc 2912 (while (and c-state-cache (>= (c-state-cache-top-lparen) upper-lim))
8cff5feb 2913 (setq scan-back-pos (car-safe (car c-state-cache)))
0ec1d2c5 2914 (setq c-state-cache (cdr c-state-cache)))
8cff5feb 2915
0ec1d2c5
AM
2916 ;; If `upper-lim' is inside the last recorded brace pair, remove its
2917 ;; RBrace and indicate we'll need to search backwards for a previous
2918 ;; brace pair.
2919 (when (and c-state-cache
2920 (consp (car c-state-cache))
2921 (> (cdar c-state-cache) upper-lim))
2922 (setcar c-state-cache (caar c-state-cache))
99f7b0a6
AM
2923 (setq scan-back-pos (car c-state-cache)
2924 cons-separated t))
0ec1d2c5
AM
2925
2926 ;; The next loop jumps forward out of a nested level of parens each
2927 ;; time round; the corresponding elements in `c-state-cache' are
2928 ;; removed. `pos' is just after the brace-pair or the open paren at
267eacb5 2929 ;; (car c-state-cache). There can be no open parens/braces/brackets
d23b0804 2930 ;; between `start-point'/`start-point-actual-macro-start' and HERE,
0ec1d2c5 2931 ;; due to the interface spec to this function.
267eacb5
AM
2932 (setq pos (if (and start-point-actual-macro-end
2933 (not (eq start-point-actual-macro-start
eb2f6eeb 2934 in-macro-start)))
267eacb5 2935 (1+ start-point-actual-macro-end) ; get outside the macro as
0ec1d2c5 2936 ; marked by a `category' text property.
267eacb5 2937 start-point))
0ec1d2c5
AM
2938 (goto-char pos)
2939 (while (and c-state-cache
c38e0c97 2940 (or (numberp (car c-state-cache)) ; Have we a { at all?
d23b0804
AM
2941 (cdr c-state-cache))
2942 (< (point) here))
0ec1d2c5
AM
2943 (cond
2944 ((null pps-state) ; first time through
2945 (setq target-depth -1))
2946 ((eq (car pps-state) target-depth) ; found closing ),},]
2947 (setq target-depth (1- (car pps-state))))
2948 ;; Do nothing when we've merely reached pps-point.
2949 )
2950
2951 ;; Scan!
2952 (setq pps-state
2953 (parse-partial-sexp
d23b0804 2954 (point) (if (< (point) pps-point) pps-point here)
0ec1d2c5
AM
2955 target-depth
2956 nil pps-state))
2957
2958 (if (= (point) pps-point)
2959 (setq pps-point-state pps-state))
2960
2961 (when (eq (car pps-state) target-depth)
2962 (setq pos (point)) ; POS is now just after an R-paren/brace.
2963 (cond
2964 ((and (consp (car c-state-cache))
2965 (eq (point) (cdar c-state-cache)))
2966 ;; We've just moved out of the paren pair containing the brace-pair
2967 ;; at (car c-state-cache). `pair-beg' is where the open paren is,
2968 ;; and is potentially where the open brace of a cons in
2969 ;; c-state-cache will be.
2970 (setq pair-beg (car-safe (cdr c-state-cache))
2971 c-state-cache (cdr-safe (cdr c-state-cache)))) ; remove {}pair + containing Lparen.
2972 ((numberp (car c-state-cache))
2973 (setq pair-beg (car c-state-cache)
2974 c-state-cache (cdr c-state-cache))) ; remove this
2975 ; containing Lparen
2976 ((numberp (cadr c-state-cache))
2977 (setq pair-beg (cadr c-state-cache)
2978 c-state-cache (cddr c-state-cache))) ; Remove a paren pair
2979 ; together with enclosed brace pair.
2980 ;; (t nil) ; Ignore an unmated Rparen.
2981 )))
2982
2983 (if (< (point) pps-point)
2984 (setq pps-state (parse-partial-sexp (point) pps-point
2985 nil nil ; TARGETDEPTH, STOPBEFORE
2986 pps-state)))
2987
2988 ;; If the last paren pair we moved out of was actually a brace pair,
2989 ;; insert it into `c-state-cache'.
2990 (when (and pair-beg (eq (char-after pair-beg) ?{))
2991 (if (consp (car-safe c-state-cache))
2992 (setq c-state-cache (cdr c-state-cache)))
2993 (setq c-state-cache (cons (cons pair-beg pos)
2994 c-state-cache)))
2995
99f7b0a6 2996 (list pos scan-back-pos cons-separated pps-state)))))
0ec1d2c5 2997
267eacb5 2998(defun c-remove-stale-state-cache-backwards (here)
0ec1d2c5
AM
2999 ;; Strip stale elements of `c-state-cache' by moving backwards through the
3000 ;; buffer, and inform the caller of the scenario detected.
3001 ;;
3002 ;; HERE is the position we're setting `c-state-cache' for.
267eacb5
AM
3003 ;; CACHE-POS (a locally bound variable) is just after the latest recorded
3004 ;; position in `c-state-cache' before HERE, or a position at or near
3005 ;; point-min which isn't in a literal.
0ec1d2c5
AM
3006 ;;
3007 ;; This function must only be called only when (> `c-state-cache-good-pos'
3008 ;; HERE). Usually the gap between CACHE-POS and HERE is large. It is thus
e1dbe924 3009 ;; optimized to eliminate (or minimize) scanning between these two
0ec1d2c5
AM
3010 ;; positions.
3011 ;;
3012 ;; Return a three element list (GOOD-POS SCAN-BACK-POS FWD-FLAG), where:
3013 ;; o - GOOD-POS is a "good position", where `c-state-cache' is valid, or
3014 ;; could become so after missing elements are inserted into
3015 ;; `c-state-cache'. This is JUST AFTER an opening or closing
3016 ;; brace/paren/bracket which is already in `c-state-cache' or just before
3017 ;; one otherwise. exceptionally (when there's no such b/p/b handy) the BOL
3018 ;; before `here''s line, or the start of the literal containing it.
3019 ;; o - SCAN-BACK-POS, if non-nil, indicates there may be a brace pair
3020 ;; preceding POS which isn't recorded in `c-state-cache'. It is a position
3021 ;; to scan backwards from.
3022 ;; o - FWD-FLAG, if non-nil, indicates there may be parens/braces between
3023 ;; POS and HERE which aren't recorded in `c-state-cache'.
3024 ;;
3025 ;; The comments in this defun use "paren" to mean parenthesis or square
3026 ;; bracket (as contrasted with a brace), and "(" and ")" likewise.
3027 ;;
267eacb5
AM
3028 ;; . {..} (..) (..) ( .. { } ) (...) ( .... . ..)
3029 ;; | | | | | |
3030 ;; CP E here D C good
3031 (let ((cache-pos (c-get-cache-scan-pos here)) ; highest position below HERE in cache (or 1)
3032 (pos c-state-cache-good-pos)
0ec1d2c5
AM
3033 pa ren ; positions of "(" and ")"
3034 dropped-cons ; whether the last element dropped from `c-state-cache'
3035 ; was a cons (representing a brace-pair)
3036 good-pos ; see above.
3037 lit ; (START . END) of a literal containing some point.
3038 here-lit-start here-lit-end ; bounds of literal containing `here'
3039 ; or `here' itself.
c0209c2c 3040 here- here+ ; start/end of macro around HERE, or HERE
0ec1d2c5 3041 (here-bol (c-point 'bol here))
c400c4d7 3042 (too-far-back (max (- here c-state-cache-too-far) (point-min))))
0ec1d2c5
AM
3043
3044 ;; Remove completely irrelevant entries from `c-state-cache'.
3045 (while (and c-state-cache
3046 (>= (setq pa (c-state-cache-top-lparen)) here))
3047 (setq dropped-cons (consp (car c-state-cache)))
3048 (setq c-state-cache (cdr c-state-cache))
3049 (setq pos pa))
3050 ;; At this stage, (> pos here);
3051 ;; (< (c-state-cache-top-lparen) here) (or is nil).
3052
c0209c2c
AM
3053 (cond
3054 ((and (consp (car c-state-cache))
3055 (> (cdar c-state-cache) here))
3056 ;; CASE 1: The top of the cache is a brace pair which now encloses
3057 ;; `here'. As good-pos, return the address. of the "{". Since we've no
3058 ;; knowledge of what's inside these braces, we have no alternative but
3059 ;; to direct the caller to scan the buffer from the opening brace.
3060 (setq pos (caar c-state-cache))
3061 (setcar c-state-cache pos)
3062 (list (1+ pos) pos t)) ; return value. We've just converted a brace pair
3063 ; entry into a { entry, so the caller needs to
3064 ; search for a brace pair before the {.
3065
3066 ;; `here' might be inside a literal. Check for this.
3067 ((progn
3068 (setq lit (c-state-literal-at here)
3069 here-lit-start (or (car lit) here)
3070 here-lit-end (or (cdr lit) here))
3071 ;; Has `here' just "newly entered" a macro?
3072 (save-excursion
3073 (goto-char here-lit-start)
3074 (if (and (c-beginning-of-macro)
3075 (or (null c-state-old-cpp-beg)
3076 (not (= (point) c-state-old-cpp-beg))))
3077 (progn
3078 (setq here- (point))
3079 (c-end-of-macro)
3080 (setq here+ (point)))
3081 (setq here- here-lit-start
3082 here+ here-lit-end)))
3083
3084 ;; `here' might be nested inside any depth of parens (or brackets but
3085 ;; not braces). Scan backwards to find the outermost such opening
3086 ;; paren, if there is one. This will be the scan position to return.
3087 (save-restriction
3088 (narrow-to-region cache-pos (point-max))
3089 (setq pos (c-state-balance-parens-backwards here- here+ pos)))
3090 nil)) ; for the cond
3091
3092 ((< pos here-lit-start)
3093 ;; CASE 2: Address of outermost ( or [ which now encloses `here', but
3094 ;; didn't enclose the (previous) `c-state-cache-good-pos'. If there is
3095 ;; a brace pair preceding this, it will already be in `c-state-cache',
3096 ;; unless there was a brace pair after it, i.e. there'll only be one to
3097 ;; scan for if we've just deleted one.
3098 (list pos (and dropped-cons pos) t)) ; Return value.
3099
3100 ;; `here' isn't enclosed in a (previously unrecorded) bracket/paren.
3101 ;; Further forward scanning isn't needed, but we still need to find a
3102 ;; GOOD-POS. Step out of all enclosing "("s on HERE's line.
3103 ((progn
0ec1d2c5
AM
3104 (save-restriction
3105 (narrow-to-region here-bol (point-max))
3106 (setq pos here-lit-start)
3107 (c-safe (while (setq pa (scan-lists pos -1 1))
3108 (setq pos pa)))) ; might signal
c0209c2c
AM
3109 nil)) ; for the cond
3110
3111 ((setq ren (c-safe-scan-lists pos -1 -1 too-far-back))
3112 ;; CASE 3: After a }/)/] before `here''s BOL.
3113 (list (1+ ren) (and dropped-cons pos) nil)) ; Return value
3114
3115 (t
3116 ;; CASE 4; Best of a bad job: BOL before `here-bol', or beginning of
3117 ;; literal containing it.
3118 (setq good-pos (c-state-lit-beg (c-point 'bopl here-bol)))
3119 (list good-pos (and dropped-cons good-pos) nil)))))
0ec1d2c5
AM
3120
3121
3122;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
3123;; Externally visible routines.
3124
3125(defun c-state-cache-init ()
3126 (setq c-state-cache nil
3127 c-state-cache-good-pos 1
3128 c-state-nonlit-pos-cache nil
3129 c-state-nonlit-pos-cache-limit 1
8ed43f15
AM
3130 c-state-semi-nonlit-pos-cache nil
3131 c-state-semi-nonlit-pos-cache-limit 1
0ec1d2c5
AM
3132 c-state-brace-pair-desert nil
3133 c-state-point-min 1
3134 c-state-point-min-lit-type nil
3135 c-state-point-min-lit-start nil
3136 c-state-min-scan-pos 1
3137 c-state-old-cpp-beg nil
3138 c-state-old-cpp-end nil)
3139 (c-state-mark-point-min-literal))
3140
e84efb70
AM
3141;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
3142;; Debugging routines to dump `c-state-cache' in a "replayable" form.
3143;; (defmacro c-sc-de (elt) ; "c-state-cache-dump-element"
3144;; `(format ,(concat "(setq " (symbol-name elt) " %s) ") ,elt))
3145;; (defmacro c-sc-qde (elt) ; "c-state-cache-quote-dump-element"
3146;; `(format ,(concat "(setq " (symbol-name elt) " '%s) ") ,elt))
3147;; (defun c-state-dump ()
3148;; ;; For debugging.
3149;; ;(message
3150;; (concat
3151;; (c-sc-qde c-state-cache)
3152;; (c-sc-de c-state-cache-good-pos)
3153;; (c-sc-qde c-state-nonlit-pos-cache)
3154;; (c-sc-de c-state-nonlit-pos-cache-limit)
3155;; (c-sc-qde c-state-brace-pair-desert)
3156;; (c-sc-de c-state-point-min)
3157;; (c-sc-de c-state-point-min-lit-type)
3158;; (c-sc-de c-state-point-min-lit-start)
3159;; (c-sc-de c-state-min-scan-pos)
3160;; (c-sc-de c-state-old-cpp-beg)
3161;; (c-sc-de c-state-old-cpp-end)))
3162;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
3163
0ec1d2c5
AM
3164(defun c-invalidate-state-cache-1 (here)
3165 ;; Invalidate all info on `c-state-cache' that applies to the buffer at HERE
3166 ;; or higher and set `c-state-cache-good-pos' accordingly. The cache is
3167 ;; left in a consistent state.
3168 ;;
3169 ;; This is much like `c-whack-state-after', but it never changes a paren
3170 ;; pair element into an open paren element. Doing that would mean that the
3171 ;; new open paren wouldn't have the required preceding paren pair element.
785eecbb 3172 ;;
0ec1d2c5
AM
3173 ;; This function is called from c-after-change.
3174
56d093a9 3175 ;; The caches of non-literals:
691e26ae
AM
3176 ;; Note that we use "<=" for the possibility of the second char of a two-char
3177 ;; comment opener being typed; this would invalidate any cache position at
3178 ;; HERE.
3179 (if (<= here c-state-nonlit-pos-cache-limit)
3180 (setq c-state-nonlit-pos-cache-limit (1- here)))
3181 (if (<= here c-state-semi-nonlit-pos-cache-limit)
3182 (setq c-state-semi-nonlit-pos-cache-limit (1- here)))
0ec1d2c5
AM
3183
3184 ;; `c-state-cache':
3185 ;; Case 1: if `here' is in a literal containing point-min, everything
3186 ;; becomes (or is already) nil.
3187 (if (or (null c-state-cache-good-pos)
3188 (< here (c-state-get-min-scan-pos)))
3189 (setq c-state-cache nil
3190 c-state-cache-good-pos nil
3191 c-state-min-scan-pos nil)
3192
c400c4d7
SM
3193 ;; Truncate `c-state-cache' and set `c-state-cache-good-pos' to a value
3194 ;; below `here'. To maintain its consistency, we may need to insert a new
3195 ;; brace pair.
e3af763f
AM
3196 (let (open-paren-in-column-0-is-defun-start
3197 (here-bol (c-point 'bol here))
0ec1d2c5
AM
3198 too-high-pa ; recorded {/(/[ next above here, or nil.
3199 dropped-cons ; was the last removed element a brace pair?
3200 pa)
3201 ;; The easy bit - knock over-the-top bits off `c-state-cache'.
3202 (while (and c-state-cache
3203 (>= (setq pa (c-state-cache-top-paren)) here))
3204 (setq dropped-cons (consp (car c-state-cache))
3205 too-high-pa (c-state-cache-top-lparen)
3206 c-state-cache (cdr c-state-cache)))
3207
3208 ;; Do we need to add in an earlier brace pair, having lopped one off?
3209 (if (and dropped-cons
3210 (< too-high-pa (+ here c-state-cache-too-far)))
d23b0804 3211 (c-append-lower-brace-pair-to-state-cache too-high-pa here here-bol))
0ec1d2c5
AM
3212 (setq c-state-cache-good-pos (or (c-state-cache-after-top-paren)
3213 (c-state-get-min-scan-pos)))))
3214
3215 ;; The brace-pair desert marker:
3216 (when (car c-state-brace-pair-desert)
3217 (if (< here (car c-state-brace-pair-desert))
3218 (setq c-state-brace-pair-desert nil)
3219 (if (< here (cdr c-state-brace-pair-desert))
3220 (setcdr c-state-brace-pair-desert here)))))
3221
3222(defun c-parse-state-1 ()
3223 ;; Find and record all noteworthy parens between some good point earlier in
3224 ;; the file and point. That good point is at least the beginning of the
3225 ;; top-level construct we are in, or the beginning of the preceding
3226 ;; top-level construct if we aren't in one.
3227 ;;
3228 ;; The returned value is a list of the noteworthy parens with the last one
3229 ;; first. If an element in the list is an integer, it's the position of an
3230 ;; open paren (of any type) which has not been closed before the point. If
3231 ;; an element is a cons, it gives the position of a closed BRACE paren
3232 ;; pair[*]; the car is the start brace position and the cdr is the position
3233 ;; following the closing brace. Only the last closed brace paren pair
3234 ;; before each open paren and before the point is recorded, and thus the
3235 ;; state never contains two cons elements in succession. When a close brace
3236 ;; has no matching open brace (e.g., the matching brace is outside the
3237 ;; visible region), it is not represented in the returned value.
3238 ;;
3239 ;; [*] N.B. The close "brace" might be a mismatching close bracket or paren.
3240 ;; This defun explicitly treats mismatching parens/braces/brackets as
3241 ;; matching. It is the open brace which makes it a "brace" pair.
3242 ;;
3243 ;; If POINT is within a macro, open parens and brace pairs within
3244 ;; THIS macro MIGHT be recorded. This depends on whether their
3245 ;; syntactic properties have been suppressed by
3246 ;; `c-neutralize-syntax-in-CPP'. This might need fixing (2008-12-11).
d9e94c22
MS
3247 ;;
3248 ;; Currently no characters which are given paren syntax with the
3249 ;; syntax-table property are recorded, i.e. angle bracket arglist
3250 ;; parens are never present here. Note that this might change.
3251 ;;
0386b551 3252 ;; BUG: This function doesn't cope entirely well with unbalanced
0ec1d2c5
AM
3253 ;; parens in macros. (2008-12-11: this has probably been resolved
3254 ;; by the function `c-neutralize-syntax-in-CPP'.) E.g. in the
3255 ;; following case the brace before the macro isn't balanced with the
3256 ;; one after it:
0386b551
AM
3257 ;;
3258 ;; {
3259 ;; #define X {
3260 ;; }
3261 ;;
0ec1d2c5
AM
3262 ;; Note to maintainers: this function DOES get called with point
3263 ;; within comments and strings, so don't assume it doesn't!
3264 ;;
0386b551 3265 ;; This function might do hidden buffer changes.
0ec1d2c5
AM
3266 (let* ((here (point))
3267 (here-bopl (c-point 'bopl))
e3af763f 3268 open-paren-in-column-0-is-defun-start
0ec1d2c5
AM
3269 strategy ; 'forward, 'backward etc..
3270 ;; Candidate positions to start scanning from:
3271 cache-pos ; highest position below HERE already existing in
3272 ; cache (or 1).
3273 good-pos
267eacb5
AM
3274 start-point ; (when scanning forward) a place below HERE where there
3275 ; are no open parens/braces between it and HERE.
0ec1d2c5
AM
3276 bopl-state
3277 res
99f7b0a6 3278 cons-separated
0ec1d2c5
AM
3279 scan-backward-pos scan-forward-p) ; used for 'backward.
3280 ;; If POINT-MIN has changed, adjust the cache
3281 (unless (= (point-min) c-state-point-min)
3282 (c-renarrow-state-cache))
3283
3284 ;; Strategy?
3285 (setq res (c-parse-state-get-strategy here c-state-cache-good-pos)
3286 strategy (car res)
267eacb5 3287 start-point (cadr res))
0ec1d2c5 3288
0ec1d2c5 3289 ;; SCAN!
d23b0804 3290 (cond
99f7b0a6 3291 ((memq strategy '(forward back-and-forward))
d23b0804
AM
3292 (setq res (c-remove-stale-state-cache start-point here here-bopl))
3293 (setq cache-pos (car res)
3294 scan-backward-pos (cadr res)
99f7b0a6
AM
3295 cons-separated (car (cddr res))
3296 bopl-state (cadr (cddr res))) ; will be nil if (< here-bopl
0ec1d2c5 3297 ; start-point)
99f7b0a6
AM
3298 (if (and scan-backward-pos
3299 (or cons-separated (eq strategy 'forward))) ;scan-backward-pos
d23b0804
AM
3300 (c-append-lower-brace-pair-to-state-cache scan-backward-pos here))
3301 (setq good-pos
3302 (c-append-to-state-cache cache-pos here))
3303 (setq c-state-cache-good-pos
3304 (if (and bopl-state
3305 (< good-pos (- here c-state-cache-too-far)))
3306 (c-state-cache-non-literal-place here-bopl bopl-state)
3307 good-pos)))
3308
3309 ((eq strategy 'backward)
3310 (setq res (c-remove-stale-state-cache-backwards here)
3311 good-pos (car res)
3312 scan-backward-pos (cadr res)
3313 scan-forward-p (car (cddr res)))
3314 (if scan-backward-pos
3315 (c-append-lower-brace-pair-to-state-cache scan-backward-pos here))
3316 (setq c-state-cache-good-pos
3317 (if scan-forward-p
3318 (c-append-to-state-cache good-pos here)
3319 good-pos)))
3320
3321 (t ; (eq strategy 'IN-LIT)
3322 (setq c-state-cache nil
3323 c-state-cache-good-pos nil))))
0ec1d2c5
AM
3324
3325 c-state-cache)
3326
3327(defun c-invalidate-state-cache (here)
3328 ;; This is a wrapper over `c-invalidate-state-cache-1'.
3329 ;;
3330 ;; It suppresses the syntactic effect of the < and > (template) brackets and
3331 ;; of all parens in preprocessor constructs, except for any such construct
3332 ;; containing point. We can then call `c-invalidate-state-cache-1' without
3333 ;; worrying further about macros and template delimiters.
3334 (c-with-<->-as-parens-suppressed
3e8f7d91
AM
3335 (if (and c-state-old-cpp-beg
3336 (< c-state-old-cpp-beg here))
0ec1d2c5 3337 (c-with-all-but-one-cpps-commented-out
3e8f7d91
AM
3338 c-state-old-cpp-beg
3339 (min c-state-old-cpp-end here)
0ec1d2c5
AM
3340 (c-invalidate-state-cache-1 here))
3341 (c-with-cpps-commented-out
3342 (c-invalidate-state-cache-1 here)))))
3343
8cff5feb
AM
3344(defmacro c-state-maybe-marker (place marker)
3345 ;; If PLACE is non-nil, return a marker marking it, otherwise nil.
3346 ;; We (re)use MARKER.
3347 `(and ,place
3348 (or ,marker (setq ,marker (make-marker)))
3349 (set-marker ,marker ,place)))
3350
0ec1d2c5
AM
3351(defun c-parse-state ()
3352 ;; This is a wrapper over `c-parse-state-1'. See that function for a
3353 ;; description of the functionality and return value.
3354 ;;
3355 ;; It suppresses the syntactic effect of the < and > (template) brackets and
3356 ;; of all parens in preprocessor constructs, except for any such construct
3357 ;; containing point. We can then call `c-parse-state-1' without worrying
3358 ;; further about macros and template delimiters.
3359 (let (here-cpp-beg here-cpp-end)
3360 (save-excursion
3361 (when (c-beginning-of-macro)
3362 (setq here-cpp-beg (point))
3363 (unless
3364 (> (setq here-cpp-end (c-syntactic-end-of-macro))
3365 here-cpp-beg)
3366 (setq here-cpp-beg nil here-cpp-end nil))))
3367 ;; FIXME!!! Put in a `condition-case' here to protect the integrity of the
3368 ;; subsystem.
3369 (prog1
3370 (c-with-<->-as-parens-suppressed
3371 (if (and here-cpp-beg (> here-cpp-end here-cpp-beg))
3372 (c-with-all-but-one-cpps-commented-out
3373 here-cpp-beg here-cpp-end
3374 (c-parse-state-1))
3375 (c-with-cpps-commented-out
3376 (c-parse-state-1))))
8cff5feb
AM
3377 (setq c-state-old-cpp-beg
3378 (c-state-maybe-marker here-cpp-beg c-state-old-cpp-beg-marker)
3379 c-state-old-cpp-end
3380 (c-state-maybe-marker here-cpp-end c-state-old-cpp-end-marker)))))
0ec1d2c5
AM
3381
3382;; Debug tool to catch cache inconsistencies. This is called from
3383;; 000tests.el.
a66cd3ee
MS
3384(defvar c-debug-parse-state nil)
3385(unless (fboundp 'c-real-parse-state)
3386 (fset 'c-real-parse-state (symbol-function 'c-parse-state)))
3387(cc-bytecomp-defun c-real-parse-state)
cd5495ff 3388
b03aabda 3389(defvar c-parse-state-point nil)
cd5495ff
AM
3390(defvar c-parse-state-state nil)
3391(defun c-record-parse-state-state ()
b03aabda 3392 (setq c-parse-state-point (point))
cd5495ff
AM
3393 (setq c-parse-state-state
3394 (mapcar
3395 (lambda (arg)
b03aabda
AM
3396 (let ((val (symbol-value arg)))
3397 (cons arg
3398 (if (consp val)
3399 (copy-tree val)
3400 val))))
cd5495ff
AM
3401 '(c-state-cache
3402 c-state-cache-good-pos
3403 c-state-nonlit-pos-cache
3404 c-state-nonlit-pos-cache-limit
8ed43f15
AM
3405 c-state-semi-nonlit-pos-cache
3406 c-state-semi-nonlit-pos-cache-limit
cd5495ff
AM
3407 c-state-brace-pair-desert
3408 c-state-point-min
3409 c-state-point-min-lit-type
3410 c-state-point-min-lit-start
3411 c-state-min-scan-pos
3412 c-state-old-cpp-beg
b03aabda
AM
3413 c-state-old-cpp-end
3414 c-parse-state-point))))
cd5495ff
AM
3415(defun c-replay-parse-state-state ()
3416 (message
3417 (concat "(setq "
3418 (mapconcat
3419 (lambda (arg)
3420 (format "%s %s%s" (car arg) (if (atom (cdr arg)) "" "'") (cdr arg)))
3421 c-parse-state-state " ")
3422 ")")))
3423
b03aabda
AM
3424(defun c-debug-parse-state-double-cons (state)
3425 (let (state-car conses-not-ok)
3426 (while state
3427 (setq state-car (car state)
3428 state (cdr state))
3429 (if (and (consp state-car)
3430 (consp (car state)))
3431 (setq conses-not-ok t)))
3432 conses-not-ok))
3433
a66cd3ee 3434(defun c-debug-parse-state ()
0ec1d2c5 3435 (let ((here (point)) (res1 (c-real-parse-state)) res2)
0386b551 3436 (let ((c-state-cache nil)
0ec1d2c5
AM
3437 (c-state-cache-good-pos 1)
3438 (c-state-nonlit-pos-cache nil)
3439 (c-state-nonlit-pos-cache-limit 1)
3440 (c-state-brace-pair-desert nil)
3441 (c-state-point-min 1)
3442 (c-state-point-min-lit-type nil)
3443 (c-state-point-min-lit-start nil)
3444 (c-state-min-scan-pos 1)
3445 (c-state-old-cpp-beg nil)
3446 (c-state-old-cpp-end nil))
a66cd3ee
MS
3447 (setq res2 (c-real-parse-state)))
3448 (unless (equal res1 res2)
0386b551
AM
3449 ;; The cache can actually go further back due to the ad-hoc way
3450 ;; the first paren is found, so try to whack off a bit of its
3451 ;; start before complaining.
cd5495ff
AM
3452 ;; (save-excursion
3453 ;; (goto-char (or (c-least-enclosing-brace res2) (point)))
3454 ;; (c-beginning-of-defun-1)
3455 ;; (while (not (or (bobp) (eq (char-after) ?{)))
3456 ;; (c-beginning-of-defun-1))
3457 ;; (unless (equal (c-whack-state-before (point) res1) res2)
3458 ;; (message (concat "c-parse-state inconsistency at %s: "
3459 ;; "using cache: %s, from scratch: %s")
3460 ;; here res1 res2)))
3461 (message (concat "c-parse-state inconsistency at %s: "
3462 "using cache: %s, from scratch: %s")
3463 here res1 res2)
3464 (message "Old state:")
3465 (c-replay-parse-state-state))
b03aabda
AM
3466
3467 (when (c-debug-parse-state-double-cons res1)
3468 (message "c-parse-state INVALIDITY at %s: %s"
3469 here res1)
3470 (message "Old state:")
3471 (c-replay-parse-state-state))
3472
cd5495ff 3473 (c-record-parse-state-state)
b03aabda
AM
3474 res2 ; res1 correct a cascading series of errors ASAP
3475 ))
0ec1d2c5 3476
a66cd3ee
MS
3477(defun c-toggle-parse-state-debug (&optional arg)
3478 (interactive "P")
3479 (setq c-debug-parse-state (c-calculate-state arg c-debug-parse-state))
3480 (fset 'c-parse-state (symbol-function (if c-debug-parse-state
3481 'c-debug-parse-state
3482 'c-real-parse-state)))
b03aabda
AM
3483 (c-keep-region-active)
3484 (message "c-debug-parse-state %sabled"
3485 (if c-debug-parse-state "en" "dis")))
0386b551
AM
3486(when c-debug-parse-state
3487 (c-toggle-parse-state-debug 1))
a66cd3ee 3488
0ec1d2c5 3489\f
d9e94c22
MS
3490(defun c-whack-state-before (bufpos paren-state)
3491 ;; Whack off any state information from PAREN-STATE which lies
3492 ;; before BUFPOS. Not destructive on PAREN-STATE.
d9e94c22
MS
3493 (let* ((newstate (list nil))
3494 (ptr newstate)
3495 car)
3496 (while paren-state
3497 (setq car (car paren-state)
3498 paren-state (cdr paren-state))
3499 (if (< (if (consp car) (car car) car) bufpos)
3500 (setq paren-state nil)
3501 (setcdr ptr (list car))
3502 (setq ptr (cdr ptr))))
3503 (cdr newstate)))
3504
3505(defun c-whack-state-after (bufpos paren-state)
3506 ;; Whack off any state information from PAREN-STATE which lies at or
3507 ;; after BUFPOS. Not destructive on PAREN-STATE.
d9e94c22
MS
3508 (catch 'done
3509 (while paren-state
3510 (let ((car (car paren-state)))
3511 (if (consp car)
3512 ;; just check the car, because in a balanced brace
3513 ;; expression, it must be impossible for the corresponding
3514 ;; close brace to be before point, but the open brace to
3515 ;; be after.
3516 (if (<= bufpos (car car))
3517 nil ; whack it off
3518 (if (< bufpos (cdr car))
3519 ;; its possible that the open brace is before
3520 ;; bufpos, but the close brace is after. In that
3521 ;; case, convert this to a non-cons element. The
3522 ;; rest of the state is before bufpos, so we're
3523 ;; done.
3524 (throw 'done (cons (car car) (cdr paren-state)))
3525 ;; we know that both the open and close braces are
3526 ;; before bufpos, so we also know that everything else
3527 ;; on state is before bufpos.
3528 (throw 'done paren-state)))
3529 (if (<= bufpos car)
3530 nil ; whack it off
3531 ;; it's before bufpos, so everything else should too.
3532 (throw 'done paren-state)))
3533 (setq paren-state (cdr paren-state)))
3534 nil)))
3535
3536(defun c-most-enclosing-brace (paren-state &optional bufpos)
3537 ;; Return the bufpos of the innermost enclosing open paren before
0386b551 3538 ;; bufpos, or nil if none was found.
d9e94c22
MS
3539 (let (enclosingp)
3540 (or bufpos (setq bufpos 134217727))
3541 (while paren-state
3542 (setq enclosingp (car paren-state)
3543 paren-state (cdr paren-state))
3544 (if (or (consp enclosingp)
3545 (>= enclosingp bufpos))
3546 (setq enclosingp nil)
d9e94c22
MS
3547 (setq paren-state nil)))
3548 enclosingp))
3549
0386b551
AM
3550(defun c-least-enclosing-brace (paren-state)
3551 ;; Return the bufpos of the outermost enclosing open paren, or nil
3552 ;; if none was found.
d9e94c22 3553 (let (pos elem)
d9e94c22
MS
3554 (while paren-state
3555 (setq elem (car paren-state)
3556 paren-state (cdr paren-state))
0386b551
AM
3557 (if (integerp elem)
3558 (setq pos elem)))
d9e94c22
MS
3559 pos))
3560
3561(defun c-safe-position (bufpos paren-state)
0386b551
AM
3562 ;; Return the closest "safe" position recorded on PAREN-STATE that
3563 ;; is higher up than BUFPOS. Return nil if PAREN-STATE doesn't
3564 ;; contain any. Return nil if BUFPOS is nil, which is useful to
3565 ;; find the closest limit before a given limit that might be nil.
d9e94c22 3566 ;;
0386b551
AM
3567 ;; A "safe" position is a position at or after a recorded open
3568 ;; paren, or after a recorded close paren. The returned position is
3569 ;; thus either the first position after a close brace, or the first
3570 ;; position after an enclosing paren, or at the enclosing paren in
3571 ;; case BUFPOS is immediately after it.
d9e94c22
MS
3572 (when bufpos
3573 (let (elem)
3574 (catch 'done
3575 (while paren-state
3576 (setq elem (car paren-state))
3577 (if (consp elem)
3578 (cond ((< (cdr elem) bufpos)
3579 (throw 'done (cdr elem)))
3580 ((< (car elem) bufpos)
3581 ;; See below.
3582 (throw 'done (min (1+ (car elem)) bufpos))))
3583 (if (< elem bufpos)
3584 ;; elem is the position at and not after the opening paren, so
3585 ;; we can go forward one more step unless it's equal to
3586 ;; bufpos. This is useful in some cases avoid an extra paren
3587 ;; level between the safe position and bufpos.
3588 (throw 'done (min (1+ elem) bufpos))))
3589 (setq paren-state (cdr paren-state)))))))
3590
3591(defun c-beginning-of-syntax ()
3592 ;; This is used for `font-lock-beginning-of-syntax-function'. It
3593 ;; goes to the closest previous point that is known to be outside
3594 ;; any string literal or comment. `c-state-cache' is used if it has
3595 ;; a position in the vicinity.
3596 (let* ((paren-state c-state-cache)
3597 elem
3598
3599 (pos (catch 'done
3600 ;; Note: Similar code in `c-safe-position'. The
3601 ;; difference is that we accept a safe position at
3602 ;; the point and don't bother to go forward past open
3603 ;; parens.
3604 (while paren-state
3605 (setq elem (car paren-state))
3606 (if (consp elem)
3607 (cond ((<= (cdr elem) (point))
3608 (throw 'done (cdr elem)))
3609 ((<= (car elem) (point))
3610 (throw 'done (car elem))))
3611 (if (<= elem (point))
3612 (throw 'done elem)))
3613 (setq paren-state (cdr paren-state)))
3614 (point-min))))
3615
3616 (if (> pos (- (point) 4000))
3617 (goto-char pos)
3618 ;; The position is far back. Try `c-beginning-of-defun-1'
3619 ;; (although we can't be entirely sure it will go to a position
3620 ;; outside a comment or string in current emacsen). FIXME:
3621 ;; Consult `syntax-ppss' here.
3622 (c-beginning-of-defun-1)
3623 (if (< (point) pos)
3624 (goto-char pos)))))
3625
3626\f
3627;; Tools for scanning identifiers and other tokens.
3628
3629(defun c-on-identifier ()
3630 "Return non-nil if the point is on or directly after an identifier.
3631Keywords are recognized and not considered identifiers. If an
3632identifier is detected, the returned value is its starting position.
0386b551
AM
3633If an identifier ends at the point and another begins at it \(can only
3634happen in Pike) then the point for the preceding one is returned.
d9e94c22 3635
0386b551
AM
3636Note that this function might do hidden buffer changes. See the
3637comment at the start of cc-engine.el for more info."
3638
3639 ;; FIXME: Shouldn't this function handle "operator" in C++?
d9e94c22
MS
3640
3641 (save-excursion
0386b551
AM
3642 (skip-syntax-backward "w_")
3643
3644 (or
3645
3646 ;; Check for a normal (non-keyword) identifier.
3647 (and (looking-at c-symbol-start)
3648 (not (looking-at c-keywords-regexp))
3649 (point))
3650
3651 (when (c-major-mode-is 'pike-mode)
3652 ;; Handle the `<operator> syntax in Pike.
3653 (let ((pos (point)))
3654 (skip-chars-backward "-!%&*+/<=>^|~[]()")
3655 (and (if (< (skip-chars-backward "`") 0)
3656 t
3657 (goto-char pos)
3658 (eq (char-after) ?\`))
3659 (looking-at c-symbol-key)
3660 (>= (match-end 0) pos)
3661 (point))))
3662
3663 ;; Handle the "operator +" syntax in C++.
3664 (when (and c-overloadable-operators-regexp
3665 (= (c-backward-token-2 0) 0))
3666
3667 (cond ((and (looking-at c-overloadable-operators-regexp)
51c9af45 3668 (or (not c-opt-op-identifier-prefix)
0386b551 3669 (and (= (c-backward-token-2 1) 0)
51c9af45 3670 (looking-at c-opt-op-identifier-prefix))))
0386b551
AM
3671 (point))
3672
3673 ((save-excursion
51c9af45
AM
3674 (and c-opt-op-identifier-prefix
3675 (looking-at c-opt-op-identifier-prefix)
0386b551
AM
3676 (= (c-forward-token-2 1) 0)
3677 (looking-at c-overloadable-operators-regexp)))
3678 (point))))
3679
3680 )))
d9e94c22
MS
3681
3682(defsubst c-simple-skip-symbol-backward ()
3683 ;; If the point is at the end of a symbol then skip backward to the
3684 ;; beginning of it. Don't move otherwise. Return non-nil if point
3685 ;; moved.
0386b551
AM
3686 ;;
3687 ;; This function might do hidden buffer changes.
d9e94c22
MS
3688 (or (< (skip-syntax-backward "w_") 0)
3689 (and (c-major-mode-is 'pike-mode)
3690 ;; Handle the `<operator> syntax in Pike.
3691 (let ((pos (point)))
2a15eb73 3692 (if (and (< (skip-chars-backward "-!%&*+/<=>^|~[]()") 0)
d9e94c22
MS
3693 (< (skip-chars-backward "`") 0)
3694 (looking-at c-symbol-key)
3695 (>= (match-end 0) pos))
3696 t
3697 (goto-char pos)
3698 nil)))))
3699
0386b551 3700(defun c-beginning-of-current-token (&optional back-limit)
d9e94c22
MS
3701 ;; Move to the beginning of the current token. Do not move if not
3702 ;; in the middle of one. BACK-LIMIT may be used to bound the
3703 ;; backward search; if given it's assumed to be at the boundary
a85fd6da 3704 ;; between two tokens. Return non-nil if the point is moved, nil
580fba94 3705 ;; otherwise.
0386b551
AM
3706 ;;
3707 ;; This function might do hidden buffer changes.
d9e94c22 3708 (let ((start (point)))
580fba94
AM
3709 (if (looking-at "\\w\\|\\s_")
3710 (skip-syntax-backward "w_" back-limit)
3711 (when (< (skip-syntax-backward ".()" back-limit) 0)
3712 (while (let ((pos (or (and (looking-at c-nonsymbol-token-regexp)
3713 (match-end 0))
3714 ;; `c-nonsymbol-token-regexp' should always match
c91c771d 3715 ;; since we've skipped backward over punctuation
580fba94
AM
3716 ;; or paren syntax, but consume one char in case
3717 ;; it doesn't so that we don't leave point before
3718 ;; some earlier incorrect token.
3719 (1+ (point)))))
3720 (if (<= pos start)
3721 (goto-char pos))))))
3722 (< (point) start)))
d9e94c22 3723
ff959bab 3724(defun c-end-of-current-token (&optional back-limit)
d9e94c22
MS
3725 ;; Move to the end of the current token. Do not move if not in the
3726 ;; middle of one. BACK-LIMIT may be used to bound the backward
3727 ;; search; if given it's assumed to be at the boundary between two
ff959bab 3728 ;; tokens. Return non-nil if the point is moved, nil otherwise.
0386b551
AM
3729 ;;
3730 ;; This function might do hidden buffer changes.
d9e94c22
MS
3731 (let ((start (point)))
3732 (cond ((< (skip-syntax-backward "w_" (1- start)) 0)
3733 (skip-syntax-forward "w_"))
3734 ((< (skip-syntax-backward ".()" back-limit) 0)
3735 (while (progn
3736 (if (looking-at c-nonsymbol-token-regexp)
3737 (goto-char (match-end 0))
3738 ;; `c-nonsymbol-token-regexp' should always match since
c91c771d 3739 ;; we've skipped backward over punctuation or paren
d9e94c22
MS
3740 ;; syntax, but move forward in case it doesn't so that
3741 ;; we don't leave point earlier than we started with.
3742 (forward-char))
ff959bab
MS
3743 (< (point) start)))))
3744 (> (point) start)))
d9e94c22
MS
3745
3746(defconst c-jump-syntax-balanced
3747 (if (memq 'gen-string-delim c-emacs-features)
3748 "\\w\\|\\s_\\|\\s\(\\|\\s\)\\|\\s\"\\|\\s|"
3749 "\\w\\|\\s_\\|\\s\(\\|\\s\)\\|\\s\""))
3750
3751(defconst c-jump-syntax-unbalanced
3752 (if (memq 'gen-string-delim c-emacs-features)
3753 "\\w\\|\\s_\\|\\s\"\\|\\s|"
3754 "\\w\\|\\s_\\|\\s\""))
3755
3756(defun c-forward-token-2 (&optional count balanced limit)
3757 "Move forward by tokens.
3758A token is defined as all symbols and identifiers which aren't
3759syntactic whitespace \(note that multicharacter tokens like \"==\" are
3760treated properly). Point is always either left at the beginning of a
3761token or not moved at all. COUNT specifies the number of tokens to
3762move; a negative COUNT moves in the opposite direction. A COUNT of 0
3763moves to the next token beginning only if not already at one. If
3764BALANCED is true, move over balanced parens, otherwise move into them.
3765Also, if BALANCED is true, never move out of an enclosing paren.
3766
3767LIMIT sets the limit for the movement and defaults to the point limit.
3768The case when LIMIT is set in the middle of a token, comment or macro
3769is handled correctly, i.e. the point won't be left there.
3770
3771Return the number of tokens left to move \(positive or negative). If
3772BALANCED is true, a move over a balanced paren counts as one. Note
3773that if COUNT is 0 and no appropriate token beginning is found, 1 will
3774be returned. Thus, a return value of 0 guarantees that point is at
3775the requested position and a return value less \(without signs) than
0386b551
AM
3776COUNT guarantees that point is at the beginning of some token.
3777
3778Note that this function might do hidden buffer changes. See the
3779comment at the start of cc-engine.el for more info."
d9e94c22
MS
3780
3781 (or count (setq count 1))
3782 (if (< count 0)
3783 (- (c-backward-token-2 (- count) balanced limit))
3784
3785 (let ((jump-syntax (if balanced
3786 c-jump-syntax-balanced
3787 c-jump-syntax-unbalanced))
3788 (last (point))
3789 (prev (point)))
3790
3791 (if (zerop count)
3792 ;; If count is zero we should jump if in the middle of a token.
3793 (c-end-of-current-token))
3794
3795 (save-restriction
3796 (if limit (narrow-to-region (point-min) limit))
3797 (if (/= (point)
3798 (progn (c-forward-syntactic-ws) (point)))
3799 ;; Skip whitespace. Count this as a move if we did in
3800 ;; fact move.
3801 (setq count (max (1- count) 0)))
3802
3803 (if (eobp)
3804 ;; Moved out of bounds. Make sure the returned count isn't zero.
3805 (progn
3806 (if (zerop count) (setq count 1))
3807 (goto-char last))
3808
3809 ;; Use `condition-case' to avoid having the limit tests
3810 ;; inside the loop.
3811 (condition-case nil
3812 (while (and
3813 (> count 0)
3814 (progn
3815 (setq last (point))
3816 (cond ((looking-at jump-syntax)
3817 (goto-char (scan-sexps (point) 1))
3818 t)
3819 ((looking-at c-nonsymbol-token-regexp)
3820 (goto-char (match-end 0))
3821 t)
3822 ;; `c-nonsymbol-token-regexp' above should always
3823 ;; match if there are correct tokens. Try to
3824 ;; widen to see if the limit was set in the
3825 ;; middle of one, else fall back to treating
3826 ;; the offending thing as a one character token.
3827 ((and limit
3828 (save-restriction
3829 (widen)
3830 (looking-at c-nonsymbol-token-regexp)))
3831 nil)
3832 (t
3833 (forward-char)
3834 t))))
3835 (c-forward-syntactic-ws)
3836 (setq prev last
3837 count (1- count)))
3838 (error (goto-char last)))
3839
3840 (when (eobp)
3841 (goto-char prev)
3842 (setq count (1+ count)))))
3843
3844 count)))
3845
3846(defun c-backward-token-2 (&optional count balanced limit)
3847 "Move backward by tokens.
3848See `c-forward-token-2' for details."
3849
3850 (or count (setq count 1))
3851 (if (< count 0)
3852 (- (c-forward-token-2 (- count) balanced limit))
3853
3854 (or limit (setq limit (point-min)))
3855 (let ((jump-syntax (if balanced
3856 c-jump-syntax-balanced
3857 c-jump-syntax-unbalanced))
3858 (last (point)))
3859
3860 (if (zerop count)
3861 ;; The count is zero so try to skip to the beginning of the
3862 ;; current token.
3863 (if (> (point)
3864 (progn (c-beginning-of-current-token) (point)))
3865 (if (< (point) limit)
3866 ;; The limit is inside the same token, so return 1.
3867 (setq count 1))
3868
3869 ;; We're not in the middle of a token. If there's
3870 ;; whitespace after the point then we must move backward,
3871 ;; so set count to 1 in that case.
3872 (and (looking-at c-syntactic-ws-start)
3873 ;; If we're looking at a '#' that might start a cpp
3874 ;; directive then we have to do a more elaborate check.
3875 (or (/= (char-after) ?#)
3876 (not c-opt-cpp-prefix)
3877 (save-excursion
3878 (and (= (point)
3879 (progn (beginning-of-line)
3880 (looking-at "[ \t]*")
3881 (match-end 0)))
3882 (or (bobp)
3883 (progn (backward-char)
3884 (not (eq (char-before) ?\\)))))))
3885 (setq count 1))))
3886
3887 ;; Use `condition-case' to avoid having to check for buffer
3888 ;; limits in `backward-char', `scan-sexps' and `goto-char' below.
3889 (condition-case nil
3890 (while (and
3891 (> count 0)
3892 (progn
3893 (c-backward-syntactic-ws)
3894 (backward-char)
3895 (if (looking-at jump-syntax)
3896 (goto-char (scan-sexps (1+ (point)) -1))
3897 ;; This can be very inefficient if there's a long
3898 ;; sequence of operator tokens without any separation.
3899 ;; That doesn't happen in practice, anyway.
3900 (c-beginning-of-current-token))
3901 (>= (point) limit)))
3902 (setq last (point)
3903 count (1- count)))
3904 (error (goto-char last)))
3905
3906 (if (< (point) limit)
3907 (goto-char last))
3908
3909 count)))
3910
3911(defun c-forward-token-1 (&optional count balanced limit)
3912 "Like `c-forward-token-2' but doesn't treat multicharacter operator
3913tokens like \"==\" as single tokens, i.e. all sequences of symbol
3914characters are jumped over character by character. This function is
3915for compatibility only; it's only a wrapper over `c-forward-token-2'."
3916 (let ((c-nonsymbol-token-regexp "\\s.\\|\\s\(\\|\\s\)"))
3917 (c-forward-token-2 count balanced limit)))
3918
3919(defun c-backward-token-1 (&optional count balanced limit)
3920 "Like `c-backward-token-2' but doesn't treat multicharacter operator
3921tokens like \"==\" as single tokens, i.e. all sequences of symbol
3922characters are jumped over character by character. This function is
3923for compatibility only; it's only a wrapper over `c-backward-token-2'."
3924 (let ((c-nonsymbol-token-regexp "\\s.\\|\\s\(\\|\\s\)"))
3925 (c-backward-token-2 count balanced limit)))
3926
3927\f
3928;; Tools for doing searches restricted to syntactically relevant text.
3929
3930(defun c-syntactic-re-search-forward (regexp &optional bound noerror
3931 paren-level not-inside-token
3932 lookbehind-submatch)
3933 "Like `re-search-forward', but only report matches that are found
3934in syntactically significant text. I.e. matches in comments, macros
3935or string literals are ignored. The start point is assumed to be
3936outside any comment, macro or string literal, or else the content of
3937that region is taken as syntactically significant text.
3938
3939If PAREN-LEVEL is non-nil, an additional restriction is added to
2a15eb73
MS
3940ignore matches in nested paren sexps. The search will also not go
3941outside the current list sexp, which has the effect that if the point
3942should be moved to BOUND when no match is found \(i.e. NOERROR is
3943neither nil nor t), then it will be at the closing paren if the end of
3944the current list sexp is encountered first.
d9e94c22
MS
3945
3946If NOT-INSIDE-TOKEN is non-nil, matches in the middle of tokens are
3947ignored. Things like multicharacter operators and special symbols
3948\(e.g. \"`()\" in Pike) are handled but currently not floating point
3949constants.
3950
3951If LOOKBEHIND-SUBMATCH is non-nil, it's taken as a number of a
3952subexpression in REGEXP. The end of that submatch is used as the
3953position to check for syntactic significance. If LOOKBEHIND-SUBMATCH
3954isn't used or if that subexpression didn't match then the start
3955position of the whole match is used instead. The \"look behind\"
3956subexpression is never tested before the starting position, so it
3957might be a good idea to include \\=\\= as a match alternative in it.
3958
3959Optimization note: Matches might be missed if the \"look behind\"
2a15eb73 3960subexpression can match the end of nonwhite syntactic whitespace,
d9e94c22 3961i.e. the end of comments or cpp directives. This since the function
2a15eb73
MS
3962skips over such things before resuming the search. It's on the other
3963hand not safe to assume that the \"look behind\" subexpression never
3964matches syntactic whitespace.
3965
3966Bug: Unbalanced parens inside cpp directives are currently not handled
3967correctly \(i.e. they don't get ignored as they should) when
0386b551
AM
3968PAREN-LEVEL is set.
3969
3970Note that this function might do hidden buffer changes. See the
3971comment at the start of cc-engine.el for more info."
d9e94c22
MS
3972
3973 (or bound (setq bound (point-max)))
3974 (if paren-level (setq paren-level -1))
3975
3976 ;;(message "c-syntactic-re-search-forward %s %s %S" (point) bound regexp)
3977
3978 (let ((start (point))
2a15eb73
MS
3979 tmp
3980 ;; Start position for the last search.
3981 search-pos
3982 ;; The `parse-partial-sexp' state between the start position
3983 ;; and the point.
3984 state
3985 ;; The current position after the last state update. The next
3986 ;; `parse-partial-sexp' continues from here.
3987 (state-pos (point))
3988 ;; The position at which to check the state and the state
3989 ;; there. This is separate from `state-pos' since we might
3990 ;; need to back up before doing the next search round.
3991 check-pos check-state
3992 ;; Last position known to end a token.
d9e94c22 3993 (last-token-end-pos (point-min))
2a15eb73
MS
3994 ;; Set when a valid match is found.
3995 found)
d9e94c22
MS
3996
3997 (condition-case err
3998 (while
3999 (and
2a15eb73
MS
4000 (progn
4001 (setq search-pos (point))
4002 (re-search-forward regexp bound noerror))
d9e94c22
MS
4003
4004 (progn
2a15eb73
MS
4005 (setq state (parse-partial-sexp
4006 state-pos (match-beginning 0) paren-level nil state)
4007 state-pos (point))
d9e94c22 4008 (if (setq check-pos (and lookbehind-submatch
2a15eb73
MS
4009 (or (not paren-level)
4010 (>= (car state) 0))
d9e94c22
MS
4011 (match-end lookbehind-submatch)))
4012 (setq check-state (parse-partial-sexp
2a15eb73
MS
4013 state-pos check-pos paren-level nil state))
4014 (setq check-pos state-pos
d9e94c22
MS
4015 check-state state))
4016
2a15eb73
MS
4017 ;; NOTE: If we got a look behind subexpression and get
4018 ;; an insignificant match in something that isn't
d9e94c22
MS
4019 ;; syntactic whitespace (i.e. strings or in nested
4020 ;; parentheses), then we can never skip more than a
2a15eb73
MS
4021 ;; single character from the match start position
4022 ;; (i.e. `state-pos' here) before continuing the
4023 ;; search. That since the look behind subexpression
4024 ;; might match the end of the insignificant region in
4025 ;; the next search.
d9e94c22
MS
4026
4027 (cond
d9e94c22
MS
4028 ((elt check-state 7)
4029 ;; Match inside a line comment. Skip to eol. Use
4030 ;; `re-search-forward' instead of `skip-chars-forward' to get
4031 ;; the right bound behavior.
4032 (re-search-forward "[\n\r]" bound noerror))
4033
4034 ((elt check-state 4)
4035 ;; Match inside a block comment. Skip to the '*/'.
4036 (search-forward "*/" bound noerror))
4037
4038 ((and (not (elt check-state 5))
4039 (eq (char-before check-pos) ?/)
2a15eb73 4040 (not (c-get-char-property (1- check-pos) 'syntax-table))
d9e94c22
MS
4041 (memq (char-after check-pos) '(?/ ?*)))
4042 ;; Match in the middle of the opener of a block or line
4043 ;; comment.
4044 (if (= (char-after check-pos) ?/)
4045 (re-search-forward "[\n\r]" bound noerror)
4046 (search-forward "*/" bound noerror)))
4047
2a15eb73
MS
4048 ;; The last `parse-partial-sexp' above might have
4049 ;; stopped short of the real check position if the end
4050 ;; of the current sexp was encountered in paren-level
4051 ;; mode. The checks above are always false in that
4052 ;; case, and since they can do better skipping in
4053 ;; lookbehind-submatch mode, we do them before
4054 ;; checking the paren level.
4055
4056 ((and paren-level
4057 (/= (setq tmp (car check-state)) 0))
4058 ;; Check the paren level first since we're short of the
4059 ;; syntactic checking position if the end of the
4060 ;; current sexp was encountered by `parse-partial-sexp'.
4061 (if (> tmp 0)
4062
4063 ;; Inside a nested paren sexp.
4064 (if lookbehind-submatch
4065 ;; See the NOTE above.
4066 (progn (goto-char state-pos) t)
4067 ;; Skip out of the paren quickly.
4068 (setq state (parse-partial-sexp state-pos bound 0 nil state)
4069 state-pos (point)))
4070
4071 ;; Have exited the current paren sexp.
4072 (if noerror
4073 (progn
4074 ;; The last `parse-partial-sexp' call above
4075 ;; has left us just after the closing paren
4076 ;; in this case, so we can modify the bound
4077 ;; to leave the point at the right position
4078 ;; upon return.
4079 (setq bound (1- (point)))
4080 nil)
4081 (signal 'search-failed (list regexp)))))
4082
4083 ((setq tmp (elt check-state 3))
4084 ;; Match inside a string.
4085 (if (or lookbehind-submatch
4086 (not (integerp tmp)))
4087 ;; See the NOTE above.
4088 (progn (goto-char state-pos) t)
4089 ;; Skip to the end of the string before continuing.
4090 (let ((ender (make-string 1 tmp)) (continue t))
4091 (while (if (search-forward ender bound noerror)
4092 (progn
4093 (setq state (parse-partial-sexp
4094 state-pos (point) nil nil state)
4095 state-pos (point))
4096 (elt state 3))
4097 (setq continue nil)))
4098 continue)))
d9e94c22
MS
4099
4100 ((save-excursion
4101 (save-match-data
4102 (c-beginning-of-macro start)))
4103 ;; Match inside a macro. Skip to the end of it.
4104 (c-end-of-macro)
4105 (cond ((<= (point) bound) t)
4106 (noerror nil)
2a15eb73 4107 (t (signal 'search-failed (list regexp)))))
d9e94c22 4108
2a15eb73
MS
4109 ((and not-inside-token
4110 (or (< check-pos last-token-end-pos)
4111 (< check-pos
4112 (save-excursion
4113 (goto-char check-pos)
4114 (save-match-data
4115 (c-end-of-current-token last-token-end-pos))
4116 (setq last-token-end-pos (point))))))
4117 ;; Inside a token.
4118 (if lookbehind-submatch
4119 ;; See the NOTE above.
4120 (goto-char state-pos)
4121 (goto-char (min last-token-end-pos bound))))
d9e94c22
MS
4122
4123 (t
4124 ;; A real match.
4125 (setq found t)
2a15eb73
MS
4126 nil)))
4127
4128 ;; Should loop to search again, but take care to avoid
4129 ;; looping on the same spot.
4130 (or (/= search-pos (point))
4131 (if (= (point) bound)
4132 (if noerror
4133 nil
4134 (signal 'search-failed (list regexp)))
4135 (forward-char)
4136 t))))
d9e94c22
MS
4137
4138 (error
4139 (goto-char start)
4140 (signal (car err) (cdr err))))
4141
2a15eb73 4142 ;;(message "c-syntactic-re-search-forward done %s" (or (match-end 0) (point)))
d9e94c22
MS
4143
4144 (if found
4145 (progn
2a15eb73
MS
4146 (goto-char (match-end 0))
4147 (match-end 0))
d9e94c22
MS
4148
4149 ;; Search failed. Set point as appropriate.
2a15eb73
MS
4150 (if (eq noerror t)
4151 (goto-char start)
4152 (goto-char bound))
d9e94c22
MS
4153 nil)))
4154
47641aac
GM
4155(defvar safe-pos-list) ; bound in c-syntactic-skip-backward
4156
d0fcee66
AM
4157(defsubst c-ssb-lit-begin ()
4158 ;; Return the start of the literal point is in, or nil.
4159 ;; We read and write the variables `safe-pos', `safe-pos-list', `state'
4160 ;; bound in the caller.
4161
4162 ;; Use `parse-partial-sexp' from a safe position down to the point to check
4163 ;; if it's outside comments and strings.
4164 (save-excursion
4165 (let ((pos (point)) safe-pos state pps-end-pos)
4166 ;; Pick a safe position as close to the point as possible.
4167 ;;
4168 ;; FIXME: Consult `syntax-ppss' here if our cache doesn't give a good
4169 ;; position.
47641aac 4170
d0fcee66
AM
4171 (while (and safe-pos-list
4172 (> (car safe-pos-list) (point)))
4173 (setq safe-pos-list (cdr safe-pos-list)))
4174 (unless (setq safe-pos (car-safe safe-pos-list))
4175 (setq safe-pos (max (or (c-safe-position
4176 (point) (or c-state-cache
4177 (c-parse-state)))
4178 0)
4179 (point-min))
4180 safe-pos-list (list safe-pos)))
4181
4182 ;; Cache positions along the way to use if we have to back up more. We
4183 ;; cache every closing paren on the same level. If the paren cache is
4184 ;; relevant in this region then we're typically already on the same
4185 ;; level as the target position. Note that we might cache positions
4186 ;; after opening parens in case safe-pos is in a nested list. That's
4187 ;; both uncommon and harmless.
4188 (while (progn
4189 (setq state (parse-partial-sexp
4190 safe-pos pos 0))
4191 (< (point) pos))
4192 (setq safe-pos (point)
4193 safe-pos-list (cons safe-pos safe-pos-list)))
4194
4195 ;; If the state contains the start of the containing sexp we cache that
4196 ;; position too, so that parse-partial-sexp in the next run has a bigger
4197 ;; chance of starting at the same level as the target position and thus
4198 ;; will get more good safe positions into the list.
4199 (if (elt state 1)
4200 (setq safe-pos (1+ (elt state 1))
4201 safe-pos-list (cons safe-pos safe-pos-list)))
4202
4203 (if (or (elt state 3) (elt state 4))
4204 ;; Inside string or comment. Continue search at the
4205 ;; beginning of it.
4206 (elt state 8)))))
4207
0386b551 4208(defun c-syntactic-skip-backward (skip-chars &optional limit paren-level)
d9e94c22
MS
4209 "Like `skip-chars-backward' but only look at syntactically relevant chars,
4210i.e. don't stop at positions inside syntactic whitespace or string
4211literals. Preprocessor directives are also ignored, with the exception
4212of the one that the point starts within, if any. If LIMIT is given,
0386b551
AM
4213it's assumed to be at a syntactically relevant position.
4214
4215If PAREN-LEVEL is non-nil, the function won't stop in nested paren
4216sexps, and the search will also not go outside the current paren sexp.
4217However, if LIMIT or the buffer limit is reached inside a nested paren
4218then the point will be left at the limit.
4219
4220Non-nil is returned if the point moved, nil otherwise.
4221
4222Note that this function might do hidden buffer changes. See the
4223comment at the start of cc-engine.el for more info."
d9e94c22
MS
4224
4225 (let ((start (point))
d0fcee66 4226 state-2
d9e94c22
MS
4227 ;; A list of syntactically relevant positions in descending
4228 ;; order. It's used to avoid scanning repeatedly over
4229 ;; potentially large regions with `parse-partial-sexp' to verify
d0fcee66 4230 ;; each position. Used in `c-ssb-lit-begin'
d9e94c22
MS
4231 safe-pos-list
4232 ;; The result from `c-beginning-of-macro' at the start position or the
4233 ;; start position itself if it isn't within a macro. Evaluated on
4234 ;; demand.
0386b551
AM
4235 start-macro-beg
4236 ;; The earliest position after the current one with the same paren
4237 ;; level. Used only when `paren-level' is set.
d0fcee66 4238 lit-beg
0386b551 4239 (paren-level-pos (point)))
d9e94c22 4240
d0fcee66
AM
4241 (while
4242 (progn
4243 ;; The next loop "tries" to find the end point each time round,
4244 ;; loops when it hasn't succeeded.
4245 (while
4246 (and
4247 (< (skip-chars-backward skip-chars limit) 0)
4248
4249 (let ((pos (point)) state-2 pps-end-pos)
4250
4251 (cond
4252 ;; Don't stop inside a literal
4253 ((setq lit-beg (c-ssb-lit-begin))
4254 (goto-char lit-beg)
4255 t)
4256
4257 ((and paren-level
4258 (save-excursion
4259 (setq state-2 (parse-partial-sexp
4260 pos paren-level-pos -1)
4261 pps-end-pos (point))
4262 (/= (car state-2) 0)))
4263 ;; Not at the right level.
4264
4265 (if (and (< (car state-2) 0)
4266 ;; We stop above if we go out of a paren.
4267 ;; Now check whether it precedes or is
4268 ;; nested in the starting sexp.
4269 (save-excursion
4270 (setq state-2
4271 (parse-partial-sexp
4272 pps-end-pos paren-level-pos
4273 nil nil state-2))
4274 (< (car state-2) 0)))
4275
4276 ;; We've stopped short of the starting position
4277 ;; so the hit was inside a nested list. Go up
4278 ;; until we are at the right level.
4279 (condition-case nil
4280 (progn
4281 (goto-char (scan-lists pos -1
4282 (- (car state-2))))
4283 (setq paren-level-pos (point))
4284 (if (and limit (>= limit paren-level-pos))
0386b551 4285 (progn
d0fcee66
AM
4286 (goto-char limit)
4287 nil)
4288 t))
4289 (error
4290 (goto-char (or limit (point-min)))
4291 nil))
4292
4293 ;; The hit was outside the list at the start
4294 ;; position. Go to the start of the list and exit.
4295 (goto-char (1+ (elt state-2 1)))
4296 nil))
4297
4298 ((c-beginning-of-macro limit)
4299 ;; Inside a macro.
4300 (if (< (point)
4301 (or start-macro-beg
4302 (setq start-macro-beg
4303 (save-excursion
4304 (goto-char start)
4305 (c-beginning-of-macro limit)
4306 (point)))))
4307 t
4308
4309 ;; It's inside the same macro we started in so it's
4310 ;; a relevant match.
4311 (goto-char pos)
4312 nil))))))
91af3942 4313
d0fcee66
AM
4314 (> (point)
4315 (progn
4316 ;; Skip syntactic ws afterwards so that we don't stop at the
4317 ;; end of a comment if `skip-chars' is something like "^/".
4318 (c-backward-syntactic-ws)
4319 (point)))))
d9e94c22 4320
0386b551
AM
4321 ;; We might want to extend this with more useful return values in
4322 ;; the future.
4323 (/= (point) start)))
4324
4325;; The following is an alternative implementation of
4326;; `c-syntactic-skip-backward' that uses backward movement to keep
4327;; track of the syntactic context. It turned out to be generally
4328;; slower than the one above which uses forward checks from earlier
4329;; safe positions.
4330;;
4331;;(defconst c-ssb-stop-re
4332;; ;; The regexp matching chars `c-syntactic-skip-backward' needs to
4333;; ;; stop at to avoid going into comments and literals.
4334;; (concat
4335;; ;; Match comment end syntax and string literal syntax. Also match
4336;; ;; '/' for block comment endings (not covered by comment end
4337;; ;; syntax).
4338;; "\\s>\\|/\\|\\s\""
4339;; (if (memq 'gen-string-delim c-emacs-features)
4340;; "\\|\\s|"
4341;; "")
4342;; (if (memq 'gen-comment-delim c-emacs-features)
4343;; "\\|\\s!"
4344;; "")))
4345;;
4346;;(defconst c-ssb-stop-paren-re
4347;; ;; Like `c-ssb-stop-re' but also stops at paren chars.
4348;; (concat c-ssb-stop-re "\\|\\s(\\|\\s)"))
4349;;
4350;;(defconst c-ssb-sexp-end-re
4351;; ;; Regexp matching the ending syntax of a complex sexp.
4352;; (concat c-string-limit-regexp "\\|\\s)"))
4353;;
4354;;(defun c-syntactic-skip-backward (skip-chars &optional limit paren-level)
4355;; "Like `skip-chars-backward' but only look at syntactically relevant chars,
4356;;i.e. don't stop at positions inside syntactic whitespace or string
4357;;literals. Preprocessor directives are also ignored. However, if the
4358;;point is within a comment, string literal or preprocessor directory to
4359;;begin with, its contents is treated as syntactically relevant chars.
4360;;If LIMIT is given, it limits the backward search and the point will be
4361;;left there if no earlier position is found.
4362;;
4363;;If PAREN-LEVEL is non-nil, the function won't stop in nested paren
4364;;sexps, and the search will also not go outside the current paren sexp.
4365;;However, if LIMIT or the buffer limit is reached inside a nested paren
4366;;then the point will be left at the limit.
4367;;
4368;;Non-nil is returned if the point moved, nil otherwise.
4369;;
4370;;Note that this function might do hidden buffer changes. See the
4371;;comment at the start of cc-engine.el for more info."
4372;;
4373;; (save-restriction
4374;; (when limit
4375;; (narrow-to-region limit (point-max)))
4376;;
4377;; (let ((start (point)))
4378;; (catch 'done
4379;; (while (let ((last-pos (point))
4380;; (stop-pos (progn
4381;; (skip-chars-backward skip-chars)
4382;; (point))))
4383;;
4384;; ;; Skip back over the same region as
4385;; ;; `skip-chars-backward' above, but keep to
4386;; ;; syntactically relevant positions.
4387;; (goto-char last-pos)
4388;; (while (and
4389;; ;; `re-search-backward' with a single char regexp
4390;; ;; should be fast.
4391;; (re-search-backward
4392;; (if paren-level c-ssb-stop-paren-re c-ssb-stop-re)
4393;; stop-pos 'move)
4394;;
4395;; (progn
4396;; (cond
4397;; ((looking-at "\\s(")
4398;; ;; `paren-level' is set and we've found the
4399;; ;; start of the containing paren.
4400;; (forward-char)
4401;; (throw 'done t))
4402;;
4403;; ((looking-at c-ssb-sexp-end-re)
4404;; ;; We're at the end of a string literal or paren
4405;; ;; sexp (if `paren-level' is set).
4406;; (forward-char)
4407;; (condition-case nil
4408;; (c-backward-sexp)
4409;; (error
4410;; (goto-char limit)
4411;; (throw 'done t))))
4412;;
4413;; (t
4414;; (forward-char)
4415;; ;; At the end of some syntactic ws or possibly
4416;; ;; after a plain '/' operator.
4417;; (let ((pos (point)))
4418;; (c-backward-syntactic-ws)
4419;; (if (= pos (point))
4420;; ;; Was a plain '/' operator. Go past it.
4421;; (backward-char)))))
4422;;
4423;; (> (point) stop-pos))))
4424;;
4425;; ;; Now the point is either at `stop-pos' or at some
4426;; ;; position further back if `stop-pos' was at a
4427;; ;; syntactically irrelevant place.
4428;;
4429;; ;; Skip additional syntactic ws so that we don't stop
4430;; ;; at the end of a comment if `skip-chars' is
4431;; ;; something like "^/".
4432;; (c-backward-syntactic-ws)
4433;;
4434;; (< (point) stop-pos))))
4435;;
4436;; ;; We might want to extend this with more useful return values
4437;; ;; in the future.
4438;; (/= (point) start))))
d9e94c22
MS
4439
4440\f
4441;; Tools for handling comments and string literals.
4442
13d49cbb 4443(defun c-in-literal (&optional lim detect-cpp)
d9e94c22
MS
4444 "Return the type of literal point is in, if any.
4445The return value is `c' if in a C-style comment, `c++' if in a C++
4446style comment, `string' if in a string literal, `pound' if DETECT-CPP
4447is non-nil and in a preprocessor line, or nil if somewhere else.
4448Optional LIM is used as the backward limit of the search. If omitted,
4449or nil, `c-beginning-of-defun' is used.
4450
4451The last point calculated is cached if the cache is enabled, i.e. if
4452`c-in-literal-cache' is bound to a two element vector.
4453
0386b551
AM
4454Note that this function might do hidden buffer changes. See the
4455comment at the start of cc-engine.el for more info."
2cc769a8
AM
4456 (save-restriction
4457 (widen)
56d093a9 4458 (let* ((safe-place (c-state-semi-safe-place (point)))
2cc769a8
AM
4459 (lit (c-state-pp-to-literal safe-place (point))))
4460 (or (cadr lit)
4461 (and detect-cpp
4462 (save-excursion (c-beginning-of-macro))
4463 'pound)))))
d9e94c22
MS
4464
4465(defun c-literal-limits (&optional lim near not-in-delimiter)
4466 "Return a cons of the beginning and end positions of the comment or
4467string surrounding point (including both delimiters), or nil if point
4468isn't in one. If LIM is non-nil, it's used as the \"safe\" position
4469to start parsing from. If NEAR is non-nil, then the limits of any
4470literal next to point is returned. \"Next to\" means there's only
4471spaces and tabs between point and the literal. The search for such a
4472literal is done first in forward direction. If NOT-IN-DELIMITER is
4473non-nil, the case when point is inside a starting delimiter won't be
a85fd6da 4474recognized. This only has effect for comments which have starting
d9e94c22
MS
4475delimiters with more than one character.
4476
0386b551
AM
4477Note that this function might do hidden buffer changes. See the
4478comment at the start of cc-engine.el for more info."
d9e94c22
MS
4479
4480 (save-excursion
4481 (let* ((pos (point))
56d093a9 4482 (lim (or lim (c-state-semi-safe-place pos)))
2cc769a8
AM
4483 (pp-to-lit (save-restriction
4484 (widen)
691e26ae 4485 (c-state-pp-to-literal lim pos not-in-delimiter)))
13d49cbb 4486 (state (car pp-to-lit))
13d49cbb 4487 (lit-limits (car (cddr pp-to-lit))))
d9e94c22 4488
13d49cbb
AM
4489 (cond
4490 (lit-limits)
13d49cbb
AM
4491
4492 (near
4493 (goto-char pos)
4494 ;; Search forward for a literal.
4495 (skip-chars-forward " \t")
4496 (cond
4497 ((looking-at c-string-limit-regexp) ; String.
4498 (cons (point) (or (c-safe (c-forward-sexp 1) (point))
4499 (point-max))))
d9e94c22 4500
13d49cbb
AM
4501 ((looking-at c-comment-start-regexp) ; Line or block comment.
4502 (cons (point) (progn (c-forward-single-comment) (point))))
4503
4504 (t
4505 ;; Search backward.
4506 (skip-chars-backward " \t")
4507
4508 (let ((end (point)) beg)
4509 (cond
4510 ((save-excursion
4511 (< (skip-syntax-backward c-string-syntax) 0)) ; String.
4512 (setq beg (c-safe (c-backward-sexp 1) (point))))
4513
4514 ((and (c-safe (forward-char -2) t)
4515 (looking-at "*/"))
4516 ;; Block comment. Due to the nature of line
4517 ;; comments, they will always be covered by the
4518 ;; normal case above.
4519 (goto-char end)
4520 (c-backward-single-comment)
4521 ;; If LIM is bogus, beg will be bogus.
4522 (setq beg (point))))
4523
4524 (if beg (cons beg end))))))
4525 ))))
d9e94c22 4526
0386b551
AM
4527;; In case external callers use this; it did have a docstring.
4528(defalias 'c-literal-limits-fast 'c-literal-limits)
d9e94c22 4529
0386b551
AM
4530(defun c-collect-line-comments (range)
4531 "If the argument is a cons of two buffer positions (such as returned by
4532`c-literal-limits'), and that range contains a C++ style line comment,
4533then an extended range is returned that contains all adjacent line
4534comments (i.e. all comments that starts in the same column with no
4535empty lines or non-whitespace characters between them). Otherwise the
4536argument is returned.
d9e94c22 4537
0386b551
AM
4538Note that this function might do hidden buffer changes. See the
4539comment at the start of cc-engine.el for more info."
d9e94c22
MS
4540
4541 (save-excursion
0386b551
AM
4542 (condition-case nil
4543 (if (and (consp range) (progn
4544 (goto-char (car range))
4545 (looking-at c-line-comment-starter)))
b414f371 4546 (let ((col (current-column))
0386b551
AM
4547 (beg (point))
4548 (bopl (c-point 'bopl))
4549 (end (cdr range)))
4550 ;; Got to take care in the backward direction to handle
4551 ;; comments which are preceded by code.
4552 (while (and (c-backward-single-comment)
4553 (>= (point) bopl)
4554 (looking-at c-line-comment-starter)
4555 (= col (current-column)))
4556 (setq beg (point)
4557 bopl (c-point 'bopl)))
4558 (goto-char end)
4559 (while (and (progn (skip-chars-forward " \t")
4560 (looking-at c-line-comment-starter))
4561 (= col (current-column))
4562 (prog1 (zerop (forward-line 1))
4563 (setq end (point)))))
4564 (cons beg end))
4565 range)
4566 (error range))))
d9e94c22
MS
4567
4568(defun c-literal-type (range)
4569 "Convenience function that given the result of `c-literal-limits',
a85fd6da
AM
4570returns nil or the type of literal that the range surrounds, one
4571of the symbols 'c, 'c++ or 'string. It's much faster than using
4572`c-in-literal' and is intended to be used when you need both the
4573type of a literal and its limits.
d9e94c22 4574
0386b551
AM
4575Note that this function might do hidden buffer changes. See the
4576comment at the start of cc-engine.el for more info."
4577
d9e94c22
MS
4578 (if (consp range)
4579 (save-excursion
4580 (goto-char (car range))
4581 (cond ((looking-at c-string-limit-regexp) 'string)
4582 ((or (looking-at "//") ; c++ line comment
4583 (and (looking-at "\\s<") ; comment starter
4584 (looking-at "#"))) ; awk comment.
4585 'c++)
4586 (t 'c))) ; Assuming the range is valid.
4587 range))
4588
9657183b
AM
4589(defsubst c-determine-limit-get-base (start try-size)
4590 ;; Get a "safe place" approximately TRY-SIZE characters before START.
4591 ;; This doesn't preserve point.
4592 (let* ((pos (max (- start try-size) (point-min)))
56d093a9 4593 (base (c-state-semi-safe-place pos))
9657183b
AM
4594 (s (parse-partial-sexp base pos)))
4595 (if (or (nth 4 s) (nth 3 s)) ; comment or string
4596 (nth 8 s)
4597 (point))))
4598
4599(defun c-determine-limit (how-far-back &optional start try-size)
4600 ;; Return a buffer position HOW-FAR-BACK non-literal characters from START
4601 ;; (default point). This is done by going back further in the buffer then
4602 ;; searching forward for literals. The position found won't be in a
4603 ;; literal. We start searching for the sought position TRY-SIZE (default
4604 ;; twice HOW-FAR-BACK) bytes back from START. This function must be fast.
4605 ;; :-)
4606 (save-excursion
4607 (let* ((start (or start (point)))
4608 (try-size (or try-size (* 2 how-far-back)))
4609 (base (c-determine-limit-get-base start try-size))
4610 (pos base)
4611
4612 (s (parse-partial-sexp pos pos)) ; null state.
4613 stack elt size
4614 (count 0))
4615 (while (< pos start)
4616 ;; Move forward one literal each time round this loop.
4617 ;; Move forward to the start of a comment or string.
4618 (setq s (parse-partial-sexp
4619 pos
4620 start
4621 nil ; target-depth
4622 nil ; stop-before
4623 s ; state
4624 'syntax-table)) ; stop-comment
4625
4626 ;; Gather details of the non-literal-bit - starting pos and size.
4627 (setq size (- (if (or (nth 4 s) (nth 3 s))
4628 (nth 8 s)
4629 (point))
4630 pos))
4631 (if (> size 0)
4632 (setq stack (cons (cons pos size) stack)))
4633
4634 ;; Move forward to the end of the comment/string.
4635 (if (or (nth 4 s) (nth 3 s))
4636 (setq s (parse-partial-sexp
4637 (point)
4638 start
4639 nil ; target-depth
4640 nil ; stop-before
4641 s ; state
4642 'syntax-table))) ; stop-comment
4643 (setq pos (point)))
c38e0c97 4644
9657183b
AM
4645 ;; Now try and find enough non-literal characters recorded on the stack.
4646 ;; Go back one recorded literal each time round this loop.
4647 (while (and (< count how-far-back)
4648 stack)
4649 (setq elt (car stack)
4650 stack (cdr stack))
4651 (setq count (+ count (cdr elt))))
4652
4653 ;; Have we found enough yet?
4654 (cond
4655 ((>= count how-far-back)
4656 (+ (car elt) (- count how-far-back)))
4657 ((eq base (point-min))
4658 (point-min))
4659 (t
4660 (c-determine-limit (- how-far-back count) base try-size))))))
2e492df3
AM
4661
4662(defun c-determine-+ve-limit (how-far &optional start-pos)
4663 ;; Return a buffer position about HOW-FAR non-literal characters forward
4664 ;; from START-POS (default point), which must not be inside a literal.
4665 (save-excursion
4666 (let ((pos (or start-pos (point)))
4667 (count how-far)
4668 (s (parse-partial-sexp (point) (point)))) ; null state
4669 (while (and (not (eobp))
4670 (> count 0))
4671 ;; Scan over counted characters.
4672 (setq s (parse-partial-sexp
4673 pos
4674 (min (+ pos count) (point-max))
4675 nil ; target-depth
4676 nil ; stop-before
4677 s ; state
4678 'syntax-table)) ; stop-comment
4679 (setq count (- count (- (point) pos) 1)
4680 pos (point))
4681 ;; Scan over literal characters.
4682 (if (nth 8 s)
4683 (setq s (parse-partial-sexp
4684 pos
4685 (point-max)
4686 nil ; target-depth
4687 nil ; stop-before
4688 s ; state
4689 'syntax-table) ; stop-comment
4690 pos (point))))
4691 (point))))
4692
d9e94c22
MS
4693\f
4694;; `c-find-decl-spots' and accompanying stuff.
4695
4696;; Variables used in `c-find-decl-spots' to cache the search done for
4697;; the first declaration in the last call. When that function starts,
4698;; it needs to back up over syntactic whitespace to look at the last
4699;; token before the region being searched. That can sometimes cause
4700;; moves back and forth over a quite large region of comments and
4701;; macros, which would be repeated for each changed character when
4702;; we're called during fontification, since font-lock refontifies the
4703;; current line for each change. Thus it's worthwhile to cache the
4704;; first match.
4705;;
4706;; `c-find-decl-syntactic-pos' is a syntactically relevant position in
4707;; the syntactic whitespace less or equal to some start position.
4708;; There's no cached value if it's nil.
4709;;
4710;; `c-find-decl-match-pos' is the match position if
4711;; `c-find-decl-prefix-search' matched before the syntactic whitespace
4712;; at `c-find-decl-syntactic-pos', or nil if there's no such match.
4713(defvar c-find-decl-syntactic-pos nil)
4714(make-variable-buffer-local 'c-find-decl-syntactic-pos)
4715(defvar c-find-decl-match-pos nil)
4716(make-variable-buffer-local 'c-find-decl-match-pos)
4717
4718(defsubst c-invalidate-find-decl-cache (change-min-pos)
4719 (and c-find-decl-syntactic-pos
4720 (< change-min-pos c-find-decl-syntactic-pos)
4721 (setq c-find-decl-syntactic-pos nil)))
4722
4723; (defface c-debug-decl-spot-face
4724; '((t (:background "Turquoise")))
4725; "Debug face to mark the spots where `c-find-decl-spots' stopped.")
4726; (defface c-debug-decl-sws-face
4727; '((t (:background "Khaki")))
4728; "Debug face to mark the syntactic whitespace between the declaration
4729; spots and the preceding token end.")
4730
4731(defmacro c-debug-put-decl-spot-faces (match-pos decl-pos)
4732 (when (facep 'c-debug-decl-spot-face)
0386b551 4733 `(c-save-buffer-state ((match-pos ,match-pos) (decl-pos ,decl-pos))
d9e94c22
MS
4734 (c-debug-add-face (max match-pos (point-min)) decl-pos
4735 'c-debug-decl-sws-face)
4736 (c-debug-add-face decl-pos (min (1+ decl-pos) (point-max))
4737 'c-debug-decl-spot-face))))
4738(defmacro c-debug-remove-decl-spot-faces (beg end)
4739 (when (facep 'c-debug-decl-spot-face)
0386b551 4740 `(c-save-buffer-state ()
d9e94c22
MS
4741 (c-debug-remove-face ,beg ,end 'c-debug-decl-spot-face)
4742 (c-debug-remove-face ,beg ,end 'c-debug-decl-sws-face))))
4743
4744(defmacro c-find-decl-prefix-search ()
4745 ;; Macro used inside `c-find-decl-spots'. It ought to be a defun,
4746 ;; but it contains lots of free variables that refer to things
4747 ;; inside `c-find-decl-spots'. The point is left at `cfd-match-pos'
4748 ;; if there is a match, otherwise at `cfd-limit'.
0386b551 4749 ;;
ece15004
AM
4750 ;; The macro moves point forward to the next putative start of a declaration
4751 ;; or cfd-limit. This decl start is the next token after a "declaration
4752 ;; prefix". The declaration prefix is the earlier of `cfd-prop-match' and
4753 ;; `cfd-re-match'. `cfd-match-pos' is set to the decl prefix.
4754 ;;
0386b551 4755 ;; This macro might do hidden buffer changes.
d9e94c22
MS
4756
4757 '(progn
4758 ;; Find the next property match position if we haven't got one already.
4759 (unless cfd-prop-match
4760 (save-excursion
4761 (while (progn
4762 (goto-char (next-single-property-change
4763 (point) 'c-type nil cfd-limit))
4764 (and (< (point) cfd-limit)
4765 (not (eq (c-get-char-property (1- (point)) 'c-type)
4766 'c-decl-end)))))
4767 (setq cfd-prop-match (point))))
4768
0386b551
AM
4769 ;; Find the next `c-decl-prefix-or-start-re' match if we haven't
4770 ;; got one already.
d9e94c22 4771 (unless cfd-re-match
0386b551
AM
4772
4773 (if (> cfd-re-match-end (point))
4774 (goto-char cfd-re-match-end))
4775
ece15004
AM
4776 ;; Each time round, the next `while' moves forward over a pseudo match
4777 ;; of `c-decl-prefix-or-start-re' which is either inside a literal, or
4778 ;; is a ":" not preceded by "public", etc.. `cfd-re-match' and
4779 ;; `cfd-re-match-end' get set.
4780 (while
4781 (progn
4782 (setq cfd-re-match-end (re-search-forward c-decl-prefix-or-start-re
4783 cfd-limit 'move))
4784 (cond
4785 ((null cfd-re-match-end)
4786 ;; No match. Finish up and exit the loop.
4787 (setq cfd-re-match cfd-limit)
4788 nil)
4789 ((c-got-face-at
4790 (if (setq cfd-re-match (match-end 1))
4791 ;; Matched the end of a token preceding a decl spot.
4792 (progn
4793 (goto-char cfd-re-match)
4794 (1- cfd-re-match))
4795 ;; Matched a token that start a decl spot.
4796 (goto-char (match-beginning 0))
4797 (point))
4798 c-literal-faces)
4799 ;; Pseudo match inside a comment or string literal. Skip out
4800 ;; of comments and string literals.
4801 (while (progn
4802 (goto-char (next-single-property-change
4803 (point) 'face nil cfd-limit))
4804 (and (< (point) cfd-limit)
4805 (c-got-face-at (point) c-literal-faces))))
4806 t) ; Continue the loop over pseudo matches.
4807 ((and (match-string 1)
4808 (string= (match-string 1) ":")
4809 (save-excursion
4810 (or (/= (c-backward-token-2 2) 0) ; no search limit. :-(
4811 (not (looking-at c-decl-start-colon-kwd-re)))))
4812 ;; Found a ":" which isn't part of "public:", etc.
4813 t)
4814 (t nil)))) ;; Found a real match. Exit the pseudo-match loop.
4815
4816 ;; If our match was at the decl start, we have to back up over the
0386b551
AM
4817 ;; preceding syntactic ws to set `cfd-match-pos' and to catch
4818 ;; any decl spots in the syntactic ws.
d9e94c22 4819 (unless cfd-re-match
0386b551
AM
4820 (c-backward-syntactic-ws)
4821 (setq cfd-re-match (point))))
d9e94c22
MS
4822
4823 ;; Choose whichever match is closer to the start.
4824 (if (< cfd-re-match cfd-prop-match)
4825 (setq cfd-match-pos cfd-re-match
4826 cfd-re-match nil)
4827 (setq cfd-match-pos cfd-prop-match
4828 cfd-prop-match nil))
4829
4830 (goto-char cfd-match-pos)
4831
4832 (when (< cfd-match-pos cfd-limit)
4833 ;; Skip forward past comments only so we don't skip macros.
4834 (c-forward-comments)
4835 ;; Set the position to continue at. We can avoid going over
4836 ;; the comments skipped above a second time, but it's possible
4837 ;; that the comment skipping has taken us past `cfd-prop-match'
4838 ;; since the property might be used inside comments.
4839 (setq cfd-continue-pos (if cfd-prop-match
4840 (min cfd-prop-match (point))
4841 (point))))))
4842
4843(defun c-find-decl-spots (cfd-limit cfd-decl-re cfd-face-checklist cfd-fun)
0386b551 4844 ;; Call CFD-FUN for each possible spot for a declaration, cast or
a85fd6da
AM
4845 ;; label from the point to CFD-LIMIT.
4846 ;;
9657183b
AM
4847 ;; CFD-FUN is called with point at the start of the spot. It's passed two
4848 ;; arguments: The first is the end position of the token preceding the spot,
4849 ;; or 0 for the implicit match at bob. The second is a flag that is t when
4850 ;; the match is inside a macro. Point should be moved forward by at least
4851 ;; one token.
4852 ;;
4853 ;; If CFD-FUN adds `c-decl-end' properties somewhere below the current spot,
4854 ;; it should return non-nil to ensure that the next search will find them.
0386b551 4855 ;;
a85fd6da 4856 ;; Such a spot is:
0386b551
AM
4857 ;; o The first token after bob.
4858 ;; o The first token after the end of submatch 1 in
4859 ;; `c-decl-prefix-or-start-re' when that submatch matches.
4860 ;; o The start of each `c-decl-prefix-or-start-re' match when
4861 ;; submatch 1 doesn't match.
5a89f0a7 4862 ;; o The first token after the end of each occurrence of the
0386b551
AM
4863 ;; `c-type' text property with the value `c-decl-end', provided
4864 ;; `c-type-decl-end-used' is set.
4865 ;;
4866 ;; Only a spot that match CFD-DECL-RE and whose face is in the
4867 ;; CFD-FACE-CHECKLIST list causes CFD-FUN to be called. The face
4868 ;; check is disabled if CFD-FACE-CHECKLIST is nil.
d9e94c22
MS
4869 ;;
4870 ;; If the match is inside a macro then the buffer is narrowed to the
4871 ;; end of it, so that CFD-FUN can investigate the following tokens
4872 ;; without matching something that begins inside a macro and ends
4873 ;; outside it. It's to avoid this work that the CFD-DECL-RE and
4874 ;; CFD-FACE-CHECKLIST checks exist.
4875 ;;
0386b551
AM
4876 ;; The spots are visited approximately in order from top to bottom.
4877 ;; It's however the positions where `c-decl-prefix-or-start-re'
4878 ;; matches and where `c-decl-end' properties are found that are in
4879 ;; order. Since the spots often are at the following token, they
4880 ;; might be visited out of order insofar as more spots are reported
4881 ;; later on within the syntactic whitespace between the match
4882 ;; positions and their spots.
4883 ;;
4884 ;; It's assumed that comments and strings are fontified in the
d9e94c22
MS
4885 ;; searched range.
4886 ;;
4887 ;; This is mainly used in fontification, and so has an elaborate
4888 ;; cache to handle repeated calls from the same start position; see
4889 ;; the variables above.
4890 ;;
4891 ;; All variables in this function begin with `cfd-' to avoid name
4892 ;; collision with the (dynamically bound) variables used in CFD-FUN.
0386b551
AM
4893 ;;
4894 ;; This function might do hidden buffer changes.
d9e94c22 4895
0386b551
AM
4896 (let ((cfd-start-pos (point))
4897 (cfd-buffer-end (point-max))
4898 ;; The end of the token preceding the decl spot last found
4899 ;; with `c-decl-prefix-or-start-re'. `cfd-limit' if there's
4900 ;; no match.
d9e94c22 4901 cfd-re-match
0386b551
AM
4902 ;; The end position of the last `c-decl-prefix-or-start-re'
4903 ;; match. If this is greater than `cfd-continue-pos', the
4904 ;; next regexp search is started here instead.
4905 (cfd-re-match-end (point-min))
4906 ;; The end of the last `c-decl-end' found by
4907 ;; `c-find-decl-prefix-search'. `cfd-limit' if there's no
4908 ;; match. If searching for the property isn't needed then we
4909 ;; disable it by setting it to `cfd-limit' directly.
d9e94c22 4910 (cfd-prop-match (unless c-type-decl-end-used cfd-limit))
0386b551
AM
4911 ;; The end of the token preceding the decl spot last found by
4912 ;; `c-find-decl-prefix-search'. 0 for the implicit match at
4913 ;; bob. `cfd-limit' if there's no match. In other words,
4914 ;; this is the minimum of `cfd-re-match' and `cfd-prop-match'.
d9e94c22
MS
4915 (cfd-match-pos cfd-limit)
4916 ;; The position to continue searching at.
4917 cfd-continue-pos
4918 ;; The position of the last "real" token we've stopped at.
4919 ;; This can be greater than `cfd-continue-pos' when we get
4920 ;; hits inside macros or at `c-decl-end' positions inside
4921 ;; comments.
4922 (cfd-token-pos 0)
4923 ;; The end position of the last entered macro.
4924 (cfd-macro-end 0))
4925
4926 ;; Initialize by finding a syntactically relevant start position
0386b551
AM
4927 ;; before the point, and do the first `c-decl-prefix-or-start-re'
4928 ;; search unless we're at bob.
d9e94c22 4929
0386b551 4930 (let (start-in-literal start-in-macro syntactic-pos)
d9e94c22
MS
4931 ;; Must back up a bit since we look for the end of the previous
4932 ;; statement or declaration, which is earlier than the first
4933 ;; returned match.
4934
0386b551
AM
4935 (cond
4936 ;; First we need to move to a syntactically relevant position.
4937 ;; Begin by backing out of comment or string literals.
4938 ((and
4939 (when (c-got-face-at (point) c-literal-faces)
4940 ;; Try to use the faces to back up to the start of the
4941 ;; literal. FIXME: What if the point is on a declaration
4942 ;; inside a comment?
4943 (while (and (not (bobp))
4944 (c-got-face-at (1- (point)) c-literal-faces))
4945 (goto-char (previous-single-property-change
4946 (point) 'face nil (point-min))))
4947
4948 ;; XEmacs doesn't fontify the quotes surrounding string
4949 ;; literals.
4950 (and (featurep 'xemacs)
4951 (eq (get-text-property (point) 'face)
4952 'font-lock-string-face)
4953 (not (bobp))
4954 (progn (backward-char)
4955 (not (looking-at c-string-limit-regexp)))
4956 (forward-char))
4957
4958 ;; Don't trust the literal to contain only literal faces
4959 ;; (the font lock package might not have fontified the
4960 ;; start of it at all, for instance) so check that we have
4961 ;; arrived at something that looks like a start or else
4962 ;; resort to `c-literal-limits'.
4963 (unless (looking-at c-literal-start-regexp)
4964 (let ((range (c-literal-limits)))
4965 (if range (goto-char (car range)))))
4966
4967 (setq start-in-literal (point)))
4968
4969 ;; The start is in a literal. If the limit is in the same
4970 ;; one we don't have to find a syntactic position etc. We
4971 ;; only check that if the limit is at or before bonl to save
4972 ;; time; it covers the by far most common case when font-lock
4973 ;; refontifies the current line only.
4974 (<= cfd-limit (c-point 'bonl cfd-start-pos))
4975 (save-excursion
4976 (goto-char cfd-start-pos)
4977 (while (progn
4978 (goto-char (next-single-property-change
4979 (point) 'face nil cfd-limit))
4980 (and (< (point) cfd-limit)
4981 (c-got-face-at (point) c-literal-faces))))
4982 (= (point) cfd-limit)))
4983
4984 ;; Completely inside a literal. Set up variables to trig the
4985 ;; (< cfd-continue-pos cfd-start-pos) case below and it'll
4986 ;; find a suitable start position.
4987 (setq cfd-continue-pos start-in-literal))
4988
4989 ;; Check if the region might be completely inside a macro, to
4990 ;; optimize that like the completely-inside-literal above.
4991 ((save-excursion
4992 (and (= (forward-line 1) 0)
4993 (bolp) ; forward-line has funny behavior at eob.
4994 (>= (point) cfd-limit)
4995 (progn (backward-char)
4996 (eq (char-before) ?\\))))
4997 ;; (Maybe) completely inside a macro. Only need to trig the
4998 ;; (< cfd-continue-pos cfd-start-pos) case below to make it
4999 ;; set things up.
5000 (setq cfd-continue-pos (1- cfd-start-pos)
5001 start-in-macro t))
d9e94c22 5002
0386b551
AM
5003 (t
5004 ;; Back out of any macro so we don't miss any declaration
5005 ;; that could follow after it.
5006 (when (c-beginning-of-macro)
5007 (setq start-in-macro t))
5008
5009 ;; Now we're at a proper syntactically relevant position so we
5010 ;; can use the cache. But first clear it if it applied
5011 ;; further down.
5012 (c-invalidate-find-decl-cache cfd-start-pos)
5013
5014 (setq syntactic-pos (point))
5015 (unless (eq syntactic-pos c-find-decl-syntactic-pos)
5016 ;; Don't have to do this if the cache is relevant here,
5017 ;; typically if the same line is refontified again. If
5018 ;; we're just some syntactic whitespace further down we can
5019 ;; still use the cache to limit the skipping.
5020 (c-backward-syntactic-ws c-find-decl-syntactic-pos))
5021
5022 ;; If we hit `c-find-decl-syntactic-pos' and
5023 ;; `c-find-decl-match-pos' is set then we install the cached
5024 ;; values. If we hit `c-find-decl-syntactic-pos' and
5025 ;; `c-find-decl-match-pos' is nil then we know there's no decl
5026 ;; prefix in the whitespace before `c-find-decl-syntactic-pos'
5027 ;; and so we can continue the search from this point. If we
5028 ;; didn't hit `c-find-decl-syntactic-pos' then we're now in
5029 ;; the right spot to begin searching anyway.
5030 (if (and (eq (point) c-find-decl-syntactic-pos)
5031 c-find-decl-match-pos)
d9e94c22
MS
5032 (setq cfd-match-pos c-find-decl-match-pos
5033 cfd-continue-pos syntactic-pos)
0386b551
AM
5034
5035 (setq c-find-decl-syntactic-pos syntactic-pos)
5036
5037 (when (if (bobp)
5038 ;; Always consider bob a match to get the first
5039 ;; declaration in the file. Do this separately instead of
5040 ;; letting `c-decl-prefix-or-start-re' match bob, so that
5041 ;; regexp always can consume at least one character to
5042 ;; ensure that we won't get stuck in an infinite loop.
5043 (setq cfd-re-match 0)
5044 (backward-char)
5045 (c-beginning-of-current-token)
5046 (< (point) cfd-limit))
5047 ;; Do an initial search now. In the bob case above it's
5048 ;; only done to search for a `c-decl-end' spot.
5049 (c-find-decl-prefix-search))
5050
5051 (setq c-find-decl-match-pos (and (< cfd-match-pos cfd-start-pos)
5052 cfd-match-pos)))))
5053
5054 ;; Advance `cfd-continue-pos' if it's before the start position.
5055 ;; The closest continue position that might have effect at or
5056 ;; after the start depends on what we started in. This also
5057 ;; finds a suitable start position in the special cases when the
5058 ;; region is completely within a literal or macro.
5059 (when (and cfd-continue-pos (< cfd-continue-pos cfd-start-pos))
5060
5061 (cond
5062 (start-in-macro
5063 ;; If we're in a macro then it's the closest preceding token
5064 ;; in the macro. Check this before `start-in-literal',
5065 ;; since if we're inside a literal in a macro, the preceding
5066 ;; token is earlier than any `c-decl-end' spot inside the
5067 ;; literal (comment).
5068 (goto-char (or start-in-literal cfd-start-pos))
5069 ;; The only syntactic ws in macros are comments.
d9e94c22 5070 (c-backward-comments)
0386b551
AM
5071 (backward-char)
5072 (c-beginning-of-current-token))
5073
5074 (start-in-literal
5075 ;; If we're in a comment it can only be the closest
5076 ;; preceding `c-decl-end' position within that comment, if
5077 ;; any. Go back to the beginning of such a property so that
5078 ;; `c-find-decl-prefix-search' will find the end of it.
5079 ;; (Can't stop at the end and install it directly on
5080 ;; `cfd-prop-match' since that variable might be cleared
5081 ;; after `cfd-fun' below.)
5082 ;;
5083 ;; Note that if the literal is a string then the property
5084 ;; search will simply skip to the beginning of it right
5085 ;; away.
5086 (if (not c-type-decl-end-used)
5087 (goto-char start-in-literal)
5088 (goto-char cfd-start-pos)
5089 (while (progn
5090 (goto-char (previous-single-property-change
5091 (point) 'c-type nil start-in-literal))
5092 (and (> (point) start-in-literal)
5093 (not (eq (c-get-char-property (point) 'c-type)
5094 'c-decl-end))))))
5095
5096 (when (= (point) start-in-literal)
5097 ;; Didn't find any property inside the comment, so we can
5098 ;; skip it entirely. (This won't skip past a string, but
5099 ;; that'll be handled quickly by the next
5100 ;; `c-find-decl-prefix-search' anyway.)
5101 (c-forward-single-comment)
5102 (if (> (point) cfd-limit)
5103 (goto-char cfd-limit))))
d9e94c22 5104
0386b551
AM
5105 (t
5106 ;; If we started in normal code, the only match that might
5107 ;; apply before the start is what we already got in
5108 ;; `cfd-match-pos' so we can continue at the start position.
5109 ;; (Note that we don't get here if the first match is below
5110 ;; it.)
5111 (goto-char cfd-start-pos)))
5112
5113 ;; Delete found matches if they are before our new continue
5114 ;; position, so that `c-find-decl-prefix-search' won't back up
5115 ;; to them later on.
5116 (setq cfd-continue-pos (point))
5117 (when (and cfd-re-match (< cfd-re-match cfd-continue-pos))
5118 (setq cfd-re-match nil))
5119 (when (and cfd-prop-match (< cfd-prop-match cfd-continue-pos))
5120 (setq cfd-prop-match nil)))
5121
5122 (if syntactic-pos
5123 ;; This is the normal case and we got a proper syntactic
5124 ;; position. If there's a match then it's always outside
5125 ;; macros and comments, so advance to the next token and set
5126 ;; `cfd-token-pos'. The loop below will later go back using
5127 ;; `cfd-continue-pos' to fix declarations inside the
5128 ;; syntactic ws.
5129 (when (and cfd-match-pos (< cfd-match-pos syntactic-pos))
5130 (goto-char syntactic-pos)
5131 (c-forward-syntactic-ws)
5132 (and cfd-continue-pos
5133 (< cfd-continue-pos (point))
5134 (setq cfd-token-pos (point))))
5135
5136 ;; Have one of the special cases when the region is completely
5137 ;; within a literal or macro. `cfd-continue-pos' is set to a
5138 ;; good start position for the search, so do it.
5139 (c-find-decl-prefix-search)))
d9e94c22 5140
51c9af45 5141 ;; Now loop. Round what? (ACM, 2006/7/5). We already got the first match.
d9e94c22
MS
5142
5143 (while (progn
5144 (while (and
5145 (< cfd-match-pos cfd-limit)
5146
5147 (or
5148 ;; Kludge to filter out matches on the "<" that
5149 ;; aren't open parens, for the sake of languages
5150 ;; that got `c-recognize-<>-arglists' set.
5151 (and (eq (char-before cfd-match-pos) ?<)
5152 (not (c-get-char-property (1- cfd-match-pos)
5153 'syntax-table)))
5154
5155 ;; If `cfd-continue-pos' is less or equal to
5156 ;; `cfd-token-pos', we've got a hit inside a macro
5157 ;; that's in the syntactic whitespace before the last
5158 ;; "real" declaration we've checked. If they're equal
5159 ;; we've arrived at the declaration a second time, so
5160 ;; there's nothing to do.
5161 (= cfd-continue-pos cfd-token-pos)
5162
5163 (progn
5164 ;; If `cfd-continue-pos' is less than `cfd-token-pos'
5165 ;; we're still searching for declarations embedded in
5166 ;; the syntactic whitespace. In that case we need
5167 ;; only to skip comments and not macros, since they
5168 ;; can't be nested, and that's already been done in
5169 ;; `c-find-decl-prefix-search'.
5170 (when (> cfd-continue-pos cfd-token-pos)
5171 (c-forward-syntactic-ws)
5172 (setq cfd-token-pos (point)))
5173
5174 ;; Continue if the following token fails the
5175 ;; CFD-DECL-RE and CFD-FACE-CHECKLIST checks.
5176 (when (or (>= (point) cfd-limit)
5177 (not (looking-at cfd-decl-re))
5178 (and cfd-face-checklist
5179 (not (c-got-face-at
5180 (point) cfd-face-checklist))))
5181 (goto-char cfd-continue-pos)
5182 t)))
5183
5184 (< (point) cfd-limit))
5185 (c-find-decl-prefix-search))
5186
5187 (< (point) cfd-limit))
5188
0386b551
AM
5189 (when (and
5190 (>= (point) cfd-start-pos)
d9e94c22 5191
0386b551
AM
5192 (progn
5193 ;; Narrow to the end of the macro if we got a hit inside
5194 ;; one, to avoid recognizing things that start inside the
5195 ;; macro and end outside it.
5196 (when (> cfd-match-pos cfd-macro-end)
5197 ;; Not in the same macro as in the previous round.
5198 (save-excursion
5199 (goto-char cfd-match-pos)
5200 (setq cfd-macro-end
5201 (if (save-excursion (and (c-beginning-of-macro)
5202 (< (point) cfd-match-pos)))
5203 (progn (c-end-of-macro)
5204 (point))
5205 0))))
5206
5207 (if (zerop cfd-macro-end)
5208 t
5209 (if (> cfd-macro-end (point))
5210 (progn (narrow-to-region (point-min) cfd-macro-end)
5211 t)
5212 ;; The matched token was the last thing in the macro,
5213 ;; so the whole match is bogus.
5214 (setq cfd-macro-end 0)
5215 nil))))
d9e94c22
MS
5216
5217 (c-debug-put-decl-spot-faces cfd-match-pos (point))
0386b551
AM
5218 (if (funcall cfd-fun cfd-match-pos (/= cfd-macro-end 0))
5219 (setq cfd-prop-match nil))
d9e94c22
MS
5220
5221 (when (/= cfd-macro-end 0)
e1dbe924 5222 ;; Restore limits if we did macro narrowing above.
d9e94c22
MS
5223 (narrow-to-region (point-min) cfd-buffer-end)))
5224
5225 (goto-char cfd-continue-pos)
5226 (if (= cfd-continue-pos cfd-limit)
5227 (setq cfd-match-pos cfd-limit)
9657183b
AM
5228 (c-find-decl-prefix-search))))) ; Moves point, sets cfd-continue-pos,
5229 ; cfd-match-pos, etc.
d9e94c22
MS
5230
5231\f
5232;; A cache for found types.
5233
5234;; Buffer local variable that contains an obarray with the types we've
5235;; found. If a declaration is recognized somewhere we record the
5236;; fully qualified identifier in it to recognize it as a type
5237;; elsewhere in the file too. This is not accurate since we do not
5238;; bother with the scoping rules of the languages, but in practice the
5239;; same name is seldom used as both a type and something else in a
5240;; file, and we only use this as a last resort in ambiguous cases (see
0386b551
AM
5241;; `c-forward-decl-or-cast-1').
5242;;
580fba94
AM
5243;; Not every type need be in this cache. However, things which have
5244;; ceased to be types must be removed from it.
5245;;
0386b551
AM
5246;; Template types in C++ are added here too but with the template
5247;; arglist replaced with "<>" in references or "<" for the one in the
5248;; primary type. E.g. the type "Foo<A,B>::Bar<C>" is stored as
5249;; "Foo<>::Bar<". This avoids storing very long strings (since C++
5250;; template specs can be fairly sized programs in themselves) and
5251;; improves the hit ratio (it's a type regardless of the template
5252;; args; it's just not the same type, but we're only interested in
5253;; recognizing types, not telling distinct types apart). Note that
5254;; template types in references are added here too; from the example
5255;; above there will also be an entry "Foo<".
d9e94c22
MS
5256(defvar c-found-types nil)
5257(make-variable-buffer-local 'c-found-types)
5258
5259(defsubst c-clear-found-types ()
5260 ;; Clears `c-found-types'.
d9e94c22
MS
5261 (setq c-found-types (make-vector 53 0)))
5262
5263(defun c-add-type (from to)
5264 ;; Add the given region as a type in `c-found-types'. If the region
5265 ;; doesn't match an existing type but there is a type which is equal
5266 ;; to the given one except that the last character is missing, then
5267 ;; the shorter type is removed. That's done to avoid adding all
5268 ;; prefixes of a type as it's being entered and font locked. This
5269 ;; doesn't cover cases like when characters are removed from a type
5270 ;; or added in the middle. We'd need the position of point when the
5271 ;; font locking is invoked to solve this well.
0386b551
AM
5272 ;;
5273 ;; This function might do hidden buffer changes.
5274 (let ((type (c-syntactic-content from to c-recognize-<>-arglists)))
5275 (unless (intern-soft type c-found-types)
5276 (unintern (substring type 0 -1) c-found-types)
5277 (intern type c-found-types))))
d9e94c22 5278
580fba94
AM
5279(defun c-unfind-type (name)
5280 ;; Remove the "NAME" from c-found-types, if present.
5281 (unintern name c-found-types))
5282
d9e94c22
MS
5283(defsubst c-check-type (from to)
5284 ;; Return non-nil if the given region contains a type in
5285 ;; `c-found-types'.
0386b551
AM
5286 ;;
5287 ;; This function might do hidden buffer changes.
5288 (intern-soft (c-syntactic-content from to c-recognize-<>-arglists)
5289 c-found-types))
d9e94c22
MS
5290
5291(defun c-list-found-types ()
5292 ;; Return all the types in `c-found-types' as a sorted list of
5293 ;; strings.
5294 (let (type-list)
5295 (mapatoms (lambda (type)
5296 (setq type-list (cons (symbol-name type)
5297 type-list)))
5298 c-found-types)
5299 (sort type-list 'string-lessp)))
a66cd3ee 5300
2f42c75f
DN
5301;; Shut up the byte compiler.
5302(defvar c-maybe-stale-found-type)
5303
580fba94
AM
5304(defun c-trim-found-types (beg end old-len)
5305 ;; An after change function which, in conjunction with the info in
5306 ;; c-maybe-stale-found-type (set in c-before-change), removes a type
5307 ;; from `c-found-types', should this type have become stale. For
5308 ;; example, this happens to "foo" when "foo \n bar();" becomes
5309 ;; "foo(); \n bar();". Such stale types, if not removed, foul up
5310 ;; the fontification.
b414f371 5311 ;;
580fba94
AM
5312 ;; Have we, perhaps, added non-ws characters to the front/back of a found
5313 ;; type?
5314 (when (> end beg)
5315 (save-excursion
5316 (when (< end (point-max))
5317 (goto-char end)
5318 (if (and (c-beginning-of-current-token) ; only moves when we started in the middle
5319 (progn (goto-char end)
5320 (c-end-of-current-token)))
5321 (c-unfind-type (buffer-substring-no-properties
5322 end (point)))))
5323 (when (> beg (point-min))
5324 (goto-char beg)
5325 (if (and (c-end-of-current-token) ; only moves when we started in the middle
5326 (progn (goto-char beg)
5327 (c-beginning-of-current-token)))
5328 (c-unfind-type (buffer-substring-no-properties
5329 (point) beg))))))
b414f371 5330
580fba94
AM
5331 (if c-maybe-stale-found-type ; e.g. (c-decl-id-start "foo" 97 107 " (* ooka) " "o")
5332 (cond
5333 ;; Changing the amount of (already existing) whitespace - don't do anything.
5334 ((and (c-partial-ws-p beg end)
5335 (or (= beg end) ; removal of WS
580fba94
AM
5336 (string-match "^[ \t\n\r\f\v]*$" (nth 5 c-maybe-stale-found-type)))))
5337
5338 ;; The syntactic relationship which defined a "found type" has been
5339 ;; destroyed.
5340 ((eq (car c-maybe-stale-found-type) 'c-decl-id-start)
5341 (c-unfind-type (cadr c-maybe-stale-found-type)))
5342;; ((eq (car c-maybe-stale-found-type) 'c-decl-type-start) FIXME!!!
5343 )))
5344
d9e94c22 5345\f
dd969a56
AM
5346;; Setting and removing syntax properties on < and > in languages (C++
5347;; and Java) where they can be template/generic delimiters as well as
5348;; their normal meaning of "less/greater than".
5349
5350;; Normally, < and > have syntax 'punctuation'. When they are found to
5351;; be delimiters, they are marked as such with the category properties
5352;; c-<-as-paren-syntax, c->-as-paren-syntax respectively.
5353
5354;; STRATEGY:
5355;;
5356;; It is impossible to determine with certainty whether a <..> pair in
5357;; C++ is two comparison operators or is template delimiters, unless
5358;; one duplicates a lot of a C++ compiler. For example, the following
5359;; code fragment:
5360;;
5361;; foo (a < b, c > d) ;
5362;;
5363;; could be a function call with two integer parameters (each a
5364;; relational expression), or it could be a constructor for class foo
5365;; taking one parameter d of templated type "a < b, c >". They are
5366;; somewhat easier to distinguish in Java.
5367;;
5368;; The strategy now (2010-01) adopted is to mark and unmark < and
5369;; > IN MATCHING PAIRS ONLY. [Previously, they were marked
5370;; individually when their context so indicated. This gave rise to
53964682 5371;; intractable problems when one of a matching pair was deleted, or
dd969a56
AM
5372;; pulled into a literal.]
5373;;
5374;; At each buffer change, the syntax-table properties are removed in a
5375;; before-change function and reapplied, when needed, in an
5376;; after-change function. It is far more important that the
5377;; properties get removed when they they are spurious than that they
5378;; be present when wanted.
5379;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
5380(defun c-clear-<-pair-props (&optional pos)
5381 ;; POS (default point) is at a < character. If it is marked with
5382 ;; open paren syntax-table text property, remove the property,
5383 ;; together with the close paren property on the matching > (if
5384 ;; any).
5385 (save-excursion
5386 (if pos
5387 (goto-char pos)
5388 (setq pos (point)))
5389 (when (equal (c-get-char-property (point) 'syntax-table)
5390 c-<-as-paren-syntax)
5391 (with-syntax-table c-no-parens-syntax-table ; ignore unbalanced [,{,(,..
5392 (c-go-list-forward))
5393 (when (equal (c-get-char-property (1- (point)) 'syntax-table)
5394 c->-as-paren-syntax) ; should always be true.
8a249abc
AM
5395 (c-clear-char-property (1- (point)) 'category))
5396 (c-clear-char-property pos 'category))))
dd969a56
AM
5397
5398(defun c-clear->-pair-props (&optional pos)
5399 ;; POS (default point) is at a > character. If it is marked with
5400 ;; close paren syntax-table property, remove the property, together
5401 ;; with the open paren property on the matching < (if any).
5402 (save-excursion
5403 (if pos
5404 (goto-char pos)
5405 (setq pos (point)))
5406 (when (equal (c-get-char-property (point) 'syntax-table)
5407 c->-as-paren-syntax)
5408 (with-syntax-table c-no-parens-syntax-table ; ignore unbalanced [,{,(,..
5409 (c-go-up-list-backward))
5410 (when (equal (c-get-char-property (point) 'syntax-table)
5411 c-<-as-paren-syntax) ; should always be true.
8a249abc
AM
5412 (c-clear-char-property (point) 'category))
5413 (c-clear-char-property pos 'category))))
dd969a56
AM
5414
5415(defun c-clear-<>-pair-props (&optional pos)
5416 ;; POS (default point) is at a < or > character. If it has an
5417 ;; open/close paren syntax-table property, remove this property both
5418 ;; from the current character and its partner (which will also be
5419 ;; thusly marked).
5420 (cond
5421 ((eq (char-after) ?\<)
5422 (c-clear-<-pair-props pos))
5423 ((eq (char-after) ?\>)
5424 (c-clear->-pair-props pos))
5425 (t (c-benign-error
5426 "c-clear-<>-pair-props called from wrong position"))))
5427
5428(defun c-clear-<-pair-props-if-match-after (lim &optional pos)
5429 ;; POS (default point) is at a < character. If it is both marked
5430 ;; with open/close paren syntax-table property, and has a matching >
5431 ;; (also marked) which is after LIM, remove the property both from
43a91810
AM
5432 ;; the current > and its partner. Return t when this happens, nil
5433 ;; when it doesn't.
dd969a56
AM
5434 (save-excursion
5435 (if pos
5436 (goto-char pos)
5437 (setq pos (point)))
5438 (when (equal (c-get-char-property (point) 'syntax-table)
5439 c-<-as-paren-syntax)
5440 (with-syntax-table c-no-parens-syntax-table ; ignore unbalanced [,{,(,..
5441 (c-go-list-forward))
5442 (when (and (>= (point) lim)
5443 (equal (c-get-char-property (1- (point)) 'syntax-table)
5444 c->-as-paren-syntax)) ; should always be true.
5445 (c-unmark-<->-as-paren (1- (point)))
43a91810
AM
5446 (c-unmark-<->-as-paren pos))
5447 t)))
dd969a56
AM
5448
5449(defun c-clear->-pair-props-if-match-before (lim &optional pos)
5450 ;; POS (default point) is at a > character. If it is both marked
5451 ;; with open/close paren syntax-table property, and has a matching <
5452 ;; (also marked) which is before LIM, remove the property both from
43a91810
AM
5453 ;; the current < and its partner. Return t when this happens, nil
5454 ;; when it doesn't.
dd969a56
AM
5455 (save-excursion
5456 (if pos
5457 (goto-char pos)
5458 (setq pos (point)))
5459 (when (equal (c-get-char-property (point) 'syntax-table)
5460 c->-as-paren-syntax)
5461 (with-syntax-table c-no-parens-syntax-table ; ignore unbalanced [,{,(,..
5462 (c-go-up-list-backward))
5463 (when (and (<= (point) lim)
5464 (equal (c-get-char-property (point) 'syntax-table)
5465 c-<-as-paren-syntax)) ; should always be true.
5466 (c-unmark-<->-as-paren (point))
43a91810
AM
5467 (c-unmark-<->-as-paren pos))
5468 t)))
dd969a56 5469
f3b554af
GM
5470;; Set by c-common-init in cc-mode.el.
5471(defvar c-new-BEG)
5472(defvar c-new-END)
5473
dd969a56
AM
5474(defun c-before-change-check-<>-operators (beg end)
5475 ;; Unmark certain pairs of "< .... >" which are currently marked as
5476 ;; template/generic delimiters. (This marking is via syntax-table
5477 ;; text properties).
5478 ;;
5479 ;; These pairs are those which are in the current "statement" (i.e.,
5480 ;; the region between the {, }, or ; before BEG and the one after
5481 ;; END), and which enclose any part of the interval (BEG END).
5482 ;;
5483 ;; Note that in C++ (?and Java), template/generic parens cannot
5484 ;; enclose a brace or semicolon, so we use these as bounds on the
5485 ;; region we must work on.
5486 ;;
5487 ;; This function is called from before-change-functions (via
5488 ;; c-get-state-before-change-functions). Thus the buffer is widened,
5489 ;; and point is undefined, both at entry and exit.
5490 ;;
5491 ;; FIXME!!! This routine ignores the possibility of macros entirely.
5492 ;; 2010-01-29.
5493 (save-excursion
5494 (let ((beg-lit-limits (progn (goto-char beg) (c-literal-limits)))
43a91810
AM
5495 (end-lit-limits (progn (goto-char end) (c-literal-limits)))
5496 new-beg new-end need-new-beg need-new-end)
dd969a56
AM
5497 ;; Locate the barrier before the changed region
5498 (goto-char (if beg-lit-limits (car beg-lit-limits) beg))
0de3da9f 5499 (c-syntactic-skip-backward "^;{}" (c-determine-limit 512))
43a91810 5500 (setq new-beg (point))
dd969a56
AM
5501
5502 ;; Remove the syntax-table properties from each pertinent <...> pair.
5503 ;; Firsly, the ones with the < before beg and > after beg.
5504 (while (c-search-forward-char-property 'category 'c-<-as-paren-syntax beg)
43a91810
AM
5505 (if (c-clear-<-pair-props-if-match-after beg (1- (point)))
5506 (setq need-new-beg t)))
dd969a56
AM
5507
5508 ;; Locate the barrier after END.
5509 (goto-char (if end-lit-limits (cdr end-lit-limits) end))
0de3da9f 5510 (c-syntactic-re-search-forward "[;{}]" (c-determine-+ve-limit 512) 'end)
43a91810 5511 (setq new-end (point))
dd969a56
AM
5512
5513 ;; Remove syntax-table properties from the remaining pertinent <...>
5514 ;; pairs, those with a > after end and < before end.
5515 (while (c-search-backward-char-property 'category 'c->-as-paren-syntax end)
43a91810
AM
5516 (if (c-clear->-pair-props-if-match-before end)
5517 (setq need-new-end t)))
5518
5519 ;; Extend the fontification region, if needed.
5520 (when need-new-beg
5521 (goto-char new-beg)
5522 (c-forward-syntactic-ws)
5523 (and (< (point) c-new-BEG) (setq c-new-BEG (point))))
5524
5525 (when need-new-end
5526 (and (> new-end c-new-END) (setq c-new-END new-end))))))
dd969a56
AM
5527
5528
d9e94c22 5529
0386b551
AM
5530(defun c-after-change-check-<>-operators (beg end)
5531 ;; This is called from `after-change-functions' when
5532 ;; c-recognize-<>-arglists' is set. It ensures that no "<" or ">"
5533 ;; chars with paren syntax become part of another operator like "<<"
5534 ;; or ">=".
5535 ;;
5536 ;; This function might do hidden buffer changes.
5537
cb694ab7
AM
5538 (save-excursion
5539 (goto-char beg)
5540 (when (or (looking-at "[<>]")
5541 (< (skip-chars-backward "<>") 0))
5542
0386b551 5543 (goto-char beg)
cb694ab7
AM
5544 (c-beginning-of-current-token)
5545 (when (and (< (point) beg)
5546 (looking-at c-<>-multichar-token-regexp)
5547 (< beg (setq beg (match-end 0))))
5548 (while (progn (skip-chars-forward "^<>" beg)
5549 (< (point) beg))
dd969a56 5550 (c-clear-<>-pair-props)
cb694ab7
AM
5551 (forward-char))))
5552
5553 (when (< beg end)
5554 (goto-char end)
0386b551
AM
5555 (when (or (looking-at "[<>]")
5556 (< (skip-chars-backward "<>") 0))
5557
cb694ab7 5558 (goto-char end)
0386b551 5559 (c-beginning-of-current-token)
cb694ab7 5560 (when (and (< (point) end)
0386b551 5561 (looking-at c-<>-multichar-token-regexp)
cb694ab7
AM
5562 (< end (setq end (match-end 0))))
5563 (while (progn (skip-chars-forward "^<>" end)
5564 (< (point) end))
dd969a56 5565 (c-clear-<>-pair-props)
cb694ab7 5566 (forward-char)))))))
d9e94c22 5567
dd969a56
AM
5568
5569\f
5570;; Handling of small scale constructs like types and names.
5571
d9e94c22
MS
5572;; Dynamically bound variable that instructs `c-forward-type' to also
5573;; treat possible types (i.e. those that it normally returns 'maybe or
5574;; 'found for) as actual types (and always return 'found for them).
5575;; This means that it records them in `c-record-type-identifiers' if
5576;; that is set, and that it adds them to `c-found-types'.
5577(defvar c-promote-possible-types nil)
5578
0386b551
AM
5579;; Dynamically bound variable that instructs `c-forward-<>-arglist' to
5580;; mark up successfully parsed arglists with paren syntax properties on
5581;; the surrounding angle brackets and with `c-<>-arg-sep' in the
5582;; `c-type' property of each argument separating comma.
5583;;
5584;; Setting this variable also makes `c-forward-<>-arglist' recurse into
5585;; all arglists for side effects (i.e. recording types), otherwise it
5586;; exploits any existing paren syntax properties to quickly jump to the
5587;; end of already parsed arglists.
5588;;
5589;; Marking up the arglists is not the default since doing that correctly
5590;; depends on a proper value for `c-restricted-<>-arglists'.
5591(defvar c-parse-and-markup-<>-arglists nil)
5592
d9e94c22 5593;; Dynamically bound variable that instructs `c-forward-<>-arglist' to
037558bf
MS
5594;; not accept arglists that contain binary operators.
5595;;
5596;; This is primarily used to handle C++ template arglists. C++
5597;; disambiguates them by checking whether the preceding name is a
5598;; template or not. We can't do that, so we assume it is a template
5599;; if it can be parsed as one. That usually works well since
5600;; comparison expressions on the forms "a < b > c" or "a < b, c > d"
5601;; in almost all cases would be pointless.
5602;;
5603;; However, in function arglists, e.g. in "foo (a < b, c > d)", we
5604;; should let the comma separate the function arguments instead. And
5605;; in a context where the value of the expression is taken, e.g. in
5606;; "if (a < b || c > d)", it's probably not a template.
5607(defvar c-restricted-<>-arglists nil)
d9e94c22 5608
0386b551
AM
5609;; Dynamically bound variables that instructs
5610;; `c-forward-keyword-clause', `c-forward-<>-arglist',
5611;; `c-forward-name', `c-forward-type', `c-forward-decl-or-cast-1', and
5612;; `c-forward-label' to record the ranges of all the type and
5613;; reference identifiers they encounter. They will build lists on
5614;; these variables where each element is a cons of the buffer
5615;; positions surrounding each identifier. This recording is only
5616;; activated when `c-record-type-identifiers' is non-nil.
d9e94c22
MS
5617;;
5618;; All known types that can't be identifiers are recorded, and also
5619;; other possible types if `c-promote-possible-types' is set.
5620;; Recording is however disabled inside angle bracket arglists that
5621;; are encountered inside names and other angle bracket arglists.
0386b551 5622;; Such occurrences are taken care of by `c-font-lock-<>-arglists'
d9e94c22
MS
5623;; instead.
5624;;
5625;; Only the names in C++ template style references (e.g. "tmpl" in
5626;; "tmpl<a,b>::foo") are recorded as references, other references
5627;; aren't handled here.
0386b551
AM
5628;;
5629;; `c-forward-label' records the label identifier(s) on
5630;; `c-record-ref-identifiers'.
d9e94c22
MS
5631(defvar c-record-type-identifiers nil)
5632(defvar c-record-ref-identifiers nil)
5633
0386b551
AM
5634;; This variable will receive a cons cell of the range of the last
5635;; single identifier symbol stepped over by `c-forward-name' if it's
5636;; successful. This is the range that should be put on one of the
5637;; record lists above by the caller. It's assigned nil if there's no
5638;; such symbol in the name.
d9e94c22
MS
5639(defvar c-last-identifier-range nil)
5640
5641(defmacro c-record-type-id (range)
5642 (if (eq (car-safe range) 'cons)
5643 ;; Always true.
5644 `(setq c-record-type-identifiers
5645 (cons ,range c-record-type-identifiers))
5646 `(let ((range ,range))
5647 (if range
5648 (setq c-record-type-identifiers
5649 (cons range c-record-type-identifiers))))))
5650
5651(defmacro c-record-ref-id (range)
5652 (if (eq (car-safe range) 'cons)
5653 ;; Always true.
5654 `(setq c-record-ref-identifiers
5655 (cons ,range c-record-ref-identifiers))
5656 `(let ((range ,range))
5657 (if range
5658 (setq c-record-ref-identifiers
5659 (cons range c-record-ref-identifiers))))))
5660
5661;; Dynamically bound variable that instructs `c-forward-type' to
5662;; record the ranges of types that only are found. Behaves otherwise
5663;; like `c-record-type-identifiers'.
5664(defvar c-record-found-types nil)
5665
5666(defmacro c-forward-keyword-prefixed-id (type)
5667 ;; Used internally in `c-forward-keyword-clause' to move forward
5668 ;; over a type (if TYPE is 'type) or a name (otherwise) which
5669 ;; possibly is prefixed by keywords and their associated clauses.
5670 ;; Try with a type/name first to not trip up on those that begin
5671 ;; with a keyword. Return t if a known or found type is moved
5672 ;; over. The point is clobbered if nil is returned. If range
5673 ;; recording is enabled, the identifier is recorded on as a type
5674 ;; if TYPE is 'type or as a reference if TYPE is 'ref.
0386b551
AM
5675 ;;
5676 ;; This macro might do hidden buffer changes.
d9e94c22
MS
5677 `(let (res)
5678 (while (if (setq res ,(if (eq type 'type)
5679 `(c-forward-type)
5680 `(c-forward-name)))
5681 nil
5682 (and (looking-at c-keywords-regexp)
0386b551 5683 (c-forward-keyword-clause 1))))
d9e94c22
MS
5684 (when (memq res '(t known found prefix))
5685 ,(when (eq type 'ref)
5686 `(when c-record-type-identifiers
5687 (c-record-ref-id c-last-identifier-range)))
5688 t)))
5689
0386b551 5690(defmacro c-forward-id-comma-list (type update-safe-pos)
d9e94c22
MS
5691 ;; Used internally in `c-forward-keyword-clause' to move forward
5692 ;; over a comma separated list of types or names using
5693 ;; `c-forward-keyword-prefixed-id'.
0386b551
AM
5694 ;;
5695 ;; This macro might do hidden buffer changes.
d9e94c22 5696 `(while (and (progn
0386b551
AM
5697 ,(when update-safe-pos
5698 `(setq safe-pos (point)))
d9e94c22
MS
5699 (eq (char-after) ?,))
5700 (progn
5701 (forward-char)
5702 (c-forward-syntactic-ws)
5703 (c-forward-keyword-prefixed-id ,type)))))
5704
0386b551
AM
5705(defun c-forward-keyword-clause (match)
5706 ;; Submatch MATCH in the current match data is assumed to surround a
5707 ;; token. If it's a keyword, move over it and any immediately
5708 ;; following clauses associated with it, stopping at the start of
5709 ;; the next token. t is returned in that case, otherwise the point
d9e94c22
MS
5710 ;; stays and nil is returned. The kind of clauses that are
5711 ;; recognized are those specified by `c-type-list-kwds',
5712 ;; `c-ref-list-kwds', `c-colon-type-list-kwds',
5713 ;; `c-paren-nontype-kwds', `c-paren-type-kwds', `c-<>-type-kwds',
5714 ;; and `c-<>-arglist-kwds'.
0386b551
AM
5715 ;;
5716 ;; This function records identifier ranges on
5717 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
5718 ;; `c-record-type-identifiers' is non-nil.
5719 ;;
5720 ;; Note that for `c-colon-type-list-kwds', which doesn't necessary
5721 ;; apply directly after the keyword, the type list is moved over
5722 ;; only when there is no unaccounted token before it (i.e. a token
5723 ;; that isn't moved over due to some other keyword list). The
5724 ;; identifier ranges in the list are still recorded if that should
5725 ;; be done, though.
5726 ;;
5727 ;; This function might do hidden buffer changes.
5728
5729 (let ((kwd-sym (c-keyword-sym (match-string match))) safe-pos pos
5730 ;; The call to `c-forward-<>-arglist' below is made after
5731 ;; `c-<>-sexp-kwds' keywords, so we're certain they actually
5732 ;; are angle bracket arglists and `c-restricted-<>-arglists'
5733 ;; should therefore be nil.
5734 (c-parse-and-markup-<>-arglists t)
5735 c-restricted-<>-arglists)
d9e94c22 5736
d9e94c22 5737 (when kwd-sym
0386b551 5738 (goto-char (match-end match))
d9e94c22
MS
5739 (c-forward-syntactic-ws)
5740 (setq safe-pos (point))
a66cd3ee 5741
d9e94c22
MS
5742 (cond
5743 ((and (c-keyword-member kwd-sym 'c-type-list-kwds)
5744 (c-forward-keyword-prefixed-id type))
5745 ;; There's a type directly after a keyword in `c-type-list-kwds'.
0386b551 5746 (c-forward-id-comma-list type t))
d9e94c22
MS
5747
5748 ((and (c-keyword-member kwd-sym 'c-ref-list-kwds)
5749 (c-forward-keyword-prefixed-id ref))
5750 ;; There's a name directly after a keyword in `c-ref-list-kwds'.
0386b551 5751 (c-forward-id-comma-list ref t))
d9e94c22
MS
5752
5753 ((and (c-keyword-member kwd-sym 'c-paren-any-kwds)
5754 (eq (char-after) ?\())
5755 ;; There's an open paren after a keyword in `c-paren-any-kwds'.
5756
5757 (forward-char)
5758 (when (and (setq pos (c-up-list-forward))
5759 (eq (char-before pos) ?\)))
5760 (when (and c-record-type-identifiers
5761 (c-keyword-member kwd-sym 'c-paren-type-kwds))
5762 ;; Use `c-forward-type' on every identifier we can find
5763 ;; inside the paren, to record the types.
5764 (while (c-syntactic-re-search-forward c-symbol-start pos t)
5765 (goto-char (match-beginning 0))
5766 (unless (c-forward-type)
5767 (looking-at c-symbol-key) ; Always matches.
5768 (goto-char (match-end 0)))))
5769
5770 (goto-char pos)
5771 (c-forward-syntactic-ws)
5772 (setq safe-pos (point))))
5773
5774 ((and (c-keyword-member kwd-sym 'c-<>-sexp-kwds)
5775 (eq (char-after) ?<)
0386b551 5776 (c-forward-<>-arglist (c-keyword-member kwd-sym 'c-<>-type-kwds)))
d9e94c22
MS
5777 (c-forward-syntactic-ws)
5778 (setq safe-pos (point)))
5779
5780 ((and (c-keyword-member kwd-sym 'c-nonsymbol-sexp-kwds)
449a2b0d
MS
5781 (not (looking-at c-symbol-start))
5782 (c-safe (c-forward-sexp) t))
d9e94c22
MS
5783 (c-forward-syntactic-ws)
5784 (setq safe-pos (point))))
5785
0386b551
AM
5786 (when (c-keyword-member kwd-sym 'c-colon-type-list-kwds)
5787 (if (eq (char-after) ?:)
5788 ;; If we are at the colon already, we move over the type
5789 ;; list after it.
5790 (progn
5791 (forward-char)
5792 (c-forward-syntactic-ws)
5793 (when (c-forward-keyword-prefixed-id type)
5794 (c-forward-id-comma-list type t)))
5795 ;; Not at the colon, so stop here. But the identifier
5796 ;; ranges in the type list later on should still be
5797 ;; recorded.
5798 (and c-record-type-identifiers
5799 (progn
5800 ;; If a keyword matched both one of the types above and
5801 ;; this one, we match `c-colon-type-list-re' after the
5802 ;; clause matched above.
5803 (goto-char safe-pos)
5804 (looking-at c-colon-type-list-re))
5805 (progn
5806 (goto-char (match-end 0))
5807 (c-forward-syntactic-ws)
5808 (c-forward-keyword-prefixed-id type))
5809 ;; There's a type after the `c-colon-type-list-re' match
5810 ;; after a keyword in `c-colon-type-list-kwds'.
5811 (c-forward-id-comma-list type nil))))
d9e94c22
MS
5812
5813 (goto-char safe-pos)
5814 t)))
5815
f3b554af
GM
5816;; cc-mode requires cc-fonts.
5817(declare-function c-fontify-recorded-types-and-refs "cc-fonts" ())
5818
0386b551
AM
5819(defun c-forward-<>-arglist (all-types)
5820 ;; The point is assumed to be at a "<". Try to treat it as the open
b4dc7d98 5821 ;; paren of an angle bracket arglist and move forward to the
0386b551
AM
5822 ;; corresponding ">". If successful, the point is left after the
5823 ;; ">" and t is returned, otherwise the point isn't moved and nil is
d9e94c22
MS
5824 ;; returned. If ALL-TYPES is t then all encountered arguments in
5825 ;; the arglist that might be types are treated as found types.
5826 ;;
0386b551
AM
5827 ;; The variable `c-parse-and-markup-<>-arglists' controls how this
5828 ;; function handles text properties on the angle brackets and argument
5829 ;; separating commas.
d9e94c22 5830 ;;
0386b551
AM
5831 ;; `c-restricted-<>-arglists' controls how lenient the template
5832 ;; arglist recognition should be.
5833 ;;
5834 ;; This function records identifier ranges on
5835 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
5836 ;; `c-record-type-identifiers' is non-nil.
5837 ;;
5838 ;; This function might do hidden buffer changes.
d9e94c22
MS
5839
5840 (let ((start (point))
5841 ;; If `c-record-type-identifiers' is set then activate
5842 ;; recording of any found types that constitute an argument in
5843 ;; the arglist.
abfc152b 5844 (c-record-found-types (if c-record-type-identifiers t)))
d9e94c22
MS
5845 (if (catch 'angle-bracket-arglist-escape
5846 (setq c-record-found-types
0386b551 5847 (c-forward-<>-arglist-recur all-types)))
d9e94c22
MS
5848 (progn
5849 (when (consp c-record-found-types)
5850 (setq c-record-type-identifiers
5851 ;; `nconc' doesn't mind that the tail of
5852 ;; `c-record-found-types' is t.
5853 (nconc c-record-found-types c-record-type-identifiers)))
452ea855 5854 (if (c-major-mode-is 'java-mode) (c-fontify-recorded-types-and-refs))
d9e94c22
MS
5855 t)
5856
5857 (goto-char start)
a66cd3ee 5858 nil)))
785eecbb 5859
0386b551 5860(defun c-forward-<>-arglist-recur (all-types)
d9e94c22 5861 ;; Recursive part of `c-forward-<>-arglist'.
0386b551
AM
5862 ;;
5863 ;; This function might do hidden buffer changes.
d9e94c22
MS
5864
5865 (let ((start (point)) res pos tmp
5866 ;; Cover this so that any recorded found type ranges are
5867 ;; automatically lost if it turns out to not be an angle
5868 ;; bracket arglist. It's propagated through the return value
5869 ;; on successful completion.
5870 (c-record-found-types c-record-found-types)
5871 ;; List that collects the positions after the argument
5872 ;; separating ',' in the arglist.
5873 arg-start-pos)
0386b551
AM
5874 ;; If the '<' has paren open syntax then we've marked it as an angle
5875 ;; bracket arglist before, so skip to the end.
5876 (if (and (not c-parse-and-markup-<>-arglists)
5877 (c-get-char-property (point) 'syntax-table))
5878
5879 (progn
5880 (forward-char)
5881 (if (and (c-go-up-list-forward)
5882 (eq (char-before) ?>))
5883 t
0386b551
AM
5884 ;; Got unmatched paren angle brackets. We don't clear the paren
5885 ;; syntax properties and retry, on the basis that it's very
5886 ;; unlikely that paren angle brackets become operators by code
5887 ;; manipulation. It's far more likely that it doesn't match due
5888 ;; to narrowing or some temporary change.
5889 (goto-char start)
5890 nil))
d9e94c22 5891
abfc152b 5892 (forward-char) ; Forward over the opening '<'.
452ea855 5893
d9e94c22 5894 (unless (looking-at c-<-op-cont-regexp)
abfc152b
AM
5895 ;; go forward one non-alphanumeric character (group) per iteration of
5896 ;; this loop.
22c3ce97 5897 (while (and
d9e94c22 5898 (progn
22c3ce97
AM
5899 (c-forward-syntactic-ws)
5900 (let ((orig-record-found-types c-record-found-types))
5901 (when (or (and c-record-type-identifiers all-types)
5902 (c-major-mode-is 'java-mode))
5903 ;; All encountered identifiers are types, so set the
5904 ;; promote flag and parse the type.
5905 (progn
5906 (c-forward-syntactic-ws)
5907 (if (looking-at "\\?")
5908 (forward-char)
5909 (when (looking-at c-identifier-start)
5910 (let ((c-promote-possible-types t)
5911 (c-record-found-types t))
5912 (c-forward-type))))
5913
5914 (c-forward-syntactic-ws)
5915
5916 (when (or (looking-at "extends")
5917 (looking-at "super"))
5918 (forward-word)
5919 (c-forward-syntactic-ws)
5920 (let ((c-promote-possible-types t)
5921 (c-record-found-types t))
5922 (c-forward-type)
5923 (c-forward-syntactic-ws))))))
5924
abfc152b 5925 (setq pos (point)) ; e.g. first token inside the '<'
22c3ce97 5926
a4ee83cc
AM
5927 ;; Note: These regexps exploit the match order in \| so
5928 ;; that "<>" is matched by "<" rather than "[^>:-]>".
5929 (c-syntactic-re-search-forward
5930 ;; Stop on ',', '|', '&', '+' and '-' to catch
5931 ;; common binary operators that could be between
5932 ;; two comparison expressions "a<b" and "c>d".
5933 "[<;{},|+&-]\\|[>)]"
5934 nil t t))
22c3ce97
AM
5935
5936 (cond
5937 ((eq (char-before) ?>)
d9e94c22
MS
5938 ;; Either an operator starting with '>' or the end of
5939 ;; the angle bracket arglist.
5940
0386b551 5941 (if (looking-at c->-op-cont-regexp)
d9e94c22 5942 (progn
0386b551
AM
5943 (goto-char (match-end 0))
5944 t) ; Continue the loop.
d9e94c22 5945
0386b551
AM
5946 ;; The angle bracket arglist is finished.
5947 (when c-parse-and-markup-<>-arglists
d9e94c22 5948 (while arg-start-pos
0386b551
AM
5949 (c-put-c-type-property (1- (car arg-start-pos))
5950 'c-<>-arg-sep)
d9e94c22
MS
5951 (setq arg-start-pos (cdr arg-start-pos)))
5952 (c-mark-<-as-paren start)
0386b551
AM
5953 (c-mark->-as-paren (1- (point))))
5954 (setq res t)
5955 nil)) ; Exit the loop.
d9e94c22
MS
5956
5957 ((eq (char-before) ?<)
5958 ;; Either an operator starting with '<' or a nested arglist.
d9e94c22
MS
5959 (setq pos (point))
5960 (let (id-start id-end subres keyword-match)
abfc152b
AM
5961 (cond
5962 ;; The '<' begins a multi-char operator.
5963 ((looking-at c-<-op-cont-regexp)
5964 (setq tmp (match-end 0))
5965 (goto-char (match-end 0)))
5966 ;; We're at a nested <.....>
5967 ((progn
5968 (setq tmp pos)
5969 (backward-char) ; to the '<'
5970 (and
5971 (save-excursion
5972 ;; There's always an identifier before an angle
5973 ;; bracket arglist, or a keyword in `c-<>-type-kwds'
5974 ;; or `c-<>-arglist-kwds'.
5975 (c-backward-syntactic-ws)
5976 (setq id-end (point))
5977 (c-simple-skip-symbol-backward)
5978 (when (or (setq keyword-match
5979 (looking-at c-opt-<>-sexp-key))
5980 (not (looking-at c-keywords-regexp)))
5981 (setq id-start (point))))
5982 (setq subres
5983 (let ((c-promote-possible-types t)
5984 (c-record-found-types t))
5985 (c-forward-<>-arglist-recur
5986 (and keyword-match
5987 (c-keyword-member
5988 (c-keyword-sym (match-string 1))
5989 'c-<>-type-kwds)))))))
d9e94c22
MS
5990
5991 ;; It was an angle bracket arglist.
5992 (setq c-record-found-types subres)
5993
5994 ;; Record the identifier before the template as a type
5995 ;; or reference depending on whether the arglist is last
5996 ;; in a qualified identifier.
5997 (when (and c-record-type-identifiers
5998 (not keyword-match))
5999 (if (and c-opt-identifier-concat-key
6000 (progn
6001 (c-forward-syntactic-ws)
6002 (looking-at c-opt-identifier-concat-key)))
6003 (c-record-ref-id (cons id-start id-end))
abfc152b
AM
6004 (c-record-type-id (cons id-start id-end)))))
6005
6006 ;; At a "less than" operator.
6007 (t
6008 (forward-char)
6009 )))
6010 t) ; carry on looping.
22c3ce97
AM
6011
6012 ((and (not c-restricted-<>-arglists)
6013 (or (and (eq (char-before) ?&)
6014 (not (eq (char-after) ?&)))
6015 (eq (char-before) ?,)))
6016 ;; Just another argument. Record the position. The
6017 ;; type check stuff that made us stop at it is at
6018 ;; the top of the loop.
d9e94c22
MS
6019 (setq arg-start-pos (cons (point) arg-start-pos)))
6020
6021 (t
6022 ;; Got a character that can't be in an angle bracket
6023 ;; arglist argument. Abort using `throw', since
6024 ;; it's useless to try to find a surrounding arglist
6025 ;; if we're nested.
6026 (throw 'angle-bracket-arglist-escape nil))))))
d9e94c22
MS
6027 (if res
6028 (or c-record-found-types t)))))
6029
0386b551
AM
6030(defun c-backward-<>-arglist (all-types &optional limit)
6031 ;; The point is assumed to be directly after a ">". Try to treat it
6032 ;; as the close paren of an angle bracket arglist and move back to
6033 ;; the corresponding "<". If successful, the point is left at
6034 ;; the "<" and t is returned, otherwise the point isn't moved and
6035 ;; nil is returned. ALL-TYPES is passed on to
6036 ;; `c-forward-<>-arglist'.
6037 ;;
6038 ;; If the optional LIMIT is given, it bounds the backward search.
6039 ;; It's then assumed to be at a syntactically relevant position.
6040 ;;
6041 ;; This is a wrapper around `c-forward-<>-arglist'. See that
6042 ;; function for more details.
6043
6044 (let ((start (point)))
6045 (backward-char)
6046 (if (and (not c-parse-and-markup-<>-arglists)
6047 (c-get-char-property (point) 'syntax-table))
6048
6049 (if (and (c-go-up-list-backward)
6050 (eq (char-after) ?<))
6051 t
6052 ;; See corresponding note in `c-forward-<>-arglist'.
6053 (goto-char start)
6054 nil)
6055
51c9af45 6056 (while (progn
0386b551
AM
6057 (c-syntactic-skip-backward "^<;{}" limit t)
6058
51c9af45
AM
6059 (and
6060 (if (eq (char-before) ?<)
6061 t
6062 ;; Stopped at bob or a char that isn't allowed in an
6063 ;; arglist, so we've failed.
6064 (goto-char start)
6065 nil)
0386b551 6066
51c9af45
AM
6067 (if (> (point)
6068 (progn (c-beginning-of-current-token)
6069 (point)))
6070 ;; If we moved then the "<" was part of some
6071 ;; multicharacter token.
6072 t
0386b551 6073
51c9af45
AM
6074 (backward-char)
6075 (let ((beg-pos (point)))
6076 (if (c-forward-<>-arglist all-types)
6077 (cond ((= (point) start)
6078 ;; Matched the arglist. Break the while.
6079 (goto-char beg-pos)
6080 nil)
6081 ((> (point) start)
6082 ;; We started from a non-paren ">" inside an
6083 ;; arglist.
6084 (goto-char start)
6085 nil)
6086 (t
6087 ;; Matched a shorter arglist. Can be a nested
6088 ;; one so continue looking.
6089 (goto-char beg-pos)
6090 t))
6091 t))))))
0386b551
AM
6092
6093 (/= (point) start))))
6094
d9e94c22
MS
6095(defun c-forward-name ()
6096 ;; Move forward over a complete name if at the beginning of one,
e15f8aaa
AM
6097 ;; stopping at the next following token. A keyword, as such,
6098 ;; doesn't count as a name. If the point is not at something that
6099 ;; is recognized as a name then it stays put.
6100 ;;
6101 ;; A name could be something as simple as "foo" in C or something as
d9e94c22
MS
6102 ;; complex as "X<Y<class A<int>::B, BIT_MAX >> b>, ::operator<> ::
6103 ;; Z<(a>b)> :: operator const X<&foo>::T Q::G<unsigned short
6104 ;; int>::*volatile const" in C++ (this function is actually little
6105 ;; more than a `looking-at' call in all modes except those that,
e15f8aaa
AM
6106 ;; like C++, have `c-recognize-<>-arglists' set).
6107 ;;
6108 ;; Return
6109 ;; o - nil if no name is found;
6110 ;; o - 'template if it's an identifier ending with an angle bracket
6111 ;; arglist;
6112 ;; o - 'operator of it's an operator identifier;
6113 ;; o - t if it's some other kind of name.
0386b551
AM
6114 ;;
6115 ;; This function records identifier ranges on
6116 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
6117 ;; `c-record-type-identifiers' is non-nil.
6118 ;;
6119 ;; This function might do hidden buffer changes.
d9e94c22 6120
0386b551 6121 (let ((pos (point)) (start (point)) res id-start id-end
d9e94c22
MS
6122 ;; Turn off `c-promote-possible-types' here since we might
6123 ;; call `c-forward-<>-arglist' and we don't want it to promote
6124 ;; every suspect thing in the arglist to a type. We're
6125 ;; typically called from `c-forward-type' in this case, and
6126 ;; the caller only wants the top level type that it finds to
6127 ;; be promoted.
6128 c-promote-possible-types)
6129 (while
6130 (and
6131 (looking-at c-identifier-key)
6132
6133 (progn
6134 ;; Check for keyword. We go to the last symbol in
6135 ;; `c-identifier-key' first.
0386b551
AM
6136 (goto-char (setq id-end (match-end 0)))
6137 (c-simple-skip-symbol-backward)
6138 (setq id-start (point))
d9e94c22
MS
6139
6140 (if (looking-at c-keywords-regexp)
6141 (when (and (c-major-mode-is 'c++-mode)
6142 (looking-at
6143 (cc-eval-when-compile
6144 (concat "\\(operator\\|\\(template\\)\\)"
6145 "\\(" (c-lang-const c-nonsymbol-key c++)
6146 "\\|$\\)")))
6147 (if (match-beginning 2)
6148 ;; "template" is only valid inside an
6149 ;; identifier if preceded by "::".
6150 (save-excursion
6151 (c-backward-syntactic-ws)
6152 (and (c-safe (backward-char 2) t)
6153 (looking-at "::")))
6154 t))
6155
6156 ;; Handle a C++ operator or template identifier.
6157 (goto-char id-end)
6158 (c-forward-syntactic-ws)
6159 (cond ((eq (char-before id-end) ?e)
6160 ;; Got "... ::template".
6161 (let ((subres (c-forward-name)))
6162 (when subres
6163 (setq pos (point)
6164 res subres))))
6165
6166 ((looking-at c-identifier-start)
6167 ;; Got a cast operator.
6168 (when (c-forward-type)
6169 (setq pos (point)
6170 res 'operator)
6171 ;; Now we should match a sequence of either
6172 ;; '*', '&' or a name followed by ":: *",
6173 ;; where each can be followed by a sequence
6174 ;; of `c-opt-type-modifier-key'.
6175 (while (cond ((looking-at "[*&]")
6176 (goto-char (match-end 0))
6177 t)
6178 ((looking-at c-identifier-start)
6179 (and (c-forward-name)
6180 (looking-at "::")
6181 (progn
6182 (goto-char (match-end 0))
6183 (c-forward-syntactic-ws)
6184 (eq (char-after) ?*))
6185 (progn
6186 (forward-char)
6187 t))))
6188 (while (progn
6189 (c-forward-syntactic-ws)
6190 (setq pos (point))
6191 (looking-at c-opt-type-modifier-key))
6192 (goto-char (match-end 1))))))
6193
6194 ((looking-at c-overloadable-operators-regexp)
6195 ;; Got some other operator.
0386b551
AM
6196 (setq c-last-identifier-range
6197 (cons (point) (match-end 0)))
d9e94c22
MS
6198 (goto-char (match-end 0))
6199 (c-forward-syntactic-ws)
6200 (setq pos (point)
6201 res 'operator)))
6202
6203 nil)
6204
0386b551
AM
6205 ;; `id-start' is equal to `id-end' if we've jumped over
6206 ;; an identifier that doesn't end with a symbol token.
6207 ;; That can occur e.g. for Java import directives on the
6208 ;; form "foo.bar.*".
6209 (when (and id-start (/= id-start id-end))
d9e94c22
MS
6210 (setq c-last-identifier-range
6211 (cons id-start id-end)))
6212 (goto-char id-end)
6213 (c-forward-syntactic-ws)
6214 (setq pos (point)
6215 res t)))
6216
6217 (progn
6218 (goto-char pos)
6219 (when (or c-opt-identifier-concat-key
6220 c-recognize-<>-arglists)
6221
6222 (cond
6223 ((and c-opt-identifier-concat-key
6224 (looking-at c-opt-identifier-concat-key))
6225 ;; Got a concatenated identifier. This handles the
6226 ;; cases with tricky syntactic whitespace that aren't
6227 ;; covered in `c-identifier-key'.
6228 (goto-char (match-end 0))
6229 (c-forward-syntactic-ws)
6230 t)
6231
6232 ((and c-recognize-<>-arglists
6233 (eq (char-after) ?<))
6234 ;; Maybe an angle bracket arglist.
452ea855
AM
6235 (when (let ((c-record-type-identifiers t)
6236 (c-record-found-types t))
0386b551
AM
6237 (c-forward-<>-arglist nil))
6238
6239 (c-add-type start (1+ pos))
d9e94c22 6240 (c-forward-syntactic-ws)
0386b551
AM
6241 (setq pos (point)
6242 c-last-identifier-range nil)
6243
d9e94c22
MS
6244 (if (and c-opt-identifier-concat-key
6245 (looking-at c-opt-identifier-concat-key))
0386b551 6246
d9e94c22
MS
6247 ;; Continue if there's an identifier concatenation
6248 ;; operator after the template argument.
6249 (progn
0386b551
AM
6250 (when (and c-record-type-identifiers id-start)
6251 (c-record-ref-id (cons id-start id-end)))
d9e94c22
MS
6252 (forward-char 2)
6253 (c-forward-syntactic-ws)
6254 t)
0386b551
AM
6255
6256 (when (and c-record-type-identifiers id-start)
6257 (c-record-type-id (cons id-start id-end)))
d9e94c22
MS
6258 (setq res 'template)
6259 nil)))
6260 )))))
6261
6262 (goto-char pos)
6263 res))
6264
e15f8aaa 6265(defun c-forward-type (&optional brace-block-too)
d9e94c22 6266 ;; Move forward over a type spec if at the beginning of one,
e15f8aaa
AM
6267 ;; stopping at the next following token. The keyword "typedef"
6268 ;; isn't part of a type spec here.
6269 ;;
6270 ;; BRACE-BLOCK-TOO, when non-nil, means move over the brace block in
6271 ;; constructs like "struct foo {...} bar ;" or "struct {...} bar;".
6272 ;; The current (2009-03-10) intention is to convert all uses of
6273 ;; `c-forward-type' to call with this parameter set, then to
6274 ;; eliminate it.
6275 ;;
6276 ;; Return
6277 ;; o - t if it's a known type that can't be a name or other
6278 ;; expression;
6279 ;; o - 'known if it's an otherwise known type (according to
6280 ;; `*-font-lock-extra-types');
6281 ;; o - 'prefix if it's a known prefix of a type;
6282 ;; o - 'found if it's a type that matches one in `c-found-types';
53964682 6283 ;; o - 'maybe if it's an identifier that might be a type; or
e15f8aaa
AM
6284 ;; o - nil if it can't be a type (the point isn't moved then).
6285 ;;
6286 ;; The point is assumed to be at the beginning of a token.
d9e94c22
MS
6287 ;;
6288 ;; Note that this function doesn't skip past the brace definition
6289 ;; that might be considered part of the type, e.g.
6290 ;; "enum {a, b, c} foo".
0386b551
AM
6291 ;;
6292 ;; This function records identifier ranges on
6293 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
6294 ;; `c-record-type-identifiers' is non-nil.
6295 ;;
6296 ;; This function might do hidden buffer changes.
0e5cf2b8
AM
6297 (when (and c-recognize-<>-arglists
6298 (looking-at "<"))
452ea855
AM
6299 (c-forward-<>-arglist t)
6300 (c-forward-syntactic-ws))
0386b551
AM
6301
6302 (let ((start (point)) pos res name-res id-start id-end id-range)
d9e94c22
MS
6303
6304 ;; Skip leading type modifiers. If any are found we know it's a
6305 ;; prefix of a type.
e15f8aaa 6306 (when c-opt-type-modifier-key ; e.g. "const" "volatile", but NOT "typedef"
d9e94c22
MS
6307 (while (looking-at c-opt-type-modifier-key)
6308 (goto-char (match-end 1))
6309 (c-forward-syntactic-ws)
6310 (setq res 'prefix)))
6311
6312 (cond
e15f8aaa
AM
6313 ((looking-at c-type-prefix-key) ; e.g. "struct", "class", but NOT
6314 ; "typedef".
d9e94c22
MS
6315 (goto-char (match-end 1))
6316 (c-forward-syntactic-ws)
6317 (setq pos (point))
e15f8aaa
AM
6318
6319 (setq name-res (c-forward-name))
6320 (setq res (not (null name-res)))
6321 (when (eq name-res t)
6322 ;; In many languages the name can be used without the
6323 ;; prefix, so we add it to `c-found-types'.
6324 (c-add-type pos (point))
6325 (when (and c-record-type-identifiers
6326 c-last-identifier-range)
6327 (c-record-type-id c-last-identifier-range)))
6328 (when (and brace-block-too
6329 (memq res '(t nil))
6330 (eq (char-after) ?\{)
6331 (save-excursion
6332 (c-safe
6333 (progn (c-forward-sexp)
6334 (c-forward-syntactic-ws)
6335 (setq pos (point))))))
6336 (goto-char pos)
6337 (setq res t))
6338 (unless res (goto-char start))) ; invalid syntax
d9e94c22
MS
6339
6340 ((progn
6341 (setq pos nil)
6342 (if (looking-at c-identifier-start)
6343 (save-excursion
6344 (setq id-start (point)
0386b551
AM
6345 name-res (c-forward-name))
6346 (when name-res
d9e94c22
MS
6347 (setq id-end (point)
6348 id-range c-last-identifier-range))))
6349 (and (cond ((looking-at c-primitive-type-key)
6350 (setq res t))
6351 ((c-with-syntax-table c-identifier-syntax-table
6352 (looking-at c-known-type-key))
6353 (setq res 'known)))
6354 (or (not id-end)
6355 (>= (save-excursion
6356 (save-match-data
6357 (goto-char (match-end 1))
6358 (c-forward-syntactic-ws)
6359 (setq pos (point))))
6360 id-end)
6361 (setq res nil))))
6362 ;; Looking at a primitive or known type identifier. We've
6363 ;; checked for a name first so that we don't go here if the
6364 ;; known type match only is a prefix of another name.
6365
6366 (setq id-end (match-end 1))
6367
6368 (when (and c-record-type-identifiers
6369 (or c-promote-possible-types (eq res t)))
6370 (c-record-type-id (cons (match-beginning 1) (match-end 1))))
6371
6372 (if (and c-opt-type-component-key
6373 (save-match-data
6374 (looking-at c-opt-type-component-key)))
6375 ;; There might be more keywords for the type.
6376 (let (safe-pos)
0386b551 6377 (c-forward-keyword-clause 1)
d9e94c22
MS
6378 (while (progn
6379 (setq safe-pos (point))
6380 (looking-at c-opt-type-component-key))
6381 (when (and c-record-type-identifiers
6382 (looking-at c-primitive-type-key))
6383 (c-record-type-id (cons (match-beginning 1)
6384 (match-end 1))))
0386b551 6385 (c-forward-keyword-clause 1))
d9e94c22
MS
6386 (if (looking-at c-primitive-type-key)
6387 (progn
6388 (when c-record-type-identifiers
6389 (c-record-type-id (cons (match-beginning 1)
6390 (match-end 1))))
0386b551 6391 (c-forward-keyword-clause 1)
d9e94c22
MS
6392 (setq res t))
6393 (goto-char safe-pos)
6394 (setq res 'prefix)))
0386b551 6395 (unless (save-match-data (c-forward-keyword-clause 1))
d9e94c22
MS
6396 (if pos
6397 (goto-char pos)
6398 (goto-char (match-end 1))
6399 (c-forward-syntactic-ws)))))
6400
0386b551
AM
6401 (name-res
6402 (cond ((eq name-res t)
d9e94c22
MS
6403 ;; A normal identifier.
6404 (goto-char id-end)
6405 (if (or res c-promote-possible-types)
6406 (progn
6407 (c-add-type id-start id-end)
0386b551 6408 (when (and c-record-type-identifiers id-range)
d9e94c22
MS
6409 (c-record-type-id id-range))
6410 (unless res
6411 (setq res 'found)))
6412 (setq res (if (c-check-type id-start id-end)
6413 ;; It's an identifier that has been used as
6414 ;; a type somewhere else.
6415 'found
6416 ;; It's an identifier that might be a type.
6417 'maybe))))
0386b551 6418 ((eq name-res 'template)
d9e94c22
MS
6419 ;; A template is a type.
6420 (goto-char id-end)
6421 (setq res t))
6422 (t
6423 ;; Otherwise it's an operator identifier, which is not a type.
6424 (goto-char start)
6425 (setq res nil)))))
6426
6427 (when res
e15f8aaa 6428 ;; Skip trailing type modifiers. If any are found we know it's
d9e94c22
MS
6429 ;; a type.
6430 (when c-opt-type-modifier-key
e15f8aaa 6431 (while (looking-at c-opt-type-modifier-key) ; e.g. "const", "volatile"
d9e94c22
MS
6432 (goto-char (match-end 1))
6433 (c-forward-syntactic-ws)
6434 (setq res t)))
d9e94c22
MS
6435 ;; Step over any type suffix operator. Do not let the existence
6436 ;; of these alter the classification of the found type, since
6437 ;; these operators typically are allowed in normal expressions
6438 ;; too.
6439 (when c-opt-type-suffix-key
6440 (while (looking-at c-opt-type-suffix-key)
6441 (goto-char (match-end 1))
6442 (c-forward-syntactic-ws)))
6443
e15f8aaa 6444 (when c-opt-type-concat-key ; Only/mainly for pike.
0386b551
AM
6445 ;; Look for a trailing operator that concatenates the type
6446 ;; with a following one, and if so step past that one through
6447 ;; a recursive call. Note that we don't record concatenated
6448 ;; types in `c-found-types' - it's the component types that
6449 ;; are recorded when appropriate.
d9e94c22
MS
6450 (setq pos (point))
6451 (let* ((c-promote-possible-types (or (memq res '(t known))
6452 c-promote-possible-types))
6453 ;; If we can't promote then set `c-record-found-types' so that
6454 ;; we can merge in the types from the second part afterwards if
6455 ;; it turns out to be a known type there.
6456 (c-record-found-types (and c-record-type-identifiers
0386b551
AM
6457 (not c-promote-possible-types)))
6458 subres)
d9e94c22
MS
6459 (if (and (looking-at c-opt-type-concat-key)
6460
6461 (progn
6462 (goto-char (match-end 1))
6463 (c-forward-syntactic-ws)
0386b551 6464 (setq subres (c-forward-type))))
d9e94c22
MS
6465
6466 (progn
6467 ;; If either operand certainly is a type then both are, but we
6468 ;; don't let the existence of the operator itself promote two
6469 ;; uncertain types to a certain one.
6470 (cond ((eq res t))
0386b551
AM
6471 ((eq subres t)
6472 (unless (eq name-res 'template)
6473 (c-add-type id-start id-end))
6474 (when (and c-record-type-identifiers id-range)
d9e94c22
MS
6475 (c-record-type-id id-range))
6476 (setq res t))
2a15eb73 6477 ((eq res 'known))
0386b551 6478 ((eq subres 'known)
2a15eb73 6479 (setq res 'known))
d9e94c22 6480 ((eq res 'found))
0386b551 6481 ((eq subres 'found)
d9e94c22
MS
6482 (setq res 'found))
6483 (t
6484 (setq res 'maybe)))
6485
6486 (when (and (eq res t)
6487 (consp c-record-found-types))
6488 ;; Merge in the ranges of any types found by the second
6489 ;; `c-forward-type'.
6490 (setq c-record-type-identifiers
6491 ;; `nconc' doesn't mind that the tail of
6492 ;; `c-record-found-types' is t.
6493 (nconc c-record-found-types
6494 c-record-type-identifiers))))
6495
6496 (goto-char pos))))
6497
6498 (when (and c-record-found-types (memq res '(known found)) id-range)
6499 (setq c-record-found-types
6500 (cons id-range c-record-found-types))))
6501
6502 ;;(message "c-forward-type %s -> %s: %s" start (point) res)
6503
6504 res))
6505
452ea855
AM
6506(defun c-forward-annotation ()
6507 ;; Used for Java code only at the moment. Assumes point is on the
6508 ;; @, moves forward an annotation. returns nil if there is no
6509 ;; annotation at point.
6510 (and (looking-at "@")
6511 (progn (forward-char) t)
6512 (c-forward-type)
6513 (progn (c-forward-syntactic-ws) t)
6514 (if (looking-at "(")
6515 (c-go-list-forward)
6516 t)))
6517
6758b6a8
GM
6518(defmacro c-pull-open-brace (ps)
6519 ;; Pull the next open brace from PS (which has the form of paren-state),
6520 ;; skipping over any brace pairs. Returns NIL when PS is exhausted.
6521 `(progn
6522 (while (consp (car ,ps))
6523 (setq ,ps (cdr ,ps)))
6524 (prog1 (car ,ps)
6525 (setq ,ps (cdr ,ps)))))
6526
3d1c39fa
AM
6527(defun c-back-over-member-initializers ()
6528 ;; Test whether we are in a C++ member initializer list, and if so, go back
6529 ;; to the introducing ":", returning the position of the opening paren of
6530 ;; the function's arglist. Otherwise return nil, leaving point unchanged.
6531 (let ((here (point))
6532 (paren-state (c-parse-state))
6533 res)
6534
6535 (setq res
6536 (catch 'done
6537 (if (not (c-at-toplevel-p))
6538 (progn
6539 (while (not (c-at-toplevel-p))
6540 (goto-char (c-pull-open-brace paren-state)))
6541 (c-backward-syntactic-ws)
6542 (when (not (c-simple-skip-symbol-backward))
6543 (throw 'done nil))
6544 (c-backward-syntactic-ws))
6545 (c-backward-syntactic-ws)
6546 (when (memq (char-before) '(?\) ?}))
6547 (when (not (c-go-list-backward))
6548 (throw 'done nil))
6549 (c-backward-syntactic-ws))
6550 (when (c-simple-skip-symbol-backward)
6551 (c-backward-syntactic-ws)))
6552
6553 (while (eq (char-before) ?,)
6554 (backward-char)
6555 (c-backward-syntactic-ws)
6556
6557 (when (not (memq (char-before) '(?\) ?})))
6558 (throw 'done nil))
6559 (when (not (c-go-list-backward))
6560 (throw 'done nil))
6561 (c-backward-syntactic-ws)
6562 (when (not (c-simple-skip-symbol-backward))
6563 (throw 'done nil))
6564 (c-backward-syntactic-ws))
6565
6566 (and
6567 (eq (char-before) ?:)
6568 (c-just-after-func-arglist-p))))
6569
6570 (or res (goto-char here))
6571 res))
6572
785eecbb 6573\f
d9e94c22
MS
6574;; Handling of large scale constructs like statements and declarations.
6575
0386b551
AM
6576;; Macro used inside `c-forward-decl-or-cast-1'. It ought to be a
6577;; defsubst or perhaps even a defun, but it contains lots of free
6578;; variables that refer to things inside `c-forward-decl-or-cast-1'.
6579(defmacro c-fdoc-shift-type-backward (&optional short)
6580 ;; `c-forward-decl-or-cast-1' can consume an arbitrary length list
6581 ;; of types when parsing a declaration, which means that it
6582 ;; sometimes consumes the identifier in the declaration as a type.
6583 ;; This is used to "backtrack" and make the last type be treated as
6584 ;; an identifier instead.
6585 `(progn
6586 ,(unless short
6587 ;; These identifiers are bound only in the inner let.
6588 '(setq identifier-type at-type
6589 identifier-start type-start
6590 got-parens nil
6591 got-identifier t
6592 got-suffix t
6593 got-suffix-after-parens id-start
6594 paren-depth 0))
6595
6596 (if (setq at-type (if (eq backup-at-type 'prefix)
6597 t
6598 backup-at-type))
6599 (setq type-start backup-type-start
6600 id-start backup-id-start)
6601 (setq type-start start-pos
6602 id-start start-pos))
6603
6604 ;; When these flags already are set we've found specifiers that
6605 ;; unconditionally signal these attributes - backtracking doesn't
6606 ;; change that. So keep them set in that case.
6607 (or at-type-decl
6608 (setq at-type-decl backup-at-type-decl))
6609 (or maybe-typeless
6610 (setq maybe-typeless backup-maybe-typeless))
6611
6612 ,(unless short
6613 ;; This identifier is bound only in the inner let.
6614 '(setq start id-start))))
6615
6616(defun c-forward-decl-or-cast-1 (preceding-token-end context last-cast-end)
6617 ;; Move forward over a declaration or a cast if at the start of one.
6618 ;; The point is assumed to be at the start of some token. Nil is
6619 ;; returned if no declaration or cast is recognized, and the point
6620 ;; is clobbered in that case.
6621 ;;
6622 ;; If a declaration is parsed:
6623 ;;
6624 ;; The point is left at the first token after the first complete
6625 ;; declarator, if there is one. The return value is a cons where
022d0cf4
AM
6626 ;; the car is the position of the first token in the declarator. (See
6627 ;; below for the cdr.)
0386b551
AM
6628 ;; Some examples:
6629 ;;
6630 ;; void foo (int a, char *b) stuff ...
6631 ;; car ^ ^ point
6632 ;; float (*a)[], b;
6633 ;; car ^ ^ point
6634 ;; unsigned int a = c_style_initializer, b;
6635 ;; car ^ ^ point
6636 ;; unsigned int a (cplusplus_style_initializer), b;
6637 ;; car ^ ^ point (might change)
6638 ;; class Foo : public Bar {}
6639 ;; car ^ ^ point
6640 ;; class PikeClass (int a, string b) stuff ...
6641 ;; car ^ ^ point
6642 ;; enum bool;
6643 ;; car ^ ^ point
6644 ;; enum bool flag;
6645 ;; car ^ ^ point
6646 ;; void cplusplus_function (int x) throw (Bad);
6647 ;; car ^ ^ point
6648 ;; Foo::Foo (int b) : Base (b) {}
6649 ;; car ^ ^ point
91af3942 6650 ;;
e15f8aaa
AM
6651 ;; The cdr of the return value is non-nil when a
6652 ;; `c-typedef-decl-kwds' specifier is found in the declaration.
6653 ;; Specifically it is a dotted pair (A . B) where B is t when a
6654 ;; `c-typedef-kwds' ("typedef") is present, and A is t when some
6655 ;; other `c-typedef-decl-kwds' (e.g. class, struct, enum)
6656 ;; specifier is present. I.e., (some of) the declared
6657 ;; identifier(s) are types.
91af3942 6658 ;;
0386b551
AM
6659 ;; If a cast is parsed:
6660 ;;
6661 ;; The point is left at the first token after the closing paren of
6662 ;; the cast. The return value is `cast'. Note that the start
6663 ;; position must be at the first token inside the cast parenthesis
6664 ;; to recognize it.
6665 ;;
6666 ;; PRECEDING-TOKEN-END is the first position after the preceding
6667 ;; token, i.e. on the other side of the syntactic ws from the point.
6668 ;; Use a value less than or equal to (point-min) if the point is at
6669 ;; the first token in (the visible part of) the buffer.
6670 ;;
6671 ;; CONTEXT is a symbol that describes the context at the point:
022d0cf4 6672 ;; 'decl In a comma-separated declaration context (typically
0386b551
AM
6673 ;; inside a function declaration arglist).
6674 ;; '<> In an angle bracket arglist.
6675 ;; 'arglist Some other type of arglist.
a85fd6da
AM
6676 ;; nil Some other context or unknown context. Includes
6677 ;; within the parens of an if, for, ... construct.
0386b551
AM
6678 ;;
6679 ;; LAST-CAST-END is the first token after the closing paren of a
6680 ;; preceding cast, or nil if none is known. If
6681 ;; `c-forward-decl-or-cast-1' is used in succession, it should be
6682 ;; the position after the closest preceding call where a cast was
6683 ;; matched. In that case it's used to discover chains of casts like
6684 ;; "(a) (b) c".
6685 ;;
6686 ;; This function records identifier ranges on
6687 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
6688 ;; `c-record-type-identifiers' is non-nil.
6689 ;;
6690 ;; This function might do hidden buffer changes.
6691
6692 (let (;; `start-pos' is used below to point to the start of the
6693 ;; first type, i.e. after any leading specifiers. It might
6694 ;; also point at the beginning of the preceding syntactic
6695 ;; whitespace.
6696 (start-pos (point))
6697 ;; Set to the result of `c-forward-type'.
6698 at-type
6699 ;; The position of the first token in what we currently
6700 ;; believe is the type in the declaration or cast, after any
6701 ;; specifiers and their associated clauses.
6702 type-start
6703 ;; The position of the first token in what we currently
6704 ;; believe is the declarator for the first identifier. Set
6705 ;; when the type is found, and moved forward over any
6706 ;; `c-decl-hangon-kwds' and their associated clauses that
6707 ;; occurs after the type.
6708 id-start
6709 ;; These store `at-type', `type-start' and `id-start' of the
6710 ;; identifier before the one in those variables. The previous
6711 ;; identifier might turn out to be the real type in a
6712 ;; declaration if the last one has to be the declarator in it.
6713 ;; If `backup-at-type' is nil then the other variables have
6714 ;; undefined values.
6715 backup-at-type backup-type-start backup-id-start
e15f8aaa
AM
6716 ;; Set if we've found a specifier (apart from "typedef") that makes
6717 ;; the defined identifier(s) types.
0386b551 6718 at-type-decl
e15f8aaa
AM
6719 ;; Set if we've a "typedef" keyword.
6720 at-typedef
0386b551
AM
6721 ;; Set if we've found a specifier that can start a declaration
6722 ;; where there's no type.
6723 maybe-typeless
6724 ;; If a specifier is found that also can be a type prefix,
6725 ;; these flags are set instead of those above. If we need to
6726 ;; back up an identifier, they are copied to the real flag
6727 ;; variables. Thus they only take effect if we fail to
6728 ;; interpret it as a type.
6729 backup-at-type-decl backup-maybe-typeless
6730 ;; Whether we've found a declaration or a cast. We might know
6731 ;; this before we've found the type in it. It's 'ids if we've
6732 ;; found two consecutive identifiers (usually a sure sign, but
6733 ;; we should allow that in labels too), and t if we've found a
6734 ;; specifier keyword (a 100% sure sign).
6735 at-decl-or-cast
6736 ;; Set when we need to back up to parse this as a declaration
6737 ;; but not as a cast.
6738 backup-if-not-cast
6739 ;; For casts, the return position.
6740 cast-end
6741 ;; Save `c-record-type-identifiers' and
6742 ;; `c-record-ref-identifiers' since ranges are recorded
6743 ;; speculatively and should be thrown away if it turns out
6744 ;; that it isn't a declaration or cast.
6745 (save-rec-type-ids c-record-type-identifiers)
6746 (save-rec-ref-ids c-record-ref-identifiers))
6747
452ea855
AM
6748 (while (c-forward-annotation)
6749 (c-forward-syntactic-ws))
6750
0386b551
AM
6751 ;; Check for a type. Unknown symbols are treated as possible
6752 ;; types, but they could also be specifiers disguised through
6753 ;; macros like __INLINE__, so we recognize both types and known
6754 ;; specifiers after them too.
6755 (while
6756 (let* ((start (point)) kwd-sym kwd-clause-end found-type)
6757
6758 ;; Look for a specifier keyword clause.
e3715033
AM
6759 (when (or (looking-at c-prefix-spec-kwds-re)
6760 (and (c-major-mode-is 'java-mode)
6761 (looking-at "@[A-Za-z0-9]+")))
e15f8aaa
AM
6762 (if (looking-at c-typedef-key)
6763 (setq at-typedef t))
0386b551
AM
6764 (setq kwd-sym (c-keyword-sym (match-string 1)))
6765 (save-excursion
6766 (c-forward-keyword-clause 1)
6767 (setq kwd-clause-end (point))))
6768
e15f8aaa 6769 (when (setq found-type (c-forward-type t)) ; brace-block-too
0386b551
AM
6770 ;; Found a known or possible type or a prefix of a known type.
6771
6772 (when at-type
6773 ;; Got two identifiers with nothing but whitespace
6774 ;; between them. That can only happen in declarations.
6775 (setq at-decl-or-cast 'ids)
6776
6777 (when (eq at-type 'found)
6778 ;; If the previous identifier is a found type we
6779 ;; record it as a real one; it might be some sort of
6780 ;; alias for a prefix like "unsigned".
6781 (save-excursion
6782 (goto-char type-start)
6783 (let ((c-promote-possible-types t))
6784 (c-forward-type)))))
6785
6786 (setq backup-at-type at-type
6787 backup-type-start type-start
6788 backup-id-start id-start
6789 at-type found-type
6790 type-start start
6791 id-start (point)
6792 ;; The previous ambiguous specifier/type turned out
6793 ;; to be a type since we've parsed another one after
6794 ;; it, so clear these backup flags.
6795 backup-at-type-decl nil
6796 backup-maybe-typeless nil))
6797
6798 (if kwd-sym
6799 (progn
6800 ;; Handle known specifier keywords and
6801 ;; `c-decl-hangon-kwds' which can occur after known
6802 ;; types.
6803
6804 (if (c-keyword-member kwd-sym 'c-decl-hangon-kwds)
6805 ;; It's a hang-on keyword that can occur anywhere.
6806 (progn
6807 (setq at-decl-or-cast t)
6808 (if at-type
6809 ;; Move the identifier start position if
6810 ;; we've passed a type.
6811 (setq id-start kwd-clause-end)
6812 ;; Otherwise treat this as a specifier and
6813 ;; move the fallback position.
6814 (setq start-pos kwd-clause-end))
6815 (goto-char kwd-clause-end))
6816
6817 ;; It's an ordinary specifier so we know that
6818 ;; anything before this can't be the type.
6819 (setq backup-at-type nil
6820 start-pos kwd-clause-end)
6821
6822 (if found-type
6823 ;; It's ambiguous whether this keyword is a
6824 ;; specifier or a type prefix, so set the backup
6825 ;; flags. (It's assumed that `c-forward-type'
6826 ;; moved further than `c-forward-keyword-clause'.)
6827 (progn
6828 (when (c-keyword-member kwd-sym 'c-typedef-decl-kwds)
6829 (setq backup-at-type-decl t))
6830 (when (c-keyword-member kwd-sym 'c-typeless-decl-kwds)
6831 (setq backup-maybe-typeless t)))
6832
6833 (when (c-keyword-member kwd-sym 'c-typedef-decl-kwds)
e15f8aaa
AM
6834 ;; This test only happens after we've scanned a type.
6835 ;; So, with valid syntax, kwd-sym can't be 'typedef.
0386b551
AM
6836 (setq at-type-decl t))
6837 (when (c-keyword-member kwd-sym 'c-typeless-decl-kwds)
6838 (setq maybe-typeless t))
6839
ee7683eb 6840 ;; Haven't matched a type so it's an unambiguous
0386b551
AM
6841 ;; specifier keyword and we know we're in a
6842 ;; declaration.
6843 (setq at-decl-or-cast t)
6844
6845 (goto-char kwd-clause-end))))
6846
6847 ;; If the type isn't known we continue so that we'll jump
6848 ;; over all specifiers and type identifiers. The reason
6849 ;; to do this for a known type prefix is to make things
6850 ;; like "unsigned INT16" work.
6851 (and found-type (not (eq found-type t))))))
6852
6853 (cond
6854 ((eq at-type t)
6855 ;; If a known type was found, we still need to skip over any
6856 ;; hangon keyword clauses after it. Otherwise it has already
6857 ;; been done in the loop above.
6858 (while (looking-at c-decl-hangon-key)
6859 (c-forward-keyword-clause 1))
6860 (setq id-start (point)))
6861
6862 ((eq at-type 'prefix)
6863 ;; A prefix type is itself a primitive type when it's not
6864 ;; followed by another type.
6865 (setq at-type t))
6866
6867 ((not at-type)
6868 ;; Got no type but set things up to continue anyway to handle
6869 ;; the various cases when a declaration doesn't start with a
6870 ;; type.
6871 (setq id-start start-pos))
6872
6873 ((and (eq at-type 'maybe)
6874 (c-major-mode-is 'c++-mode))
6875 ;; If it's C++ then check if the last "type" ends on the form
6876 ;; "foo::foo" or "foo::~foo", i.e. if it's the name of a
6877 ;; (con|de)structor.
6878 (save-excursion
6879 (let (name end-2 end-1)
6880 (goto-char id-start)
6881 (c-backward-syntactic-ws)
6882 (setq end-2 (point))
6883 (when (and
6884 (c-simple-skip-symbol-backward)
6885 (progn
6886 (setq name
6887 (buffer-substring-no-properties (point) end-2))
6888 ;; Cheating in the handling of syntactic ws below.
6889 (< (skip-chars-backward ":~ \t\n\r\v\f") 0))
6890 (progn
6891 (setq end-1 (point))
6892 (c-simple-skip-symbol-backward))
6893 (>= (point) type-start)
6894 (equal (buffer-substring-no-properties (point) end-1)
6895 name))
6896 ;; It is a (con|de)structor name. In that case the
6897 ;; declaration is typeless so zap out any preceding
6898 ;; identifier(s) that we might have taken as types.
6899 (goto-char type-start)
6900 (setq at-type nil
6901 backup-at-type nil
6902 id-start type-start))))))
6903
6904 ;; Check for and step over a type decl expression after the thing
6905 ;; that is or might be a type. This can't be skipped since we
6906 ;; need the correct end position of the declarator for
6907 ;; `max-type-decl-end-*'.
6908 (let ((start (point)) (paren-depth 0) pos
6909 ;; True if there's a non-open-paren match of
6910 ;; `c-type-decl-prefix-key'.
6911 got-prefix
6912 ;; True if the declarator is surrounded by a parenthesis pair.
6913 got-parens
6914 ;; True if there is an identifier in the declarator.
6915 got-identifier
6916 ;; True if there's a non-close-paren match of
6917 ;; `c-type-decl-suffix-key'.
6918 got-suffix
6919 ;; True if there's a prefix match outside the outermost
6920 ;; paren pair that surrounds the declarator.
6921 got-prefix-before-parens
1379f2c5 6922 ;; True if there's a suffix match outside the outermost
0386b551
AM
6923 ;; paren pair that surrounds the declarator. The value is
6924 ;; the position of the first suffix match.
6925 got-suffix-after-parens
6926 ;; True if we've parsed the type decl to a token that is
6927 ;; known to end declarations in this context.
6928 at-decl-end
6929 ;; The earlier values of `at-type' and `type-start' if we've
6930 ;; shifted the type backwards.
6931 identifier-type identifier-start
6932 ;; If `c-parse-and-markup-<>-arglists' is set we need to
6933 ;; turn it off during the name skipping below to avoid
6934 ;; getting `c-type' properties that might be bogus. That
6935 ;; can happen since we don't know if
6936 ;; `c-restricted-<>-arglists' will be correct inside the
6937 ;; arglist paren that gets entered.
033f22dd
AM
6938 c-parse-and-markup-<>-arglists
6939 ;; Start of the identifier for which `got-identifier' was set.
6940 name-start)
0386b551
AM
6941
6942 (goto-char id-start)
6943
6944 ;; Skip over type decl prefix operators. (Note similar code in
6945 ;; `c-font-lock-declarators'.)
e8dd0787
AM
6946 (if (and c-recognize-typeless-decls
6947 (equal c-type-decl-prefix-key "\\<\\>"))
6948 (when (eq (char-after) ?\()
0386b551
AM
6949 (progn
6950 (setq paren-depth (1+ paren-depth))
e8dd0787
AM
6951 (forward-char)))
6952 (while (and (looking-at c-type-decl-prefix-key)
6953 (if (and (c-major-mode-is 'c++-mode)
6954 (match-beginning 3))
6955 ;; If the third submatch matches in C++ then
6956 ;; we're looking at an identifier that's a
6957 ;; prefix only if it specifies a member pointer.
033f22dd
AM
6958 (when (progn (setq pos (point))
6959 (setq got-identifier (c-forward-name)))
6960 (setq name-start pos)
e8dd0787
AM
6961 (if (looking-at "\\(::\\)")
6962 ;; We only check for a trailing "::" and
6963 ;; let the "*" that should follow be
6964 ;; matched in the next round.
6965 (progn (setq got-identifier nil) t)
6966 ;; It turned out to be the real identifier,
6967 ;; so stop.
6968 nil))
6969 t))
6970
6971 (if (eq (char-after) ?\()
6972 (progn
6973 (setq paren-depth (1+ paren-depth))
6974 (forward-char))
6975 (unless got-prefix-before-parens
6976 (setq got-prefix-before-parens (= paren-depth 0)))
6977 (setq got-prefix t)
6978 (goto-char (match-end 1)))
6979 (c-forward-syntactic-ws)))
0386b551
AM
6980
6981 (setq got-parens (> paren-depth 0))
6982
6983 ;; Skip over an identifier.
6984 (or got-identifier
6985 (and (looking-at c-identifier-start)
033f22dd
AM
6986 (setq pos (point))
6987 (setq got-identifier (c-forward-name))
6988 (setq name-start pos)))
0386b551
AM
6989
6990 ;; Skip over type decl suffix operators.
6991 (while (if (looking-at c-type-decl-suffix-key)
6992
6993 (if (eq (char-after) ?\))
6994 (when (> paren-depth 0)
6995 (setq paren-depth (1- paren-depth))
6996 (forward-char)
6997 t)
6998 (when (if (save-match-data (looking-at "\\s\("))
6999 (c-safe (c-forward-sexp 1) t)
7000 (goto-char (match-end 1))
7001 t)
7002 (when (and (not got-suffix-after-parens)
7003 (= paren-depth 0))
7004 (setq got-suffix-after-parens (match-beginning 0)))
7005 (setq got-suffix t)))
7006
7007 ;; No suffix matched. We might have matched the
7008 ;; identifier as a type and the open paren of a
7009 ;; function arglist as a type decl prefix. In that
7010 ;; case we should "backtrack": Reinterpret the last
7011 ;; type as the identifier, move out of the arglist and
7012 ;; continue searching for suffix operators.
7013 ;;
7014 ;; Do this even if there's no preceding type, to cope
7015 ;; with old style function declarations in K&R C,
7016 ;; (con|de)structors in C++ and `c-typeless-decl-kwds'
7017 ;; style declarations. That isn't applicable in an
7018 ;; arglist context, though.
7019 (when (and (= paren-depth 1)
7020 (not got-prefix-before-parens)
7021 (not (eq at-type t))
7022 (or backup-at-type
7023 maybe-typeless
7024 backup-maybe-typeless
7025 (when c-recognize-typeless-decls
7026 (not context)))
7027 (setq pos (c-up-list-forward (point)))
7028 (eq (char-before pos) ?\)))
7029 (c-fdoc-shift-type-backward)
7030 (goto-char pos)
7031 t))
7032
7033 (c-forward-syntactic-ws))
7034
7035 (when (and (or maybe-typeless backup-maybe-typeless)
7036 (not got-identifier)
7037 (not got-prefix)
7038 at-type)
7039 ;; Have found no identifier but `c-typeless-decl-kwds' has
7040 ;; matched so we know we're inside a declaration. The
7041 ;; preceding type must be the identifier instead.
7042 (c-fdoc-shift-type-backward))
7043
7044 (setq
7045 at-decl-or-cast
7046 (catch 'at-decl-or-cast
7047
a85fd6da 7048 ;; CASE 1
0386b551
AM
7049 (when (> paren-depth 0)
7050 ;; Encountered something inside parens that isn't matched by
7051 ;; the `c-type-decl-*' regexps, so it's not a type decl
7052 ;; expression. Try to skip out to the same paren depth to
7053 ;; not confuse the cast check below.
7054 (c-safe (goto-char (scan-lists (point) 1 paren-depth)))
7055 ;; If we've found a specifier keyword then it's a
7056 ;; declaration regardless.
7057 (throw 'at-decl-or-cast (eq at-decl-or-cast t)))
7058
7059 (setq at-decl-end
7060 (looking-at (cond ((eq context '<>) "[,>]")
7061 (context "[,\)]")
7062 (t "[,;]"))))
7063
7064 ;; Now we've collected info about various characteristics of
7065 ;; the construct we're looking at. Below follows a decision
7066 ;; tree based on that. It's ordered to check more certain
7067 ;; signs before less certain ones.
7068
7069 (if got-identifier
7070 (progn
7071
a85fd6da 7072 ;; CASE 2
0386b551
AM
7073 (when (and (or at-type maybe-typeless)
7074 (not (or got-prefix got-parens)))
7075 ;; Got another identifier directly after the type, so it's a
7076 ;; declaration.
7077 (throw 'at-decl-or-cast t))
7078
0386b551 7079
033f22dd
AM
7080 (when (and got-parens
7081 (not got-prefix)
7082 ;; (not got-suffix-after-parens)
7083 (or backup-at-type
7084 maybe-typeless
7085 backup-maybe-typeless
7086 (eq at-decl-or-cast t)
7087 (save-excursion
7088 (goto-char name-start)
7089 (not (memq (c-forward-type) '(nil maybe))))))
7090 ;; Got a declaration of the form "foo bar (gnu);" or "bar
7091 ;; (gnu);" where we've recognized "bar" as the type and "gnu"
7092 ;; as the declarator. In this case it's however more likely
7093 ;; that "bar" is the declarator and "gnu" a function argument
7094 ;; or initializer (if `c-recognize-paren-inits' is set),
7095 ;; since the parens around "gnu" would be superfluous if it's
7096 ;; a declarator. Shift the type one step backward.
7097 (c-fdoc-shift-type-backward)))
7098
7099 ;; Found no identifier.
0386b551
AM
7100 (if backup-at-type
7101 (progn
7102
b248a85d
AM
7103
7104 ;; CASE 3
7105 (when (= (point) start)
7106 ;; Got a plain list of identifiers. If a colon follows it's
7107 ;; a valid label, or maybe a bitfield. Otherwise the last
7108 ;; one probably is the declared identifier and we should
7109 ;; back up to the previous type, providing it isn't a cast.
452ea855
AM
7110 (if (and (eq (char-after) ?:)
7111 (not (c-major-mode-is 'java-mode)))
b248a85d
AM
7112 (cond
7113 ;; If we've found a specifier keyword then it's a
7114 ;; declaration regardless.
7115 ((eq at-decl-or-cast t)
7116 (throw 'at-decl-or-cast t))
7117 ((and c-has-bitfields
7118 (eq at-decl-or-cast 'ids)) ; bitfield.
7119 (setq backup-if-not-cast t)
7120 (throw 'at-decl-or-cast t)))
7121
7122 (setq backup-if-not-cast t)
7123 (throw 'at-decl-or-cast t)))
0386b551 7124
a85fd6da 7125 ;; CASE 4
0386b551
AM
7126 (when (and got-suffix
7127 (not got-prefix)
7128 (not got-parens))
7129 ;; Got a plain list of identifiers followed by some suffix.
7130 ;; If this isn't a cast then the last identifier probably is
7131 ;; the declared one and we should back up to the previous
7132 ;; type.
7133 (setq backup-if-not-cast t)
7134 (throw 'at-decl-or-cast t)))
7135
a85fd6da 7136 ;; CASE 5
0386b551
AM
7137 (when (eq at-type t)
7138 ;; If the type is known we know that there can't be any
7139 ;; identifier somewhere else, and it's only in declarations in
7140 ;; e.g. function prototypes and in casts that the identifier may
7141 ;; be left out.
7142 (throw 'at-decl-or-cast t))
7143
7144 (when (= (point) start)
7145 ;; Only got a single identifier (parsed as a type so far).
a85fd6da 7146 ;; CASE 6
0386b551
AM
7147 (if (and
7148 ;; Check that the identifier isn't at the start of an
7149 ;; expression.
7150 at-decl-end
7151 (cond
7152 ((eq context 'decl)
7153 ;; Inside an arglist that contains declarations. If K&R
7154 ;; style declarations and parenthesis style initializers
7155 ;; aren't allowed then the single identifier must be a
7156 ;; type, else we require that it's known or found
7157 ;; (primitive types are handled above).
7158 (or (and (not c-recognize-knr-p)
7159 (not c-recognize-paren-inits))
7160 (memq at-type '(known found))))
7161 ((eq context '<>)
7162 ;; Inside a template arglist. Accept known and found
7163 ;; types; other identifiers could just as well be
7164 ;; constants in C++.
7165 (memq at-type '(known found)))))
7166 (throw 'at-decl-or-cast t)
a85fd6da 7167 ;; CASE 7
0386b551
AM
7168 ;; Can't be a valid declaration or cast, but if we've found a
7169 ;; specifier it can't be anything else either, so treat it as
7170 ;; an invalid/unfinished declaration or cast.
7171 (throw 'at-decl-or-cast at-decl-or-cast))))
7172
7173 (if (and got-parens
7174 (not got-prefix)
7175 (not context)
7176 (not (eq at-type t))
7177 (or backup-at-type
7178 maybe-typeless
7179 backup-maybe-typeless
7180 (when c-recognize-typeless-decls
7181 (or (not got-suffix)
7182 (not (looking-at
7183 c-after-suffixed-type-maybe-decl-key))))))
7184 ;; Got an empty paren pair and a preceding type that probably
7185 ;; really is the identifier. Shift the type backwards to make
7186 ;; the last one the identifier. This is analogous to the
7187 ;; "backtracking" done inside the `c-type-decl-suffix-key' loop
7188 ;; above.
7189 ;;
7190 ;; Exception: In addition to the conditions in that
7191 ;; "backtracking" code, do not shift backward if we're not
7192 ;; looking at either `c-after-suffixed-type-decl-key' or "[;,]".
7193 ;; Since there's no preceding type, the shift would mean that
7194 ;; the declaration is typeless. But if the regexp doesn't match
7195 ;; then we will simply fall through in the tests below and not
7196 ;; recognize it at all, so it's better to try it as an abstract
7197 ;; declarator instead.
7198 (c-fdoc-shift-type-backward)
7199
7200 ;; Still no identifier.
a85fd6da 7201 ;; CASE 8
0386b551
AM
7202 (when (and got-prefix (or got-parens got-suffix))
7203 ;; Require `got-prefix' together with either `got-parens' or
7204 ;; `got-suffix' to recognize it as an abstract declarator:
7205 ;; `got-parens' only is probably an empty function call.
7206 ;; `got-suffix' only can build an ordinary expression together
7207 ;; with the preceding identifier which we've taken as a type.
7208 ;; We could actually accept on `got-prefix' only, but that can
7209 ;; easily occur temporarily while writing an expression so we
7210 ;; avoid that case anyway. We could do a better job if we knew
7211 ;; the point when the fontification was invoked.
7212 (throw 'at-decl-or-cast t))
7213
a85fd6da 7214 ;; CASE 9
0386b551
AM
7215 (when (and at-type
7216 (not got-prefix)
7217 (not got-parens)
7218 got-suffix-after-parens
7219 (eq (char-after got-suffix-after-parens) ?\())
7220 ;; Got a type, no declarator but a paren suffix. I.e. it's a
fa463103 7221 ;; normal function call after all (or perhaps a C++ style object
0386b551
AM
7222 ;; instantiation expression).
7223 (throw 'at-decl-or-cast nil))))
7224
a85fd6da 7225 ;; CASE 10
0386b551
AM
7226 (when at-decl-or-cast
7227 ;; By now we've located the type in the declaration that we know
7228 ;; we're in.
7229 (throw 'at-decl-or-cast t))
7230
a85fd6da 7231 ;; CASE 11
0386b551
AM
7232 (when (and got-identifier
7233 (not context)
7234 (looking-at c-after-suffixed-type-decl-key)
7235 (if (and got-parens
7236 (not got-prefix)
7237 (not got-suffix)
7238 (not (eq at-type t)))
7239 ;; Shift the type backward in the case that there's a
7240 ;; single identifier inside parens. That can only
7241 ;; occur in K&R style function declarations so it's
7242 ;; more likely that it really is a function call.
7243 ;; Therefore we only do this after
7244 ;; `c-after-suffixed-type-decl-key' has matched.
7245 (progn (c-fdoc-shift-type-backward) t)
7246 got-suffix-after-parens))
7247 ;; A declaration according to `c-after-suffixed-type-decl-key'.
7248 (throw 'at-decl-or-cast t))
7249
a85fd6da 7250 ;; CASE 12
0386b551
AM
7251 (when (and (or got-prefix (not got-parens))
7252 (memq at-type '(t known)))
7253 ;; It's a declaration if a known type precedes it and it can't be a
7254 ;; function call.
7255 (throw 'at-decl-or-cast t))
7256
7257 ;; If we get here we can't tell if this is a type decl or a normal
7258 ;; expression by looking at it alone. (That's under the assumption
7259 ;; that normal expressions always can look like type decl expressions,
7260 ;; which isn't really true but the cases where it doesn't hold are so
7261 ;; uncommon (e.g. some placements of "const" in C++) it's not worth
7262 ;; the effort to look for them.)
7263
3af1c868 7264;;; 2008-04-16: commented out the next form, to allow the function to recognize
2cdb7698
AM
7265;;; "foo (int bar)" in CC (an implicit type (in class foo) without a semicolon)
7266;;; as a(n almost complete) declaration, enabling it to be fontified.
7267 ;; CASE 13
7268 ;; (unless (or at-decl-end (looking-at "=[^=]"))
7269 ;; If this is a declaration it should end here or its initializer(*)
7270 ;; should start here, so check for allowed separation tokens. Note
7271 ;; that this rule doesn't work e.g. with a K&R arglist after a
7272 ;; function header.
7273 ;;
7274 ;; *) Don't check for C++ style initializers using parens
7275 ;; since those already have been matched as suffixes.
7276 ;;
7277 ;; If `at-decl-or-cast' is then we've found some other sign that
7278 ;; it's a declaration or cast, so then it's probably an
7279 ;; invalid/unfinished one.
7280 ;; (throw 'at-decl-or-cast at-decl-or-cast))
0386b551
AM
7281
7282 ;; Below are tests that only should be applied when we're certain to
7283 ;; not have parsed halfway through an expression.
7284
a85fd6da 7285 ;; CASE 14
0386b551
AM
7286 (when (memq at-type '(t known))
7287 ;; The expression starts with a known type so treat it as a
7288 ;; declaration.
7289 (throw 'at-decl-or-cast t))
7290
a85fd6da 7291 ;; CASE 15
0386b551
AM
7292 (when (and (c-major-mode-is 'c++-mode)
7293 ;; In C++ we check if the identifier is a known type, since
7294 ;; (con|de)structors use the class name as identifier.
7295 ;; We've always shifted over the identifier as a type and
7296 ;; then backed up again in this case.
7297 identifier-type
7298 (or (memq identifier-type '(found known))
7299 (and (eq (char-after identifier-start) ?~)
7300 ;; `at-type' probably won't be 'found for
7301 ;; destructors since the "~" is then part of the
7302 ;; type name being checked against the list of
7303 ;; known types, so do a check without that
7304 ;; operator.
7305 (or (save-excursion
7306 (goto-char (1+ identifier-start))
7307 (c-forward-syntactic-ws)
7308 (c-with-syntax-table
7309 c-identifier-syntax-table
7310 (looking-at c-known-type-key)))
7311 (save-excursion
7312 (goto-char (1+ identifier-start))
7313 ;; We have already parsed the type earlier,
7314 ;; so it'd be possible to cache the end
7315 ;; position instead of redoing it here, but
7316 ;; then we'd need to keep track of another
7317 ;; position everywhere.
7318 (c-check-type (point)
7319 (progn (c-forward-type)
7320 (point))))))))
7321 (throw 'at-decl-or-cast t))
7322
7323 (if got-identifier
7324 (progn
a85fd6da 7325 ;; CASE 16
0386b551
AM
7326 (when (and got-prefix-before-parens
7327 at-type
7328 (or at-decl-end (looking-at "=[^=]"))
7329 (not context)
7330 (not got-suffix))
7331 ;; Got something like "foo * bar;". Since we're not inside an
7332 ;; arglist it would be a meaningless expression because the
7333 ;; result isn't used. We therefore choose to recognize it as
7334 ;; a declaration. Do not allow a suffix since it could then
7335 ;; be a function call.
7336 (throw 'at-decl-or-cast t))
7337
a85fd6da 7338 ;; CASE 17
0386b551
AM
7339 (when (and (or got-suffix-after-parens
7340 (looking-at "=[^=]"))
7341 (eq at-type 'found)
7342 (not (eq context 'arglist)))
7343 ;; Got something like "a (*b) (c);" or "a (b) = c;". It could
7344 ;; be an odd expression or it could be a declaration. Treat
7345 ;; it as a declaration if "a" has been used as a type
7346 ;; somewhere else (if it's a known type we won't get here).
7347 (throw 'at-decl-or-cast t)))
7348
a85fd6da 7349 ;; CASE 18
0386b551
AM
7350 (when (and context
7351 (or got-prefix
7352 (and (eq context 'decl)
7353 (not c-recognize-paren-inits)
7354 (or got-parens got-suffix))))
7355 ;; Got a type followed by an abstract declarator. If `got-prefix'
7356 ;; is set it's something like "a *" without anything after it. If
7357 ;; `got-parens' or `got-suffix' is set it's "a()", "a[]", "a()[]",
7358 ;; or similar, which we accept only if the context rules out
7359 ;; expressions.
7360 (throw 'at-decl-or-cast t)))
7361
7362 ;; If we had a complete symbol table here (which rules out
7363 ;; `c-found-types') we should return t due to the disambiguation rule
7364 ;; (in at least C++) that anything that can be parsed as a declaration
7365 ;; is a declaration. Now we're being more defensive and prefer to
7366 ;; highlight things like "foo (bar);" as a declaration only if we're
7367 ;; inside an arglist that contains declarations.
7368 (eq context 'decl))))
7369
7370 ;; The point is now after the type decl expression.
7371
7372 (cond
7373 ;; Check for a cast.
7374 ((save-excursion
7375 (and
7376 c-cast-parens
7377
7378 ;; Should be the first type/identifier in a cast paren.
7379 (> preceding-token-end (point-min))
7380 (memq (char-before preceding-token-end) c-cast-parens)
7381
7382 ;; The closing paren should follow.
7383 (progn
7384 (c-forward-syntactic-ws)
7385 (looking-at "\\s\)"))
7386
7387 ;; There should be a primary expression after it.
7388 (let (pos)
7389 (forward-char)
7390 (c-forward-syntactic-ws)
7391 (setq cast-end (point))
7392 (and (looking-at c-primary-expr-regexp)
7393 (progn
7394 (setq pos (match-end 0))
7395 (or
7396 ;; Check if the expression begins with a prefix keyword.
7397 (match-beginning 2)
7398 (if (match-beginning 1)
7399 ;; Expression begins with an ambiguous operator. Treat
7400 ;; it as a cast if it's a type decl or if we've
7401 ;; recognized the type somewhere else.
7402 (or at-decl-or-cast
7403 (memq at-type '(t known found)))
7404 ;; Unless it's a keyword, it's the beginning of a primary
7405 ;; expression.
7406 (not (looking-at c-keywords-regexp)))))
7407 ;; If `c-primary-expr-regexp' matched a nonsymbol token, check
7408 ;; that it matched a whole one so that we don't e.g. confuse
7409 ;; the operator '-' with '->'. It's ok if it matches further,
7410 ;; though, since it e.g. can match the float '.5' while the
7411 ;; operator regexp only matches '.'.
7412 (or (not (looking-at c-nonsymbol-token-regexp))
7413 (<= (match-end 0) pos))))
7414
7415 ;; There should either be a cast before it or something that isn't an
7416 ;; identifier or close paren.
7417 (> preceding-token-end (point-min))
7418 (progn
7419 (goto-char (1- preceding-token-end))
7420 (or (eq (point) last-cast-end)
7421 (progn
7422 (c-backward-syntactic-ws)
7423 (if (< (skip-syntax-backward "w_") 0)
7424 ;; It's a symbol. Accept it only if it's one of the
7425 ;; keywords that can precede an expression (without
7426 ;; surrounding parens).
7427 (looking-at c-simple-stmt-key)
7428 (and
7429 ;; Check that it isn't a close paren (block close is ok,
7430 ;; though).
7431 (not (memq (char-before) '(?\) ?\])))
7432 ;; Check that it isn't a nonsymbol identifier.
7433 (not (c-on-identifier)))))))))
7434
7435 ;; Handle the cast.
7436 (when (and c-record-type-identifiers at-type (not (eq at-type t)))
7437 (let ((c-promote-possible-types t))
7438 (goto-char type-start)
7439 (c-forward-type)))
7440
7441 (goto-char cast-end)
7442 'cast)
7443
7444 (at-decl-or-cast
7445 ;; We're at a declaration. Highlight the type and the following
7446 ;; declarators.
7447
7448 (when backup-if-not-cast
7449 (c-fdoc-shift-type-backward t))
7450
7451 (when (and (eq context 'decl) (looking-at ","))
7452 ;; Make sure to propagate the `c-decl-arg-start' property to
7453 ;; the next argument if it's set in this one, to cope with
7454 ;; interactive refontification.
7455 (c-put-c-type-property (point) 'c-decl-arg-start))
7456
ee3ce8a7
AM
7457 (when (and c-record-type-identifiers at-type ;; (not (eq at-type t))
7458 ;; There seems no reason to exclude a token from
7459 ;; fontification just because it's "a known type that can't
7460 ;; be a name or other expression". 2013-09-18.
7461 )
0386b551
AM
7462 (let ((c-promote-possible-types t))
7463 (save-excursion
7464 (goto-char type-start)
7465 (c-forward-type))))
7466
e15f8aaa
AM
7467 (cons id-start
7468 (and (or at-type-decl at-typedef)
7469 (cons at-type-decl at-typedef))))
0386b551
AM
7470
7471 (t
7472 ;; False alarm. Restore the recorded ranges.
7473 (setq c-record-type-identifiers save-rec-type-ids
7474 c-record-ref-identifiers save-rec-ref-ids)
7475 nil))))
7476
7477(defun c-forward-label (&optional assume-markup preceding-token-end limit)
51c9af45 7478 ;; Assuming that point is at the beginning of a token, check if it starts a
1379f2c5
AM
7479 ;; label and if so move over it and return non-nil (t in default situations,
7480 ;; specific symbols (see below) for interesting situations), otherwise don't
7481 ;; move and return nil. "Label" here means "most things with a colon".
51c9af45
AM
7482 ;;
7483 ;; More precisely, a "label" is regarded as one of:
1379f2c5
AM
7484 ;; (i) a goto target like "foo:" - returns the symbol `goto-target';
7485 ;; (ii) A case label - either the entire construct "case FOO:", or just the
7486 ;; bare "case", should the colon be missing. We return t;
7487 ;; (iii) a keyword which needs a colon, like "default:" or "private:"; We
7488 ;; return t;
51c9af45 7489 ;; (iv) One of QT's "extended" C++ variants of
1379f2c5
AM
7490 ;; "private:"/"protected:"/"public:"/"more:" looking like "public slots:".
7491 ;; Returns the symbol `qt-2kwds-colon'.
7492 ;; (v) QT's construct "signals:". Returns the symbol `qt-1kwd-colon'.
82ba65cf 7493 ;; (vi) One of the keywords matched by `c-opt-extra-label-key' (without any
51c9af45 7494 ;; colon). Currently (2006-03), this applies only to Objective C's
1379f2c5 7495 ;; keywords "@private", "@protected", and "@public". Returns t.
51c9af45 7496 ;;
e1dbe924 7497 ;; One of the things which will NOT be recognized as a label is a bit-field
51c9af45
AM
7498 ;; element of a struct, something like "int foo:5".
7499 ;;
7500 ;; The end of the label is taken to be just after the colon, or the end of
7501 ;; the first submatch in `c-opt-extra-label-key'. The point is directly
7502 ;; after the end on return. The terminating char gets marked with
7503 ;; `c-decl-end' to improve recognition of the following declaration or
7504 ;; statement.
0386b551
AM
7505 ;;
7506 ;; If ASSUME-MARKUP is non-nil, it's assumed that the preceding
51c9af45 7507 ;; label, if any, has already been marked up like that.
0386b551
AM
7508 ;;
7509 ;; If PRECEDING-TOKEN-END is given, it should be the first position
7510 ;; after the preceding token, i.e. on the other side of the
7511 ;; syntactic ws from the point. Use a value less than or equal to
7512 ;; (point-min) if the point is at the first token in (the visible
7513 ;; part of) the buffer.
7514 ;;
7515 ;; The optional LIMIT limits the forward scan for the colon.
7516 ;;
7517 ;; This function records the ranges of the label symbols on
7518 ;; `c-record-ref-identifiers' if `c-record-type-identifiers' (!) is
7519 ;; non-nil.
7520 ;;
7521 ;; This function might do hidden buffer changes.
7522
51c9af45 7523 (let ((start (point))
1379f2c5 7524 label-end
51c9af45 7525 qt-symbol-idx
1379f2c5 7526 macro-start ; if we're in one.
f412a567
AM
7527 label-type
7528 kwd)
0386b551 7529 (cond
b414f371 7530 ;; "case" or "default" (Doesn't apply to AWK).
0386b551
AM
7531 ((looking-at c-label-kwds-regexp)
7532 (let ((kwd-end (match-end 1)))
7533 ;; Record only the keyword itself for fontification, since in
7534 ;; case labels the following is a constant expression and not
7535 ;; a label.
7536 (when c-record-type-identifiers
7537 (c-record-ref-id (cons (match-beginning 1) kwd-end)))
7538
7539 ;; Find the label end.
7540 (goto-char kwd-end)
1379f2c5
AM
7541 (setq label-type
7542 (if (and (c-syntactic-re-search-forward
7543 ;; Stop on chars that aren't allowed in expressions,
7544 ;; and on operator chars that would be meaningless
7545 ;; there. FIXME: This doesn't cope with ?: operators.
7546 "[;{=,@]\\|\\(\\=\\|[^:]\\):\\([^:]\\|\\'\\)"
7547 limit t t nil 1)
7548 (match-beginning 2))
7549
7550 (progn ; there's a proper :
7551 (goto-char (match-beginning 2)) ; just after the :
7552 (c-put-c-type-property (1- (point)) 'c-decl-end)
7553 t)
7554
7555 ;; It's an unfinished label. We consider the keyword enough
7556 ;; to recognize it as a label, so that it gets fontified.
7557 ;; Leave the point at the end of it, but don't put any
7558 ;; `c-decl-end' marker.
7559 (goto-char kwd-end)
7560 t))))
0386b551 7561
51c9af45 7562 ;; @private, @protected, @public, in Objective C, or similar.
0386b551
AM
7563 ((and c-opt-extra-label-key
7564 (looking-at c-opt-extra-label-key))
7565 ;; For a `c-opt-extra-label-key' match, we record the whole
7566 ;; thing for fontification. That's to get the leading '@' in
7567 ;; Objective-C protection labels fontified.
7568 (goto-char (match-end 1))
7569 (when c-record-type-identifiers
7570 (c-record-ref-id (cons (match-beginning 1) (point))))
7571 (c-put-c-type-property (1- (point)) 'c-decl-end)
1379f2c5 7572 (setq label-type t))
0386b551 7573
51c9af45
AM
7574 ;; All other cases of labels.
7575 ((and c-recognize-colon-labels ; nil for AWK and IDL, otherwise t.
0386b551
AM
7576
7577 ;; A colon label must have something before the colon.
7578 (not (eq (char-after) ?:))
7579
7580 ;; Check that we're not after a token that can't precede a label.
7581 (or
7582 ;; Trivially succeeds when there's no preceding token.
536610a4 7583 ;; Succeeds when we're at a virtual semicolon.
0386b551
AM
7584 (if preceding-token-end
7585 (<= preceding-token-end (point-min))
7586 (save-excursion
7587 (c-backward-syntactic-ws)
7588 (setq preceding-token-end (point))
536610a4
AM
7589 (or (bobp)
7590 (c-at-vsemi-p))))
0386b551
AM
7591
7592 ;; Check if we're after a label, if we're after a closing
7593 ;; paren that belong to statement, and with
7594 ;; `c-label-prefix-re'. It's done in different order
7595 ;; depending on `assume-markup' since the checks have
7596 ;; different expensiveness.
7597 (if assume-markup
7598 (or
7599 (eq (c-get-char-property (1- preceding-token-end) 'c-type)
7600 'c-decl-end)
7601
7602 (save-excursion
7603 (goto-char (1- preceding-token-end))
7604 (c-beginning-of-current-token)
51c9af45
AM
7605 (or (looking-at c-label-prefix-re)
7606 (looking-at c-block-stmt-1-key)))
0386b551
AM
7607
7608 (and (eq (char-before preceding-token-end) ?\))
7609 (c-after-conditional)))
7610
7611 (or
7612 (save-excursion
7613 (goto-char (1- preceding-token-end))
7614 (c-beginning-of-current-token)
51c9af45
AM
7615 (or (looking-at c-label-prefix-re)
7616 (looking-at c-block-stmt-1-key)))
0386b551
AM
7617
7618 (cond
7619 ((eq (char-before preceding-token-end) ?\))
7620 (c-after-conditional))
7621
7622 ((eq (char-before preceding-token-end) ?:)
7623 ;; Might be after another label, so check it recursively.
51c9af45
AM
7624 (save-restriction
7625 (save-excursion
7626 (goto-char (1- preceding-token-end))
7627 ;; Essentially the same as the
7628 ;; `c-syntactic-re-search-forward' regexp below.
7629 (setq macro-start
7630 (save-excursion (and (c-beginning-of-macro)
7631 (point))))
7632 (if macro-start (narrow-to-region macro-start (point-max)))
7633 (c-syntactic-skip-backward "^-]:?;}=*/%&|,<>!@+" nil t)
7634 ;; Note: the following should work instead of the
7635 ;; narrow-to-region above. Investigate why not,
7636 ;; sometime. ACM, 2006-03-31.
7637 ;; (c-syntactic-skip-backward "^-]:?;}=*/%&|,<>!@+"
7638 ;; macro-start t)
7639 (let ((pte (point))
7640 ;; If the caller turned on recording for us,
7641 ;; it shouldn't apply when we check the
7642 ;; preceding label.
7643 c-record-type-identifiers)
7644 ;; A label can't start at a cpp directive. Check for
7645 ;; this, since c-forward-syntactic-ws would foul up on it.
7646 (unless (and c-opt-cpp-prefix (looking-at c-opt-cpp-prefix))
7647 (c-forward-syntactic-ws)
7648 (c-forward-label nil pte start))))))))))
7649
1379f2c5 7650 ;; Point is still at the beginning of the possible label construct.
b414f371 7651 ;;
51c9af45
AM
7652 ;; Check that the next nonsymbol token is ":", or that we're in one
7653 ;; of QT's "slots" declarations. Allow '(' for the sake of macro
7654 ;; arguments. FIXME: Should build this regexp from the language
7655 ;; constants.
1379f2c5
AM
7656 (cond
7657 ;; public: protected: private:
7658 ((and
7659 (c-major-mode-is 'c++-mode)
7660 (search-forward-regexp
7661 "\\=p\\(r\\(ivate\\|otected\\)\\|ublic\\)\\>[^_]" nil t)
7662 (progn (backward-char)
7663 (c-forward-syntactic-ws limit)
7664 (looking-at ":\\([^:]\\|\\'\\)"))) ; A single colon.
7665 (forward-char)
7666 (setq label-type t))
7667 ;; QT double keyword like "protected slots:" or goto target.
7668 ((progn (goto-char start) nil))
7669 ((when (c-syntactic-re-search-forward
7670 "[ \t\n[:?;{=*/%&|,<>!@+-]" limit t t) ; not at EOB
7671 (backward-char)
7672 (setq label-end (point))
7673 (setq qt-symbol-idx
7674 (and (c-major-mode-is 'c++-mode)
7675 (string-match
7676 "\\(p\\(r\\(ivate\\|otected\\)\\|ublic\\)\\|more\\)\\>"
b414f371 7677 (buffer-substring start (point)))))
1379f2c5
AM
7678 (c-forward-syntactic-ws limit)
7679 (cond
7680 ((looking-at ":\\([^:]\\|\\'\\)") ; A single colon.
7681 (forward-char)
7682 (setq label-type
f412a567
AM
7683 (if (or (string= "signals" ; Special QT macro
7684 (setq kwd (buffer-substring-no-properties start label-end)))
7685 (string= "Q_SIGNALS" kwd))
1379f2c5
AM
7686 'qt-1kwd-colon
7687 'goto-target)))
7688 ((and qt-symbol-idx
f412a567 7689 (search-forward-regexp "\\=\\(slots\\|Q_SLOTS\\)\\>" limit t)
1379f2c5
AM
7690 (progn (c-forward-syntactic-ws limit)
7691 (looking-at ":\\([^:]\\|\\'\\)"))) ; A single colon
7692 (forward-char)
7693 (setq label-type 'qt-2kwds-colon)))))))
0386b551
AM
7694
7695 (save-restriction
7696 (narrow-to-region start (point))
7697
7698 ;; Check that `c-nonlabel-token-key' doesn't match anywhere.
7699 (catch 'check-label
7700 (goto-char start)
7701 (while (progn
7702 (when (looking-at c-nonlabel-token-key)
7703 (goto-char start)
1379f2c5 7704 (setq label-type nil)
0386b551
AM
7705 (throw 'check-label nil))
7706 (and (c-safe (c-forward-sexp)
7707 (c-forward-syntactic-ws)
7708 t)
7709 (not (eobp)))))
7710
7711 ;; Record the identifiers in the label for fontification, unless
7712 ;; it begins with `c-label-kwds' in which case the following
7713 ;; identifiers are part of a (constant) expression that
7714 ;; shouldn't be fontified.
7715 (when (and c-record-type-identifiers
7716 (progn (goto-char start)
7717 (not (looking-at c-label-kwds-regexp))))
7718 (while (c-syntactic-re-search-forward c-symbol-key nil t)
7719 (c-record-ref-id (cons (match-beginning 0)
7720 (match-end 0)))))
7721
7722 (c-put-c-type-property (1- (point-max)) 'c-decl-end)
82ba65cf 7723 (goto-char (point-max)))))
0386b551
AM
7724
7725 (t
7726 ;; Not a label.
1379f2c5
AM
7727 (goto-char start)))
7728 label-type))
0386b551
AM
7729
7730(defun c-forward-objc-directive ()
7731 ;; Assuming the point is at the beginning of a token, try to move
7732 ;; forward to the end of the Objective-C directive that starts
7733 ;; there. Return t if a directive was fully recognized, otherwise
7734 ;; the point is moved as far as one could be successfully parsed and
7735 ;; nil is returned.
7736 ;;
7737 ;; This function records identifier ranges on
7738 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
7739 ;; `c-record-type-identifiers' is non-nil.
7740 ;;
7741 ;; This function might do hidden buffer changes.
7742
7743 (let ((start (point))
7744 start-char
7745 (c-promote-possible-types t)
667ced3a 7746 lim
0386b551
AM
7747 ;; Turn off recognition of angle bracket arglists while parsing
7748 ;; types here since the protocol reference list might then be
7749 ;; considered part of the preceding name or superclass-name.
7750 c-recognize-<>-arglists)
7751
7752 (if (or
7753 (when (looking-at
7754 (eval-when-compile
7755 (c-make-keywords-re t
7756 (append (c-lang-const c-protection-kwds objc)
7757 '("@end"))
7758 'objc-mode)))
7759 (goto-char (match-end 1))
7760 t)
7761
7762 (and
7763 (looking-at
7764 (eval-when-compile
7765 (c-make-keywords-re t
7766 '("@interface" "@implementation" "@protocol")
7767 'objc-mode)))
7768
7769 ;; Handle the name of the class itself.
7770 (progn
cb694ab7
AM
7771; (c-forward-token-2) ; 2006/1/13 This doesn't move if the token's
7772; at EOB.
7773 (goto-char (match-end 0))
667ced3a 7774 (setq lim (point))
cb694ab7 7775 (c-skip-ws-forward)
0386b551
AM
7776 (c-forward-type))
7777
7778 (catch 'break
7779 ;; Look for ": superclass-name" or "( category-name )".
7780 (when (looking-at "[:\(]")
7781 (setq start-char (char-after))
7782 (forward-char)
7783 (c-forward-syntactic-ws)
7784 (unless (c-forward-type) (throw 'break nil))
7785 (when (eq start-char ?\()
7786 (unless (eq (char-after) ?\)) (throw 'break nil))
7787 (forward-char)
7788 (c-forward-syntactic-ws)))
7789
7790 ;; Look for a protocol reference list.
7791 (if (eq (char-after) ?<)
7792 (let ((c-recognize-<>-arglists t)
7793 (c-parse-and-markup-<>-arglists t)
7794 c-restricted-<>-arglists)
7795 (c-forward-<>-arglist t))
7796 t))))
7797
7798 (progn
667ced3a 7799 (c-backward-syntactic-ws lim)
0386b551
AM
7800 (c-clear-c-type-property start (1- (point)) 'c-decl-end)
7801 (c-put-c-type-property (1- (point)) 'c-decl-end)
7802 t)
7803
7804 (c-clear-c-type-property start (point) 'c-decl-end)
7805 nil)))
7806
785eecbb
RS
7807(defun c-beginning-of-inheritance-list (&optional lim)
7808 ;; Go to the first non-whitespace after the colon that starts a
7809 ;; multiple inheritance introduction. Optional LIM is the farthest
7810 ;; back we should search.
0386b551
AM
7811 ;;
7812 ;; This function might do hidden buffer changes.
7813 (c-with-syntax-table c++-template-syntax-table
7814 (c-backward-token-2 0 t lim)
7815 (while (and (or (looking-at c-symbol-start)
7816 (looking-at "[<,]\\|::"))
7817 (zerop (c-backward-token-2 1 t lim))))))
785eecbb 7818
785eecbb
RS
7819(defun c-in-method-def-p ()
7820 ;; Return nil if we aren't in a method definition, otherwise the
7821 ;; position of the initial [+-].
0386b551
AM
7822 ;;
7823 ;; This function might do hidden buffer changes.
785eecbb
RS
7824 (save-excursion
7825 (beginning-of-line)
a66cd3ee
MS
7826 (and c-opt-method-key
7827 (looking-at c-opt-method-key)
785eecbb
RS
7828 (point))
7829 ))
7830
a66cd3ee
MS
7831;; Contributed by Kevin Ryde <user42@zip.com.au>.
7832(defun c-in-gcc-asm-p ()
7833 ;; Return non-nil if point is within a gcc \"asm\" block.
7834 ;;
7835 ;; This should be called with point inside an argument list.
7836 ;;
7837 ;; Only one level of enclosing parentheses is considered, so for
7838 ;; instance `nil' is returned when in a function call within an asm
7839 ;; operand.
0386b551
AM
7840 ;;
7841 ;; This function might do hidden buffer changes.
a66cd3ee
MS
7842
7843 (and c-opt-asm-stmt-key
7844 (save-excursion
7845 (beginning-of-line)
7846 (backward-up-list 1)
7847 (c-beginning-of-statement-1 (point-min) nil t)
7848 (looking-at c-opt-asm-stmt-key))))
7849
abb7e5cf 7850(defun c-at-toplevel-p ()
a85fd6da
AM
7851 "Return a determination as to whether point is \"at the top level\".
7852Informally, \"at the top level\" is anywhere where you can write
7853a function.
7854
7855More precisely, being at the top-level means that point is either
7856outside any enclosing block (such as a function definition), or
7857directly inside a class, namespace or other block that contains
7858another declaration level.
abb7e5cf
SM
7859
7860If point is not at the top-level (e.g. it is inside a method
7861definition), then nil is returned. Otherwise, if point is at a
7862top-level not enclosed within a class definition, t is returned.
7863Otherwise, a 2-vector is returned where the zeroth element is the
7864buffer position of the start of the class declaration, and the first
7865element is the buffer position of the enclosing class's opening
0386b551
AM
7866brace.
7867
7868Note that this function might do hidden buffer changes. See the
7869comment at the start of cc-engine.el for more info."
a66cd3ee
MS
7870 (let ((paren-state (c-parse-state)))
7871 (or (not (c-most-enclosing-brace paren-state))
7872 (c-search-uplist-for-classkey paren-state))))
7873
d9e94c22 7874(defun c-just-after-func-arglist-p (&optional lim)
0386b551
AM
7875 ;; Return non-nil if the point is in the region after the argument
7876 ;; list of a function and its opening brace (or semicolon in case it
7877 ;; got no body). If there are K&R style argument declarations in
7878 ;; that region, the point has to be inside the first one for this
7879 ;; function to recognize it.
a66cd3ee 7880 ;;
0386b551
AM
7881 ;; If successful, the point is moved to the first token after the
7882 ;; function header (see `c-forward-decl-or-cast-1' for details) and
7883 ;; the position of the opening paren of the function arglist is
7884 ;; returned.
7885 ;;
7886 ;; The point is clobbered if not successful.
7887 ;;
7888 ;; LIM is used as bound for backward buffer searches.
7889 ;;
7890 ;; This function might do hidden buffer changes.
7891
7892 (let ((beg (point)) end id-start)
7893 (and
7894 (eq (c-beginning-of-statement-1 lim) 'same)
7895
2e492df3
AM
7896 (not (and (c-major-mode-is 'objc-mode)
7897 (c-forward-objc-directive)))
0386b551
AM
7898
7899 (setq id-start
7900 (car-safe (c-forward-decl-or-cast-1 (c-point 'bosws) nil nil)))
7901 (< id-start beg)
7902
7903 ;; There should not be a '=' or ',' between beg and the
7904 ;; start of the declaration since that means we were in the
7905 ;; "expression part" of the declaration.
7906 (or (> (point) beg)
7907 (not (looking-at "[=,]")))
7908
7909 (save-excursion
7910 ;; Check that there's an arglist paren in the
7911 ;; declaration.
7912 (goto-char id-start)
7913 (cond ((eq (char-after) ?\()
7914 ;; The declarator is a paren expression, so skip past it
7915 ;; so that we don't get stuck on that instead of the
7916 ;; function arglist.
7917 (c-forward-sexp))
51c9af45
AM
7918 ((and c-opt-op-identifier-prefix
7919 (looking-at c-opt-op-identifier-prefix))
0386b551
AM
7920 ;; Don't trip up on "operator ()".
7921 (c-forward-token-2 2 t)))
7922 (and (< (point) beg)
7923 (c-syntactic-re-search-forward "(" beg t t)
7924 (1- (point)))))))
785eecbb 7925
a66cd3ee
MS
7926(defun c-in-knr-argdecl (&optional lim)
7927 ;; Return the position of the first argument declaration if point is
7928 ;; inside a K&R style argument declaration list, nil otherwise.
7929 ;; `c-recognize-knr-p' is not checked. If LIM is non-nil, it's a
7930 ;; position that bounds the backward search for the argument list.
7931 ;;
9cf17ef1
AM
7932 ;; Point must be within a possible K&R region, e.g. just before a top-level
7933 ;; "{". It must be outside of parens and brackets. The test can return
7934 ;; false positives otherwise.
0386b551
AM
7935 ;;
7936 ;; This function might do hidden buffer changes.
d9e94c22 7937
a66cd3ee
MS
7938 (save-excursion
7939 (save-restriction
9cf17ef1
AM
7940 ;; If we're in a macro, our search range is restricted to it. Narrow to
7941 ;; the searchable range.
98d7371e
AM
7942 (let* ((macro-start (save-excursion (and (c-beginning-of-macro) (point))))
7943 (macro-end (save-excursion (and macro-start (c-end-of-macro) (point))))
7944 (low-lim (max (or lim (point-min)) (or macro-start (point-min))))
6b6481ed 7945 before-lparen after-rparen
98d7371e
AM
7946 (pp-count-out 20)) ; Max number of paren/brace constructs before
7947 ; we give up
7948 (narrow-to-region low-lim (or macro-end (point-max)))
9cf17ef1
AM
7949
7950 ;; Search backwards for the defun's argument list. We give up if we
98d7371e
AM
7951 ;; encounter a "}" (end of a previous defun) an "=" (which can't be in
7952 ;; a knr region) or BOB.
9cf17ef1
AM
7953 ;;
7954 ;; The criterion for a paren structure being the arg list is:
7955 ;; o - there is non-WS stuff after it but before any "{"; AND
7956 ;; o - the token after it isn't a ";" AND
7957 ;; o - it is preceded by either an identifier (the function name) or
7958 ;; a macro expansion like "DEFUN (...)"; AND
7959 ;; o - its content is a non-empty comma-separated list of identifiers
7960 ;; (an empty arg list won't have a knr region).
7961 ;;
7962 ;; The following snippet illustrates these rules:
7963 ;; int foo (bar, baz, yuk)
7964 ;; int bar [] ;
7965 ;; int (*baz) (my_type) ;
7966 ;; int (*) (void) (*yuk) (void) ;
7967 ;; {
7968
7969 (catch 'knr
6b6481ed
AM
7970 (while (> pp-count-out 0) ; go back one paren/bracket pair each time.
7971 (setq pp-count-out (1- pp-count-out))
98d7371e 7972 (c-syntactic-skip-backward "^)]}=")
9cf17ef1
AM
7973 (cond ((eq (char-before) ?\))
7974 (setq after-rparen (point)))
10489fcb
AM
7975 ((eq (char-before) ?\])
7976 (setq after-rparen nil))
98d7371e
AM
7977 (t ; either } (hit previous defun) or = or no more
7978 ; parens/brackets.
10489fcb 7979 (throw 'knr nil)))
9cf17ef1
AM
7980
7981 (if after-rparen
7982 ;; We're inside a paren. Could it be our argument list....?
7983 (if
7984 (and
7985 (progn
7986 (goto-char after-rparen)
7987 (unless (c-go-list-backward) (throw 'knr nil)) ;
7988 ;; FIXME!!! What about macros between the parens? 2007/01/20
7989 (setq before-lparen (point)))
d9e94c22 7990
9cf17ef1
AM
7991 ;; It can't be the arg list if next token is ; or {
7992 (progn (goto-char after-rparen)
7993 (c-forward-syntactic-ws)
98d7371e 7994 (not (memq (char-after) '(?\; ?\{ ?\=))))
d9e94c22 7995
9cf17ef1
AM
7996 ;; Is the thing preceding the list an identifier (the
7997 ;; function name), or a macro expansion?
7998 (progn
7999 (goto-char before-lparen)
8000 (eq (c-backward-token-2) 0)
98d7371e 8001 (or (eq (c-on-identifier) (point))
9cf17ef1
AM
8002 (and (eq (char-after) ?\))
8003 (c-go-up-list-backward)
8004 (eq (c-backward-token-2) 0)
98d7371e 8005 (eq (c-on-identifier) (point)))))
9cf17ef1
AM
8006
8007 ;; Have we got a non-empty list of comma-separated
8008 ;; identifiers?
8009 (progn
8010 (goto-char before-lparen)
8011 (c-forward-token-2) ; to first token inside parens
8012 (and
8013 (c-on-identifier)
8014 (c-forward-token-2)
8015 (catch 'id-list
8016 (while (eq (char-after) ?\,)
8017 (c-forward-token-2)
8018 (unless (c-on-identifier) (throw 'id-list nil))
8019 (c-forward-token-2))
8020 (eq (char-after) ?\))))))
8021
8022 ;; ...Yes. We've identified the function's argument list.
8023 (throw 'knr
8024 (progn (goto-char after-rparen)
8025 (c-forward-syntactic-ws)
8026 (point)))
8027
8028 ;; ...No. The current parens aren't the function's arg list.
8029 (goto-char before-lparen))
8030
8031 (or (c-go-list-backward) ; backwards over [ .... ]
8032 (throw 'knr nil)))))))))
785eecbb
RS
8033
8034(defun c-skip-conditional ()
8035 ;; skip forward over conditional at point, including any predicate
8036 ;; statements in parentheses. No error checking is performed.
0386b551
AM
8037 ;;
8038 ;; This function might do hidden buffer changes.
0ec8351b
BW
8039 (c-forward-sexp (cond
8040 ;; else if()
a66cd3ee
MS
8041 ((looking-at (concat "\\<else"
8042 "\\([ \t\n]\\|\\\\\n\\)+"
8043 "if\\>\\([^_]\\|$\\)"))
8044 3)
0ec8351b 8045 ;; do, else, try, finally
a66cd3ee
MS
8046 ((looking-at (concat "\\<\\("
8047 "do\\|else\\|try\\|finally"
8048 "\\)\\>\\([^_]\\|$\\)"))
130c507e 8049 1)
ce8c7486 8050 ;; for, if, while, switch, catch, synchronized, foreach
0ec8351b 8051 (t 2))))
785eecbb 8052
a66cd3ee
MS
8053(defun c-after-conditional (&optional lim)
8054 ;; If looking at the token after a conditional then return the
8055 ;; position of its start, otherwise return nil.
0386b551
AM
8056 ;;
8057 ;; This function might do hidden buffer changes.
a66cd3ee 8058 (save-excursion
d9e94c22 8059 (and (zerop (c-backward-token-2 1 t lim))
a66cd3ee
MS
8060 (or (looking-at c-block-stmt-1-key)
8061 (and (eq (char-after) ?\()
d9e94c22 8062 (zerop (c-backward-token-2 1 t lim))
5ab78d3d
AM
8063 (or (looking-at c-block-stmt-2-key)
8064 (looking-at c-block-stmt-1-2-key))))
a66cd3ee
MS
8065 (point))))
8066
0386b551
AM
8067(defun c-after-special-operator-id (&optional lim)
8068 ;; If the point is after an operator identifier that isn't handled
8069 ;; like an ordinary symbol (i.e. like "operator =" in C++) then the
8070 ;; position of the start of that identifier is returned. nil is
8071 ;; returned otherwise. The point may be anywhere in the syntactic
8072 ;; whitespace after the last token of the operator identifier.
8073 ;;
8074 ;; This function might do hidden buffer changes.
8075 (save-excursion
8076 (and c-overloadable-operators-regexp
8077 (zerop (c-backward-token-2 1 nil lim))
8078 (looking-at c-overloadable-operators-regexp)
51c9af45 8079 (or (not c-opt-op-identifier-prefix)
0386b551
AM
8080 (and
8081 (zerop (c-backward-token-2 1 nil lim))
51c9af45 8082 (looking-at c-opt-op-identifier-prefix)))
0386b551
AM
8083 (point))))
8084
a66cd3ee
MS
8085(defsubst c-backward-to-block-anchor (&optional lim)
8086 ;; Assuming point is at a brace that opens a statement block of some
8087 ;; kind, move to the proper anchor point for that block. It might
8088 ;; need to be adjusted further by c-add-stmt-syntax, but the
8089 ;; position at return is suitable as start position for that
8090 ;; function.
0386b551
AM
8091 ;;
8092 ;; This function might do hidden buffer changes.
a66cd3ee
MS
8093 (unless (= (point) (c-point 'boi))
8094 (let ((start (c-after-conditional lim)))
8095 (if start
8096 (goto-char start)))))
8097
037558bf 8098(defsubst c-backward-to-decl-anchor (&optional lim)
a66cd3ee
MS
8099 ;; Assuming point is at a brace that opens the block of a top level
8100 ;; declaration of some kind, move to the proper anchor point for
8101 ;; that block.
0386b551
AM
8102 ;;
8103 ;; This function might do hidden buffer changes.
a66cd3ee 8104 (unless (= (point) (c-point 'boi))
037558bf 8105 (c-beginning-of-statement-1 lim)))
a66cd3ee 8106
ff959bab 8107(defun c-search-decl-header-end ()
a66cd3ee
MS
8108 ;; Search forward for the end of the "header" of the current
8109 ;; declaration. That's the position where the definition body
8110 ;; starts, or the first variable initializer, or the ending
8111 ;; semicolon. I.e. search forward for the closest following
8112 ;; (syntactically relevant) '{', '=' or ';' token. Point is left
8113 ;; _after_ the first found token, or at point-max if none is found.
0386b551
AM
8114 ;;
8115 ;; This function might do hidden buffer changes.
ff959bab
MS
8116
8117 (let ((base (point)))
8118 (if (c-major-mode-is 'c++-mode)
8119
8120 ;; In C++ we need to take special care to handle operator
8121 ;; tokens and those pesky template brackets.
8122 (while (and
8123 (c-syntactic-re-search-forward "[;{<=]" nil 'move t t)
8124 (or
8125 (c-end-of-current-token base)
8126 ;; Handle operator identifiers, i.e. ignore any
8127 ;; operator token preceded by "operator".
8128 (save-excursion
8129 (and (c-safe (c-backward-sexp) t)
51c9af45 8130 (looking-at c-opt-op-identifier-prefix)))
ff959bab
MS
8131 (and (eq (char-before) ?<)
8132 (c-with-syntax-table c++-template-syntax-table
8133 (if (c-safe (goto-char (c-up-list-forward (point))))
8134 t
8135 (goto-char (point-max))
8136 nil)))))
8137 (setq base (point)))
8138
8139 (while (and
8140 (c-syntactic-re-search-forward "[;{=]" nil 'move t t)
8141 (c-end-of-current-token base))
8142 (setq base (point))))))
a66cd3ee
MS
8143
8144(defun c-beginning-of-decl-1 (&optional lim)
8145 ;; Go to the beginning of the current declaration, or the beginning
8146 ;; of the previous one if already at the start of it. Point won't
0386b551 8147 ;; be moved out of any surrounding paren. Return a cons cell of the
a66cd3ee
MS
8148 ;; form (MOVE . KNR-POS). MOVE is like the return value from
8149 ;; `c-beginning-of-statement-1'. If point skipped over some K&R
8150 ;; style argument declarations (and they are to be recognized) then
8151 ;; KNR-POS is set to the start of the first such argument
8152 ;; declaration, otherwise KNR-POS is nil. If LIM is non-nil, it's a
8153 ;; position that bounds the backward search.
8154 ;;
8155 ;; NB: Cases where the declaration continues after the block, as in
8156 ;; "struct foo { ... } bar;", are currently recognized as two
8157 ;; declarations, e.g. "struct foo { ... }" and "bar;" in this case.
0386b551
AM
8158 ;;
8159 ;; This function might do hidden buffer changes.
a66cd3ee
MS
8160 (catch 'return
8161 (let* ((start (point))
d9e94c22 8162 (last-stmt-start (point))
0386b551 8163 (move (c-beginning-of-statement-1 lim nil t)))
a66cd3ee 8164
a66cd3ee
MS
8165 ;; `c-beginning-of-statement-1' stops at a block start, but we
8166 ;; want to continue if the block doesn't begin a top level
2a15eb73
MS
8167 ;; construct, i.e. if it isn't preceded by ';', '}', ':', bob,
8168 ;; or an open paren.
d9e94c22 8169 (let ((beg (point)) tentative-move)
51c9af45
AM
8170 ;; Go back one "statement" each time round the loop until we're just
8171 ;; after a ;, }, or :, or at BOB or the start of a macro or start of
8172 ;; an ObjC method. This will move over a multiple declaration whose
8173 ;; components are comma separated.
d9e94c22
MS
8174 (while (and
8175 ;; Must check with c-opt-method-key in ObjC mode.
8176 (not (and c-opt-method-key
8177 (looking-at c-opt-method-key)))
8178 (/= last-stmt-start (point))
8179 (progn
8180 (c-backward-syntactic-ws lim)
8181 (not (memq (char-before) '(?\; ?} ?: nil))))
2a15eb73
MS
8182 (save-excursion
8183 (backward-char)
8184 (not (looking-at "\\s(")))
d9e94c22
MS
8185 ;; Check that we don't move from the first thing in a
8186 ;; macro to its header.
8187 (not (eq (setq tentative-move
0386b551 8188 (c-beginning-of-statement-1 lim nil t))
d9e94c22
MS
8189 'macro)))
8190 (setq last-stmt-start beg
8191 beg (point)
8192 move tentative-move))
8193 (goto-char beg))
8194
8195 (when c-recognize-knr-p
8196 (let ((fallback-pos (point)) knr-argdecl-start)
8197 ;; Handle K&R argdecls. Back up after the "statement" jumped
8198 ;; over by `c-beginning-of-statement-1', unless it was the
8199 ;; function body, in which case we're sitting on the opening
8200 ;; brace now. Then test if we're in a K&R argdecl region and
8201 ;; that we started at the other side of the first argdecl in
8202 ;; it.
8203 (unless (eq (char-after) ?{)
8204 (goto-char last-stmt-start))
8205 (if (and (setq knr-argdecl-start (c-in-knr-argdecl lim))
8206 (< knr-argdecl-start start)
8207 (progn
8208 (goto-char knr-argdecl-start)
0386b551 8209 (not (eq (c-beginning-of-statement-1 lim nil t) 'macro))))
d9e94c22
MS
8210 (throw 'return
8211 (cons (if (eq (char-after fallback-pos) ?{)
8212 'previous
8213 'same)
8214 knr-argdecl-start))
8215 (goto-char fallback-pos))))
8216
51c9af45
AM
8217 ;; `c-beginning-of-statement-1' counts each brace block as a separate
8218 ;; statement, so the result will be 'previous if we've moved over any.
8219 ;; So change our result back to 'same if necessary.
8220 ;;
8221 ;; If they were brace list initializers we might not have moved over a
8222 ;; declaration boundary though, so change it to 'same if we've moved
8223 ;; past a '=' before '{', but not ';'. (This ought to be integrated
8224 ;; into `c-beginning-of-statement-1', so we avoid this extra pass which
8225 ;; potentially can search over a large amount of text.). Take special
8226 ;; pains not to get mislead by C++'s "operator=", and the like.
d9e94c22
MS
8227 (if (and (eq move 'previous)
8228 (c-with-syntax-table (if (c-major-mode-is 'c++-mode)
8229 c++-template-syntax-table
8230 (syntax-table))
8231 (save-excursion
51c9af45
AM
8232 (and
8233 (progn
8234 (while ; keep going back to "[;={"s until we either find
8235 ; no more, or get to one which isn't an "operator ="
8236 (and (c-syntactic-re-search-forward "[;={]" start t t t)
8237 (eq (char-before) ?=)
8238 c-overloadable-operators-regexp
8239 c-opt-op-identifier-prefix
8240 (save-excursion
8241 (eq (c-backward-token-2) 0)
8242 (looking-at c-overloadable-operators-regexp)
8243 (eq (c-backward-token-2) 0)
8244 (looking-at c-opt-op-identifier-prefix))))
8245 (eq (char-before) ?=))
8246 (c-syntactic-re-search-forward "[;{]" start t t)
8247 (eq (char-before) ?{)
8248 (c-safe (goto-char (c-up-list-forward (point))) t)
8249 (not (c-syntactic-re-search-forward ";" start t t))))))
d9e94c22
MS
8250 (cons 'same nil)
8251 (cons move nil)))))
a66cd3ee
MS
8252
8253(defun c-end-of-decl-1 ()
8254 ;; Assuming point is at the start of a declaration (as detected by
8255 ;; e.g. `c-beginning-of-decl-1'), go to the end of it. Unlike
8256 ;; `c-beginning-of-decl-1', this function handles the case when a
8257 ;; block is followed by identifiers in e.g. struct declarations in C
8258 ;; or C++. If a proper end was found then t is returned, otherwise
8259 ;; point is moved as far as possible within the current sexp and nil
8260 ;; is returned. This function doesn't handle macros; use
8261 ;; `c-end-of-macro' instead in those cases.
0386b551
AM
8262 ;;
8263 ;; This function might do hidden buffer changes.
ce8c7486 8264 (let ((start (point))
a66cd3ee
MS
8265 (decl-syntax-table (if (c-major-mode-is 'c++-mode)
8266 c++-template-syntax-table
8267 (syntax-table))))
8268 (catch 'return
8269 (c-search-decl-header-end)
8270
8271 (when (and c-recognize-knr-p
8272 (eq (char-before) ?\;)
8273 (c-in-knr-argdecl start))
8274 ;; Stopped at the ';' in a K&R argdecl section which is
8275 ;; detected using the same criteria as in
8276 ;; `c-beginning-of-decl-1'. Move to the following block
8277 ;; start.
d9e94c22 8278 (c-syntactic-re-search-forward "{" nil 'move t))
a66cd3ee
MS
8279
8280 (when (eq (char-before) ?{)
8281 ;; Encountered a block in the declaration. Jump over it.
8282 (condition-case nil
8283 (goto-char (c-up-list-forward (point)))
d9e94c22
MS
8284 (error (goto-char (point-max))
8285 (throw 'return nil)))
a66cd3ee
MS
8286 (if (or (not c-opt-block-decls-with-vars-key)
8287 (save-excursion
8288 (c-with-syntax-table decl-syntax-table
8289 (let ((lim (point)))
8290 (goto-char start)
b3cf7e18
MS
8291 (not (and
8292 ;; Check for `c-opt-block-decls-with-vars-key'
8293 ;; before the first paren.
8294 (c-syntactic-re-search-forward
d9e94c22 8295 (concat "[;=\(\[{]\\|\\("
b3cf7e18
MS
8296 c-opt-block-decls-with-vars-key
8297 "\\)")
d9e94c22 8298 lim t t t)
b3cf7e18
MS
8299 (match-beginning 1)
8300 (not (eq (char-before) ?_))
d9e94c22
MS
8301 ;; Check that the first following paren is
8302 ;; the block.
8303 (c-syntactic-re-search-forward "[;=\(\[{]"
8304 lim t t t)
b3cf7e18 8305 (eq (char-before) ?{)))))))
a66cd3ee
MS
8306 ;; The declaration doesn't have any of the
8307 ;; `c-opt-block-decls-with-vars' keywords in the
8308 ;; beginning, so it ends here at the end of the block.
8309 (throw 'return t)))
8310
8311 (c-with-syntax-table decl-syntax-table
8312 (while (progn
8313 (if (eq (char-before) ?\;)
8314 (throw 'return t))
d9e94c22 8315 (c-syntactic-re-search-forward ";" nil 'move t))))
a66cd3ee 8316 nil)))
ce8c7486 8317
0386b551
AM
8318(defun c-looking-at-decl-block (containing-sexp goto-start &optional limit)
8319 ;; Assuming the point is at an open brace, check if it starts a
8320 ;; block that contains another declaration level, i.e. that isn't a
8321 ;; statement block or a brace list, and if so return non-nil.
8322 ;;
8323 ;; If the check is successful, the return value is the start of the
8324 ;; keyword that tells what kind of construct it is, i.e. typically
8325 ;; what `c-decl-block-key' matched. Also, if GOTO-START is set then
8326 ;; the point will be at the start of the construct, before any
8327 ;; leading specifiers, otherwise it's at the returned position.
8328 ;;
8329 ;; The point is clobbered if the check is unsuccessful.
8330 ;;
8331 ;; CONTAINING-SEXP is the position of the open of the surrounding
8332 ;; paren, or nil if none.
8333 ;;
8334 ;; The optional LIMIT limits the backward search for the start of
8335 ;; the construct. It's assumed to be at a syntactically relevant
8336 ;; position.
8337 ;;
8338 ;; If any template arglists are found in the searched region before
8339 ;; the open brace, they get marked with paren syntax.
8340 ;;
8341 ;; This function might do hidden buffer changes.
8342
8343 (let ((open-brace (point)) kwd-start first-specifier-pos)
8344 (c-syntactic-skip-backward c-block-prefix-charset limit t)
8345
8346 (when (and c-recognize-<>-arglists
8347 (eq (char-before) ?>))
8348 ;; Could be at the end of a template arglist.
8349 (let ((c-parse-and-markup-<>-arglists t)
8350 (c-disallow-comma-in-<>-arglists
8351 (and containing-sexp
8352 (not (eq (char-after containing-sexp) ?{)))))
8353 (while (and
8354 (c-backward-<>-arglist nil limit)
8355 (progn
8356 (c-syntactic-skip-backward c-block-prefix-charset limit t)
8357 (eq (char-before) ?>))))))
8358
8359 ;; Note: Can't get bogus hits inside template arglists below since they
8360 ;; have gotten paren syntax above.
8361 (when (and
8362 ;; If `goto-start' is set we begin by searching for the
8363 ;; first possible position of a leading specifier list.
8364 ;; The `c-decl-block-key' search continues from there since
8365 ;; we know it can't match earlier.
8366 (if goto-start
8367 (when (c-syntactic-re-search-forward c-symbol-start
8368 open-brace t t)
8369 (goto-char (setq first-specifier-pos (match-beginning 0)))
8370 t)
8371 t)
8372
8373 (cond
8374 ((c-syntactic-re-search-forward c-decl-block-key open-brace t t t)
8375 (goto-char (setq kwd-start (match-beginning 0)))
8376 (or
8377
8378 ;; Found a keyword that can't be a type?
8379 (match-beginning 1)
8380
8381 ;; Can be a type too, in which case it's the return type of a
8382 ;; function (under the assumption that no declaration level
8383 ;; block construct starts with a type).
8384 (not (c-forward-type))
8385
8386 ;; Jumped over a type, but it could be a declaration keyword
8387 ;; followed by the declared identifier that we've jumped over
8388 ;; instead (e.g. in "class Foo {"). If it indeed is a type
8389 ;; then we should be at the declarator now, so check for a
8390 ;; valid declarator start.
8391 ;;
8392 ;; Note: This doesn't cope with the case when a declared
8393 ;; identifier is followed by e.g. '(' in a language where '('
8394 ;; also might be part of a declarator expression. Currently
8395 ;; there's no such language.
8396 (not (or (looking-at c-symbol-start)
8397 (looking-at c-type-decl-prefix-key)))))
8398
8399 ;; In Pike a list of modifiers may be followed by a brace
8400 ;; to make them apply to many identifiers. Note that the
8401 ;; match data will be empty on return in this case.
8402 ((and (c-major-mode-is 'pike-mode)
8403 (progn
8404 (goto-char open-brace)
8405 (= (c-backward-token-2) 0))
8406 (looking-at c-specifier-key)
8407 ;; Use this variant to avoid yet another special regexp.
8408 (c-keyword-member (c-keyword-sym (match-string 1))
8409 'c-modifier-kwds))
8410 (setq kwd-start (point))
8411 t)))
8412
8413 ;; Got a match.
8414
8415 (if goto-start
8416 ;; Back up over any preceding specifiers and their clauses
8417 ;; by going forward from `first-specifier-pos', which is the
8418 ;; earliest possible position where the specifier list can
8419 ;; start.
8420 (progn
8421 (goto-char first-specifier-pos)
8422
8423 (while (< (point) kwd-start)
8424 (if (looking-at c-symbol-key)
8425 ;; Accept any plain symbol token on the ground that
8426 ;; it's a specifier masked through a macro (just
8427 ;; like `c-forward-decl-or-cast-1' skip forward over
8428 ;; such tokens).
8429 ;;
8430 ;; Could be more restrictive wrt invalid keywords,
8431 ;; but that'd only occur in invalid code so there's
8432 ;; no use spending effort on it.
8433 (let ((end (match-end 0)))
8434 (unless (c-forward-keyword-clause 0)
8435 (goto-char end)
8436 (c-forward-syntactic-ws)))
8437
8438 ;; Can't parse a declaration preamble and is still
8439 ;; before `kwd-start'. That means `first-specifier-pos'
8440 ;; was in some earlier construct. Search again.
8441 (if (c-syntactic-re-search-forward c-symbol-start
8442 kwd-start 'move t)
8443 (goto-char (setq first-specifier-pos (match-beginning 0)))
8444 ;; Got no preamble before the block declaration keyword.
8445 (setq first-specifier-pos kwd-start))))
8446
8447 (goto-char first-specifier-pos))
8448 (goto-char kwd-start))
8449
8450 kwd-start)))
ce8c7486 8451
a66cd3ee 8452(defun c-search-uplist-for-classkey (paren-state)
0386b551
AM
8453 ;; Check if the closest containing paren sexp is a declaration
8454 ;; block, returning a 2 element vector in that case. Aref 0
8455 ;; contains the bufpos at boi of the class key line, and aref 1
8456 ;; contains the bufpos of the open brace. This function is an
8457 ;; obsolete wrapper for `c-looking-at-decl-block'.
8458 ;;
8459 ;; This function might do hidden buffer changes.
8460 (let ((open-paren-pos (c-most-enclosing-brace paren-state)))
8461 (when open-paren-pos
8462 (save-excursion
8463 (goto-char open-paren-pos)
8464 (when (and (eq (char-after) ?{)
8465 (c-looking-at-decl-block
8466 (c-safe-position open-paren-pos paren-state)
8467 nil))
8468 (back-to-indentation)
8469 (vector (point) open-paren-pos))))))
785eecbb 8470
020716e1
AM
8471(defun c-most-enclosing-decl-block (paren-state)
8472 ;; Return the buffer position of the most enclosing decl-block brace (in the
8473 ;; sense of c-looking-at-decl-block) in the PAREN-STATE structure, or nil if
8474 ;; none was found.
8475 (let* ((open-brace (c-pull-open-brace paren-state))
8476 (next-open-brace (c-pull-open-brace paren-state)))
8477 (while (and open-brace
8478 (save-excursion
8479 (goto-char open-brace)
8480 (not (c-looking-at-decl-block next-open-brace nil))))
8481 (setq open-brace next-open-brace
8482 next-open-brace (c-pull-open-brace paren-state)))
8483 open-brace))
8484
9657183b
AM
8485(defun c-cheap-inside-bracelist-p (paren-state)
8486 ;; Return the position of the L-brace if point is inside a brace list
8487 ;; initialization of an array, etc. This is an approximate function,
8488 ;; designed for speed over accuracy. It will not find every bracelist, but
8489 ;; a non-nil result is reliable. We simply search for "= {" (naturally with
8490 ;; syntactic whitespace allowed). PAREN-STATE is the normal thing that it
8491 ;; is everywhere else.
8492 (let (b-pos)
8493 (save-excursion
8494 (while
8495 (and (setq b-pos (c-pull-open-brace paren-state))
8496 (progn (goto-char b-pos)
8497 (c-backward-sws)
8498 (c-backward-token-2)
8499 (not (looking-at "=")))))
8500 b-pos)))
8501
7a068717
AM
8502(defun c-backward-over-enum-header ()
8503 ;; We're at a "{". Move back to the enum-like keyword that starts this
8504 ;; declaration and return t, otherwise don't move and return nil.
8505 (let ((here (point))
8506 up-sexp-pos before-identifier)
8507 (while
8508 (and
8509 (eq (c-backward-token-2) 0)
8510 (or (not (looking-at "\\s)"))
8511 (c-go-up-list-backward))
8512 (cond
87756ca9
AM
8513 ((and (looking-at c-symbol-key) (c-on-identifier)
8514 (not before-identifier))
7a068717
AM
8515 (setq before-identifier t))
8516 ((and before-identifier
87756ca9
AM
8517 (or (eq (char-after) ?,)
8518 (looking-at c-postfix-decl-spec-key)))
7a068717
AM
8519 (setq before-identifier nil)
8520 t)
8521 ((looking-at c-brace-list-key) nil)
8522 ((and c-recognize-<>-arglists
8523 (eq (char-after) ?<)
8524 (looking-at "\\s("))
8525 t)
8526 (t nil))))
8527 (or (looking-at c-brace-list-key)
8528 (progn (goto-char here) nil))))
8529
a66cd3ee 8530(defun c-inside-bracelist-p (containing-sexp paren-state)
785eecbb
RS
8531 ;; return the buffer position of the beginning of the brace list
8532 ;; statement if we're inside a brace list, otherwise return nil.
8533 ;; CONTAINING-SEXP is the buffer pos of the innermost containing
0386b551 8534 ;; paren. PAREN-STATE is the remainder of the state of enclosing
130c507e 8535 ;; braces
785eecbb
RS
8536 ;;
8537 ;; N.B.: This algorithm can potentially get confused by cpp macros
0386b551 8538 ;; placed in inconvenient locations. It's a trade-off we make for
785eecbb 8539 ;; speed.
0386b551
AM
8540 ;;
8541 ;; This function might do hidden buffer changes.
785eecbb 8542 (or
d9e94c22 8543 ;; This will pick up brace list declarations.
7a068717
AM
8544 (save-excursion
8545 (goto-char containing-sexp)
8546 (c-backward-over-enum-header))
785eecbb
RS
8547 ;; this will pick up array/aggregate init lists, even if they are nested.
8548 (save-excursion
0ec8351b
BW
8549 (let ((class-key
8550 ;; Pike can have class definitions anywhere, so we must
8551 ;; check for the class key here.
8552 (and (c-major-mode-is 'pike-mode)
a66cd3ee 8553 c-decl-block-key))
8a51e842 8554 bufpos braceassignp lim next-containing macro-start)
785eecbb
RS
8555 (while (and (not bufpos)
8556 containing-sexp)
8a51e842 8557 (when paren-state
a66cd3ee
MS
8558 (if (consp (car paren-state))
8559 (setq lim (cdr (car paren-state))
8560 paren-state (cdr paren-state))
8561 (setq lim (car paren-state)))
8562 (when paren-state
8563 (setq next-containing (car paren-state)
8564 paren-state (cdr paren-state))))
785eecbb 8565 (goto-char containing-sexp)
a66cd3ee
MS
8566 (if (c-looking-at-inexpr-block next-containing next-containing)
8567 ;; We're in an in-expression block of some kind. Do not
8568 ;; check nesting. We deliberately set the limit to the
8569 ;; containing sexp, so that c-looking-at-inexpr-block
8570 ;; doesn't check for an identifier before it.
0ec8351b
BW
8571 (setq containing-sexp nil)
8572 ;; see if the open brace is preceded by = or [...] in
8573 ;; this statement, but watch out for operator=
a66cd3ee 8574 (setq braceassignp 'dontknow)
d9e94c22 8575 (c-backward-token-2 1 t lim)
6393fef2 8576 ;; Checks to do only on the first sexp before the brace.
d9e94c22 8577 (when (and c-opt-inexpr-brace-list-key
6393fef2
RS
8578 (eq (char-after) ?\[))
8579 ;; In Java, an initialization brace list may follow
8580 ;; directly after "new Foo[]", so check for a "new"
8581 ;; earlier.
8582 (while (eq braceassignp 'dontknow)
8583 (setq braceassignp
d9e94c22
MS
8584 (cond ((/= (c-backward-token-2 1 t lim) 0) nil)
8585 ((looking-at c-opt-inexpr-brace-list-key) t)
6393fef2
RS
8586 ((looking-at "\\sw\\|\\s_\\|[.[]")
8587 ;; Carry on looking if this is an
8588 ;; identifier (may contain "." in Java)
8589 ;; or another "[]" sexp.
8590 'dontknow)
8591 (t nil)))))
8592 ;; Checks to do on all sexps before the brace, up to the
8593 ;; beginning of the statement.
8594 (while (eq braceassignp 'dontknow)
0ec8351b
BW
8595 (cond ((eq (char-after) ?\;)
8596 (setq braceassignp nil))
8597 ((and class-key
8598 (looking-at class-key))
8599 (setq braceassignp nil))
8600 ((eq (char-after) ?=)
8601 ;; We've seen a =, but must check earlier tokens so
8602 ;; that it isn't something that should be ignored.
8603 (setq braceassignp 'maybe)
8604 (while (and (eq braceassignp 'maybe)
d9e94c22 8605 (zerop (c-backward-token-2 1 t lim)))
0ec8351b
BW
8606 (setq braceassignp
8607 (cond
8608 ;; Check for operator =
51c9af45
AM
8609 ((and c-opt-op-identifier-prefix
8610 (looking-at c-opt-op-identifier-prefix))
0386b551 8611 nil)
130c507e
GM
8612 ;; Check for `<opchar>= in Pike.
8613 ((and (c-major-mode-is 'pike-mode)
8614 (or (eq (char-after) ?`)
8615 ;; Special case for Pikes
8616 ;; `[]=, since '[' is not in
8617 ;; the punctuation class.
8618 (and (eq (char-after) ?\[)
8619 (eq (char-before) ?`))))
8620 nil)
0ec8351b
BW
8621 ((looking-at "\\s.") 'maybe)
8622 ;; make sure we're not in a C++ template
8623 ;; argument assignment
a66cd3ee
MS
8624 ((and
8625 (c-major-mode-is 'c++-mode)
8626 (save-excursion
8627 (let ((here (point))
8628 (pos< (progn
8629 (skip-chars-backward "^<>")
8630 (point))))
8631 (and (eq (char-before) ?<)
8632 (not (c-crosses-statement-barrier-p
8633 pos< here))
8634 (not (c-in-literal))
8635 ))))
0ec8351b 8636 nil)
6393fef2 8637 (t t))))))
8a51e842
AM
8638 (if (and (eq braceassignp 'dontknow)
8639 (/= (c-backward-token-2 1 t lim) 0))
8640 (setq braceassignp nil)))
8641 (cond
8642 (braceassignp
8643 ;; We've hit the beginning of the aggregate list.
8644 (c-beginning-of-statement-1
8645 (c-most-enclosing-brace paren-state))
8646 (setq bufpos (point)))
8647 ((eq (char-after) ?\;)
8648 ;; Brace lists can't contain a semicolon, so we're done.
8649 (setq containing-sexp nil))
8650 ((and (setq macro-start (point))
8651 (c-forward-to-cpp-define-body)
8652 (eq (point) containing-sexp))
8653 ;; We've a macro whose expansion starts with the '{'.
8654 ;; Heuristically, if we have a ';' in it we've not got a
8655 ;; brace list, otherwise we have.
8656 (let ((macro-end (progn (c-end-of-macro) (point))))
8657 (goto-char containing-sexp)
8658 (forward-char)
8659 (if (and (c-syntactic-re-search-forward "[;,]" macro-end t t)
8660 (eq (char-before) ?\;))
8661 (setq bufpos nil
8662 containing-sexp nil)
8663 (setq bufpos macro-start))))
8664 (t
8665 ;; Go up one level
8666 (setq containing-sexp next-containing
8667 lim nil
8668 next-containing nil)))))
8669
785eecbb
RS
8670 bufpos))
8671 ))
8672
0ec8351b 8673(defun c-looking-at-special-brace-list (&optional lim)
c38e0c97
PE
8674 ;; If we're looking at the start of a pike-style list, i.e., `({ })',
8675 ;; `([ ])', `(< >)', etc., a cons of a cons of its starting and ending
0ec8351b
BW
8676 ;; positions and its entry in c-special-brace-lists is returned, nil
8677 ;; otherwise. The ending position is nil if the list is still open.
8678 ;; LIM is the limit for forward search. The point may either be at
8679 ;; the `(' or at the following paren character. Tries to check the
8680 ;; matching closer, but assumes it's correct if no balanced paren is
8681 ;; found (i.e. the case `({ ... } ... )' is detected as _not_ being
8682 ;; a special brace list).
0386b551
AM
8683 ;;
8684 ;; This function might do hidden buffer changes.
0ec8351b
BW
8685 (if c-special-brace-lists
8686 (condition-case ()
8687 (save-excursion
8688 (let ((beg (point))
2a15eb73 8689 inner-beg end type)
0ec8351b
BW
8690 (c-forward-syntactic-ws)
8691 (if (eq (char-after) ?\()
8692 (progn
8693 (forward-char 1)
8694 (c-forward-syntactic-ws)
2a15eb73 8695 (setq inner-beg (point))
0ec8351b
BW
8696 (setq type (assq (char-after) c-special-brace-lists)))
8697 (if (setq type (assq (char-after) c-special-brace-lists))
8698 (progn
2a15eb73 8699 (setq inner-beg (point))
0ec8351b
BW
8700 (c-backward-syntactic-ws)
8701 (forward-char -1)
8702 (setq beg (if (eq (char-after) ?\()
8703 (point)
8704 nil)))))
8705 (if (and beg type)
2a15eb73
MS
8706 (if (and (c-safe
8707 (goto-char beg)
8708 (c-forward-sexp 1)
8709 (setq end (point))
8710 (= (char-before) ?\)))
8711 (c-safe
8712 (goto-char inner-beg)
8713 (if (looking-at "\\s(")
8714 ;; Check balancing of the inner paren
8715 ;; below.
8716 (progn
0ec8351b 8717 (c-forward-sexp 1)
2a15eb73
MS
8718 t)
8719 ;; If the inner char isn't a paren then
8720 ;; we can't check balancing, so just
8721 ;; check the char before the outer
8722 ;; closing paren.
8723 (goto-char end)
8724 (backward-char)
8725 (c-backward-syntactic-ws)
8726 (= (char-before) (cdr type)))))
0ec8351b
BW
8727 (if (or (/= (char-syntax (char-before)) ?\))
8728 (= (progn
8729 (c-forward-syntactic-ws)
8730 (point))
8731 (1- end)))
8732 (cons (cons beg end) type))
8733 (cons (list beg) type)))))
8734 (error nil))))
8735
a66cd3ee
MS
8736(defun c-looking-at-bos (&optional lim)
8737 ;; Return non-nil if between two statements or declarations, assuming
8738 ;; point is not inside a literal or comment.
0386b551
AM
8739 ;;
8740 ;; Obsolete - `c-at-statement-start-p' or `c-at-expression-start-p'
8741 ;; are recommended instead.
8742 ;;
8743 ;; This function might do hidden buffer changes.
8744 (c-at-statement-start-p))
efbc652a 8745(make-obsolete 'c-looking-at-bos 'c-at-statement-start-p "22.1")
0386b551
AM
8746
8747(defun c-looking-at-inexpr-block (lim containing-sexp &optional check-at-end)
8748 ;; Return non-nil if we're looking at the beginning of a block
0ec8351b
BW
8749 ;; inside an expression. The value returned is actually a cons of
8750 ;; either 'inlambda, 'inexpr-statement or 'inexpr-class and the
0386b551
AM
8751 ;; position of the beginning of the construct.
8752 ;;
8753 ;; LIM limits the backward search. CONTAINING-SEXP is the start
8754 ;; position of the closest containing list. If it's nil, the
8755 ;; containing paren isn't used to decide whether we're inside an
8756 ;; expression or not. If both LIM and CONTAINING-SEXP are used, LIM
8757 ;; needs to be farther back.
8758 ;;
8759 ;; If CHECK-AT-END is non-nil then extra checks at the end of the
8760 ;; brace block might be done. It should only be used when the
8761 ;; construct can be assumed to be complete, i.e. when the original
8762 ;; starting position was further down than that.
8763 ;;
8764 ;; This function might do hidden buffer changes.
8765
0ec8351b 8766 (save-excursion
0386b551 8767 (let ((res 'maybe) passed-paren
a66cd3ee
MS
8768 (closest-lim (or containing-sexp lim (point-min)))
8769 ;; Look at the character after point only as a last resort
8770 ;; when we can't disambiguate.
8771 (block-follows (and (eq (char-after) ?{) (point))))
0386b551 8772
a66cd3ee
MS
8773 (while (and (eq res 'maybe)
8774 (progn (c-backward-syntactic-ws)
8775 (> (point) closest-lim))
8776 (not (bobp))
8777 (progn (backward-char)
8778 (looking-at "[\]\).]\\|\\w\\|\\s_"))
0386b551
AM
8779 (c-safe (forward-char)
8780 (goto-char (scan-sexps (point) -1))))
8781
a66cd3ee 8782 (setq res
0386b551
AM
8783 (if (looking-at c-keywords-regexp)
8784 (let ((kw-sym (c-keyword-sym (match-string 1))))
8785 (cond
8786 ((and block-follows
8787 (c-keyword-member kw-sym 'c-inexpr-class-kwds))
8788 (and (not (eq passed-paren ?\[))
8789 (or (not (looking-at c-class-key))
8790 ;; If the class definition is at the start of
8791 ;; a statement, we don't consider it an
8792 ;; in-expression class.
8793 (let ((prev (point)))
8794 (while (and
8795 (= (c-backward-token-2 1 nil closest-lim) 0)
8796 (eq (char-syntax (char-after)) ?w))
8797 (setq prev (point)))
8798 (goto-char prev)
8799 (not (c-at-statement-start-p)))
8800 ;; Also, in Pike we treat it as an
8801 ;; in-expression class if it's used in an
8802 ;; object clone expression.
8803 (save-excursion
8804 (and check-at-end
8805 (c-major-mode-is 'pike-mode)
8806 (progn (goto-char block-follows)
8807 (zerop (c-forward-token-2 1 t)))
8808 (eq (char-after) ?\())))
8809 (cons 'inexpr-class (point))))
8810 ((c-keyword-member kw-sym 'c-inexpr-block-kwds)
8811 (when (not passed-paren)
8812 (cons 'inexpr-statement (point))))
8813 ((c-keyword-member kw-sym 'c-lambda-kwds)
8814 (when (or (not passed-paren)
8815 (eq passed-paren ?\())
8816 (cons 'inlambda (point))))
8817 ((c-keyword-member kw-sym 'c-block-stmt-kwds)
8818 nil)
8819 (t
8820 'maybe)))
8821
8822 (if (looking-at "\\s(")
8823 (if passed-paren
8824 (if (and (eq passed-paren ?\[)
8825 (eq (char-after) ?\[))
8826 ;; Accept several square bracket sexps for
8827 ;; Java array initializations.
8828 'maybe)
8829 (setq passed-paren (char-after))
8830 'maybe)
8831 'maybe))))
8832
a66cd3ee 8833 (if (eq res 'maybe)
0386b551
AM
8834 (when (and c-recognize-paren-inexpr-blocks
8835 block-follows
a66cd3ee
MS
8836 containing-sexp
8837 (eq (char-after containing-sexp) ?\())
8838 (goto-char containing-sexp)
8839 (if (or (save-excursion
8840 (c-backward-syntactic-ws lim)
8841 (and (> (point) (or lim (point-min)))
8842 (c-on-identifier)))
8843 (and c-special-brace-lists
8844 (c-looking-at-special-brace-list)))
8845 nil
8846 (cons 'inexpr-statement (point))))
0386b551 8847
a66cd3ee 8848 res))))
0ec8351b 8849
a66cd3ee 8850(defun c-looking-at-inexpr-block-backward (paren-state)
0ec8351b
BW
8851 ;; Returns non-nil if we're looking at the end of an in-expression
8852 ;; block, otherwise the same as `c-looking-at-inexpr-block'.
a66cd3ee 8853 ;; PAREN-STATE is the paren state relevant at the current position.
0386b551
AM
8854 ;;
8855 ;; This function might do hidden buffer changes.
0ec8351b 8856 (save-excursion
a66cd3ee
MS
8857 ;; We currently only recognize a block.
8858 (let ((here (point))
8859 (elem (car-safe paren-state))
8860 containing-sexp)
8861 (when (and (consp elem)
8862 (progn (goto-char (cdr elem))
8863 (c-forward-syntactic-ws here)
8864 (= (point) here)))
8865 (goto-char (car elem))
8866 (if (setq paren-state (cdr paren-state))
8867 (setq containing-sexp (car-safe paren-state)))
8868 (c-looking-at-inexpr-block (c-safe-position containing-sexp
8869 paren-state)
8870 containing-sexp)))))
0ec8351b 8871
536610a4
AM
8872(defun c-at-macro-vsemi-p (&optional pos)
8873 ;; Is there a "virtual semicolon" at POS or point?
8874 ;; (See cc-defs.el for full details of "virtual semicolons".)
8875 ;;
8876 ;; This is true when point is at the last non syntactic WS position on the
8877 ;; line, there is a macro call last on the line, and this particular macro's
8878 ;; name is defined by the regexp `c-vs-macro-regexp' as not needing a
8879 ;; semicolon.
8880 (save-excursion
8881 (save-restriction
8882 (widen)
8883 (if pos
8884 (goto-char pos)
8885 (setq pos (point)))
8886 (and
8887 c-macro-with-semi-re
536610a4
AM
8888 (eq (skip-chars-backward " \t") 0)
8889
8890 ;; Check we've got nothing after this except comments and empty lines
8891 ;; joined by escaped EOLs.
8892 (skip-chars-forward " \t") ; always returns non-nil.
8893 (progn
8894 (while ; go over 1 block comment per iteration.
8895 (and
8896 (looking-at "\\(\\\\[\n\r][ \t]*\\)*")
8897 (goto-char (match-end 0))
8898 (cond
8899 ((looking-at c-block-comment-start-regexp)
8900 (and (forward-comment 1)
8901 (skip-chars-forward " \t"))) ; always returns non-nil
8902 ((looking-at c-line-comment-start-regexp)
8903 (end-of-line)
8904 nil)
8905 (t nil))))
8906 (eolp))
91af3942 8907
536610a4
AM
8908 (goto-char pos)
8909 (progn (c-backward-syntactic-ws)
8910 (eq (point) pos))
8911
8912 ;; Check for one of the listed macros being before point.
8913 (or (not (eq (char-before) ?\)))
8914 (when (c-go-list-backward)
8915 (c-backward-syntactic-ws)
8916 t))
8917 (c-simple-skip-symbol-backward)
2e492df3
AM
8918 (looking-at c-macro-with-semi-re)
8919 (goto-char pos)
c38e0c97 8920 (not (c-in-literal)))))) ; The most expensive check last.
536610a4
AM
8921
8922(defun c-macro-vsemi-status-unknown-p () t) ; See cc-defs.el.
8923
785eecbb 8924\f
d9e94c22
MS
8925;; `c-guess-basic-syntax' and the functions that precedes it below
8926;; implements the main decision tree for determining the syntactic
8927;; analysis of the current line of code.
8928
8929;; Dynamically bound to t when `c-guess-basic-syntax' is called during
8930;; auto newline analysis.
8931(defvar c-auto-newline-analysis nil)
8932
82ba65cf
AM
8933(defun c-brace-anchor-point (bracepos)
8934 ;; BRACEPOS is the position of a brace in a construct like "namespace
8935 ;; Bar {". Return the anchor point in this construct; this is the
8936 ;; earliest symbol on the brace's line which isn't earlier than
8937 ;; "namespace".
8938 ;;
8939 ;; Currently (2007-08-17), "like namespace" means "matches
8940 ;; c-other-block-decl-kwds". It doesn't work with "class" or "struct"
8941 ;; or anything like that.
8942 (save-excursion
8943 (let ((boi (c-point 'boi bracepos)))
8944 (goto-char bracepos)
8945 (while (and (> (point) boi)
8946 (not (looking-at c-other-decl-block-key)))
8947 (c-backward-token-2))
8948 (if (> (point) boi) (point) boi))))
8949
d9e94c22
MS
8950(defsubst c-add-syntax (symbol &rest args)
8951 ;; A simple function to prepend a new syntax element to
8952 ;; `c-syntactic-context'. Using `setq' on it is unsafe since it
8953 ;; should always be dynamically bound but since we read it first
8954 ;; we'll fail properly anyway if this function is misused.
8955 (setq c-syntactic-context (cons (cons symbol args)
8956 c-syntactic-context)))
8957
8958(defsubst c-append-syntax (symbol &rest args)
8959 ;; Like `c-add-syntax' but appends to the end of the syntax list.
8960 ;; (Normally not necessary.)
8961 (setq c-syntactic-context (nconc c-syntactic-context
8962 (list (cons symbol args)))))
a66cd3ee
MS
8963
8964(defun c-add-stmt-syntax (syntax-symbol
d9e94c22 8965 syntax-extra-args
a66cd3ee
MS
8966 stop-at-boi-only
8967 containing-sexp
d9e94c22 8968 paren-state)
51c9af45
AM
8969 ;; Add the indicated SYNTAX-SYMBOL to `c-syntactic-context', extending it as
8970 ;; needed with further syntax elements of the types `substatement',
8971 ;; `inexpr-statement', `arglist-cont-nonempty', `statement-block-intro', and
8972 ;; `defun-block-intro'.
b414f371 8973 ;;
a66cd3ee
MS
8974 ;; Do the generic processing to anchor the given syntax symbol on
8975 ;; the preceding statement: Skip over any labels and containing
8976 ;; statements on the same line, and then search backward until we
8977 ;; find a statement or block start that begins at boi without a
8978 ;; label or comment.
8979 ;;
8980 ;; Point is assumed to be at the prospective anchor point for the
8981 ;; given SYNTAX-SYMBOL. More syntax entries are added if we need to
82ba65cf
AM
8982 ;; skip past open parens and containing statements. Most of the added
8983 ;; syntax elements will get the same anchor point - the exception is
8984 ;; for an anchor in a construct like "namespace"[*] - this is as early
8985 ;; as possible in the construct but on the same line as the {.
8986 ;;
8987 ;; [*] i.e. with a keyword matching c-other-block-decl-kwds.
a66cd3ee 8988 ;;
d9e94c22
MS
8989 ;; SYNTAX-EXTRA-ARGS are a list of the extra arguments for the
8990 ;; syntax symbol. They are appended after the anchor point.
8991 ;;
0386b551
AM
8992 ;; If STOP-AT-BOI-ONLY is nil, we can stop in the middle of the line
8993 ;; if the current statement starts there.
a66cd3ee 8994 ;;
0386b551
AM
8995 ;; Note: It's not a problem if PAREN-STATE "overshoots"
8996 ;; CONTAINING-SEXP, i.e. contains info about parens further down.
8997 ;;
8998 ;; This function might do hidden buffer changes.
8999
a66cd3ee
MS
9000 (if (= (point) (c-point 'boi))
9001 ;; This is by far the most common case, so let's give it special
9002 ;; treatment.
d9e94c22 9003 (apply 'c-add-syntax syntax-symbol (point) syntax-extra-args)
a66cd3ee 9004
0386b551 9005 (let ((syntax-last c-syntactic-context)
d9e94c22 9006 (boi (c-point 'boi))
0386b551
AM
9007 ;; Set when we're on a label, so that we don't stop there.
9008 ;; FIXME: To be complete we should check if we're on a label
9009 ;; now at the start.
9010 on-label)
d9e94c22 9011
82ba65cf
AM
9012 ;; Use point as the anchor point for "namespace", "extern", etc.
9013 (apply 'c-add-syntax syntax-symbol
9014 (if (rassq syntax-symbol c-other-decl-block-key-in-symbols-alist)
9015 (point) nil)
9016 syntax-extra-args)
3ce79cf5 9017
0386b551
AM
9018 ;; Loop while we have to back out of containing blocks.
9019 (while
9020 (and
9021 (catch 'back-up-block
9022
9023 ;; Loop while we have to back up statements.
9024 (while (or (/= (point) boi)
9025 on-label
9026 (looking-at c-comment-start-regexp))
9027
9028 ;; Skip past any comments that stands between the
9029 ;; statement start and boi.
9030 (let ((savepos (point)))
9031 (while (and (/= savepos boi)
9032 (c-backward-single-comment))
9033 (setq savepos (point)
9034 boi (c-point 'boi)))
9035 (goto-char savepos))
9036
9037 ;; Skip to the beginning of this statement or backward
9038 ;; another one.
9039 (let ((old-pos (point))
9040 (old-boi boi)
9041 (step-type (c-beginning-of-statement-1 containing-sexp)))
9042 (setq boi (c-point 'boi)
9043 on-label (eq step-type 'label))
9044
9045 (cond ((= (point) old-pos)
9046 ;; If we didn't move we're at the start of a block and
9047 ;; have to continue outside it.
9048 (throw 'back-up-block t))
9049
9050 ((and (eq step-type 'up)
9051 (>= (point) old-boi)
9052 (looking-at "else\\>[^_]")
9053 (save-excursion
9054 (goto-char old-pos)
9055 (looking-at "if\\>[^_]")))
9056 ;; Special case to avoid deeper and deeper indentation
9057 ;; of "else if" clauses.
9058 )
9059
9060 ((and (not stop-at-boi-only)
9061 (/= old-pos old-boi)
9062 (memq step-type '(up previous)))
9063 ;; If stop-at-boi-only is nil, we shouldn't back up
9064 ;; over previous or containing statements to try to
9065 ;; reach boi, so go back to the last position and
9066 ;; exit.
9067 (goto-char old-pos)
9068 (throw 'back-up-block nil))
9069
9070 (t
9071 (if (and (not stop-at-boi-only)
9072 (memq step-type '(up previous beginning)))
9073 ;; If we've moved into another statement then we
9074 ;; should no longer try to stop in the middle of a
9075 ;; line.
9076 (setq stop-at-boi-only t))
9077
9078 ;; Record this as a substatement if we skipped up one
9079 ;; level.
9080 (when (eq step-type 'up)
9081 (c-add-syntax 'substatement nil))))
9082 )))
9083
9084 containing-sexp)
9085
9086 ;; Now we have to go out of this block.
9087 (goto-char containing-sexp)
9088
9089 ;; Don't stop in the middle of a special brace list opener
9090 ;; like "({".
9091 (when c-special-brace-lists
9092 (let ((special-list (c-looking-at-special-brace-list)))
9093 (when (and special-list
9094 (< (car (car special-list)) (point)))
9095 (setq containing-sexp (car (car special-list)))
9096 (goto-char containing-sexp))))
9097
9098 (setq paren-state (c-whack-state-after containing-sexp paren-state)
9099 containing-sexp (c-most-enclosing-brace paren-state)
9100 boi (c-point 'boi))
9101
9102 ;; Analyze the construct in front of the block we've stepped out
9103 ;; from and add the right syntactic element for it.
9104 (let ((paren-pos (point))
9105 (paren-char (char-after))
9106 step-type)
9107
9108 (if (eq paren-char ?\()
9109 ;; Stepped out of a parenthesis block, so we're in an
9110 ;; expression now.
9111 (progn
9112 (when (/= paren-pos boi)
9113 (if (and c-recognize-paren-inexpr-blocks
3ce79cf5 9114 (progn
0386b551
AM
9115 (c-backward-syntactic-ws containing-sexp)
9116 (or (not (looking-at "\\>"))
9117 (not (c-on-identifier))))
9118 (save-excursion
9119 (goto-char (1+ paren-pos))
9120 (c-forward-syntactic-ws)
9121 (eq (char-after) ?{)))
9122 ;; Stepped out of an in-expression statement. This
9123 ;; syntactic element won't get an anchor pos.
9124 (c-add-syntax 'inexpr-statement)
9125
9126 ;; A parenthesis normally belongs to an arglist.
9127 (c-add-syntax 'arglist-cont-nonempty nil paren-pos)))
9128
9129 (goto-char (max boi
9130 (if containing-sexp
9131 (1+ containing-sexp)
9132 (point-min))))
9133 (setq step-type 'same
9134 on-label nil))
9135
82ba65cf 9136 ;; Stepped out of a brace block.
0386b551
AM
9137 (setq step-type (c-beginning-of-statement-1 containing-sexp)
9138 on-label (eq step-type 'label))
9139
9140 (if (and (eq step-type 'same)
9141 (/= paren-pos (point)))
82ba65cf
AM
9142 (let (inexpr)
9143 (cond
9144 ((save-excursion
9145 (goto-char paren-pos)
9146 (setq inexpr (c-looking-at-inexpr-block
9147 (c-safe-position containing-sexp paren-state)
9148 containing-sexp)))
9149 (c-add-syntax (if (eq (car inexpr) 'inlambda)
9150 'defun-block-intro
9151 'statement-block-intro)
9152 nil))
9153 ((looking-at c-other-decl-block-key)
9154 (c-add-syntax
9155 (cdr (assoc (match-string 1)
9156 c-other-decl-block-key-in-symbols-alist))
9157 (max (c-point 'boi paren-pos) (point))))
9158 (t (c-add-syntax 'defun-block-intro nil))))
b414f371 9159
82ba65cf 9160 (c-add-syntax 'statement-block-intro nil)))
0386b551
AM
9161
9162 (if (= paren-pos boi)
9163 ;; Always done if the open brace was at boi. The
9164 ;; c-beginning-of-statement-1 call above is necessary
9165 ;; anyway, to decide the type of block-intro to add.
9166 (goto-char paren-pos)
9167 (setq boi (c-point 'boi)))
9168 ))
a66cd3ee 9169
d9e94c22
MS
9170 ;; Fill in the current point as the anchor for all the symbols
9171 ;; added above.
82ba65cf 9172 (let ((p c-syntactic-context) q)
d9e94c22 9173 (while (not (eq p syntax-last))
82ba65cf
AM
9174 (setq q (cdr (car p))) ; e.g. (nil 28) [from (arglist-cont-nonempty nil 28)]
9175 (while q
9176 (unless (car q)
9177 (setcar q (point)))
9178 (setq q (cdr q)))
d9e94c22 9179 (setq p (cdr p))))
a66cd3ee
MS
9180 )))
9181
0386b551
AM
9182(defun c-add-class-syntax (symbol
9183 containing-decl-open
9184 containing-decl-start
9185 containing-decl-kwd
9186 paren-state)
a66cd3ee
MS
9187 ;; The inclass and class-close syntactic symbols are added in
9188 ;; several places and some work is needed to fix everything.
9189 ;; Therefore it's collected here.
0386b551
AM
9190 ;;
9191 ;; This function might do hidden buffer changes.
9192 (goto-char containing-decl-open)
9193 (if (and (eq symbol 'inclass) (= (point) (c-point 'boi)))
9194 (progn
9195 (c-add-syntax symbol containing-decl-open)
9196 containing-decl-open)
9197 (goto-char containing-decl-start)
9198 ;; Ought to use `c-add-stmt-syntax' instead of backing up to boi
9199 ;; here, but we have to do like this for compatibility.
9200 (back-to-indentation)
9201 (c-add-syntax symbol (point))
9202 (if (and (c-keyword-member containing-decl-kwd
9203 'c-inexpr-class-kwds)
9204 (/= containing-decl-start (c-point 'boi containing-decl-start)))
9205 (c-add-syntax 'inexpr-class))
9206 (point)))
a66cd3ee
MS
9207
9208(defun c-guess-continued-construct (indent-point
9209 char-after-ip
9210 beg-of-same-or-containing-stmt
9211 containing-sexp
9212 paren-state)
9213 ;; This function contains the decision tree reached through both
9214 ;; cases 18 and 10. It's a continued statement or top level
9215 ;; construct of some kind.
0386b551
AM
9216 ;;
9217 ;; This function might do hidden buffer changes.
d9e94c22 9218
452ea855 9219 (let (special-brace-list placeholder)
a66cd3ee
MS
9220 (goto-char indent-point)
9221 (skip-chars-forward " \t")
d9e94c22 9222
a66cd3ee
MS
9223 (cond
9224 ;; (CASE A removed.)
9225 ;; CASE B: open braces for class or brace-lists
9226 ((setq special-brace-list
9227 (or (and c-special-brace-lists
9228 (c-looking-at-special-brace-list))
9229 (eq char-after-ip ?{)))
d9e94c22 9230
a66cd3ee
MS
9231 (cond
9232 ;; CASE B.1: class-open
9233 ((save-excursion
0386b551
AM
9234 (and (eq (char-after) ?{)
9235 (c-looking-at-decl-block containing-sexp t)
9236 (setq beg-of-same-or-containing-stmt (point))))
a66cd3ee 9237 (c-add-syntax 'class-open beg-of-same-or-containing-stmt))
d9e94c22 9238
a66cd3ee
MS
9239 ;; CASE B.2: brace-list-open
9240 ((or (consp special-brace-list)
9241 (save-excursion
9242 (goto-char beg-of-same-or-containing-stmt)
d9e94c22
MS
9243 (c-syntactic-re-search-forward "=\\([^=]\\|$\\)"
9244 indent-point t t t)))
a66cd3ee 9245 ;; The most semantically accurate symbol here is
d9e94c22
MS
9246 ;; brace-list-open, but we normally report it simply as a
9247 ;; statement-cont. The reason is that one normally adjusts
9248 ;; brace-list-open for brace lists as top-level constructs,
9249 ;; and brace lists inside statements is a completely different
9250 ;; context. C.f. case 5A.3.
a66cd3ee 9251 (c-beginning-of-statement-1 containing-sexp)
d9e94c22
MS
9252 (c-add-stmt-syntax (if c-auto-newline-analysis
9253 ;; Turn off the dwim above when we're
9254 ;; analyzing the nature of the brace
9255 ;; for the auto newline feature.
9256 'brace-list-open
9257 'statement-cont)
0386b551 9258 nil nil
d9e94c22
MS
9259 containing-sexp paren-state))
9260
a66cd3ee
MS
9261 ;; CASE B.3: The body of a function declared inside a normal
9262 ;; block. Can occur e.g. in Pike and when using gcc
037558bf 9263 ;; extensions, but watch out for macros followed by blocks.
d9e94c22 9264 ;; C.f. cases E, 16F and 17G.
0386b551 9265 ((and (not (c-at-statement-start-p))
d9e94c22 9266 (eq (c-beginning-of-statement-1 containing-sexp nil nil t)
037558bf
MS
9267 'same)
9268 (save-excursion
0386b551
AM
9269 (let ((c-recognize-typeless-decls nil))
9270 ;; Turn off recognition of constructs that lacks a
9271 ;; type in this case, since that's more likely to be
9272 ;; a macro followed by a block.
9273 (c-forward-decl-or-cast-1 (c-point 'bosws) nil nil))))
9274 (c-add-stmt-syntax 'defun-open nil t
d9e94c22
MS
9275 containing-sexp paren-state))
9276
037558bf
MS
9277 ;; CASE B.4: Continued statement with block open. The most
9278 ;; accurate analysis is perhaps `statement-cont' together with
9279 ;; `block-open' but we play DWIM and use `substatement-open'
58179cce 9280 ;; instead. The rationale is that this typically is a macro
037558bf
MS
9281 ;; followed by a block which makes it very similar to a
9282 ;; statement with a substatement block.
a66cd3ee 9283 (t
0386b551 9284 (c-add-stmt-syntax 'substatement-open nil nil
037558bf 9285 containing-sexp paren-state))
a66cd3ee 9286 ))
d9e94c22 9287
a66cd3ee 9288 ;; CASE C: iostream insertion or extraction operator
d9e94c22 9289 ((and (looking-at "\\(<<\\|>>\\)\\([^=]\\|$\\)")
a66cd3ee
MS
9290 (save-excursion
9291 (goto-char beg-of-same-or-containing-stmt)
d9e94c22
MS
9292 ;; If there is no preceding streamop in the statement
9293 ;; then indent this line as a normal statement-cont.
9294 (when (c-syntactic-re-search-forward
9295 "\\(<<\\|>>\\)\\([^=]\\|$\\)" indent-point 'move t t)
a66cd3ee
MS
9296 (c-add-syntax 'stream-op (c-point 'boi))
9297 t))))
d9e94c22
MS
9298
9299 ;; CASE E: In the "K&R region" of a function declared inside a
9300 ;; normal block. C.f. case B.3.
9301 ((and (save-excursion
9302 ;; Check that the next token is a '{'. This works as
9303 ;; long as no language that allows nested function
037558bf
MS
9304 ;; definitions allows stuff like member init lists, K&R
9305 ;; declarations or throws clauses there.
d9e94c22
MS
9306 ;;
9307 ;; Note that we do a forward search for something ahead
9308 ;; of the indentation line here. That's not good since
9309 ;; the user might not have typed it yet. Unfortunately
9310 ;; it's exceedingly tricky to recognize a function
9311 ;; prototype in a code block without resorting to this.
9312 (c-forward-syntactic-ws)
9313 (eq (char-after) ?{))
0386b551 9314 (not (c-at-statement-start-p))
d9e94c22 9315 (eq (c-beginning-of-statement-1 containing-sexp nil nil t)
037558bf
MS
9316 'same)
9317 (save-excursion
0386b551
AM
9318 (let ((c-recognize-typeless-decls nil))
9319 ;; Turn off recognition of constructs that lacks a
9320 ;; type in this case, since that's more likely to be
9321 ;; a macro followed by a block.
9322 (c-forward-decl-or-cast-1 (c-point 'bosws) nil nil))))
9323 (c-add-stmt-syntax 'func-decl-cont nil t
d9e94c22
MS
9324 containing-sexp paren-state))
9325
452ea855
AM
9326 ;;CASE F: continued statement and the only preceding items are
9327 ;;annotations.
9328 ((and (c-major-mode-is 'java-mode)
9329 (setq placeholder (point))
9330 (c-beginning-of-statement-1)
9331 (progn
9332 (while (and (c-forward-annotation)
9333 (< (point) placeholder))
9334 (c-forward-syntactic-ws))
9335 t)
9336 (prog1
9337 (>= (point) placeholder)
9338 (goto-char placeholder)))
9339 (c-beginning-of-statement-1 containing-sexp)
9340 (c-add-syntax 'annotation-var-cont (point)))
9341
e6597158
AM
9342 ;; CASE G: a template list continuation?
9343 ;; Mostly a duplication of case 5D.3 to fix templates-19:
9344 ((and (c-major-mode-is 'c++-mode)
9345 (save-excursion
9346 (goto-char indent-point)
9347 (c-with-syntax-table c++-template-syntax-table
9348 (setq placeholder (c-up-list-backward)))
9349 (and placeholder
9350 (eq (char-after placeholder) ?<)
9351 (/= (char-before placeholder) ?<)
9352 (progn
9353 (goto-char (1+ placeholder))
9354 (not (looking-at c-<-op-cont-regexp))))))
9355 (c-with-syntax-table c++-template-syntax-table
9356 (goto-char placeholder)
9357 (c-beginning-of-statement-1 containing-sexp t)
9358 (if (save-excursion
9359 (c-backward-syntactic-ws containing-sexp)
9360 (eq (char-before) ?<))
9361 ;; In a nested template arglist.
9362 (progn
9363 (goto-char placeholder)
9364 (c-syntactic-skip-backward "^,;" containing-sexp t)
9365 (c-forward-syntactic-ws))
9366 (back-to-indentation)))
9367 ;; FIXME: Should use c-add-stmt-syntax, but it's not yet
9368 ;; template aware.
9369 (c-add-syntax 'template-args-cont (point) placeholder))
91af3942 9370
a66cd3ee
MS
9371 ;; CASE D: continued statement.
9372 (t
9373 (c-beginning-of-statement-1 containing-sexp)
0386b551 9374 (c-add-stmt-syntax 'statement-cont nil nil
d9e94c22 9375 containing-sexp paren-state))
a66cd3ee 9376 )))
785eecbb 9377
0386b551
AM
9378;; The next autoload was added by RMS on 2005/8/9 - don't know why (ACM,
9379;; 2005/11/29).
4f047c58 9380;;;###autoload
785eecbb 9381(defun c-guess-basic-syntax ()
0386b551 9382 "Return the syntactic context of the current line."
785eecbb 9383 (save-excursion
9d1f0f55
AM
9384 (beginning-of-line)
9385 (c-save-buffer-state
9386 ((indent-point (point))
9387 (case-fold-search nil)
e3af763f
AM
9388 open-paren-in-column-0-is-defun-start
9389 ;; A whole ugly bunch of various temporary variables. Have
9d1f0f55
AM
9390 ;; to declare them here since it's not possible to declare
9391 ;; a variable with only the scope of a cond test and the
9392 ;; following result clauses, and most of this function is a
9393 ;; single gigantic cond. :P
9394 literal char-before-ip before-ws-ip char-after-ip macro-start
9395 in-macro-expr c-syntactic-context placeholder c-in-literal-cache
9396 step-type tmpsymbol keyword injava-inher special-brace-list tmp-pos
9397 containing-<
9398 ;; The following record some positions for the containing
9399 ;; declaration block if we're directly within one:
9400 ;; `containing-decl-open' is the position of the open
9401 ;; brace. `containing-decl-start' is the start of the
9402 ;; declaration. `containing-decl-kwd' is the keyword
9403 ;; symbol of the keyword that tells what kind of block it
9404 ;; is.
9405 containing-decl-open
9406 containing-decl-start
9407 containing-decl-kwd
9408 ;; The open paren of the closest surrounding sexp or nil if
9409 ;; there is none.
9410 containing-sexp
9411 ;; The position after the closest preceding brace sexp
9412 ;; (nested sexps are ignored), or the position after
9413 ;; `containing-sexp' if there is none, or (point-min) if
9414 ;; `containing-sexp' is nil.
9415 lim
9416 ;; The paren state outside `containing-sexp', or at
9417 ;; `indent-point' if `containing-sexp' is nil.
9418 (paren-state (c-parse-state))
9419 ;; There's always at most one syntactic element which got
9420 ;; an anchor pos. It's stored in syntactic-relpos.
9421 syntactic-relpos
9422 (c-stmt-delim-chars c-stmt-delim-chars))
9423
9424 ;; Check if we're directly inside an enclosing declaration
9425 ;; level block.
9426 (when (and (setq containing-sexp
9427 (c-most-enclosing-brace paren-state))
9428 (progn
9429 (goto-char containing-sexp)
9430 (eq (char-after) ?{))
9431 (setq placeholder
9432 (c-looking-at-decl-block
9433 (c-most-enclosing-brace paren-state
9434 containing-sexp)
9435 t)))
9436 (setq containing-decl-open containing-sexp
9437 containing-decl-start (point)
9438 containing-sexp nil)
9439 (goto-char placeholder)
9440 (setq containing-decl-kwd (and (looking-at c-keywords-regexp)
9441 (c-keyword-sym (match-string 1)))))
9442
9443 ;; Init some position variables.
9444 (if c-state-cache
9445 (progn
9446 (setq containing-sexp (car paren-state)
9447 paren-state (cdr paren-state))
9448 (if (consp containing-sexp)
9449 (progn
9450 (setq lim (cdr containing-sexp))
9451 (if (cdr c-state-cache)
9452 ;; Ignore balanced paren. The next entry
9453 ;; can't be another one.
9454 (setq containing-sexp (car (cdr c-state-cache))
9455 paren-state (cdr paren-state))
9456 ;; If there is no surrounding open paren then
9457 ;; put the last balanced pair back on paren-state.
9458 (setq paren-state (cons containing-sexp paren-state)
9459 containing-sexp nil)))
9460 (setq lim (1+ containing-sexp))))
9461 (setq lim (point-min)))
9462
9463 ;; If we're in a parenthesis list then ',' delimits the
9464 ;; "statements" rather than being an operator (with the
9465 ;; exception of the "for" clause). This difference is
9466 ;; typically only noticeable when statements are used in macro
9467 ;; arglists.
9468 (when (and containing-sexp
9469 (eq (char-after containing-sexp) ?\())
9470 (setq c-stmt-delim-chars c-stmt-delim-chars-with-comma))
9471 ;; cache char before and after indent point, and move point to
9472 ;; the most likely position to perform the majority of tests
9473 (goto-char indent-point)
9474 (c-backward-syntactic-ws lim)
9475 (setq before-ws-ip (point)
9476 char-before-ip (char-before))
9477 (goto-char indent-point)
9478 (skip-chars-forward " \t")
9479 (setq char-after-ip (char-after))
9480
9481 ;; are we in a literal?
9482 (setq literal (c-in-literal lim))
9483
9484 ;; now figure out syntactic qualities of the current line
9485 (cond
0386b551 9486
9d1f0f55
AM
9487 ;; CASE 1: in a string.
9488 ((eq literal 'string)
9489 (c-add-syntax 'string (c-point 'bopl)))
9490
9491 ;; CASE 2: in a C or C++ style comment.
9492 ((and (memq literal '(c c++))
9493 ;; This is a kludge for XEmacs where we use
9494 ;; `buffer-syntactic-context', which doesn't correctly
9495 ;; recognize "\*/" to end a block comment.
9496 ;; `parse-partial-sexp' which is used by
9497 ;; `c-literal-limits' will however do that in most
9498 ;; versions, which results in that we get nil from
9499 ;; `c-literal-limits' even when `c-in-literal' claims
9500 ;; we're inside a comment.
9501 (setq placeholder (c-literal-limits lim)))
9502 (c-add-syntax literal (car placeholder)))
9503
9504 ;; CASE 3: in a cpp preprocessor macro continuation.
9505 ((and (save-excursion
9506 (when (c-beginning-of-macro)
9507 (setq macro-start (point))))
9508 (/= macro-start (c-point 'boi))
9509 (progn
9510 (setq tmpsymbol 'cpp-macro-cont)
9511 (or (not c-syntactic-indentation-in-macros)
9512 (save-excursion
9513 (goto-char macro-start)
9514 ;; If at the beginning of the body of a #define
9515 ;; directive then analyze as cpp-define-intro
9516 ;; only. Go on with the syntactic analysis
9517 ;; otherwise. in-macro-expr is set if we're in a
9518 ;; cpp expression, i.e. before the #define body
9519 ;; or anywhere in a non-#define directive.
9520 (if (c-forward-to-cpp-define-body)
9521 (let ((indent-boi (c-point 'boi indent-point)))
9522 (setq in-macro-expr (> (point) indent-boi)
9523 tmpsymbol 'cpp-define-intro)
9524 (= (point) indent-boi))
9525 (setq in-macro-expr t)
9526 nil)))))
9527 (c-add-syntax tmpsymbol macro-start)
9528 (setq macro-start nil))
9529
9530 ;; CASE 11: an else clause?
9531 ((looking-at "else\\>[^_]")
9532 (c-beginning-of-statement-1 containing-sexp)
9533 (c-add-stmt-syntax 'else-clause nil t
9534 containing-sexp paren-state))
0386b551 9535
9d1f0f55
AM
9536 ;; CASE 12: while closure of a do/while construct?
9537 ((and (looking-at "while\\>[^_]")
9538 (save-excursion
9539 (prog1 (eq (c-beginning-of-statement-1 containing-sexp)
9540 'beginning)
9541 (setq placeholder (point)))))
9542 (goto-char placeholder)
9543 (c-add-stmt-syntax 'do-while-closure nil t
9544 containing-sexp paren-state))
0386b551 9545
9d1f0f55
AM
9546 ;; CASE 13: A catch or finally clause? This case is simpler
9547 ;; than if-else and do-while, because a block is required
9548 ;; after every try, catch and finally.
9549 ((save-excursion
9550 (and (cond ((c-major-mode-is 'c++-mode)
9551 (looking-at "catch\\>[^_]"))
9552 ((c-major-mode-is 'java-mode)
9553 (looking-at "\\(catch\\|finally\\)\\>[^_]")))
9554 (and (c-safe (c-backward-syntactic-ws)
9555 (c-backward-sexp)
9556 t)
9557 (eq (char-after) ?{)
9558 (c-safe (c-backward-syntactic-ws)
9559 (c-backward-sexp)
9560 t)
9561 (if (eq (char-after) ?\()
9562 (c-safe (c-backward-sexp) t)
9563 t))
9564 (looking-at "\\(try\\|catch\\)\\>[^_]")
9565 (setq placeholder (point))))
9566 (goto-char placeholder)
9567 (c-add-stmt-syntax 'catch-clause nil t
9568 containing-sexp paren-state))
0386b551 9569
9d1f0f55
AM
9570 ;; CASE 18: A substatement we can recognize by keyword.
9571 ((save-excursion
9572 (and c-opt-block-stmt-key
9573 (not (eq char-before-ip ?\;))
9574 (not (c-at-vsemi-p before-ws-ip))
9575 (not (memq char-after-ip '(?\) ?\] ?,)))
9576 (or (not (eq char-before-ip ?}))
9577 (c-looking-at-inexpr-block-backward c-state-cache))
9578 (> (point)
9579 (progn
9580 ;; Ought to cache the result from the
9581 ;; c-beginning-of-statement-1 calls here.
9582 (setq placeholder (point))
9583 (while (eq (setq step-type
9584 (c-beginning-of-statement-1 lim))
9585 'label))
9586 (if (eq step-type 'previous)
9587 (goto-char placeholder)
a66cd3ee 9588 (setq placeholder (point))
9d1f0f55
AM
9589 (if (and (eq step-type 'same)
9590 (not (looking-at c-opt-block-stmt-key)))
9591 ;; Step up to the containing statement if we
9592 ;; stayed in the same one.
9593 (let (step)
9594 (while (eq
9595 (setq step
9596 (c-beginning-of-statement-1 lim))
9597 'label))
9598 (if (eq step 'up)
9599 (setq placeholder (point))
fa463103 9600 ;; There was no containing statement after all.
9d1f0f55
AM
9601 (goto-char placeholder)))))
9602 placeholder))
9603 (if (looking-at c-block-stmt-2-key)
9604 ;; Require a parenthesis after these keywords.
9605 ;; Necessary to catch e.g. synchronized in Java,
9606 ;; which can be used both as statement and
9607 ;; modifier.
9608 (and (zerop (c-forward-token-2 1 nil))
9609 (eq (char-after) ?\())
9610 (looking-at c-opt-block-stmt-key))))
9611
9612 (if (eq step-type 'up)
9613 ;; CASE 18A: Simple substatement.
9614 (progn
9615 (goto-char placeholder)
9616 (cond
9617 ((eq char-after-ip ?{)
9618 (c-add-stmt-syntax 'substatement-open nil nil
9619 containing-sexp paren-state))
9620 ((save-excursion
9621 (goto-char indent-point)
9622 (back-to-indentation)
9623 (c-forward-label))
9624 (c-add-stmt-syntax 'substatement-label nil nil
9625 containing-sexp paren-state))
9626 (t
9627 (c-add-stmt-syntax 'substatement nil nil
9628 containing-sexp paren-state))))
9629
9630 ;; CASE 18B: Some other substatement. This is shared
9631 ;; with case 10.
9632 (c-guess-continued-construct indent-point
9633 char-after-ip
9634 placeholder
9635 lim
9636 paren-state)))
0386b551 9637
9d1f0f55
AM
9638 ;; CASE 14: A case or default label
9639 ((looking-at c-label-kwds-regexp)
9640 (if containing-sexp
9641 (progn
9642 (goto-char containing-sexp)
9643 (setq lim (c-most-enclosing-brace c-state-cache
9644 containing-sexp))
9645 (c-backward-to-block-anchor lim)
9646 (c-add-stmt-syntax 'case-label nil t lim paren-state))
9647 ;; Got a bogus label at the top level. In lack of better
9648 ;; alternatives, anchor it on (point-min).
9649 (c-add-syntax 'case-label (point-min))))
0386b551 9650
9d1f0f55
AM
9651 ;; CASE 15: any other label
9652 ((save-excursion
130c507e 9653 (back-to-indentation)
9d1f0f55
AM
9654 (and (not (looking-at c-syntactic-ws-start))
9655 (c-forward-label)))
9656 (cond (containing-decl-open
9657 (setq placeholder (c-add-class-syntax 'inclass
9658 containing-decl-open
9659 containing-decl-start
9660 containing-decl-kwd
9661 paren-state))
9662 ;; Append access-label with the same anchor point as
9663 ;; inclass gets.
9664 (c-append-syntax 'access-label placeholder))
9665
9666 (containing-sexp
9667 (goto-char containing-sexp)
9668 (setq lim (c-most-enclosing-brace c-state-cache
9669 containing-sexp))
9670 (save-excursion
9671 (setq tmpsymbol
9672 (if (and (eq (c-beginning-of-statement-1 lim) 'up)
9673 (looking-at "switch\\>[^_]"))
9674 ;; If the surrounding statement is a switch then
9675 ;; let's analyze all labels as switch labels, so
9676 ;; that they get lined up consistently.
9677 'case-label
9678 'label)))
9679 (c-backward-to-block-anchor lim)
9680 (c-add-stmt-syntax tmpsymbol nil t lim paren-state))
0386b551 9681
9d1f0f55
AM
9682 (t
9683 ;; A label on the top level. Treat it as a class
9684 ;; context. (point-min) is the closest we get to the
9685 ;; class open brace.
9686 (c-add-syntax 'access-label (point-min)))))
9687
9688 ;; CASE 4: In-expression statement. C.f. cases 7B, 16A and
9689 ;; 17E.
9690 ((setq placeholder (c-looking-at-inexpr-block
9691 (c-safe-position containing-sexp paren-state)
9692 containing-sexp
9693 ;; Have to turn on the heuristics after
9694 ;; the point even though it doesn't work
9695 ;; very well. C.f. test case class-16.pike.
9696 t))
9697 (setq tmpsymbol (assq (car placeholder)
9698 '((inexpr-class . class-open)
9699 (inexpr-statement . block-open))))
9700 (if tmpsymbol
9701 ;; It's a statement block or an anonymous class.
9702 (setq tmpsymbol (cdr tmpsymbol))
9703 ;; It's a Pike lambda. Check whether we are between the
9704 ;; lambda keyword and the argument list or at the defun
9705 ;; opener.
9706 (setq tmpsymbol (if (eq char-after-ip ?{)
9707 'inline-open
9708 'lambda-intro-cont)))
9709 (goto-char (cdr placeholder))
9710 (back-to-indentation)
9711 (c-add-stmt-syntax tmpsymbol nil t
9712 (c-most-enclosing-brace c-state-cache (point))
9713 paren-state)
9714 (unless (eq (point) (cdr placeholder))
9715 (c-add-syntax (car placeholder))))
0386b551 9716
9d1f0f55
AM
9717 ;; CASE 5: Line is inside a declaration level block or at top level.
9718 ((or containing-decl-open (null containing-sexp))
9719 (cond
0386b551 9720
9d1f0f55
AM
9721 ;; CASE 5A: we are looking at a defun, brace list, class,
9722 ;; or inline-inclass method opening brace
9723 ((setq special-brace-list
9724 (or (and c-special-brace-lists
9725 (c-looking-at-special-brace-list))
9726 (eq char-after-ip ?{)))
9727 (cond
0386b551 9728
9d1f0f55
AM
9729 ;; CASE 5A.1: Non-class declaration block open.
9730 ((save-excursion
9731 (let (tmp)
9732 (and (eq char-after-ip ?{)
9733 (setq tmp (c-looking-at-decl-block containing-sexp t))
ce8c7486 9734 (progn
9d1f0f55
AM
9735 (setq placeholder (point))
9736 (goto-char tmp)
9737 (looking-at c-symbol-key))
9738 (c-keyword-member
9739 (c-keyword-sym (setq keyword (match-string 0)))
9740 'c-other-block-decl-kwds))))
9741 (goto-char placeholder)
9742 (c-add-stmt-syntax
9743 (if (string-equal keyword "extern")
9744 ;; Special case for extern-lang-open.
9745 'extern-lang-open
9746 (intern (concat keyword "-open")))
9747 nil t containing-sexp paren-state))
0386b551 9748
9d1f0f55 9749 ;; CASE 5A.2: we are looking at a class opening brace
0386b551 9750 ((save-excursion
9d1f0f55
AM
9751 (goto-char indent-point)
9752 (skip-chars-forward " \t")
9753 (and (eq (char-after) ?{)
9754 (c-looking-at-decl-block containing-sexp t)
9755 (setq placeholder (point))))
9756 (c-add-syntax 'class-open placeholder))
9757
9758 ;; CASE 5A.3: brace list open
9759 ((save-excursion
9760 (c-beginning-of-decl-1 lim)
9761 (while (looking-at c-specifier-key)
9762 (goto-char (match-end 1))
9763 (c-forward-syntactic-ws indent-point))
9764 (setq placeholder (c-point 'boi))
9765 (or (consp special-brace-list)
9766 (and (or (save-excursion
9767 (goto-char indent-point)
9768 (setq tmpsymbol nil)
9769 (while (and (> (point) placeholder)
9770 (zerop (c-backward-token-2 1 t))
ace2989a 9771 (not (looking-at "=\\([^=]\\|$\\)")))
9d1f0f55
AM
9772 (and c-opt-inexpr-brace-list-key
9773 (not tmpsymbol)
9774 (looking-at c-opt-inexpr-brace-list-key)
9775 (setq tmpsymbol 'topmost-intro-cont)))
ace2989a 9776 (looking-at "=\\([^=]\\|$\\)"))
9d1f0f55
AM
9777 (looking-at c-brace-list-key))
9778 (save-excursion
9779 (while (and (< (point) indent-point)
9780 (zerop (c-forward-token-2 1 t))
9781 (not (memq (char-after) '(?\; ?\()))))
9782 (not (memq (char-after) '(?\; ?\()))
9783 ))))
9784 (if (and (not c-auto-newline-analysis)
9785 (c-major-mode-is 'java-mode)
9786 (eq tmpsymbol 'topmost-intro-cont))
9787 ;; We're in Java and have found that the open brace
9788 ;; belongs to a "new Foo[]" initialization list,
9789 ;; which means the brace list is part of an
9790 ;; expression and not a top level definition. We
9791 ;; therefore treat it as any topmost continuation
9792 ;; even though the semantically correct symbol still
9793 ;; is brace-list-open, on the same grounds as in
9794 ;; case B.2.
9795 (progn
9796 (c-beginning-of-statement-1 lim)
9797 (c-add-syntax 'topmost-intro-cont (c-point 'boi)))
9798 (c-add-syntax 'brace-list-open placeholder)))
9799
9800 ;; CASE 5A.4: inline defun open
9801 ((and containing-decl-open
9802 (not (c-keyword-member containing-decl-kwd
9803 'c-other-block-decl-kwds)))
9804 (c-add-syntax 'inline-open)
9805 (c-add-class-syntax 'inclass
0386b551
AM
9806 containing-decl-open
9807 containing-decl-start
9808 containing-decl-kwd
9809 paren-state))
9810
9d1f0f55
AM
9811 ;; CASE 5A.5: ordinary defun open
9812 (t
9813 (save-excursion
9814 (c-beginning-of-decl-1 lim)
9815 (while (looking-at c-specifier-key)
9816 (goto-char (match-end 1))
9817 (c-forward-syntactic-ws indent-point))
9818 (c-add-syntax 'defun-open (c-point 'boi))
9819 ;; Bogus to use bol here, but it's the legacy. (Resolved,
9820 ;; 2007-11-09)
9821 ))))
9822
3d1c39fa
AM
9823 ;; CASE 5R: Member init list. (Used to be part of CASE 5B.1)
9824 ;; Note there is no limit on the backward search here, since member
9825 ;; init lists can, in practice, be very large.
9d1f0f55 9826 ((save-excursion
3d1c39fa 9827 (when (setq placeholder (c-back-over-member-initializers))
9d1f0f55 9828 (setq tmp-pos (point))))
3d1c39fa 9829 (if (= (c-point 'bosws) (1+ tmp-pos))
9d1f0f55
AM
9830 (progn
9831 ;; There is no preceding member init clause.
9832 ;; Indent relative to the beginning of indentation
9833 ;; for the topmost-intro line that contains the
9834 ;; prototype's open paren.
9835 (goto-char placeholder)
9836 (c-add-syntax 'member-init-intro (c-point 'boi)))
9837 ;; Indent relative to the first member init clause.
9838 (goto-char (1+ tmp-pos))
9839 (c-forward-syntactic-ws)
9840 (c-add-syntax 'member-init-cont (point))))
9841
3d1c39fa
AM
9842 ;; CASE 5B: After a function header but before the body (or
9843 ;; the ending semicolon if there's no body).
9844 ((save-excursion
9845 (when (setq placeholder (c-just-after-func-arglist-p
9846 (max lim (c-determine-limit 500))))
9847 (setq tmp-pos (point))))
9848 (cond
9849
9850 ;; CASE 5B.1: Member init list.
9851 ((eq (char-after tmp-pos) ?:)
9852 ;; There is no preceding member init clause.
9853 ;; Indent relative to the beginning of indentation
9854 ;; for the topmost-intro line that contains the
9855 ;; prototype's open paren.
9856 (goto-char placeholder)
9857 (c-add-syntax 'member-init-intro (c-point 'boi)))
9858
9d1f0f55 9859 ;; CASE 5B.2: K&R arg decl intro
785eecbb 9860 ((and c-recognize-knr-p
9d1f0f55
AM
9861 (c-in-knr-argdecl lim))
9862 (c-beginning-of-statement-1 lim)
9863 (c-add-syntax 'knr-argdecl-intro (c-point 'boi))
9864 (if containing-decl-open
9865 (c-add-class-syntax 'inclass
9866 containing-decl-open
9867 containing-decl-start
9868 containing-decl-kwd
9869 paren-state)))
9870
9871 ;; CASE 5B.4: Nether region after a C++ or Java func
9872 ;; decl, which could include a `throws' declaration.
9873 (t
9874 (c-beginning-of-statement-1 lim)
9875 (c-add-syntax 'func-decl-cont (c-point 'boi))
9876 )))
9877
9878 ;; CASE 5C: inheritance line. could be first inheritance
9879 ;; line, or continuation of a multiple inheritance
9880 ((or (and (c-major-mode-is 'c++-mode)
9881 (progn
9882 (when (eq char-after-ip ?,)
9883 (skip-chars-forward " \t")
9884 (forward-char))
9885 (looking-at c-opt-postfix-decl-spec-key)))
9886 (and (or (eq char-before-ip ?:)
9887 ;; watch out for scope operator
9888 (save-excursion
9889 (and (eq char-after-ip ?:)
9890 (c-safe (forward-char 1) t)
9891 (not (eq (char-after) ?:))
9892 )))
a66cd3ee 9893 (save-excursion
869455d4
AM
9894 (c-beginning-of-statement-1 lim)
9895 (when (looking-at c-opt-<>-sexp-key)
9896 (goto-char (match-end 1))
9897 (c-forward-syntactic-ws)
9898 (c-forward-<>-arglist nil)
9899 (c-forward-syntactic-ws))
9d1f0f55
AM
9900 (looking-at c-class-key)))
9901 ;; for Java
9902 (and (c-major-mode-is 'java-mode)
9903 (let ((fence (save-excursion
9904 (c-beginning-of-statement-1 lim)
9905 (point)))
9906 cont done)
9907 (save-excursion
9908 (while (not done)
9909 (cond ((looking-at c-opt-postfix-decl-spec-key)
9910 (setq injava-inher (cons cont (point))
9911 done t))
9912 ((or (not (c-safe (c-forward-sexp -1) t))
9913 (<= (point) fence))
9914 (setq done t))
9915 )
9916 (setq cont t)))
9917 injava-inher)
9918 (not (c-crosses-statement-barrier-p (cdr injava-inher)
9919 (point)))
9920 ))
9921 (cond
0386b551 9922
9d1f0f55
AM
9923 ;; CASE 5C.1: non-hanging colon on an inher intro
9924 ((eq char-after-ip ?:)
9925 (c-beginning-of-statement-1 lim)
9926 (c-add-syntax 'inher-intro (c-point 'boi))
9927 ;; don't add inclass symbol since relative point already
9928 ;; contains any class offset
9929 )
0386b551 9930
9d1f0f55
AM
9931 ;; CASE 5C.2: hanging colon on an inher intro
9932 ((eq char-before-ip ?:)
9933 (c-beginning-of-statement-1 lim)
9934 (c-add-syntax 'inher-intro (c-point 'boi))
0386b551 9935 (if containing-decl-open
9d1f0f55
AM
9936 (c-add-class-syntax 'inclass
9937 containing-decl-open
9938 containing-decl-start
9939 containing-decl-kwd
9940 paren-state)))
9941
9942 ;; CASE 5C.3: in a Java implements/extends
9943 (injava-inher
9944 (let ((where (cdr injava-inher))
9945 (cont (car injava-inher)))
9946 (goto-char where)
9947 (cond ((looking-at "throws\\>[^_]")
9948 (c-add-syntax 'func-decl-cont
9949 (progn (c-beginning-of-statement-1 lim)
9950 (c-point 'boi))))
9951 (cont (c-add-syntax 'inher-cont where))
9952 (t (c-add-syntax 'inher-intro
9953 (progn (goto-char (cdr injava-inher))
9954 (c-beginning-of-statement-1 lim)
9955 (point))))
9956 )))
9957
9958 ;; CASE 5C.4: a continued inheritance line
9959 (t
9960 (c-beginning-of-inheritance-list lim)
9961 (c-add-syntax 'inher-cont (point))
9962 ;; don't add inclass symbol since relative point already
9963 ;; contains any class offset
9964 )))
9965
b7aea0ff
AM
9966 ;; CASE 5P: AWK pattern or function or continuation
9967 ;; thereof.
9968 ((c-major-mode-is 'awk-mode)
9969 (setq placeholder (point))
9970 (c-add-stmt-syntax
9971 (if (and (eq (c-beginning-of-statement-1) 'same)
9972 (/= (point) placeholder))
9973 'topmost-intro-cont
9974 'topmost-intro)
9975 nil nil
9976 containing-sexp paren-state))
9977
9d1f0f55
AM
9978 ;; CASE 5D: this could be a top-level initialization, a
9979 ;; member init list continuation, or a template argument
9980 ;; list continuation.
9981 ((save-excursion
9982 ;; Note: We use the fact that lim is always after any
9983 ;; preceding brace sexp.
9984 (if c-recognize-<>-arglists
9985 (while (and
9986 (progn
9987 (c-syntactic-skip-backward "^;,=<>" lim t)
9988 (> (point) lim))
9989 (or
9990 (when c-overloadable-operators-regexp
9991 (when (setq placeholder (c-after-special-operator-id lim))
9992 (goto-char placeholder)
9993 t))
9994 (cond
9995 ((eq (char-before) ?>)
9996 (or (c-backward-<>-arglist nil lim)
9997 (backward-char))
9998 t)
9999 ((eq (char-before) ?<)
10000 (backward-char)
10001 (if (save-excursion
10002 (c-forward-<>-arglist nil))
10003 (progn (forward-char)
10004 nil)
10005 t))
10006 (t nil)))))
10007 ;; NB: No c-after-special-operator-id stuff in this
10008 ;; clause - we assume only C++ needs it.
10009 (c-syntactic-skip-backward "^;,=" lim t))
10010 (memq (char-before) '(?, ?= ?<)))
10011 (cond
10012
10013 ;; CASE 5D.3: perhaps a template list continuation?
10014 ((and (c-major-mode-is 'c++-mode)
10015 (save-excursion
10016 (save-restriction
10017 (c-with-syntax-table c++-template-syntax-table
10018 (goto-char indent-point)
10019 (setq placeholder (c-up-list-backward))
10020 (and placeholder
10021 (eq (char-after placeholder) ?<))))))
10022 (c-with-syntax-table c++-template-syntax-table
10023 (goto-char placeholder)
10024 (c-beginning-of-statement-1 lim t)
10025 (if (save-excursion
10026 (c-backward-syntactic-ws lim)
10027 (eq (char-before) ?<))
10028 ;; In a nested template arglist.
10029 (progn
10030 (goto-char placeholder)
10031 (c-syntactic-skip-backward "^,;" lim t)
10032 (c-forward-syntactic-ws))
10033 (back-to-indentation)))
10034 ;; FIXME: Should use c-add-stmt-syntax, but it's not yet
10035 ;; template aware.
10036 (c-add-syntax 'template-args-cont (point) placeholder))
10037
10038 ;; CASE 5D.4: perhaps a multiple inheritance line?
10039 ((and (c-major-mode-is 'c++-mode)
3efc2cd7 10040 (save-excursion
785eecbb 10041 (c-beginning-of-statement-1 lim)
9d1f0f55
AM
10042 (setq placeholder (point))
10043 (if (looking-at "static\\>[^_]")
10044 (c-forward-token-2 1 nil indent-point))
10045 (and (looking-at c-class-key)
10046 (zerop (c-forward-token-2 2 nil indent-point))
10047 (if (eq (char-after) ?<)
10048 (c-with-syntax-table c++-template-syntax-table
10049 (zerop (c-forward-token-2 1 t indent-point)))
10050 t)
10051 (eq (char-after) ?:))))
10052 (goto-char placeholder)
10053 (c-add-syntax 'inher-cont (c-point 'boi)))
10054
10055 ;; CASE 5D.5: Continuation of the "expression part" of a
e1dbe924 10056 ;; top level construct. Or, perhaps, an unrecognized construct.
9d1f0f55
AM
10057 (t
10058 (while (and (setq placeholder (point))
2e492df3 10059 (eq (car (c-beginning-of-decl-1 containing-sexp)) ; Can't use `lim' here.
9d1f0f55
AM
10060 'same)
10061 (save-excursion
10062 (c-backward-syntactic-ws)
10063 (eq (char-before) ?}))
10064 (< (point) placeholder)))
10065 (c-add-stmt-syntax
10066 (cond
e1dbe924 10067 ((eq (point) placeholder) 'statement) ; unrecognized construct
9d1f0f55
AM
10068 ;; A preceding comma at the top level means that a
10069 ;; new variable declaration starts here. Use
10070 ;; topmost-intro-cont for it, for consistency with
10071 ;; the first variable declaration. C.f. case 5N.
10072 ((eq char-before-ip ?,) 'topmost-intro-cont)
10073 (t 'statement-cont))
10074 nil nil containing-sexp paren-state))
10075 ))
0386b551 10076
9d1f0f55
AM
10077 ;; CASE 5F: Close of a non-class declaration level block.
10078 ((and (eq char-after-ip ?})
10079 (c-keyword-member containing-decl-kwd
10080 'c-other-block-decl-kwds))
10081 ;; This is inconsistent: Should use `containing-decl-open'
10082 ;; here if it's at boi, like in case 5J.
10083 (goto-char containing-decl-start)
10084 (c-add-stmt-syntax
10085 (if (string-equal (symbol-name containing-decl-kwd) "extern")
10086 ;; Special case for compatibility with the
10087 ;; extern-lang syntactic symbols.
10088 'extern-lang-close
10089 (intern (concat (symbol-name containing-decl-kwd)
10090 "-close")))
10091 nil t
10092 (c-most-enclosing-brace paren-state (point))
10093 paren-state))
10094
10095 ;; CASE 5G: we are looking at the brace which closes the
10096 ;; enclosing nested class decl
10097 ((and containing-sexp
10098 (eq char-after-ip ?})
10099 (eq containing-decl-open containing-sexp))
10100 (c-add-class-syntax 'class-close
10101 containing-decl-open
10102 containing-decl-start
10103 containing-decl-kwd
10104 paren-state))
10105
10106 ;; CASE 5H: we could be looking at subsequent knr-argdecls
10107 ((and c-recognize-knr-p
10108 (not containing-sexp) ; can't be knr inside braces.
10109 (not (eq char-before-ip ?}))
10110 (save-excursion
10111 (setq placeholder (cdr (c-beginning-of-decl-1 lim)))
10112 (and placeholder
10113 ;; Do an extra check to avoid tripping up on
10114 ;; statements that occur in invalid contexts
10115 ;; (e.g. in macro bodies where we don't really
10116 ;; know the context of what we're looking at).
10117 (not (and c-opt-block-stmt-key
10118 (looking-at c-opt-block-stmt-key)))))
10119 (< placeholder indent-point))
10120 (goto-char placeholder)
10121 (c-add-syntax 'knr-argdecl (point)))
10122
10123 ;; CASE 5I: ObjC method definition.
10124 ((and c-opt-method-key
10125 (looking-at c-opt-method-key))
10126 (c-beginning-of-statement-1 nil t)
10127 (if (= (point) indent-point)
10128 ;; Handle the case when it's the first (non-comment)
10129 ;; thing in the buffer. Can't look for a 'same return
10130 ;; value from cbos1 since ObjC directives currently
10131 ;; aren't recognized fully, so that we get 'same
10132 ;; instead of 'previous if it moved over a preceding
10133 ;; directive.
10134 (goto-char (point-min)))
10135 (c-add-syntax 'objc-method-intro (c-point 'boi)))
10136
9d1f0f55
AM
10137 ;; CASE 5N: At a variable declaration that follows a class
10138 ;; definition or some other block declaration that doesn't
10139 ;; end at the closing '}'. C.f. case 5D.5.
10140 ((progn
10141 (c-backward-syntactic-ws lim)
10142 (and (eq (char-before) ?})
10143 (save-excursion
10144 (let ((start (point)))
10145 (if (and c-state-cache
10146 (consp (car c-state-cache))
10147 (eq (cdar c-state-cache) (point)))
10148 ;; Speed up the backward search a bit.
10149 (goto-char (caar c-state-cache)))
2e492df3 10150 (c-beginning-of-decl-1 containing-sexp) ; Can't use `lim' here.
9d1f0f55
AM
10151 (setq placeholder (point))
10152 (if (= start (point))
10153 ;; The '}' is unbalanced.
10154 nil
10155 (c-end-of-decl-1)
10156 (>= (point) indent-point))))))
10157 (goto-char placeholder)
10158 (c-add-stmt-syntax 'topmost-intro-cont nil nil
10159 containing-sexp paren-state))
10160
10161 ;; NOTE: The point is at the end of the previous token here.
10162
10163 ;; CASE 5J: we are at the topmost level, make
10164 ;; sure we skip back past any access specifiers
10165 ((and
10166 ;; A macro continuation line is never at top level.
10167 (not (and macro-start
10168 (> indent-point macro-start)))
10169 (save-excursion
10170 (setq placeholder (point))
10171 (or (memq char-before-ip '(?\; ?{ ?} nil))
10172 (c-at-vsemi-p before-ws-ip)
10173 (when (and (eq char-before-ip ?:)
10174 (eq (c-beginning-of-statement-1 lim)
10175 'label))
10176 (c-backward-syntactic-ws lim)
10177 (setq placeholder (point)))
10178 (and (c-major-mode-is 'objc-mode)
10179 (catch 'not-in-directive
10180 (c-beginning-of-statement-1 lim)
10181 (setq placeholder (point))
10182 (while (and (c-forward-objc-directive)
10183 (< (point) indent-point))
10184 (c-forward-syntactic-ws)
10185 (if (>= (point) indent-point)
10186 (throw 'not-in-directive t))
10187 (setq placeholder (point)))
10188 nil)))))
10189 ;; For historic reasons we anchor at bol of the last
10190 ;; line of the previous declaration. That's clearly
10191 ;; highly bogus and useless, and it makes our lives hard
10192 ;; to remain compatible. :P
10193 (goto-char placeholder)
10194 (c-add-syntax 'topmost-intro (c-point 'bol))
10195 (if containing-decl-open
10196 (if (c-keyword-member containing-decl-kwd
10197 'c-other-block-decl-kwds)
10198 (progn
10199 (goto-char (c-brace-anchor-point containing-decl-open))
10200 (c-add-stmt-syntax
10201 (if (string-equal (symbol-name containing-decl-kwd)
10202 "extern")
10203 ;; Special case for compatibility with the
10204 ;; extern-lang syntactic symbols.
10205 'inextern-lang
10206 (intern (concat "in"
10207 (symbol-name containing-decl-kwd))))
10208 nil t
10209 (c-most-enclosing-brace paren-state (point))
10210 paren-state))
10211 (c-add-class-syntax 'inclass
10212 containing-decl-open
10213 containing-decl-start
10214 containing-decl-kwd
10215 paren-state)))
10216 (when (and c-syntactic-indentation-in-macros
10217 macro-start
10218 (/= macro-start (c-point 'boi indent-point)))
10219 (c-add-syntax 'cpp-define-intro)
10220 (setq macro-start nil)))
10221
10222 ;; CASE 5K: we are at an ObjC method definition
10223 ;; continuation line.
10224 ((and c-opt-method-key
10225 (save-excursion
10226 (c-beginning-of-statement-1 lim)
10227 (beginning-of-line)
10228 (when (looking-at c-opt-method-key)
10229 (setq placeholder (point)))))
10230 (c-add-syntax 'objc-method-args-cont placeholder))
10231
10232 ;; CASE 5L: we are at the first argument of a template
452ea855
AM
10233 ;; arglist that begins on the previous line.
10234 ((and c-recognize-<>-arglists
10235 (eq (char-before) ?<)
10236 (not (and c-overloadable-operators-regexp
10237 (c-after-special-operator-id lim))))
10238 (c-beginning-of-statement-1 (c-safe-position (point) paren-state))
10239 (c-add-syntax 'template-args-cont (c-point 'boi)))
10240
10241 ;; CASE 5Q: we are at a statement within a macro.
10242 (macro-start
10243 (c-beginning-of-statement-1 containing-sexp)
10244 (c-add-stmt-syntax 'statement nil t containing-sexp paren-state))
10245
40ba43b4 10246 ;;CASE 5N: We are at a topmost continuation line and the only
9d1f0f55 10247 ;;preceding items are annotations.
452ea855
AM
10248 ((and (c-major-mode-is 'java-mode)
10249 (setq placeholder (point))
9d1f0f55
AM
10250 (c-beginning-of-statement-1)
10251 (progn
452ea855 10252 (while (and (c-forward-annotation))
9d1f0f55
AM
10253 (c-forward-syntactic-ws))
10254 t)
10255 (prog1
10256 (>= (point) placeholder)
10257 (goto-char placeholder)))
10258 (c-add-syntax 'annotation-top-cont (c-point 'boi)))
452ea855
AM
10259
10260 ;; CASE 5M: we are at a topmost continuation line
10261 (t
10262 (c-beginning-of-statement-1 (c-safe-position (point) paren-state))
9d1f0f55
AM
10263 (when (c-major-mode-is 'objc-mode)
10264 (setq placeholder (point))
10265 (while (and (c-forward-objc-directive)
10266 (< (point) indent-point))
10267 (c-forward-syntactic-ws)
10268 (setq placeholder (point)))
10269 (goto-char placeholder))
10270 (c-add-syntax 'topmost-intro-cont (c-point 'boi)))
10271 ))
0386b551 10272
0386b551 10273
452ea855 10274 ;; (CASE 6 has been removed.)
d0fcee66 10275
452ea855
AM
10276 ;; CASE 7: line is an expression, not a statement. Most
10277 ;; likely we are either in a function prototype or a function
10278 ;; call argument list
10279 ((not (or (and c-special-brace-lists
10280 (save-excursion
10281 (goto-char containing-sexp)
10282 (c-looking-at-special-brace-list)))
10283 (eq (char-after containing-sexp) ?{)))
10284 (cond
0386b551 10285
452ea855 10286 ;; CASE 7A: we are looking at the arglist closing paren.
9d1f0f55
AM
10287 ;; C.f. case 7F.
10288 ((memq char-after-ip '(?\) ?\]))
10289 (goto-char containing-sexp)
10290 (setq placeholder (c-point 'boi))
10291 (if (and (c-safe (backward-up-list 1) t)
10292 (>= (point) placeholder))
10293 (progn
10294 (forward-char)
10295 (skip-chars-forward " \t"))
10296 (goto-char placeholder))
10297 (c-add-stmt-syntax 'arglist-close (list containing-sexp) t
452ea855
AM
10298 (c-most-enclosing-brace paren-state (point))
10299 paren-state))
0386b551 10300
452ea855
AM
10301 ;; CASE 7B: Looking at the opening brace of an
10302 ;; in-expression block or brace list. C.f. cases 4, 16A
10303 ;; and 17E.
9d1f0f55
AM
10304 ((and (eq char-after-ip ?{)
10305 (progn
10306 (setq placeholder (c-inside-bracelist-p (point)
10307 paren-state))
10308 (if placeholder
10309 (setq tmpsymbol '(brace-list-open . inexpr-class))
10310 (setq tmpsymbol '(block-open . inexpr-statement)
10311 placeholder
10312 (cdr-safe (c-looking-at-inexpr-block
10313 (c-safe-position containing-sexp
10314 paren-state)
10315 containing-sexp)))
10316 ;; placeholder is nil if it's a block directly in
10317 ;; a function arglist. That makes us skip out of
10318 ;; this case.
10319 )))
10320 (goto-char placeholder)
10321 (back-to-indentation)
10322 (c-add-stmt-syntax (car tmpsymbol) nil t
10323 (c-most-enclosing-brace paren-state (point))
10324 paren-state)
10325 (if (/= (point) placeholder)
10326 (c-add-syntax (cdr tmpsymbol))))
10327
10328 ;; CASE 7C: we are looking at the first argument in an empty
10329 ;; argument list. Use arglist-close if we're actually
10330 ;; looking at a close paren or bracket.
10331 ((memq char-before-ip '(?\( ?\[))
10332 (goto-char containing-sexp)
10333 (setq placeholder (c-point 'boi))
10334 (if (and (c-safe (backward-up-list 1) t)
10335 (>= (point) placeholder))
10336 (progn
10337 (forward-char)
10338 (skip-chars-forward " \t"))
10339 (goto-char placeholder))
10340 (c-add-stmt-syntax 'arglist-intro (list containing-sexp) t
10341 (c-most-enclosing-brace paren-state (point))
10342 paren-state))
0386b551 10343
9d1f0f55
AM
10344 ;; CASE 7D: we are inside a conditional test clause. treat
10345 ;; these things as statements
10346 ((progn
785eecbb 10347 (goto-char containing-sexp)
9d1f0f55
AM
10348 (and (c-safe (c-forward-sexp -1) t)
10349 (looking-at "\\<for\\>[^_]")))
10350 (goto-char (1+ containing-sexp))
10351 (c-forward-syntactic-ws indent-point)
10352 (if (eq char-before-ip ?\;)
10353 (c-add-syntax 'statement (point))
10354 (c-add-syntax 'statement-cont (point))
10355 ))
10356
10357 ;; CASE 7E: maybe a continued ObjC method call. This is the
10358 ;; case when we are inside a [] bracketed exp, and what
10359 ;; precede the opening bracket is not an identifier.
10360 ((and c-opt-method-key
10361 (eq (char-after containing-sexp) ?\[)
10362 (progn
10363 (goto-char (1- containing-sexp))
10364 (c-backward-syntactic-ws (c-point 'bod))
10365 (if (not (looking-at c-symbol-key))
10366 (c-add-syntax 'objc-method-call-cont containing-sexp))
10367 )))
0386b551 10368
9d1f0f55
AM
10369 ;; CASE 7F: we are looking at an arglist continuation line,
10370 ;; but the preceding argument is on the same line as the
10371 ;; opening paren. This case includes multi-line
10372 ;; mathematical paren groupings, but we could be on a
10373 ;; for-list continuation line. C.f. case 7A.
10374 ((progn
785eecbb 10375 (goto-char (1+ containing-sexp))
9d1f0f55
AM
10376 (< (save-excursion
10377 (c-forward-syntactic-ws)
10378 (point))
10379 (c-point 'bonl)))
10380 (goto-char containing-sexp) ; paren opening the arglist
10381 (setq placeholder (c-point 'boi))
10382 (if (and (c-safe (backward-up-list 1) t)
10383 (>= (point) placeholder))
10384 (progn
10385 (forward-char)
10386 (skip-chars-forward " \t"))
10387 (goto-char placeholder))
10388 (c-add-stmt-syntax 'arglist-cont-nonempty (list containing-sexp) t
10389 (c-most-enclosing-brace c-state-cache (point))
10390 paren-state))
0386b551 10391
9d1f0f55
AM
10392 ;; CASE 7G: we are looking at just a normal arglist
10393 ;; continuation line
10394 (t (c-forward-syntactic-ws indent-point)
10395 (c-add-syntax 'arglist-cont (c-point 'boi)))
10396 ))
0386b551 10397
9d1f0f55
AM
10398 ;; CASE 8: func-local multi-inheritance line
10399 ((and (c-major-mode-is 'c++-mode)
10400 (save-excursion
10401 (goto-char indent-point)
10402 (skip-chars-forward " \t")
10403 (looking-at c-opt-postfix-decl-spec-key)))
10404 (goto-char indent-point)
10405 (skip-chars-forward " \t")
10406 (cond
0386b551 10407
9d1f0f55
AM
10408 ;; CASE 8A: non-hanging colon on an inher intro
10409 ((eq char-after-ip ?:)
10410 (c-backward-syntactic-ws lim)
10411 (c-add-syntax 'inher-intro (c-point 'boi)))
0386b551 10412
9d1f0f55
AM
10413 ;; CASE 8B: hanging colon on an inher intro
10414 ((eq char-before-ip ?:)
10415 (c-add-syntax 'inher-intro (c-point 'boi)))
0386b551 10416
9d1f0f55
AM
10417 ;; CASE 8C: a continued inheritance line
10418 (t
10419 (c-beginning-of-inheritance-list lim)
10420 (c-add-syntax 'inher-cont (point))
10421 )))
10422
10423 ;; CASE 9: we are inside a brace-list
10424 ((and (not (c-major-mode-is 'awk-mode)) ; Maybe this isn't needed (ACM, 2002/3/29)
10425 (setq special-brace-list
10426 (or (and c-special-brace-lists ;;;; ALWAYS NIL FOR AWK!!
10427 (save-excursion
10428 (goto-char containing-sexp)
10429 (c-looking-at-special-brace-list)))
10430 (c-inside-bracelist-p containing-sexp paren-state))))
10431 (cond
0386b551 10432
9d1f0f55
AM
10433 ;; CASE 9A: In the middle of a special brace list opener.
10434 ((and (consp special-brace-list)
10435 (save-excursion
10436 (goto-char containing-sexp)
10437 (eq (char-after) ?\())
10438 (eq char-after-ip (car (cdr special-brace-list))))
10439 (goto-char (car (car special-brace-list)))
10440 (skip-chars-backward " \t")
10441 (if (and (bolp)
10442 (assoc 'statement-cont
10443 (setq placeholder (c-guess-basic-syntax))))
10444 (setq c-syntactic-context placeholder)
10445 (c-beginning-of-statement-1
10446 (c-safe-position (1- containing-sexp) paren-state))
10447 (c-forward-token-2 0)
10448 (while (looking-at c-specifier-key)
10449 (goto-char (match-end 1))
10450 (c-forward-syntactic-ws))
10451 (c-add-syntax 'brace-list-open (c-point 'boi))))
10452
10453 ;; CASE 9B: brace-list-close brace
10454 ((if (consp special-brace-list)
10455 ;; Check special brace list closer.
10456 (progn
10457 (goto-char (car (car special-brace-list)))
10458 (save-excursion
10459 (goto-char indent-point)
10460 (back-to-indentation)
10461 (or
10462 ;; We were between the special close char and the `)'.
10463 (and (eq (char-after) ?\))
10464 (eq (1+ (point)) (cdr (car special-brace-list))))
10465 ;; We were before the special close char.
10466 (and (eq (char-after) (cdr (cdr special-brace-list)))
10467 (zerop (c-forward-token-2))
10468 (eq (1+ (point)) (cdr (car special-brace-list)))))))
10469 ;; Normal brace list check.
10470 (and (eq char-after-ip ?})
10471 (c-safe (goto-char (c-up-list-backward (point))) t)
10472 (= (point) containing-sexp)))
10473 (if (eq (point) (c-point 'boi))
10474 (c-add-syntax 'brace-list-close (point))
10475 (setq lim (c-most-enclosing-brace c-state-cache (point)))
10476 (c-beginning-of-statement-1 lim)
10477 (c-add-stmt-syntax 'brace-list-close nil t lim paren-state)))
0386b551 10478
9d1f0f55
AM
10479 (t
10480 ;; Prepare for the rest of the cases below by going to the
10481 ;; token following the opening brace
10482 (if (consp special-brace-list)
10483 (progn
10484 (goto-char (car (car special-brace-list)))
10485 (c-forward-token-2 1 nil indent-point))
10486 (goto-char containing-sexp))
10487 (forward-char)
10488 (let ((start (point)))
10489 (c-forward-syntactic-ws indent-point)
10490 (goto-char (max start (c-point 'bol))))
10491 (c-skip-ws-forward indent-point)
785eecbb 10492 (cond
0386b551 10493
9d1f0f55
AM
10494 ;; CASE 9C: we're looking at the first line in a brace-list
10495 ((= (point) indent-point)
10496 (if (consp special-brace-list)
10497 (goto-char (car (car special-brace-list)))
10498 (goto-char containing-sexp))
a66cd3ee 10499 (if (eq (point) (c-point 'boi))
9d1f0f55 10500 (c-add-syntax 'brace-list-intro (point))
a66cd3ee
MS
10501 (setq lim (c-most-enclosing-brace c-state-cache (point)))
10502 (c-beginning-of-statement-1 lim)
9d1f0f55 10503 (c-add-stmt-syntax 'brace-list-intro nil t lim paren-state)))
0386b551 10504
9d1f0f55
AM
10505 ;; CASE 9D: this is just a later brace-list-entry or
10506 ;; brace-entry-open
10507 (t (if (or (eq char-after-ip ?{)
10508 (and c-special-brace-lists
10509 (save-excursion
10510 (goto-char indent-point)
10511 (c-forward-syntactic-ws (c-point 'eol))
10512 (c-looking-at-special-brace-list (point)))))
10513 (c-add-syntax 'brace-entry-open (point))
10514 (c-add-syntax 'brace-list-entry (point))
10515 ))
10516 ))))
10517
10518 ;; CASE 10: A continued statement or top level construct.
10519 ((and (not (memq char-before-ip '(?\; ?:)))
10520 (not (c-at-vsemi-p before-ws-ip))
10521 (or (not (eq char-before-ip ?}))
10522 (c-looking-at-inexpr-block-backward c-state-cache))
10523 (> (point)
10524 (save-excursion
10525 (c-beginning-of-statement-1 containing-sexp)
10526 (setq placeholder (point))))
10527 (/= placeholder containing-sexp))
10528 ;; This is shared with case 18.
10529 (c-guess-continued-construct indent-point
10530 char-after-ip
10531 placeholder
10532 containing-sexp
10533 paren-state))
10534
10535 ;; CASE 16: block close brace, possibly closing the defun or
10536 ;; the class
10537 ((eq char-after-ip ?})
10538 ;; From here on we have the next containing sexp in lim.
10539 (setq lim (c-most-enclosing-brace paren-state))
10540 (goto-char containing-sexp)
10541 (cond
0386b551 10542
9d1f0f55
AM
10543 ;; CASE 16E: Closing a statement block? This catches
10544 ;; cases where it's preceded by a statement keyword,
10545 ;; which works even when used in an "invalid" context,
10546 ;; e.g. a macro argument.
10547 ((c-after-conditional)
10548 (c-backward-to-block-anchor lim)
10549 (c-add-stmt-syntax 'block-close nil t lim paren-state))
10550
10551 ;; CASE 16A: closing a lambda defun or an in-expression
10552 ;; block? C.f. cases 4, 7B and 17E.
10553 ((setq placeholder (c-looking-at-inexpr-block
10554 (c-safe-position containing-sexp paren-state)
10555 nil))
10556 (setq tmpsymbol (if (eq (car placeholder) 'inlambda)
10557 'inline-close
10558 'block-close))
a66cd3ee 10559 (goto-char containing-sexp)
9d1f0f55
AM
10560 (back-to-indentation)
10561 (if (= containing-sexp (point))
10562 (c-add-syntax tmpsymbol (point))
10563 (goto-char (cdr placeholder))
10564 (back-to-indentation)
10565 (c-add-stmt-syntax tmpsymbol nil t
10566 (c-most-enclosing-brace paren-state (point))
10567 paren-state)
10568 (if (/= (point) (cdr placeholder))
10569 (c-add-syntax (car placeholder)))))
0386b551 10570
9d1f0f55
AM
10571 ;; CASE 16B: does this close an inline or a function in
10572 ;; a non-class declaration level block?
10573 ((save-excursion
10574 (and lim
10575 (progn
10576 (goto-char lim)
10577 (c-looking-at-decl-block
10578 (c-most-enclosing-brace paren-state lim)
10579 nil))
10580 (setq placeholder (point))))
10581 (c-backward-to-decl-anchor lim)
10582 (back-to-indentation)
10583 (if (save-excursion
10584 (goto-char placeholder)
10585 (looking-at c-other-decl-block-key))
10586 (c-add-syntax 'defun-close (point))
10587 (c-add-syntax 'inline-close (point))))
10588
10589 ;; CASE 16F: Can be a defun-close of a function declared
10590 ;; in a statement block, e.g. in Pike or when using gcc
10591 ;; extensions, but watch out for macros followed by
10592 ;; blocks. Let it through to be handled below.
10593 ;; C.f. cases B.3 and 17G.
10594 ((save-excursion
10595 (and (not (c-at-statement-start-p))
10596 (eq (c-beginning-of-statement-1 lim nil nil t) 'same)
10597 (setq placeholder (point))
10598 (let ((c-recognize-typeless-decls nil))
10599 ;; Turn off recognition of constructs that
10600 ;; lacks a type in this case, since that's more
10601 ;; likely to be a macro followed by a block.
10602 (c-forward-decl-or-cast-1 (c-point 'bosws) nil nil))))
10603 (back-to-indentation)
10604 (if (/= (point) containing-sexp)
10605 (goto-char placeholder))
10606 (c-add-stmt-syntax 'defun-close nil t lim paren-state))
10607
10608 ;; CASE 16C: If there is an enclosing brace then this is
10609 ;; a block close since defun closes inside declaration
10610 ;; level blocks have been handled above.
10611 (lim
10612 ;; If the block is preceded by a case/switch label on
10613 ;; the same line, we anchor at the first preceding label
10614 ;; at boi. The default handling in c-add-stmt-syntax
10615 ;; really fixes it better, but we do like this to keep
10616 ;; the indentation compatible with version 5.28 and
10617 ;; earlier. C.f. case 17H.
10618 (while (and (/= (setq placeholder (point)) (c-point 'boi))
10619 (eq (c-beginning-of-statement-1 lim) 'label)))
10620 (goto-char placeholder)
10621 (if (looking-at c-label-kwds-regexp)
10622 (c-add-syntax 'block-close (point))
10623 (goto-char containing-sexp)
10624 ;; c-backward-to-block-anchor not necessary here; those
10625 ;; situations are handled in case 16E above.
10626 (c-add-stmt-syntax 'block-close nil t lim paren-state)))
0386b551 10627
9d1f0f55 10628 ;; CASE 16D: Only top level defun close left.
785eecbb 10629 (t
9d1f0f55
AM
10630 (goto-char containing-sexp)
10631 (c-backward-to-decl-anchor lim)
10632 (c-add-stmt-syntax 'defun-close nil nil
10633 (c-most-enclosing-brace paren-state)
10634 paren-state))
10635 ))
0386b551 10636
804a116c
AM
10637 ;; CASE 19: line is an expression, not a statement, and is directly
10638 ;; contained by a template delimiter. Most likely, we are in a
10639 ;; template arglist within a statement. This case is based on CASE
10640 ;; 7. At some point in the future, we may wish to create more
10641 ;; syntactic symbols such as `template-intro',
10642 ;; `template-cont-nonempty', etc., and distinguish between them as we
10643 ;; do for `arglist-intro' etc. (2009-12-07).
10644 ((and c-recognize-<>-arglists
10645 (setq containing-< (c-up-list-backward indent-point containing-sexp))
10646 (eq (char-after containing-<) ?\<))
10647 (setq placeholder (c-point 'boi containing-<))
10648 (goto-char containing-sexp) ; Most nested Lbrace/Lparen (but not
10649 ; '<') before indent-point.
10650 (if (>= (point) placeholder)
10651 (progn
10652 (forward-char)
10653 (skip-chars-forward " \t"))
10654 (goto-char placeholder))
10655 (c-add-stmt-syntax 'template-args-cont (list containing-<) t
10656 (c-most-enclosing-brace c-state-cache (point))
10657 paren-state))
10658
9d1f0f55
AM
10659 ;; CASE 17: Statement or defun catchall.
10660 (t
10661 (goto-char indent-point)
10662 ;; Back up statements until we find one that starts at boi.
10663 (while (let* ((prev-point (point))
10664 (last-step-type (c-beginning-of-statement-1
10665 containing-sexp)))
10666 (if (= (point) prev-point)
10667 (progn
10668 (setq step-type (or step-type last-step-type))
10669 nil)
10670 (setq step-type last-step-type)
10671 (/= (point) (c-point 'boi)))))
10672 (cond
0386b551 10673
9d1f0f55
AM
10674 ;; CASE 17B: continued statement
10675 ((and (eq step-type 'same)
10676 (/= (point) indent-point))
10677 (c-add-stmt-syntax 'statement-cont nil nil
10678 containing-sexp paren-state))
10679
10680 ;; CASE 17A: After a case/default label?
10681 ((progn
10682 (while (and (eq step-type 'label)
10683 (not (looking-at c-label-kwds-regexp)))
10684 (setq step-type
10685 (c-beginning-of-statement-1 containing-sexp)))
10686 (eq step-type 'label))
10687 (c-add-stmt-syntax (if (eq char-after-ip ?{)
10688 'statement-case-open
10689 'statement-case-intro)
10690 nil t containing-sexp paren-state))
10691
10692 ;; CASE 17D: any old statement
10693 ((progn
10694 (while (eq step-type 'label)
10695 (setq step-type
10696 (c-beginning-of-statement-1 containing-sexp)))
10697 (eq step-type 'previous))
10698 (c-add-stmt-syntax 'statement nil t
10699 containing-sexp paren-state)
10700 (if (eq char-after-ip ?{)
10701 (c-add-syntax 'block-open)))
10702
10703 ;; CASE 17I: Inside a substatement block.
10704 ((progn
10705 ;; The following tests are all based on containing-sexp.
10706 (goto-char containing-sexp)
10707 ;; From here on we have the next containing sexp in lim.
10708 (setq lim (c-most-enclosing-brace paren-state containing-sexp))
10709 (c-after-conditional))
10710 (c-backward-to-block-anchor lim)
10711 (c-add-stmt-syntax 'statement-block-intro nil t
10712 lim paren-state)
10713 (if (eq char-after-ip ?{)
10714 (c-add-syntax 'block-open)))
10715
10716 ;; CASE 17E: first statement in an in-expression block.
10717 ;; C.f. cases 4, 7B and 16A.
10718 ((setq placeholder (c-looking-at-inexpr-block
10719 (c-safe-position containing-sexp paren-state)
10720 nil))
10721 (setq tmpsymbol (if (eq (car placeholder) 'inlambda)
10722 'defun-block-intro
10723 'statement-block-intro))
10724 (back-to-indentation)
10725 (if (= containing-sexp (point))
10726 (c-add-syntax tmpsymbol (point))
10727 (goto-char (cdr placeholder))
a66cd3ee 10728 (back-to-indentation)
9d1f0f55
AM
10729 (c-add-stmt-syntax tmpsymbol nil t
10730 (c-most-enclosing-brace c-state-cache (point))
10731 paren-state)
10732 (if (/= (point) (cdr placeholder))
10733 (c-add-syntax (car placeholder))))
10734 (if (eq char-after-ip ?{)
10735 (c-add-syntax 'block-open)))
10736
10737 ;; CASE 17F: first statement in an inline, or first
10738 ;; statement in a top-level defun. we can tell this is it
10739 ;; if there are no enclosing braces that haven't been
10740 ;; narrowed out by a class (i.e. don't use bod here).
10741 ((save-excursion
10742 (or (not (setq placeholder (c-most-enclosing-brace
10743 paren-state)))
10744 (and (progn
10745 (goto-char placeholder)
10746 (eq (char-after) ?{))
10747 (c-looking-at-decl-block (c-most-enclosing-brace
10748 paren-state (point))
10749 nil))))
10750 (c-backward-to-decl-anchor lim)
10751 (back-to-indentation)
10752 (c-add-syntax 'defun-block-intro (point)))
0386b551 10753
9d1f0f55
AM
10754 ;; CASE 17G: First statement in a function declared inside
10755 ;; a normal block. This can occur in Pike and with
10756 ;; e.g. the gcc extensions, but watch out for macros
10757 ;; followed by blocks. C.f. cases B.3 and 16F.
10758 ((save-excursion
10759 (and (not (c-at-statement-start-p))
10760 (eq (c-beginning-of-statement-1 lim nil nil t) 'same)
10761 (setq placeholder (point))
10762 (let ((c-recognize-typeless-decls nil))
10763 ;; Turn off recognition of constructs that lacks
10764 ;; a type in this case, since that's more likely
10765 ;; to be a macro followed by a block.
10766 (c-forward-decl-or-cast-1 (c-point 'bosws) nil nil))))
10767 (back-to-indentation)
10768 (if (/= (point) containing-sexp)
10769 (goto-char placeholder))
10770 (c-add-stmt-syntax 'defun-block-intro nil t
10771 lim paren-state))
0386b551 10772
9d1f0f55
AM
10773 ;; CASE 17H: First statement in a block.
10774 (t
10775 ;; If the block is preceded by a case/switch label on the
10776 ;; same line, we anchor at the first preceding label at
10777 ;; boi. The default handling in c-add-stmt-syntax is
10778 ;; really fixes it better, but we do like this to keep the
10779 ;; indentation compatible with version 5.28 and earlier.
10780 ;; C.f. case 16C.
10781 (while (and (/= (setq placeholder (point)) (c-point 'boi))
10782 (eq (c-beginning-of-statement-1 lim) 'label)))
10783 (goto-char placeholder)
10784 (if (looking-at c-label-kwds-regexp)
10785 (c-add-syntax 'statement-block-intro (point))
10786 (goto-char containing-sexp)
10787 ;; c-backward-to-block-anchor not necessary here; those
10788 ;; situations are handled in case 17I above.
10789 (c-add-stmt-syntax 'statement-block-intro nil t
10790 lim paren-state))
10791 (if (eq char-after-ip ?{)
10792 (c-add-syntax 'block-open)))
10793 ))
10794 )
10795
10796 ;; now we need to look at any modifiers
10797 (goto-char indent-point)
10798 (skip-chars-forward " \t")
10799
10800 ;; are we looking at a comment only line?
10801 (when (and (looking-at c-comment-start-regexp)
10802 (/= (c-forward-token-2 0 nil (c-point 'eol)) 0))
10803 (c-append-syntax 'comment-intro))
10804
10805 ;; we might want to give additional offset to friends (in C++).
10806 (when (and c-opt-friend-key
10807 (looking-at c-opt-friend-key))
10808 (c-append-syntax 'friend))
10809
10810 ;; Set syntactic-relpos.
10811 (let ((p c-syntactic-context))
10812 (while (and p
10813 (if (integerp (c-langelem-pos (car p)))
10814 (progn
10815 (setq syntactic-relpos (c-langelem-pos (car p)))
10816 nil)
10817 t))
10818 (setq p (cdr p))))
0386b551 10819
9d1f0f55
AM
10820 ;; Start of or a continuation of a preprocessor directive?
10821 (if (and macro-start
10822 (eq macro-start (c-point 'boi))
10823 (not (and (c-major-mode-is 'pike-mode)
10824 (eq (char-after (1+ macro-start)) ?\"))))
10825 (c-append-syntax 'cpp-macro)
10826 (when (and c-syntactic-indentation-in-macros macro-start)
10827 (if in-macro-expr
10828 (when (or
10829 (< syntactic-relpos macro-start)
10830 (not (or
10831 (assq 'arglist-intro c-syntactic-context)
10832 (assq 'arglist-cont c-syntactic-context)
10833 (assq 'arglist-cont-nonempty c-syntactic-context)
10834 (assq 'arglist-close c-syntactic-context))))
10835 ;; If inside a cpp expression, i.e. anywhere in a
10836 ;; cpp directive except a #define body, we only let
10837 ;; through the syntactic analysis that is internal
10838 ;; in the expression. That means the arglist
10839 ;; elements, if they are anchored inside the cpp
10840 ;; expression.
10841 (setq c-syntactic-context nil)
10842 (c-add-syntax 'cpp-macro-cont macro-start))
10843 (when (and (eq macro-start syntactic-relpos)
10844 (not (assq 'cpp-define-intro c-syntactic-context))
10845 (save-excursion
10846 (goto-char macro-start)
10847 (or (not (c-forward-to-cpp-define-body))
10848 (<= (point) (c-point 'boi indent-point)))))
10849 ;; Inside a #define body and the syntactic analysis is
10850 ;; anchored on the start of the #define. In this case
10851 ;; we add cpp-define-intro to get the extra
10852 ;; indentation of the #define body.
10853 (c-add-syntax 'cpp-define-intro)))))
10854
10855 ;; return the syntax
10856 c-syntactic-context)))
785eecbb
RS
10857
10858\f
d9e94c22 10859;; Indentation calculation.
785eecbb 10860
130c507e
GM
10861(defun c-evaluate-offset (offset langelem symbol)
10862 ;; offset can be a number, a function, a variable, a list, or one of
10863 ;; the symbols + or -
0386b551
AM
10864 ;;
10865 ;; This function might do hidden buffer changes.
10866 (let ((res
10867 (cond
10868 ((numberp offset) offset)
10869 ((vectorp offset) offset)
10870 ((null offset) nil)
10871
10872 ((eq offset '+) c-basic-offset)
10873 ((eq offset '-) (- c-basic-offset))
10874 ((eq offset '++) (* 2 c-basic-offset))
10875 ((eq offset '--) (* 2 (- c-basic-offset)))
10876 ((eq offset '*) (/ c-basic-offset 2))
10877 ((eq offset '/) (/ (- c-basic-offset) 2))
10878
10879 ((functionp offset)
10880 (c-evaluate-offset
10881 (funcall offset
10882 (cons (c-langelem-sym langelem)
10883 (c-langelem-pos langelem)))
10884 langelem symbol))
10885
10886 ((listp offset)
10887 (cond
10888 ((eq (car offset) 'quote)
10889 (c-benign-error "The offset %S for %s was mistakenly quoted"
10890 offset symbol)
10891 nil)
10892
10893 ((memq (car offset) '(min max))
10894 (let (res val (method (car offset)))
10895 (setq offset (cdr offset))
10896 (while offset
10897 (setq val (c-evaluate-offset (car offset) langelem symbol))
10898 (cond
10899 ((not val))
10900 ((not res)
10901 (setq res val))
10902 ((integerp val)
10903 (if (vectorp res)
10904 (c-benign-error "\
10905Error evaluating offset %S for %s: \
10906Cannot combine absolute offset %S with relative %S in `%s' method"
10907 (car offset) symbol res val method)
10908 (setq res (funcall method res val))))
10909 (t
10910 (if (integerp res)
10911 (c-benign-error "\
10912Error evaluating offset %S for %s: \
10913Cannot combine relative offset %S with absolute %S in `%s' method"
10914 (car offset) symbol res val method)
10915 (setq res (vector (funcall method (aref res 0)
10916 (aref val 0)))))))
10917 (setq offset (cdr offset)))
10918 res))
10919
10920 ((eq (car offset) 'add)
10921 (let (res val)
10922 (setq offset (cdr offset))
10923 (while offset
10924 (setq val (c-evaluate-offset (car offset) langelem symbol))
10925 (cond
10926 ((not val))
10927 ((not res)
10928 (setq res val))
10929 ((integerp val)
10930 (if (vectorp res)
10931 (setq res (vector (+ (aref res 0) val)))
10932 (setq res (+ res val))))
10933 (t
10934 (if (vectorp res)
10935 (c-benign-error "\
10936Error evaluating offset %S for %s: \
10937Cannot combine absolute offsets %S and %S in `add' method"
10938 (car offset) symbol res val)
10939 (setq res val)))) ; Override.
10940 (setq offset (cdr offset)))
10941 res))
10942
10943 (t
10944 (let (res)
10945 (when (eq (car offset) 'first)
10946 (setq offset (cdr offset)))
10947 (while (and (not res) offset)
10948 (setq res (c-evaluate-offset (car offset) langelem symbol)
10949 offset (cdr offset)))
10950 res))))
10951
10952 ((and (symbolp offset) (boundp offset))
10953 (symbol-value offset))
10954
10955 (t
10956 (c-benign-error "Unknown offset format %S for %s" offset symbol)
10957 nil))))
10958
10959 (if (or (null res) (integerp res)
10960 (and (vectorp res) (= (length res) 1) (integerp (aref res 0))))
10961 res
10962 (c-benign-error "Error evaluating offset %S for %s: Got invalid value %S"
10963 offset symbol res)
10964 nil)))
130c507e 10965
d9e94c22
MS
10966(defun c-calc-offset (langelem)
10967 ;; Get offset from LANGELEM which is a list beginning with the
10968 ;; syntactic symbol and followed by any analysis data it provides.
10969 ;; That data may be zero or more elements, but if at least one is
0386b551
AM
10970 ;; given then the first is the anchor position (or nil). The symbol
10971 ;; is matched against `c-offsets-alist' and the offset calculated
10972 ;; from that is returned.
10973 ;;
10974 ;; This function might do hidden buffer changes.
10975 (let* ((symbol (c-langelem-sym langelem))
130c507e
GM
10976 (match (assq symbol c-offsets-alist))
10977 (offset (cdr-safe match)))
a66cd3ee
MS
10978 (if match
10979 (setq offset (c-evaluate-offset offset langelem symbol))
10980 (if c-strict-syntax-p
10981 (c-benign-error "No offset found for syntactic symbol %s" symbol))
10982 (setq offset 0))
130c507e
GM
10983 (if (vectorp offset)
10984 offset
a66cd3ee
MS
10985 (or (and (numberp offset) offset)
10986 (and (symbolp offset) (symbol-value offset))
10987 0))
130c507e
GM
10988 ))
10989
d9e94c22
MS
10990(defun c-get-offset (langelem)
10991 ;; This is a compatibility wrapper for `c-calc-offset' in case
10992 ;; someone is calling it directly. It takes an old style syntactic
0386b551
AM
10993 ;; element on the form (SYMBOL . ANCHOR-POS) and converts it to the
10994 ;; new list form.
10995 ;;
10996 ;; This function might do hidden buffer changes.
10997 (if (c-langelem-pos langelem)
10998 (c-calc-offset (list (c-langelem-sym langelem)
10999 (c-langelem-pos langelem)))
d9e94c22
MS
11000 (c-calc-offset langelem)))
11001
130c507e 11002(defun c-get-syntactic-indentation (langelems)
d9e94c22
MS
11003 ;; Calculate the syntactic indentation from a syntactic description
11004 ;; as returned by `c-guess-syntax'.
11005 ;;
0386b551 11006 ;; Note that topmost-intro always has an anchor position at bol, for
a66cd3ee
MS
11007 ;; historical reasons. It's often used together with other symbols
11008 ;; that has more sane positions. Since we always use the first
0386b551
AM
11009 ;; found anchor position, we rely on that these other symbols always
11010 ;; precede topmost-intro in the LANGELEMS list.
11011 ;;
11012 ;; This function might do hidden buffer changes.
a66cd3ee 11013 (let ((indent 0) anchor)
d9e94c22
MS
11014
11015 (while langelems
11016 (let* ((c-syntactic-element (car langelems))
11017 (res (c-calc-offset c-syntactic-element)))
11018
11019 (if (vectorp res)
11020 ;; Got an absolute column that overrides any indentation
11021 ;; we've collected so far, but not the relative
11022 ;; indentation we might get for the nested structures
11023 ;; further down the langelems list.
11024 (setq indent (elt res 0)
11025 anchor (point-min)) ; A position at column 0.
11026
11027 ;; Got a relative change of the current calculated
11028 ;; indentation.
11029 (setq indent (+ indent res))
11030
11031 ;; Use the anchor position from the first syntactic
11032 ;; element with one.
11033 (unless anchor
0386b551 11034 (setq anchor (c-langelem-pos (car langelems)))))
d9e94c22
MS
11035
11036 (setq langelems (cdr langelems))))
11037
11038 (if anchor
11039 (+ indent (save-excursion
11040 (goto-char anchor)
11041 (current-column)))
11042 indent)))
51f606de 11043
785eecbb 11044\f
130c507e 11045(cc-provide 'cc-engine)
3afbc435 11046
785eecbb 11047;;; cc-engine.el ends here