Fix indentation/fontification of Java enum with "implements"/generic.
[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
ab422c4d 3;; Copyright (C) 1985, 1987, 1992-2013 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)))
1264 ;; At a CPP construct?
1265 ((and c-opt-cpp-symbol (looking-at c-opt-cpp-symbol)
1266 (save-excursion
1267 (forward-line 0)
1268 (looking-at c-opt-cpp-prefix)))
1269 (c-end-of-macro))
1270 ((memq (char-after) non-skip-list)
1271 (throw 'done (point)))))
1272 ;; In trailing space after an as yet undetected virtual semicolon?
1273 (c-backward-syntactic-ws from)
b5eb9035
AM
1274 (when (and (bolp) (not (bobp))) ; Can happen in AWK Mode with an
1275 ; unterminated string/regexp.
1276 (backward-char))
536610a4
AM
1277 (if (and (< (point) to)
1278 (c-at-vsemi-p))
1279 (point)
1280 nil))))))
0386b551
AM
1281
1282(defun c-at-statement-start-p ()
1283 "Return non-nil if the point is at the first token in a statement
1284or somewhere in the syntactic whitespace before it.
1285
1286A \"statement\" here is not restricted to those inside code blocks.
1287Any kind of declaration-like construct that occur outside function
1288bodies is also considered a \"statement\".
1289
1290Note that this function might do hidden buffer changes. See the
1291comment at the start of cc-engine.el for more info."
1292
1293 (save-excursion
1294 (let ((end (point))
1295 c-maybe-labelp)
1296 (c-syntactic-skip-backward (substring c-stmt-delim-chars 1) nil t)
1297 (or (bobp)
1298 (eq (char-before) ?})
1299 (and (eq (char-before) ?{)
1300 (not (and c-special-brace-lists
1301 (progn (backward-char)
1302 (c-looking-at-special-brace-list)))))
1303 (c-crosses-statement-barrier-p (point) end)))))
1304
1305(defun c-at-expression-start-p ()
1306 "Return non-nil if the point is at the first token in an expression or
1307statement, or somewhere in the syntactic whitespace before it.
1308
1309An \"expression\" here is a bit different from the normal language
1310grammar sense: It's any sequence of expression tokens except commas,
1311unless they are enclosed inside parentheses of some kind. Also, an
1312expression never continues past an enclosing parenthesis, but it might
1313contain parenthesis pairs of any sort except braces.
1314
1315Since expressions never cross statement boundaries, this function also
1316recognizes statement beginnings, just like `c-at-statement-start-p'.
1317
1318Note that this function might do hidden buffer changes. See the
1319comment at the start of cc-engine.el for more info."
1320
1321 (save-excursion
1322 (let ((end (point))
1323 (c-stmt-delim-chars c-stmt-delim-chars-with-comma)
1324 c-maybe-labelp)
1325 (c-syntactic-skip-backward (substring c-stmt-delim-chars 1) nil t)
1326 (or (bobp)
1327 (memq (char-before) '(?{ ?}))
1328 (save-excursion (backward-char)
1329 (looking-at "\\s("))
1330 (c-crosses-statement-barrier-p (point) end)))))
785eecbb
RS
1331
1332\f
d9e94c22
MS
1333;; A set of functions that covers various idiosyncrasies in
1334;; implementations of `forward-comment'.
1335
1336;; Note: Some emacsen considers incorrectly that any line comment
1337;; ending with a backslash continues to the next line. I can't think
1338;; of any way to work around that in a reliable way without changing
1339;; the buffer, though. Suggestions welcome. ;) (No, temporarily
1340;; changing the syntax for backslash doesn't work since we must treat
1341;; escapes in string literals correctly.)
1342
1343(defun c-forward-single-comment ()
1344 "Move forward past whitespace and the closest following comment, if any.
1345Return t if a comment was found, nil otherwise. In either case, the
1346point is moved past the following whitespace. Line continuations,
1347i.e. a backslashes followed by line breaks, are treated as whitespace.
1348The line breaks that end line comments are considered to be the
1349comment enders, so the point will be put on the beginning of the next
1350line if it moved past a line comment.
1351
1352This function does not do any hidden buffer changes."
1353
1354 (let ((start (point)))
1355 (when (looking-at "\\([ \t\n\r\f\v]\\|\\\\[\n\r]\\)+")
1356 (goto-char (match-end 0)))
1357
1358 (when (forward-comment 1)
1359 (if (eobp)
1360 ;; Some emacsen (e.g. XEmacs 21) return t when moving
1361 ;; forwards at eob.
1362 nil
1363
1364 ;; Emacs includes the ending newline in a b-style (c++)
1365 ;; comment, but XEmacs doesn't. We depend on the Emacs
1366 ;; behavior (which also is symmetric).
1367 (if (and (eolp) (elt (parse-partial-sexp start (point)) 7))
1368 (condition-case nil (forward-char 1)))
1369
1370 t))))
1371
1372(defsubst c-forward-comments ()
1373 "Move forward past all following whitespace and comments.
1374Line continuations, i.e. a backslashes followed by line breaks, are
1375treated as whitespace.
1376
0386b551
AM
1377Note that this function might do hidden buffer changes. See the
1378comment at the start of cc-engine.el for more info."
d9e94c22
MS
1379
1380 (while (or
1381 ;; If forward-comment in at least XEmacs 21 is given a large
1382 ;; positive value, it'll loop all the way through if it hits
1383 ;; eob.
1384 (and (forward-comment 5)
1385 ;; Some emacsen (e.g. XEmacs 21) return t when moving
1386 ;; forwards at eob.
1387 (not (eobp)))
1388
1389 (when (looking-at "\\\\[\n\r]")
1390 (forward-char 2)
1391 t))))
1392
1393(defun c-backward-single-comment ()
1394 "Move backward past whitespace and the closest preceding comment, if any.
1395Return t if a comment was found, nil otherwise. In either case, the
1396point is moved past the preceding whitespace. Line continuations,
1397i.e. a backslashes followed by line breaks, are treated as whitespace.
1398The line breaks that end line comments are considered to be the
1399comment enders, so the point cannot be at the end of the same line to
1400move over a line comment.
1401
1402This function does not do any hidden buffer changes."
1403
1404 (let ((start (point)))
1405 ;; When we got newline terminated comments, forward-comment in all
1406 ;; supported emacsen so far will stop at eol of each line not
1407 ;; ending with a comment when moving backwards. This corrects for
1408 ;; that, and at the same time handles line continuations.
1409 (while (progn
1410 (skip-chars-backward " \t\n\r\f\v")
1411 (and (looking-at "[\n\r]")
0386b551 1412 (eq (char-before) ?\\)))
d9e94c22
MS
1413 (backward-char))
1414
1415 (if (bobp)
1416 ;; Some emacsen (e.g. Emacs 19.34) return t when moving
1417 ;; backwards at bob.
1418 nil
1419
1420 ;; Leave point after the closest following newline if we've
1421 ;; backed up over any above, since forward-comment won't move
1422 ;; backward over a line comment if point is at the end of the
1423 ;; same line.
1424 (re-search-forward "\\=\\s *[\n\r]" start t)
1425
bf2c1571 1426 (if (if (let (open-paren-in-column-0-is-defun-start) (forward-comment -1))
d9e94c22
MS
1427 (if (eolp)
1428 ;; If forward-comment above succeeded and we're at eol
1429 ;; then the newline we moved over above didn't end a
1430 ;; line comment, so we give it another go.
bf2c1571
AM
1431 (let (open-paren-in-column-0-is-defun-start)
1432 (forward-comment -1))
d9e94c22
MS
1433 t))
1434
1435 ;; Emacs <= 20 and XEmacs move back over the closer of a
1436 ;; block comment that lacks an opener.
1437 (if (looking-at "\\*/")
1438 (progn (forward-char 2) nil)
1439 t)))))
1440
1441(defsubst c-backward-comments ()
1442 "Move backward past all preceding whitespace and comments.
1443Line continuations, i.e. a backslashes followed by line breaks, are
1444treated as whitespace. The line breaks that end line comments are
1445considered to be the comment enders, so the point cannot be at the end
0386b551
AM
1446of the same line to move over a line comment. Unlike
1447c-backward-syntactic-ws, this function doesn't move back over
1448preprocessor directives.
d9e94c22 1449
0386b551
AM
1450Note that this function might do hidden buffer changes. See the
1451comment at the start of cc-engine.el for more info."
d9e94c22
MS
1452
1453 (let ((start (point)))
1454 (while (and
0386b551 1455 ;; `forward-comment' in some emacsen (e.g. XEmacs 21.4)
d9e94c22
MS
1456 ;; return t when moving backwards at bob.
1457 (not (bobp))
1458
14b8c3d9
AM
1459 (if (let (open-paren-in-column-0-is-defun-start moved-comment)
1460 (while
1461 (and (not (setq moved-comment (forward-comment -1)))
1462 ;; Cope specifically with ^M^J here -
1463 ;; forward-comment sometimes gets stuck after ^Ms,
1464 ;; sometimes after ^M^J.
1465 (or
1466 (when (eq (char-before) ?\r)
1467 (backward-char)
1468 t)
1469 (when (and (eq (char-before) ?\n)
1470 (eq (char-before (1- (point))) ?\r))
1471 (backward-char 2)
1472 t))))
1473 moved-comment)
d9e94c22
MS
1474 (if (looking-at "\\*/")
1475 ;; Emacs <= 20 and XEmacs move back over the
1476 ;; closer of a block comment that lacks an opener.
1477 (progn (forward-char 2) nil)
1478 t)
1479
1480 ;; XEmacs treats line continuations as whitespace but
1481 ;; only in the backward direction, which seems a bit
1482 ;; odd. Anyway, this is necessary for Emacs.
1483 (when (and (looking-at "[\n\r]")
1484 (eq (char-before) ?\\)
1485 (< (point) start))
1486 (backward-char)
1487 t))))))
1488
1489\f
d9e94c22 1490;; Tools for skipping over syntactic whitespace.
a66cd3ee 1491
d9e94c22
MS
1492;; The following functions use text properties to cache searches over
1493;; large regions of syntactic whitespace. It works as follows:
1494;;
1495;; o If a syntactic whitespace region contains anything but simple
1496;; whitespace (i.e. space, tab and line breaks), the text property
1497;; `c-in-sws' is put over it. At places where we have stopped
1498;; within that region there's also a `c-is-sws' text property.
1499;; That since there typically are nested whitespace inside that
1500;; must be handled separately, e.g. whitespace inside a comment or
1501;; cpp directive. Thus, from one point with `c-is-sws' it's safe
1502;; to jump to another point with that property within the same
1503;; `c-in-sws' region. It can be likened to a ladder where
1504;; `c-in-sws' marks the bars and `c-is-sws' the rungs.
1505;;
1506;; o The `c-is-sws' property is put on the simple whitespace chars at
1507;; a "rung position" and also maybe on the first following char.
1508;; As many characters as can be conveniently found in this range
1509;; are marked, but no assumption can be made that the whole range
1510;; is marked (it could be clobbered by later changes, for
1511;; instance).
1512;;
1513;; Note that some part of the beginning of a sequence of simple
1514;; whitespace might be part of the end of a preceding line comment
1515;; or cpp directive and must not be considered part of the "rung".
1516;; Such whitespace is some amount of horizontal whitespace followed
1517;; by a newline. In the case of cpp directives it could also be
1518;; two newlines with horizontal whitespace between them.
1519;;
1520;; The reason to include the first following char is to cope with
1521;; "rung positions" that doesn't have any ordinary whitespace. If
1522;; `c-is-sws' is put on a token character it does not have
1523;; `c-in-sws' set simultaneously. That's the only case when that
1524;; can occur, and the reason for not extending the `c-in-sws'
1525;; region to cover it is that the `c-in-sws' region could then be
1526;; accidentally merged with a following one if the token is only
1527;; one character long.
1528;;
1529;; o On buffer changes the `c-in-sws' and `c-is-sws' properties are
1530;; removed in the changed region. If the change was inside
1531;; syntactic whitespace that means that the "ladder" is broken, but
1532;; a later call to `c-forward-sws' or `c-backward-sws' will use the
1533;; parts on either side and use an ordinary search only to "repair"
1534;; the gap.
1535;;
1536;; Special care needs to be taken if a region is removed: If there
1537;; are `c-in-sws' on both sides of it which do not connect inside
1538;; the region then they can't be joined. If e.g. a marked macro is
1539;; broken, syntactic whitespace inside the new text might be
1540;; marked. If those marks would become connected with the old
1541;; `c-in-sws' range around the macro then we could get a ladder
1542;; with one end outside the macro and the other at some whitespace
1543;; within it.
1544;;
1545;; The main motivation for this system is to increase the speed in
1546;; skipping over the large whitespace regions that can occur at the
1547;; top level in e.g. header files that contain a lot of comments and
1548;; cpp directives. For small comments inside code it's probably
1549;; slower than using `forward-comment' straightforwardly, but speed is
1550;; not a significant factor there anyway.
1551
1552; (defface c-debug-is-sws-face
1553; '((t (:background "GreenYellow")))
1554; "Debug face to mark the `c-is-sws' property.")
1555; (defface c-debug-in-sws-face
1556; '((t (:underline t)))
1557; "Debug face to mark the `c-in-sws' property.")
1558
1559; (defun c-debug-put-sws-faces ()
1560; ;; Put the sws debug faces on all the `c-is-sws' and `c-in-sws'
1561; ;; properties in the buffer.
1562; (interactive)
1563; (save-excursion
0386b551 1564; (c-save-buffer-state (in-face)
d9e94c22
MS
1565; (goto-char (point-min))
1566; (setq in-face (if (get-text-property (point) 'c-is-sws)
1567; (point)))
1568; (while (progn
1569; (goto-char (next-single-property-change
1570; (point) 'c-is-sws nil (point-max)))
1571; (if in-face
1572; (progn
1573; (c-debug-add-face in-face (point) 'c-debug-is-sws-face)
1574; (setq in-face nil))
1575; (setq in-face (point)))
1576; (not (eobp))))
1577; (goto-char (point-min))
1578; (setq in-face (if (get-text-property (point) 'c-in-sws)
1579; (point)))
1580; (while (progn
1581; (goto-char (next-single-property-change
1582; (point) 'c-in-sws nil (point-max)))
1583; (if in-face
1584; (progn
1585; (c-debug-add-face in-face (point) 'c-debug-in-sws-face)
1586; (setq in-face nil))
1587; (setq in-face (point)))
1588; (not (eobp)))))))
1589
1590(defmacro c-debug-sws-msg (&rest args)
1591 ;;`(message ,@args)
1592 )
1593
1594(defmacro c-put-is-sws (beg end)
0386b551 1595 ;; This macro does a hidden buffer change.
d9e94c22
MS
1596 `(let ((beg ,beg) (end ,end))
1597 (put-text-property beg end 'c-is-sws t)
1598 ,@(when (facep 'c-debug-is-sws-face)
1599 `((c-debug-add-face beg end 'c-debug-is-sws-face)))))
1600
1601(defmacro c-put-in-sws (beg end)
0386b551 1602 ;; This macro does a hidden buffer change.
d9e94c22
MS
1603 `(let ((beg ,beg) (end ,end))
1604 (put-text-property beg end 'c-in-sws t)
1605 ,@(when (facep 'c-debug-is-sws-face)
1606 `((c-debug-add-face beg end 'c-debug-in-sws-face)))))
1607
1608(defmacro c-remove-is-sws (beg end)
0386b551 1609 ;; This macro does a hidden buffer change.
d9e94c22
MS
1610 `(let ((beg ,beg) (end ,end))
1611 (remove-text-properties beg end '(c-is-sws nil))
1612 ,@(when (facep 'c-debug-is-sws-face)
1613 `((c-debug-remove-face beg end 'c-debug-is-sws-face)))))
1614
1615(defmacro c-remove-in-sws (beg end)
0386b551 1616 ;; This macro does a hidden buffer change.
d9e94c22
MS
1617 `(let ((beg ,beg) (end ,end))
1618 (remove-text-properties beg end '(c-in-sws nil))
1619 ,@(when (facep 'c-debug-is-sws-face)
1620 `((c-debug-remove-face beg end 'c-debug-in-sws-face)))))
1621
1622(defmacro c-remove-is-and-in-sws (beg end)
0386b551 1623 ;; This macro does a hidden buffer change.
d9e94c22
MS
1624 `(let ((beg ,beg) (end ,end))
1625 (remove-text-properties beg end '(c-is-sws nil c-in-sws nil))
1626 ,@(when (facep 'c-debug-is-sws-face)
1627 `((c-debug-remove-face beg end 'c-debug-is-sws-face)
1628 (c-debug-remove-face beg end 'c-debug-in-sws-face)))))
1629
1630(defsubst c-invalidate-sws-region-after (beg end)
1631 ;; Called from `after-change-functions'. Note that if
1632 ;; `c-forward-sws' or `c-backward-sws' are used outside
1633 ;; `c-save-buffer-state' or similar then this will remove the cache
1634 ;; properties right after they're added.
0386b551
AM
1635 ;;
1636 ;; This function does hidden buffer changes.
d9e94c22
MS
1637
1638 (save-excursion
1639 ;; Adjust the end to remove the properties in any following simple
1640 ;; ws up to and including the next line break, if there is any
1641 ;; after the changed region. This is necessary e.g. when a rung
1642 ;; marked empty line is converted to a line comment by inserting
1643 ;; "//" before the line break. In that case the line break would
1644 ;; keep the rung mark which could make a later `c-backward-sws'
1645 ;; move into the line comment instead of over it.
1646 (goto-char end)
1647 (skip-chars-forward " \t\f\v")
1648 (when (and (eolp) (not (eobp)))
1649 (setq end (1+ (point)))))
1650
1651 (when (and (= beg end)
1652 (get-text-property beg 'c-in-sws)
2a15eb73 1653 (> beg (point-min))
d9e94c22
MS
1654 (get-text-property (1- beg) 'c-in-sws))
1655 ;; Ensure that an `c-in-sws' range gets broken. Note that it isn't
1656 ;; safe to keep a range that was continuous before the change. E.g:
1657 ;;
1658 ;; #define foo
1659 ;; \
1660 ;; bar
1661 ;;
1662 ;; There can be a "ladder" between "#" and "b". Now, if the newline
1663 ;; after "foo" is removed then "bar" will become part of the cpp
1664 ;; directive instead of a syntactically relevant token. In that
1665 ;; case there's no longer syntactic ws from "#" to "b".
1666 (setq beg (1- beg)))
1667
1668 (c-debug-sws-msg "c-invalidate-sws-region-after [%s..%s]" beg end)
1669 (c-remove-is-and-in-sws beg end))
1670
1671(defun c-forward-sws ()
1672 ;; Used by `c-forward-syntactic-ws' to implement the unbounded search.
0386b551
AM
1673 ;;
1674 ;; This function might do hidden buffer changes.
d9e94c22
MS
1675
1676 (let (;; `rung-pos' is set to a position as early as possible in the
1677 ;; unmarked part of the simple ws region.
1678 (rung-pos (point)) next-rung-pos rung-end-pos last-put-in-sws-pos
1679 rung-is-marked next-rung-is-marked simple-ws-end
1680 ;; `safe-start' is set when it's safe to cache the start position.
1681 ;; It's not set if we've initially skipped over comments and line
1682 ;; continuations since we might have gone out through the end of a
1683 ;; macro then. This provision makes `c-forward-sws' not populate the
1684 ;; cache in the majority of cases, but otoh is `c-backward-sws' by far
1685 ;; more common.
1686 safe-start)
1687
1688 ;; Skip simple ws and do a quick check on the following character to see
1689 ;; if it's anything that can't start syntactic ws, so we can bail out
1690 ;; early in the majority of cases when there just are a few ws chars.
1691 (skip-chars-forward " \t\n\r\f\v")
1692 (when (looking-at c-syntactic-ws-start)
1693
1694 (setq rung-end-pos (min (1+ (point)) (point-max)))
1695 (if (setq rung-is-marked (text-property-any rung-pos rung-end-pos
1696 'c-is-sws t))
1697 ;; Find the last rung position to avoid setting properties in all
1698 ;; the cases when the marked rung is complete.
1699 ;; (`next-single-property-change' is certain to move at least one
1700 ;; step forward.)
1701 (setq rung-pos (1- (next-single-property-change
1702 rung-is-marked 'c-is-sws nil rung-end-pos)))
1703 ;; Got no marked rung here. Since the simple ws might have started
1704 ;; inside a line comment or cpp directive we must set `rung-pos' as
1705 ;; high as possible.
1706 (setq rung-pos (point)))
1707
2143fa32 1708 (with-silent-modifications
d9e94c22
MS
1709 (while
1710 (progn
1711 (while
1712 (when (and rung-is-marked
1713 (get-text-property (point) 'c-in-sws))
1714
1715 ;; The following search is the main reason that `c-in-sws'
1716 ;; and `c-is-sws' aren't combined to one property.
1717 (goto-char (next-single-property-change
1718 (point) 'c-in-sws nil (point-max)))
1719 (unless (get-text-property (point) 'c-is-sws)
1720 ;; If the `c-in-sws' region extended past the last
1721 ;; `c-is-sws' char we have to go back a bit.
1722 (or (get-text-property (1- (point)) 'c-is-sws)
1723 (goto-char (previous-single-property-change
1724 (point) 'c-is-sws)))
1725 (backward-char))
1726
1727 (c-debug-sws-msg
1728 "c-forward-sws cached move %s -> %s (max %s)"
1729 rung-pos (point) (point-max))
1730
1731 (setq rung-pos (point))
1732 (and (> (skip-chars-forward " \t\n\r\f\v") 0)
1733 (not (eobp))))
1734
1735 ;; We'll loop here if there is simple ws after the last rung.
1736 ;; That means that there's been some change in it and it's
1737 ;; possible that we've stepped into another ladder, so extend
1738 ;; the previous one to join with it if there is one, and try to
1739 ;; use the cache again.
1740 (c-debug-sws-msg
1741 "c-forward-sws extending rung with [%s..%s] (max %s)"
1742 (1+ rung-pos) (1+ (point)) (point-max))
1743 (unless (get-text-property (point) 'c-is-sws)
1744 ;; Remove any `c-in-sws' property from the last char of
1745 ;; the rung before we mark it with `c-is-sws', so that we
1746 ;; won't connect with the remains of a broken "ladder".
1747 (c-remove-in-sws (point) (1+ (point))))
1748 (c-put-is-sws (1+ rung-pos)
1749 (1+ (point)))
1750 (c-put-in-sws rung-pos
1751 (setq rung-pos (point)
1752 last-put-in-sws-pos rung-pos)))
1753
1754 (setq simple-ws-end (point))
1755 (c-forward-comments)
1756
1757 (cond
1758 ((/= (point) simple-ws-end)
1759 ;; Skipped over comments. Don't cache at eob in case the buffer
1760 ;; is narrowed.
1761 (not (eobp)))
1762
1763 ((save-excursion
1764 (and c-opt-cpp-prefix
1765 (looking-at c-opt-cpp-start)
1766 (progn (skip-chars-backward " \t")
1767 (bolp))
1768 (or (bobp)
1769 (progn (backward-char)
1770 (not (eq (char-before) ?\\))))))
1771 ;; Skip a preprocessor directive.
1772 (end-of-line)
1773 (while (and (eq (char-before) ?\\)
1774 (= (forward-line 1) 0))
1775 (end-of-line))
1776 (forward-line 1)
1777 (setq safe-start t)
1778 ;; Don't cache at eob in case the buffer is narrowed.
1779 (not (eobp)))))
1780
1781 ;; We've searched over a piece of non-white syntactic ws. See if this
1782 ;; can be cached.
1783 (setq next-rung-pos (point))
1784 (skip-chars-forward " \t\n\r\f\v")
1785 (setq rung-end-pos (min (1+ (point)) (point-max)))
1786
1787 (if (or
1788 ;; Cache if we haven't skipped comments only, and if we started
1789 ;; either from a marked rung or from a completely uncached
1790 ;; position.
1791 (and safe-start
1792 (or rung-is-marked
1793 (not (get-text-property simple-ws-end 'c-in-sws))))
1794
1795 ;; See if there's a marked rung in the encountered simple ws. If
1796 ;; so then we can cache, unless `safe-start' is nil. Even then
1797 ;; we need to do this to check if the cache can be used for the
1798 ;; next step.
1799 (and (setq next-rung-is-marked
1800 (text-property-any next-rung-pos rung-end-pos
1801 'c-is-sws t))
1802 safe-start))
b2acd789 1803
0ec8351b 1804 (progn
d9e94c22
MS
1805 (c-debug-sws-msg
1806 "c-forward-sws caching [%s..%s] - [%s..%s] (max %s)"
1807 rung-pos (1+ simple-ws-end) next-rung-pos rung-end-pos
1808 (point-max))
1809
1810 ;; Remove the properties for any nested ws that might be cached.
1811 ;; Only necessary for `c-is-sws' since `c-in-sws' will be set
1812 ;; anyway.
1813 (c-remove-is-sws (1+ simple-ws-end) next-rung-pos)
1814 (unless (and rung-is-marked (= rung-pos simple-ws-end))
1815 (c-put-is-sws rung-pos
1816 (1+ simple-ws-end))
1817 (setq rung-is-marked t))
1818 (c-put-in-sws rung-pos
1819 (setq rung-pos (point)
1820 last-put-in-sws-pos rung-pos))
1821 (unless (get-text-property (1- rung-end-pos) 'c-is-sws)
1822 ;; Remove any `c-in-sws' property from the last char of
1823 ;; the rung before we mark it with `c-is-sws', so that we
1824 ;; won't connect with the remains of a broken "ladder".
1825 (c-remove-in-sws (1- rung-end-pos) rung-end-pos))
1826 (c-put-is-sws next-rung-pos
1827 rung-end-pos))
1828
1829 (c-debug-sws-msg
1830 "c-forward-sws not caching [%s..%s] - [%s..%s] (max %s)"
1831 rung-pos (1+ simple-ws-end) next-rung-pos rung-end-pos
1832 (point-max))
1833
1834 ;; Set `rung-pos' for the next rung. It's the same thing here as
1835 ;; initially, except that the rung position is set as early as
1836 ;; possible since we can't be in the ending ws of a line comment or
1837 ;; cpp directive now.
1838 (if (setq rung-is-marked next-rung-is-marked)
1839 (setq rung-pos (1- (next-single-property-change
1840 rung-is-marked 'c-is-sws nil rung-end-pos)))
1841 (setq rung-pos next-rung-pos))
1842 (setq safe-start t)))
1843
1844 ;; Make sure that the newly marked `c-in-sws' region doesn't connect to
1845 ;; another one after the point (which might occur when editing inside a
1846 ;; comment or macro).
1847 (when (eq last-put-in-sws-pos (point))
1848 (cond ((< last-put-in-sws-pos (point-max))
1849 (c-debug-sws-msg
1850 "c-forward-sws clearing at %s for cache separation"
1851 last-put-in-sws-pos)
1852 (c-remove-in-sws last-put-in-sws-pos
1853 (1+ last-put-in-sws-pos)))
1854 (t
1855 ;; If at eob we have to clear the last character before the end
1856 ;; instead since the buffer might be narrowed and there might
1857 ;; be a `c-in-sws' after (point-max). In this case it's
1858 ;; necessary to clear both properties.
1859 (c-debug-sws-msg
1860 "c-forward-sws clearing thoroughly at %s for cache separation"
1861 (1- last-put-in-sws-pos))
1862 (c-remove-is-and-in-sws (1- last-put-in-sws-pos)
1863 last-put-in-sws-pos))))
2143fa32 1864 ))))
b2acd789 1865
d9e94c22
MS
1866(defun c-backward-sws ()
1867 ;; Used by `c-backward-syntactic-ws' to implement the unbounded search.
0386b551
AM
1868 ;;
1869 ;; This function might do hidden buffer changes.
d9e94c22
MS
1870
1871 (let (;; `rung-pos' is set to a position as late as possible in the unmarked
1872 ;; part of the simple ws region.
1873 (rung-pos (point)) next-rung-pos last-put-in-sws-pos
1874 rung-is-marked simple-ws-beg cmt-skip-pos)
1875
1876 ;; Skip simple horizontal ws and do a quick check on the preceding
91af3942 1877 ;; character to see if it's anything that can't end syntactic ws, so we can
d9e94c22
MS
1878 ;; bail out early in the majority of cases when there just are a few ws
1879 ;; chars. Newlines are complicated in the backward direction, so we can't
1880 ;; skip over them.
1881 (skip-chars-backward " \t\f")
1882 (when (and (not (bobp))
1883 (save-excursion
1884 (backward-char)
1885 (looking-at c-syntactic-ws-end)))
1886
1887 ;; Try to find a rung position in the simple ws preceding point, so that
1888 ;; we can get a cache hit even if the last bit of the simple ws has
1889 ;; changed recently.
1890 (setq simple-ws-beg (point))
1891 (skip-chars-backward " \t\n\r\f\v")
1892 (if (setq rung-is-marked (text-property-any
1893 (point) (min (1+ rung-pos) (point-max))
1894 'c-is-sws t))
1895 ;; `rung-pos' will be the earliest marked position, which means that
1896 ;; there might be later unmarked parts in the simple ws region.
1897 ;; It's not worth the effort to fix that; the last part of the
1898 ;; simple ws is also typically edited often, so it could be wasted.
1899 (goto-char (setq rung-pos rung-is-marked))
1900 (goto-char simple-ws-beg))
1901
2143fa32 1902 (with-silent-modifications
d9e94c22
MS
1903 (while
1904 (progn
1905 (while
1906 (when (and rung-is-marked
1907 (not (bobp))
1908 (get-text-property (1- (point)) 'c-in-sws))
1909
1910 ;; The following search is the main reason that `c-in-sws'
1911 ;; and `c-is-sws' aren't combined to one property.
1912 (goto-char (previous-single-property-change
1913 (point) 'c-in-sws nil (point-min)))
1914 (unless (get-text-property (point) 'c-is-sws)
1915 ;; If the `c-in-sws' region extended past the first
1916 ;; `c-is-sws' char we have to go forward a bit.
1917 (goto-char (next-single-property-change
1918 (point) 'c-is-sws)))
1919
1920 (c-debug-sws-msg
1921 "c-backward-sws cached move %s <- %s (min %s)"
1922 (point) rung-pos (point-min))
1923
1924 (setq rung-pos (point))
1925 (if (and (< (min (skip-chars-backward " \t\f\v")
1926 (progn
1927 (setq simple-ws-beg (point))
1928 (skip-chars-backward " \t\n\r\f\v")))
1929 0)
1930 (setq rung-is-marked
1931 (text-property-any (point) rung-pos
1932 'c-is-sws t)))
1933 t
1934 (goto-char simple-ws-beg)
1935 nil))
1936
1937 ;; We'll loop here if there is simple ws before the first rung.
1938 ;; That means that there's been some change in it and it's
1939 ;; possible that we've stepped into another ladder, so extend
1940 ;; the previous one to join with it if there is one, and try to
1941 ;; use the cache again.
1942 (c-debug-sws-msg
1943 "c-backward-sws extending rung with [%s..%s] (min %s)"
1944 rung-is-marked rung-pos (point-min))
1945 (unless (get-text-property (1- rung-pos) 'c-is-sws)
1946 ;; Remove any `c-in-sws' property from the last char of
1947 ;; the rung before we mark it with `c-is-sws', so that we
1948 ;; won't connect with the remains of a broken "ladder".
1949 (c-remove-in-sws (1- rung-pos) rung-pos))
1950 (c-put-is-sws rung-is-marked
1951 rung-pos)
1952 (c-put-in-sws rung-is-marked
1953 (1- rung-pos))
1954 (setq rung-pos rung-is-marked
1955 last-put-in-sws-pos rung-pos))
1956
1957 (c-backward-comments)
1958 (setq cmt-skip-pos (point))
a66cd3ee 1959
d9e94c22
MS
1960 (cond
1961 ((and c-opt-cpp-prefix
1962 (/= cmt-skip-pos simple-ws-beg)
1963 (c-beginning-of-macro))
1964 ;; Inside a cpp directive. See if it should be skipped over.
1965 (let ((cpp-beg (point)))
1966
1967 ;; Move back over all line continuations in the region skipped
1968 ;; over by `c-backward-comments'. If we go past it then we
1969 ;; started inside the cpp directive.
1970 (goto-char simple-ws-beg)
1971 (beginning-of-line)
1972 (while (and (> (point) cmt-skip-pos)
1973 (progn (backward-char)
1974 (eq (char-before) ?\\)))
1975 (beginning-of-line))
1976
1977 (if (< (point) cmt-skip-pos)
1978 ;; Don't move past the cpp directive if we began inside
1979 ;; it. Note that the position at the end of the last line
1980 ;; of the macro is also considered to be within it.
1981 (progn (goto-char cmt-skip-pos)
1982 nil)
1983
1984 ;; It's worthwhile to spend a little bit of effort on finding
1985 ;; the end of the macro, to get a good `simple-ws-beg'
1986 ;; position for the cache. Note that `c-backward-comments'
1987 ;; could have stepped over some comments before going into
1988 ;; the macro, and then `simple-ws-beg' must be kept on the
1989 ;; same side of those comments.
1990 (goto-char simple-ws-beg)
1991 (skip-chars-backward " \t\n\r\f\v")
1992 (if (eq (char-before) ?\\)
1993 (forward-char))
1994 (forward-line 1)
1995 (if (< (point) simple-ws-beg)
1996 ;; Might happen if comments after the macro were skipped
1997 ;; over.
1998 (setq simple-ws-beg (point)))
1999
2000 (goto-char cpp-beg)
2001 t)))
2002
2003 ((/= (save-excursion
2004 (skip-chars-forward " \t\n\r\f\v" simple-ws-beg)
2005 (setq next-rung-pos (point)))
2006 simple-ws-beg)
2007 ;; Skipped over comments. Must put point at the end of
2008 ;; the simple ws at point since we might be after a line
2009 ;; comment or cpp directive that's been partially
2010 ;; narrowed out, and we can't risk marking the simple ws
2011 ;; at the end of it.
2012 (goto-char next-rung-pos)
2013 t)))
2014
2015 ;; We've searched over a piece of non-white syntactic ws. See if this
2016 ;; can be cached.
2017 (setq next-rung-pos (point))
2018 (skip-chars-backward " \t\f\v")
2019
2020 (if (or
2021 ;; Cache if we started either from a marked rung or from a
2022 ;; completely uncached position.
2023 rung-is-marked
2024 (not (get-text-property (1- simple-ws-beg) 'c-in-sws))
2025
2026 ;; Cache if there's a marked rung in the encountered simple ws.
2027 (save-excursion
2028 (skip-chars-backward " \t\n\r\f\v")
2029 (text-property-any (point) (min (1+ next-rung-pos) (point-max))
2030 'c-is-sws t)))
a66cd3ee 2031
d9e94c22
MS
2032 (progn
2033 (c-debug-sws-msg
2034 "c-backward-sws caching [%s..%s] - [%s..%s] (min %s)"
2035 (point) (1+ next-rung-pos)
2036 simple-ws-beg (min (1+ rung-pos) (point-max))
2037 (point-min))
2038
2039 ;; Remove the properties for any nested ws that might be cached.
2040 ;; Only necessary for `c-is-sws' since `c-in-sws' will be set
2041 ;; anyway.
2042 (c-remove-is-sws (1+ next-rung-pos) simple-ws-beg)
2043 (unless (and rung-is-marked (= simple-ws-beg rung-pos))
2044 (let ((rung-end-pos (min (1+ rung-pos) (point-max))))
2045 (unless (get-text-property (1- rung-end-pos) 'c-is-sws)
2046 ;; Remove any `c-in-sws' property from the last char of
2047 ;; the rung before we mark it with `c-is-sws', so that we
2048 ;; won't connect with the remains of a broken "ladder".
2049 (c-remove-in-sws (1- rung-end-pos) rung-end-pos))
2050 (c-put-is-sws simple-ws-beg
2051 rung-end-pos)
2052 (setq rung-is-marked t)))
2053 (c-put-in-sws (setq simple-ws-beg (point)
2054 last-put-in-sws-pos simple-ws-beg)
2055 rung-pos)
2056 (c-put-is-sws (setq rung-pos simple-ws-beg)
2057 (1+ next-rung-pos)))
2058
2059 (c-debug-sws-msg
2060 "c-backward-sws not caching [%s..%s] - [%s..%s] (min %s)"
2061 (point) (1+ next-rung-pos)
2062 simple-ws-beg (min (1+ rung-pos) (point-max))
2063 (point-min))
2064 (setq rung-pos next-rung-pos
2065 simple-ws-beg (point))
2066 ))
2067
2068 ;; Make sure that the newly marked `c-in-sws' region doesn't connect to
2069 ;; another one before the point (which might occur when editing inside a
2070 ;; comment or macro).
2071 (when (eq last-put-in-sws-pos (point))
2072 (cond ((< (point-min) last-put-in-sws-pos)
2073 (c-debug-sws-msg
2074 "c-backward-sws clearing at %s for cache separation"
2075 (1- last-put-in-sws-pos))
2076 (c-remove-in-sws (1- last-put-in-sws-pos)
2077 last-put-in-sws-pos))
2078 ((> (point-min) 1)
2079 ;; If at bob and the buffer is narrowed, we have to clear the
2080 ;; character we're standing on instead since there might be a
2081 ;; `c-in-sws' before (point-min). In this case it's necessary
2082 ;; to clear both properties.
2083 (c-debug-sws-msg
2084 "c-backward-sws clearing thoroughly at %s for cache separation"
2085 last-put-in-sws-pos)
2086 (c-remove-is-and-in-sws last-put-in-sws-pos
2087 (1+ last-put-in-sws-pos)))))
2143fa32 2088 ))))
785eecbb 2089
d9e94c22 2090\f
580fba94
AM
2091;; Other whitespace tools
2092(defun c-partial-ws-p (beg end)
2093 ;; Is the region (beg end) WS, and is there WS (or BOB/EOB) next to the
2094 ;; region? This is a "heuristic" function. .....
b414f371 2095 ;;
c421028a
AM
2096 ;; The motivation for the second bit is to check whether removing this
2097 ;; region would coalesce two symbols.
580fba94
AM
2098 ;;
2099 ;; FIXME!!! This function doesn't check virtual semicolons in any way. Be
2100 ;; careful about using this function for, e.g. AWK. (2007/3/7)
2101 (save-excursion
2102 (let ((end+1 (min (1+ end) (point-max))))
2103 (or (progn (goto-char (max (point-min) (1- beg)))
2104 (c-skip-ws-forward end)
2105 (eq (point) end))
2106 (progn (goto-char beg)
2107 (c-skip-ws-forward end+1)
2108 (eq (point) end+1))))))
2109\f
0386b551 2110;; A system for finding noteworthy parens before the point.
e1c458ae 2111
0ec1d2c5
AM
2112(defconst c-state-cache-too-far 5000)
2113;; A maximum comfortable scanning distance, e.g. between
2114;; `c-state-cache-good-pos' and "HERE" (where we call c-parse-state). When
0d26e0b6 2115;; this distance is exceeded, we take "emergency measures", e.g. by clearing
0ec1d2c5
AM
2116;; the cache and starting again from point-min or a beginning of defun. This
2117;; value can be tuned for efficiency or set to a lower value for testing.
2118
d9e94c22
MS
2119(defvar c-state-cache nil)
2120(make-variable-buffer-local 'c-state-cache)
2121;; The state cache used by `c-parse-state' to cut down the amount of
0ec1d2c5
AM
2122;; searching. It's the result from some earlier `c-parse-state' call. See
2123;; `c-parse-state''s doc string for details of its structure.
0386b551 2124;;
d9e94c22
MS
2125;; The use of the cached info is more effective if the next
2126;; `c-parse-state' call is on a line close by the one the cached state
2127;; was made at; the cache can actually slow down a little if the
2128;; cached state was made very far back in the buffer. The cache is
2129;; most effective if `c-parse-state' is used on each line while moving
2130;; forward.
e1c458ae 2131
0386b551
AM
2132(defvar c-state-cache-good-pos 1)
2133(make-variable-buffer-local 'c-state-cache-good-pos)
0ec1d2c5
AM
2134;; This is a position where `c-state-cache' is known to be correct, or
2135;; nil (see below). It's a position inside one of the recorded unclosed
2136;; parens or the top level, but not further nested inside any literal or
2137;; subparen that is closed before the last recorded position.
0386b551
AM
2138;;
2139;; The exact position is chosen to try to be close to yet earlier than
2140;; the position where `c-state-cache' will be called next. Right now
2141;; the heuristic is to set it to the position after the last found
2142;; closing paren (of any type) before the line on which
2143;; `c-parse-state' was called. That is chosen primarily to work well
2144;; with refontification of the current line.
0ec1d2c5
AM
2145;;
2146;; 2009-07-28: When `c-state-point-min' and the last position where
2147;; `c-parse-state' or for which `c-invalidate-state-cache' was called, are
2148;; both in the same literal, there is no such "good position", and
2149;; c-state-cache-good-pos is then nil. This is the ONLY circumstance in which
2150;; it can be nil. In this case, `c-state-point-min-literal' will be non-nil.
2151;;
2152;; 2009-06-12: In a brace desert, c-state-cache-good-pos may also be in
2153;; the middle of the desert, as long as it is not within a brace pair
2154;; recorded in `c-state-cache' or a paren/bracket pair.
2155
2156
2157;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2158;; We maintain a simple cache of positions which aren't in a literal, so as to
2159;; speed up testing for non-literality.
9657183b 2160(defconst c-state-nonlit-pos-interval 3000)
0ec1d2c5
AM
2161;; The approximate interval between entries in `c-state-nonlit-pos-cache'.
2162
2163(defvar c-state-nonlit-pos-cache nil)
2164(make-variable-buffer-local 'c-state-nonlit-pos-cache)
e84efb70
AM
2165;; A list of buffer positions which are known not to be in a literal or a cpp
2166;; construct. This is ordered with higher positions at the front of the list.
2167;; Only those which are less than `c-state-nonlit-pos-cache-limit' are valid.
0ec1d2c5
AM
2168
2169(defvar c-state-nonlit-pos-cache-limit 1)
2170(make-variable-buffer-local 'c-state-nonlit-pos-cache-limit)
2171;; An upper limit on valid entries in `c-state-nonlit-pos-cache'. This is
2172;; reduced by buffer changes, and increased by invocations of
2173;; `c-state-literal-at'.
2174
56d093a9
AM
2175(defvar c-state-semi-nonlit-pos-cache nil)
2176(make-variable-buffer-local 'c-state-semi-nonlit-pos-cache)
2177;; A list of buffer positions which are known not to be in a literal. This is
2178;; ordered with higher positions at the front of the list. Only those which
2179;; are less than `c-state-semi-nonlit-pos-cache-limit' are valid.
2180
2181(defvar c-state-semi-nonlit-pos-cache-limit 1)
2182(make-variable-buffer-local 'c-state-semi-nonlit-pos-cache-limit)
2183;; An upper limit on valid entries in `c-state-semi-nonlit-pos-cache'. This is
2184;; reduced by buffer changes, and increased by invocations of
f003f294 2185;; `c-state-literal-at'. FIXME!!!
56d093a9 2186
691e26ae 2187(defsubst c-state-pp-to-literal (from to &optional not-in-delimiter)
13d49cbb
AM
2188 ;; Do a parse-partial-sexp from FROM to TO, returning either
2189 ;; (STATE TYPE (BEG . END)) if TO is in a literal; or
2190 ;; (STATE) otherwise,
2191 ;; where STATE is the parsing state at TO, TYPE is the type of the literal
2192 ;; (one of 'c, 'c++, 'string) and (BEG . END) is the boundaries of the literal.
2193 ;;
691e26ae
AM
2194 ;; Unless NOT-IN-DELIMITER is non-nil, when TO is inside a two-character
2195 ;; comment opener, this is recognized as being in a comment literal.
2196 ;;
13d49cbb
AM
2197 ;; Only elements 3 (in a string), 4 (in a comment), 5 (following a quote),
2198 ;; 7 (comment type) and 8 (start of comment/string) (and possibly 9) of
2199 ;; STATE are valid.
0ec1d2c5 2200 (save-excursion
13d49cbb 2201 (let ((s (parse-partial-sexp from to))
691e26ae
AM
2202 ty co-st)
2203 (cond
2204 ((or (nth 3 s) (nth 4 s)) ; in a string or comment
13d49cbb
AM
2205 (setq ty (cond
2206 ((nth 3 s) 'string)
691e26ae 2207 ((nth 7 s) 'c++)
13d49cbb 2208 (t 'c)))
0ec1d2c5 2209 (parse-partial-sexp (point) (point-max)
691e26ae
AM
2210 nil ; TARGETDEPTH
2211 nil ; STOPBEFORE
2212 s ; OLDSTATE
2213 'syntax-table) ; stop at end of literal
2214 `(,s ,ty (,(nth 8 s) . ,(point))))
2215
2216 ((and (not not-in-delimiter) ; inside a comment starter
2217 (not (bobp))
2218 (progn (backward-char)
2219 (looking-at c-comment-start-regexp)))
2220 (setq ty (if (looking-at c-block-comment-start-regexp) 'c 'c++)
2221 co-st (point))
2222 (forward-comment 1)
2223 `(,s ,ty (,co-st . ,(point))))
2224
2225 (t `(,s))))))
13d49cbb
AM
2226
2227(defun c-state-safe-place (here)
2228 ;; Return a buffer position before HERE which is "safe", i.e. outside any
2229 ;; string, comment, or macro.
e84efb70
AM
2230 ;;
2231 ;; NOTE: This function manipulates `c-state-nonlit-pos-cache'. This cache
2232 ;; MAY NOT contain any positions within macros, since macros are frequently
2233 ;; turned into comments by use of the `c-cpp-delimiter' category properties.
2234 ;; We cannot rely on this mechanism whilst determining a cache pos since
2235 ;; this function is also called from outwith `c-parse-state'.
0ec1d2c5
AM
2236 (save-restriction
2237 (widen)
2238 (save-excursion
2239 (let ((c c-state-nonlit-pos-cache)
56d093a9 2240 pos npos high-pos lit macro-beg macro-end)
0ec1d2c5
AM
2241 ;; Trim the cache to take account of buffer changes.
2242 (while (and c (> (car c) c-state-nonlit-pos-cache-limit))
2243 (setq c (cdr c)))
2244 (setq c-state-nonlit-pos-cache c)
2245
2246 (while (and c (> (car c) here))
56d093a9 2247 (setq high-pos (car c))
0ec1d2c5
AM
2248 (setq c (cdr c)))
2249 (setq pos (or (car c) (point-min)))
2250
56d093a9
AM
2251 (unless high-pos
2252 (while
2253 ;; Add an element to `c-state-nonlit-pos-cache' each iteration.
2254 (and
2255 (<= (setq npos (+ pos c-state-nonlit-pos-interval)) here)
2256
2257 ;; Test for being in a literal. If so, go to after it.
2258 (progn
2259 (setq lit (car (cddr (c-state-pp-to-literal pos npos))))
2260 (or (null lit)
2261 (prog1 (<= (cdr lit) here)
2262 (setq npos (cdr lit)))))
2263
2264 ;; Test for being in a macro. If so, go to after it.
2265 (progn
2266 (goto-char npos)
2267 (setq macro-beg
2268 (and (c-beginning-of-macro) (/= (point) npos) (point)))
2269 (when macro-beg
2270 (c-syntactic-end-of-macro)
2271 (or (eobp) (forward-char))
2272 (setq macro-end (point)))
2273 (or (null macro-beg)
2274 (prog1 (<= macro-end here)
2275 (setq npos macro-end)))))
2276
2277 (setq pos npos)
2278 (setq c-state-nonlit-pos-cache (cons pos c-state-nonlit-pos-cache)))
2279 ;; Add one extra element above HERE so as to to avoid the previous
2280 ;; expensive calculation when the next call is close to the current
2281 ;; one. This is especially useful when inside a large macro.
2282 (setq c-state-nonlit-pos-cache (cons npos c-state-nonlit-pos-cache)))
0ec1d2c5
AM
2283
2284 (if (> pos c-state-nonlit-pos-cache-limit)
2285 (setq c-state-nonlit-pos-cache-limit pos))
13d49cbb 2286 pos))))
9858f6c3 2287
56d093a9
AM
2288(defun c-state-semi-safe-place (here)
2289 ;; Return a buffer position before HERE which is "safe", i.e. outside any
2290 ;; string or comment. It may be in a macro.
2291 (save-restriction
2292 (widen)
2293 (save-excursion
2294 (let ((c c-state-semi-nonlit-pos-cache)
2295 pos npos high-pos lit macro-beg macro-end)
2296 ;; Trim the cache to take account of buffer changes.
2297 (while (and c (> (car c) c-state-semi-nonlit-pos-cache-limit))
2298 (setq c (cdr c)))
2299 (setq c-state-semi-nonlit-pos-cache c)
c38e0c97 2300
56d093a9
AM
2301 (while (and c (> (car c) here))
2302 (setq high-pos (car c))
2303 (setq c (cdr c)))
2304 (setq pos (or (car c) (point-min)))
c38e0c97 2305
56d093a9
AM
2306 (unless high-pos
2307 (while
2308 ;; Add an element to `c-state-semi-nonlit-pos-cache' each iteration.
2309 (and
2310 (<= (setq npos (+ pos c-state-nonlit-pos-interval)) here)
c38e0c97 2311
56d093a9
AM
2312 ;; Test for being in a literal. If so, go to after it.
2313 (progn
2314 (setq lit (car (cddr (c-state-pp-to-literal pos npos))))
2315 (or (null lit)
2316 (prog1 (<= (cdr lit) here)
2317 (setq npos (cdr lit))))))
c38e0c97 2318
56d093a9
AM
2319 (setq pos npos)
2320 (setq c-state-semi-nonlit-pos-cache
2321 (cons pos c-state-semi-nonlit-pos-cache))))
2322
2323 (if (> pos c-state-semi-nonlit-pos-cache-limit)
2324 (setq c-state-semi-nonlit-pos-cache-limit pos))
2325 pos))))
2326
13d49cbb
AM
2327(defun c-state-literal-at (here)
2328 ;; If position HERE is inside a literal, return (START . END), the
2329 ;; boundaries of the literal (which may be outside the accessible bit of the
2330 ;; buffer). Otherwise, return nil.
2331 ;;
2332 ;; This function is almost the same as `c-literal-limits'. Previously, it
9858f6c3 2333 ;; differed in that it was a lower level function, and that it rigorously
13d49cbb
AM
2334 ;; followed the syntax from BOB. `c-literal-limits' is now (2011-12)
2335 ;; virtually identical to this function.
2336 (save-restriction
2337 (widen)
2338 (save-excursion
2339 (let ((pos (c-state-safe-place here)))
2340 (car (cddr (c-state-pp-to-literal pos here)))))))
0ec1d2c5
AM
2341
2342(defsubst c-state-lit-beg (pos)
2343 ;; Return the start of the literal containing POS, or POS itself.
2344 (or (car (c-state-literal-at pos))
2345 pos))
2346
2347(defsubst c-state-cache-non-literal-place (pos state)
9c4aeabf 2348 ;; Return a position outside of a string/comment/macro at or before POS.
0ec1d2c5 2349 ;; STATE is the parse-partial-sexp state at POS.
9c4aeabf
AM
2350 (let ((res (if (or (nth 3 state) ; in a string?
2351 (nth 4 state)) ; in a comment?
2352 (nth 8 state)
2353 pos)))
2354 (save-excursion
2355 (goto-char res)
2356 (if (c-beginning-of-macro)
2357 (point)
2358 res))))
0ec1d2c5
AM
2359
2360;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2361;; Stuff to do with point-min, and coping with any literal there.
2362(defvar c-state-point-min 1)
2363(make-variable-buffer-local 'c-state-point-min)
2364;; This is (point-min) when `c-state-cache' was last calculated. A change of
2365;; narrowing is likely to affect the parens that are visible before the point.
2366
2367(defvar c-state-point-min-lit-type nil)
2368(make-variable-buffer-local 'c-state-point-min-lit-type)
2369(defvar c-state-point-min-lit-start nil)
2370(make-variable-buffer-local 'c-state-point-min-lit-start)
2371;; These two variables define the literal, if any, containing point-min.
2372;; Their values are, respectively, 'string, c, or c++, and the start of the
2373;; literal. If there's no literal there, they're both nil.
2374
2375(defvar c-state-min-scan-pos 1)
2376(make-variable-buffer-local 'c-state-min-scan-pos)
2377;; This is the earliest buffer-pos from which scanning can be done. It is
2378;; either the end of the literal containing point-min, or point-min itself.
2379;; It becomes nil if the buffer is changed earlier than this point.
2380(defun c-state-get-min-scan-pos ()
2381 ;; Return the lowest valid scanning pos. This will be the end of the
2382 ;; literal enclosing point-min, or point-min itself.
2383 (or c-state-min-scan-pos
2384 (save-restriction
2385 (save-excursion
2386 (widen)
2387 (goto-char c-state-point-min-lit-start)
2388 (if (eq c-state-point-min-lit-type 'string)
2389 (forward-sexp)
2390 (forward-comment 1))
2391 (setq c-state-min-scan-pos (point))))))
2392
2393(defun c-state-mark-point-min-literal ()
2394 ;; Determine the properties of any literal containing POINT-MIN, setting the
2395 ;; variables `c-state-point-min-lit-type', `c-state-point-min-lit-start',
2396 ;; and `c-state-min-scan-pos' accordingly. The return value is meaningless.
2397 (let ((p-min (point-min))
2398 lit)
2399 (save-restriction
2400 (widen)
2401 (setq lit (c-state-literal-at p-min))
2402 (if lit
2403 (setq c-state-point-min-lit-type
2404 (save-excursion
2405 (goto-char (car lit))
2406 (cond
2407 ((looking-at c-block-comment-start-regexp) 'c)
2408 ((looking-at c-line-comment-starter) 'c++)
2409 (t 'string)))
2410 c-state-point-min-lit-start (car lit)
2411 c-state-min-scan-pos (cdr lit))
2412 (setq c-state-point-min-lit-type nil
2413 c-state-point-min-lit-start nil
2414 c-state-min-scan-pos p-min)))))
2415
2416
2417;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2418;; A variable which signals a brace dessert - helpful for reducing the number
2419;; of fruitless backward scans.
2420(defvar c-state-brace-pair-desert nil)
2421(make-variable-buffer-local 'c-state-brace-pair-desert)
e84efb70 2422;; Used only in `c-append-lower-brace-pair-to-state-cache'. It is set when
0ec1d2c5
AM
2423;; that defun has searched backwards for a brace pair and not found one. Its
2424;; value is either nil or a cons (PA . FROM), where PA is the position of the
2425;; enclosing opening paren/brace/bracket which bounds the backwards search (or
2426;; nil when at top level) and FROM is where the backward search started. It
2427;; is reset to nil in `c-invalidate-state-cache'.
2428
2429
2430;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2431;; Lowish level functions/macros which work directly on `c-state-cache', or a
2432;; list of like structure.
2433(defmacro c-state-cache-top-lparen (&optional cache)
2434 ;; Return the address of the top left brace/bracket/paren recorded in CACHE
2435 ;; (default `c-state-cache') (or nil).
2436 (let ((cash (or cache 'c-state-cache)))
2437 `(if (consp (car ,cash))
2438 (caar ,cash)
2439 (car ,cash))))
2440
2441(defmacro c-state-cache-top-paren (&optional cache)
2442 ;; Return the address of the latest brace/bracket/paren (whether left or
2443 ;; right) recorded in CACHE (default `c-state-cache') or nil.
2444 (let ((cash (or cache 'c-state-cache)))
2445 `(if (consp (car ,cash))
2446 (cdar ,cash)
2447 (car ,cash))))
2448
2449(defmacro c-state-cache-after-top-paren (&optional cache)
2450 ;; Return the position just after the latest brace/bracket/paren (whether
2451 ;; left or right) recorded in CACHE (default `c-state-cache') or nil.
2452 (let ((cash (or cache 'c-state-cache)))
2453 `(if (consp (car ,cash))
2454 (cdar ,cash)
2455 (and (car ,cash)
2456 (1+ (car ,cash))))))
2457
2458(defun c-get-cache-scan-pos (here)
2459 ;; From the state-cache, determine the buffer position from which we might
2460 ;; scan forward to HERE to update this cache. This position will be just
2461 ;; after a paren/brace/bracket recorded in the cache, if possible, otherwise
2462 ;; return the earliest position in the accessible region which isn't within
2463 ;; a literal. If the visible portion of the buffer is entirely within a
2464 ;; literal, return NIL.
2465 (let ((c c-state-cache) elt)
2466 ;(while (>= (or (c-state-cache-top-lparen c) 1) here)
2467 (while (and c
2468 (>= (c-state-cache-top-lparen c) here))
2469 (setq c (cdr c)))
2470
2471 (setq elt (car c))
2472 (cond
2473 ((consp elt)
2474 (if (> (cdr elt) here)
2475 (1+ (car elt))
2476 (cdr elt)))
2477 (elt (1+ elt))
2478 ((<= (c-state-get-min-scan-pos) here)
2479 (c-state-get-min-scan-pos))
2480 (t nil))))
2481
2482;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2483;; Variables which keep track of preprocessor constructs.
8cff5feb
AM
2484(defvar c-state-old-cpp-beg-marker nil)
2485(make-variable-buffer-local 'c-state-old-cpp-beg-marker)
0ec1d2c5
AM
2486(defvar c-state-old-cpp-beg nil)
2487(make-variable-buffer-local 'c-state-old-cpp-beg)
8cff5feb
AM
2488(defvar c-state-old-cpp-end-marker nil)
2489(make-variable-buffer-local 'c-state-old-cpp-end-marker)
0ec1d2c5
AM
2490(defvar c-state-old-cpp-end nil)
2491(make-variable-buffer-local 'c-state-old-cpp-end)
2492;; These are the limits of the macro containing point at the previous call of
2493;; `c-parse-state', or nil.
2494
2495;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
c7015153 2496;; Defuns which analyze the buffer, yet don't change `c-state-cache'.
c0209c2c
AM
2497(defun c-state-balance-parens-backwards (here- here+ top)
2498 ;; Return the position of the opening paren/brace/bracket before HERE- which
2499 ;; matches the outermost close p/b/b between HERE+ and TOP. Except when
2500 ;; there's a macro, HERE- and HERE+ are the same. Like this:
0ec1d2c5 2501 ;;
c0209c2c
AM
2502 ;; ............................................
2503 ;; | |
2504 ;; ( [ ( .........#macro.. ) ( ) ] )
2505 ;; ^ ^ ^ ^
2506 ;; | | | |
2507 ;; return HERE- HERE+ TOP
0ec1d2c5
AM
2508 ;;
2509 ;; If there aren't enough opening paren/brace/brackets, return the position
c0209c2c 2510 ;; of the outermost one found, or HERE- if there are none. If there are no
da6062e6 2511 ;; closing p/b/bs between HERE+ and TOP, return HERE-. HERE-/+ and TOP
c0209c2c
AM
2512 ;; must not be inside literals. Only the accessible portion of the buffer
2513 ;; will be scanned.
2514
2515 ;; PART 1: scan from `here+' up to `top', accumulating ")"s which enclose
2516 ;; `here'. Go round the next loop each time we pass over such a ")". These
2517 ;; probably match "("s before `here-'.
0ec1d2c5
AM
2518 (let (pos pa ren+1 lonely-rens)
2519 (save-excursion
2520 (save-restriction
2521 (narrow-to-region (point-min) top) ; This can move point, sometimes.
c0209c2c 2522 (setq pos here+)
0ec1d2c5
AM
2523 (c-safe
2524 (while
2525 (setq ren+1 (scan-lists pos 1 1)) ; might signal
2526 (setq lonely-rens (cons ren+1 lonely-rens)
2527 pos ren+1)))))
2528
c0209c2c 2529 ;; PART 2: Scan back before `here-' searching for the "("s
0ec1d2c5
AM
2530 ;; matching/mismatching the ")"s found above. We only need to direct the
2531 ;; caller to scan when we've encountered unmatched right parens.
c0209c2c
AM
2532 (setq pos here-)
2533 (when lonely-rens
2534 (c-safe
2535 (while
2536 (and lonely-rens ; actual values aren't used.
2537 (setq pa (scan-lists pos -1 1)))
2538 (setq pos pa)
2539 (setq lonely-rens (cdr lonely-rens)))))
2540 pos))
0ec1d2c5
AM
2541
2542(defun c-parse-state-get-strategy (here good-pos)
2543 ;; Determine the scanning strategy for adjusting `c-parse-state', attempting
53964682 2544 ;; to minimize the amount of scanning. HERE is the pertinent position in
0ec1d2c5
AM
2545 ;; the buffer, GOOD-POS is a position where `c-state-cache' (possibly with
2546 ;; its head trimmed) is known to be good, or nil if there is no such
2547 ;; position.
2548 ;;
2549 ;; The return value is a list, one of the following:
2550 ;;
267eacb5
AM
2551 ;; o - ('forward START-POINT) - scan forward from START-POINT,
2552 ;; which is not less than the highest position in `c-state-cache' below here.
2553 ;; o - ('backward nil) - scan backwards (from HERE).
267eacb5 2554 ;; o - ('IN-LIT nil) - point is inside the literal containing point-min.
0ec1d2c5 2555 (let ((cache-pos (c-get-cache-scan-pos here)) ; highest position below HERE in cache (or 1)
d23b0804
AM
2556 strategy ; 'forward, 'backward, or 'IN-LIT.
2557 start-point)
0ec1d2c5
AM
2558 (setq good-pos (or good-pos (c-state-get-min-scan-pos)))
2559 (cond
2560 ((< here (c-state-get-min-scan-pos))
d23b0804 2561 (setq strategy 'IN-LIT))
0ec1d2c5
AM
2562 ((<= good-pos here)
2563 (setq strategy 'forward
d23b0804 2564 start-point (max good-pos cache-pos)))
0ec1d2c5 2565 ((< (- good-pos here) (- here cache-pos)) ; FIXME!!! ; apply some sort of weighting.
d23b0804 2566 (setq strategy 'backward))
0ec1d2c5
AM
2567 (t
2568 (setq strategy 'forward
d23b0804
AM
2569 start-point cache-pos)))
2570 (list strategy (and (eq strategy 'forward) start-point))))
0ec1d2c5
AM
2571
2572
2573;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2574;; Routines which change `c-state-cache' and associated values.
2575(defun c-renarrow-state-cache ()
2576 ;; The region (more precisely, point-min) has changed since we
2577 ;; calculated `c-state-cache'. Amend `c-state-cache' accordingly.
2578 (if (< (point-min) c-state-point-min)
2579 ;; If point-min has MOVED BACKWARDS then we drop the state completely.
2580 ;; It would be possible to do a better job here and recalculate the top
2581 ;; only.
2582 (progn
2583 (c-state-mark-point-min-literal)
2584 (setq c-state-cache nil
2585 c-state-cache-good-pos c-state-min-scan-pos
2586 c-state-brace-pair-desert nil))
2587
2588 ;; point-min has MOVED FORWARD.
2589
2590 ;; Is the new point-min inside a (different) literal?
2591 (unless (and c-state-point-min-lit-start ; at prev. point-min
2592 (< (point-min) (c-state-get-min-scan-pos)))
2593 (c-state-mark-point-min-literal))
2594
2595 ;; Cut off a bit of the tail from `c-state-cache'.
2596 (let ((ptr (cons nil c-state-cache))
2597 pa)
2598 (while (and (setq pa (c-state-cache-top-lparen (cdr ptr)))
2599 (>= pa (point-min)))
2600 (setq ptr (cdr ptr)))
2601
2602 (when (consp ptr)
2603 (if (eq (cdr ptr) c-state-cache)
2604 (setq c-state-cache nil
2605 c-state-cache-good-pos c-state-min-scan-pos)
2606 (setcdr ptr nil)
2607 (setq c-state-cache-good-pos (1+ (c-state-cache-top-lparen))))
2608 )))
2609
2610 (setq c-state-point-min (point-min)))
2611
d23b0804 2612(defun c-append-lower-brace-pair-to-state-cache (from here &optional upper-lim)
583e23bd
AM
2613 ;; If there is a brace pair preceding FROM in the buffer, at the same level
2614 ;; of nesting (not necessarily immediately preceding), push a cons onto
2615 ;; `c-state-cache' to represent it. FROM must not be inside a literal. If
2616 ;; UPPER-LIM is non-nil, we append the highest brace pair whose "}" is below
2617 ;; UPPER-LIM.
0ec1d2c5
AM
2618 ;;
2619 ;; Return non-nil when this has been done.
2620 ;;
583e23bd
AM
2621 ;; The situation it copes with is this transformation:
2622 ;;
2623 ;; OLD: { (.) {...........}
2624 ;; ^ ^
2625 ;; FROM HERE
c38e0c97 2626 ;;
583e23bd
AM
2627 ;; NEW: { {....} (.) {.........
2628 ;; ^ ^ ^
2629 ;; LOWER BRACE PAIR HERE or HERE
c38e0c97 2630 ;;
0ec1d2c5
AM
2631 ;; This routine should be fast. Since it can get called a LOT, we maintain
2632 ;; `c-state-brace-pair-desert', a small cache of "failures", such that we
2633 ;; reduce the time wasted in repeated fruitless searches in brace deserts.
2634 (save-excursion
2635 (save-restriction
d23b0804 2636 (let* (new-cons
8cff5feb
AM
2637 (cache-pos (c-state-cache-top-lparen)) ; might be nil.
2638 (macro-start-or-from
2639 (progn (goto-char from)
2640 (c-beginning-of-macro)
2641 (point)))
2642 (bra ; Position of "{".
2643 ;; Don't start scanning in the middle of a CPP construct unless
2644 ;; it contains HERE - these constructs, in Emacs, are "commented
2645 ;; out" with category properties.
2646 (if (eq (c-get-char-property macro-start-or-from 'category)
2647 'c-cpp-delimiter)
2648 macro-start-or-from
2649 from))
2650 ce) ; Position of "}"
0ec1d2c5
AM
2651 (or upper-lim (setq upper-lim from))
2652
2653 ;; If we're essentially repeating a fruitless search, just give up.
2654 (unless (and c-state-brace-pair-desert
2655 (eq cache-pos (car c-state-brace-pair-desert))
b03aabda
AM
2656 (or (null (car c-state-brace-pair-desert))
2657 (> from (car c-state-brace-pair-desert)))
0ec1d2c5 2658 (<= from (cdr c-state-brace-pair-desert)))
b03aabda 2659 ;; DESERT-LIM. Avoid repeated searching through the cached desert.
583e23bd
AM
2660 (let ((desert-lim
2661 (and c-state-brace-pair-desert
2662 (eq cache-pos (car c-state-brace-pair-desert))
b03aabda 2663 (>= from (cdr c-state-brace-pair-desert))
583e23bd
AM
2664 (cdr c-state-brace-pair-desert)))
2665 ;; CACHE-LIM. This limit will be necessary when an opening
2666 ;; paren at `cache-pos' has just had its matching close paren
b03aabda
AM
2667 ;; inserted into the buffer. `cache-pos' continues to be a
2668 ;; search bound, even though the algorithm below would skip
2669 ;; over the new paren pair.
583e23bd
AM
2670 (cache-lim (and cache-pos (< cache-pos from) cache-pos)))
2671 (narrow-to-region
2672 (cond
2673 ((and desert-lim cache-lim)
2674 (max desert-lim cache-lim))
2675 (desert-lim)
2676 (cache-lim)
2677 ((point-min)))
d23b0804
AM
2678 ;; The top limit is EOB to ensure that `bra' is inside the
2679 ;; accessible part of the buffer at the next scan operation.
2680 (1+ (buffer-size))))
0ec1d2c5
AM
2681
2682 ;; In the next pair of nested loops, the inner one moves back past a
2683 ;; pair of (mis-)matching parens or brackets; the outer one moves
2684 ;; back over a sequence of unmatched close brace/paren/bracket each
2685 ;; time round.
2686 (while
2687 (progn
2688 (c-safe
2689 (while
2690 (and (setq ce (scan-lists bra -1 -1)) ; back past )/]/}; might signal
2691 (setq bra (scan-lists ce -1 1)) ; back past (/[/{; might signal
267eacb5
AM
2692 (or (> bra here) ;(> ce here)
2693 (and
2694 (< ce here)
2695 (or (not (eq (char-after bra) ?\{))
2696 (and (goto-char bra)
2697 (c-beginning-of-macro)
2698 (< (point) macro-start-or-from))))))))
0ec1d2c5
AM
2699 (and ce (< ce bra)))
2700 (setq bra ce)) ; If we just backed over an unbalanced closing
2701 ; brace, ignore it.
2702
267eacb5 2703 (if (and ce (< ce here) (< bra ce) (eq (char-after bra) ?\{))
0ec1d2c5
AM
2704 ;; We've found the desired brace-pair.
2705 (progn
2706 (setq new-cons (cons bra (1+ ce)))
2707 (cond
2708 ((consp (car c-state-cache))
2709 (setcar c-state-cache new-cons))
2710 ((and (numberp (car c-state-cache)) ; probably never happens
2711 (< ce (car c-state-cache)))
2712 (setcdr c-state-cache
2713 (cons new-cons (cdr c-state-cache))))
2714 (t (setq c-state-cache (cons new-cons c-state-cache)))))
2715
583e23bd 2716 ;; We haven't found a brace pair. Record this in the cache.
267eacb5
AM
2717 (setq c-state-brace-pair-desert
2718 (cons (if (and ce (< bra ce) (> ce here)) ; {..} straddling HERE?
2719 bra
2720 (point-min))
2721 (min here from)))))))))
0ec1d2c5
AM
2722
2723(defsubst c-state-push-any-brace-pair (bra+1 macro-start-or-here)
2724 ;; If BRA+1 is nil, do nothing. Otherwise, BRA+1 is the buffer position
2725 ;; following a {, and that brace has a (mis-)matching } (or ]), and we
2726 ;; "push" "a" brace pair onto `c-state-cache'.
2727 ;;
2728 ;; Here "push" means overwrite the top element if it's itself a brace-pair,
2729 ;; otherwise push it normally.
2730 ;;
2731 ;; The brace pair we push is normally the one surrounding BRA+1, but if the
2732 ;; latter is inside a macro, not being a macro containing
2733 ;; MACRO-START-OR-HERE, we scan backwards through the buffer for a non-macro
2734 ;; base pair. This latter case is assumed to be rare.
2735 ;;
2736 ;; Note: POINT is not preserved in this routine.
2737 (if bra+1
2738 (if (or (> bra+1 macro-start-or-here)
2739 (progn (goto-char bra+1)
2740 (not (c-beginning-of-macro))))
2741 (setq c-state-cache
2742 (cons (cons (1- bra+1)
2743 (scan-lists bra+1 1 1))
2744 (if (consp (car c-state-cache))
2745 (cdr c-state-cache)
2746 c-state-cache)))
d23b0804 2747 ;; N.B. This defsubst codes one method for the simple, normal case,
0ec1d2c5 2748 ;; and a more sophisticated, slower way for the general case. Don't
e1dbe924 2749 ;; eliminate this defsubst - it's a speed optimization.
d23b0804 2750 (c-append-lower-brace-pair-to-state-cache (1- bra+1) (point-max)))))
0ec1d2c5 2751
d23b0804
AM
2752(defun c-append-to-state-cache (from here)
2753 ;; Scan the buffer from FROM to HERE, adding elements into `c-state-cache'
2754 ;; for braces etc. Return a candidate for `c-state-cache-good-pos'.
0ec1d2c5
AM
2755 ;;
2756 ;; FROM must be after the latest brace/paren/bracket in `c-state-cache', if
2757 ;; any. Typically, it is immediately after it. It must not be inside a
2758 ;; literal.
d23b0804 2759 (let ((here-bol (c-point 'bol here))
0ec1d2c5 2760 (macro-start-or-here
d23b0804 2761 (save-excursion (goto-char here)
0ec1d2c5
AM
2762 (if (c-beginning-of-macro)
2763 (point)
d23b0804 2764 here)))
0ec1d2c5
AM
2765 pa+1 ; pos just after an opening PAren (or brace).
2766 (ren+1 from) ; usually a pos just after an closing paREN etc.
2767 ; Is actually the pos. to scan for a (/{/[ from,
2768 ; which sometimes is after a silly )/}/].
2769 paren+1 ; Pos after some opening or closing paren.
2770 paren+1s ; A list of `paren+1's; used to determine a
2771 ; good-pos.
2772 bra+1 ce+1 ; just after L/R bra-ces.
2773 bra+1s ; list of OLD values of bra+1.
2774 mstart) ; start of a macro.
2775
2776 (save-excursion
d23b0804
AM
2777 (save-restriction
2778 (narrow-to-region (point-min) here)
2779 ;; Each time round the following loop, we enter a successively deeper
2780 ;; level of brace/paren nesting. (Except sometimes we "continue at
2781 ;; the existing level".) `pa+1' is a pos inside an opening
2782 ;; brace/paren/bracket, usually just after it.
2783 (while
2784 (progn
2785 ;; Each time round the next loop moves forward over an opening then
2786 ;; a closing brace/bracket/paren. This loop is white hot, so it
2787 ;; plays ugly tricks to go fast. DON'T PUT ANYTHING INTO THIS
2788 ;; LOOP WHICH ISN'T ABSOLUTELY NECESSARY!!! It terminates when a
2789 ;; call of `scan-lists' signals an error, which happens when there
2790 ;; are no more b/b/p's to scan.
2791 (c-safe
2792 (while t
2793 (setq pa+1 (scan-lists ren+1 1 -1) ; Into (/{/[; might signal
2794 paren+1s (cons pa+1 paren+1s))
2795 (setq ren+1 (scan-lists pa+1 1 1)) ; Out of )/}/]; might signal
2796 (if (and (eq (char-before pa+1) ?{)) ; Check for a macro later.
2797 (setq bra+1 pa+1))
2798 (setcar paren+1s ren+1)))
2799
2800 (if (and pa+1 (> pa+1 ren+1))
2801 ;; We've just entered a deeper nesting level.
2802 (progn
2803 ;; Insert the brace pair (if present) and the single open
2804 ;; paren/brace/bracket into `c-state-cache' It cannot be
2805 ;; inside a macro, except one around point, because of what
2806 ;; `c-neutralize-syntax-in-CPP' has done.
2807 (c-state-push-any-brace-pair bra+1 macro-start-or-here)
2808 ;; Insert the opening brace/bracket/paren position.
2809 (setq c-state-cache (cons (1- pa+1) c-state-cache))
2810 ;; Clear admin stuff for the next more nested part of the scan.
2811 (setq ren+1 pa+1 pa+1 nil bra+1 nil bra+1s nil)
2812 t) ; Carry on the loop
2813
2814 ;; All open p/b/b's at this nesting level, if any, have probably
2815 ;; been closed by matching/mismatching ones. We're probably
2816 ;; finished - we just need to check for having found an
2817 ;; unmatched )/}/], which we ignore. Such a )/}/] can't be in a
2818 ;; macro, due the action of `c-neutralize-syntax-in-CPP'.
2819 (c-safe (setq ren+1 (scan-lists ren+1 1 1)))))) ; acts as loop control.
2820
2821 ;; Record the final, innermost, brace-pair if there is one.
2822 (c-state-push-any-brace-pair bra+1 macro-start-or-here)
2823
2824 ;; Determine a good pos
2825 (while (and (setq paren+1 (car paren+1s))
2826 (> (if (> paren+1 macro-start-or-here)
2827 paren+1
2828 (goto-char paren+1)
2829 (setq mstart (and (c-beginning-of-macro)
2830 (point)))
2831 (or mstart paren+1))
2832 here-bol))
2833 (setq paren+1s (cdr paren+1s)))
2834 (cond
2835 ((and paren+1 mstart)
2836 (min paren+1 mstart))
2837 (paren+1)
2838 (t from))))))
2839
2840(defun c-remove-stale-state-cache (start-point here pps-point)
0ec1d2c5 2841 ;; Remove stale entries from the `c-cache-state', i.e. those which will
d23b0804
AM
2842 ;; not be in it when it is amended for position HERE. Additionally, the
2843 ;; "outermost" open-brace entry before HERE will be converted to a cons if
2844 ;; the matching close-brace is scanned.
0ec1d2c5 2845 ;;
267eacb5 2846 ;; START-POINT is a "maximal" "safe position" - there must be no open
d23b0804 2847 ;; parens/braces/brackets between START-POINT and HERE.
0ec1d2c5
AM
2848 ;;
2849 ;; As a second thing, calculate the result of parse-partial-sexp at
267eacb5 2850 ;; PPS-POINT, w.r.t. START-POINT. The motivation here is that
0ec1d2c5 2851 ;; `c-state-cache-good-pos' may become PPS-POINT, but the caller may need to
267eacb5 2852 ;; adjust it to get outside a string/comment. (Sorry about this! The code
0ec1d2c5
AM
2853 ;; needs to be FAST).
2854 ;;
2855 ;; Return a list (GOOD-POS SCAN-BACK-POS PPS-STATE), where
2856 ;; o - GOOD-POS is a position where the new value `c-state-cache' is known
2857 ;; to be good (we aim for this to be as high as possible);
2858 ;; o - SCAN-BACK-POS, if not nil, indicates there may be a brace pair
2859 ;; preceding POS which needs to be recorded in `c-state-cache'. It is a
267eacb5
AM
2860 ;; position to scan backwards from. It is the position of the "{" of the
2861 ;; last element to be removed from `c-state-cache', when that elt is a
2862 ;; cons, otherwise nil.
0ec1d2c5 2863 ;; o - PPS-STATE is the parse-partial-sexp state at PPS-POINT.
d23b0804
AM
2864 (save-excursion
2865 (save-restriction
2866 (narrow-to-region 1 (point-max))
2867 (let* ((in-macro-start ; start of macro containing HERE or nil.
0ec1d2c5 2868 (save-excursion
d23b0804 2869 (goto-char here)
0ec1d2c5
AM
2870 (and (c-beginning-of-macro)
2871 (point))))
267eacb5
AM
2872 (start-point-actual-macro-start ; Start of macro containing
2873 ; start-point or nil
d23b0804 2874 (and (< start-point here)
0ec1d2c5 2875 (save-excursion
267eacb5 2876 (goto-char start-point)
0ec1d2c5
AM
2877 (and (c-beginning-of-macro)
2878 (point)))))
267eacb5 2879 (start-point-actual-macro-end ; End of this macro, (maybe
d23b0804 2880 ; HERE), or nil.
267eacb5 2881 (and start-point-actual-macro-start
0ec1d2c5 2882 (save-excursion
267eacb5 2883 (goto-char start-point-actual-macro-start)
0ec1d2c5
AM
2884 (c-end-of-macro)
2885 (point))))
267eacb5 2886 pps-state ; Will be 9 or 10 elements long.
0ec1d2c5
AM
2887 pos
2888 upper-lim ; ,beyond which `c-state-cache' entries are removed
2889 scan-back-pos
2890 pair-beg pps-point-state target-depth)
2891
d23b0804
AM
2892 ;; Remove entries beyond HERE. Also remove any entries inside
2893 ;; a macro, unless HERE is in the same macro.
0ec1d2c5
AM
2894 (setq upper-lim
2895 (if (or (null c-state-old-cpp-beg)
d23b0804
AM
2896 (and (> here c-state-old-cpp-beg)
2897 (< here c-state-old-cpp-end)))
2898 here
2899 (min here c-state-old-cpp-beg)))
657071fc 2900 (while (and c-state-cache (>= (c-state-cache-top-lparen) upper-lim))
8cff5feb 2901 (setq scan-back-pos (car-safe (car c-state-cache)))
0ec1d2c5 2902 (setq c-state-cache (cdr c-state-cache)))
8cff5feb 2903
0ec1d2c5
AM
2904 ;; If `upper-lim' is inside the last recorded brace pair, remove its
2905 ;; RBrace and indicate we'll need to search backwards for a previous
2906 ;; brace pair.
2907 (when (and c-state-cache
2908 (consp (car c-state-cache))
2909 (> (cdar c-state-cache) upper-lim))
2910 (setcar c-state-cache (caar c-state-cache))
2911 (setq scan-back-pos (car c-state-cache)))
2912
2913 ;; The next loop jumps forward out of a nested level of parens each
2914 ;; time round; the corresponding elements in `c-state-cache' are
2915 ;; removed. `pos' is just after the brace-pair or the open paren at
267eacb5 2916 ;; (car c-state-cache). There can be no open parens/braces/brackets
d23b0804 2917 ;; between `start-point'/`start-point-actual-macro-start' and HERE,
0ec1d2c5 2918 ;; due to the interface spec to this function.
267eacb5
AM
2919 (setq pos (if (and start-point-actual-macro-end
2920 (not (eq start-point-actual-macro-start
eb2f6eeb 2921 in-macro-start)))
267eacb5 2922 (1+ start-point-actual-macro-end) ; get outside the macro as
0ec1d2c5 2923 ; marked by a `category' text property.
267eacb5 2924 start-point))
0ec1d2c5
AM
2925 (goto-char pos)
2926 (while (and c-state-cache
c38e0c97 2927 (or (numberp (car c-state-cache)) ; Have we a { at all?
d23b0804
AM
2928 (cdr c-state-cache))
2929 (< (point) here))
0ec1d2c5
AM
2930 (cond
2931 ((null pps-state) ; first time through
2932 (setq target-depth -1))
2933 ((eq (car pps-state) target-depth) ; found closing ),},]
2934 (setq target-depth (1- (car pps-state))))
2935 ;; Do nothing when we've merely reached pps-point.
2936 )
2937
2938 ;; Scan!
2939 (setq pps-state
2940 (parse-partial-sexp
d23b0804 2941 (point) (if (< (point) pps-point) pps-point here)
0ec1d2c5
AM
2942 target-depth
2943 nil pps-state))
2944
2945 (if (= (point) pps-point)
2946 (setq pps-point-state pps-state))
2947
2948 (when (eq (car pps-state) target-depth)
2949 (setq pos (point)) ; POS is now just after an R-paren/brace.
2950 (cond
2951 ((and (consp (car c-state-cache))
2952 (eq (point) (cdar c-state-cache)))
2953 ;; We've just moved out of the paren pair containing the brace-pair
2954 ;; at (car c-state-cache). `pair-beg' is where the open paren is,
2955 ;; and is potentially where the open brace of a cons in
2956 ;; c-state-cache will be.
2957 (setq pair-beg (car-safe (cdr c-state-cache))
2958 c-state-cache (cdr-safe (cdr c-state-cache)))) ; remove {}pair + containing Lparen.
2959 ((numberp (car c-state-cache))
2960 (setq pair-beg (car c-state-cache)
2961 c-state-cache (cdr c-state-cache))) ; remove this
2962 ; containing Lparen
2963 ((numberp (cadr c-state-cache))
2964 (setq pair-beg (cadr c-state-cache)
2965 c-state-cache (cddr c-state-cache))) ; Remove a paren pair
2966 ; together with enclosed brace pair.
2967 ;; (t nil) ; Ignore an unmated Rparen.
2968 )))
2969
2970 (if (< (point) pps-point)
2971 (setq pps-state (parse-partial-sexp (point) pps-point
2972 nil nil ; TARGETDEPTH, STOPBEFORE
2973 pps-state)))
2974
2975 ;; If the last paren pair we moved out of was actually a brace pair,
2976 ;; insert it into `c-state-cache'.
2977 (when (and pair-beg (eq (char-after pair-beg) ?{))
2978 (if (consp (car-safe c-state-cache))
2979 (setq c-state-cache (cdr c-state-cache)))
2980 (setq c-state-cache (cons (cons pair-beg pos)
2981 c-state-cache)))
2982
2983 (list pos scan-back-pos pps-state)))))
2984
267eacb5 2985(defun c-remove-stale-state-cache-backwards (here)
0ec1d2c5
AM
2986 ;; Strip stale elements of `c-state-cache' by moving backwards through the
2987 ;; buffer, and inform the caller of the scenario detected.
2988 ;;
2989 ;; HERE is the position we're setting `c-state-cache' for.
267eacb5
AM
2990 ;; CACHE-POS (a locally bound variable) is just after the latest recorded
2991 ;; position in `c-state-cache' before HERE, or a position at or near
2992 ;; point-min which isn't in a literal.
0ec1d2c5
AM
2993 ;;
2994 ;; This function must only be called only when (> `c-state-cache-good-pos'
2995 ;; HERE). Usually the gap between CACHE-POS and HERE is large. It is thus
e1dbe924 2996 ;; optimized to eliminate (or minimize) scanning between these two
0ec1d2c5
AM
2997 ;; positions.
2998 ;;
2999 ;; Return a three element list (GOOD-POS SCAN-BACK-POS FWD-FLAG), where:
3000 ;; o - GOOD-POS is a "good position", where `c-state-cache' is valid, or
3001 ;; could become so after missing elements are inserted into
3002 ;; `c-state-cache'. This is JUST AFTER an opening or closing
3003 ;; brace/paren/bracket which is already in `c-state-cache' or just before
3004 ;; one otherwise. exceptionally (when there's no such b/p/b handy) the BOL
3005 ;; before `here''s line, or the start of the literal containing it.
3006 ;; o - SCAN-BACK-POS, if non-nil, indicates there may be a brace pair
3007 ;; preceding POS which isn't recorded in `c-state-cache'. It is a position
3008 ;; to scan backwards from.
3009 ;; o - FWD-FLAG, if non-nil, indicates there may be parens/braces between
3010 ;; POS and HERE which aren't recorded in `c-state-cache'.
3011 ;;
3012 ;; The comments in this defun use "paren" to mean parenthesis or square
3013 ;; bracket (as contrasted with a brace), and "(" and ")" likewise.
3014 ;;
267eacb5
AM
3015 ;; . {..} (..) (..) ( .. { } ) (...) ( .... . ..)
3016 ;; | | | | | |
3017 ;; CP E here D C good
3018 (let ((cache-pos (c-get-cache-scan-pos here)) ; highest position below HERE in cache (or 1)
3019 (pos c-state-cache-good-pos)
0ec1d2c5
AM
3020 pa ren ; positions of "(" and ")"
3021 dropped-cons ; whether the last element dropped from `c-state-cache'
3022 ; was a cons (representing a brace-pair)
3023 good-pos ; see above.
3024 lit ; (START . END) of a literal containing some point.
3025 here-lit-start here-lit-end ; bounds of literal containing `here'
3026 ; or `here' itself.
c0209c2c 3027 here- here+ ; start/end of macro around HERE, or HERE
0ec1d2c5 3028 (here-bol (c-point 'bol here))
c400c4d7 3029 (too-far-back (max (- here c-state-cache-too-far) (point-min))))
0ec1d2c5
AM
3030
3031 ;; Remove completely irrelevant entries from `c-state-cache'.
3032 (while (and c-state-cache
3033 (>= (setq pa (c-state-cache-top-lparen)) here))
3034 (setq dropped-cons (consp (car c-state-cache)))
3035 (setq c-state-cache (cdr c-state-cache))
3036 (setq pos pa))
3037 ;; At this stage, (> pos here);
3038 ;; (< (c-state-cache-top-lparen) here) (or is nil).
3039
c0209c2c
AM
3040 (cond
3041 ((and (consp (car c-state-cache))
3042 (> (cdar c-state-cache) here))
3043 ;; CASE 1: The top of the cache is a brace pair which now encloses
3044 ;; `here'. As good-pos, return the address. of the "{". Since we've no
3045 ;; knowledge of what's inside these braces, we have no alternative but
3046 ;; to direct the caller to scan the buffer from the opening brace.
3047 (setq pos (caar c-state-cache))
3048 (setcar c-state-cache pos)
3049 (list (1+ pos) pos t)) ; return value. We've just converted a brace pair
3050 ; entry into a { entry, so the caller needs to
3051 ; search for a brace pair before the {.
3052
3053 ;; `here' might be inside a literal. Check for this.
3054 ((progn
3055 (setq lit (c-state-literal-at here)
3056 here-lit-start (or (car lit) here)
3057 here-lit-end (or (cdr lit) here))
3058 ;; Has `here' just "newly entered" a macro?
3059 (save-excursion
3060 (goto-char here-lit-start)
3061 (if (and (c-beginning-of-macro)
3062 (or (null c-state-old-cpp-beg)
3063 (not (= (point) c-state-old-cpp-beg))))
3064 (progn
3065 (setq here- (point))
3066 (c-end-of-macro)
3067 (setq here+ (point)))
3068 (setq here- here-lit-start
3069 here+ here-lit-end)))
3070
3071 ;; `here' might be nested inside any depth of parens (or brackets but
3072 ;; not braces). Scan backwards to find the outermost such opening
3073 ;; paren, if there is one. This will be the scan position to return.
3074 (save-restriction
3075 (narrow-to-region cache-pos (point-max))
3076 (setq pos (c-state-balance-parens-backwards here- here+ pos)))
3077 nil)) ; for the cond
3078
3079 ((< pos here-lit-start)
3080 ;; CASE 2: Address of outermost ( or [ which now encloses `here', but
3081 ;; didn't enclose the (previous) `c-state-cache-good-pos'. If there is
3082 ;; a brace pair preceding this, it will already be in `c-state-cache',
3083 ;; unless there was a brace pair after it, i.e. there'll only be one to
3084 ;; scan for if we've just deleted one.
3085 (list pos (and dropped-cons pos) t)) ; Return value.
3086
3087 ;; `here' isn't enclosed in a (previously unrecorded) bracket/paren.
3088 ;; Further forward scanning isn't needed, but we still need to find a
3089 ;; GOOD-POS. Step out of all enclosing "("s on HERE's line.
3090 ((progn
0ec1d2c5
AM
3091 (save-restriction
3092 (narrow-to-region here-bol (point-max))
3093 (setq pos here-lit-start)
3094 (c-safe (while (setq pa (scan-lists pos -1 1))
3095 (setq pos pa)))) ; might signal
c0209c2c
AM
3096 nil)) ; for the cond
3097
3098 ((setq ren (c-safe-scan-lists pos -1 -1 too-far-back))
3099 ;; CASE 3: After a }/)/] before `here''s BOL.
3100 (list (1+ ren) (and dropped-cons pos) nil)) ; Return value
3101
3102 (t
3103 ;; CASE 4; Best of a bad job: BOL before `here-bol', or beginning of
3104 ;; literal containing it.
3105 (setq good-pos (c-state-lit-beg (c-point 'bopl here-bol)))
3106 (list good-pos (and dropped-cons good-pos) nil)))))
0ec1d2c5
AM
3107
3108
3109;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
3110;; Externally visible routines.
3111
3112(defun c-state-cache-init ()
3113 (setq c-state-cache nil
3114 c-state-cache-good-pos 1
3115 c-state-nonlit-pos-cache nil
3116 c-state-nonlit-pos-cache-limit 1
8ed43f15
AM
3117 c-state-semi-nonlit-pos-cache nil
3118 c-state-semi-nonlit-pos-cache-limit 1
0ec1d2c5
AM
3119 c-state-brace-pair-desert nil
3120 c-state-point-min 1
3121 c-state-point-min-lit-type nil
3122 c-state-point-min-lit-start nil
3123 c-state-min-scan-pos 1
3124 c-state-old-cpp-beg nil
3125 c-state-old-cpp-end nil)
3126 (c-state-mark-point-min-literal))
3127
e84efb70
AM
3128;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
3129;; Debugging routines to dump `c-state-cache' in a "replayable" form.
3130;; (defmacro c-sc-de (elt) ; "c-state-cache-dump-element"
3131;; `(format ,(concat "(setq " (symbol-name elt) " %s) ") ,elt))
3132;; (defmacro c-sc-qde (elt) ; "c-state-cache-quote-dump-element"
3133;; `(format ,(concat "(setq " (symbol-name elt) " '%s) ") ,elt))
3134;; (defun c-state-dump ()
3135;; ;; For debugging.
3136;; ;(message
3137;; (concat
3138;; (c-sc-qde c-state-cache)
3139;; (c-sc-de c-state-cache-good-pos)
3140;; (c-sc-qde c-state-nonlit-pos-cache)
3141;; (c-sc-de c-state-nonlit-pos-cache-limit)
3142;; (c-sc-qde c-state-brace-pair-desert)
3143;; (c-sc-de c-state-point-min)
3144;; (c-sc-de c-state-point-min-lit-type)
3145;; (c-sc-de c-state-point-min-lit-start)
3146;; (c-sc-de c-state-min-scan-pos)
3147;; (c-sc-de c-state-old-cpp-beg)
3148;; (c-sc-de c-state-old-cpp-end)))
3149;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
3150
0ec1d2c5
AM
3151(defun c-invalidate-state-cache-1 (here)
3152 ;; Invalidate all info on `c-state-cache' that applies to the buffer at HERE
3153 ;; or higher and set `c-state-cache-good-pos' accordingly. The cache is
3154 ;; left in a consistent state.
3155 ;;
3156 ;; This is much like `c-whack-state-after', but it never changes a paren
3157 ;; pair element into an open paren element. Doing that would mean that the
3158 ;; new open paren wouldn't have the required preceding paren pair element.
785eecbb 3159 ;;
0ec1d2c5
AM
3160 ;; This function is called from c-after-change.
3161
56d093a9 3162 ;; The caches of non-literals:
691e26ae
AM
3163 ;; Note that we use "<=" for the possibility of the second char of a two-char
3164 ;; comment opener being typed; this would invalidate any cache position at
3165 ;; HERE.
3166 (if (<= here c-state-nonlit-pos-cache-limit)
3167 (setq c-state-nonlit-pos-cache-limit (1- here)))
3168 (if (<= here c-state-semi-nonlit-pos-cache-limit)
3169 (setq c-state-semi-nonlit-pos-cache-limit (1- here)))
0ec1d2c5
AM
3170
3171 ;; `c-state-cache':
3172 ;; Case 1: if `here' is in a literal containing point-min, everything
3173 ;; becomes (or is already) nil.
3174 (if (or (null c-state-cache-good-pos)
3175 (< here (c-state-get-min-scan-pos)))
3176 (setq c-state-cache nil
3177 c-state-cache-good-pos nil
3178 c-state-min-scan-pos nil)
3179
c400c4d7
SM
3180 ;; Truncate `c-state-cache' and set `c-state-cache-good-pos' to a value
3181 ;; below `here'. To maintain its consistency, we may need to insert a new
3182 ;; brace pair.
0ec1d2c5
AM
3183 (let ((here-bol (c-point 'bol here))
3184 too-high-pa ; recorded {/(/[ next above here, or nil.
3185 dropped-cons ; was the last removed element a brace pair?
3186 pa)
3187 ;; The easy bit - knock over-the-top bits off `c-state-cache'.
3188 (while (and c-state-cache
3189 (>= (setq pa (c-state-cache-top-paren)) here))
3190 (setq dropped-cons (consp (car c-state-cache))
3191 too-high-pa (c-state-cache-top-lparen)
3192 c-state-cache (cdr c-state-cache)))
3193
3194 ;; Do we need to add in an earlier brace pair, having lopped one off?
3195 (if (and dropped-cons
3196 (< too-high-pa (+ here c-state-cache-too-far)))
d23b0804 3197 (c-append-lower-brace-pair-to-state-cache too-high-pa here here-bol))
0ec1d2c5
AM
3198 (setq c-state-cache-good-pos (or (c-state-cache-after-top-paren)
3199 (c-state-get-min-scan-pos)))))
3200
3201 ;; The brace-pair desert marker:
3202 (when (car c-state-brace-pair-desert)
3203 (if (< here (car c-state-brace-pair-desert))
3204 (setq c-state-brace-pair-desert nil)
3205 (if (< here (cdr c-state-brace-pair-desert))
3206 (setcdr c-state-brace-pair-desert here)))))
3207
3208(defun c-parse-state-1 ()
3209 ;; Find and record all noteworthy parens between some good point earlier in
3210 ;; the file and point. That good point is at least the beginning of the
3211 ;; top-level construct we are in, or the beginning of the preceding
3212 ;; top-level construct if we aren't in one.
3213 ;;
3214 ;; The returned value is a list of the noteworthy parens with the last one
3215 ;; first. If an element in the list is an integer, it's the position of an
3216 ;; open paren (of any type) which has not been closed before the point. If
3217 ;; an element is a cons, it gives the position of a closed BRACE paren
3218 ;; pair[*]; the car is the start brace position and the cdr is the position
3219 ;; following the closing brace. Only the last closed brace paren pair
3220 ;; before each open paren and before the point is recorded, and thus the
3221 ;; state never contains two cons elements in succession. When a close brace
3222 ;; has no matching open brace (e.g., the matching brace is outside the
3223 ;; visible region), it is not represented in the returned value.
3224 ;;
3225 ;; [*] N.B. The close "brace" might be a mismatching close bracket or paren.
3226 ;; This defun explicitly treats mismatching parens/braces/brackets as
3227 ;; matching. It is the open brace which makes it a "brace" pair.
3228 ;;
3229 ;; If POINT is within a macro, open parens and brace pairs within
3230 ;; THIS macro MIGHT be recorded. This depends on whether their
3231 ;; syntactic properties have been suppressed by
3232 ;; `c-neutralize-syntax-in-CPP'. This might need fixing (2008-12-11).
d9e94c22
MS
3233 ;;
3234 ;; Currently no characters which are given paren syntax with the
3235 ;; syntax-table property are recorded, i.e. angle bracket arglist
3236 ;; parens are never present here. Note that this might change.
3237 ;;
0386b551 3238 ;; BUG: This function doesn't cope entirely well with unbalanced
0ec1d2c5
AM
3239 ;; parens in macros. (2008-12-11: this has probably been resolved
3240 ;; by the function `c-neutralize-syntax-in-CPP'.) E.g. in the
3241 ;; following case the brace before the macro isn't balanced with the
3242 ;; one after it:
0386b551
AM
3243 ;;
3244 ;; {
3245 ;; #define X {
3246 ;; }
3247 ;;
0ec1d2c5
AM
3248 ;; Note to maintainers: this function DOES get called with point
3249 ;; within comments and strings, so don't assume it doesn't!
3250 ;;
0386b551 3251 ;; This function might do hidden buffer changes.
0ec1d2c5
AM
3252 (let* ((here (point))
3253 (here-bopl (c-point 'bopl))
3254 strategy ; 'forward, 'backward etc..
3255 ;; Candidate positions to start scanning from:
3256 cache-pos ; highest position below HERE already existing in
3257 ; cache (or 1).
3258 good-pos
267eacb5
AM
3259 start-point ; (when scanning forward) a place below HERE where there
3260 ; are no open parens/braces between it and HERE.
0ec1d2c5
AM
3261 bopl-state
3262 res
3263 scan-backward-pos scan-forward-p) ; used for 'backward.
3264 ;; If POINT-MIN has changed, adjust the cache
3265 (unless (= (point-min) c-state-point-min)
3266 (c-renarrow-state-cache))
3267
3268 ;; Strategy?
3269 (setq res (c-parse-state-get-strategy here c-state-cache-good-pos)
3270 strategy (car res)
267eacb5 3271 start-point (cadr res))
0ec1d2c5 3272
0ec1d2c5 3273 ;; SCAN!
d23b0804
AM
3274 (cond
3275 ((eq strategy 'forward)
3276 (setq res (c-remove-stale-state-cache start-point here here-bopl))
3277 (setq cache-pos (car res)
3278 scan-backward-pos (cadr res)
3279 bopl-state (car (cddr res))) ; will be nil if (< here-bopl
0ec1d2c5 3280 ; start-point)
d23b0804
AM
3281 (if scan-backward-pos
3282 (c-append-lower-brace-pair-to-state-cache scan-backward-pos here))
3283 (setq good-pos
3284 (c-append-to-state-cache cache-pos here))
3285 (setq c-state-cache-good-pos
3286 (if (and bopl-state
3287 (< good-pos (- here c-state-cache-too-far)))
3288 (c-state-cache-non-literal-place here-bopl bopl-state)
3289 good-pos)))
3290
3291 ((eq strategy 'backward)
3292 (setq res (c-remove-stale-state-cache-backwards here)
3293 good-pos (car res)
3294 scan-backward-pos (cadr res)
3295 scan-forward-p (car (cddr res)))
3296 (if scan-backward-pos
3297 (c-append-lower-brace-pair-to-state-cache scan-backward-pos here))
3298 (setq c-state-cache-good-pos
3299 (if scan-forward-p
3300 (c-append-to-state-cache good-pos here)
3301 good-pos)))
3302
3303 (t ; (eq strategy 'IN-LIT)
3304 (setq c-state-cache nil
3305 c-state-cache-good-pos nil))))
0ec1d2c5
AM
3306
3307 c-state-cache)
3308
3309(defun c-invalidate-state-cache (here)
3310 ;; This is a wrapper over `c-invalidate-state-cache-1'.
3311 ;;
3312 ;; It suppresses the syntactic effect of the < and > (template) brackets and
3313 ;; of all parens in preprocessor constructs, except for any such construct
3314 ;; containing point. We can then call `c-invalidate-state-cache-1' without
3315 ;; worrying further about macros and template delimiters.
3316 (c-with-<->-as-parens-suppressed
3e8f7d91
AM
3317 (if (and c-state-old-cpp-beg
3318 (< c-state-old-cpp-beg here))
0ec1d2c5 3319 (c-with-all-but-one-cpps-commented-out
3e8f7d91
AM
3320 c-state-old-cpp-beg
3321 (min c-state-old-cpp-end here)
0ec1d2c5
AM
3322 (c-invalidate-state-cache-1 here))
3323 (c-with-cpps-commented-out
3324 (c-invalidate-state-cache-1 here)))))
3325
8cff5feb
AM
3326(defmacro c-state-maybe-marker (place marker)
3327 ;; If PLACE is non-nil, return a marker marking it, otherwise nil.
3328 ;; We (re)use MARKER.
3329 `(and ,place
3330 (or ,marker (setq ,marker (make-marker)))
3331 (set-marker ,marker ,place)))
3332
0ec1d2c5
AM
3333(defun c-parse-state ()
3334 ;; This is a wrapper over `c-parse-state-1'. See that function for a
3335 ;; description of the functionality and return value.
3336 ;;
3337 ;; It suppresses the syntactic effect of the < and > (template) brackets and
3338 ;; of all parens in preprocessor constructs, except for any such construct
3339 ;; containing point. We can then call `c-parse-state-1' without worrying
3340 ;; further about macros and template delimiters.
3341 (let (here-cpp-beg here-cpp-end)
3342 (save-excursion
3343 (when (c-beginning-of-macro)
3344 (setq here-cpp-beg (point))
3345 (unless
3346 (> (setq here-cpp-end (c-syntactic-end-of-macro))
3347 here-cpp-beg)
3348 (setq here-cpp-beg nil here-cpp-end nil))))
3349 ;; FIXME!!! Put in a `condition-case' here to protect the integrity of the
3350 ;; subsystem.
3351 (prog1
3352 (c-with-<->-as-parens-suppressed
3353 (if (and here-cpp-beg (> here-cpp-end here-cpp-beg))
3354 (c-with-all-but-one-cpps-commented-out
3355 here-cpp-beg here-cpp-end
3356 (c-parse-state-1))
3357 (c-with-cpps-commented-out
3358 (c-parse-state-1))))
8cff5feb
AM
3359 (setq c-state-old-cpp-beg
3360 (c-state-maybe-marker here-cpp-beg c-state-old-cpp-beg-marker)
3361 c-state-old-cpp-end
3362 (c-state-maybe-marker here-cpp-end c-state-old-cpp-end-marker)))))
0ec1d2c5
AM
3363
3364;; Debug tool to catch cache inconsistencies. This is called from
3365;; 000tests.el.
a66cd3ee
MS
3366(defvar c-debug-parse-state nil)
3367(unless (fboundp 'c-real-parse-state)
3368 (fset 'c-real-parse-state (symbol-function 'c-parse-state)))
3369(cc-bytecomp-defun c-real-parse-state)
cd5495ff 3370
b03aabda 3371(defvar c-parse-state-point nil)
cd5495ff
AM
3372(defvar c-parse-state-state nil)
3373(defun c-record-parse-state-state ()
b03aabda 3374 (setq c-parse-state-point (point))
cd5495ff
AM
3375 (setq c-parse-state-state
3376 (mapcar
3377 (lambda (arg)
b03aabda
AM
3378 (let ((val (symbol-value arg)))
3379 (cons arg
3380 (if (consp val)
3381 (copy-tree val)
3382 val))))
cd5495ff
AM
3383 '(c-state-cache
3384 c-state-cache-good-pos
3385 c-state-nonlit-pos-cache
3386 c-state-nonlit-pos-cache-limit
8ed43f15
AM
3387 c-state-semi-nonlit-pos-cache
3388 c-state-semi-nonlit-pos-cache-limit
cd5495ff
AM
3389 c-state-brace-pair-desert
3390 c-state-point-min
3391 c-state-point-min-lit-type
3392 c-state-point-min-lit-start
3393 c-state-min-scan-pos
3394 c-state-old-cpp-beg
b03aabda
AM
3395 c-state-old-cpp-end
3396 c-parse-state-point))))
cd5495ff
AM
3397(defun c-replay-parse-state-state ()
3398 (message
3399 (concat "(setq "
3400 (mapconcat
3401 (lambda (arg)
3402 (format "%s %s%s" (car arg) (if (atom (cdr arg)) "" "'") (cdr arg)))
3403 c-parse-state-state " ")
3404 ")")))
3405
b03aabda
AM
3406(defun c-debug-parse-state-double-cons (state)
3407 (let (state-car conses-not-ok)
3408 (while state
3409 (setq state-car (car state)
3410 state (cdr state))
3411 (if (and (consp state-car)
3412 (consp (car state)))
3413 (setq conses-not-ok t)))
3414 conses-not-ok))
3415
a66cd3ee 3416(defun c-debug-parse-state ()
0ec1d2c5 3417 (let ((here (point)) (res1 (c-real-parse-state)) res2)
0386b551 3418 (let ((c-state-cache nil)
0ec1d2c5
AM
3419 (c-state-cache-good-pos 1)
3420 (c-state-nonlit-pos-cache nil)
3421 (c-state-nonlit-pos-cache-limit 1)
3422 (c-state-brace-pair-desert nil)
3423 (c-state-point-min 1)
3424 (c-state-point-min-lit-type nil)
3425 (c-state-point-min-lit-start nil)
3426 (c-state-min-scan-pos 1)
3427 (c-state-old-cpp-beg nil)
3428 (c-state-old-cpp-end nil))
a66cd3ee
MS
3429 (setq res2 (c-real-parse-state)))
3430 (unless (equal res1 res2)
0386b551
AM
3431 ;; The cache can actually go further back due to the ad-hoc way
3432 ;; the first paren is found, so try to whack off a bit of its
3433 ;; start before complaining.
cd5495ff
AM
3434 ;; (save-excursion
3435 ;; (goto-char (or (c-least-enclosing-brace res2) (point)))
3436 ;; (c-beginning-of-defun-1)
3437 ;; (while (not (or (bobp) (eq (char-after) ?{)))
3438 ;; (c-beginning-of-defun-1))
3439 ;; (unless (equal (c-whack-state-before (point) res1) res2)
3440 ;; (message (concat "c-parse-state inconsistency at %s: "
3441 ;; "using cache: %s, from scratch: %s")
3442 ;; here res1 res2)))
3443 (message (concat "c-parse-state inconsistency at %s: "
3444 "using cache: %s, from scratch: %s")
3445 here res1 res2)
3446 (message "Old state:")
3447 (c-replay-parse-state-state))
b03aabda
AM
3448
3449 (when (c-debug-parse-state-double-cons res1)
3450 (message "c-parse-state INVALIDITY at %s: %s"
3451 here res1)
3452 (message "Old state:")
3453 (c-replay-parse-state-state))
3454
cd5495ff 3455 (c-record-parse-state-state)
b03aabda
AM
3456 res2 ; res1 correct a cascading series of errors ASAP
3457 ))
0ec1d2c5 3458
a66cd3ee
MS
3459(defun c-toggle-parse-state-debug (&optional arg)
3460 (interactive "P")
3461 (setq c-debug-parse-state (c-calculate-state arg c-debug-parse-state))
3462 (fset 'c-parse-state (symbol-function (if c-debug-parse-state
3463 'c-debug-parse-state
3464 'c-real-parse-state)))
b03aabda
AM
3465 (c-keep-region-active)
3466 (message "c-debug-parse-state %sabled"
3467 (if c-debug-parse-state "en" "dis")))
0386b551
AM
3468(when c-debug-parse-state
3469 (c-toggle-parse-state-debug 1))
a66cd3ee 3470
0ec1d2c5 3471\f
d9e94c22
MS
3472(defun c-whack-state-before (bufpos paren-state)
3473 ;; Whack off any state information from PAREN-STATE which lies
3474 ;; before BUFPOS. Not destructive on PAREN-STATE.
d9e94c22
MS
3475 (let* ((newstate (list nil))
3476 (ptr newstate)
3477 car)
3478 (while paren-state
3479 (setq car (car paren-state)
3480 paren-state (cdr paren-state))
3481 (if (< (if (consp car) (car car) car) bufpos)
3482 (setq paren-state nil)
3483 (setcdr ptr (list car))
3484 (setq ptr (cdr ptr))))
3485 (cdr newstate)))
3486
3487(defun c-whack-state-after (bufpos paren-state)
3488 ;; Whack off any state information from PAREN-STATE which lies at or
3489 ;; after BUFPOS. Not destructive on PAREN-STATE.
d9e94c22
MS
3490 (catch 'done
3491 (while paren-state
3492 (let ((car (car paren-state)))
3493 (if (consp car)
3494 ;; just check the car, because in a balanced brace
3495 ;; expression, it must be impossible for the corresponding
3496 ;; close brace to be before point, but the open brace to
3497 ;; be after.
3498 (if (<= bufpos (car car))
3499 nil ; whack it off
3500 (if (< bufpos (cdr car))
3501 ;; its possible that the open brace is before
3502 ;; bufpos, but the close brace is after. In that
3503 ;; case, convert this to a non-cons element. The
3504 ;; rest of the state is before bufpos, so we're
3505 ;; done.
3506 (throw 'done (cons (car car) (cdr paren-state)))
3507 ;; we know that both the open and close braces are
3508 ;; before bufpos, so we also know that everything else
3509 ;; on state is before bufpos.
3510 (throw 'done paren-state)))
3511 (if (<= bufpos car)
3512 nil ; whack it off
3513 ;; it's before bufpos, so everything else should too.
3514 (throw 'done paren-state)))
3515 (setq paren-state (cdr paren-state)))
3516 nil)))
3517
3518(defun c-most-enclosing-brace (paren-state &optional bufpos)
3519 ;; Return the bufpos of the innermost enclosing open paren before
0386b551 3520 ;; bufpos, or nil if none was found.
d9e94c22
MS
3521 (let (enclosingp)
3522 (or bufpos (setq bufpos 134217727))
3523 (while paren-state
3524 (setq enclosingp (car paren-state)
3525 paren-state (cdr paren-state))
3526 (if (or (consp enclosingp)
3527 (>= enclosingp bufpos))
3528 (setq enclosingp nil)
d9e94c22
MS
3529 (setq paren-state nil)))
3530 enclosingp))
3531
0386b551
AM
3532(defun c-least-enclosing-brace (paren-state)
3533 ;; Return the bufpos of the outermost enclosing open paren, or nil
3534 ;; if none was found.
d9e94c22 3535 (let (pos elem)
d9e94c22
MS
3536 (while paren-state
3537 (setq elem (car paren-state)
3538 paren-state (cdr paren-state))
0386b551
AM
3539 (if (integerp elem)
3540 (setq pos elem)))
d9e94c22
MS
3541 pos))
3542
3543(defun c-safe-position (bufpos paren-state)
0386b551
AM
3544 ;; Return the closest "safe" position recorded on PAREN-STATE that
3545 ;; is higher up than BUFPOS. Return nil if PAREN-STATE doesn't
3546 ;; contain any. Return nil if BUFPOS is nil, which is useful to
3547 ;; find the closest limit before a given limit that might be nil.
d9e94c22 3548 ;;
0386b551
AM
3549 ;; A "safe" position is a position at or after a recorded open
3550 ;; paren, or after a recorded close paren. The returned position is
3551 ;; thus either the first position after a close brace, or the first
3552 ;; position after an enclosing paren, or at the enclosing paren in
3553 ;; case BUFPOS is immediately after it.
d9e94c22
MS
3554 (when bufpos
3555 (let (elem)
3556 (catch 'done
3557 (while paren-state
3558 (setq elem (car paren-state))
3559 (if (consp elem)
3560 (cond ((< (cdr elem) bufpos)
3561 (throw 'done (cdr elem)))
3562 ((< (car elem) bufpos)
3563 ;; See below.
3564 (throw 'done (min (1+ (car elem)) bufpos))))
3565 (if (< elem bufpos)
3566 ;; elem is the position at and not after the opening paren, so
3567 ;; we can go forward one more step unless it's equal to
3568 ;; bufpos. This is useful in some cases avoid an extra paren
3569 ;; level between the safe position and bufpos.
3570 (throw 'done (min (1+ elem) bufpos))))
3571 (setq paren-state (cdr paren-state)))))))
3572
3573(defun c-beginning-of-syntax ()
3574 ;; This is used for `font-lock-beginning-of-syntax-function'. It
3575 ;; goes to the closest previous point that is known to be outside
3576 ;; any string literal or comment. `c-state-cache' is used if it has
3577 ;; a position in the vicinity.
3578 (let* ((paren-state c-state-cache)
3579 elem
3580
3581 (pos (catch 'done
3582 ;; Note: Similar code in `c-safe-position'. The
3583 ;; difference is that we accept a safe position at
3584 ;; the point and don't bother to go forward past open
3585 ;; parens.
3586 (while paren-state
3587 (setq elem (car paren-state))
3588 (if (consp elem)
3589 (cond ((<= (cdr elem) (point))
3590 (throw 'done (cdr elem)))
3591 ((<= (car elem) (point))
3592 (throw 'done (car elem))))
3593 (if (<= elem (point))
3594 (throw 'done elem)))
3595 (setq paren-state (cdr paren-state)))
3596 (point-min))))
3597
3598 (if (> pos (- (point) 4000))
3599 (goto-char pos)
3600 ;; The position is far back. Try `c-beginning-of-defun-1'
3601 ;; (although we can't be entirely sure it will go to a position
3602 ;; outside a comment or string in current emacsen). FIXME:
3603 ;; Consult `syntax-ppss' here.
3604 (c-beginning-of-defun-1)
3605 (if (< (point) pos)
3606 (goto-char pos)))))
3607
3608\f
3609;; Tools for scanning identifiers and other tokens.
3610
3611(defun c-on-identifier ()
3612 "Return non-nil if the point is on or directly after an identifier.
3613Keywords are recognized and not considered identifiers. If an
3614identifier is detected, the returned value is its starting position.
0386b551
AM
3615If an identifier ends at the point and another begins at it \(can only
3616happen in Pike) then the point for the preceding one is returned.
d9e94c22 3617
0386b551
AM
3618Note that this function might do hidden buffer changes. See the
3619comment at the start of cc-engine.el for more info."
3620
3621 ;; FIXME: Shouldn't this function handle "operator" in C++?
d9e94c22
MS
3622
3623 (save-excursion
0386b551
AM
3624 (skip-syntax-backward "w_")
3625
3626 (or
3627
3628 ;; Check for a normal (non-keyword) identifier.
3629 (and (looking-at c-symbol-start)
3630 (not (looking-at c-keywords-regexp))
3631 (point))
3632
3633 (when (c-major-mode-is 'pike-mode)
3634 ;; Handle the `<operator> syntax in Pike.
3635 (let ((pos (point)))
3636 (skip-chars-backward "-!%&*+/<=>^|~[]()")
3637 (and (if (< (skip-chars-backward "`") 0)
3638 t
3639 (goto-char pos)
3640 (eq (char-after) ?\`))
3641 (looking-at c-symbol-key)
3642 (>= (match-end 0) pos)
3643 (point))))
3644
3645 ;; Handle the "operator +" syntax in C++.
3646 (when (and c-overloadable-operators-regexp
3647 (= (c-backward-token-2 0) 0))
3648
3649 (cond ((and (looking-at c-overloadable-operators-regexp)
51c9af45 3650 (or (not c-opt-op-identifier-prefix)
0386b551 3651 (and (= (c-backward-token-2 1) 0)
51c9af45 3652 (looking-at c-opt-op-identifier-prefix))))
0386b551
AM
3653 (point))
3654
3655 ((save-excursion
51c9af45
AM
3656 (and c-opt-op-identifier-prefix
3657 (looking-at c-opt-op-identifier-prefix)
0386b551
AM
3658 (= (c-forward-token-2 1) 0)
3659 (looking-at c-overloadable-operators-regexp)))
3660 (point))))
3661
3662 )))
d9e94c22
MS
3663
3664(defsubst c-simple-skip-symbol-backward ()
3665 ;; If the point is at the end of a symbol then skip backward to the
3666 ;; beginning of it. Don't move otherwise. Return non-nil if point
3667 ;; moved.
0386b551
AM
3668 ;;
3669 ;; This function might do hidden buffer changes.
d9e94c22
MS
3670 (or (< (skip-syntax-backward "w_") 0)
3671 (and (c-major-mode-is 'pike-mode)
3672 ;; Handle the `<operator> syntax in Pike.
3673 (let ((pos (point)))
2a15eb73 3674 (if (and (< (skip-chars-backward "-!%&*+/<=>^|~[]()") 0)
d9e94c22
MS
3675 (< (skip-chars-backward "`") 0)
3676 (looking-at c-symbol-key)
3677 (>= (match-end 0) pos))
3678 t
3679 (goto-char pos)
3680 nil)))))
3681
0386b551 3682(defun c-beginning-of-current-token (&optional back-limit)
d9e94c22
MS
3683 ;; Move to the beginning of the current token. Do not move if not
3684 ;; in the middle of one. BACK-LIMIT may be used to bound the
3685 ;; backward search; if given it's assumed to be at the boundary
a85fd6da 3686 ;; between two tokens. Return non-nil if the point is moved, nil
580fba94 3687 ;; otherwise.
0386b551
AM
3688 ;;
3689 ;; This function might do hidden buffer changes.
d9e94c22 3690 (let ((start (point)))
580fba94
AM
3691 (if (looking-at "\\w\\|\\s_")
3692 (skip-syntax-backward "w_" back-limit)
3693 (when (< (skip-syntax-backward ".()" back-limit) 0)
3694 (while (let ((pos (or (and (looking-at c-nonsymbol-token-regexp)
3695 (match-end 0))
3696 ;; `c-nonsymbol-token-regexp' should always match
3697 ;; since we've skipped backward over punctuator
3698 ;; or paren syntax, but consume one char in case
3699 ;; it doesn't so that we don't leave point before
3700 ;; some earlier incorrect token.
3701 (1+ (point)))))
3702 (if (<= pos start)
3703 (goto-char pos))))))
3704 (< (point) start)))
d9e94c22 3705
ff959bab 3706(defun c-end-of-current-token (&optional back-limit)
d9e94c22
MS
3707 ;; Move to the end of the current token. Do not move if not in the
3708 ;; middle of one. BACK-LIMIT may be used to bound the backward
3709 ;; search; if given it's assumed to be at the boundary between two
ff959bab 3710 ;; tokens. Return non-nil if the point is moved, nil otherwise.
0386b551
AM
3711 ;;
3712 ;; This function might do hidden buffer changes.
d9e94c22
MS
3713 (let ((start (point)))
3714 (cond ((< (skip-syntax-backward "w_" (1- start)) 0)
3715 (skip-syntax-forward "w_"))
3716 ((< (skip-syntax-backward ".()" back-limit) 0)
3717 (while (progn
3718 (if (looking-at c-nonsymbol-token-regexp)
3719 (goto-char (match-end 0))
3720 ;; `c-nonsymbol-token-regexp' should always match since
3721 ;; we've skipped backward over punctuator or paren
3722 ;; syntax, but move forward in case it doesn't so that
3723 ;; we don't leave point earlier than we started with.
3724 (forward-char))
ff959bab
MS
3725 (< (point) start)))))
3726 (> (point) start)))
d9e94c22
MS
3727
3728(defconst c-jump-syntax-balanced
3729 (if (memq 'gen-string-delim c-emacs-features)
3730 "\\w\\|\\s_\\|\\s\(\\|\\s\)\\|\\s\"\\|\\s|"
3731 "\\w\\|\\s_\\|\\s\(\\|\\s\)\\|\\s\""))
3732
3733(defconst c-jump-syntax-unbalanced
3734 (if (memq 'gen-string-delim c-emacs-features)
3735 "\\w\\|\\s_\\|\\s\"\\|\\s|"
3736 "\\w\\|\\s_\\|\\s\""))
3737
3738(defun c-forward-token-2 (&optional count balanced limit)
3739 "Move forward by tokens.
3740A token is defined as all symbols and identifiers which aren't
3741syntactic whitespace \(note that multicharacter tokens like \"==\" are
3742treated properly). Point is always either left at the beginning of a
3743token or not moved at all. COUNT specifies the number of tokens to
3744move; a negative COUNT moves in the opposite direction. A COUNT of 0
3745moves to the next token beginning only if not already at one. If
3746BALANCED is true, move over balanced parens, otherwise move into them.
3747Also, if BALANCED is true, never move out of an enclosing paren.
3748
3749LIMIT sets the limit for the movement and defaults to the point limit.
3750The case when LIMIT is set in the middle of a token, comment or macro
3751is handled correctly, i.e. the point won't be left there.
3752
3753Return the number of tokens left to move \(positive or negative). If
3754BALANCED is true, a move over a balanced paren counts as one. Note
3755that if COUNT is 0 and no appropriate token beginning is found, 1 will
3756be returned. Thus, a return value of 0 guarantees that point is at
3757the requested position and a return value less \(without signs) than
0386b551
AM
3758COUNT guarantees that point is at the beginning of some token.
3759
3760Note that this function might do hidden buffer changes. See the
3761comment at the start of cc-engine.el for more info."
d9e94c22
MS
3762
3763 (or count (setq count 1))
3764 (if (< count 0)
3765 (- (c-backward-token-2 (- count) balanced limit))
3766
3767 (let ((jump-syntax (if balanced
3768 c-jump-syntax-balanced
3769 c-jump-syntax-unbalanced))
3770 (last (point))
3771 (prev (point)))
3772
3773 (if (zerop count)
3774 ;; If count is zero we should jump if in the middle of a token.
3775 (c-end-of-current-token))
3776
3777 (save-restriction
3778 (if limit (narrow-to-region (point-min) limit))
3779 (if (/= (point)
3780 (progn (c-forward-syntactic-ws) (point)))
3781 ;; Skip whitespace. Count this as a move if we did in
3782 ;; fact move.
3783 (setq count (max (1- count) 0)))
3784
3785 (if (eobp)
3786 ;; Moved out of bounds. Make sure the returned count isn't zero.
3787 (progn
3788 (if (zerop count) (setq count 1))
3789 (goto-char last))
3790
3791 ;; Use `condition-case' to avoid having the limit tests
3792 ;; inside the loop.
3793 (condition-case nil
3794 (while (and
3795 (> count 0)
3796 (progn
3797 (setq last (point))
3798 (cond ((looking-at jump-syntax)
3799 (goto-char (scan-sexps (point) 1))
3800 t)
3801 ((looking-at c-nonsymbol-token-regexp)
3802 (goto-char (match-end 0))
3803 t)
3804 ;; `c-nonsymbol-token-regexp' above should always
3805 ;; match if there are correct tokens. Try to
3806 ;; widen to see if the limit was set in the
3807 ;; middle of one, else fall back to treating
3808 ;; the offending thing as a one character token.
3809 ((and limit
3810 (save-restriction
3811 (widen)
3812 (looking-at c-nonsymbol-token-regexp)))
3813 nil)
3814 (t
3815 (forward-char)
3816 t))))
3817 (c-forward-syntactic-ws)
3818 (setq prev last
3819 count (1- count)))
3820 (error (goto-char last)))
3821
3822 (when (eobp)
3823 (goto-char prev)
3824 (setq count (1+ count)))))
3825
3826 count)))
3827
3828(defun c-backward-token-2 (&optional count balanced limit)
3829 "Move backward by tokens.
3830See `c-forward-token-2' for details."
3831
3832 (or count (setq count 1))
3833 (if (< count 0)
3834 (- (c-forward-token-2 (- count) balanced limit))
3835
3836 (or limit (setq limit (point-min)))
3837 (let ((jump-syntax (if balanced
3838 c-jump-syntax-balanced
3839 c-jump-syntax-unbalanced))
3840 (last (point)))
3841
3842 (if (zerop count)
3843 ;; The count is zero so try to skip to the beginning of the
3844 ;; current token.
3845 (if (> (point)
3846 (progn (c-beginning-of-current-token) (point)))
3847 (if (< (point) limit)
3848 ;; The limit is inside the same token, so return 1.
3849 (setq count 1))
3850
3851 ;; We're not in the middle of a token. If there's
3852 ;; whitespace after the point then we must move backward,
3853 ;; so set count to 1 in that case.
3854 (and (looking-at c-syntactic-ws-start)
3855 ;; If we're looking at a '#' that might start a cpp
3856 ;; directive then we have to do a more elaborate check.
3857 (or (/= (char-after) ?#)
3858 (not c-opt-cpp-prefix)
3859 (save-excursion
3860 (and (= (point)
3861 (progn (beginning-of-line)
3862 (looking-at "[ \t]*")
3863 (match-end 0)))
3864 (or (bobp)
3865 (progn (backward-char)
3866 (not (eq (char-before) ?\\)))))))
3867 (setq count 1))))
3868
3869 ;; Use `condition-case' to avoid having to check for buffer
3870 ;; limits in `backward-char', `scan-sexps' and `goto-char' below.
3871 (condition-case nil
3872 (while (and
3873 (> count 0)
3874 (progn
3875 (c-backward-syntactic-ws)
3876 (backward-char)
3877 (if (looking-at jump-syntax)
3878 (goto-char (scan-sexps (1+ (point)) -1))
3879 ;; This can be very inefficient if there's a long
3880 ;; sequence of operator tokens without any separation.
3881 ;; That doesn't happen in practice, anyway.
3882 (c-beginning-of-current-token))
3883 (>= (point) limit)))
3884 (setq last (point)
3885 count (1- count)))
3886 (error (goto-char last)))
3887
3888 (if (< (point) limit)
3889 (goto-char last))
3890
3891 count)))
3892
3893(defun c-forward-token-1 (&optional count balanced limit)
3894 "Like `c-forward-token-2' but doesn't treat multicharacter operator
3895tokens like \"==\" as single tokens, i.e. all sequences of symbol
3896characters are jumped over character by character. This function is
3897for compatibility only; it's only a wrapper over `c-forward-token-2'."
3898 (let ((c-nonsymbol-token-regexp "\\s.\\|\\s\(\\|\\s\)"))
3899 (c-forward-token-2 count balanced limit)))
3900
3901(defun c-backward-token-1 (&optional count balanced limit)
3902 "Like `c-backward-token-2' but doesn't treat multicharacter operator
3903tokens like \"==\" as single tokens, i.e. all sequences of symbol
3904characters are jumped over character by character. This function is
3905for compatibility only; it's only a wrapper over `c-backward-token-2'."
3906 (let ((c-nonsymbol-token-regexp "\\s.\\|\\s\(\\|\\s\)"))
3907 (c-backward-token-2 count balanced limit)))
3908
3909\f
3910;; Tools for doing searches restricted to syntactically relevant text.
3911
3912(defun c-syntactic-re-search-forward (regexp &optional bound noerror
3913 paren-level not-inside-token
3914 lookbehind-submatch)
3915 "Like `re-search-forward', but only report matches that are found
3916in syntactically significant text. I.e. matches in comments, macros
3917or string literals are ignored. The start point is assumed to be
3918outside any comment, macro or string literal, or else the content of
3919that region is taken as syntactically significant text.
3920
3921If PAREN-LEVEL is non-nil, an additional restriction is added to
2a15eb73
MS
3922ignore matches in nested paren sexps. The search will also not go
3923outside the current list sexp, which has the effect that if the point
3924should be moved to BOUND when no match is found \(i.e. NOERROR is
3925neither nil nor t), then it will be at the closing paren if the end of
3926the current list sexp is encountered first.
d9e94c22
MS
3927
3928If NOT-INSIDE-TOKEN is non-nil, matches in the middle of tokens are
3929ignored. Things like multicharacter operators and special symbols
3930\(e.g. \"`()\" in Pike) are handled but currently not floating point
3931constants.
3932
3933If LOOKBEHIND-SUBMATCH is non-nil, it's taken as a number of a
3934subexpression in REGEXP. The end of that submatch is used as the
3935position to check for syntactic significance. If LOOKBEHIND-SUBMATCH
3936isn't used or if that subexpression didn't match then the start
3937position of the whole match is used instead. The \"look behind\"
3938subexpression is never tested before the starting position, so it
3939might be a good idea to include \\=\\= as a match alternative in it.
3940
3941Optimization note: Matches might be missed if the \"look behind\"
2a15eb73 3942subexpression can match the end of nonwhite syntactic whitespace,
d9e94c22 3943i.e. the end of comments or cpp directives. This since the function
2a15eb73
MS
3944skips over such things before resuming the search. It's on the other
3945hand not safe to assume that the \"look behind\" subexpression never
3946matches syntactic whitespace.
3947
3948Bug: Unbalanced parens inside cpp directives are currently not handled
3949correctly \(i.e. they don't get ignored as they should) when
0386b551
AM
3950PAREN-LEVEL is set.
3951
3952Note that this function might do hidden buffer changes. See the
3953comment at the start of cc-engine.el for more info."
d9e94c22
MS
3954
3955 (or bound (setq bound (point-max)))
3956 (if paren-level (setq paren-level -1))
3957
3958 ;;(message "c-syntactic-re-search-forward %s %s %S" (point) bound regexp)
3959
3960 (let ((start (point))
2a15eb73
MS
3961 tmp
3962 ;; Start position for the last search.
3963 search-pos
3964 ;; The `parse-partial-sexp' state between the start position
3965 ;; and the point.
3966 state
3967 ;; The current position after the last state update. The next
3968 ;; `parse-partial-sexp' continues from here.
3969 (state-pos (point))
3970 ;; The position at which to check the state and the state
3971 ;; there. This is separate from `state-pos' since we might
3972 ;; need to back up before doing the next search round.
3973 check-pos check-state
3974 ;; Last position known to end a token.
d9e94c22 3975 (last-token-end-pos (point-min))
2a15eb73
MS
3976 ;; Set when a valid match is found.
3977 found)
d9e94c22
MS
3978
3979 (condition-case err
3980 (while
3981 (and
2a15eb73
MS
3982 (progn
3983 (setq search-pos (point))
3984 (re-search-forward regexp bound noerror))
d9e94c22
MS
3985
3986 (progn
2a15eb73
MS
3987 (setq state (parse-partial-sexp
3988 state-pos (match-beginning 0) paren-level nil state)
3989 state-pos (point))
d9e94c22 3990 (if (setq check-pos (and lookbehind-submatch
2a15eb73
MS
3991 (or (not paren-level)
3992 (>= (car state) 0))
d9e94c22
MS
3993 (match-end lookbehind-submatch)))
3994 (setq check-state (parse-partial-sexp
2a15eb73
MS
3995 state-pos check-pos paren-level nil state))
3996 (setq check-pos state-pos
d9e94c22
MS
3997 check-state state))
3998
2a15eb73
MS
3999 ;; NOTE: If we got a look behind subexpression and get
4000 ;; an insignificant match in something that isn't
d9e94c22
MS
4001 ;; syntactic whitespace (i.e. strings or in nested
4002 ;; parentheses), then we can never skip more than a
2a15eb73
MS
4003 ;; single character from the match start position
4004 ;; (i.e. `state-pos' here) before continuing the
4005 ;; search. That since the look behind subexpression
4006 ;; might match the end of the insignificant region in
4007 ;; the next search.
d9e94c22
MS
4008
4009 (cond
d9e94c22
MS
4010 ((elt check-state 7)
4011 ;; Match inside a line comment. Skip to eol. Use
4012 ;; `re-search-forward' instead of `skip-chars-forward' to get
4013 ;; the right bound behavior.
4014 (re-search-forward "[\n\r]" bound noerror))
4015
4016 ((elt check-state 4)
4017 ;; Match inside a block comment. Skip to the '*/'.
4018 (search-forward "*/" bound noerror))
4019
4020 ((and (not (elt check-state 5))
4021 (eq (char-before check-pos) ?/)
2a15eb73 4022 (not (c-get-char-property (1- check-pos) 'syntax-table))
d9e94c22
MS
4023 (memq (char-after check-pos) '(?/ ?*)))
4024 ;; Match in the middle of the opener of a block or line
4025 ;; comment.
4026 (if (= (char-after check-pos) ?/)
4027 (re-search-forward "[\n\r]" bound noerror)
4028 (search-forward "*/" bound noerror)))
4029
2a15eb73
MS
4030 ;; The last `parse-partial-sexp' above might have
4031 ;; stopped short of the real check position if the end
4032 ;; of the current sexp was encountered in paren-level
4033 ;; mode. The checks above are always false in that
4034 ;; case, and since they can do better skipping in
4035 ;; lookbehind-submatch mode, we do them before
4036 ;; checking the paren level.
4037
4038 ((and paren-level
4039 (/= (setq tmp (car check-state)) 0))
4040 ;; Check the paren level first since we're short of the
4041 ;; syntactic checking position if the end of the
4042 ;; current sexp was encountered by `parse-partial-sexp'.
4043 (if (> tmp 0)
4044
4045 ;; Inside a nested paren sexp.
4046 (if lookbehind-submatch
4047 ;; See the NOTE above.
4048 (progn (goto-char state-pos) t)
4049 ;; Skip out of the paren quickly.
4050 (setq state (parse-partial-sexp state-pos bound 0 nil state)
4051 state-pos (point)))
4052
4053 ;; Have exited the current paren sexp.
4054 (if noerror
4055 (progn
4056 ;; The last `parse-partial-sexp' call above
4057 ;; has left us just after the closing paren
4058 ;; in this case, so we can modify the bound
4059 ;; to leave the point at the right position
4060 ;; upon return.
4061 (setq bound (1- (point)))
4062 nil)
4063 (signal 'search-failed (list regexp)))))
4064
4065 ((setq tmp (elt check-state 3))
4066 ;; Match inside a string.
4067 (if (or lookbehind-submatch
4068 (not (integerp tmp)))
4069 ;; See the NOTE above.
4070 (progn (goto-char state-pos) t)
4071 ;; Skip to the end of the string before continuing.
4072 (let ((ender (make-string 1 tmp)) (continue t))
4073 (while (if (search-forward ender bound noerror)
4074 (progn
4075 (setq state (parse-partial-sexp
4076 state-pos (point) nil nil state)
4077 state-pos (point))
4078 (elt state 3))
4079 (setq continue nil)))
4080 continue)))
d9e94c22
MS
4081
4082 ((save-excursion
4083 (save-match-data
4084 (c-beginning-of-macro start)))
4085 ;; Match inside a macro. Skip to the end of it.
4086 (c-end-of-macro)
4087 (cond ((<= (point) bound) t)
4088 (noerror nil)
2a15eb73 4089 (t (signal 'search-failed (list regexp)))))
d9e94c22 4090
2a15eb73
MS
4091 ((and not-inside-token
4092 (or (< check-pos last-token-end-pos)
4093 (< check-pos
4094 (save-excursion
4095 (goto-char check-pos)
4096 (save-match-data
4097 (c-end-of-current-token last-token-end-pos))
4098 (setq last-token-end-pos (point))))))
4099 ;; Inside a token.
4100 (if lookbehind-submatch
4101 ;; See the NOTE above.
4102 (goto-char state-pos)
4103 (goto-char (min last-token-end-pos bound))))
d9e94c22
MS
4104
4105 (t
4106 ;; A real match.
4107 (setq found t)
2a15eb73
MS
4108 nil)))
4109
4110 ;; Should loop to search again, but take care to avoid
4111 ;; looping on the same spot.
4112 (or (/= search-pos (point))
4113 (if (= (point) bound)
4114 (if noerror
4115 nil
4116 (signal 'search-failed (list regexp)))
4117 (forward-char)
4118 t))))
d9e94c22
MS
4119
4120 (error
4121 (goto-char start)
4122 (signal (car err) (cdr err))))
4123
2a15eb73 4124 ;;(message "c-syntactic-re-search-forward done %s" (or (match-end 0) (point)))
d9e94c22
MS
4125
4126 (if found
4127 (progn
2a15eb73
MS
4128 (goto-char (match-end 0))
4129 (match-end 0))
d9e94c22
MS
4130
4131 ;; Search failed. Set point as appropriate.
2a15eb73
MS
4132 (if (eq noerror t)
4133 (goto-char start)
4134 (goto-char bound))
d9e94c22
MS
4135 nil)))
4136
47641aac
GM
4137(defvar safe-pos-list) ; bound in c-syntactic-skip-backward
4138
d0fcee66
AM
4139(defsubst c-ssb-lit-begin ()
4140 ;; Return the start of the literal point is in, or nil.
4141 ;; We read and write the variables `safe-pos', `safe-pos-list', `state'
4142 ;; bound in the caller.
4143
4144 ;; Use `parse-partial-sexp' from a safe position down to the point to check
4145 ;; if it's outside comments and strings.
4146 (save-excursion
4147 (let ((pos (point)) safe-pos state pps-end-pos)
4148 ;; Pick a safe position as close to the point as possible.
4149 ;;
4150 ;; FIXME: Consult `syntax-ppss' here if our cache doesn't give a good
4151 ;; position.
47641aac 4152
d0fcee66
AM
4153 (while (and safe-pos-list
4154 (> (car safe-pos-list) (point)))
4155 (setq safe-pos-list (cdr safe-pos-list)))
4156 (unless (setq safe-pos (car-safe safe-pos-list))
4157 (setq safe-pos (max (or (c-safe-position
4158 (point) (or c-state-cache
4159 (c-parse-state)))
4160 0)
4161 (point-min))
4162 safe-pos-list (list safe-pos)))
4163
4164 ;; Cache positions along the way to use if we have to back up more. We
4165 ;; cache every closing paren on the same level. If the paren cache is
4166 ;; relevant in this region then we're typically already on the same
4167 ;; level as the target position. Note that we might cache positions
4168 ;; after opening parens in case safe-pos is in a nested list. That's
4169 ;; both uncommon and harmless.
4170 (while (progn
4171 (setq state (parse-partial-sexp
4172 safe-pos pos 0))
4173 (< (point) pos))
4174 (setq safe-pos (point)
4175 safe-pos-list (cons safe-pos safe-pos-list)))
4176
4177 ;; If the state contains the start of the containing sexp we cache that
4178 ;; position too, so that parse-partial-sexp in the next run has a bigger
4179 ;; chance of starting at the same level as the target position and thus
4180 ;; will get more good safe positions into the list.
4181 (if (elt state 1)
4182 (setq safe-pos (1+ (elt state 1))
4183 safe-pos-list (cons safe-pos safe-pos-list)))
4184
4185 (if (or (elt state 3) (elt state 4))
4186 ;; Inside string or comment. Continue search at the
4187 ;; beginning of it.
4188 (elt state 8)))))
4189
0386b551 4190(defun c-syntactic-skip-backward (skip-chars &optional limit paren-level)
d9e94c22
MS
4191 "Like `skip-chars-backward' but only look at syntactically relevant chars,
4192i.e. don't stop at positions inside syntactic whitespace or string
4193literals. Preprocessor directives are also ignored, with the exception
4194of the one that the point starts within, if any. If LIMIT is given,
0386b551
AM
4195it's assumed to be at a syntactically relevant position.
4196
4197If PAREN-LEVEL is non-nil, the function won't stop in nested paren
4198sexps, and the search will also not go outside the current paren sexp.
4199However, if LIMIT or the buffer limit is reached inside a nested paren
4200then the point will be left at the limit.
4201
4202Non-nil is returned if the point moved, nil otherwise.
4203
4204Note that this function might do hidden buffer changes. See the
4205comment at the start of cc-engine.el for more info."
d9e94c22
MS
4206
4207 (let ((start (point))
d0fcee66 4208 state-2
d9e94c22
MS
4209 ;; A list of syntactically relevant positions in descending
4210 ;; order. It's used to avoid scanning repeatedly over
4211 ;; potentially large regions with `parse-partial-sexp' to verify
d0fcee66 4212 ;; each position. Used in `c-ssb-lit-begin'
d9e94c22
MS
4213 safe-pos-list
4214 ;; The result from `c-beginning-of-macro' at the start position or the
4215 ;; start position itself if it isn't within a macro. Evaluated on
4216 ;; demand.
0386b551
AM
4217 start-macro-beg
4218 ;; The earliest position after the current one with the same paren
4219 ;; level. Used only when `paren-level' is set.
d0fcee66 4220 lit-beg
0386b551 4221 (paren-level-pos (point)))
d9e94c22 4222
d0fcee66
AM
4223 (while
4224 (progn
4225 ;; The next loop "tries" to find the end point each time round,
4226 ;; loops when it hasn't succeeded.
4227 (while
4228 (and
4229 (< (skip-chars-backward skip-chars limit) 0)
4230
4231 (let ((pos (point)) state-2 pps-end-pos)
4232
4233 (cond
4234 ;; Don't stop inside a literal
4235 ((setq lit-beg (c-ssb-lit-begin))
4236 (goto-char lit-beg)
4237 t)
4238
4239 ((and paren-level
4240 (save-excursion
4241 (setq state-2 (parse-partial-sexp
4242 pos paren-level-pos -1)
4243 pps-end-pos (point))
4244 (/= (car state-2) 0)))
4245 ;; Not at the right level.
4246
4247 (if (and (< (car state-2) 0)
4248 ;; We stop above if we go out of a paren.
4249 ;; Now check whether it precedes or is
4250 ;; nested in the starting sexp.
4251 (save-excursion
4252 (setq state-2
4253 (parse-partial-sexp
4254 pps-end-pos paren-level-pos
4255 nil nil state-2))
4256 (< (car state-2) 0)))
4257
4258 ;; We've stopped short of the starting position
4259 ;; so the hit was inside a nested list. Go up
4260 ;; until we are at the right level.
4261 (condition-case nil
4262 (progn
4263 (goto-char (scan-lists pos -1
4264 (- (car state-2))))
4265 (setq paren-level-pos (point))
4266 (if (and limit (>= limit paren-level-pos))
0386b551 4267 (progn
d0fcee66
AM
4268 (goto-char limit)
4269 nil)
4270 t))
4271 (error
4272 (goto-char (or limit (point-min)))
4273 nil))
4274
4275 ;; The hit was outside the list at the start
4276 ;; position. Go to the start of the list and exit.
4277 (goto-char (1+ (elt state-2 1)))
4278 nil))
4279
4280 ((c-beginning-of-macro limit)
4281 ;; Inside a macro.
4282 (if (< (point)
4283 (or start-macro-beg
4284 (setq start-macro-beg
4285 (save-excursion
4286 (goto-char start)
4287 (c-beginning-of-macro limit)
4288 (point)))))
4289 t
4290
4291 ;; It's inside the same macro we started in so it's
4292 ;; a relevant match.
4293 (goto-char pos)
4294 nil))))))
91af3942 4295
d0fcee66
AM
4296 (> (point)
4297 (progn
4298 ;; Skip syntactic ws afterwards so that we don't stop at the
4299 ;; end of a comment if `skip-chars' is something like "^/".
4300 (c-backward-syntactic-ws)
4301 (point)))))
d9e94c22 4302
0386b551
AM
4303 ;; We might want to extend this with more useful return values in
4304 ;; the future.
4305 (/= (point) start)))
4306
4307;; The following is an alternative implementation of
4308;; `c-syntactic-skip-backward' that uses backward movement to keep
4309;; track of the syntactic context. It turned out to be generally
4310;; slower than the one above which uses forward checks from earlier
4311;; safe positions.
4312;;
4313;;(defconst c-ssb-stop-re
4314;; ;; The regexp matching chars `c-syntactic-skip-backward' needs to
4315;; ;; stop at to avoid going into comments and literals.
4316;; (concat
4317;; ;; Match comment end syntax and string literal syntax. Also match
4318;; ;; '/' for block comment endings (not covered by comment end
4319;; ;; syntax).
4320;; "\\s>\\|/\\|\\s\""
4321;; (if (memq 'gen-string-delim c-emacs-features)
4322;; "\\|\\s|"
4323;; "")
4324;; (if (memq 'gen-comment-delim c-emacs-features)
4325;; "\\|\\s!"
4326;; "")))
4327;;
4328;;(defconst c-ssb-stop-paren-re
4329;; ;; Like `c-ssb-stop-re' but also stops at paren chars.
4330;; (concat c-ssb-stop-re "\\|\\s(\\|\\s)"))
4331;;
4332;;(defconst c-ssb-sexp-end-re
4333;; ;; Regexp matching the ending syntax of a complex sexp.
4334;; (concat c-string-limit-regexp "\\|\\s)"))
4335;;
4336;;(defun c-syntactic-skip-backward (skip-chars &optional limit paren-level)
4337;; "Like `skip-chars-backward' but only look at syntactically relevant chars,
4338;;i.e. don't stop at positions inside syntactic whitespace or string
4339;;literals. Preprocessor directives are also ignored. However, if the
4340;;point is within a comment, string literal or preprocessor directory to
4341;;begin with, its contents is treated as syntactically relevant chars.
4342;;If LIMIT is given, it limits the backward search and the point will be
4343;;left there if no earlier position is found.
4344;;
4345;;If PAREN-LEVEL is non-nil, the function won't stop in nested paren
4346;;sexps, and the search will also not go outside the current paren sexp.
4347;;However, if LIMIT or the buffer limit is reached inside a nested paren
4348;;then the point will be left at the limit.
4349;;
4350;;Non-nil is returned if the point moved, nil otherwise.
4351;;
4352;;Note that this function might do hidden buffer changes. See the
4353;;comment at the start of cc-engine.el for more info."
4354;;
4355;; (save-restriction
4356;; (when limit
4357;; (narrow-to-region limit (point-max)))
4358;;
4359;; (let ((start (point)))
4360;; (catch 'done
4361;; (while (let ((last-pos (point))
4362;; (stop-pos (progn
4363;; (skip-chars-backward skip-chars)
4364;; (point))))
4365;;
4366;; ;; Skip back over the same region as
4367;; ;; `skip-chars-backward' above, but keep to
4368;; ;; syntactically relevant positions.
4369;; (goto-char last-pos)
4370;; (while (and
4371;; ;; `re-search-backward' with a single char regexp
4372;; ;; should be fast.
4373;; (re-search-backward
4374;; (if paren-level c-ssb-stop-paren-re c-ssb-stop-re)
4375;; stop-pos 'move)
4376;;
4377;; (progn
4378;; (cond
4379;; ((looking-at "\\s(")
4380;; ;; `paren-level' is set and we've found the
4381;; ;; start of the containing paren.
4382;; (forward-char)
4383;; (throw 'done t))
4384;;
4385;; ((looking-at c-ssb-sexp-end-re)
4386;; ;; We're at the end of a string literal or paren
4387;; ;; sexp (if `paren-level' is set).
4388;; (forward-char)
4389;; (condition-case nil
4390;; (c-backward-sexp)
4391;; (error
4392;; (goto-char limit)
4393;; (throw 'done t))))
4394;;
4395;; (t
4396;; (forward-char)
4397;; ;; At the end of some syntactic ws or possibly
4398;; ;; after a plain '/' operator.
4399;; (let ((pos (point)))
4400;; (c-backward-syntactic-ws)
4401;; (if (= pos (point))
4402;; ;; Was a plain '/' operator. Go past it.
4403;; (backward-char)))))
4404;;
4405;; (> (point) stop-pos))))
4406;;
4407;; ;; Now the point is either at `stop-pos' or at some
4408;; ;; position further back if `stop-pos' was at a
4409;; ;; syntactically irrelevant place.
4410;;
4411;; ;; Skip additional syntactic ws so that we don't stop
4412;; ;; at the end of a comment if `skip-chars' is
4413;; ;; something like "^/".
4414;; (c-backward-syntactic-ws)
4415;;
4416;; (< (point) stop-pos))))
4417;;
4418;; ;; We might want to extend this with more useful return values
4419;; ;; in the future.
4420;; (/= (point) start))))
d9e94c22
MS
4421
4422\f
4423;; Tools for handling comments and string literals.
4424
13d49cbb 4425(defun c-in-literal (&optional lim detect-cpp)
d9e94c22
MS
4426 "Return the type of literal point is in, if any.
4427The return value is `c' if in a C-style comment, `c++' if in a C++
4428style comment, `string' if in a string literal, `pound' if DETECT-CPP
4429is non-nil and in a preprocessor line, or nil if somewhere else.
4430Optional LIM is used as the backward limit of the search. If omitted,
4431or nil, `c-beginning-of-defun' is used.
4432
4433The last point calculated is cached if the cache is enabled, i.e. if
4434`c-in-literal-cache' is bound to a two element vector.
4435
0386b551
AM
4436Note that this function might do hidden buffer changes. See the
4437comment at the start of cc-engine.el for more info."
2cc769a8
AM
4438 (save-restriction
4439 (widen)
56d093a9 4440 (let* ((safe-place (c-state-semi-safe-place (point)))
2cc769a8
AM
4441 (lit (c-state-pp-to-literal safe-place (point))))
4442 (or (cadr lit)
4443 (and detect-cpp
4444 (save-excursion (c-beginning-of-macro))
4445 'pound)))))
d9e94c22
MS
4446
4447(defun c-literal-limits (&optional lim near not-in-delimiter)
4448 "Return a cons of the beginning and end positions of the comment or
4449string surrounding point (including both delimiters), or nil if point
4450isn't in one. If LIM is non-nil, it's used as the \"safe\" position
4451to start parsing from. If NEAR is non-nil, then the limits of any
4452literal next to point is returned. \"Next to\" means there's only
4453spaces and tabs between point and the literal. The search for such a
4454literal is done first in forward direction. If NOT-IN-DELIMITER is
4455non-nil, the case when point is inside a starting delimiter won't be
a85fd6da 4456recognized. This only has effect for comments which have starting
d9e94c22
MS
4457delimiters with more than one character.
4458
0386b551
AM
4459Note that this function might do hidden buffer changes. See the
4460comment at the start of cc-engine.el for more info."
d9e94c22
MS
4461
4462 (save-excursion
4463 (let* ((pos (point))
56d093a9 4464 (lim (or lim (c-state-semi-safe-place pos)))
2cc769a8
AM
4465 (pp-to-lit (save-restriction
4466 (widen)
691e26ae 4467 (c-state-pp-to-literal lim pos not-in-delimiter)))
13d49cbb 4468 (state (car pp-to-lit))
13d49cbb 4469 (lit-limits (car (cddr pp-to-lit))))
d9e94c22 4470
13d49cbb
AM
4471 (cond
4472 (lit-limits)
13d49cbb
AM
4473
4474 (near
4475 (goto-char pos)
4476 ;; Search forward for a literal.
4477 (skip-chars-forward " \t")
4478 (cond
4479 ((looking-at c-string-limit-regexp) ; String.
4480 (cons (point) (or (c-safe (c-forward-sexp 1) (point))
4481 (point-max))))
d9e94c22 4482
13d49cbb
AM
4483 ((looking-at c-comment-start-regexp) ; Line or block comment.
4484 (cons (point) (progn (c-forward-single-comment) (point))))
4485
4486 (t
4487 ;; Search backward.
4488 (skip-chars-backward " \t")
4489
4490 (let ((end (point)) beg)
4491 (cond
4492 ((save-excursion
4493 (< (skip-syntax-backward c-string-syntax) 0)) ; String.
4494 (setq beg (c-safe (c-backward-sexp 1) (point))))
4495
4496 ((and (c-safe (forward-char -2) t)
4497 (looking-at "*/"))
4498 ;; Block comment. Due to the nature of line
4499 ;; comments, they will always be covered by the
4500 ;; normal case above.
4501 (goto-char end)
4502 (c-backward-single-comment)
4503 ;; If LIM is bogus, beg will be bogus.
4504 (setq beg (point))))
4505
4506 (if beg (cons beg end))))))
4507 ))))
d9e94c22 4508
0386b551
AM
4509;; In case external callers use this; it did have a docstring.
4510(defalias 'c-literal-limits-fast 'c-literal-limits)
d9e94c22 4511
0386b551
AM
4512(defun c-collect-line-comments (range)
4513 "If the argument is a cons of two buffer positions (such as returned by
4514`c-literal-limits'), and that range contains a C++ style line comment,
4515then an extended range is returned that contains all adjacent line
4516comments (i.e. all comments that starts in the same column with no
4517empty lines or non-whitespace characters between them). Otherwise the
4518argument is returned.
d9e94c22 4519
0386b551
AM
4520Note that this function might do hidden buffer changes. See the
4521comment at the start of cc-engine.el for more info."
d9e94c22
MS
4522
4523 (save-excursion
0386b551
AM
4524 (condition-case nil
4525 (if (and (consp range) (progn
4526 (goto-char (car range))
4527 (looking-at c-line-comment-starter)))
b414f371 4528 (let ((col (current-column))
0386b551
AM
4529 (beg (point))
4530 (bopl (c-point 'bopl))
4531 (end (cdr range)))
4532 ;; Got to take care in the backward direction to handle
4533 ;; comments which are preceded by code.
4534 (while (and (c-backward-single-comment)
4535 (>= (point) bopl)
4536 (looking-at c-line-comment-starter)
4537 (= col (current-column)))
4538 (setq beg (point)
4539 bopl (c-point 'bopl)))
4540 (goto-char end)
4541 (while (and (progn (skip-chars-forward " \t")
4542 (looking-at c-line-comment-starter))
4543 (= col (current-column))
4544 (prog1 (zerop (forward-line 1))
4545 (setq end (point)))))
4546 (cons beg end))
4547 range)
4548 (error range))))
d9e94c22
MS
4549
4550(defun c-literal-type (range)
4551 "Convenience function that given the result of `c-literal-limits',
a85fd6da
AM
4552returns nil or the type of literal that the range surrounds, one
4553of the symbols 'c, 'c++ or 'string. It's much faster than using
4554`c-in-literal' and is intended to be used when you need both the
4555type of a literal and its limits.
d9e94c22 4556
0386b551
AM
4557Note that this function might do hidden buffer changes. See the
4558comment at the start of cc-engine.el for more info."
4559
d9e94c22
MS
4560 (if (consp range)
4561 (save-excursion
4562 (goto-char (car range))
4563 (cond ((looking-at c-string-limit-regexp) 'string)
4564 ((or (looking-at "//") ; c++ line comment
4565 (and (looking-at "\\s<") ; comment starter
4566 (looking-at "#"))) ; awk comment.
4567 'c++)
4568 (t 'c))) ; Assuming the range is valid.
4569 range))
4570
9657183b
AM
4571(defsubst c-determine-limit-get-base (start try-size)
4572 ;; Get a "safe place" approximately TRY-SIZE characters before START.
4573 ;; This doesn't preserve point.
4574 (let* ((pos (max (- start try-size) (point-min)))
56d093a9 4575 (base (c-state-semi-safe-place pos))
9657183b
AM
4576 (s (parse-partial-sexp base pos)))
4577 (if (or (nth 4 s) (nth 3 s)) ; comment or string
4578 (nth 8 s)
4579 (point))))
4580
4581(defun c-determine-limit (how-far-back &optional start try-size)
4582 ;; Return a buffer position HOW-FAR-BACK non-literal characters from START
4583 ;; (default point). This is done by going back further in the buffer then
4584 ;; searching forward for literals. The position found won't be in a
4585 ;; literal. We start searching for the sought position TRY-SIZE (default
4586 ;; twice HOW-FAR-BACK) bytes back from START. This function must be fast.
4587 ;; :-)
4588 (save-excursion
4589 (let* ((start (or start (point)))
4590 (try-size (or try-size (* 2 how-far-back)))
4591 (base (c-determine-limit-get-base start try-size))
4592 (pos base)
4593
4594 (s (parse-partial-sexp pos pos)) ; null state.
4595 stack elt size
4596 (count 0))
4597 (while (< pos start)
4598 ;; Move forward one literal each time round this loop.
4599 ;; Move forward to the start of a comment or string.
4600 (setq s (parse-partial-sexp
4601 pos
4602 start
4603 nil ; target-depth
4604 nil ; stop-before
4605 s ; state
4606 'syntax-table)) ; stop-comment
4607
4608 ;; Gather details of the non-literal-bit - starting pos and size.
4609 (setq size (- (if (or (nth 4 s) (nth 3 s))
4610 (nth 8 s)
4611 (point))
4612 pos))
4613 (if (> size 0)
4614 (setq stack (cons (cons pos size) stack)))
4615
4616 ;; Move forward to the end of the comment/string.
4617 (if (or (nth 4 s) (nth 3 s))
4618 (setq s (parse-partial-sexp
4619 (point)
4620 start
4621 nil ; target-depth
4622 nil ; stop-before
4623 s ; state
4624 'syntax-table))) ; stop-comment
4625 (setq pos (point)))
c38e0c97 4626
9657183b
AM
4627 ;; Now try and find enough non-literal characters recorded on the stack.
4628 ;; Go back one recorded literal each time round this loop.
4629 (while (and (< count how-far-back)
4630 stack)
4631 (setq elt (car stack)
4632 stack (cdr stack))
4633 (setq count (+ count (cdr elt))))
4634
4635 ;; Have we found enough yet?
4636 (cond
4637 ((>= count how-far-back)
4638 (+ (car elt) (- count how-far-back)))
4639 ((eq base (point-min))
4640 (point-min))
4641 (t
4642 (c-determine-limit (- how-far-back count) base try-size))))))
2e492df3
AM
4643
4644(defun c-determine-+ve-limit (how-far &optional start-pos)
4645 ;; Return a buffer position about HOW-FAR non-literal characters forward
4646 ;; from START-POS (default point), which must not be inside a literal.
4647 (save-excursion
4648 (let ((pos (or start-pos (point)))
4649 (count how-far)
4650 (s (parse-partial-sexp (point) (point)))) ; null state
4651 (while (and (not (eobp))
4652 (> count 0))
4653 ;; Scan over counted characters.
4654 (setq s (parse-partial-sexp
4655 pos
4656 (min (+ pos count) (point-max))
4657 nil ; target-depth
4658 nil ; stop-before
4659 s ; state
4660 'syntax-table)) ; stop-comment
4661 (setq count (- count (- (point) pos) 1)
4662 pos (point))
4663 ;; Scan over literal characters.
4664 (if (nth 8 s)
4665 (setq s (parse-partial-sexp
4666 pos
4667 (point-max)
4668 nil ; target-depth
4669 nil ; stop-before
4670 s ; state
4671 'syntax-table) ; stop-comment
4672 pos (point))))
4673 (point))))
4674
d9e94c22
MS
4675\f
4676;; `c-find-decl-spots' and accompanying stuff.
4677
4678;; Variables used in `c-find-decl-spots' to cache the search done for
4679;; the first declaration in the last call. When that function starts,
4680;; it needs to back up over syntactic whitespace to look at the last
4681;; token before the region being searched. That can sometimes cause
4682;; moves back and forth over a quite large region of comments and
4683;; macros, which would be repeated for each changed character when
4684;; we're called during fontification, since font-lock refontifies the
4685;; current line for each change. Thus it's worthwhile to cache the
4686;; first match.
4687;;
4688;; `c-find-decl-syntactic-pos' is a syntactically relevant position in
4689;; the syntactic whitespace less or equal to some start position.
4690;; There's no cached value if it's nil.
4691;;
4692;; `c-find-decl-match-pos' is the match position if
4693;; `c-find-decl-prefix-search' matched before the syntactic whitespace
4694;; at `c-find-decl-syntactic-pos', or nil if there's no such match.
4695(defvar c-find-decl-syntactic-pos nil)
4696(make-variable-buffer-local 'c-find-decl-syntactic-pos)
4697(defvar c-find-decl-match-pos nil)
4698(make-variable-buffer-local 'c-find-decl-match-pos)
4699
4700(defsubst c-invalidate-find-decl-cache (change-min-pos)
4701 (and c-find-decl-syntactic-pos
4702 (< change-min-pos c-find-decl-syntactic-pos)
4703 (setq c-find-decl-syntactic-pos nil)))
4704
4705; (defface c-debug-decl-spot-face
4706; '((t (:background "Turquoise")))
4707; "Debug face to mark the spots where `c-find-decl-spots' stopped.")
4708; (defface c-debug-decl-sws-face
4709; '((t (:background "Khaki")))
4710; "Debug face to mark the syntactic whitespace between the declaration
4711; spots and the preceding token end.")
4712
4713(defmacro c-debug-put-decl-spot-faces (match-pos decl-pos)
4714 (when (facep 'c-debug-decl-spot-face)
0386b551 4715 `(c-save-buffer-state ((match-pos ,match-pos) (decl-pos ,decl-pos))
d9e94c22
MS
4716 (c-debug-add-face (max match-pos (point-min)) decl-pos
4717 'c-debug-decl-sws-face)
4718 (c-debug-add-face decl-pos (min (1+ decl-pos) (point-max))
4719 'c-debug-decl-spot-face))))
4720(defmacro c-debug-remove-decl-spot-faces (beg end)
4721 (when (facep 'c-debug-decl-spot-face)
0386b551 4722 `(c-save-buffer-state ()
d9e94c22
MS
4723 (c-debug-remove-face ,beg ,end 'c-debug-decl-spot-face)
4724 (c-debug-remove-face ,beg ,end 'c-debug-decl-sws-face))))
4725
4726(defmacro c-find-decl-prefix-search ()
4727 ;; Macro used inside `c-find-decl-spots'. It ought to be a defun,
4728 ;; but it contains lots of free variables that refer to things
4729 ;; inside `c-find-decl-spots'. The point is left at `cfd-match-pos'
4730 ;; if there is a match, otherwise at `cfd-limit'.
0386b551 4731 ;;
ece15004
AM
4732 ;; The macro moves point forward to the next putative start of a declaration
4733 ;; or cfd-limit. This decl start is the next token after a "declaration
4734 ;; prefix". The declaration prefix is the earlier of `cfd-prop-match' and
4735 ;; `cfd-re-match'. `cfd-match-pos' is set to the decl prefix.
4736 ;;
0386b551 4737 ;; This macro might do hidden buffer changes.
d9e94c22
MS
4738
4739 '(progn
4740 ;; Find the next property match position if we haven't got one already.
4741 (unless cfd-prop-match
4742 (save-excursion
4743 (while (progn
4744 (goto-char (next-single-property-change
4745 (point) 'c-type nil cfd-limit))
4746 (and (< (point) cfd-limit)
4747 (not (eq (c-get-char-property (1- (point)) 'c-type)
4748 'c-decl-end)))))
4749 (setq cfd-prop-match (point))))
4750
0386b551
AM
4751 ;; Find the next `c-decl-prefix-or-start-re' match if we haven't
4752 ;; got one already.
d9e94c22 4753 (unless cfd-re-match
0386b551
AM
4754
4755 (if (> cfd-re-match-end (point))
4756 (goto-char cfd-re-match-end))
4757
ece15004
AM
4758 ;; Each time round, the next `while' moves forward over a pseudo match
4759 ;; of `c-decl-prefix-or-start-re' which is either inside a literal, or
4760 ;; is a ":" not preceded by "public", etc.. `cfd-re-match' and
4761 ;; `cfd-re-match-end' get set.
4762 (while
4763 (progn
4764 (setq cfd-re-match-end (re-search-forward c-decl-prefix-or-start-re
4765 cfd-limit 'move))
4766 (cond
4767 ((null cfd-re-match-end)
4768 ;; No match. Finish up and exit the loop.
4769 (setq cfd-re-match cfd-limit)
4770 nil)
4771 ((c-got-face-at
4772 (if (setq cfd-re-match (match-end 1))
4773 ;; Matched the end of a token preceding a decl spot.
4774 (progn
4775 (goto-char cfd-re-match)
4776 (1- cfd-re-match))
4777 ;; Matched a token that start a decl spot.
4778 (goto-char (match-beginning 0))
4779 (point))
4780 c-literal-faces)
4781 ;; Pseudo match inside a comment or string literal. Skip out
4782 ;; of comments and string literals.
4783 (while (progn
4784 (goto-char (next-single-property-change
4785 (point) 'face nil cfd-limit))
4786 (and (< (point) cfd-limit)
4787 (c-got-face-at (point) c-literal-faces))))
4788 t) ; Continue the loop over pseudo matches.
4789 ((and (match-string 1)
4790 (string= (match-string 1) ":")
4791 (save-excursion
4792 (or (/= (c-backward-token-2 2) 0) ; no search limit. :-(
4793 (not (looking-at c-decl-start-colon-kwd-re)))))
4794 ;; Found a ":" which isn't part of "public:", etc.
4795 t)
4796 (t nil)))) ;; Found a real match. Exit the pseudo-match loop.
4797
4798 ;; If our match was at the decl start, we have to back up over the
0386b551
AM
4799 ;; preceding syntactic ws to set `cfd-match-pos' and to catch
4800 ;; any decl spots in the syntactic ws.
d9e94c22 4801 (unless cfd-re-match
0386b551
AM
4802 (c-backward-syntactic-ws)
4803 (setq cfd-re-match (point))))
d9e94c22
MS
4804
4805 ;; Choose whichever match is closer to the start.
4806 (if (< cfd-re-match cfd-prop-match)
4807 (setq cfd-match-pos cfd-re-match
4808 cfd-re-match nil)
4809 (setq cfd-match-pos cfd-prop-match
4810 cfd-prop-match nil))
4811
4812 (goto-char cfd-match-pos)
4813
4814 (when (< cfd-match-pos cfd-limit)
4815 ;; Skip forward past comments only so we don't skip macros.
4816 (c-forward-comments)
4817 ;; Set the position to continue at. We can avoid going over
4818 ;; the comments skipped above a second time, but it's possible
4819 ;; that the comment skipping has taken us past `cfd-prop-match'
4820 ;; since the property might be used inside comments.
4821 (setq cfd-continue-pos (if cfd-prop-match
4822 (min cfd-prop-match (point))
4823 (point))))))
4824
4825(defun c-find-decl-spots (cfd-limit cfd-decl-re cfd-face-checklist cfd-fun)
0386b551 4826 ;; Call CFD-FUN for each possible spot for a declaration, cast or
a85fd6da
AM
4827 ;; label from the point to CFD-LIMIT.
4828 ;;
9657183b
AM
4829 ;; CFD-FUN is called with point at the start of the spot. It's passed two
4830 ;; arguments: The first is the end position of the token preceding the spot,
4831 ;; or 0 for the implicit match at bob. The second is a flag that is t when
4832 ;; the match is inside a macro. Point should be moved forward by at least
4833 ;; one token.
4834 ;;
4835 ;; If CFD-FUN adds `c-decl-end' properties somewhere below the current spot,
4836 ;; it should return non-nil to ensure that the next search will find them.
0386b551 4837 ;;
a85fd6da 4838 ;; Such a spot is:
0386b551
AM
4839 ;; o The first token after bob.
4840 ;; o The first token after the end of submatch 1 in
4841 ;; `c-decl-prefix-or-start-re' when that submatch matches.
4842 ;; o The start of each `c-decl-prefix-or-start-re' match when
4843 ;; submatch 1 doesn't match.
5a89f0a7 4844 ;; o The first token after the end of each occurrence of the
0386b551
AM
4845 ;; `c-type' text property with the value `c-decl-end', provided
4846 ;; `c-type-decl-end-used' is set.
4847 ;;
4848 ;; Only a spot that match CFD-DECL-RE and whose face is in the
4849 ;; CFD-FACE-CHECKLIST list causes CFD-FUN to be called. The face
4850 ;; check is disabled if CFD-FACE-CHECKLIST is nil.
d9e94c22
MS
4851 ;;
4852 ;; If the match is inside a macro then the buffer is narrowed to the
4853 ;; end of it, so that CFD-FUN can investigate the following tokens
4854 ;; without matching something that begins inside a macro and ends
4855 ;; outside it. It's to avoid this work that the CFD-DECL-RE and
4856 ;; CFD-FACE-CHECKLIST checks exist.
4857 ;;
0386b551
AM
4858 ;; The spots are visited approximately in order from top to bottom.
4859 ;; It's however the positions where `c-decl-prefix-or-start-re'
4860 ;; matches and where `c-decl-end' properties are found that are in
4861 ;; order. Since the spots often are at the following token, they
4862 ;; might be visited out of order insofar as more spots are reported
4863 ;; later on within the syntactic whitespace between the match
4864 ;; positions and their spots.
4865 ;;
4866 ;; It's assumed that comments and strings are fontified in the
d9e94c22
MS
4867 ;; searched range.
4868 ;;
4869 ;; This is mainly used in fontification, and so has an elaborate
4870 ;; cache to handle repeated calls from the same start position; see
4871 ;; the variables above.
4872 ;;
4873 ;; All variables in this function begin with `cfd-' to avoid name
4874 ;; collision with the (dynamically bound) variables used in CFD-FUN.
0386b551
AM
4875 ;;
4876 ;; This function might do hidden buffer changes.
d9e94c22 4877
0386b551
AM
4878 (let ((cfd-start-pos (point))
4879 (cfd-buffer-end (point-max))
4880 ;; The end of the token preceding the decl spot last found
4881 ;; with `c-decl-prefix-or-start-re'. `cfd-limit' if there's
4882 ;; no match.
d9e94c22 4883 cfd-re-match
0386b551
AM
4884 ;; The end position of the last `c-decl-prefix-or-start-re'
4885 ;; match. If this is greater than `cfd-continue-pos', the
4886 ;; next regexp search is started here instead.
4887 (cfd-re-match-end (point-min))
4888 ;; The end of the last `c-decl-end' found by
4889 ;; `c-find-decl-prefix-search'. `cfd-limit' if there's no
4890 ;; match. If searching for the property isn't needed then we
4891 ;; disable it by setting it to `cfd-limit' directly.
d9e94c22 4892 (cfd-prop-match (unless c-type-decl-end-used cfd-limit))
0386b551
AM
4893 ;; The end of the token preceding the decl spot last found by
4894 ;; `c-find-decl-prefix-search'. 0 for the implicit match at
4895 ;; bob. `cfd-limit' if there's no match. In other words,
4896 ;; this is the minimum of `cfd-re-match' and `cfd-prop-match'.
d9e94c22
MS
4897 (cfd-match-pos cfd-limit)
4898 ;; The position to continue searching at.
4899 cfd-continue-pos
4900 ;; The position of the last "real" token we've stopped at.
4901 ;; This can be greater than `cfd-continue-pos' when we get
4902 ;; hits inside macros or at `c-decl-end' positions inside
4903 ;; comments.
4904 (cfd-token-pos 0)
4905 ;; The end position of the last entered macro.
4906 (cfd-macro-end 0))
4907
4908 ;; Initialize by finding a syntactically relevant start position
0386b551
AM
4909 ;; before the point, and do the first `c-decl-prefix-or-start-re'
4910 ;; search unless we're at bob.
d9e94c22 4911
0386b551 4912 (let (start-in-literal start-in-macro syntactic-pos)
d9e94c22
MS
4913 ;; Must back up a bit since we look for the end of the previous
4914 ;; statement or declaration, which is earlier than the first
4915 ;; returned match.
4916
0386b551
AM
4917 (cond
4918 ;; First we need to move to a syntactically relevant position.
4919 ;; Begin by backing out of comment or string literals.
4920 ((and
4921 (when (c-got-face-at (point) c-literal-faces)
4922 ;; Try to use the faces to back up to the start of the
4923 ;; literal. FIXME: What if the point is on a declaration
4924 ;; inside a comment?
4925 (while (and (not (bobp))
4926 (c-got-face-at (1- (point)) c-literal-faces))
4927 (goto-char (previous-single-property-change
4928 (point) 'face nil (point-min))))
4929
4930 ;; XEmacs doesn't fontify the quotes surrounding string
4931 ;; literals.
4932 (and (featurep 'xemacs)
4933 (eq (get-text-property (point) 'face)
4934 'font-lock-string-face)
4935 (not (bobp))
4936 (progn (backward-char)
4937 (not (looking-at c-string-limit-regexp)))
4938 (forward-char))
4939
4940 ;; Don't trust the literal to contain only literal faces
4941 ;; (the font lock package might not have fontified the
4942 ;; start of it at all, for instance) so check that we have
4943 ;; arrived at something that looks like a start or else
4944 ;; resort to `c-literal-limits'.
4945 (unless (looking-at c-literal-start-regexp)
4946 (let ((range (c-literal-limits)))
4947 (if range (goto-char (car range)))))
4948
4949 (setq start-in-literal (point)))
4950
4951 ;; The start is in a literal. If the limit is in the same
4952 ;; one we don't have to find a syntactic position etc. We
4953 ;; only check that if the limit is at or before bonl to save
4954 ;; time; it covers the by far most common case when font-lock
4955 ;; refontifies the current line only.
4956 (<= cfd-limit (c-point 'bonl cfd-start-pos))
4957 (save-excursion
4958 (goto-char cfd-start-pos)
4959 (while (progn
4960 (goto-char (next-single-property-change
4961 (point) 'face nil cfd-limit))
4962 (and (< (point) cfd-limit)
4963 (c-got-face-at (point) c-literal-faces))))
4964 (= (point) cfd-limit)))
4965
4966 ;; Completely inside a literal. Set up variables to trig the
4967 ;; (< cfd-continue-pos cfd-start-pos) case below and it'll
4968 ;; find a suitable start position.
4969 (setq cfd-continue-pos start-in-literal))
4970
4971 ;; Check if the region might be completely inside a macro, to
4972 ;; optimize that like the completely-inside-literal above.
4973 ((save-excursion
4974 (and (= (forward-line 1) 0)
4975 (bolp) ; forward-line has funny behavior at eob.
4976 (>= (point) cfd-limit)
4977 (progn (backward-char)
4978 (eq (char-before) ?\\))))
4979 ;; (Maybe) completely inside a macro. Only need to trig the
4980 ;; (< cfd-continue-pos cfd-start-pos) case below to make it
4981 ;; set things up.
4982 (setq cfd-continue-pos (1- cfd-start-pos)
4983 start-in-macro t))
d9e94c22 4984
0386b551
AM
4985 (t
4986 ;; Back out of any macro so we don't miss any declaration
4987 ;; that could follow after it.
4988 (when (c-beginning-of-macro)
4989 (setq start-in-macro t))
4990
4991 ;; Now we're at a proper syntactically relevant position so we
4992 ;; can use the cache. But first clear it if it applied
4993 ;; further down.
4994 (c-invalidate-find-decl-cache cfd-start-pos)
4995
4996 (setq syntactic-pos (point))
4997 (unless (eq syntactic-pos c-find-decl-syntactic-pos)
4998 ;; Don't have to do this if the cache is relevant here,
4999 ;; typically if the same line is refontified again. If
5000 ;; we're just some syntactic whitespace further down we can
5001 ;; still use the cache to limit the skipping.
5002 (c-backward-syntactic-ws c-find-decl-syntactic-pos))
5003
5004 ;; If we hit `c-find-decl-syntactic-pos' and
5005 ;; `c-find-decl-match-pos' is set then we install the cached
5006 ;; values. If we hit `c-find-decl-syntactic-pos' and
5007 ;; `c-find-decl-match-pos' is nil then we know there's no decl
5008 ;; prefix in the whitespace before `c-find-decl-syntactic-pos'
5009 ;; and so we can continue the search from this point. If we
5010 ;; didn't hit `c-find-decl-syntactic-pos' then we're now in
5011 ;; the right spot to begin searching anyway.
5012 (if (and (eq (point) c-find-decl-syntactic-pos)
5013 c-find-decl-match-pos)
d9e94c22
MS
5014 (setq cfd-match-pos c-find-decl-match-pos
5015 cfd-continue-pos syntactic-pos)
0386b551
AM
5016
5017 (setq c-find-decl-syntactic-pos syntactic-pos)
5018
5019 (when (if (bobp)
5020 ;; Always consider bob a match to get the first
5021 ;; declaration in the file. Do this separately instead of
5022 ;; letting `c-decl-prefix-or-start-re' match bob, so that
5023 ;; regexp always can consume at least one character to
5024 ;; ensure that we won't get stuck in an infinite loop.
5025 (setq cfd-re-match 0)
5026 (backward-char)
5027 (c-beginning-of-current-token)
5028 (< (point) cfd-limit))
5029 ;; Do an initial search now. In the bob case above it's
5030 ;; only done to search for a `c-decl-end' spot.
5031 (c-find-decl-prefix-search))
5032
5033 (setq c-find-decl-match-pos (and (< cfd-match-pos cfd-start-pos)
5034 cfd-match-pos)))))
5035
5036 ;; Advance `cfd-continue-pos' if it's before the start position.
5037 ;; The closest continue position that might have effect at or
5038 ;; after the start depends on what we started in. This also
5039 ;; finds a suitable start position in the special cases when the
5040 ;; region is completely within a literal or macro.
5041 (when (and cfd-continue-pos (< cfd-continue-pos cfd-start-pos))
5042
5043 (cond
5044 (start-in-macro
5045 ;; If we're in a macro then it's the closest preceding token
5046 ;; in the macro. Check this before `start-in-literal',
5047 ;; since if we're inside a literal in a macro, the preceding
5048 ;; token is earlier than any `c-decl-end' spot inside the
5049 ;; literal (comment).
5050 (goto-char (or start-in-literal cfd-start-pos))
5051 ;; The only syntactic ws in macros are comments.
d9e94c22 5052 (c-backward-comments)
0386b551
AM
5053 (backward-char)
5054 (c-beginning-of-current-token))
5055
5056 (start-in-literal
5057 ;; If we're in a comment it can only be the closest
5058 ;; preceding `c-decl-end' position within that comment, if
5059 ;; any. Go back to the beginning of such a property so that
5060 ;; `c-find-decl-prefix-search' will find the end of it.
5061 ;; (Can't stop at the end and install it directly on
5062 ;; `cfd-prop-match' since that variable might be cleared
5063 ;; after `cfd-fun' below.)
5064 ;;
5065 ;; Note that if the literal is a string then the property
5066 ;; search will simply skip to the beginning of it right
5067 ;; away.
5068 (if (not c-type-decl-end-used)
5069 (goto-char start-in-literal)
5070 (goto-char cfd-start-pos)
5071 (while (progn
5072 (goto-char (previous-single-property-change
5073 (point) 'c-type nil start-in-literal))
5074 (and (> (point) start-in-literal)
5075 (not (eq (c-get-char-property (point) 'c-type)
5076 'c-decl-end))))))
5077
5078 (when (= (point) start-in-literal)
5079 ;; Didn't find any property inside the comment, so we can
5080 ;; skip it entirely. (This won't skip past a string, but
5081 ;; that'll be handled quickly by the next
5082 ;; `c-find-decl-prefix-search' anyway.)
5083 (c-forward-single-comment)
5084 (if (> (point) cfd-limit)
5085 (goto-char cfd-limit))))
d9e94c22 5086
0386b551
AM
5087 (t
5088 ;; If we started in normal code, the only match that might
5089 ;; apply before the start is what we already got in
5090 ;; `cfd-match-pos' so we can continue at the start position.
5091 ;; (Note that we don't get here if the first match is below
5092 ;; it.)
5093 (goto-char cfd-start-pos)))
5094
5095 ;; Delete found matches if they are before our new continue
5096 ;; position, so that `c-find-decl-prefix-search' won't back up
5097 ;; to them later on.
5098 (setq cfd-continue-pos (point))
5099 (when (and cfd-re-match (< cfd-re-match cfd-continue-pos))
5100 (setq cfd-re-match nil))
5101 (when (and cfd-prop-match (< cfd-prop-match cfd-continue-pos))
5102 (setq cfd-prop-match nil)))
5103
5104 (if syntactic-pos
5105 ;; This is the normal case and we got a proper syntactic
5106 ;; position. If there's a match then it's always outside
5107 ;; macros and comments, so advance to the next token and set
5108 ;; `cfd-token-pos'. The loop below will later go back using
5109 ;; `cfd-continue-pos' to fix declarations inside the
5110 ;; syntactic ws.
5111 (when (and cfd-match-pos (< cfd-match-pos syntactic-pos))
5112 (goto-char syntactic-pos)
5113 (c-forward-syntactic-ws)
5114 (and cfd-continue-pos
5115 (< cfd-continue-pos (point))
5116 (setq cfd-token-pos (point))))
5117
5118 ;; Have one of the special cases when the region is completely
5119 ;; within a literal or macro. `cfd-continue-pos' is set to a
5120 ;; good start position for the search, so do it.
5121 (c-find-decl-prefix-search)))
d9e94c22 5122
51c9af45 5123 ;; Now loop. Round what? (ACM, 2006/7/5). We already got the first match.
d9e94c22
MS
5124
5125 (while (progn
5126 (while (and
5127 (< cfd-match-pos cfd-limit)
5128
5129 (or
5130 ;; Kludge to filter out matches on the "<" that
5131 ;; aren't open parens, for the sake of languages
5132 ;; that got `c-recognize-<>-arglists' set.
5133 (and (eq (char-before cfd-match-pos) ?<)
5134 (not (c-get-char-property (1- cfd-match-pos)
5135 'syntax-table)))
5136
5137 ;; If `cfd-continue-pos' is less or equal to
5138 ;; `cfd-token-pos', we've got a hit inside a macro
5139 ;; that's in the syntactic whitespace before the last
5140 ;; "real" declaration we've checked. If they're equal
5141 ;; we've arrived at the declaration a second time, so
5142 ;; there's nothing to do.
5143 (= cfd-continue-pos cfd-token-pos)
5144
5145 (progn
5146 ;; If `cfd-continue-pos' is less than `cfd-token-pos'
5147 ;; we're still searching for declarations embedded in
5148 ;; the syntactic whitespace. In that case we need
5149 ;; only to skip comments and not macros, since they
5150 ;; can't be nested, and that's already been done in
5151 ;; `c-find-decl-prefix-search'.
5152 (when (> cfd-continue-pos cfd-token-pos)
5153 (c-forward-syntactic-ws)
5154 (setq cfd-token-pos (point)))
5155
5156 ;; Continue if the following token fails the
5157 ;; CFD-DECL-RE and CFD-FACE-CHECKLIST checks.
5158 (when (or (>= (point) cfd-limit)
5159 (not (looking-at cfd-decl-re))
5160 (and cfd-face-checklist
5161 (not (c-got-face-at
5162 (point) cfd-face-checklist))))
5163 (goto-char cfd-continue-pos)
5164 t)))
5165
5166 (< (point) cfd-limit))
5167 (c-find-decl-prefix-search))
5168
5169 (< (point) cfd-limit))
5170
0386b551
AM
5171 (when (and
5172 (>= (point) cfd-start-pos)
d9e94c22 5173
0386b551
AM
5174 (progn
5175 ;; Narrow to the end of the macro if we got a hit inside
5176 ;; one, to avoid recognizing things that start inside the
5177 ;; macro and end outside it.
5178 (when (> cfd-match-pos cfd-macro-end)
5179 ;; Not in the same macro as in the previous round.
5180 (save-excursion
5181 (goto-char cfd-match-pos)
5182 (setq cfd-macro-end
5183 (if (save-excursion (and (c-beginning-of-macro)
5184 (< (point) cfd-match-pos)))
5185 (progn (c-end-of-macro)
5186 (point))
5187 0))))
5188
5189 (if (zerop cfd-macro-end)
5190 t
5191 (if (> cfd-macro-end (point))
5192 (progn (narrow-to-region (point-min) cfd-macro-end)
5193 t)
5194 ;; The matched token was the last thing in the macro,
5195 ;; so the whole match is bogus.
5196 (setq cfd-macro-end 0)
5197 nil))))
d9e94c22
MS
5198
5199 (c-debug-put-decl-spot-faces cfd-match-pos (point))
0386b551
AM
5200 (if (funcall cfd-fun cfd-match-pos (/= cfd-macro-end 0))
5201 (setq cfd-prop-match nil))
d9e94c22
MS
5202
5203 (when (/= cfd-macro-end 0)
e1dbe924 5204 ;; Restore limits if we did macro narrowing above.
d9e94c22
MS
5205 (narrow-to-region (point-min) cfd-buffer-end)))
5206
5207 (goto-char cfd-continue-pos)
5208 (if (= cfd-continue-pos cfd-limit)
5209 (setq cfd-match-pos cfd-limit)
9657183b
AM
5210 (c-find-decl-prefix-search))))) ; Moves point, sets cfd-continue-pos,
5211 ; cfd-match-pos, etc.
d9e94c22
MS
5212
5213\f
5214;; A cache for found types.
5215
5216;; Buffer local variable that contains an obarray with the types we've
5217;; found. If a declaration is recognized somewhere we record the
5218;; fully qualified identifier in it to recognize it as a type
5219;; elsewhere in the file too. This is not accurate since we do not
5220;; bother with the scoping rules of the languages, but in practice the
5221;; same name is seldom used as both a type and something else in a
5222;; file, and we only use this as a last resort in ambiguous cases (see
0386b551
AM
5223;; `c-forward-decl-or-cast-1').
5224;;
580fba94
AM
5225;; Not every type need be in this cache. However, things which have
5226;; ceased to be types must be removed from it.
5227;;
0386b551
AM
5228;; Template types in C++ are added here too but with the template
5229;; arglist replaced with "<>" in references or "<" for the one in the
5230;; primary type. E.g. the type "Foo<A,B>::Bar<C>" is stored as
5231;; "Foo<>::Bar<". This avoids storing very long strings (since C++
5232;; template specs can be fairly sized programs in themselves) and
5233;; improves the hit ratio (it's a type regardless of the template
5234;; args; it's just not the same type, but we're only interested in
5235;; recognizing types, not telling distinct types apart). Note that
5236;; template types in references are added here too; from the example
5237;; above there will also be an entry "Foo<".
d9e94c22
MS
5238(defvar c-found-types nil)
5239(make-variable-buffer-local 'c-found-types)
5240
5241(defsubst c-clear-found-types ()
5242 ;; Clears `c-found-types'.
d9e94c22
MS
5243 (setq c-found-types (make-vector 53 0)))
5244
5245(defun c-add-type (from to)
5246 ;; Add the given region as a type in `c-found-types'. If the region
5247 ;; doesn't match an existing type but there is a type which is equal
5248 ;; to the given one except that the last character is missing, then
5249 ;; the shorter type is removed. That's done to avoid adding all
5250 ;; prefixes of a type as it's being entered and font locked. This
5251 ;; doesn't cover cases like when characters are removed from a type
5252 ;; or added in the middle. We'd need the position of point when the
5253 ;; font locking is invoked to solve this well.
0386b551
AM
5254 ;;
5255 ;; This function might do hidden buffer changes.
5256 (let ((type (c-syntactic-content from to c-recognize-<>-arglists)))
5257 (unless (intern-soft type c-found-types)
5258 (unintern (substring type 0 -1) c-found-types)
5259 (intern type c-found-types))))
d9e94c22 5260
580fba94
AM
5261(defun c-unfind-type (name)
5262 ;; Remove the "NAME" from c-found-types, if present.
5263 (unintern name c-found-types))
5264
d9e94c22
MS
5265(defsubst c-check-type (from to)
5266 ;; Return non-nil if the given region contains a type in
5267 ;; `c-found-types'.
0386b551
AM
5268 ;;
5269 ;; This function might do hidden buffer changes.
5270 (intern-soft (c-syntactic-content from to c-recognize-<>-arglists)
5271 c-found-types))
d9e94c22
MS
5272
5273(defun c-list-found-types ()
5274 ;; Return all the types in `c-found-types' as a sorted list of
5275 ;; strings.
5276 (let (type-list)
5277 (mapatoms (lambda (type)
5278 (setq type-list (cons (symbol-name type)
5279 type-list)))
5280 c-found-types)
5281 (sort type-list 'string-lessp)))
a66cd3ee 5282
2f42c75f
DN
5283;; Shut up the byte compiler.
5284(defvar c-maybe-stale-found-type)
5285
580fba94
AM
5286(defun c-trim-found-types (beg end old-len)
5287 ;; An after change function which, in conjunction with the info in
5288 ;; c-maybe-stale-found-type (set in c-before-change), removes a type
5289 ;; from `c-found-types', should this type have become stale. For
5290 ;; example, this happens to "foo" when "foo \n bar();" becomes
5291 ;; "foo(); \n bar();". Such stale types, if not removed, foul up
5292 ;; the fontification.
b414f371 5293 ;;
580fba94
AM
5294 ;; Have we, perhaps, added non-ws characters to the front/back of a found
5295 ;; type?
5296 (when (> end beg)
5297 (save-excursion
5298 (when (< end (point-max))
5299 (goto-char end)
5300 (if (and (c-beginning-of-current-token) ; only moves when we started in the middle
5301 (progn (goto-char end)
5302 (c-end-of-current-token)))
5303 (c-unfind-type (buffer-substring-no-properties
5304 end (point)))))
5305 (when (> beg (point-min))
5306 (goto-char beg)
5307 (if (and (c-end-of-current-token) ; only moves when we started in the middle
5308 (progn (goto-char beg)
5309 (c-beginning-of-current-token)))
5310 (c-unfind-type (buffer-substring-no-properties
5311 (point) beg))))))
b414f371 5312
580fba94
AM
5313 (if c-maybe-stale-found-type ; e.g. (c-decl-id-start "foo" 97 107 " (* ooka) " "o")
5314 (cond
5315 ;; Changing the amount of (already existing) whitespace - don't do anything.
5316 ((and (c-partial-ws-p beg end)
5317 (or (= beg end) ; removal of WS
580fba94
AM
5318 (string-match "^[ \t\n\r\f\v]*$" (nth 5 c-maybe-stale-found-type)))))
5319
5320 ;; The syntactic relationship which defined a "found type" has been
5321 ;; destroyed.
5322 ((eq (car c-maybe-stale-found-type) 'c-decl-id-start)
5323 (c-unfind-type (cadr c-maybe-stale-found-type)))
5324;; ((eq (car c-maybe-stale-found-type) 'c-decl-type-start) FIXME!!!
5325 )))
5326
d9e94c22 5327\f
dd969a56
AM
5328;; Setting and removing syntax properties on < and > in languages (C++
5329;; and Java) where they can be template/generic delimiters as well as
5330;; their normal meaning of "less/greater than".
5331
5332;; Normally, < and > have syntax 'punctuation'. When they are found to
5333;; be delimiters, they are marked as such with the category properties
5334;; c-<-as-paren-syntax, c->-as-paren-syntax respectively.
5335
5336;; STRATEGY:
5337;;
5338;; It is impossible to determine with certainty whether a <..> pair in
5339;; C++ is two comparison operators or is template delimiters, unless
5340;; one duplicates a lot of a C++ compiler. For example, the following
5341;; code fragment:
5342;;
5343;; foo (a < b, c > d) ;
5344;;
5345;; could be a function call with two integer parameters (each a
5346;; relational expression), or it could be a constructor for class foo
5347;; taking one parameter d of templated type "a < b, c >". They are
5348;; somewhat easier to distinguish in Java.
5349;;
5350;; The strategy now (2010-01) adopted is to mark and unmark < and
5351;; > IN MATCHING PAIRS ONLY. [Previously, they were marked
5352;; individually when their context so indicated. This gave rise to
53964682 5353;; intractable problems when one of a matching pair was deleted, or
dd969a56
AM
5354;; pulled into a literal.]
5355;;
5356;; At each buffer change, the syntax-table properties are removed in a
5357;; before-change function and reapplied, when needed, in an
5358;; after-change function. It is far more important that the
5359;; properties get removed when they they are spurious than that they
5360;; be present when wanted.
5361;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
5362(defun c-clear-<-pair-props (&optional pos)
5363 ;; POS (default point) is at a < character. If it is marked with
5364 ;; open paren syntax-table text property, remove the property,
5365 ;; together with the close paren property on the matching > (if
5366 ;; any).
5367 (save-excursion
5368 (if pos
5369 (goto-char pos)
5370 (setq pos (point)))
5371 (when (equal (c-get-char-property (point) 'syntax-table)
5372 c-<-as-paren-syntax)
5373 (with-syntax-table c-no-parens-syntax-table ; ignore unbalanced [,{,(,..
5374 (c-go-list-forward))
5375 (when (equal (c-get-char-property (1- (point)) 'syntax-table)
5376 c->-as-paren-syntax) ; should always be true.
8a249abc
AM
5377 (c-clear-char-property (1- (point)) 'category))
5378 (c-clear-char-property pos 'category))))
dd969a56
AM
5379
5380(defun c-clear->-pair-props (&optional pos)
5381 ;; POS (default point) is at a > character. If it is marked with
5382 ;; close paren syntax-table property, remove the property, together
5383 ;; with the open paren property on the matching < (if any).
5384 (save-excursion
5385 (if pos
5386 (goto-char pos)
5387 (setq pos (point)))
5388 (when (equal (c-get-char-property (point) 'syntax-table)
5389 c->-as-paren-syntax)
5390 (with-syntax-table c-no-parens-syntax-table ; ignore unbalanced [,{,(,..
5391 (c-go-up-list-backward))
5392 (when (equal (c-get-char-property (point) 'syntax-table)
5393 c-<-as-paren-syntax) ; should always be true.
8a249abc
AM
5394 (c-clear-char-property (point) 'category))
5395 (c-clear-char-property pos 'category))))
dd969a56
AM
5396
5397(defun c-clear-<>-pair-props (&optional pos)
5398 ;; POS (default point) is at a < or > character. If it has an
5399 ;; open/close paren syntax-table property, remove this property both
5400 ;; from the current character and its partner (which will also be
5401 ;; thusly marked).
5402 (cond
5403 ((eq (char-after) ?\<)
5404 (c-clear-<-pair-props pos))
5405 ((eq (char-after) ?\>)
5406 (c-clear->-pair-props pos))
5407 (t (c-benign-error
5408 "c-clear-<>-pair-props called from wrong position"))))
5409
5410(defun c-clear-<-pair-props-if-match-after (lim &optional pos)
5411 ;; POS (default point) is at a < character. If it is both marked
5412 ;; with open/close paren syntax-table property, and has a matching >
5413 ;; (also marked) which is after LIM, remove the property both from
43a91810
AM
5414 ;; the current > and its partner. Return t when this happens, nil
5415 ;; when it doesn't.
dd969a56
AM
5416 (save-excursion
5417 (if pos
5418 (goto-char pos)
5419 (setq pos (point)))
5420 (when (equal (c-get-char-property (point) 'syntax-table)
5421 c-<-as-paren-syntax)
5422 (with-syntax-table c-no-parens-syntax-table ; ignore unbalanced [,{,(,..
5423 (c-go-list-forward))
5424 (when (and (>= (point) lim)
5425 (equal (c-get-char-property (1- (point)) 'syntax-table)
5426 c->-as-paren-syntax)) ; should always be true.
5427 (c-unmark-<->-as-paren (1- (point)))
43a91810
AM
5428 (c-unmark-<->-as-paren pos))
5429 t)))
dd969a56
AM
5430
5431(defun c-clear->-pair-props-if-match-before (lim &optional pos)
5432 ;; POS (default point) is at a > character. If it is both marked
5433 ;; with open/close paren syntax-table property, and has a matching <
5434 ;; (also marked) which is before LIM, remove the property both from
43a91810
AM
5435 ;; the current < and its partner. Return t when this happens, nil
5436 ;; when it doesn't.
dd969a56
AM
5437 (save-excursion
5438 (if pos
5439 (goto-char pos)
5440 (setq pos (point)))
5441 (when (equal (c-get-char-property (point) 'syntax-table)
5442 c->-as-paren-syntax)
5443 (with-syntax-table c-no-parens-syntax-table ; ignore unbalanced [,{,(,..
5444 (c-go-up-list-backward))
5445 (when (and (<= (point) lim)
5446 (equal (c-get-char-property (point) 'syntax-table)
5447 c-<-as-paren-syntax)) ; should always be true.
5448 (c-unmark-<->-as-paren (point))
43a91810
AM
5449 (c-unmark-<->-as-paren pos))
5450 t)))
dd969a56 5451
f3b554af
GM
5452;; Set by c-common-init in cc-mode.el.
5453(defvar c-new-BEG)
5454(defvar c-new-END)
5455
dd969a56
AM
5456(defun c-before-change-check-<>-operators (beg end)
5457 ;; Unmark certain pairs of "< .... >" which are currently marked as
5458 ;; template/generic delimiters. (This marking is via syntax-table
5459 ;; text properties).
5460 ;;
5461 ;; These pairs are those which are in the current "statement" (i.e.,
5462 ;; the region between the {, }, or ; before BEG and the one after
5463 ;; END), and which enclose any part of the interval (BEG END).
5464 ;;
5465 ;; Note that in C++ (?and Java), template/generic parens cannot
5466 ;; enclose a brace or semicolon, so we use these as bounds on the
5467 ;; region we must work on.
5468 ;;
5469 ;; This function is called from before-change-functions (via
5470 ;; c-get-state-before-change-functions). Thus the buffer is widened,
5471 ;; and point is undefined, both at entry and exit.
5472 ;;
5473 ;; FIXME!!! This routine ignores the possibility of macros entirely.
5474 ;; 2010-01-29.
5475 (save-excursion
5476 (let ((beg-lit-limits (progn (goto-char beg) (c-literal-limits)))
43a91810
AM
5477 (end-lit-limits (progn (goto-char end) (c-literal-limits)))
5478 new-beg new-end need-new-beg need-new-end)
dd969a56
AM
5479 ;; Locate the barrier before the changed region
5480 (goto-char (if beg-lit-limits (car beg-lit-limits) beg))
0de3da9f 5481 (c-syntactic-skip-backward "^;{}" (c-determine-limit 512))
43a91810 5482 (setq new-beg (point))
dd969a56
AM
5483
5484 ;; Remove the syntax-table properties from each pertinent <...> pair.
5485 ;; Firsly, the ones with the < before beg and > after beg.
5486 (while (c-search-forward-char-property 'category 'c-<-as-paren-syntax beg)
43a91810
AM
5487 (if (c-clear-<-pair-props-if-match-after beg (1- (point)))
5488 (setq need-new-beg t)))
dd969a56
AM
5489
5490 ;; Locate the barrier after END.
5491 (goto-char (if end-lit-limits (cdr end-lit-limits) end))
0de3da9f 5492 (c-syntactic-re-search-forward "[;{}]" (c-determine-+ve-limit 512) 'end)
43a91810 5493 (setq new-end (point))
dd969a56
AM
5494
5495 ;; Remove syntax-table properties from the remaining pertinent <...>
5496 ;; pairs, those with a > after end and < before end.
5497 (while (c-search-backward-char-property 'category 'c->-as-paren-syntax end)
43a91810
AM
5498 (if (c-clear->-pair-props-if-match-before end)
5499 (setq need-new-end t)))
5500
5501 ;; Extend the fontification region, if needed.
5502 (when need-new-beg
5503 (goto-char new-beg)
5504 (c-forward-syntactic-ws)
5505 (and (< (point) c-new-BEG) (setq c-new-BEG (point))))
5506
5507 (when need-new-end
5508 (and (> new-end c-new-END) (setq c-new-END new-end))))))
dd969a56
AM
5509
5510
d9e94c22 5511
0386b551
AM
5512(defun c-after-change-check-<>-operators (beg end)
5513 ;; This is called from `after-change-functions' when
5514 ;; c-recognize-<>-arglists' is set. It ensures that no "<" or ">"
5515 ;; chars with paren syntax become part of another operator like "<<"
5516 ;; or ">=".
5517 ;;
5518 ;; This function might do hidden buffer changes.
5519
cb694ab7
AM
5520 (save-excursion
5521 (goto-char beg)
5522 (when (or (looking-at "[<>]")
5523 (< (skip-chars-backward "<>") 0))
5524
0386b551 5525 (goto-char beg)
cb694ab7
AM
5526 (c-beginning-of-current-token)
5527 (when (and (< (point) beg)
5528 (looking-at c-<>-multichar-token-regexp)
5529 (< beg (setq beg (match-end 0))))
5530 (while (progn (skip-chars-forward "^<>" beg)
5531 (< (point) beg))
dd969a56 5532 (c-clear-<>-pair-props)
cb694ab7
AM
5533 (forward-char))))
5534
5535 (when (< beg end)
5536 (goto-char end)
0386b551
AM
5537 (when (or (looking-at "[<>]")
5538 (< (skip-chars-backward "<>") 0))
5539
cb694ab7 5540 (goto-char end)
0386b551 5541 (c-beginning-of-current-token)
cb694ab7 5542 (when (and (< (point) end)
0386b551 5543 (looking-at c-<>-multichar-token-regexp)
cb694ab7
AM
5544 (< end (setq end (match-end 0))))
5545 (while (progn (skip-chars-forward "^<>" end)
5546 (< (point) end))
dd969a56 5547 (c-clear-<>-pair-props)
cb694ab7 5548 (forward-char)))))))
d9e94c22 5549
dd969a56
AM
5550
5551\f
5552;; Handling of small scale constructs like types and names.
5553
d9e94c22
MS
5554;; Dynamically bound variable that instructs `c-forward-type' to also
5555;; treat possible types (i.e. those that it normally returns 'maybe or
5556;; 'found for) as actual types (and always return 'found for them).
5557;; This means that it records them in `c-record-type-identifiers' if
5558;; that is set, and that it adds them to `c-found-types'.
5559(defvar c-promote-possible-types nil)
5560
0386b551
AM
5561;; Dynamically bound variable that instructs `c-forward-<>-arglist' to
5562;; mark up successfully parsed arglists with paren syntax properties on
5563;; the surrounding angle brackets and with `c-<>-arg-sep' in the
5564;; `c-type' property of each argument separating comma.
5565;;
5566;; Setting this variable also makes `c-forward-<>-arglist' recurse into
5567;; all arglists for side effects (i.e. recording types), otherwise it
5568;; exploits any existing paren syntax properties to quickly jump to the
5569;; end of already parsed arglists.
5570;;
5571;; Marking up the arglists is not the default since doing that correctly
5572;; depends on a proper value for `c-restricted-<>-arglists'.
5573(defvar c-parse-and-markup-<>-arglists nil)
5574
d9e94c22 5575;; Dynamically bound variable that instructs `c-forward-<>-arglist' to
037558bf
MS
5576;; not accept arglists that contain binary operators.
5577;;
5578;; This is primarily used to handle C++ template arglists. C++
5579;; disambiguates them by checking whether the preceding name is a
5580;; template or not. We can't do that, so we assume it is a template
5581;; if it can be parsed as one. That usually works well since
5582;; comparison expressions on the forms "a < b > c" or "a < b, c > d"
5583;; in almost all cases would be pointless.
5584;;
5585;; However, in function arglists, e.g. in "foo (a < b, c > d)", we
5586;; should let the comma separate the function arguments instead. And
5587;; in a context where the value of the expression is taken, e.g. in
5588;; "if (a < b || c > d)", it's probably not a template.
5589(defvar c-restricted-<>-arglists nil)
d9e94c22 5590
0386b551
AM
5591;; Dynamically bound variables that instructs
5592;; `c-forward-keyword-clause', `c-forward-<>-arglist',
5593;; `c-forward-name', `c-forward-type', `c-forward-decl-or-cast-1', and
5594;; `c-forward-label' to record the ranges of all the type and
5595;; reference identifiers they encounter. They will build lists on
5596;; these variables where each element is a cons of the buffer
5597;; positions surrounding each identifier. This recording is only
5598;; activated when `c-record-type-identifiers' is non-nil.
d9e94c22
MS
5599;;
5600;; All known types that can't be identifiers are recorded, and also
5601;; other possible types if `c-promote-possible-types' is set.
5602;; Recording is however disabled inside angle bracket arglists that
5603;; are encountered inside names and other angle bracket arglists.
0386b551 5604;; Such occurrences are taken care of by `c-font-lock-<>-arglists'
d9e94c22
MS
5605;; instead.
5606;;
5607;; Only the names in C++ template style references (e.g. "tmpl" in
5608;; "tmpl<a,b>::foo") are recorded as references, other references
5609;; aren't handled here.
0386b551
AM
5610;;
5611;; `c-forward-label' records the label identifier(s) on
5612;; `c-record-ref-identifiers'.
d9e94c22
MS
5613(defvar c-record-type-identifiers nil)
5614(defvar c-record-ref-identifiers nil)
5615
0386b551
AM
5616;; This variable will receive a cons cell of the range of the last
5617;; single identifier symbol stepped over by `c-forward-name' if it's
5618;; successful. This is the range that should be put on one of the
5619;; record lists above by the caller. It's assigned nil if there's no
5620;; such symbol in the name.
d9e94c22
MS
5621(defvar c-last-identifier-range nil)
5622
5623(defmacro c-record-type-id (range)
5624 (if (eq (car-safe range) 'cons)
5625 ;; Always true.
5626 `(setq c-record-type-identifiers
5627 (cons ,range c-record-type-identifiers))
5628 `(let ((range ,range))
5629 (if range
5630 (setq c-record-type-identifiers
5631 (cons range c-record-type-identifiers))))))
5632
5633(defmacro c-record-ref-id (range)
5634 (if (eq (car-safe range) 'cons)
5635 ;; Always true.
5636 `(setq c-record-ref-identifiers
5637 (cons ,range c-record-ref-identifiers))
5638 `(let ((range ,range))
5639 (if range
5640 (setq c-record-ref-identifiers
5641 (cons range c-record-ref-identifiers))))))
5642
5643;; Dynamically bound variable that instructs `c-forward-type' to
5644;; record the ranges of types that only are found. Behaves otherwise
5645;; like `c-record-type-identifiers'.
5646(defvar c-record-found-types nil)
5647
5648(defmacro c-forward-keyword-prefixed-id (type)
5649 ;; Used internally in `c-forward-keyword-clause' to move forward
5650 ;; over a type (if TYPE is 'type) or a name (otherwise) which
5651 ;; possibly is prefixed by keywords and their associated clauses.
5652 ;; Try with a type/name first to not trip up on those that begin
5653 ;; with a keyword. Return t if a known or found type is moved
5654 ;; over. The point is clobbered if nil is returned. If range
5655 ;; recording is enabled, the identifier is recorded on as a type
5656 ;; if TYPE is 'type or as a reference if TYPE is 'ref.
0386b551
AM
5657 ;;
5658 ;; This macro might do hidden buffer changes.
d9e94c22
MS
5659 `(let (res)
5660 (while (if (setq res ,(if (eq type 'type)
5661 `(c-forward-type)
5662 `(c-forward-name)))
5663 nil
5664 (and (looking-at c-keywords-regexp)
0386b551 5665 (c-forward-keyword-clause 1))))
d9e94c22
MS
5666 (when (memq res '(t known found prefix))
5667 ,(when (eq type 'ref)
5668 `(when c-record-type-identifiers
5669 (c-record-ref-id c-last-identifier-range)))
5670 t)))
5671
0386b551 5672(defmacro c-forward-id-comma-list (type update-safe-pos)
d9e94c22
MS
5673 ;; Used internally in `c-forward-keyword-clause' to move forward
5674 ;; over a comma separated list of types or names using
5675 ;; `c-forward-keyword-prefixed-id'.
0386b551
AM
5676 ;;
5677 ;; This macro might do hidden buffer changes.
d9e94c22 5678 `(while (and (progn
0386b551
AM
5679 ,(when update-safe-pos
5680 `(setq safe-pos (point)))
d9e94c22
MS
5681 (eq (char-after) ?,))
5682 (progn
5683 (forward-char)
5684 (c-forward-syntactic-ws)
5685 (c-forward-keyword-prefixed-id ,type)))))
5686
0386b551
AM
5687(defun c-forward-keyword-clause (match)
5688 ;; Submatch MATCH in the current match data is assumed to surround a
5689 ;; token. If it's a keyword, move over it and any immediately
5690 ;; following clauses associated with it, stopping at the start of
5691 ;; the next token. t is returned in that case, otherwise the point
d9e94c22
MS
5692 ;; stays and nil is returned. The kind of clauses that are
5693 ;; recognized are those specified by `c-type-list-kwds',
5694 ;; `c-ref-list-kwds', `c-colon-type-list-kwds',
5695 ;; `c-paren-nontype-kwds', `c-paren-type-kwds', `c-<>-type-kwds',
5696 ;; and `c-<>-arglist-kwds'.
0386b551
AM
5697 ;;
5698 ;; This function records identifier ranges on
5699 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
5700 ;; `c-record-type-identifiers' is non-nil.
5701 ;;
5702 ;; Note that for `c-colon-type-list-kwds', which doesn't necessary
5703 ;; apply directly after the keyword, the type list is moved over
5704 ;; only when there is no unaccounted token before it (i.e. a token
5705 ;; that isn't moved over due to some other keyword list). The
5706 ;; identifier ranges in the list are still recorded if that should
5707 ;; be done, though.
5708 ;;
5709 ;; This function might do hidden buffer changes.
5710
5711 (let ((kwd-sym (c-keyword-sym (match-string match))) safe-pos pos
5712 ;; The call to `c-forward-<>-arglist' below is made after
5713 ;; `c-<>-sexp-kwds' keywords, so we're certain they actually
5714 ;; are angle bracket arglists and `c-restricted-<>-arglists'
5715 ;; should therefore be nil.
5716 (c-parse-and-markup-<>-arglists t)
5717 c-restricted-<>-arglists)
d9e94c22 5718
d9e94c22 5719 (when kwd-sym
0386b551 5720 (goto-char (match-end match))
d9e94c22
MS
5721 (c-forward-syntactic-ws)
5722 (setq safe-pos (point))
a66cd3ee 5723
d9e94c22
MS
5724 (cond
5725 ((and (c-keyword-member kwd-sym 'c-type-list-kwds)
5726 (c-forward-keyword-prefixed-id type))
5727 ;; There's a type directly after a keyword in `c-type-list-kwds'.
0386b551 5728 (c-forward-id-comma-list type t))
d9e94c22
MS
5729
5730 ((and (c-keyword-member kwd-sym 'c-ref-list-kwds)
5731 (c-forward-keyword-prefixed-id ref))
5732 ;; There's a name directly after a keyword in `c-ref-list-kwds'.
0386b551 5733 (c-forward-id-comma-list ref t))
d9e94c22
MS
5734
5735 ((and (c-keyword-member kwd-sym 'c-paren-any-kwds)
5736 (eq (char-after) ?\())
5737 ;; There's an open paren after a keyword in `c-paren-any-kwds'.
5738
5739 (forward-char)
5740 (when (and (setq pos (c-up-list-forward))
5741 (eq (char-before pos) ?\)))
5742 (when (and c-record-type-identifiers
5743 (c-keyword-member kwd-sym 'c-paren-type-kwds))
5744 ;; Use `c-forward-type' on every identifier we can find
5745 ;; inside the paren, to record the types.
5746 (while (c-syntactic-re-search-forward c-symbol-start pos t)
5747 (goto-char (match-beginning 0))
5748 (unless (c-forward-type)
5749 (looking-at c-symbol-key) ; Always matches.
5750 (goto-char (match-end 0)))))
5751
5752 (goto-char pos)
5753 (c-forward-syntactic-ws)
5754 (setq safe-pos (point))))
5755
5756 ((and (c-keyword-member kwd-sym 'c-<>-sexp-kwds)
5757 (eq (char-after) ?<)
0386b551 5758 (c-forward-<>-arglist (c-keyword-member kwd-sym 'c-<>-type-kwds)))
d9e94c22
MS
5759 (c-forward-syntactic-ws)
5760 (setq safe-pos (point)))
5761
5762 ((and (c-keyword-member kwd-sym 'c-nonsymbol-sexp-kwds)
449a2b0d
MS
5763 (not (looking-at c-symbol-start))
5764 (c-safe (c-forward-sexp) t))
d9e94c22
MS
5765 (c-forward-syntactic-ws)
5766 (setq safe-pos (point))))
5767
0386b551
AM
5768 (when (c-keyword-member kwd-sym 'c-colon-type-list-kwds)
5769 (if (eq (char-after) ?:)
5770 ;; If we are at the colon already, we move over the type
5771 ;; list after it.
5772 (progn
5773 (forward-char)
5774 (c-forward-syntactic-ws)
5775 (when (c-forward-keyword-prefixed-id type)
5776 (c-forward-id-comma-list type t)))
5777 ;; Not at the colon, so stop here. But the identifier
5778 ;; ranges in the type list later on should still be
5779 ;; recorded.
5780 (and c-record-type-identifiers
5781 (progn
5782 ;; If a keyword matched both one of the types above and
5783 ;; this one, we match `c-colon-type-list-re' after the
5784 ;; clause matched above.
5785 (goto-char safe-pos)
5786 (looking-at c-colon-type-list-re))
5787 (progn
5788 (goto-char (match-end 0))
5789 (c-forward-syntactic-ws)
5790 (c-forward-keyword-prefixed-id type))
5791 ;; There's a type after the `c-colon-type-list-re' match
5792 ;; after a keyword in `c-colon-type-list-kwds'.
5793 (c-forward-id-comma-list type nil))))
d9e94c22
MS
5794
5795 (goto-char safe-pos)
5796 t)))
5797
f3b554af
GM
5798;; cc-mode requires cc-fonts.
5799(declare-function c-fontify-recorded-types-and-refs "cc-fonts" ())
5800
0386b551
AM
5801(defun c-forward-<>-arglist (all-types)
5802 ;; The point is assumed to be at a "<". Try to treat it as the open
b4dc7d98 5803 ;; paren of an angle bracket arglist and move forward to the
0386b551
AM
5804 ;; corresponding ">". If successful, the point is left after the
5805 ;; ">" and t is returned, otherwise the point isn't moved and nil is
d9e94c22
MS
5806 ;; returned. If ALL-TYPES is t then all encountered arguments in
5807 ;; the arglist that might be types are treated as found types.
5808 ;;
0386b551
AM
5809 ;; The variable `c-parse-and-markup-<>-arglists' controls how this
5810 ;; function handles text properties on the angle brackets and argument
5811 ;; separating commas.
d9e94c22 5812 ;;
0386b551
AM
5813 ;; `c-restricted-<>-arglists' controls how lenient the template
5814 ;; arglist recognition should be.
5815 ;;
5816 ;; This function records identifier ranges on
5817 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
5818 ;; `c-record-type-identifiers' is non-nil.
5819 ;;
5820 ;; This function might do hidden buffer changes.
d9e94c22
MS
5821
5822 (let ((start (point))
5823 ;; If `c-record-type-identifiers' is set then activate
5824 ;; recording of any found types that constitute an argument in
5825 ;; the arglist.
abfc152b 5826 (c-record-found-types (if c-record-type-identifiers t)))
d9e94c22
MS
5827 (if (catch 'angle-bracket-arglist-escape
5828 (setq c-record-found-types
0386b551 5829 (c-forward-<>-arglist-recur all-types)))
d9e94c22
MS
5830 (progn
5831 (when (consp c-record-found-types)
5832 (setq c-record-type-identifiers
5833 ;; `nconc' doesn't mind that the tail of
5834 ;; `c-record-found-types' is t.
5835 (nconc c-record-found-types c-record-type-identifiers)))
452ea855 5836 (if (c-major-mode-is 'java-mode) (c-fontify-recorded-types-and-refs))
d9e94c22
MS
5837 t)
5838
5839 (goto-char start)
a66cd3ee 5840 nil)))
785eecbb 5841
0386b551 5842(defun c-forward-<>-arglist-recur (all-types)
d9e94c22 5843 ;; Recursive part of `c-forward-<>-arglist'.
0386b551
AM
5844 ;;
5845 ;; This function might do hidden buffer changes.
d9e94c22
MS
5846
5847 (let ((start (point)) res pos tmp
5848 ;; Cover this so that any recorded found type ranges are
5849 ;; automatically lost if it turns out to not be an angle
5850 ;; bracket arglist. It's propagated through the return value
5851 ;; on successful completion.
5852 (c-record-found-types c-record-found-types)
5853 ;; List that collects the positions after the argument
5854 ;; separating ',' in the arglist.
5855 arg-start-pos)
0386b551
AM
5856 ;; If the '<' has paren open syntax then we've marked it as an angle
5857 ;; bracket arglist before, so skip to the end.
5858 (if (and (not c-parse-and-markup-<>-arglists)
5859 (c-get-char-property (point) 'syntax-table))
5860
5861 (progn
5862 (forward-char)
5863 (if (and (c-go-up-list-forward)
5864 (eq (char-before) ?>))
5865 t
0386b551
AM
5866 ;; Got unmatched paren angle brackets. We don't clear the paren
5867 ;; syntax properties and retry, on the basis that it's very
5868 ;; unlikely that paren angle brackets become operators by code
5869 ;; manipulation. It's far more likely that it doesn't match due
5870 ;; to narrowing or some temporary change.
5871 (goto-char start)
5872 nil))
d9e94c22 5873
abfc152b 5874 (forward-char) ; Forward over the opening '<'.
452ea855 5875
d9e94c22 5876 (unless (looking-at c-<-op-cont-regexp)
abfc152b
AM
5877 ;; go forward one non-alphanumeric character (group) per iteration of
5878 ;; this loop.
22c3ce97 5879 (while (and
d9e94c22 5880 (progn
22c3ce97
AM
5881 (c-forward-syntactic-ws)
5882 (let ((orig-record-found-types c-record-found-types))
5883 (when (or (and c-record-type-identifiers all-types)
5884 (c-major-mode-is 'java-mode))
5885 ;; All encountered identifiers are types, so set the
5886 ;; promote flag and parse the type.
5887 (progn
5888 (c-forward-syntactic-ws)
5889 (if (looking-at "\\?")
5890 (forward-char)
5891 (when (looking-at c-identifier-start)
5892 (let ((c-promote-possible-types t)
5893 (c-record-found-types t))
5894 (c-forward-type))))
5895
5896 (c-forward-syntactic-ws)
5897
5898 (when (or (looking-at "extends")
5899 (looking-at "super"))
5900 (forward-word)
5901 (c-forward-syntactic-ws)
5902 (let ((c-promote-possible-types t)
5903 (c-record-found-types t))
5904 (c-forward-type)
5905 (c-forward-syntactic-ws))))))
5906
abfc152b 5907 (setq pos (point)) ; e.g. first token inside the '<'
22c3ce97 5908
a4ee83cc
AM
5909 ;; Note: These regexps exploit the match order in \| so
5910 ;; that "<>" is matched by "<" rather than "[^>:-]>".
5911 (c-syntactic-re-search-forward
5912 ;; Stop on ',', '|', '&', '+' and '-' to catch
5913 ;; common binary operators that could be between
5914 ;; two comparison expressions "a<b" and "c>d".
5915 "[<;{},|+&-]\\|[>)]"
5916 nil t t))
22c3ce97
AM
5917
5918 (cond
5919 ((eq (char-before) ?>)
d9e94c22
MS
5920 ;; Either an operator starting with '>' or the end of
5921 ;; the angle bracket arglist.
5922
0386b551 5923 (if (looking-at c->-op-cont-regexp)
d9e94c22 5924 (progn
0386b551
AM
5925 (goto-char (match-end 0))
5926 t) ; Continue the loop.
d9e94c22 5927
0386b551
AM
5928 ;; The angle bracket arglist is finished.
5929 (when c-parse-and-markup-<>-arglists
d9e94c22 5930 (while arg-start-pos
0386b551
AM
5931 (c-put-c-type-property (1- (car arg-start-pos))
5932 'c-<>-arg-sep)
d9e94c22
MS
5933 (setq arg-start-pos (cdr arg-start-pos)))
5934 (c-mark-<-as-paren start)
0386b551
AM
5935 (c-mark->-as-paren (1- (point))))
5936 (setq res t)
5937 nil)) ; Exit the loop.
d9e94c22
MS
5938
5939 ((eq (char-before) ?<)
5940 ;; Either an operator starting with '<' or a nested arglist.
d9e94c22
MS
5941 (setq pos (point))
5942 (let (id-start id-end subres keyword-match)
abfc152b
AM
5943 (cond
5944 ;; The '<' begins a multi-char operator.
5945 ((looking-at c-<-op-cont-regexp)
5946 (setq tmp (match-end 0))
5947 (goto-char (match-end 0)))
5948 ;; We're at a nested <.....>
5949 ((progn
5950 (setq tmp pos)
5951 (backward-char) ; to the '<'
5952 (and
5953 (save-excursion
5954 ;; There's always an identifier before an angle
5955 ;; bracket arglist, or a keyword in `c-<>-type-kwds'
5956 ;; or `c-<>-arglist-kwds'.
5957 (c-backward-syntactic-ws)
5958 (setq id-end (point))
5959 (c-simple-skip-symbol-backward)
5960 (when (or (setq keyword-match
5961 (looking-at c-opt-<>-sexp-key))
5962 (not (looking-at c-keywords-regexp)))
5963 (setq id-start (point))))
5964 (setq subres
5965 (let ((c-promote-possible-types t)
5966 (c-record-found-types t))
5967 (c-forward-<>-arglist-recur
5968 (and keyword-match
5969 (c-keyword-member
5970 (c-keyword-sym (match-string 1))
5971 'c-<>-type-kwds)))))))
d9e94c22
MS
5972
5973 ;; It was an angle bracket arglist.
5974 (setq c-record-found-types subres)
5975
5976 ;; Record the identifier before the template as a type
5977 ;; or reference depending on whether the arglist is last
5978 ;; in a qualified identifier.
5979 (when (and c-record-type-identifiers
5980 (not keyword-match))
5981 (if (and c-opt-identifier-concat-key
5982 (progn
5983 (c-forward-syntactic-ws)
5984 (looking-at c-opt-identifier-concat-key)))
5985 (c-record-ref-id (cons id-start id-end))
abfc152b
AM
5986 (c-record-type-id (cons id-start id-end)))))
5987
5988 ;; At a "less than" operator.
5989 (t
5990 (forward-char)
5991 )))
5992 t) ; carry on looping.
22c3ce97
AM
5993
5994 ((and (not c-restricted-<>-arglists)
5995 (or (and (eq (char-before) ?&)
5996 (not (eq (char-after) ?&)))
5997 (eq (char-before) ?,)))
5998 ;; Just another argument. Record the position. The
5999 ;; type check stuff that made us stop at it is at
6000 ;; the top of the loop.
d9e94c22
MS
6001 (setq arg-start-pos (cons (point) arg-start-pos)))
6002
6003 (t
6004 ;; Got a character that can't be in an angle bracket
6005 ;; arglist argument. Abort using `throw', since
6006 ;; it's useless to try to find a surrounding arglist
6007 ;; if we're nested.
6008 (throw 'angle-bracket-arglist-escape nil))))))
d9e94c22
MS
6009 (if res
6010 (or c-record-found-types t)))))
6011
0386b551
AM
6012(defun c-backward-<>-arglist (all-types &optional limit)
6013 ;; The point is assumed to be directly after a ">". Try to treat it
6014 ;; as the close paren of an angle bracket arglist and move back to
6015 ;; the corresponding "<". If successful, the point is left at
6016 ;; the "<" and t is returned, otherwise the point isn't moved and
6017 ;; nil is returned. ALL-TYPES is passed on to
6018 ;; `c-forward-<>-arglist'.
6019 ;;
6020 ;; If the optional LIMIT is given, it bounds the backward search.
6021 ;; It's then assumed to be at a syntactically relevant position.
6022 ;;
6023 ;; This is a wrapper around `c-forward-<>-arglist'. See that
6024 ;; function for more details.
6025
6026 (let ((start (point)))
6027 (backward-char)
6028 (if (and (not c-parse-and-markup-<>-arglists)
6029 (c-get-char-property (point) 'syntax-table))
6030
6031 (if (and (c-go-up-list-backward)
6032 (eq (char-after) ?<))
6033 t
6034 ;; See corresponding note in `c-forward-<>-arglist'.
6035 (goto-char start)
6036 nil)
6037
51c9af45 6038 (while (progn
0386b551
AM
6039 (c-syntactic-skip-backward "^<;{}" limit t)
6040
51c9af45
AM
6041 (and
6042 (if (eq (char-before) ?<)
6043 t
6044 ;; Stopped at bob or a char that isn't allowed in an
6045 ;; arglist, so we've failed.
6046 (goto-char start)
6047 nil)
0386b551 6048
51c9af45
AM
6049 (if (> (point)
6050 (progn (c-beginning-of-current-token)
6051 (point)))
6052 ;; If we moved then the "<" was part of some
6053 ;; multicharacter token.
6054 t
0386b551 6055
51c9af45
AM
6056 (backward-char)
6057 (let ((beg-pos (point)))
6058 (if (c-forward-<>-arglist all-types)
6059 (cond ((= (point) start)
6060 ;; Matched the arglist. Break the while.
6061 (goto-char beg-pos)
6062 nil)
6063 ((> (point) start)
6064 ;; We started from a non-paren ">" inside an
6065 ;; arglist.
6066 (goto-char start)
6067 nil)
6068 (t
6069 ;; Matched a shorter arglist. Can be a nested
6070 ;; one so continue looking.
6071 (goto-char beg-pos)
6072 t))
6073 t))))))
0386b551
AM
6074
6075 (/= (point) start))))
6076
d9e94c22
MS
6077(defun c-forward-name ()
6078 ;; Move forward over a complete name if at the beginning of one,
e15f8aaa
AM
6079 ;; stopping at the next following token. A keyword, as such,
6080 ;; doesn't count as a name. If the point is not at something that
6081 ;; is recognized as a name then it stays put.
6082 ;;
6083 ;; A name could be something as simple as "foo" in C or something as
d9e94c22
MS
6084 ;; complex as "X<Y<class A<int>::B, BIT_MAX >> b>, ::operator<> ::
6085 ;; Z<(a>b)> :: operator const X<&foo>::T Q::G<unsigned short
6086 ;; int>::*volatile const" in C++ (this function is actually little
6087 ;; more than a `looking-at' call in all modes except those that,
e15f8aaa
AM
6088 ;; like C++, have `c-recognize-<>-arglists' set).
6089 ;;
6090 ;; Return
6091 ;; o - nil if no name is found;
6092 ;; o - 'template if it's an identifier ending with an angle bracket
6093 ;; arglist;
6094 ;; o - 'operator of it's an operator identifier;
6095 ;; o - t if it's some other kind of name.
0386b551
AM
6096 ;;
6097 ;; This function records identifier ranges on
6098 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
6099 ;; `c-record-type-identifiers' is non-nil.
6100 ;;
6101 ;; This function might do hidden buffer changes.
d9e94c22 6102
0386b551 6103 (let ((pos (point)) (start (point)) res id-start id-end
d9e94c22
MS
6104 ;; Turn off `c-promote-possible-types' here since we might
6105 ;; call `c-forward-<>-arglist' and we don't want it to promote
6106 ;; every suspect thing in the arglist to a type. We're
6107 ;; typically called from `c-forward-type' in this case, and
6108 ;; the caller only wants the top level type that it finds to
6109 ;; be promoted.
6110 c-promote-possible-types)
6111 (while
6112 (and
6113 (looking-at c-identifier-key)
6114
6115 (progn
6116 ;; Check for keyword. We go to the last symbol in
6117 ;; `c-identifier-key' first.
0386b551
AM
6118 (goto-char (setq id-end (match-end 0)))
6119 (c-simple-skip-symbol-backward)
6120 (setq id-start (point))
d9e94c22
MS
6121
6122 (if (looking-at c-keywords-regexp)
6123 (when (and (c-major-mode-is 'c++-mode)
6124 (looking-at
6125 (cc-eval-when-compile
6126 (concat "\\(operator\\|\\(template\\)\\)"
6127 "\\(" (c-lang-const c-nonsymbol-key c++)
6128 "\\|$\\)")))
6129 (if (match-beginning 2)
6130 ;; "template" is only valid inside an
6131 ;; identifier if preceded by "::".
6132 (save-excursion
6133 (c-backward-syntactic-ws)
6134 (and (c-safe (backward-char 2) t)
6135 (looking-at "::")))
6136 t))
6137
6138 ;; Handle a C++ operator or template identifier.
6139 (goto-char id-end)
6140 (c-forward-syntactic-ws)
6141 (cond ((eq (char-before id-end) ?e)
6142 ;; Got "... ::template".
6143 (let ((subres (c-forward-name)))
6144 (when subres
6145 (setq pos (point)
6146 res subres))))
6147
6148 ((looking-at c-identifier-start)
6149 ;; Got a cast operator.
6150 (when (c-forward-type)
6151 (setq pos (point)
6152 res 'operator)
6153 ;; Now we should match a sequence of either
6154 ;; '*', '&' or a name followed by ":: *",
6155 ;; where each can be followed by a sequence
6156 ;; of `c-opt-type-modifier-key'.
6157 (while (cond ((looking-at "[*&]")
6158 (goto-char (match-end 0))
6159 t)
6160 ((looking-at c-identifier-start)
6161 (and (c-forward-name)
6162 (looking-at "::")
6163 (progn
6164 (goto-char (match-end 0))
6165 (c-forward-syntactic-ws)
6166 (eq (char-after) ?*))
6167 (progn
6168 (forward-char)
6169 t))))
6170 (while (progn
6171 (c-forward-syntactic-ws)
6172 (setq pos (point))
6173 (looking-at c-opt-type-modifier-key))
6174 (goto-char (match-end 1))))))
6175
6176 ((looking-at c-overloadable-operators-regexp)
6177 ;; Got some other operator.
0386b551
AM
6178 (setq c-last-identifier-range
6179 (cons (point) (match-end 0)))
d9e94c22
MS
6180 (goto-char (match-end 0))
6181 (c-forward-syntactic-ws)
6182 (setq pos (point)
6183 res 'operator)))
6184
6185 nil)
6186
0386b551
AM
6187 ;; `id-start' is equal to `id-end' if we've jumped over
6188 ;; an identifier that doesn't end with a symbol token.
6189 ;; That can occur e.g. for Java import directives on the
6190 ;; form "foo.bar.*".
6191 (when (and id-start (/= id-start id-end))
d9e94c22
MS
6192 (setq c-last-identifier-range
6193 (cons id-start id-end)))
6194 (goto-char id-end)
6195 (c-forward-syntactic-ws)
6196 (setq pos (point)
6197 res t)))
6198
6199 (progn
6200 (goto-char pos)
6201 (when (or c-opt-identifier-concat-key
6202 c-recognize-<>-arglists)
6203
6204 (cond
6205 ((and c-opt-identifier-concat-key
6206 (looking-at c-opt-identifier-concat-key))
6207 ;; Got a concatenated identifier. This handles the
6208 ;; cases with tricky syntactic whitespace that aren't
6209 ;; covered in `c-identifier-key'.
6210 (goto-char (match-end 0))
6211 (c-forward-syntactic-ws)
6212 t)
6213
6214 ((and c-recognize-<>-arglists
6215 (eq (char-after) ?<))
6216 ;; Maybe an angle bracket arglist.
452ea855
AM
6217 (when (let ((c-record-type-identifiers t)
6218 (c-record-found-types t))
0386b551
AM
6219 (c-forward-<>-arglist nil))
6220
6221 (c-add-type start (1+ pos))
d9e94c22 6222 (c-forward-syntactic-ws)
0386b551
AM
6223 (setq pos (point)
6224 c-last-identifier-range nil)
6225
d9e94c22
MS
6226 (if (and c-opt-identifier-concat-key
6227 (looking-at c-opt-identifier-concat-key))
0386b551 6228
d9e94c22
MS
6229 ;; Continue if there's an identifier concatenation
6230 ;; operator after the template argument.
6231 (progn
0386b551
AM
6232 (when (and c-record-type-identifiers id-start)
6233 (c-record-ref-id (cons id-start id-end)))
d9e94c22
MS
6234 (forward-char 2)
6235 (c-forward-syntactic-ws)
6236 t)
0386b551
AM
6237
6238 (when (and c-record-type-identifiers id-start)
6239 (c-record-type-id (cons id-start id-end)))
d9e94c22
MS
6240 (setq res 'template)
6241 nil)))
6242 )))))
6243
6244 (goto-char pos)
6245 res))
6246
e15f8aaa 6247(defun c-forward-type (&optional brace-block-too)
d9e94c22 6248 ;; Move forward over a type spec if at the beginning of one,
e15f8aaa
AM
6249 ;; stopping at the next following token. The keyword "typedef"
6250 ;; isn't part of a type spec here.
6251 ;;
6252 ;; BRACE-BLOCK-TOO, when non-nil, means move over the brace block in
6253 ;; constructs like "struct foo {...} bar ;" or "struct {...} bar;".
6254 ;; The current (2009-03-10) intention is to convert all uses of
6255 ;; `c-forward-type' to call with this parameter set, then to
6256 ;; eliminate it.
6257 ;;
6258 ;; Return
6259 ;; o - t if it's a known type that can't be a name or other
6260 ;; expression;
6261 ;; o - 'known if it's an otherwise known type (according to
6262 ;; `*-font-lock-extra-types');
6263 ;; o - 'prefix if it's a known prefix of a type;
6264 ;; o - 'found if it's a type that matches one in `c-found-types';
53964682 6265 ;; o - 'maybe if it's an identifier that might be a type; or
e15f8aaa
AM
6266 ;; o - nil if it can't be a type (the point isn't moved then).
6267 ;;
6268 ;; The point is assumed to be at the beginning of a token.
d9e94c22
MS
6269 ;;
6270 ;; Note that this function doesn't skip past the brace definition
6271 ;; that might be considered part of the type, e.g.
6272 ;; "enum {a, b, c} foo".
0386b551
AM
6273 ;;
6274 ;; This function records identifier ranges on
6275 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
6276 ;; `c-record-type-identifiers' is non-nil.
6277 ;;
6278 ;; This function might do hidden buffer changes.
0e5cf2b8
AM
6279 (when (and c-recognize-<>-arglists
6280 (looking-at "<"))
452ea855
AM
6281 (c-forward-<>-arglist t)
6282 (c-forward-syntactic-ws))
0386b551
AM
6283
6284 (let ((start (point)) pos res name-res id-start id-end id-range)
d9e94c22
MS
6285
6286 ;; Skip leading type modifiers. If any are found we know it's a
6287 ;; prefix of a type.
e15f8aaa 6288 (when c-opt-type-modifier-key ; e.g. "const" "volatile", but NOT "typedef"
d9e94c22
MS
6289 (while (looking-at c-opt-type-modifier-key)
6290 (goto-char (match-end 1))
6291 (c-forward-syntactic-ws)
6292 (setq res 'prefix)))
6293
6294 (cond
e15f8aaa
AM
6295 ((looking-at c-type-prefix-key) ; e.g. "struct", "class", but NOT
6296 ; "typedef".
d9e94c22
MS
6297 (goto-char (match-end 1))
6298 (c-forward-syntactic-ws)
6299 (setq pos (point))
e15f8aaa
AM
6300
6301 (setq name-res (c-forward-name))
6302 (setq res (not (null name-res)))
6303 (when (eq name-res t)
6304 ;; In many languages the name can be used without the
6305 ;; prefix, so we add it to `c-found-types'.
6306 (c-add-type pos (point))
6307 (when (and c-record-type-identifiers
6308 c-last-identifier-range)
6309 (c-record-type-id c-last-identifier-range)))
6310 (when (and brace-block-too
6311 (memq res '(t nil))
6312 (eq (char-after) ?\{)
6313 (save-excursion
6314 (c-safe
6315 (progn (c-forward-sexp)
6316 (c-forward-syntactic-ws)
6317 (setq pos (point))))))
6318 (goto-char pos)
6319 (setq res t))
6320 (unless res (goto-char start))) ; invalid syntax
d9e94c22
MS
6321
6322 ((progn
6323 (setq pos nil)
6324 (if (looking-at c-identifier-start)
6325 (save-excursion
6326 (setq id-start (point)
0386b551
AM
6327 name-res (c-forward-name))
6328 (when name-res
d9e94c22
MS
6329 (setq id-end (point)
6330 id-range c-last-identifier-range))))
6331 (and (cond ((looking-at c-primitive-type-key)
6332 (setq res t))
6333 ((c-with-syntax-table c-identifier-syntax-table
6334 (looking-at c-known-type-key))
6335 (setq res 'known)))
6336 (or (not id-end)
6337 (>= (save-excursion
6338 (save-match-data
6339 (goto-char (match-end 1))
6340 (c-forward-syntactic-ws)
6341 (setq pos (point))))
6342 id-end)
6343 (setq res nil))))
6344 ;; Looking at a primitive or known type identifier. We've
6345 ;; checked for a name first so that we don't go here if the
6346 ;; known type match only is a prefix of another name.
6347
6348 (setq id-end (match-end 1))
6349
6350 (when (and c-record-type-identifiers
6351 (or c-promote-possible-types (eq res t)))
6352 (c-record-type-id (cons (match-beginning 1) (match-end 1))))
6353
6354 (if (and c-opt-type-component-key
6355 (save-match-data
6356 (looking-at c-opt-type-component-key)))
6357 ;; There might be more keywords for the type.
6358 (let (safe-pos)
0386b551 6359 (c-forward-keyword-clause 1)
d9e94c22
MS
6360 (while (progn
6361 (setq safe-pos (point))
6362 (looking-at c-opt-type-component-key))
6363 (when (and c-record-type-identifiers
6364 (looking-at c-primitive-type-key))
6365 (c-record-type-id (cons (match-beginning 1)
6366 (match-end 1))))
0386b551 6367 (c-forward-keyword-clause 1))
d9e94c22
MS
6368 (if (looking-at c-primitive-type-key)
6369 (progn
6370 (when c-record-type-identifiers
6371 (c-record-type-id (cons (match-beginning 1)
6372 (match-end 1))))
0386b551 6373 (c-forward-keyword-clause 1)
d9e94c22
MS
6374 (setq res t))
6375 (goto-char safe-pos)
6376 (setq res 'prefix)))
0386b551 6377 (unless (save-match-data (c-forward-keyword-clause 1))
d9e94c22
MS
6378 (if pos
6379 (goto-char pos)
6380 (goto-char (match-end 1))
6381 (c-forward-syntactic-ws)))))
6382
0386b551
AM
6383 (name-res
6384 (cond ((eq name-res t)
d9e94c22
MS
6385 ;; A normal identifier.
6386 (goto-char id-end)
6387 (if (or res c-promote-possible-types)
6388 (progn
6389 (c-add-type id-start id-end)
0386b551 6390 (when (and c-record-type-identifiers id-range)
d9e94c22
MS
6391 (c-record-type-id id-range))
6392 (unless res
6393 (setq res 'found)))
6394 (setq res (if (c-check-type id-start id-end)
6395 ;; It's an identifier that has been used as
6396 ;; a type somewhere else.
6397 'found
6398 ;; It's an identifier that might be a type.
6399 'maybe))))
0386b551 6400 ((eq name-res 'template)
d9e94c22
MS
6401 ;; A template is a type.
6402 (goto-char id-end)
6403 (setq res t))
6404 (t
6405 ;; Otherwise it's an operator identifier, which is not a type.
6406 (goto-char start)
6407 (setq res nil)))))
6408
6409 (when res
e15f8aaa 6410 ;; Skip trailing type modifiers. If any are found we know it's
d9e94c22
MS
6411 ;; a type.
6412 (when c-opt-type-modifier-key
e15f8aaa 6413 (while (looking-at c-opt-type-modifier-key) ; e.g. "const", "volatile"
d9e94c22
MS
6414 (goto-char (match-end 1))
6415 (c-forward-syntactic-ws)
6416 (setq res t)))
d9e94c22
MS
6417 ;; Step over any type suffix operator. Do not let the existence
6418 ;; of these alter the classification of the found type, since
6419 ;; these operators typically are allowed in normal expressions
6420 ;; too.
6421 (when c-opt-type-suffix-key
6422 (while (looking-at c-opt-type-suffix-key)
6423 (goto-char (match-end 1))
6424 (c-forward-syntactic-ws)))
6425
e15f8aaa 6426 (when c-opt-type-concat-key ; Only/mainly for pike.
0386b551
AM
6427 ;; Look for a trailing operator that concatenates the type
6428 ;; with a following one, and if so step past that one through
6429 ;; a recursive call. Note that we don't record concatenated
6430 ;; types in `c-found-types' - it's the component types that
6431 ;; are recorded when appropriate.
d9e94c22
MS
6432 (setq pos (point))
6433 (let* ((c-promote-possible-types (or (memq res '(t known))
6434 c-promote-possible-types))
6435 ;; If we can't promote then set `c-record-found-types' so that
6436 ;; we can merge in the types from the second part afterwards if
6437 ;; it turns out to be a known type there.
6438 (c-record-found-types (and c-record-type-identifiers
0386b551
AM
6439 (not c-promote-possible-types)))
6440 subres)
d9e94c22
MS
6441 (if (and (looking-at c-opt-type-concat-key)
6442
6443 (progn
6444 (goto-char (match-end 1))
6445 (c-forward-syntactic-ws)
0386b551 6446 (setq subres (c-forward-type))))
d9e94c22
MS
6447
6448 (progn
6449 ;; If either operand certainly is a type then both are, but we
6450 ;; don't let the existence of the operator itself promote two
6451 ;; uncertain types to a certain one.
6452 (cond ((eq res t))
0386b551
AM
6453 ((eq subres t)
6454 (unless (eq name-res 'template)
6455 (c-add-type id-start id-end))
6456 (when (and c-record-type-identifiers id-range)
d9e94c22
MS
6457 (c-record-type-id id-range))
6458 (setq res t))
2a15eb73 6459 ((eq res 'known))
0386b551 6460 ((eq subres 'known)
2a15eb73 6461 (setq res 'known))
d9e94c22 6462 ((eq res 'found))
0386b551 6463 ((eq subres 'found)
d9e94c22
MS
6464 (setq res 'found))
6465 (t
6466 (setq res 'maybe)))
6467
6468 (when (and (eq res t)
6469 (consp c-record-found-types))
6470 ;; Merge in the ranges of any types found by the second
6471 ;; `c-forward-type'.
6472 (setq c-record-type-identifiers
6473 ;; `nconc' doesn't mind that the tail of
6474 ;; `c-record-found-types' is t.
6475 (nconc c-record-found-types
6476 c-record-type-identifiers))))
6477
6478 (goto-char pos))))
6479
6480 (when (and c-record-found-types (memq res '(known found)) id-range)
6481 (setq c-record-found-types
6482 (cons id-range c-record-found-types))))
6483
6484 ;;(message "c-forward-type %s -> %s: %s" start (point) res)
6485
6486 res))
6487
452ea855
AM
6488(defun c-forward-annotation ()
6489 ;; Used for Java code only at the moment. Assumes point is on the
6490 ;; @, moves forward an annotation. returns nil if there is no
6491 ;; annotation at point.
6492 (and (looking-at "@")
6493 (progn (forward-char) t)
6494 (c-forward-type)
6495 (progn (c-forward-syntactic-ws) t)
6496 (if (looking-at "(")
6497 (c-go-list-forward)
6498 t)))
6499
6758b6a8
GM
6500(defmacro c-pull-open-brace (ps)
6501 ;; Pull the next open brace from PS (which has the form of paren-state),
6502 ;; skipping over any brace pairs. Returns NIL when PS is exhausted.
6503 `(progn
6504 (while (consp (car ,ps))
6505 (setq ,ps (cdr ,ps)))
6506 (prog1 (car ,ps)
6507 (setq ,ps (cdr ,ps)))))
6508
3d1c39fa
AM
6509(defun c-back-over-member-initializers ()
6510 ;; Test whether we are in a C++ member initializer list, and if so, go back
6511 ;; to the introducing ":", returning the position of the opening paren of
6512 ;; the function's arglist. Otherwise return nil, leaving point unchanged.
6513 (let ((here (point))
6514 (paren-state (c-parse-state))
6515 res)
6516
6517 (setq res
6518 (catch 'done
6519 (if (not (c-at-toplevel-p))
6520 (progn
6521 (while (not (c-at-toplevel-p))
6522 (goto-char (c-pull-open-brace paren-state)))
6523 (c-backward-syntactic-ws)
6524 (when (not (c-simple-skip-symbol-backward))
6525 (throw 'done nil))
6526 (c-backward-syntactic-ws))
6527 (c-backward-syntactic-ws)
6528 (when (memq (char-before) '(?\) ?}))
6529 (when (not (c-go-list-backward))
6530 (throw 'done nil))
6531 (c-backward-syntactic-ws))
6532 (when (c-simple-skip-symbol-backward)
6533 (c-backward-syntactic-ws)))
6534
6535 (while (eq (char-before) ?,)
6536 (backward-char)
6537 (c-backward-syntactic-ws)
6538
6539 (when (not (memq (char-before) '(?\) ?})))
6540 (throw 'done nil))
6541 (when (not (c-go-list-backward))
6542 (throw 'done nil))
6543 (c-backward-syntactic-ws)
6544 (when (not (c-simple-skip-symbol-backward))
6545 (throw 'done nil))
6546 (c-backward-syntactic-ws))
6547
6548 (and
6549 (eq (char-before) ?:)
6550 (c-just-after-func-arglist-p))))
6551
6552 (or res (goto-char here))
6553 res))
6554
785eecbb 6555\f
d9e94c22
MS
6556;; Handling of large scale constructs like statements and declarations.
6557
0386b551
AM
6558;; Macro used inside `c-forward-decl-or-cast-1'. It ought to be a
6559;; defsubst or perhaps even a defun, but it contains lots of free
6560;; variables that refer to things inside `c-forward-decl-or-cast-1'.
6561(defmacro c-fdoc-shift-type-backward (&optional short)
6562 ;; `c-forward-decl-or-cast-1' can consume an arbitrary length list
6563 ;; of types when parsing a declaration, which means that it
6564 ;; sometimes consumes the identifier in the declaration as a type.
6565 ;; This is used to "backtrack" and make the last type be treated as
6566 ;; an identifier instead.
6567 `(progn
6568 ,(unless short
6569 ;; These identifiers are bound only in the inner let.
6570 '(setq identifier-type at-type
6571 identifier-start type-start
6572 got-parens nil
6573 got-identifier t
6574 got-suffix t
6575 got-suffix-after-parens id-start
6576 paren-depth 0))
6577
6578 (if (setq at-type (if (eq backup-at-type 'prefix)
6579 t
6580 backup-at-type))
6581 (setq type-start backup-type-start
6582 id-start backup-id-start)
6583 (setq type-start start-pos
6584 id-start start-pos))
6585
6586 ;; When these flags already are set we've found specifiers that
6587 ;; unconditionally signal these attributes - backtracking doesn't
6588 ;; change that. So keep them set in that case.
6589 (or at-type-decl
6590 (setq at-type-decl backup-at-type-decl))
6591 (or maybe-typeless
6592 (setq maybe-typeless backup-maybe-typeless))
6593
6594 ,(unless short
6595 ;; This identifier is bound only in the inner let.
6596 '(setq start id-start))))
6597
6598(defun c-forward-decl-or-cast-1 (preceding-token-end context last-cast-end)
6599 ;; Move forward over a declaration or a cast if at the start of one.
6600 ;; The point is assumed to be at the start of some token. Nil is
6601 ;; returned if no declaration or cast is recognized, and the point
6602 ;; is clobbered in that case.
6603 ;;
6604 ;; If a declaration is parsed:
6605 ;;
6606 ;; The point is left at the first token after the first complete
6607 ;; declarator, if there is one. The return value is a cons where
022d0cf4
AM
6608 ;; the car is the position of the first token in the declarator. (See
6609 ;; below for the cdr.)
0386b551
AM
6610 ;; Some examples:
6611 ;;
6612 ;; void foo (int a, char *b) stuff ...
6613 ;; car ^ ^ point
6614 ;; float (*a)[], b;
6615 ;; car ^ ^ point
6616 ;; unsigned int a = c_style_initializer, b;
6617 ;; car ^ ^ point
6618 ;; unsigned int a (cplusplus_style_initializer), b;
6619 ;; car ^ ^ point (might change)
6620 ;; class Foo : public Bar {}
6621 ;; car ^ ^ point
6622 ;; class PikeClass (int a, string b) stuff ...
6623 ;; car ^ ^ point
6624 ;; enum bool;
6625 ;; car ^ ^ point
6626 ;; enum bool flag;
6627 ;; car ^ ^ point
6628 ;; void cplusplus_function (int x) throw (Bad);
6629 ;; car ^ ^ point
6630 ;; Foo::Foo (int b) : Base (b) {}
6631 ;; car ^ ^ point
91af3942 6632 ;;
e15f8aaa
AM
6633 ;; The cdr of the return value is non-nil when a
6634 ;; `c-typedef-decl-kwds' specifier is found in the declaration.
6635 ;; Specifically it is a dotted pair (A . B) where B is t when a
6636 ;; `c-typedef-kwds' ("typedef") is present, and A is t when some
6637 ;; other `c-typedef-decl-kwds' (e.g. class, struct, enum)
6638 ;; specifier is present. I.e., (some of) the declared
6639 ;; identifier(s) are types.
91af3942 6640 ;;
0386b551
AM
6641 ;; If a cast is parsed:
6642 ;;
6643 ;; The point is left at the first token after the closing paren of
6644 ;; the cast. The return value is `cast'. Note that the start
6645 ;; position must be at the first token inside the cast parenthesis
6646 ;; to recognize it.
6647 ;;
6648 ;; PRECEDING-TOKEN-END is the first position after the preceding
6649 ;; token, i.e. on the other side of the syntactic ws from the point.
6650 ;; Use a value less than or equal to (point-min) if the point is at
6651 ;; the first token in (the visible part of) the buffer.
6652 ;;
6653 ;; CONTEXT is a symbol that describes the context at the point:
022d0cf4 6654 ;; 'decl In a comma-separated declaration context (typically
0386b551
AM
6655 ;; inside a function declaration arglist).
6656 ;; '<> In an angle bracket arglist.
6657 ;; 'arglist Some other type of arglist.
a85fd6da
AM
6658 ;; nil Some other context or unknown context. Includes
6659 ;; within the parens of an if, for, ... construct.
0386b551
AM
6660 ;;
6661 ;; LAST-CAST-END is the first token after the closing paren of a
6662 ;; preceding cast, or nil if none is known. If
6663 ;; `c-forward-decl-or-cast-1' is used in succession, it should be
6664 ;; the position after the closest preceding call where a cast was
6665 ;; matched. In that case it's used to discover chains of casts like
6666 ;; "(a) (b) c".
6667 ;;
6668 ;; This function records identifier ranges on
6669 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
6670 ;; `c-record-type-identifiers' is non-nil.
6671 ;;
6672 ;; This function might do hidden buffer changes.
6673
6674 (let (;; `start-pos' is used below to point to the start of the
6675 ;; first type, i.e. after any leading specifiers. It might
6676 ;; also point at the beginning of the preceding syntactic
6677 ;; whitespace.
6678 (start-pos (point))
6679 ;; Set to the result of `c-forward-type'.
6680 at-type
6681 ;; The position of the first token in what we currently
6682 ;; believe is the type in the declaration or cast, after any
6683 ;; specifiers and their associated clauses.
6684 type-start
6685 ;; The position of the first token in what we currently
6686 ;; believe is the declarator for the first identifier. Set
6687 ;; when the type is found, and moved forward over any
6688 ;; `c-decl-hangon-kwds' and their associated clauses that
6689 ;; occurs after the type.
6690 id-start
6691 ;; These store `at-type', `type-start' and `id-start' of the
6692 ;; identifier before the one in those variables. The previous
6693 ;; identifier might turn out to be the real type in a
6694 ;; declaration if the last one has to be the declarator in it.
6695 ;; If `backup-at-type' is nil then the other variables have
6696 ;; undefined values.
6697 backup-at-type backup-type-start backup-id-start
e15f8aaa
AM
6698 ;; Set if we've found a specifier (apart from "typedef") that makes
6699 ;; the defined identifier(s) types.
0386b551 6700 at-type-decl
e15f8aaa
AM
6701 ;; Set if we've a "typedef" keyword.
6702 at-typedef
0386b551
AM
6703 ;; Set if we've found a specifier that can start a declaration
6704 ;; where there's no type.
6705 maybe-typeless
6706 ;; If a specifier is found that also can be a type prefix,
6707 ;; these flags are set instead of those above. If we need to
6708 ;; back up an identifier, they are copied to the real flag
6709 ;; variables. Thus they only take effect if we fail to
6710 ;; interpret it as a type.
6711 backup-at-type-decl backup-maybe-typeless
6712 ;; Whether we've found a declaration or a cast. We might know
6713 ;; this before we've found the type in it. It's 'ids if we've
6714 ;; found two consecutive identifiers (usually a sure sign, but
6715 ;; we should allow that in labels too), and t if we've found a
6716 ;; specifier keyword (a 100% sure sign).
6717 at-decl-or-cast
6718 ;; Set when we need to back up to parse this as a declaration
6719 ;; but not as a cast.
6720 backup-if-not-cast
6721 ;; For casts, the return position.
6722 cast-end
6723 ;; Save `c-record-type-identifiers' and
6724 ;; `c-record-ref-identifiers' since ranges are recorded
6725 ;; speculatively and should be thrown away if it turns out
6726 ;; that it isn't a declaration or cast.
6727 (save-rec-type-ids c-record-type-identifiers)
6728 (save-rec-ref-ids c-record-ref-identifiers))
6729
452ea855
AM
6730 (while (c-forward-annotation)
6731 (c-forward-syntactic-ws))
6732
0386b551
AM
6733 ;; Check for a type. Unknown symbols are treated as possible
6734 ;; types, but they could also be specifiers disguised through
6735 ;; macros like __INLINE__, so we recognize both types and known
6736 ;; specifiers after them too.
6737 (while
6738 (let* ((start (point)) kwd-sym kwd-clause-end found-type)
6739
6740 ;; Look for a specifier keyword clause.
e3715033
AM
6741 (when (or (looking-at c-prefix-spec-kwds-re)
6742 (and (c-major-mode-is 'java-mode)
6743 (looking-at "@[A-Za-z0-9]+")))
e15f8aaa
AM
6744 (if (looking-at c-typedef-key)
6745 (setq at-typedef t))
0386b551
AM
6746 (setq kwd-sym (c-keyword-sym (match-string 1)))
6747 (save-excursion
6748 (c-forward-keyword-clause 1)
6749 (setq kwd-clause-end (point))))
6750
e15f8aaa 6751 (when (setq found-type (c-forward-type t)) ; brace-block-too
0386b551
AM
6752 ;; Found a known or possible type or a prefix of a known type.
6753
6754 (when at-type
6755 ;; Got two identifiers with nothing but whitespace
6756 ;; between them. That can only happen in declarations.
6757 (setq at-decl-or-cast 'ids)
6758
6759 (when (eq at-type 'found)
6760 ;; If the previous identifier is a found type we
6761 ;; record it as a real one; it might be some sort of
6762 ;; alias for a prefix like "unsigned".
6763 (save-excursion
6764 (goto-char type-start)
6765 (let ((c-promote-possible-types t))
6766 (c-forward-type)))))
6767
6768 (setq backup-at-type at-type
6769 backup-type-start type-start
6770 backup-id-start id-start
6771 at-type found-type
6772 type-start start
6773 id-start (point)
6774 ;; The previous ambiguous specifier/type turned out
6775 ;; to be a type since we've parsed another one after
6776 ;; it, so clear these backup flags.
6777 backup-at-type-decl nil
6778 backup-maybe-typeless nil))
6779
6780 (if kwd-sym
6781 (progn
6782 ;; Handle known specifier keywords and
6783 ;; `c-decl-hangon-kwds' which can occur after known
6784 ;; types.
6785
6786 (if (c-keyword-member kwd-sym 'c-decl-hangon-kwds)
6787 ;; It's a hang-on keyword that can occur anywhere.
6788 (progn
6789 (setq at-decl-or-cast t)
6790 (if at-type
6791 ;; Move the identifier start position if
6792 ;; we've passed a type.
6793 (setq id-start kwd-clause-end)
6794 ;; Otherwise treat this as a specifier and
6795 ;; move the fallback position.
6796 (setq start-pos kwd-clause-end))
6797 (goto-char kwd-clause-end))
6798
6799 ;; It's an ordinary specifier so we know that
6800 ;; anything before this can't be the type.
6801 (setq backup-at-type nil
6802 start-pos kwd-clause-end)
6803
6804 (if found-type
6805 ;; It's ambiguous whether this keyword is a
6806 ;; specifier or a type prefix, so set the backup
6807 ;; flags. (It's assumed that `c-forward-type'
6808 ;; moved further than `c-forward-keyword-clause'.)
6809 (progn
6810 (when (c-keyword-member kwd-sym 'c-typedef-decl-kwds)
6811 (setq backup-at-type-decl t))
6812 (when (c-keyword-member kwd-sym 'c-typeless-decl-kwds)
6813 (setq backup-maybe-typeless t)))
6814
6815 (when (c-keyword-member kwd-sym 'c-typedef-decl-kwds)
e15f8aaa
AM
6816 ;; This test only happens after we've scanned a type.
6817 ;; So, with valid syntax, kwd-sym can't be 'typedef.
0386b551
AM
6818 (setq at-type-decl t))
6819 (when (c-keyword-member kwd-sym 'c-typeless-decl-kwds)
6820 (setq maybe-typeless t))
6821
ee7683eb 6822 ;; Haven't matched a type so it's an unambiguous
0386b551
AM
6823 ;; specifier keyword and we know we're in a
6824 ;; declaration.
6825 (setq at-decl-or-cast t)
6826
6827 (goto-char kwd-clause-end))))
6828
6829 ;; If the type isn't known we continue so that we'll jump
6830 ;; over all specifiers and type identifiers. The reason
6831 ;; to do this for a known type prefix is to make things
6832 ;; like "unsigned INT16" work.
6833 (and found-type (not (eq found-type t))))))
6834
6835 (cond
6836 ((eq at-type t)
6837 ;; If a known type was found, we still need to skip over any
6838 ;; hangon keyword clauses after it. Otherwise it has already
6839 ;; been done in the loop above.
6840 (while (looking-at c-decl-hangon-key)
6841 (c-forward-keyword-clause 1))
6842 (setq id-start (point)))
6843
6844 ((eq at-type 'prefix)
6845 ;; A prefix type is itself a primitive type when it's not
6846 ;; followed by another type.
6847 (setq at-type t))
6848
6849 ((not at-type)
6850 ;; Got no type but set things up to continue anyway to handle
6851 ;; the various cases when a declaration doesn't start with a
6852 ;; type.
6853 (setq id-start start-pos))
6854
6855 ((and (eq at-type 'maybe)
6856 (c-major-mode-is 'c++-mode))
6857 ;; If it's C++ then check if the last "type" ends on the form
6858 ;; "foo::foo" or "foo::~foo", i.e. if it's the name of a
6859 ;; (con|de)structor.
6860 (save-excursion
6861 (let (name end-2 end-1)
6862 (goto-char id-start)
6863 (c-backward-syntactic-ws)
6864 (setq end-2 (point))
6865 (when (and
6866 (c-simple-skip-symbol-backward)
6867 (progn
6868 (setq name
6869 (buffer-substring-no-properties (point) end-2))
6870 ;; Cheating in the handling of syntactic ws below.
6871 (< (skip-chars-backward ":~ \t\n\r\v\f") 0))
6872 (progn
6873 (setq end-1 (point))
6874 (c-simple-skip-symbol-backward))
6875 (>= (point) type-start)
6876 (equal (buffer-substring-no-properties (point) end-1)
6877 name))
6878 ;; It is a (con|de)structor name. In that case the
6879 ;; declaration is typeless so zap out any preceding
6880 ;; identifier(s) that we might have taken as types.
6881 (goto-char type-start)
6882 (setq at-type nil
6883 backup-at-type nil
6884 id-start type-start))))))
6885
6886 ;; Check for and step over a type decl expression after the thing
6887 ;; that is or might be a type. This can't be skipped since we
6888 ;; need the correct end position of the declarator for
6889 ;; `max-type-decl-end-*'.
6890 (let ((start (point)) (paren-depth 0) pos
6891 ;; True if there's a non-open-paren match of
6892 ;; `c-type-decl-prefix-key'.
6893 got-prefix
6894 ;; True if the declarator is surrounded by a parenthesis pair.
6895 got-parens
6896 ;; True if there is an identifier in the declarator.
6897 got-identifier
6898 ;; True if there's a non-close-paren match of
6899 ;; `c-type-decl-suffix-key'.
6900 got-suffix
6901 ;; True if there's a prefix match outside the outermost
6902 ;; paren pair that surrounds the declarator.
6903 got-prefix-before-parens
1379f2c5 6904 ;; True if there's a suffix match outside the outermost
0386b551
AM
6905 ;; paren pair that surrounds the declarator. The value is
6906 ;; the position of the first suffix match.
6907 got-suffix-after-parens
6908 ;; True if we've parsed the type decl to a token that is
6909 ;; known to end declarations in this context.
6910 at-decl-end
6911 ;; The earlier values of `at-type' and `type-start' if we've
6912 ;; shifted the type backwards.
6913 identifier-type identifier-start
6914 ;; If `c-parse-and-markup-<>-arglists' is set we need to
6915 ;; turn it off during the name skipping below to avoid
6916 ;; getting `c-type' properties that might be bogus. That
6917 ;; can happen since we don't know if
6918 ;; `c-restricted-<>-arglists' will be correct inside the
6919 ;; arglist paren that gets entered.
6920 c-parse-and-markup-<>-arglists)
6921
6922 (goto-char id-start)
6923
6924 ;; Skip over type decl prefix operators. (Note similar code in
6925 ;; `c-font-lock-declarators'.)
e8dd0787
AM
6926 (if (and c-recognize-typeless-decls
6927 (equal c-type-decl-prefix-key "\\<\\>"))
6928 (when (eq (char-after) ?\()
0386b551
AM
6929 (progn
6930 (setq paren-depth (1+ paren-depth))
e8dd0787
AM
6931 (forward-char)))
6932 (while (and (looking-at c-type-decl-prefix-key)
6933 (if (and (c-major-mode-is 'c++-mode)
6934 (match-beginning 3))
6935 ;; If the third submatch matches in C++ then
6936 ;; we're looking at an identifier that's a
6937 ;; prefix only if it specifies a member pointer.
6938 (when (setq got-identifier (c-forward-name))
6939 (if (looking-at "\\(::\\)")
6940 ;; We only check for a trailing "::" and
6941 ;; let the "*" that should follow be
6942 ;; matched in the next round.
6943 (progn (setq got-identifier nil) t)
6944 ;; It turned out to be the real identifier,
6945 ;; so stop.
6946 nil))
6947 t))
6948
6949 (if (eq (char-after) ?\()
6950 (progn
6951 (setq paren-depth (1+ paren-depth))
6952 (forward-char))
6953 (unless got-prefix-before-parens
6954 (setq got-prefix-before-parens (= paren-depth 0)))
6955 (setq got-prefix t)
6956 (goto-char (match-end 1)))
6957 (c-forward-syntactic-ws)))
0386b551
AM
6958
6959 (setq got-parens (> paren-depth 0))
6960
6961 ;; Skip over an identifier.
6962 (or got-identifier
6963 (and (looking-at c-identifier-start)
6964 (setq got-identifier (c-forward-name))))
6965
6966 ;; Skip over type decl suffix operators.
6967 (while (if (looking-at c-type-decl-suffix-key)
6968
6969 (if (eq (char-after) ?\))
6970 (when (> paren-depth 0)
6971 (setq paren-depth (1- paren-depth))
6972 (forward-char)
6973 t)
6974 (when (if (save-match-data (looking-at "\\s\("))
6975 (c-safe (c-forward-sexp 1) t)
6976 (goto-char (match-end 1))
6977 t)
6978 (when (and (not got-suffix-after-parens)
6979 (= paren-depth 0))
6980 (setq got-suffix-after-parens (match-beginning 0)))
6981 (setq got-suffix t)))
6982
6983 ;; No suffix matched. We might have matched the
6984 ;; identifier as a type and the open paren of a
6985 ;; function arglist as a type decl prefix. In that
6986 ;; case we should "backtrack": Reinterpret the last
6987 ;; type as the identifier, move out of the arglist and
6988 ;; continue searching for suffix operators.
6989 ;;
6990 ;; Do this even if there's no preceding type, to cope
6991 ;; with old style function declarations in K&R C,
6992 ;; (con|de)structors in C++ and `c-typeless-decl-kwds'
6993 ;; style declarations. That isn't applicable in an
6994 ;; arglist context, though.
6995 (when (and (= paren-depth 1)
6996 (not got-prefix-before-parens)
6997 (not (eq at-type t))
6998 (or backup-at-type
6999 maybe-typeless
7000 backup-maybe-typeless
7001 (when c-recognize-typeless-decls
7002 (not context)))
7003 (setq pos (c-up-list-forward (point)))
7004 (eq (char-before pos) ?\)))
7005 (c-fdoc-shift-type-backward)
7006 (goto-char pos)
7007 t))
7008
7009 (c-forward-syntactic-ws))
7010
7011 (when (and (or maybe-typeless backup-maybe-typeless)
7012 (not got-identifier)
7013 (not got-prefix)
7014 at-type)
7015 ;; Have found no identifier but `c-typeless-decl-kwds' has
7016 ;; matched so we know we're inside a declaration. The
7017 ;; preceding type must be the identifier instead.
7018 (c-fdoc-shift-type-backward))
7019
7020 (setq
7021 at-decl-or-cast
7022 (catch 'at-decl-or-cast
7023
a85fd6da 7024 ;; CASE 1
0386b551
AM
7025 (when (> paren-depth 0)
7026 ;; Encountered something inside parens that isn't matched by
7027 ;; the `c-type-decl-*' regexps, so it's not a type decl
7028 ;; expression. Try to skip out to the same paren depth to
7029 ;; not confuse the cast check below.
7030 (c-safe (goto-char (scan-lists (point) 1 paren-depth)))
7031 ;; If we've found a specifier keyword then it's a
7032 ;; declaration regardless.
7033 (throw 'at-decl-or-cast (eq at-decl-or-cast t)))
7034
7035 (setq at-decl-end
7036 (looking-at (cond ((eq context '<>) "[,>]")
7037 (context "[,\)]")
7038 (t "[,;]"))))
7039
7040 ;; Now we've collected info about various characteristics of
7041 ;; the construct we're looking at. Below follows a decision
7042 ;; tree based on that. It's ordered to check more certain
7043 ;; signs before less certain ones.
7044
7045 (if got-identifier
7046 (progn
7047
a85fd6da 7048 ;; CASE 2
0386b551
AM
7049 (when (and (or at-type maybe-typeless)
7050 (not (or got-prefix got-parens)))
7051 ;; Got another identifier directly after the type, so it's a
7052 ;; declaration.
7053 (throw 'at-decl-or-cast t))
7054
7055 (when (and got-parens
7056 (not got-prefix)
7057 (not got-suffix-after-parens)
7058 (or backup-at-type
7059 maybe-typeless
7060 backup-maybe-typeless))
7061 ;; Got a declaration of the form "foo bar (gnu);" where we've
7062 ;; recognized "bar" as the type and "gnu" as the declarator.
7063 ;; In this case it's however more likely that "bar" is the
7064 ;; declarator and "gnu" a function argument or initializer (if
7065 ;; `c-recognize-paren-inits' is set), since the parens around
7066 ;; "gnu" would be superfluous if it's a declarator. Shift the
7067 ;; type one step backward.
7068 (c-fdoc-shift-type-backward)))
7069
7070 ;; Found no identifier.
7071
7072 (if backup-at-type
7073 (progn
7074
b248a85d
AM
7075
7076 ;; CASE 3
7077 (when (= (point) start)
7078 ;; Got a plain list of identifiers. If a colon follows it's
7079 ;; a valid label, or maybe a bitfield. Otherwise the last
7080 ;; one probably is the declared identifier and we should
7081 ;; back up to the previous type, providing it isn't a cast.
452ea855
AM
7082 (if (and (eq (char-after) ?:)
7083 (not (c-major-mode-is 'java-mode)))
b248a85d
AM
7084 (cond
7085 ;; If we've found a specifier keyword then it's a
7086 ;; declaration regardless.
7087 ((eq at-decl-or-cast t)
7088 (throw 'at-decl-or-cast t))
7089 ((and c-has-bitfields
7090 (eq at-decl-or-cast 'ids)) ; bitfield.
7091 (setq backup-if-not-cast t)
7092 (throw 'at-decl-or-cast t)))
7093
7094 (setq backup-if-not-cast t)
7095 (throw 'at-decl-or-cast t)))
0386b551 7096
a85fd6da 7097 ;; CASE 4
0386b551
AM
7098 (when (and got-suffix
7099 (not got-prefix)
7100 (not got-parens))
7101 ;; Got a plain list of identifiers followed by some suffix.
7102 ;; If this isn't a cast then the last identifier probably is
7103 ;; the declared one and we should back up to the previous
7104 ;; type.
7105 (setq backup-if-not-cast t)
7106 (throw 'at-decl-or-cast t)))
7107
a85fd6da 7108 ;; CASE 5
0386b551
AM
7109 (when (eq at-type t)
7110 ;; If the type is known we know that there can't be any
7111 ;; identifier somewhere else, and it's only in declarations in
7112 ;; e.g. function prototypes and in casts that the identifier may
7113 ;; be left out.
7114 (throw 'at-decl-or-cast t))
7115
7116 (when (= (point) start)
7117 ;; Only got a single identifier (parsed as a type so far).
a85fd6da 7118 ;; CASE 6
0386b551
AM
7119 (if (and
7120 ;; Check that the identifier isn't at the start of an
7121 ;; expression.
7122 at-decl-end
7123 (cond
7124 ((eq context 'decl)
7125 ;; Inside an arglist that contains declarations. If K&R
7126 ;; style declarations and parenthesis style initializers
7127 ;; aren't allowed then the single identifier must be a
7128 ;; type, else we require that it's known or found
7129 ;; (primitive types are handled above).
7130 (or (and (not c-recognize-knr-p)
7131 (not c-recognize-paren-inits))
7132 (memq at-type '(known found))))
7133 ((eq context '<>)
7134 ;; Inside a template arglist. Accept known and found
7135 ;; types; other identifiers could just as well be
7136 ;; constants in C++.
7137 (memq at-type '(known found)))))
7138 (throw 'at-decl-or-cast t)
a85fd6da 7139 ;; CASE 7
0386b551
AM
7140 ;; Can't be a valid declaration or cast, but if we've found a
7141 ;; specifier it can't be anything else either, so treat it as
7142 ;; an invalid/unfinished declaration or cast.
7143 (throw 'at-decl-or-cast at-decl-or-cast))))
7144
7145 (if (and got-parens
7146 (not got-prefix)
7147 (not context)
7148 (not (eq at-type t))
7149 (or backup-at-type
7150 maybe-typeless
7151 backup-maybe-typeless
7152 (when c-recognize-typeless-decls
7153 (or (not got-suffix)
7154 (not (looking-at
7155 c-after-suffixed-type-maybe-decl-key))))))
7156 ;; Got an empty paren pair and a preceding type that probably
7157 ;; really is the identifier. Shift the type backwards to make
7158 ;; the last one the identifier. This is analogous to the
7159 ;; "backtracking" done inside the `c-type-decl-suffix-key' loop
7160 ;; above.
7161 ;;
7162 ;; Exception: In addition to the conditions in that
7163 ;; "backtracking" code, do not shift backward if we're not
7164 ;; looking at either `c-after-suffixed-type-decl-key' or "[;,]".
7165 ;; Since there's no preceding type, the shift would mean that
7166 ;; the declaration is typeless. But if the regexp doesn't match
7167 ;; then we will simply fall through in the tests below and not
7168 ;; recognize it at all, so it's better to try it as an abstract
7169 ;; declarator instead.
7170 (c-fdoc-shift-type-backward)
7171
7172 ;; Still no identifier.
a85fd6da 7173 ;; CASE 8
0386b551
AM
7174 (when (and got-prefix (or got-parens got-suffix))
7175 ;; Require `got-prefix' together with either `got-parens' or
7176 ;; `got-suffix' to recognize it as an abstract declarator:
7177 ;; `got-parens' only is probably an empty function call.
7178 ;; `got-suffix' only can build an ordinary expression together
7179 ;; with the preceding identifier which we've taken as a type.
7180 ;; We could actually accept on `got-prefix' only, but that can
7181 ;; easily occur temporarily while writing an expression so we
7182 ;; avoid that case anyway. We could do a better job if we knew
7183 ;; the point when the fontification was invoked.
7184 (throw 'at-decl-or-cast t))
7185
a85fd6da 7186 ;; CASE 9
0386b551
AM
7187 (when (and at-type
7188 (not got-prefix)
7189 (not got-parens)
7190 got-suffix-after-parens
7191 (eq (char-after got-suffix-after-parens) ?\())
7192 ;; Got a type, no declarator but a paren suffix. I.e. it's a
fa463103 7193 ;; normal function call after all (or perhaps a C++ style object
0386b551
AM
7194 ;; instantiation expression).
7195 (throw 'at-decl-or-cast nil))))
7196
a85fd6da 7197 ;; CASE 10
0386b551
AM
7198 (when at-decl-or-cast
7199 ;; By now we've located the type in the declaration that we know
7200 ;; we're in.
7201 (throw 'at-decl-or-cast t))
7202
a85fd6da 7203 ;; CASE 11
0386b551
AM
7204 (when (and got-identifier
7205 (not context)
7206 (looking-at c-after-suffixed-type-decl-key)
7207 (if (and got-parens
7208 (not got-prefix)
7209 (not got-suffix)
7210 (not (eq at-type t)))
7211 ;; Shift the type backward in the case that there's a
7212 ;; single identifier inside parens. That can only
7213 ;; occur in K&R style function declarations so it's
7214 ;; more likely that it really is a function call.
7215 ;; Therefore we only do this after
7216 ;; `c-after-suffixed-type-decl-key' has matched.
7217 (progn (c-fdoc-shift-type-backward) t)
7218 got-suffix-after-parens))
7219 ;; A declaration according to `c-after-suffixed-type-decl-key'.
7220 (throw 'at-decl-or-cast t))
7221
a85fd6da 7222 ;; CASE 12
0386b551
AM
7223 (when (and (or got-prefix (not got-parens))
7224 (memq at-type '(t known)))
7225 ;; It's a declaration if a known type precedes it and it can't be a
7226 ;; function call.
7227 (throw 'at-decl-or-cast t))
7228
7229 ;; If we get here we can't tell if this is a type decl or a normal
7230 ;; expression by looking at it alone. (That's under the assumption
7231 ;; that normal expressions always can look like type decl expressions,
7232 ;; which isn't really true but the cases where it doesn't hold are so
7233 ;; uncommon (e.g. some placements of "const" in C++) it's not worth
7234 ;; the effort to look for them.)
7235
3af1c868 7236;;; 2008-04-16: commented out the next form, to allow the function to recognize
2cdb7698
AM
7237;;; "foo (int bar)" in CC (an implicit type (in class foo) without a semicolon)
7238;;; as a(n almost complete) declaration, enabling it to be fontified.
7239 ;; CASE 13
7240 ;; (unless (or at-decl-end (looking-at "=[^=]"))
7241 ;; If this is a declaration it should end here or its initializer(*)
7242 ;; should start here, so check for allowed separation tokens. Note
7243 ;; that this rule doesn't work e.g. with a K&R arglist after a
7244 ;; function header.
7245 ;;
7246 ;; *) Don't check for C++ style initializers using parens
7247 ;; since those already have been matched as suffixes.
7248 ;;
7249 ;; If `at-decl-or-cast' is then we've found some other sign that
7250 ;; it's a declaration or cast, so then it's probably an
7251 ;; invalid/unfinished one.
7252 ;; (throw 'at-decl-or-cast at-decl-or-cast))
0386b551
AM
7253
7254 ;; Below are tests that only should be applied when we're certain to
7255 ;; not have parsed halfway through an expression.
7256
a85fd6da 7257 ;; CASE 14
0386b551
AM
7258 (when (memq at-type '(t known))
7259 ;; The expression starts with a known type so treat it as a
7260 ;; declaration.
7261 (throw 'at-decl-or-cast t))
7262
a85fd6da 7263 ;; CASE 15
0386b551
AM
7264 (when (and (c-major-mode-is 'c++-mode)
7265 ;; In C++ we check if the identifier is a known type, since
7266 ;; (con|de)structors use the class name as identifier.
7267 ;; We've always shifted over the identifier as a type and
7268 ;; then backed up again in this case.
7269 identifier-type
7270 (or (memq identifier-type '(found known))
7271 (and (eq (char-after identifier-start) ?~)
7272 ;; `at-type' probably won't be 'found for
7273 ;; destructors since the "~" is then part of the
7274 ;; type name being checked against the list of
7275 ;; known types, so do a check without that
7276 ;; operator.
7277 (or (save-excursion
7278 (goto-char (1+ identifier-start))
7279 (c-forward-syntactic-ws)
7280 (c-with-syntax-table
7281 c-identifier-syntax-table
7282 (looking-at c-known-type-key)))
7283 (save-excursion
7284 (goto-char (1+ identifier-start))
7285 ;; We have already parsed the type earlier,
7286 ;; so it'd be possible to cache the end
7287 ;; position instead of redoing it here, but
7288 ;; then we'd need to keep track of another
7289 ;; position everywhere.
7290 (c-check-type (point)
7291 (progn (c-forward-type)
7292 (point))))))))
7293 (throw 'at-decl-or-cast t))
7294
7295 (if got-identifier
7296 (progn
a85fd6da 7297 ;; CASE 16
0386b551
AM
7298 (when (and got-prefix-before-parens
7299 at-type
7300 (or at-decl-end (looking-at "=[^=]"))
7301 (not context)
7302 (not got-suffix))
7303 ;; Got something like "foo * bar;". Since we're not inside an
7304 ;; arglist it would be a meaningless expression because the
7305 ;; result isn't used. We therefore choose to recognize it as
7306 ;; a declaration. Do not allow a suffix since it could then
7307 ;; be a function call.
7308 (throw 'at-decl-or-cast t))
7309
a85fd6da 7310 ;; CASE 17
0386b551
AM
7311 (when (and (or got-suffix-after-parens
7312 (looking-at "=[^=]"))
7313 (eq at-type 'found)
7314 (not (eq context 'arglist)))
7315 ;; Got something like "a (*b) (c);" or "a (b) = c;". It could
7316 ;; be an odd expression or it could be a declaration. Treat
7317 ;; it as a declaration if "a" has been used as a type
7318 ;; somewhere else (if it's a known type we won't get here).
7319 (throw 'at-decl-or-cast t)))
7320
a85fd6da 7321 ;; CASE 18
0386b551
AM
7322 (when (and context
7323 (or got-prefix
7324 (and (eq context 'decl)
7325 (not c-recognize-paren-inits)
7326 (or got-parens got-suffix))))
7327 ;; Got a type followed by an abstract declarator. If `got-prefix'
7328 ;; is set it's something like "a *" without anything after it. If
7329 ;; `got-parens' or `got-suffix' is set it's "a()", "a[]", "a()[]",
7330 ;; or similar, which we accept only if the context rules out
7331 ;; expressions.
7332 (throw 'at-decl-or-cast t)))
7333
7334 ;; If we had a complete symbol table here (which rules out
7335 ;; `c-found-types') we should return t due to the disambiguation rule
7336 ;; (in at least C++) that anything that can be parsed as a declaration
7337 ;; is a declaration. Now we're being more defensive and prefer to
7338 ;; highlight things like "foo (bar);" as a declaration only if we're
7339 ;; inside an arglist that contains declarations.
7340 (eq context 'decl))))
7341
7342 ;; The point is now after the type decl expression.
7343
7344 (cond
7345 ;; Check for a cast.
7346 ((save-excursion
7347 (and
7348 c-cast-parens
7349
7350 ;; Should be the first type/identifier in a cast paren.
7351 (> preceding-token-end (point-min))
7352 (memq (char-before preceding-token-end) c-cast-parens)
7353
7354 ;; The closing paren should follow.
7355 (progn
7356 (c-forward-syntactic-ws)
7357 (looking-at "\\s\)"))
7358
7359 ;; There should be a primary expression after it.
7360 (let (pos)
7361 (forward-char)
7362 (c-forward-syntactic-ws)
7363 (setq cast-end (point))
7364 (and (looking-at c-primary-expr-regexp)
7365 (progn
7366 (setq pos (match-end 0))
7367 (or
7368 ;; Check if the expression begins with a prefix keyword.
7369 (match-beginning 2)
7370 (if (match-beginning 1)
7371 ;; Expression begins with an ambiguous operator. Treat
7372 ;; it as a cast if it's a type decl or if we've
7373 ;; recognized the type somewhere else.
7374 (or at-decl-or-cast
7375 (memq at-type '(t known found)))
7376 ;; Unless it's a keyword, it's the beginning of a primary
7377 ;; expression.
7378 (not (looking-at c-keywords-regexp)))))
7379 ;; If `c-primary-expr-regexp' matched a nonsymbol token, check
7380 ;; that it matched a whole one so that we don't e.g. confuse
7381 ;; the operator '-' with '->'. It's ok if it matches further,
7382 ;; though, since it e.g. can match the float '.5' while the
7383 ;; operator regexp only matches '.'.
7384 (or (not (looking-at c-nonsymbol-token-regexp))
7385 (<= (match-end 0) pos))))
7386
7387 ;; There should either be a cast before it or something that isn't an
7388 ;; identifier or close paren.
7389 (> preceding-token-end (point-min))
7390 (progn
7391 (goto-char (1- preceding-token-end))
7392 (or (eq (point) last-cast-end)
7393 (progn
7394 (c-backward-syntactic-ws)
7395 (if (< (skip-syntax-backward "w_") 0)
7396 ;; It's a symbol. Accept it only if it's one of the
7397 ;; keywords that can precede an expression (without
7398 ;; surrounding parens).
7399 (looking-at c-simple-stmt-key)
7400 (and
7401 ;; Check that it isn't a close paren (block close is ok,
7402 ;; though).
7403 (not (memq (char-before) '(?\) ?\])))
7404 ;; Check that it isn't a nonsymbol identifier.
7405 (not (c-on-identifier)))))))))
7406
7407 ;; Handle the cast.
7408 (when (and c-record-type-identifiers at-type (not (eq at-type t)))
7409 (let ((c-promote-possible-types t))
7410 (goto-char type-start)
7411 (c-forward-type)))
7412
7413 (goto-char cast-end)
7414 'cast)
7415
7416 (at-decl-or-cast
7417 ;; We're at a declaration. Highlight the type and the following
7418 ;; declarators.
7419
7420 (when backup-if-not-cast
7421 (c-fdoc-shift-type-backward t))
7422
7423 (when (and (eq context 'decl) (looking-at ","))
7424 ;; Make sure to propagate the `c-decl-arg-start' property to
7425 ;; the next argument if it's set in this one, to cope with
7426 ;; interactive refontification.
7427 (c-put-c-type-property (point) 'c-decl-arg-start))
7428
ee3ce8a7
AM
7429 (when (and c-record-type-identifiers at-type ;; (not (eq at-type t))
7430 ;; There seems no reason to exclude a token from
7431 ;; fontification just because it's "a known type that can't
7432 ;; be a name or other expression". 2013-09-18.
7433 )
0386b551
AM
7434 (let ((c-promote-possible-types t))
7435 (save-excursion
7436 (goto-char type-start)
7437 (c-forward-type))))
7438
e15f8aaa
AM
7439 (cons id-start
7440 (and (or at-type-decl at-typedef)
7441 (cons at-type-decl at-typedef))))
0386b551
AM
7442
7443 (t
7444 ;; False alarm. Restore the recorded ranges.
7445 (setq c-record-type-identifiers save-rec-type-ids
7446 c-record-ref-identifiers save-rec-ref-ids)
7447 nil))))
7448
7449(defun c-forward-label (&optional assume-markup preceding-token-end limit)
51c9af45 7450 ;; Assuming that point is at the beginning of a token, check if it starts a
1379f2c5
AM
7451 ;; label and if so move over it and return non-nil (t in default situations,
7452 ;; specific symbols (see below) for interesting situations), otherwise don't
7453 ;; move and return nil. "Label" here means "most things with a colon".
51c9af45
AM
7454 ;;
7455 ;; More precisely, a "label" is regarded as one of:
1379f2c5
AM
7456 ;; (i) a goto target like "foo:" - returns the symbol `goto-target';
7457 ;; (ii) A case label - either the entire construct "case FOO:", or just the
7458 ;; bare "case", should the colon be missing. We return t;
7459 ;; (iii) a keyword which needs a colon, like "default:" or "private:"; We
7460 ;; return t;
51c9af45 7461 ;; (iv) One of QT's "extended" C++ variants of
1379f2c5
AM
7462 ;; "private:"/"protected:"/"public:"/"more:" looking like "public slots:".
7463 ;; Returns the symbol `qt-2kwds-colon'.
7464 ;; (v) QT's construct "signals:". Returns the symbol `qt-1kwd-colon'.
82ba65cf 7465 ;; (vi) One of the keywords matched by `c-opt-extra-label-key' (without any
51c9af45 7466 ;; colon). Currently (2006-03), this applies only to Objective C's
1379f2c5 7467 ;; keywords "@private", "@protected", and "@public". Returns t.
51c9af45 7468 ;;
e1dbe924 7469 ;; One of the things which will NOT be recognized as a label is a bit-field
51c9af45
AM
7470 ;; element of a struct, something like "int foo:5".
7471 ;;
7472 ;; The end of the label is taken to be just after the colon, or the end of
7473 ;; the first submatch in `c-opt-extra-label-key'. The point is directly
7474 ;; after the end on return. The terminating char gets marked with
7475 ;; `c-decl-end' to improve recognition of the following declaration or
7476 ;; statement.
0386b551
AM
7477 ;;
7478 ;; If ASSUME-MARKUP is non-nil, it's assumed that the preceding
51c9af45 7479 ;; label, if any, has already been marked up like that.
0386b551
AM
7480 ;;
7481 ;; If PRECEDING-TOKEN-END is given, it should be the first position
7482 ;; after the preceding token, i.e. on the other side of the
7483 ;; syntactic ws from the point. Use a value less than or equal to
7484 ;; (point-min) if the point is at the first token in (the visible
7485 ;; part of) the buffer.
7486 ;;
7487 ;; The optional LIMIT limits the forward scan for the colon.
7488 ;;
7489 ;; This function records the ranges of the label symbols on
7490 ;; `c-record-ref-identifiers' if `c-record-type-identifiers' (!) is
7491 ;; non-nil.
7492 ;;
7493 ;; This function might do hidden buffer changes.
7494
51c9af45 7495 (let ((start (point))
1379f2c5 7496 label-end
51c9af45 7497 qt-symbol-idx
1379f2c5 7498 macro-start ; if we're in one.
f412a567
AM
7499 label-type
7500 kwd)
0386b551 7501 (cond
b414f371 7502 ;; "case" or "default" (Doesn't apply to AWK).
0386b551
AM
7503 ((looking-at c-label-kwds-regexp)
7504 (let ((kwd-end (match-end 1)))
7505 ;; Record only the keyword itself for fontification, since in
7506 ;; case labels the following is a constant expression and not
7507 ;; a label.
7508 (when c-record-type-identifiers
7509 (c-record-ref-id (cons (match-beginning 1) kwd-end)))
7510
7511 ;; Find the label end.
7512 (goto-char kwd-end)
1379f2c5
AM
7513 (setq label-type
7514 (if (and (c-syntactic-re-search-forward
7515 ;; Stop on chars that aren't allowed in expressions,
7516 ;; and on operator chars that would be meaningless
7517 ;; there. FIXME: This doesn't cope with ?: operators.
7518 "[;{=,@]\\|\\(\\=\\|[^:]\\):\\([^:]\\|\\'\\)"
7519 limit t t nil 1)
7520 (match-beginning 2))
7521
7522 (progn ; there's a proper :
7523 (goto-char (match-beginning 2)) ; just after the :
7524 (c-put-c-type-property (1- (point)) 'c-decl-end)
7525 t)
7526
7527 ;; It's an unfinished label. We consider the keyword enough
7528 ;; to recognize it as a label, so that it gets fontified.
7529 ;; Leave the point at the end of it, but don't put any
7530 ;; `c-decl-end' marker.
7531 (goto-char kwd-end)
7532 t))))
0386b551 7533
51c9af45 7534 ;; @private, @protected, @public, in Objective C, or similar.
0386b551
AM
7535 ((and c-opt-extra-label-key
7536 (looking-at c-opt-extra-label-key))
7537 ;; For a `c-opt-extra-label-key' match, we record the whole
7538 ;; thing for fontification. That's to get the leading '@' in
7539 ;; Objective-C protection labels fontified.
7540 (goto-char (match-end 1))
7541 (when c-record-type-identifiers
7542 (c-record-ref-id (cons (match-beginning 1) (point))))
7543 (c-put-c-type-property (1- (point)) 'c-decl-end)
1379f2c5 7544 (setq label-type t))
0386b551 7545
51c9af45
AM
7546 ;; All other cases of labels.
7547 ((and c-recognize-colon-labels ; nil for AWK and IDL, otherwise t.
0386b551
AM
7548
7549 ;; A colon label must have something before the colon.
7550 (not (eq (char-after) ?:))
7551
7552 ;; Check that we're not after a token that can't precede a label.
7553 (or
7554 ;; Trivially succeeds when there's no preceding token.
536610a4 7555 ;; Succeeds when we're at a virtual semicolon.
0386b551
AM
7556 (if preceding-token-end
7557 (<= preceding-token-end (point-min))
7558 (save-excursion
7559 (c-backward-syntactic-ws)
7560 (setq preceding-token-end (point))
536610a4
AM
7561 (or (bobp)
7562 (c-at-vsemi-p))))
0386b551
AM
7563
7564 ;; Check if we're after a label, if we're after a closing
7565 ;; paren that belong to statement, and with
7566 ;; `c-label-prefix-re'. It's done in different order
7567 ;; depending on `assume-markup' since the checks have
7568 ;; different expensiveness.
7569 (if assume-markup
7570 (or
7571 (eq (c-get-char-property (1- preceding-token-end) 'c-type)
7572 'c-decl-end)
7573
7574 (save-excursion
7575 (goto-char (1- preceding-token-end))
7576 (c-beginning-of-current-token)
51c9af45
AM
7577 (or (looking-at c-label-prefix-re)
7578 (looking-at c-block-stmt-1-key)))
0386b551
AM
7579
7580 (and (eq (char-before preceding-token-end) ?\))
7581 (c-after-conditional)))
7582
7583 (or
7584 (save-excursion
7585 (goto-char (1- preceding-token-end))
7586 (c-beginning-of-current-token)
51c9af45
AM
7587 (or (looking-at c-label-prefix-re)
7588 (looking-at c-block-stmt-1-key)))
0386b551
AM
7589
7590 (cond
7591 ((eq (char-before preceding-token-end) ?\))
7592 (c-after-conditional))
7593
7594 ((eq (char-before preceding-token-end) ?:)
7595 ;; Might be after another label, so check it recursively.
51c9af45
AM
7596 (save-restriction
7597 (save-excursion
7598 (goto-char (1- preceding-token-end))
7599 ;; Essentially the same as the
7600 ;; `c-syntactic-re-search-forward' regexp below.
7601 (setq macro-start
7602 (save-excursion (and (c-beginning-of-macro)
7603 (point))))
7604 (if macro-start (narrow-to-region macro-start (point-max)))
7605 (c-syntactic-skip-backward "^-]:?;}=*/%&|,<>!@+" nil t)
7606 ;; Note: the following should work instead of the
7607 ;; narrow-to-region above. Investigate why not,
7608 ;; sometime. ACM, 2006-03-31.
7609 ;; (c-syntactic-skip-backward "^-]:?;}=*/%&|,<>!@+"
7610 ;; macro-start t)
7611 (let ((pte (point))
7612 ;; If the caller turned on recording for us,
7613 ;; it shouldn't apply when we check the
7614 ;; preceding label.
7615 c-record-type-identifiers)
7616 ;; A label can't start at a cpp directive. Check for
7617 ;; this, since c-forward-syntactic-ws would foul up on it.
7618 (unless (and c-opt-cpp-prefix (looking-at c-opt-cpp-prefix))
7619 (c-forward-syntactic-ws)
7620 (c-forward-label nil pte start))))))))))
7621
1379f2c5 7622 ;; Point is still at the beginning of the possible label construct.
b414f371 7623 ;;
51c9af45
AM
7624 ;; Check that the next nonsymbol token is ":", or that we're in one
7625 ;; of QT's "slots" declarations. Allow '(' for the sake of macro
7626 ;; arguments. FIXME: Should build this regexp from the language
7627 ;; constants.
1379f2c5
AM
7628 (cond
7629 ;; public: protected: private:
7630 ((and
7631 (c-major-mode-is 'c++-mode)
7632 (search-forward-regexp
7633 "\\=p\\(r\\(ivate\\|otected\\)\\|ublic\\)\\>[^_]" nil t)
7634 (progn (backward-char)
7635 (c-forward-syntactic-ws limit)
7636 (looking-at ":\\([^:]\\|\\'\\)"))) ; A single colon.
7637 (forward-char)
7638 (setq label-type t))
7639 ;; QT double keyword like "protected slots:" or goto target.
7640 ((progn (goto-char start) nil))
7641 ((when (c-syntactic-re-search-forward
7642 "[ \t\n[:?;{=*/%&|,<>!@+-]" limit t t) ; not at EOB
7643 (backward-char)
7644 (setq label-end (point))
7645 (setq qt-symbol-idx
7646 (and (c-major-mode-is 'c++-mode)
7647 (string-match
7648 "\\(p\\(r\\(ivate\\|otected\\)\\|ublic\\)\\|more\\)\\>"
b414f371 7649 (buffer-substring start (point)))))
1379f2c5
AM
7650 (c-forward-syntactic-ws limit)
7651 (cond
7652 ((looking-at ":\\([^:]\\|\\'\\)") ; A single colon.
7653 (forward-char)
7654 (setq label-type
f412a567
AM
7655 (if (or (string= "signals" ; Special QT macro
7656 (setq kwd (buffer-substring-no-properties start label-end)))
7657 (string= "Q_SIGNALS" kwd))
1379f2c5
AM
7658 'qt-1kwd-colon
7659 'goto-target)))
7660 ((and qt-symbol-idx
f412a567 7661 (search-forward-regexp "\\=\\(slots\\|Q_SLOTS\\)\\>" limit t)
1379f2c5
AM
7662 (progn (c-forward-syntactic-ws limit)
7663 (looking-at ":\\([^:]\\|\\'\\)"))) ; A single colon
7664 (forward-char)
7665 (setq label-type 'qt-2kwds-colon)))))))
0386b551
AM
7666
7667 (save-restriction
7668 (narrow-to-region start (point))
7669
7670 ;; Check that `c-nonlabel-token-key' doesn't match anywhere.
7671 (catch 'check-label
7672 (goto-char start)
7673 (while (progn
7674 (when (looking-at c-nonlabel-token-key)
7675 (goto-char start)
1379f2c5 7676 (setq label-type nil)
0386b551
AM
7677 (throw 'check-label nil))
7678 (and (c-safe (c-forward-sexp)
7679 (c-forward-syntactic-ws)
7680 t)
7681 (not (eobp)))))
7682
7683 ;; Record the identifiers in the label for fontification, unless
7684 ;; it begins with `c-label-kwds' in which case the following
7685 ;; identifiers are part of a (constant) expression that
7686 ;; shouldn't be fontified.
7687 (when (and c-record-type-identifiers
7688 (progn (goto-char start)
7689 (not (looking-at c-label-kwds-regexp))))
7690 (while (c-syntactic-re-search-forward c-symbol-key nil t)
7691 (c-record-ref-id (cons (match-beginning 0)
7692 (match-end 0)))))
7693
7694 (c-put-c-type-property (1- (point-max)) 'c-decl-end)
82ba65cf 7695 (goto-char (point-max)))))
0386b551
AM
7696
7697 (t
7698 ;; Not a label.
1379f2c5
AM
7699 (goto-char start)))
7700 label-type))
0386b551
AM
7701
7702(defun c-forward-objc-directive ()
7703 ;; Assuming the point is at the beginning of a token, try to move
7704 ;; forward to the end of the Objective-C directive that starts
7705 ;; there. Return t if a directive was fully recognized, otherwise
7706 ;; the point is moved as far as one could be successfully parsed and
7707 ;; nil is returned.
7708 ;;
7709 ;; This function records identifier ranges on
7710 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
7711 ;; `c-record-type-identifiers' is non-nil.
7712 ;;
7713 ;; This function might do hidden buffer changes.
7714
7715 (let ((start (point))
7716 start-char
7717 (c-promote-possible-types t)
667ced3a 7718 lim
0386b551
AM
7719 ;; Turn off recognition of angle bracket arglists while parsing
7720 ;; types here since the protocol reference list might then be
7721 ;; considered part of the preceding name or superclass-name.
7722 c-recognize-<>-arglists)
7723
7724 (if (or
7725 (when (looking-at
7726 (eval-when-compile
7727 (c-make-keywords-re t
7728 (append (c-lang-const c-protection-kwds objc)
7729 '("@end"))
7730 'objc-mode)))
7731 (goto-char (match-end 1))
7732 t)
7733
7734 (and
7735 (looking-at
7736 (eval-when-compile
7737 (c-make-keywords-re t
7738 '("@interface" "@implementation" "@protocol")
7739 'objc-mode)))
7740
7741 ;; Handle the name of the class itself.
7742 (progn
cb694ab7
AM
7743; (c-forward-token-2) ; 2006/1/13 This doesn't move if the token's
7744; at EOB.
7745 (goto-char (match-end 0))
667ced3a 7746 (setq lim (point))
cb694ab7 7747 (c-skip-ws-forward)
0386b551
AM
7748 (c-forward-type))
7749
7750 (catch 'break
7751 ;; Look for ": superclass-name" or "( category-name )".
7752 (when (looking-at "[:\(]")
7753 (setq start-char (char-after))
7754 (forward-char)
7755 (c-forward-syntactic-ws)
7756 (unless (c-forward-type) (throw 'break nil))
7757 (when (eq start-char ?\()
7758 (unless (eq (char-after) ?\)) (throw 'break nil))
7759 (forward-char)
7760 (c-forward-syntactic-ws)))
7761
7762 ;; Look for a protocol reference list.
7763 (if (eq (char-after) ?<)
7764 (let ((c-recognize-<>-arglists t)
7765 (c-parse-and-markup-<>-arglists t)
7766 c-restricted-<>-arglists)
7767 (c-forward-<>-arglist t))
7768 t))))
7769
7770 (progn
667ced3a 7771 (c-backward-syntactic-ws lim)
0386b551
AM
7772 (c-clear-c-type-property start (1- (point)) 'c-decl-end)
7773 (c-put-c-type-property (1- (point)) 'c-decl-end)
7774 t)
7775
7776 (c-clear-c-type-property start (point) 'c-decl-end)
7777 nil)))
7778
785eecbb
RS
7779(defun c-beginning-of-inheritance-list (&optional lim)
7780 ;; Go to the first non-whitespace after the colon that starts a
7781 ;; multiple inheritance introduction. Optional LIM is the farthest
7782 ;; back we should search.
0386b551
AM
7783 ;;
7784 ;; This function might do hidden buffer changes.
7785 (c-with-syntax-table c++-template-syntax-table
7786 (c-backward-token-2 0 t lim)
7787 (while (and (or (looking-at c-symbol-start)
7788 (looking-at "[<,]\\|::"))
7789 (zerop (c-backward-token-2 1 t lim))))))
785eecbb 7790
785eecbb
RS
7791(defun c-in-method-def-p ()
7792 ;; Return nil if we aren't in a method definition, otherwise the
7793 ;; position of the initial [+-].
0386b551
AM
7794 ;;
7795 ;; This function might do hidden buffer changes.
785eecbb
RS
7796 (save-excursion
7797 (beginning-of-line)
a66cd3ee
MS
7798 (and c-opt-method-key
7799 (looking-at c-opt-method-key)
785eecbb
RS
7800 (point))
7801 ))
7802
a66cd3ee
MS
7803;; Contributed by Kevin Ryde <user42@zip.com.au>.
7804(defun c-in-gcc-asm-p ()
7805 ;; Return non-nil if point is within a gcc \"asm\" block.
7806 ;;
7807 ;; This should be called with point inside an argument list.
7808 ;;
7809 ;; Only one level of enclosing parentheses is considered, so for
7810 ;; instance `nil' is returned when in a function call within an asm
7811 ;; operand.
0386b551
AM
7812 ;;
7813 ;; This function might do hidden buffer changes.
a66cd3ee
MS
7814
7815 (and c-opt-asm-stmt-key
7816 (save-excursion
7817 (beginning-of-line)
7818 (backward-up-list 1)
7819 (c-beginning-of-statement-1 (point-min) nil t)
7820 (looking-at c-opt-asm-stmt-key))))
7821
abb7e5cf 7822(defun c-at-toplevel-p ()
a85fd6da
AM
7823 "Return a determination as to whether point is \"at the top level\".
7824Informally, \"at the top level\" is anywhere where you can write
7825a function.
7826
7827More precisely, being at the top-level means that point is either
7828outside any enclosing block (such as a function definition), or
7829directly inside a class, namespace or other block that contains
7830another declaration level.
abb7e5cf
SM
7831
7832If point is not at the top-level (e.g. it is inside a method
7833definition), then nil is returned. Otherwise, if point is at a
7834top-level not enclosed within a class definition, t is returned.
7835Otherwise, a 2-vector is returned where the zeroth element is the
7836buffer position of the start of the class declaration, and the first
7837element is the buffer position of the enclosing class's opening
0386b551
AM
7838brace.
7839
7840Note that this function might do hidden buffer changes. See the
7841comment at the start of cc-engine.el for more info."
a66cd3ee
MS
7842 (let ((paren-state (c-parse-state)))
7843 (or (not (c-most-enclosing-brace paren-state))
7844 (c-search-uplist-for-classkey paren-state))))
7845
d9e94c22 7846(defun c-just-after-func-arglist-p (&optional lim)
0386b551
AM
7847 ;; Return non-nil if the point is in the region after the argument
7848 ;; list of a function and its opening brace (or semicolon in case it
7849 ;; got no body). If there are K&R style argument declarations in
7850 ;; that region, the point has to be inside the first one for this
7851 ;; function to recognize it.
a66cd3ee 7852 ;;
0386b551
AM
7853 ;; If successful, the point is moved to the first token after the
7854 ;; function header (see `c-forward-decl-or-cast-1' for details) and
7855 ;; the position of the opening paren of the function arglist is
7856 ;; returned.
7857 ;;
7858 ;; The point is clobbered if not successful.
7859 ;;
7860 ;; LIM is used as bound for backward buffer searches.
7861 ;;
7862 ;; This function might do hidden buffer changes.
7863
7864 (let ((beg (point)) end id-start)
7865 (and
7866 (eq (c-beginning-of-statement-1 lim) 'same)
7867
2e492df3
AM
7868 (not (and (c-major-mode-is 'objc-mode)
7869 (c-forward-objc-directive)))
0386b551
AM
7870
7871 (setq id-start
7872 (car-safe (c-forward-decl-or-cast-1 (c-point 'bosws) nil nil)))
7873 (< id-start beg)
7874
7875 ;; There should not be a '=' or ',' between beg and the
7876 ;; start of the declaration since that means we were in the
7877 ;; "expression part" of the declaration.
7878 (or (> (point) beg)
7879 (not (looking-at "[=,]")))
7880
7881 (save-excursion
7882 ;; Check that there's an arglist paren in the
7883 ;; declaration.
7884 (goto-char id-start)
7885 (cond ((eq (char-after) ?\()
7886 ;; The declarator is a paren expression, so skip past it
7887 ;; so that we don't get stuck on that instead of the
7888 ;; function arglist.
7889 (c-forward-sexp))
51c9af45
AM
7890 ((and c-opt-op-identifier-prefix
7891 (looking-at c-opt-op-identifier-prefix))
0386b551
AM
7892 ;; Don't trip up on "operator ()".
7893 (c-forward-token-2 2 t)))
7894 (and (< (point) beg)
7895 (c-syntactic-re-search-forward "(" beg t t)
7896 (1- (point)))))))
785eecbb 7897
a66cd3ee
MS
7898(defun c-in-knr-argdecl (&optional lim)
7899 ;; Return the position of the first argument declaration if point is
7900 ;; inside a K&R style argument declaration list, nil otherwise.
7901 ;; `c-recognize-knr-p' is not checked. If LIM is non-nil, it's a
7902 ;; position that bounds the backward search for the argument list.
7903 ;;
9cf17ef1
AM
7904 ;; Point must be within a possible K&R region, e.g. just before a top-level
7905 ;; "{". It must be outside of parens and brackets. The test can return
7906 ;; false positives otherwise.
0386b551
AM
7907 ;;
7908 ;; This function might do hidden buffer changes.
d9e94c22 7909
a66cd3ee
MS
7910 (save-excursion
7911 (save-restriction
9cf17ef1
AM
7912 ;; If we're in a macro, our search range is restricted to it. Narrow to
7913 ;; the searchable range.
98d7371e
AM
7914 (let* ((macro-start (save-excursion (and (c-beginning-of-macro) (point))))
7915 (macro-end (save-excursion (and macro-start (c-end-of-macro) (point))))
7916 (low-lim (max (or lim (point-min)) (or macro-start (point-min))))
6b6481ed 7917 before-lparen after-rparen
98d7371e
AM
7918 (pp-count-out 20)) ; Max number of paren/brace constructs before
7919 ; we give up
7920 (narrow-to-region low-lim (or macro-end (point-max)))
9cf17ef1
AM
7921
7922 ;; Search backwards for the defun's argument list. We give up if we
98d7371e
AM
7923 ;; encounter a "}" (end of a previous defun) an "=" (which can't be in
7924 ;; a knr region) or BOB.
9cf17ef1
AM
7925 ;;
7926 ;; The criterion for a paren structure being the arg list is:
7927 ;; o - there is non-WS stuff after it but before any "{"; AND
7928 ;; o - the token after it isn't a ";" AND
7929 ;; o - it is preceded by either an identifier (the function name) or
7930 ;; a macro expansion like "DEFUN (...)"; AND
7931 ;; o - its content is a non-empty comma-separated list of identifiers
7932 ;; (an empty arg list won't have a knr region).
7933 ;;
7934 ;; The following snippet illustrates these rules:
7935 ;; int foo (bar, baz, yuk)
7936 ;; int bar [] ;
7937 ;; int (*baz) (my_type) ;
7938 ;; int (*) (void) (*yuk) (void) ;
7939 ;; {
7940
7941 (catch 'knr
6b6481ed
AM
7942 (while (> pp-count-out 0) ; go back one paren/bracket pair each time.
7943 (setq pp-count-out (1- pp-count-out))
98d7371e 7944 (c-syntactic-skip-backward "^)]}=")
9cf17ef1
AM
7945 (cond ((eq (char-before) ?\))
7946 (setq after-rparen (point)))
10489fcb
AM
7947 ((eq (char-before) ?\])
7948 (setq after-rparen nil))
98d7371e
AM
7949 (t ; either } (hit previous defun) or = or no more
7950 ; parens/brackets.
10489fcb 7951 (throw 'knr nil)))
9cf17ef1
AM
7952
7953 (if after-rparen
7954 ;; We're inside a paren. Could it be our argument list....?
7955 (if
7956 (and
7957 (progn
7958 (goto-char after-rparen)
7959 (unless (c-go-list-backward) (throw 'knr nil)) ;
7960 ;; FIXME!!! What about macros between the parens? 2007/01/20
7961 (setq before-lparen (point)))
d9e94c22 7962
9cf17ef1
AM
7963 ;; It can't be the arg list if next token is ; or {
7964 (progn (goto-char after-rparen)
7965 (c-forward-syntactic-ws)
98d7371e 7966 (not (memq (char-after) '(?\; ?\{ ?\=))))
d9e94c22 7967
9cf17ef1
AM
7968 ;; Is the thing preceding the list an identifier (the
7969 ;; function name), or a macro expansion?
7970 (progn
7971 (goto-char before-lparen)
7972 (eq (c-backward-token-2) 0)
98d7371e 7973 (or (eq (c-on-identifier) (point))
9cf17ef1
AM
7974 (and (eq (char-after) ?\))
7975 (c-go-up-list-backward)
7976 (eq (c-backward-token-2) 0)
98d7371e 7977 (eq (c-on-identifier) (point)))))
9cf17ef1
AM
7978
7979 ;; Have we got a non-empty list of comma-separated
7980 ;; identifiers?
7981 (progn
7982 (goto-char before-lparen)
7983 (c-forward-token-2) ; to first token inside parens
7984 (and
7985 (c-on-identifier)
7986 (c-forward-token-2)
7987 (catch 'id-list
7988 (while (eq (char-after) ?\,)
7989 (c-forward-token-2)
7990 (unless (c-on-identifier) (throw 'id-list nil))
7991 (c-forward-token-2))
7992 (eq (char-after) ?\))))))
7993
7994 ;; ...Yes. We've identified the function's argument list.
7995 (throw 'knr
7996 (progn (goto-char after-rparen)
7997 (c-forward-syntactic-ws)
7998 (point)))
7999
8000 ;; ...No. The current parens aren't the function's arg list.
8001 (goto-char before-lparen))
8002
8003 (or (c-go-list-backward) ; backwards over [ .... ]
8004 (throw 'knr nil)))))))))
785eecbb
RS
8005
8006(defun c-skip-conditional ()
8007 ;; skip forward over conditional at point, including any predicate
8008 ;; statements in parentheses. No error checking is performed.
0386b551
AM
8009 ;;
8010 ;; This function might do hidden buffer changes.
0ec8351b
BW
8011 (c-forward-sexp (cond
8012 ;; else if()
a66cd3ee
MS
8013 ((looking-at (concat "\\<else"
8014 "\\([ \t\n]\\|\\\\\n\\)+"
8015 "if\\>\\([^_]\\|$\\)"))
8016 3)
0ec8351b 8017 ;; do, else, try, finally
a66cd3ee
MS
8018 ((looking-at (concat "\\<\\("
8019 "do\\|else\\|try\\|finally"
8020 "\\)\\>\\([^_]\\|$\\)"))
130c507e 8021 1)
ce8c7486 8022 ;; for, if, while, switch, catch, synchronized, foreach
0ec8351b 8023 (t 2))))
785eecbb 8024
a66cd3ee
MS
8025(defun c-after-conditional (&optional lim)
8026 ;; If looking at the token after a conditional then return the
8027 ;; position of its start, otherwise return nil.
0386b551
AM
8028 ;;
8029 ;; This function might do hidden buffer changes.
a66cd3ee 8030 (save-excursion
d9e94c22 8031 (and (zerop (c-backward-token-2 1 t lim))
a66cd3ee
MS
8032 (or (looking-at c-block-stmt-1-key)
8033 (and (eq (char-after) ?\()
d9e94c22 8034 (zerop (c-backward-token-2 1 t lim))
5ab78d3d
AM
8035 (or (looking-at c-block-stmt-2-key)
8036 (looking-at c-block-stmt-1-2-key))))
a66cd3ee
MS
8037 (point))))
8038
0386b551
AM
8039(defun c-after-special-operator-id (&optional lim)
8040 ;; If the point is after an operator identifier that isn't handled
8041 ;; like an ordinary symbol (i.e. like "operator =" in C++) then the
8042 ;; position of the start of that identifier is returned. nil is
8043 ;; returned otherwise. The point may be anywhere in the syntactic
8044 ;; whitespace after the last token of the operator identifier.
8045 ;;
8046 ;; This function might do hidden buffer changes.
8047 (save-excursion
8048 (and c-overloadable-operators-regexp
8049 (zerop (c-backward-token-2 1 nil lim))
8050 (looking-at c-overloadable-operators-regexp)
51c9af45 8051 (or (not c-opt-op-identifier-prefix)
0386b551
AM
8052 (and
8053 (zerop (c-backward-token-2 1 nil lim))
51c9af45 8054 (looking-at c-opt-op-identifier-prefix)))
0386b551
AM
8055 (point))))
8056
a66cd3ee
MS
8057(defsubst c-backward-to-block-anchor (&optional lim)
8058 ;; Assuming point is at a brace that opens a statement block of some
8059 ;; kind, move to the proper anchor point for that block. It might
8060 ;; need to be adjusted further by c-add-stmt-syntax, but the
8061 ;; position at return is suitable as start position for that
8062 ;; function.
0386b551
AM
8063 ;;
8064 ;; This function might do hidden buffer changes.
a66cd3ee
MS
8065 (unless (= (point) (c-point 'boi))
8066 (let ((start (c-after-conditional lim)))
8067 (if start
8068 (goto-char start)))))
8069
037558bf 8070(defsubst c-backward-to-decl-anchor (&optional lim)
a66cd3ee
MS
8071 ;; Assuming point is at a brace that opens the block of a top level
8072 ;; declaration of some kind, move to the proper anchor point for
8073 ;; that block.
0386b551
AM
8074 ;;
8075 ;; This function might do hidden buffer changes.
a66cd3ee 8076 (unless (= (point) (c-point 'boi))
037558bf 8077 (c-beginning-of-statement-1 lim)))
a66cd3ee 8078
ff959bab 8079(defun c-search-decl-header-end ()
a66cd3ee
MS
8080 ;; Search forward for the end of the "header" of the current
8081 ;; declaration. That's the position where the definition body
8082 ;; starts, or the first variable initializer, or the ending
8083 ;; semicolon. I.e. search forward for the closest following
8084 ;; (syntactically relevant) '{', '=' or ';' token. Point is left
8085 ;; _after_ the first found token, or at point-max if none is found.
0386b551
AM
8086 ;;
8087 ;; This function might do hidden buffer changes.
ff959bab
MS
8088
8089 (let ((base (point)))
8090 (if (c-major-mode-is 'c++-mode)
8091
8092 ;; In C++ we need to take special care to handle operator
8093 ;; tokens and those pesky template brackets.
8094 (while (and
8095 (c-syntactic-re-search-forward "[;{<=]" nil 'move t t)
8096 (or
8097 (c-end-of-current-token base)
8098 ;; Handle operator identifiers, i.e. ignore any
8099 ;; operator token preceded by "operator".
8100 (save-excursion
8101 (and (c-safe (c-backward-sexp) t)
51c9af45 8102 (looking-at c-opt-op-identifier-prefix)))
ff959bab
MS
8103 (and (eq (char-before) ?<)
8104 (c-with-syntax-table c++-template-syntax-table
8105 (if (c-safe (goto-char (c-up-list-forward (point))))
8106 t
8107 (goto-char (point-max))
8108 nil)))))
8109 (setq base (point)))
8110
8111 (while (and
8112 (c-syntactic-re-search-forward "[;{=]" nil 'move t t)
8113 (c-end-of-current-token base))
8114 (setq base (point))))))
a66cd3ee
MS
8115
8116(defun c-beginning-of-decl-1 (&optional lim)
8117 ;; Go to the beginning of the current declaration, or the beginning
8118 ;; of the previous one if already at the start of it. Point won't
0386b551 8119 ;; be moved out of any surrounding paren. Return a cons cell of the
a66cd3ee
MS
8120 ;; form (MOVE . KNR-POS). MOVE is like the return value from
8121 ;; `c-beginning-of-statement-1'. If point skipped over some K&R
8122 ;; style argument declarations (and they are to be recognized) then
8123 ;; KNR-POS is set to the start of the first such argument
8124 ;; declaration, otherwise KNR-POS is nil. If LIM is non-nil, it's a
8125 ;; position that bounds the backward search.
8126 ;;
8127 ;; NB: Cases where the declaration continues after the block, as in
8128 ;; "struct foo { ... } bar;", are currently recognized as two
8129 ;; declarations, e.g. "struct foo { ... }" and "bar;" in this case.
0386b551
AM
8130 ;;
8131 ;; This function might do hidden buffer changes.
a66cd3ee
MS
8132 (catch 'return
8133 (let* ((start (point))
d9e94c22 8134 (last-stmt-start (point))
0386b551 8135 (move (c-beginning-of-statement-1 lim nil t)))
a66cd3ee 8136
a66cd3ee
MS
8137 ;; `c-beginning-of-statement-1' stops at a block start, but we
8138 ;; want to continue if the block doesn't begin a top level
2a15eb73
MS
8139 ;; construct, i.e. if it isn't preceded by ';', '}', ':', bob,
8140 ;; or an open paren.
d9e94c22 8141 (let ((beg (point)) tentative-move)
51c9af45
AM
8142 ;; Go back one "statement" each time round the loop until we're just
8143 ;; after a ;, }, or :, or at BOB or the start of a macro or start of
8144 ;; an ObjC method. This will move over a multiple declaration whose
8145 ;; components are comma separated.
d9e94c22
MS
8146 (while (and
8147 ;; Must check with c-opt-method-key in ObjC mode.
8148 (not (and c-opt-method-key
8149 (looking-at c-opt-method-key)))
8150 (/= last-stmt-start (point))
8151 (progn
8152 (c-backward-syntactic-ws lim)
8153 (not (memq (char-before) '(?\; ?} ?: nil))))
2a15eb73
MS
8154 (save-excursion
8155 (backward-char)
8156 (not (looking-at "\\s(")))
d9e94c22
MS
8157 ;; Check that we don't move from the first thing in a
8158 ;; macro to its header.
8159 (not (eq (setq tentative-move
0386b551 8160 (c-beginning-of-statement-1 lim nil t))
d9e94c22
MS
8161 'macro)))
8162 (setq last-stmt-start beg
8163 beg (point)
8164 move tentative-move))
8165 (goto-char beg))
8166
8167 (when c-recognize-knr-p
8168 (let ((fallback-pos (point)) knr-argdecl-start)
8169 ;; Handle K&R argdecls. Back up after the "statement" jumped
8170 ;; over by `c-beginning-of-statement-1', unless it was the
8171 ;; function body, in which case we're sitting on the opening
8172 ;; brace now. Then test if we're in a K&R argdecl region and
8173 ;; that we started at the other side of the first argdecl in
8174 ;; it.
8175 (unless (eq (char-after) ?{)
8176 (goto-char last-stmt-start))
8177 (if (and (setq knr-argdecl-start (c-in-knr-argdecl lim))
8178 (< knr-argdecl-start start)
8179 (progn
8180 (goto-char knr-argdecl-start)
0386b551 8181 (not (eq (c-beginning-of-statement-1 lim nil t) 'macro))))
d9e94c22
MS
8182 (throw 'return
8183 (cons (if (eq (char-after fallback-pos) ?{)
8184 'previous
8185 'same)
8186 knr-argdecl-start))
8187 (goto-char fallback-pos))))
8188
51c9af45
AM
8189 ;; `c-beginning-of-statement-1' counts each brace block as a separate
8190 ;; statement, so the result will be 'previous if we've moved over any.
8191 ;; So change our result back to 'same if necessary.
8192 ;;
8193 ;; If they were brace list initializers we might not have moved over a
8194 ;; declaration boundary though, so change it to 'same if we've moved
8195 ;; past a '=' before '{', but not ';'. (This ought to be integrated
8196 ;; into `c-beginning-of-statement-1', so we avoid this extra pass which
8197 ;; potentially can search over a large amount of text.). Take special
8198 ;; pains not to get mislead by C++'s "operator=", and the like.
d9e94c22
MS
8199 (if (and (eq move 'previous)
8200 (c-with-syntax-table (if (c-major-mode-is 'c++-mode)
8201 c++-template-syntax-table
8202 (syntax-table))
8203 (save-excursion
51c9af45
AM
8204 (and
8205 (progn
8206 (while ; keep going back to "[;={"s until we either find
8207 ; no more, or get to one which isn't an "operator ="
8208 (and (c-syntactic-re-search-forward "[;={]" start t t t)
8209 (eq (char-before) ?=)
8210 c-overloadable-operators-regexp
8211 c-opt-op-identifier-prefix
8212 (save-excursion
8213 (eq (c-backward-token-2) 0)
8214 (looking-at c-overloadable-operators-regexp)
8215 (eq (c-backward-token-2) 0)
8216 (looking-at c-opt-op-identifier-prefix))))
8217 (eq (char-before) ?=))
8218 (c-syntactic-re-search-forward "[;{]" start t t)
8219 (eq (char-before) ?{)
8220 (c-safe (goto-char (c-up-list-forward (point))) t)
8221 (not (c-syntactic-re-search-forward ";" start t t))))))
d9e94c22
MS
8222 (cons 'same nil)
8223 (cons move nil)))))
a66cd3ee
MS
8224
8225(defun c-end-of-decl-1 ()
8226 ;; Assuming point is at the start of a declaration (as detected by
8227 ;; e.g. `c-beginning-of-decl-1'), go to the end of it. Unlike
8228 ;; `c-beginning-of-decl-1', this function handles the case when a
8229 ;; block is followed by identifiers in e.g. struct declarations in C
8230 ;; or C++. If a proper end was found then t is returned, otherwise
8231 ;; point is moved as far as possible within the current sexp and nil
8232 ;; is returned. This function doesn't handle macros; use
8233 ;; `c-end-of-macro' instead in those cases.
0386b551
AM
8234 ;;
8235 ;; This function might do hidden buffer changes.
ce8c7486 8236 (let ((start (point))
a66cd3ee
MS
8237 (decl-syntax-table (if (c-major-mode-is 'c++-mode)
8238 c++-template-syntax-table
8239 (syntax-table))))
8240 (catch 'return
8241 (c-search-decl-header-end)
8242
8243 (when (and c-recognize-knr-p
8244 (eq (char-before) ?\;)
8245 (c-in-knr-argdecl start))
8246 ;; Stopped at the ';' in a K&R argdecl section which is
8247 ;; detected using the same criteria as in
8248 ;; `c-beginning-of-decl-1'. Move to the following block
8249 ;; start.
d9e94c22 8250 (c-syntactic-re-search-forward "{" nil 'move t))
a66cd3ee
MS
8251
8252 (when (eq (char-before) ?{)
8253 ;; Encountered a block in the declaration. Jump over it.
8254 (condition-case nil
8255 (goto-char (c-up-list-forward (point)))
d9e94c22
MS
8256 (error (goto-char (point-max))
8257 (throw 'return nil)))
a66cd3ee
MS
8258 (if (or (not c-opt-block-decls-with-vars-key)
8259 (save-excursion
8260 (c-with-syntax-table decl-syntax-table
8261 (let ((lim (point)))
8262 (goto-char start)
b3cf7e18
MS
8263 (not (and
8264 ;; Check for `c-opt-block-decls-with-vars-key'
8265 ;; before the first paren.
8266 (c-syntactic-re-search-forward
d9e94c22 8267 (concat "[;=\(\[{]\\|\\("
b3cf7e18
MS
8268 c-opt-block-decls-with-vars-key
8269 "\\)")
d9e94c22 8270 lim t t t)
b3cf7e18
MS
8271 (match-beginning 1)
8272 (not (eq (char-before) ?_))
d9e94c22
MS
8273 ;; Check that the first following paren is
8274 ;; the block.
8275 (c-syntactic-re-search-forward "[;=\(\[{]"
8276 lim t t t)
b3cf7e18 8277 (eq (char-before) ?{)))))))
a66cd3ee
MS
8278 ;; The declaration doesn't have any of the
8279 ;; `c-opt-block-decls-with-vars' keywords in the
8280 ;; beginning, so it ends here at the end of the block.
8281 (throw 'return t)))
8282
8283 (c-with-syntax-table decl-syntax-table
8284 (while (progn
8285 (if (eq (char-before) ?\;)
8286 (throw 'return t))
d9e94c22 8287 (c-syntactic-re-search-forward ";" nil 'move t))))
a66cd3ee 8288 nil)))
ce8c7486 8289
0386b551
AM
8290(defun c-looking-at-decl-block (containing-sexp goto-start &optional limit)
8291 ;; Assuming the point is at an open brace, check if it starts a
8292 ;; block that contains another declaration level, i.e. that isn't a
8293 ;; statement block or a brace list, and if so return non-nil.
8294 ;;
8295 ;; If the check is successful, the return value is the start of the
8296 ;; keyword that tells what kind of construct it is, i.e. typically
8297 ;; what `c-decl-block-key' matched. Also, if GOTO-START is set then
8298 ;; the point will be at the start of the construct, before any
8299 ;; leading specifiers, otherwise it's at the returned position.
8300 ;;
8301 ;; The point is clobbered if the check is unsuccessful.
8302 ;;
8303 ;; CONTAINING-SEXP is the position of the open of the surrounding
8304 ;; paren, or nil if none.
8305 ;;
8306 ;; The optional LIMIT limits the backward search for the start of
8307 ;; the construct. It's assumed to be at a syntactically relevant
8308 ;; position.
8309 ;;
8310 ;; If any template arglists are found in the searched region before
8311 ;; the open brace, they get marked with paren syntax.
8312 ;;
8313 ;; This function might do hidden buffer changes.
8314
8315 (let ((open-brace (point)) kwd-start first-specifier-pos)
8316 (c-syntactic-skip-backward c-block-prefix-charset limit t)
8317
8318 (when (and c-recognize-<>-arglists
8319 (eq (char-before) ?>))
8320 ;; Could be at the end of a template arglist.
8321 (let ((c-parse-and-markup-<>-arglists t)
8322 (c-disallow-comma-in-<>-arglists
8323 (and containing-sexp
8324 (not (eq (char-after containing-sexp) ?{)))))
8325 (while (and
8326 (c-backward-<>-arglist nil limit)
8327 (progn
8328 (c-syntactic-skip-backward c-block-prefix-charset limit t)
8329 (eq (char-before) ?>))))))
8330
8331 ;; Note: Can't get bogus hits inside template arglists below since they
8332 ;; have gotten paren syntax above.
8333 (when (and
8334 ;; If `goto-start' is set we begin by searching for the
8335 ;; first possible position of a leading specifier list.
8336 ;; The `c-decl-block-key' search continues from there since
8337 ;; we know it can't match earlier.
8338 (if goto-start
8339 (when (c-syntactic-re-search-forward c-symbol-start
8340 open-brace t t)
8341 (goto-char (setq first-specifier-pos (match-beginning 0)))
8342 t)
8343 t)
8344
8345 (cond
8346 ((c-syntactic-re-search-forward c-decl-block-key open-brace t t t)
8347 (goto-char (setq kwd-start (match-beginning 0)))
8348 (or
8349
8350 ;; Found a keyword that can't be a type?
8351 (match-beginning 1)
8352
8353 ;; Can be a type too, in which case it's the return type of a
8354 ;; function (under the assumption that no declaration level
8355 ;; block construct starts with a type).
8356 (not (c-forward-type))
8357
8358 ;; Jumped over a type, but it could be a declaration keyword
8359 ;; followed by the declared identifier that we've jumped over
8360 ;; instead (e.g. in "class Foo {"). If it indeed is a type
8361 ;; then we should be at the declarator now, so check for a
8362 ;; valid declarator start.
8363 ;;
8364 ;; Note: This doesn't cope with the case when a declared
8365 ;; identifier is followed by e.g. '(' in a language where '('
8366 ;; also might be part of a declarator expression. Currently
8367 ;; there's no such language.
8368 (not (or (looking-at c-symbol-start)
8369 (looking-at c-type-decl-prefix-key)))))
8370
8371 ;; In Pike a list of modifiers may be followed by a brace
8372 ;; to make them apply to many identifiers. Note that the
8373 ;; match data will be empty on return in this case.
8374 ((and (c-major-mode-is 'pike-mode)
8375 (progn
8376 (goto-char open-brace)
8377 (= (c-backward-token-2) 0))
8378 (looking-at c-specifier-key)
8379 ;; Use this variant to avoid yet another special regexp.
8380 (c-keyword-member (c-keyword-sym (match-string 1))
8381 'c-modifier-kwds))
8382 (setq kwd-start (point))
8383 t)))
8384
8385 ;; Got a match.
8386
8387 (if goto-start
8388 ;; Back up over any preceding specifiers and their clauses
8389 ;; by going forward from `first-specifier-pos', which is the
8390 ;; earliest possible position where the specifier list can
8391 ;; start.
8392 (progn
8393 (goto-char first-specifier-pos)
8394
8395 (while (< (point) kwd-start)
8396 (if (looking-at c-symbol-key)
8397 ;; Accept any plain symbol token on the ground that
8398 ;; it's a specifier masked through a macro (just
8399 ;; like `c-forward-decl-or-cast-1' skip forward over
8400 ;; such tokens).
8401 ;;
8402 ;; Could be more restrictive wrt invalid keywords,
8403 ;; but that'd only occur in invalid code so there's
8404 ;; no use spending effort on it.
8405 (let ((end (match-end 0)))
8406 (unless (c-forward-keyword-clause 0)
8407 (goto-char end)
8408 (c-forward-syntactic-ws)))
8409
8410 ;; Can't parse a declaration preamble and is still
8411 ;; before `kwd-start'. That means `first-specifier-pos'
8412 ;; was in some earlier construct. Search again.
8413 (if (c-syntactic-re-search-forward c-symbol-start
8414 kwd-start 'move t)
8415 (goto-char (setq first-specifier-pos (match-beginning 0)))
8416 ;; Got no preamble before the block declaration keyword.
8417 (setq first-specifier-pos kwd-start))))
8418
8419 (goto-char first-specifier-pos))
8420 (goto-char kwd-start))
8421
8422 kwd-start)))
ce8c7486 8423
a66cd3ee 8424(defun c-search-uplist-for-classkey (paren-state)
0386b551
AM
8425 ;; Check if the closest containing paren sexp is a declaration
8426 ;; block, returning a 2 element vector in that case. Aref 0
8427 ;; contains the bufpos at boi of the class key line, and aref 1
8428 ;; contains the bufpos of the open brace. This function is an
8429 ;; obsolete wrapper for `c-looking-at-decl-block'.
8430 ;;
8431 ;; This function might do hidden buffer changes.
8432 (let ((open-paren-pos (c-most-enclosing-brace paren-state)))
8433 (when open-paren-pos
8434 (save-excursion
8435 (goto-char open-paren-pos)
8436 (when (and (eq (char-after) ?{)
8437 (c-looking-at-decl-block
8438 (c-safe-position open-paren-pos paren-state)
8439 nil))
8440 (back-to-indentation)
8441 (vector (point) open-paren-pos))))))
785eecbb 8442
020716e1
AM
8443(defun c-most-enclosing-decl-block (paren-state)
8444 ;; Return the buffer position of the most enclosing decl-block brace (in the
8445 ;; sense of c-looking-at-decl-block) in the PAREN-STATE structure, or nil if
8446 ;; none was found.
8447 (let* ((open-brace (c-pull-open-brace paren-state))
8448 (next-open-brace (c-pull-open-brace paren-state)))
8449 (while (and open-brace
8450 (save-excursion
8451 (goto-char open-brace)
8452 (not (c-looking-at-decl-block next-open-brace nil))))
8453 (setq open-brace next-open-brace
8454 next-open-brace (c-pull-open-brace paren-state)))
8455 open-brace))
8456
9657183b
AM
8457(defun c-cheap-inside-bracelist-p (paren-state)
8458 ;; Return the position of the L-brace if point is inside a brace list
8459 ;; initialization of an array, etc. This is an approximate function,
8460 ;; designed for speed over accuracy. It will not find every bracelist, but
8461 ;; a non-nil result is reliable. We simply search for "= {" (naturally with
8462 ;; syntactic whitespace allowed). PAREN-STATE is the normal thing that it
8463 ;; is everywhere else.
8464 (let (b-pos)
8465 (save-excursion
8466 (while
8467 (and (setq b-pos (c-pull-open-brace paren-state))
8468 (progn (goto-char b-pos)
8469 (c-backward-sws)
8470 (c-backward-token-2)
8471 (not (looking-at "=")))))
8472 b-pos)))
8473
7a068717
AM
8474(defun c-backward-over-enum-header ()
8475 ;; We're at a "{". Move back to the enum-like keyword that starts this
8476 ;; declaration and return t, otherwise don't move and return nil.
8477 (let ((here (point))
8478 up-sexp-pos before-identifier)
8479 (while
8480 (and
8481 (eq (c-backward-token-2) 0)
8482 (or (not (looking-at "\\s)"))
8483 (c-go-up-list-backward))
8484 (cond
8485 ((and (looking-at c-symbol-key) (c-on-identifier))
8486 (setq before-identifier t))
8487 ((and before-identifier
8488 (looking-at c-postfix-decl-spec-key))
8489 (setq before-identifier nil)
8490 t)
8491 ((looking-at c-brace-list-key) nil)
8492 ((and c-recognize-<>-arglists
8493 (eq (char-after) ?<)
8494 (looking-at "\\s("))
8495 t)
8496 (t nil))))
8497 (or (looking-at c-brace-list-key)
8498 (progn (goto-char here) nil))))
8499
a66cd3ee 8500(defun c-inside-bracelist-p (containing-sexp paren-state)
785eecbb
RS
8501 ;; return the buffer position of the beginning of the brace list
8502 ;; statement if we're inside a brace list, otherwise return nil.
8503 ;; CONTAINING-SEXP is the buffer pos of the innermost containing
0386b551 8504 ;; paren. PAREN-STATE is the remainder of the state of enclosing
130c507e 8505 ;; braces
785eecbb
RS
8506 ;;
8507 ;; N.B.: This algorithm can potentially get confused by cpp macros
0386b551 8508 ;; placed in inconvenient locations. It's a trade-off we make for
785eecbb 8509 ;; speed.
0386b551
AM
8510 ;;
8511 ;; This function might do hidden buffer changes.
785eecbb 8512 (or
d9e94c22 8513 ;; This will pick up brace list declarations.
7a068717
AM
8514 (save-excursion
8515 (goto-char containing-sexp)
8516 (c-backward-over-enum-header))
785eecbb
RS
8517 ;; this will pick up array/aggregate init lists, even if they are nested.
8518 (save-excursion
0ec8351b
BW
8519 (let ((class-key
8520 ;; Pike can have class definitions anywhere, so we must
8521 ;; check for the class key here.
8522 (and (c-major-mode-is 'pike-mode)
a66cd3ee 8523 c-decl-block-key))
8a51e842 8524 bufpos braceassignp lim next-containing macro-start)
785eecbb
RS
8525 (while (and (not bufpos)
8526 containing-sexp)
8a51e842 8527 (when paren-state
a66cd3ee
MS
8528 (if (consp (car paren-state))
8529 (setq lim (cdr (car paren-state))
8530 paren-state (cdr paren-state))
8531 (setq lim (car paren-state)))
8532 (when paren-state
8533 (setq next-containing (car paren-state)
8534 paren-state (cdr paren-state))))
785eecbb 8535 (goto-char containing-sexp)
a66cd3ee
MS
8536 (if (c-looking-at-inexpr-block next-containing next-containing)
8537 ;; We're in an in-expression block of some kind. Do not
8538 ;; check nesting. We deliberately set the limit to the
8539 ;; containing sexp, so that c-looking-at-inexpr-block
8540 ;; doesn't check for an identifier before it.
0ec8351b
BW
8541 (setq containing-sexp nil)
8542 ;; see if the open brace is preceded by = or [...] in
8543 ;; this statement, but watch out for operator=
a66cd3ee 8544 (setq braceassignp 'dontknow)
d9e94c22 8545 (c-backward-token-2 1 t lim)
6393fef2 8546 ;; Checks to do only on the first sexp before the brace.
d9e94c22 8547 (when (and c-opt-inexpr-brace-list-key
6393fef2
RS
8548 (eq (char-after) ?\[))
8549 ;; In Java, an initialization brace list may follow
8550 ;; directly after "new Foo[]", so check for a "new"
8551 ;; earlier.
8552 (while (eq braceassignp 'dontknow)
8553 (setq braceassignp
d9e94c22
MS
8554 (cond ((/= (c-backward-token-2 1 t lim) 0) nil)
8555 ((looking-at c-opt-inexpr-brace-list-key) t)
6393fef2
RS
8556 ((looking-at "\\sw\\|\\s_\\|[.[]")
8557 ;; Carry on looking if this is an
8558 ;; identifier (may contain "." in Java)
8559 ;; or another "[]" sexp.
8560 'dontknow)
8561 (t nil)))))
8562 ;; Checks to do on all sexps before the brace, up to the
8563 ;; beginning of the statement.
8564 (while (eq braceassignp 'dontknow)
0ec8351b
BW
8565 (cond ((eq (char-after) ?\;)
8566 (setq braceassignp nil))
8567 ((and class-key
8568 (looking-at class-key))
8569 (setq braceassignp nil))
8570 ((eq (char-after) ?=)
8571 ;; We've seen a =, but must check earlier tokens so
8572 ;; that it isn't something that should be ignored.
8573 (setq braceassignp 'maybe)
8574 (while (and (eq braceassignp 'maybe)
d9e94c22 8575 (zerop (c-backward-token-2 1 t lim)))
0ec8351b
BW
8576 (setq braceassignp
8577 (cond
8578 ;; Check for operator =
51c9af45
AM
8579 ((and c-opt-op-identifier-prefix
8580 (looking-at c-opt-op-identifier-prefix))
0386b551 8581 nil)
130c507e
GM
8582 ;; Check for `<opchar>= in Pike.
8583 ((and (c-major-mode-is 'pike-mode)
8584 (or (eq (char-after) ?`)
8585 ;; Special case for Pikes
8586 ;; `[]=, since '[' is not in
8587 ;; the punctuation class.
8588 (and (eq (char-after) ?\[)
8589 (eq (char-before) ?`))))
8590 nil)
0ec8351b
BW
8591 ((looking-at "\\s.") 'maybe)
8592 ;; make sure we're not in a C++ template
8593 ;; argument assignment
a66cd3ee
MS
8594 ((and
8595 (c-major-mode-is 'c++-mode)
8596 (save-excursion
8597 (let ((here (point))
8598 (pos< (progn
8599 (skip-chars-backward "^<>")
8600 (point))))
8601 (and (eq (char-before) ?<)
8602 (not (c-crosses-statement-barrier-p
8603 pos< here))
8604 (not (c-in-literal))
8605 ))))
0ec8351b 8606 nil)
6393fef2 8607 (t t))))))
8a51e842
AM
8608 (if (and (eq braceassignp 'dontknow)
8609 (/= (c-backward-token-2 1 t lim) 0))
8610 (setq braceassignp nil)))
8611 (cond
8612 (braceassignp
8613 ;; We've hit the beginning of the aggregate list.
8614 (c-beginning-of-statement-1
8615 (c-most-enclosing-brace paren-state))
8616 (setq bufpos (point)))
8617 ((eq (char-after) ?\;)
8618 ;; Brace lists can't contain a semicolon, so we're done.
8619 (setq containing-sexp nil))
8620 ((and (setq macro-start (point))
8621 (c-forward-to-cpp-define-body)
8622 (eq (point) containing-sexp))
8623 ;; We've a macro whose expansion starts with the '{'.
8624 ;; Heuristically, if we have a ';' in it we've not got a
8625 ;; brace list, otherwise we have.
8626 (let ((macro-end (progn (c-end-of-macro) (point))))
8627 (goto-char containing-sexp)
8628 (forward-char)
8629 (if (and (c-syntactic-re-search-forward "[;,]" macro-end t t)
8630 (eq (char-before) ?\;))
8631 (setq bufpos nil
8632 containing-sexp nil)
8633 (setq bufpos macro-start))))
8634 (t
8635 ;; Go up one level
8636 (setq containing-sexp next-containing
8637 lim nil
8638 next-containing nil)))))
8639
785eecbb
RS
8640 bufpos))
8641 ))
8642
0ec8351b 8643(defun c-looking-at-special-brace-list (&optional lim)
c38e0c97
PE
8644 ;; If we're looking at the start of a pike-style list, i.e., `({ })',
8645 ;; `([ ])', `(< >)', etc., a cons of a cons of its starting and ending
0ec8351b
BW
8646 ;; positions and its entry in c-special-brace-lists is returned, nil
8647 ;; otherwise. The ending position is nil if the list is still open.
8648 ;; LIM is the limit for forward search. The point may either be at
8649 ;; the `(' or at the following paren character. Tries to check the
8650 ;; matching closer, but assumes it's correct if no balanced paren is
8651 ;; found (i.e. the case `({ ... } ... )' is detected as _not_ being
8652 ;; a special brace list).
0386b551
AM
8653 ;;
8654 ;; This function might do hidden buffer changes.
0ec8351b
BW
8655 (if c-special-brace-lists
8656 (condition-case ()
8657 (save-excursion
8658 (let ((beg (point))
2a15eb73 8659 inner-beg end type)
0ec8351b
BW
8660 (c-forward-syntactic-ws)
8661 (if (eq (char-after) ?\()
8662 (progn
8663 (forward-char 1)
8664 (c-forward-syntactic-ws)
2a15eb73 8665 (setq inner-beg (point))
0ec8351b
BW
8666 (setq type (assq (char-after) c-special-brace-lists)))
8667 (if (setq type (assq (char-after) c-special-brace-lists))
8668 (progn
2a15eb73 8669 (setq inner-beg (point))
0ec8351b
BW
8670 (c-backward-syntactic-ws)
8671 (forward-char -1)
8672 (setq beg (if (eq (char-after) ?\()
8673 (point)
8674 nil)))))
8675 (if (and beg type)
2a15eb73
MS
8676 (if (and (c-safe
8677 (goto-char beg)
8678 (c-forward-sexp 1)
8679 (setq end (point))
8680 (= (char-before) ?\)))
8681 (c-safe
8682 (goto-char inner-beg)
8683 (if (looking-at "\\s(")
8684 ;; Check balancing of the inner paren
8685 ;; below.
8686 (progn
0ec8351b 8687 (c-forward-sexp 1)
2a15eb73
MS
8688 t)
8689 ;; If the inner char isn't a paren then
8690 ;; we can't check balancing, so just
8691 ;; check the char before the outer
8692 ;; closing paren.
8693 (goto-char end)
8694 (backward-char)
8695 (c-backward-syntactic-ws)
8696 (= (char-before) (cdr type)))))
0ec8351b
BW
8697 (if (or (/= (char-syntax (char-before)) ?\))
8698 (= (progn
8699 (c-forward-syntactic-ws)
8700 (point))
8701 (1- end)))
8702 (cons (cons beg end) type))
8703 (cons (list beg) type)))))
8704 (error nil))))
8705
a66cd3ee
MS
8706(defun c-looking-at-bos (&optional lim)
8707 ;; Return non-nil if between two statements or declarations, assuming
8708 ;; point is not inside a literal or comment.
0386b551
AM
8709 ;;
8710 ;; Obsolete - `c-at-statement-start-p' or `c-at-expression-start-p'
8711 ;; are recommended instead.
8712 ;;
8713 ;; This function might do hidden buffer changes.
8714 (c-at-statement-start-p))
efbc652a 8715(make-obsolete 'c-looking-at-bos 'c-at-statement-start-p "22.1")
0386b551
AM
8716
8717(defun c-looking-at-inexpr-block (lim containing-sexp &optional check-at-end)
8718 ;; Return non-nil if we're looking at the beginning of a block
0ec8351b
BW
8719 ;; inside an expression. The value returned is actually a cons of
8720 ;; either 'inlambda, 'inexpr-statement or 'inexpr-class and the
0386b551
AM
8721 ;; position of the beginning of the construct.
8722 ;;
8723 ;; LIM limits the backward search. CONTAINING-SEXP is the start
8724 ;; position of the closest containing list. If it's nil, the
8725 ;; containing paren isn't used to decide whether we're inside an
8726 ;; expression or not. If both LIM and CONTAINING-SEXP are used, LIM
8727 ;; needs to be farther back.
8728 ;;
8729 ;; If CHECK-AT-END is non-nil then extra checks at the end of the
8730 ;; brace block might be done. It should only be used when the
8731 ;; construct can be assumed to be complete, i.e. when the original
8732 ;; starting position was further down than that.
8733 ;;
8734 ;; This function might do hidden buffer changes.
8735
0ec8351b 8736 (save-excursion
0386b551 8737 (let ((res 'maybe) passed-paren
a66cd3ee
MS
8738 (closest-lim (or containing-sexp lim (point-min)))
8739 ;; Look at the character after point only as a last resort
8740 ;; when we can't disambiguate.
8741 (block-follows (and (eq (char-after) ?{) (point))))
0386b551 8742
a66cd3ee
MS
8743 (while (and (eq res 'maybe)
8744 (progn (c-backward-syntactic-ws)
8745 (> (point) closest-lim))
8746 (not (bobp))
8747 (progn (backward-char)
8748 (looking-at "[\]\).]\\|\\w\\|\\s_"))
0386b551
AM
8749 (c-safe (forward-char)
8750 (goto-char (scan-sexps (point) -1))))
8751
a66cd3ee 8752 (setq res
0386b551
AM
8753 (if (looking-at c-keywords-regexp)
8754 (let ((kw-sym (c-keyword-sym (match-string 1))))
8755 (cond
8756 ((and block-follows
8757 (c-keyword-member kw-sym 'c-inexpr-class-kwds))
8758 (and (not (eq passed-paren ?\[))
8759 (or (not (looking-at c-class-key))
8760 ;; If the class definition is at the start of
8761 ;; a statement, we don't consider it an
8762 ;; in-expression class.
8763 (let ((prev (point)))
8764 (while (and
8765 (= (c-backward-token-2 1 nil closest-lim) 0)
8766 (eq (char-syntax (char-after)) ?w))
8767 (setq prev (point)))
8768 (goto-char prev)
8769 (not (c-at-statement-start-p)))
8770 ;; Also, in Pike we treat it as an
8771 ;; in-expression class if it's used in an
8772 ;; object clone expression.
8773 (save-excursion
8774 (and check-at-end
8775 (c-major-mode-is 'pike-mode)
8776 (progn (goto-char block-follows)
8777 (zerop (c-forward-token-2 1 t)))
8778 (eq (char-after) ?\())))
8779 (cons 'inexpr-class (point))))
8780 ((c-keyword-member kw-sym 'c-inexpr-block-kwds)
8781 (when (not passed-paren)
8782 (cons 'inexpr-statement (point))))
8783 ((c-keyword-member kw-sym 'c-lambda-kwds)
8784 (when (or (not passed-paren)
8785 (eq passed-paren ?\())
8786 (cons 'inlambda (point))))
8787 ((c-keyword-member kw-sym 'c-block-stmt-kwds)
8788 nil)
8789 (t
8790 'maybe)))
8791
8792 (if (looking-at "\\s(")
8793 (if passed-paren
8794 (if (and (eq passed-paren ?\[)
8795 (eq (char-after) ?\[))
8796 ;; Accept several square bracket sexps for
8797 ;; Java array initializations.
8798 'maybe)
8799 (setq passed-paren (char-after))
8800 'maybe)
8801 'maybe))))
8802
a66cd3ee 8803 (if (eq res 'maybe)
0386b551
AM
8804 (when (and c-recognize-paren-inexpr-blocks
8805 block-follows
a66cd3ee
MS
8806 containing-sexp
8807 (eq (char-after containing-sexp) ?\())
8808 (goto-char containing-sexp)
8809 (if (or (save-excursion
8810 (c-backward-syntactic-ws lim)
8811 (and (> (point) (or lim (point-min)))
8812 (c-on-identifier)))
8813 (and c-special-brace-lists
8814 (c-looking-at-special-brace-list)))
8815 nil
8816 (cons 'inexpr-statement (point))))
0386b551 8817
a66cd3ee 8818 res))))
0ec8351b 8819
a66cd3ee 8820(defun c-looking-at-inexpr-block-backward (paren-state)
0ec8351b
BW
8821 ;; Returns non-nil if we're looking at the end of an in-expression
8822 ;; block, otherwise the same as `c-looking-at-inexpr-block'.
a66cd3ee 8823 ;; PAREN-STATE is the paren state relevant at the current position.
0386b551
AM
8824 ;;
8825 ;; This function might do hidden buffer changes.
0ec8351b 8826 (save-excursion
a66cd3ee
MS
8827 ;; We currently only recognize a block.
8828 (let ((here (point))
8829 (elem (car-safe paren-state))
8830 containing-sexp)
8831 (when (and (consp elem)
8832 (progn (goto-char (cdr elem))
8833 (c-forward-syntactic-ws here)
8834 (= (point) here)))
8835 (goto-char (car elem))
8836 (if (setq paren-state (cdr paren-state))
8837 (setq containing-sexp (car-safe paren-state)))
8838 (c-looking-at-inexpr-block (c-safe-position containing-sexp
8839 paren-state)
8840 containing-sexp)))))
0ec8351b 8841
536610a4
AM
8842(defun c-at-macro-vsemi-p (&optional pos)
8843 ;; Is there a "virtual semicolon" at POS or point?
8844 ;; (See cc-defs.el for full details of "virtual semicolons".)
8845 ;;
8846 ;; This is true when point is at the last non syntactic WS position on the
8847 ;; line, there is a macro call last on the line, and this particular macro's
8848 ;; name is defined by the regexp `c-vs-macro-regexp' as not needing a
8849 ;; semicolon.
8850 (save-excursion
8851 (save-restriction
8852 (widen)
8853 (if pos
8854 (goto-char pos)
8855 (setq pos (point)))
8856 (and
8857 c-macro-with-semi-re
536610a4
AM
8858 (eq (skip-chars-backward " \t") 0)
8859
8860 ;; Check we've got nothing after this except comments and empty lines
8861 ;; joined by escaped EOLs.
8862 (skip-chars-forward " \t") ; always returns non-nil.
8863 (progn
8864 (while ; go over 1 block comment per iteration.
8865 (and
8866 (looking-at "\\(\\\\[\n\r][ \t]*\\)*")
8867 (goto-char (match-end 0))
8868 (cond
8869 ((looking-at c-block-comment-start-regexp)
8870 (and (forward-comment 1)
8871 (skip-chars-forward " \t"))) ; always returns non-nil
8872 ((looking-at c-line-comment-start-regexp)
8873 (end-of-line)
8874 nil)
8875 (t nil))))
8876 (eolp))
91af3942 8877
536610a4
AM
8878 (goto-char pos)
8879 (progn (c-backward-syntactic-ws)
8880 (eq (point) pos))
8881
8882 ;; Check for one of the listed macros being before point.
8883 (or (not (eq (char-before) ?\)))
8884 (when (c-go-list-backward)
8885 (c-backward-syntactic-ws)
8886 t))
8887 (c-simple-skip-symbol-backward)
2e492df3
AM
8888 (looking-at c-macro-with-semi-re)
8889 (goto-char pos)
c38e0c97 8890 (not (c-in-literal)))))) ; The most expensive check last.
536610a4
AM
8891
8892(defun c-macro-vsemi-status-unknown-p () t) ; See cc-defs.el.
8893
785eecbb 8894\f
d9e94c22
MS
8895;; `c-guess-basic-syntax' and the functions that precedes it below
8896;; implements the main decision tree for determining the syntactic
8897;; analysis of the current line of code.
8898
8899;; Dynamically bound to t when `c-guess-basic-syntax' is called during
8900;; auto newline analysis.
8901(defvar c-auto-newline-analysis nil)
8902
82ba65cf
AM
8903(defun c-brace-anchor-point (bracepos)
8904 ;; BRACEPOS is the position of a brace in a construct like "namespace
8905 ;; Bar {". Return the anchor point in this construct; this is the
8906 ;; earliest symbol on the brace's line which isn't earlier than
8907 ;; "namespace".
8908 ;;
8909 ;; Currently (2007-08-17), "like namespace" means "matches
8910 ;; c-other-block-decl-kwds". It doesn't work with "class" or "struct"
8911 ;; or anything like that.
8912 (save-excursion
8913 (let ((boi (c-point 'boi bracepos)))
8914 (goto-char bracepos)
8915 (while (and (> (point) boi)
8916 (not (looking-at c-other-decl-block-key)))
8917 (c-backward-token-2))
8918 (if (> (point) boi) (point) boi))))
8919
d9e94c22
MS
8920(defsubst c-add-syntax (symbol &rest args)
8921 ;; A simple function to prepend a new syntax element to
8922 ;; `c-syntactic-context'. Using `setq' on it is unsafe since it
8923 ;; should always be dynamically bound but since we read it first
8924 ;; we'll fail properly anyway if this function is misused.
8925 (setq c-syntactic-context (cons (cons symbol args)
8926 c-syntactic-context)))
8927
8928(defsubst c-append-syntax (symbol &rest args)
8929 ;; Like `c-add-syntax' but appends to the end of the syntax list.
8930 ;; (Normally not necessary.)
8931 (setq c-syntactic-context (nconc c-syntactic-context
8932 (list (cons symbol args)))))
a66cd3ee
MS
8933
8934(defun c-add-stmt-syntax (syntax-symbol
d9e94c22 8935 syntax-extra-args
a66cd3ee
MS
8936 stop-at-boi-only
8937 containing-sexp
d9e94c22 8938 paren-state)
51c9af45
AM
8939 ;; Add the indicated SYNTAX-SYMBOL to `c-syntactic-context', extending it as
8940 ;; needed with further syntax elements of the types `substatement',
8941 ;; `inexpr-statement', `arglist-cont-nonempty', `statement-block-intro', and
8942 ;; `defun-block-intro'.
b414f371 8943 ;;
a66cd3ee
MS
8944 ;; Do the generic processing to anchor the given syntax symbol on
8945 ;; the preceding statement: Skip over any labels and containing
8946 ;; statements on the same line, and then search backward until we
8947 ;; find a statement or block start that begins at boi without a
8948 ;; label or comment.
8949 ;;
8950 ;; Point is assumed to be at the prospective anchor point for the
8951 ;; given SYNTAX-SYMBOL. More syntax entries are added if we need to
82ba65cf
AM
8952 ;; skip past open parens and containing statements. Most of the added
8953 ;; syntax elements will get the same anchor point - the exception is
8954 ;; for an anchor in a construct like "namespace"[*] - this is as early
8955 ;; as possible in the construct but on the same line as the {.
8956 ;;
8957 ;; [*] i.e. with a keyword matching c-other-block-decl-kwds.
a66cd3ee 8958 ;;
d9e94c22
MS
8959 ;; SYNTAX-EXTRA-ARGS are a list of the extra arguments for the
8960 ;; syntax symbol. They are appended after the anchor point.
8961 ;;
0386b551
AM
8962 ;; If STOP-AT-BOI-ONLY is nil, we can stop in the middle of the line
8963 ;; if the current statement starts there.
a66cd3ee 8964 ;;
0386b551
AM
8965 ;; Note: It's not a problem if PAREN-STATE "overshoots"
8966 ;; CONTAINING-SEXP, i.e. contains info about parens further down.
8967 ;;
8968 ;; This function might do hidden buffer changes.
8969
a66cd3ee
MS
8970 (if (= (point) (c-point 'boi))
8971 ;; This is by far the most common case, so let's give it special
8972 ;; treatment.
d9e94c22 8973 (apply 'c-add-syntax syntax-symbol (point) syntax-extra-args)
a66cd3ee 8974
0386b551 8975 (let ((syntax-last c-syntactic-context)
d9e94c22 8976 (boi (c-point 'boi))
0386b551
AM
8977 ;; Set when we're on a label, so that we don't stop there.
8978 ;; FIXME: To be complete we should check if we're on a label
8979 ;; now at the start.
8980 on-label)
d9e94c22 8981
82ba65cf
AM
8982 ;; Use point as the anchor point for "namespace", "extern", etc.
8983 (apply 'c-add-syntax syntax-symbol
8984 (if (rassq syntax-symbol c-other-decl-block-key-in-symbols-alist)
8985 (point) nil)
8986 syntax-extra-args)
3ce79cf5 8987
0386b551
AM
8988 ;; Loop while we have to back out of containing blocks.
8989 (while
8990 (and
8991 (catch 'back-up-block
8992
8993 ;; Loop while we have to back up statements.
8994 (while (or (/= (point) boi)
8995 on-label
8996 (looking-at c-comment-start-regexp))
8997
8998 ;; Skip past any comments that stands between the
8999 ;; statement start and boi.
9000 (let ((savepos (point)))
9001 (while (and (/= savepos boi)
9002 (c-backward-single-comment))
9003 (setq savepos (point)
9004 boi (c-point 'boi)))
9005 (goto-char savepos))
9006
9007 ;; Skip to the beginning of this statement or backward
9008 ;; another one.
9009 (let ((old-pos (point))
9010 (old-boi boi)
9011 (step-type (c-beginning-of-statement-1 containing-sexp)))
9012 (setq boi (c-point 'boi)
9013 on-label (eq step-type 'label))
9014
9015 (cond ((= (point) old-pos)
9016 ;; If we didn't move we're at the start of a block and
9017 ;; have to continue outside it.
9018 (throw 'back-up-block t))
9019
9020 ((and (eq step-type 'up)
9021 (>= (point) old-boi)
9022 (looking-at "else\\>[^_]")
9023 (save-excursion
9024 (goto-char old-pos)
9025 (looking-at "if\\>[^_]")))
9026 ;; Special case to avoid deeper and deeper indentation
9027 ;; of "else if" clauses.
9028 )
9029
9030 ((and (not stop-at-boi-only)
9031 (/= old-pos old-boi)
9032 (memq step-type '(up previous)))
9033 ;; If stop-at-boi-only is nil, we shouldn't back up
9034 ;; over previous or containing statements to try to
9035 ;; reach boi, so go back to the last position and
9036 ;; exit.
9037 (goto-char old-pos)
9038 (throw 'back-up-block nil))
9039
9040 (t
9041 (if (and (not stop-at-boi-only)
9042 (memq step-type '(up previous beginning)))
9043 ;; If we've moved into another statement then we
9044 ;; should no longer try to stop in the middle of a
9045 ;; line.
9046 (setq stop-at-boi-only t))
9047
9048 ;; Record this as a substatement if we skipped up one
9049 ;; level.
9050 (when (eq step-type 'up)
9051 (c-add-syntax 'substatement nil))))
9052 )))
9053
9054 containing-sexp)
9055
9056 ;; Now we have to go out of this block.
9057 (goto-char containing-sexp)
9058
9059 ;; Don't stop in the middle of a special brace list opener
9060 ;; like "({".
9061 (when c-special-brace-lists
9062 (let ((special-list (c-looking-at-special-brace-list)))
9063 (when (and special-list
9064 (< (car (car special-list)) (point)))
9065 (setq containing-sexp (car (car special-list)))
9066 (goto-char containing-sexp))))
9067
9068 (setq paren-state (c-whack-state-after containing-sexp paren-state)
9069 containing-sexp (c-most-enclosing-brace paren-state)
9070 boi (c-point 'boi))
9071
9072 ;; Analyze the construct in front of the block we've stepped out
9073 ;; from and add the right syntactic element for it.
9074 (let ((paren-pos (point))
9075 (paren-char (char-after))
9076 step-type)
9077
9078 (if (eq paren-char ?\()
9079 ;; Stepped out of a parenthesis block, so we're in an
9080 ;; expression now.
9081 (progn
9082 (when (/= paren-pos boi)
9083 (if (and c-recognize-paren-inexpr-blocks
3ce79cf5 9084 (progn
0386b551
AM
9085 (c-backward-syntactic-ws containing-sexp)
9086 (or (not (looking-at "\\>"))
9087 (not (c-on-identifier))))
9088 (save-excursion
9089 (goto-char (1+ paren-pos))
9090 (c-forward-syntactic-ws)
9091 (eq (char-after) ?{)))
9092 ;; Stepped out of an in-expression statement. This
9093 ;; syntactic element won't get an anchor pos.
9094 (c-add-syntax 'inexpr-statement)
9095
9096 ;; A parenthesis normally belongs to an arglist.
9097 (c-add-syntax 'arglist-cont-nonempty nil paren-pos)))
9098
9099 (goto-char (max boi
9100 (if containing-sexp
9101 (1+ containing-sexp)
9102 (point-min))))
9103 (setq step-type 'same
9104 on-label nil))
9105
82ba65cf 9106 ;; Stepped out of a brace block.
0386b551
AM
9107 (setq step-type (c-beginning-of-statement-1 containing-sexp)
9108 on-label (eq step-type 'label))
9109
9110 (if (and (eq step-type 'same)
9111 (/= paren-pos (point)))
82ba65cf
AM
9112 (let (inexpr)
9113 (cond
9114 ((save-excursion
9115 (goto-char paren-pos)
9116 (setq inexpr (c-looking-at-inexpr-block
9117 (c-safe-position containing-sexp paren-state)
9118 containing-sexp)))
9119 (c-add-syntax (if (eq (car inexpr) 'inlambda)
9120 'defun-block-intro
9121 'statement-block-intro)
9122 nil))
9123 ((looking-at c-other-decl-block-key)
9124 (c-add-syntax
9125 (cdr (assoc (match-string 1)
9126 c-other-decl-block-key-in-symbols-alist))
9127 (max (c-point 'boi paren-pos) (point))))
9128 (t (c-add-syntax 'defun-block-intro nil))))
b414f371 9129
82ba65cf 9130 (c-add-syntax 'statement-block-intro nil)))
0386b551
AM
9131
9132 (if (= paren-pos boi)
9133 ;; Always done if the open brace was at boi. The
9134 ;; c-beginning-of-statement-1 call above is necessary
9135 ;; anyway, to decide the type of block-intro to add.
9136 (goto-char paren-pos)
9137 (setq boi (c-point 'boi)))
9138 ))
a66cd3ee 9139
d9e94c22
MS
9140 ;; Fill in the current point as the anchor for all the symbols
9141 ;; added above.
82ba65cf 9142 (let ((p c-syntactic-context) q)
d9e94c22 9143 (while (not (eq p syntax-last))
82ba65cf
AM
9144 (setq q (cdr (car p))) ; e.g. (nil 28) [from (arglist-cont-nonempty nil 28)]
9145 (while q
9146 (unless (car q)
9147 (setcar q (point)))
9148 (setq q (cdr q)))
d9e94c22 9149 (setq p (cdr p))))
a66cd3ee
MS
9150 )))
9151
0386b551
AM
9152(defun c-add-class-syntax (symbol
9153 containing-decl-open
9154 containing-decl-start
9155 containing-decl-kwd
9156 paren-state)
a66cd3ee
MS
9157 ;; The inclass and class-close syntactic symbols are added in
9158 ;; several places and some work is needed to fix everything.
9159 ;; Therefore it's collected here.
0386b551
AM
9160 ;;
9161 ;; This function might do hidden buffer changes.
9162 (goto-char containing-decl-open)
9163 (if (and (eq symbol 'inclass) (= (point) (c-point 'boi)))
9164 (progn
9165 (c-add-syntax symbol containing-decl-open)
9166 containing-decl-open)
9167 (goto-char containing-decl-start)
9168 ;; Ought to use `c-add-stmt-syntax' instead of backing up to boi
9169 ;; here, but we have to do like this for compatibility.
9170 (back-to-indentation)
9171 (c-add-syntax symbol (point))
9172 (if (and (c-keyword-member containing-decl-kwd
9173 'c-inexpr-class-kwds)
9174 (/= containing-decl-start (c-point 'boi containing-decl-start)))
9175 (c-add-syntax 'inexpr-class))
9176 (point)))
a66cd3ee
MS
9177
9178(defun c-guess-continued-construct (indent-point
9179 char-after-ip
9180 beg-of-same-or-containing-stmt
9181 containing-sexp
9182 paren-state)
9183 ;; This function contains the decision tree reached through both
9184 ;; cases 18 and 10. It's a continued statement or top level
9185 ;; construct of some kind.
0386b551
AM
9186 ;;
9187 ;; This function might do hidden buffer changes.
d9e94c22 9188
452ea855 9189 (let (special-brace-list placeholder)
a66cd3ee
MS
9190 (goto-char indent-point)
9191 (skip-chars-forward " \t")
d9e94c22 9192
a66cd3ee
MS
9193 (cond
9194 ;; (CASE A removed.)
9195 ;; CASE B: open braces for class or brace-lists
9196 ((setq special-brace-list
9197 (or (and c-special-brace-lists
9198 (c-looking-at-special-brace-list))
9199 (eq char-after-ip ?{)))
d9e94c22 9200
a66cd3ee
MS
9201 (cond
9202 ;; CASE B.1: class-open
9203 ((save-excursion
0386b551
AM
9204 (and (eq (char-after) ?{)
9205 (c-looking-at-decl-block containing-sexp t)
9206 (setq beg-of-same-or-containing-stmt (point))))
a66cd3ee 9207 (c-add-syntax 'class-open beg-of-same-or-containing-stmt))
d9e94c22 9208
a66cd3ee
MS
9209 ;; CASE B.2: brace-list-open
9210 ((or (consp special-brace-list)
9211 (save-excursion
9212 (goto-char beg-of-same-or-containing-stmt)
d9e94c22
MS
9213 (c-syntactic-re-search-forward "=\\([^=]\\|$\\)"
9214 indent-point t t t)))
a66cd3ee 9215 ;; The most semantically accurate symbol here is
d9e94c22
MS
9216 ;; brace-list-open, but we normally report it simply as a
9217 ;; statement-cont. The reason is that one normally adjusts
9218 ;; brace-list-open for brace lists as top-level constructs,
9219 ;; and brace lists inside statements is a completely different
9220 ;; context. C.f. case 5A.3.
a66cd3ee 9221 (c-beginning-of-statement-1 containing-sexp)
d9e94c22
MS
9222 (c-add-stmt-syntax (if c-auto-newline-analysis
9223 ;; Turn off the dwim above when we're
9224 ;; analyzing the nature of the brace
9225 ;; for the auto newline feature.
9226 'brace-list-open
9227 'statement-cont)
0386b551 9228 nil nil
d9e94c22
MS
9229 containing-sexp paren-state))
9230
a66cd3ee
MS
9231 ;; CASE B.3: The body of a function declared inside a normal
9232 ;; block. Can occur e.g. in Pike and when using gcc
037558bf 9233 ;; extensions, but watch out for macros followed by blocks.
d9e94c22 9234 ;; C.f. cases E, 16F and 17G.
0386b551 9235 ((and (not (c-at-statement-start-p))
d9e94c22 9236 (eq (c-beginning-of-statement-1 containing-sexp nil nil t)
037558bf
MS
9237 'same)
9238 (save-excursion
0386b551
AM
9239 (let ((c-recognize-typeless-decls nil))
9240 ;; Turn off recognition of constructs that lacks a
9241 ;; type in this case, since that's more likely to be
9242 ;; a macro followed by a block.
9243 (c-forward-decl-or-cast-1 (c-point 'bosws) nil nil))))
9244 (c-add-stmt-syntax 'defun-open nil t
d9e94c22
MS
9245 containing-sexp paren-state))
9246
037558bf
MS
9247 ;; CASE B.4: Continued statement with block open. The most
9248 ;; accurate analysis is perhaps `statement-cont' together with
9249 ;; `block-open' but we play DWIM and use `substatement-open'
58179cce 9250 ;; instead. The rationale is that this typically is a macro
037558bf
MS
9251 ;; followed by a block which makes it very similar to a
9252 ;; statement with a substatement block.
a66cd3ee 9253 (t
0386b551 9254 (c-add-stmt-syntax 'substatement-open nil nil
037558bf 9255 containing-sexp paren-state))
a66cd3ee 9256 ))
d9e94c22 9257
a66cd3ee 9258 ;; CASE C: iostream insertion or extraction operator
d9e94c22 9259 ((and (looking-at "\\(<<\\|>>\\)\\([^=]\\|$\\)")
a66cd3ee
MS
9260 (save-excursion
9261 (goto-char beg-of-same-or-containing-stmt)
d9e94c22
MS
9262 ;; If there is no preceding streamop in the statement
9263 ;; then indent this line as a normal statement-cont.
9264 (when (c-syntactic-re-search-forward
9265 "\\(<<\\|>>\\)\\([^=]\\|$\\)" indent-point 'move t t)
a66cd3ee
MS
9266 (c-add-syntax 'stream-op (c-point 'boi))
9267 t))))
d9e94c22
MS
9268
9269 ;; CASE E: In the "K&R region" of a function declared inside a
9270 ;; normal block. C.f. case B.3.
9271 ((and (save-excursion
9272 ;; Check that the next token is a '{'. This works as
9273 ;; long as no language that allows nested function
037558bf
MS
9274 ;; definitions allows stuff like member init lists, K&R
9275 ;; declarations or throws clauses there.
d9e94c22
MS
9276 ;;
9277 ;; Note that we do a forward search for something ahead
9278 ;; of the indentation line here. That's not good since
9279 ;; the user might not have typed it yet. Unfortunately
9280 ;; it's exceedingly tricky to recognize a function
9281 ;; prototype in a code block without resorting to this.
9282 (c-forward-syntactic-ws)
9283 (eq (char-after) ?{))
0386b551 9284 (not (c-at-statement-start-p))
d9e94c22 9285 (eq (c-beginning-of-statement-1 containing-sexp nil nil t)
037558bf
MS
9286 'same)
9287 (save-excursion
0386b551
AM
9288 (let ((c-recognize-typeless-decls nil))
9289 ;; Turn off recognition of constructs that lacks a
9290 ;; type in this case, since that's more likely to be
9291 ;; a macro followed by a block.
9292 (c-forward-decl-or-cast-1 (c-point 'bosws) nil nil))))
9293 (c-add-stmt-syntax 'func-decl-cont nil t
d9e94c22
MS
9294 containing-sexp paren-state))
9295
452ea855
AM
9296 ;;CASE F: continued statement and the only preceding items are
9297 ;;annotations.
9298 ((and (c-major-mode-is 'java-mode)
9299 (setq placeholder (point))
9300 (c-beginning-of-statement-1)
9301 (progn
9302 (while (and (c-forward-annotation)
9303 (< (point) placeholder))
9304 (c-forward-syntactic-ws))
9305 t)
9306 (prog1
9307 (>= (point) placeholder)
9308 (goto-char placeholder)))
9309 (c-beginning-of-statement-1 containing-sexp)
9310 (c-add-syntax 'annotation-var-cont (point)))
9311
e6597158
AM
9312 ;; CASE G: a template list continuation?
9313 ;; Mostly a duplication of case 5D.3 to fix templates-19:
9314 ((and (c-major-mode-is 'c++-mode)
9315 (save-excursion
9316 (goto-char indent-point)
9317 (c-with-syntax-table c++-template-syntax-table
9318 (setq placeholder (c-up-list-backward)))
9319 (and placeholder
9320 (eq (char-after placeholder) ?<)
9321 (/= (char-before placeholder) ?<)
9322 (progn
9323 (goto-char (1+ placeholder))
9324 (not (looking-at c-<-op-cont-regexp))))))
9325 (c-with-syntax-table c++-template-syntax-table
9326 (goto-char placeholder)
9327 (c-beginning-of-statement-1 containing-sexp t)
9328 (if (save-excursion
9329 (c-backward-syntactic-ws containing-sexp)
9330 (eq (char-before) ?<))
9331 ;; In a nested template arglist.
9332 (progn
9333 (goto-char placeholder)
9334 (c-syntactic-skip-backward "^,;" containing-sexp t)
9335 (c-forward-syntactic-ws))
9336 (back-to-indentation)))
9337 ;; FIXME: Should use c-add-stmt-syntax, but it's not yet
9338 ;; template aware.
9339 (c-add-syntax 'template-args-cont (point) placeholder))
91af3942 9340
a66cd3ee
MS
9341 ;; CASE D: continued statement.
9342 (t
9343 (c-beginning-of-statement-1 containing-sexp)
0386b551 9344 (c-add-stmt-syntax 'statement-cont nil nil
d9e94c22 9345 containing-sexp paren-state))
a66cd3ee 9346 )))
785eecbb 9347
0386b551
AM
9348;; The next autoload was added by RMS on 2005/8/9 - don't know why (ACM,
9349;; 2005/11/29).
4f047c58 9350;;;###autoload
785eecbb 9351(defun c-guess-basic-syntax ()
0386b551 9352 "Return the syntactic context of the current line."
785eecbb 9353 (save-excursion
9d1f0f55
AM
9354 (beginning-of-line)
9355 (c-save-buffer-state
9356 ((indent-point (point))
9357 (case-fold-search nil)
9358 ;; A whole ugly bunch of various temporary variables. Have
9359 ;; to declare them here since it's not possible to declare
9360 ;; a variable with only the scope of a cond test and the
9361 ;; following result clauses, and most of this function is a
9362 ;; single gigantic cond. :P
9363 literal char-before-ip before-ws-ip char-after-ip macro-start
9364 in-macro-expr c-syntactic-context placeholder c-in-literal-cache
9365 step-type tmpsymbol keyword injava-inher special-brace-list tmp-pos
9366 containing-<
9367 ;; The following record some positions for the containing
9368 ;; declaration block if we're directly within one:
9369 ;; `containing-decl-open' is the position of the open
9370 ;; brace. `containing-decl-start' is the start of the
9371 ;; declaration. `containing-decl-kwd' is the keyword
9372 ;; symbol of the keyword that tells what kind of block it
9373 ;; is.
9374 containing-decl-open
9375 containing-decl-start
9376 containing-decl-kwd
9377 ;; The open paren of the closest surrounding sexp or nil if
9378 ;; there is none.
9379 containing-sexp
9380 ;; The position after the closest preceding brace sexp
9381 ;; (nested sexps are ignored), or the position after
9382 ;; `containing-sexp' if there is none, or (point-min) if
9383 ;; `containing-sexp' is nil.
9384 lim
9385 ;; The paren state outside `containing-sexp', or at
9386 ;; `indent-point' if `containing-sexp' is nil.
9387 (paren-state (c-parse-state))
9388 ;; There's always at most one syntactic element which got
9389 ;; an anchor pos. It's stored in syntactic-relpos.
9390 syntactic-relpos
9391 (c-stmt-delim-chars c-stmt-delim-chars))
9392
9393 ;; Check if we're directly inside an enclosing declaration
9394 ;; level block.
9395 (when (and (setq containing-sexp
9396 (c-most-enclosing-brace paren-state))
9397 (progn
9398 (goto-char containing-sexp)
9399 (eq (char-after) ?{))
9400 (setq placeholder
9401 (c-looking-at-decl-block
9402 (c-most-enclosing-brace paren-state
9403 containing-sexp)
9404 t)))
9405 (setq containing-decl-open containing-sexp
9406 containing-decl-start (point)
9407 containing-sexp nil)
9408 (goto-char placeholder)
9409 (setq containing-decl-kwd (and (looking-at c-keywords-regexp)
9410 (c-keyword-sym (match-string 1)))))
9411
9412 ;; Init some position variables.
9413 (if c-state-cache
9414 (progn
9415 (setq containing-sexp (car paren-state)
9416 paren-state (cdr paren-state))
9417 (if (consp containing-sexp)
9418 (progn
9419 (setq lim (cdr containing-sexp))
9420 (if (cdr c-state-cache)
9421 ;; Ignore balanced paren. The next entry
9422 ;; can't be another one.
9423 (setq containing-sexp (car (cdr c-state-cache))
9424 paren-state (cdr paren-state))
9425 ;; If there is no surrounding open paren then
9426 ;; put the last balanced pair back on paren-state.
9427 (setq paren-state (cons containing-sexp paren-state)
9428 containing-sexp nil)))
9429 (setq lim (1+ containing-sexp))))
9430 (setq lim (point-min)))
9431
9432 ;; If we're in a parenthesis list then ',' delimits the
9433 ;; "statements" rather than being an operator (with the
9434 ;; exception of the "for" clause). This difference is
9435 ;; typically only noticeable when statements are used in macro
9436 ;; arglists.
9437 (when (and containing-sexp
9438 (eq (char-after containing-sexp) ?\())
9439 (setq c-stmt-delim-chars c-stmt-delim-chars-with-comma))
9440 ;; cache char before and after indent point, and move point to
9441 ;; the most likely position to perform the majority of tests
9442 (goto-char indent-point)
9443 (c-backward-syntactic-ws lim)
9444 (setq before-ws-ip (point)
9445 char-before-ip (char-before))
9446 (goto-char indent-point)
9447 (skip-chars-forward " \t")
9448 (setq char-after-ip (char-after))
9449
9450 ;; are we in a literal?
9451 (setq literal (c-in-literal lim))
9452
9453 ;; now figure out syntactic qualities of the current line
9454 (cond
0386b551 9455
9d1f0f55
AM
9456 ;; CASE 1: in a string.
9457 ((eq literal 'string)
9458 (c-add-syntax 'string (c-point 'bopl)))
9459
9460 ;; CASE 2: in a C or C++ style comment.
9461 ((and (memq literal '(c c++))
9462 ;; This is a kludge for XEmacs where we use
9463 ;; `buffer-syntactic-context', which doesn't correctly
9464 ;; recognize "\*/" to end a block comment.
9465 ;; `parse-partial-sexp' which is used by
9466 ;; `c-literal-limits' will however do that in most
9467 ;; versions, which results in that we get nil from
9468 ;; `c-literal-limits' even when `c-in-literal' claims
9469 ;; we're inside a comment.
9470 (setq placeholder (c-literal-limits lim)))
9471 (c-add-syntax literal (car placeholder)))
9472
9473 ;; CASE 3: in a cpp preprocessor macro continuation.
9474 ((and (save-excursion
9475 (when (c-beginning-of-macro)
9476 (setq macro-start (point))))
9477 (/= macro-start (c-point 'boi))
9478 (progn
9479 (setq tmpsymbol 'cpp-macro-cont)
9480 (or (not c-syntactic-indentation-in-macros)
9481 (save-excursion
9482 (goto-char macro-start)
9483 ;; If at the beginning of the body of a #define
9484 ;; directive then analyze as cpp-define-intro
9485 ;; only. Go on with the syntactic analysis
9486 ;; otherwise. in-macro-expr is set if we're in a
9487 ;; cpp expression, i.e. before the #define body
9488 ;; or anywhere in a non-#define directive.
9489 (if (c-forward-to-cpp-define-body)
9490 (let ((indent-boi (c-point 'boi indent-point)))
9491 (setq in-macro-expr (> (point) indent-boi)
9492 tmpsymbol 'cpp-define-intro)
9493 (= (point) indent-boi))
9494 (setq in-macro-expr t)
9495 nil)))))
9496 (c-add-syntax tmpsymbol macro-start)
9497 (setq macro-start nil))
9498
9499 ;; CASE 11: an else clause?
9500 ((looking-at "else\\>[^_]")
9501 (c-beginning-of-statement-1 containing-sexp)
9502 (c-add-stmt-syntax 'else-clause nil t
9503 containing-sexp paren-state))
0386b551 9504
9d1f0f55
AM
9505 ;; CASE 12: while closure of a do/while construct?
9506 ((and (looking-at "while\\>[^_]")
9507 (save-excursion
9508 (prog1 (eq (c-beginning-of-statement-1 containing-sexp)
9509 'beginning)
9510 (setq placeholder (point)))))
9511 (goto-char placeholder)
9512 (c-add-stmt-syntax 'do-while-closure nil t
9513 containing-sexp paren-state))
0386b551 9514
9d1f0f55
AM
9515 ;; CASE 13: A catch or finally clause? This case is simpler
9516 ;; than if-else and do-while, because a block is required
9517 ;; after every try, catch and finally.
9518 ((save-excursion
9519 (and (cond ((c-major-mode-is 'c++-mode)
9520 (looking-at "catch\\>[^_]"))
9521 ((c-major-mode-is 'java-mode)
9522 (looking-at "\\(catch\\|finally\\)\\>[^_]")))
9523 (and (c-safe (c-backward-syntactic-ws)
9524 (c-backward-sexp)
9525 t)
9526 (eq (char-after) ?{)
9527 (c-safe (c-backward-syntactic-ws)
9528 (c-backward-sexp)
9529 t)
9530 (if (eq (char-after) ?\()
9531 (c-safe (c-backward-sexp) t)
9532 t))
9533 (looking-at "\\(try\\|catch\\)\\>[^_]")
9534 (setq placeholder (point))))
9535 (goto-char placeholder)
9536 (c-add-stmt-syntax 'catch-clause nil t
9537 containing-sexp paren-state))
0386b551 9538
9d1f0f55
AM
9539 ;; CASE 18: A substatement we can recognize by keyword.
9540 ((save-excursion
9541 (and c-opt-block-stmt-key
9542 (not (eq char-before-ip ?\;))
9543 (not (c-at-vsemi-p before-ws-ip))
9544 (not (memq char-after-ip '(?\) ?\] ?,)))
9545 (or (not (eq char-before-ip ?}))
9546 (c-looking-at-inexpr-block-backward c-state-cache))
9547 (> (point)
9548 (progn
9549 ;; Ought to cache the result from the
9550 ;; c-beginning-of-statement-1 calls here.
9551 (setq placeholder (point))
9552 (while (eq (setq step-type
9553 (c-beginning-of-statement-1 lim))
9554 'label))
9555 (if (eq step-type 'previous)
9556 (goto-char placeholder)
a66cd3ee 9557 (setq placeholder (point))
9d1f0f55
AM
9558 (if (and (eq step-type 'same)
9559 (not (looking-at c-opt-block-stmt-key)))
9560 ;; Step up to the containing statement if we
9561 ;; stayed in the same one.
9562 (let (step)
9563 (while (eq
9564 (setq step
9565 (c-beginning-of-statement-1 lim))
9566 'label))
9567 (if (eq step 'up)
9568 (setq placeholder (point))
fa463103 9569 ;; There was no containing statement after all.
9d1f0f55
AM
9570 (goto-char placeholder)))))
9571 placeholder))
9572 (if (looking-at c-block-stmt-2-key)
9573 ;; Require a parenthesis after these keywords.
9574 ;; Necessary to catch e.g. synchronized in Java,
9575 ;; which can be used both as statement and
9576 ;; modifier.
9577 (and (zerop (c-forward-token-2 1 nil))
9578 (eq (char-after) ?\())
9579 (looking-at c-opt-block-stmt-key))))
9580
9581 (if (eq step-type 'up)
9582 ;; CASE 18A: Simple substatement.
9583 (progn
9584 (goto-char placeholder)
9585 (cond
9586 ((eq char-after-ip ?{)
9587 (c-add-stmt-syntax 'substatement-open nil nil
9588 containing-sexp paren-state))
9589 ((save-excursion
9590 (goto-char indent-point)
9591 (back-to-indentation)
9592 (c-forward-label))
9593 (c-add-stmt-syntax 'substatement-label nil nil
9594 containing-sexp paren-state))
9595 (t
9596 (c-add-stmt-syntax 'substatement nil nil
9597 containing-sexp paren-state))))
9598
9599 ;; CASE 18B: Some other substatement. This is shared
9600 ;; with case 10.
9601 (c-guess-continued-construct indent-point
9602 char-after-ip
9603 placeholder
9604 lim
9605 paren-state)))
0386b551 9606
9d1f0f55
AM
9607 ;; CASE 14: A case or default label
9608 ((looking-at c-label-kwds-regexp)
9609 (if containing-sexp
9610 (progn
9611 (goto-char containing-sexp)
9612 (setq lim (c-most-enclosing-brace c-state-cache
9613 containing-sexp))
9614 (c-backward-to-block-anchor lim)
9615 (c-add-stmt-syntax 'case-label nil t lim paren-state))
9616 ;; Got a bogus label at the top level. In lack of better
9617 ;; alternatives, anchor it on (point-min).
9618 (c-add-syntax 'case-label (point-min))))
0386b551 9619
9d1f0f55
AM
9620 ;; CASE 15: any other label
9621 ((save-excursion
130c507e 9622 (back-to-indentation)
9d1f0f55
AM
9623 (and (not (looking-at c-syntactic-ws-start))
9624 (c-forward-label)))
9625 (cond (containing-decl-open
9626 (setq placeholder (c-add-class-syntax 'inclass
9627 containing-decl-open
9628 containing-decl-start
9629 containing-decl-kwd
9630 paren-state))
9631 ;; Append access-label with the same anchor point as
9632 ;; inclass gets.
9633 (c-append-syntax 'access-label placeholder))
9634
9635 (containing-sexp
9636 (goto-char containing-sexp)
9637 (setq lim (c-most-enclosing-brace c-state-cache
9638 containing-sexp))
9639 (save-excursion
9640 (setq tmpsymbol
9641 (if (and (eq (c-beginning-of-statement-1 lim) 'up)
9642 (looking-at "switch\\>[^_]"))
9643 ;; If the surrounding statement is a switch then
9644 ;; let's analyze all labels as switch labels, so
9645 ;; that they get lined up consistently.
9646 'case-label
9647 'label)))
9648 (c-backward-to-block-anchor lim)
9649 (c-add-stmt-syntax tmpsymbol nil t lim paren-state))
0386b551 9650
9d1f0f55
AM
9651 (t
9652 ;; A label on the top level. Treat it as a class
9653 ;; context. (point-min) is the closest we get to the
9654 ;; class open brace.
9655 (c-add-syntax 'access-label (point-min)))))
9656
9657 ;; CASE 4: In-expression statement. C.f. cases 7B, 16A and
9658 ;; 17E.
9659 ((setq placeholder (c-looking-at-inexpr-block
9660 (c-safe-position containing-sexp paren-state)
9661 containing-sexp
9662 ;; Have to turn on the heuristics after
9663 ;; the point even though it doesn't work
9664 ;; very well. C.f. test case class-16.pike.
9665 t))
9666 (setq tmpsymbol (assq (car placeholder)
9667 '((inexpr-class . class-open)
9668 (inexpr-statement . block-open))))
9669 (if tmpsymbol
9670 ;; It's a statement block or an anonymous class.
9671 (setq tmpsymbol (cdr tmpsymbol))
9672 ;; It's a Pike lambda. Check whether we are between the
9673 ;; lambda keyword and the argument list or at the defun
9674 ;; opener.
9675 (setq tmpsymbol (if (eq char-after-ip ?{)
9676 'inline-open
9677 'lambda-intro-cont)))
9678 (goto-char (cdr placeholder))
9679 (back-to-indentation)
9680 (c-add-stmt-syntax tmpsymbol nil t
9681 (c-most-enclosing-brace c-state-cache (point))
9682 paren-state)
9683 (unless (eq (point) (cdr placeholder))
9684 (c-add-syntax (car placeholder))))
0386b551 9685
9d1f0f55
AM
9686 ;; CASE 5: Line is inside a declaration level block or at top level.
9687 ((or containing-decl-open (null containing-sexp))
9688 (cond
0386b551 9689
9d1f0f55
AM
9690 ;; CASE 5A: we are looking at a defun, brace list, class,
9691 ;; or inline-inclass method opening brace
9692 ((setq special-brace-list
9693 (or (and c-special-brace-lists
9694 (c-looking-at-special-brace-list))
9695 (eq char-after-ip ?{)))
9696 (cond
0386b551 9697
9d1f0f55
AM
9698 ;; CASE 5A.1: Non-class declaration block open.
9699 ((save-excursion
9700 (let (tmp)
9701 (and (eq char-after-ip ?{)
9702 (setq tmp (c-looking-at-decl-block containing-sexp t))
ce8c7486 9703 (progn
9d1f0f55
AM
9704 (setq placeholder (point))
9705 (goto-char tmp)
9706 (looking-at c-symbol-key))
9707 (c-keyword-member
9708 (c-keyword-sym (setq keyword (match-string 0)))
9709 'c-other-block-decl-kwds))))
9710 (goto-char placeholder)
9711 (c-add-stmt-syntax
9712 (if (string-equal keyword "extern")
9713 ;; Special case for extern-lang-open.
9714 'extern-lang-open
9715 (intern (concat keyword "-open")))
9716 nil t containing-sexp paren-state))
0386b551 9717
9d1f0f55 9718 ;; CASE 5A.2: we are looking at a class opening brace
0386b551 9719 ((save-excursion
9d1f0f55
AM
9720 (goto-char indent-point)
9721 (skip-chars-forward " \t")
9722 (and (eq (char-after) ?{)
9723 (c-looking-at-decl-block containing-sexp t)
9724 (setq placeholder (point))))
9725 (c-add-syntax 'class-open placeholder))
9726
9727 ;; CASE 5A.3: brace list open
9728 ((save-excursion
9729 (c-beginning-of-decl-1 lim)
9730 (while (looking-at c-specifier-key)
9731 (goto-char (match-end 1))
9732 (c-forward-syntactic-ws indent-point))
9733 (setq placeholder (c-point 'boi))
9734 (or (consp special-brace-list)
9735 (and (or (save-excursion
9736 (goto-char indent-point)
9737 (setq tmpsymbol nil)
9738 (while (and (> (point) placeholder)
9739 (zerop (c-backward-token-2 1 t))
ace2989a 9740 (not (looking-at "=\\([^=]\\|$\\)")))
9d1f0f55
AM
9741 (and c-opt-inexpr-brace-list-key
9742 (not tmpsymbol)
9743 (looking-at c-opt-inexpr-brace-list-key)
9744 (setq tmpsymbol 'topmost-intro-cont)))
ace2989a 9745 (looking-at "=\\([^=]\\|$\\)"))
9d1f0f55
AM
9746 (looking-at c-brace-list-key))
9747 (save-excursion
9748 (while (and (< (point) indent-point)
9749 (zerop (c-forward-token-2 1 t))
9750 (not (memq (char-after) '(?\; ?\()))))
9751 (not (memq (char-after) '(?\; ?\()))
9752 ))))
9753 (if (and (not c-auto-newline-analysis)
9754 (c-major-mode-is 'java-mode)
9755 (eq tmpsymbol 'topmost-intro-cont))
9756 ;; We're in Java and have found that the open brace
9757 ;; belongs to a "new Foo[]" initialization list,
9758 ;; which means the brace list is part of an
9759 ;; expression and not a top level definition. We
9760 ;; therefore treat it as any topmost continuation
9761 ;; even though the semantically correct symbol still
9762 ;; is brace-list-open, on the same grounds as in
9763 ;; case B.2.
9764 (progn
9765 (c-beginning-of-statement-1 lim)
9766 (c-add-syntax 'topmost-intro-cont (c-point 'boi)))
9767 (c-add-syntax 'brace-list-open placeholder)))
9768
9769 ;; CASE 5A.4: inline defun open
9770 ((and containing-decl-open
9771 (not (c-keyword-member containing-decl-kwd
9772 'c-other-block-decl-kwds)))
9773 (c-add-syntax 'inline-open)
9774 (c-add-class-syntax 'inclass
0386b551
AM
9775 containing-decl-open
9776 containing-decl-start
9777 containing-decl-kwd
9778 paren-state))
9779
9d1f0f55
AM
9780 ;; CASE 5A.5: ordinary defun open
9781 (t
9782 (save-excursion
9783 (c-beginning-of-decl-1 lim)
9784 (while (looking-at c-specifier-key)
9785 (goto-char (match-end 1))
9786 (c-forward-syntactic-ws indent-point))
9787 (c-add-syntax 'defun-open (c-point 'boi))
9788 ;; Bogus to use bol here, but it's the legacy. (Resolved,
9789 ;; 2007-11-09)
9790 ))))
9791
3d1c39fa
AM
9792 ;; CASE 5R: Member init list. (Used to be part of CASE 5B.1)
9793 ;; Note there is no limit on the backward search here, since member
9794 ;; init lists can, in practice, be very large.
9d1f0f55 9795 ((save-excursion
3d1c39fa 9796 (when (setq placeholder (c-back-over-member-initializers))
9d1f0f55 9797 (setq tmp-pos (point))))
3d1c39fa 9798 (if (= (c-point 'bosws) (1+ tmp-pos))
9d1f0f55
AM
9799 (progn
9800 ;; There is no preceding member init clause.
9801 ;; Indent relative to the beginning of indentation
9802 ;; for the topmost-intro line that contains the
9803 ;; prototype's open paren.
9804 (goto-char placeholder)
9805 (c-add-syntax 'member-init-intro (c-point 'boi)))
9806 ;; Indent relative to the first member init clause.
9807 (goto-char (1+ tmp-pos))
9808 (c-forward-syntactic-ws)
9809 (c-add-syntax 'member-init-cont (point))))
9810
3d1c39fa
AM
9811 ;; CASE 5B: After a function header but before the body (or
9812 ;; the ending semicolon if there's no body).
9813 ((save-excursion
9814 (when (setq placeholder (c-just-after-func-arglist-p
9815 (max lim (c-determine-limit 500))))
9816 (setq tmp-pos (point))))
9817 (cond
9818
9819 ;; CASE 5B.1: Member init list.
9820 ((eq (char-after tmp-pos) ?:)
9821 ;; There is no preceding member init clause.
9822 ;; Indent relative to the beginning of indentation
9823 ;; for the topmost-intro line that contains the
9824 ;; prototype's open paren.
9825 (goto-char placeholder)
9826 (c-add-syntax 'member-init-intro (c-point 'boi)))
9827
9d1f0f55 9828 ;; CASE 5B.2: K&R arg decl intro
785eecbb 9829 ((and c-recognize-knr-p
9d1f0f55
AM
9830 (c-in-knr-argdecl lim))
9831 (c-beginning-of-statement-1 lim)
9832 (c-add-syntax 'knr-argdecl-intro (c-point 'boi))
9833 (if containing-decl-open
9834 (c-add-class-syntax 'inclass
9835 containing-decl-open
9836 containing-decl-start
9837 containing-decl-kwd
9838 paren-state)))
9839
9840 ;; CASE 5B.4: Nether region after a C++ or Java func
9841 ;; decl, which could include a `throws' declaration.
9842 (t
9843 (c-beginning-of-statement-1 lim)
9844 (c-add-syntax 'func-decl-cont (c-point 'boi))
9845 )))
9846
9847 ;; CASE 5C: inheritance line. could be first inheritance
9848 ;; line, or continuation of a multiple inheritance
9849 ((or (and (c-major-mode-is 'c++-mode)
9850 (progn
9851 (when (eq char-after-ip ?,)
9852 (skip-chars-forward " \t")
9853 (forward-char))
9854 (looking-at c-opt-postfix-decl-spec-key)))
9855 (and (or (eq char-before-ip ?:)
9856 ;; watch out for scope operator
9857 (save-excursion
9858 (and (eq char-after-ip ?:)
9859 (c-safe (forward-char 1) t)
9860 (not (eq (char-after) ?:))
9861 )))
a66cd3ee 9862 (save-excursion
869455d4
AM
9863 (c-beginning-of-statement-1 lim)
9864 (when (looking-at c-opt-<>-sexp-key)
9865 (goto-char (match-end 1))
9866 (c-forward-syntactic-ws)
9867 (c-forward-<>-arglist nil)
9868 (c-forward-syntactic-ws))
9d1f0f55
AM
9869 (looking-at c-class-key)))
9870 ;; for Java
9871 (and (c-major-mode-is 'java-mode)
9872 (let ((fence (save-excursion
9873 (c-beginning-of-statement-1 lim)
9874 (point)))
9875 cont done)
9876 (save-excursion
9877 (while (not done)
9878 (cond ((looking-at c-opt-postfix-decl-spec-key)
9879 (setq injava-inher (cons cont (point))
9880 done t))
9881 ((or (not (c-safe (c-forward-sexp -1) t))
9882 (<= (point) fence))
9883 (setq done t))
9884 )
9885 (setq cont t)))
9886 injava-inher)
9887 (not (c-crosses-statement-barrier-p (cdr injava-inher)
9888 (point)))
9889 ))
9890 (cond
0386b551 9891
9d1f0f55
AM
9892 ;; CASE 5C.1: non-hanging colon on an inher intro
9893 ((eq char-after-ip ?:)
9894 (c-beginning-of-statement-1 lim)
9895 (c-add-syntax 'inher-intro (c-point 'boi))
9896 ;; don't add inclass symbol since relative point already
9897 ;; contains any class offset
9898 )
0386b551 9899
9d1f0f55
AM
9900 ;; CASE 5C.2: hanging colon on an inher intro
9901 ((eq char-before-ip ?:)
9902 (c-beginning-of-statement-1 lim)
9903 (c-add-syntax 'inher-intro (c-point 'boi))
0386b551 9904 (if containing-decl-open
9d1f0f55
AM
9905 (c-add-class-syntax 'inclass
9906 containing-decl-open
9907 containing-decl-start
9908 containing-decl-kwd
9909 paren-state)))
9910
9911 ;; CASE 5C.3: in a Java implements/extends
9912 (injava-inher
9913 (let ((where (cdr injava-inher))
9914 (cont (car injava-inher)))
9915 (goto-char where)
9916 (cond ((looking-at "throws\\>[^_]")
9917 (c-add-syntax 'func-decl-cont
9918 (progn (c-beginning-of-statement-1 lim)
9919 (c-point 'boi))))
9920 (cont (c-add-syntax 'inher-cont where))
9921 (t (c-add-syntax 'inher-intro
9922 (progn (goto-char (cdr injava-inher))
9923 (c-beginning-of-statement-1 lim)
9924 (point))))
9925 )))
9926
9927 ;; CASE 5C.4: a continued inheritance line
9928 (t
9929 (c-beginning-of-inheritance-list lim)
9930 (c-add-syntax 'inher-cont (point))
9931 ;; don't add inclass symbol since relative point already
9932 ;; contains any class offset
9933 )))
9934
b7aea0ff
AM
9935 ;; CASE 5P: AWK pattern or function or continuation
9936 ;; thereof.
9937 ((c-major-mode-is 'awk-mode)
9938 (setq placeholder (point))
9939 (c-add-stmt-syntax
9940 (if (and (eq (c-beginning-of-statement-1) 'same)
9941 (/= (point) placeholder))
9942 'topmost-intro-cont
9943 'topmost-intro)
9944 nil nil
9945 containing-sexp paren-state))
9946
9d1f0f55
AM
9947 ;; CASE 5D: this could be a top-level initialization, a
9948 ;; member init list continuation, or a template argument
9949 ;; list continuation.
9950 ((save-excursion
9951 ;; Note: We use the fact that lim is always after any
9952 ;; preceding brace sexp.
9953 (if c-recognize-<>-arglists
9954 (while (and
9955 (progn
9956 (c-syntactic-skip-backward "^;,=<>" lim t)
9957 (> (point) lim))
9958 (or
9959 (when c-overloadable-operators-regexp
9960 (when (setq placeholder (c-after-special-operator-id lim))
9961 (goto-char placeholder)
9962 t))
9963 (cond
9964 ((eq (char-before) ?>)
9965 (or (c-backward-<>-arglist nil lim)
9966 (backward-char))
9967 t)
9968 ((eq (char-before) ?<)
9969 (backward-char)
9970 (if (save-excursion
9971 (c-forward-<>-arglist nil))
9972 (progn (forward-char)
9973 nil)
9974 t))
9975 (t nil)))))
9976 ;; NB: No c-after-special-operator-id stuff in this
9977 ;; clause - we assume only C++ needs it.
9978 (c-syntactic-skip-backward "^;,=" lim t))
9979 (memq (char-before) '(?, ?= ?<)))
9980 (cond
9981
9982 ;; CASE 5D.3: perhaps a template list continuation?
9983 ((and (c-major-mode-is 'c++-mode)
9984 (save-excursion
9985 (save-restriction
9986 (c-with-syntax-table c++-template-syntax-table
9987 (goto-char indent-point)
9988 (setq placeholder (c-up-list-backward))
9989 (and placeholder
9990 (eq (char-after placeholder) ?<))))))
9991 (c-with-syntax-table c++-template-syntax-table
9992 (goto-char placeholder)
9993 (c-beginning-of-statement-1 lim t)
9994 (if (save-excursion
9995 (c-backward-syntactic-ws lim)
9996 (eq (char-before) ?<))
9997 ;; In a nested template arglist.
9998 (progn
9999 (goto-char placeholder)
10000 (c-syntactic-skip-backward "^,;" lim t)
10001 (c-forward-syntactic-ws))
10002 (back-to-indentation)))
10003 ;; FIXME: Should use c-add-stmt-syntax, but it's not yet
10004 ;; template aware.
10005 (c-add-syntax 'template-args-cont (point) placeholder))
10006
10007 ;; CASE 5D.4: perhaps a multiple inheritance line?
10008 ((and (c-major-mode-is 'c++-mode)
3efc2cd7 10009 (save-excursion
785eecbb 10010 (c-beginning-of-statement-1 lim)
9d1f0f55
AM
10011 (setq placeholder (point))
10012 (if (looking-at "static\\>[^_]")
10013 (c-forward-token-2 1 nil indent-point))
10014 (and (looking-at c-class-key)
10015 (zerop (c-forward-token-2 2 nil indent-point))
10016 (if (eq (char-after) ?<)
10017 (c-with-syntax-table c++-template-syntax-table
10018 (zerop (c-forward-token-2 1 t indent-point)))
10019 t)
10020 (eq (char-after) ?:))))
10021 (goto-char placeholder)
10022 (c-add-syntax 'inher-cont (c-point 'boi)))
10023
10024 ;; CASE 5D.5: Continuation of the "expression part" of a
e1dbe924 10025 ;; top level construct. Or, perhaps, an unrecognized construct.
9d1f0f55
AM
10026 (t
10027 (while (and (setq placeholder (point))
2e492df3 10028 (eq (car (c-beginning-of-decl-1 containing-sexp)) ; Can't use `lim' here.
9d1f0f55
AM
10029 'same)
10030 (save-excursion
10031 (c-backward-syntactic-ws)
10032 (eq (char-before) ?}))
10033 (< (point) placeholder)))
10034 (c-add-stmt-syntax
10035 (cond
e1dbe924 10036 ((eq (point) placeholder) 'statement) ; unrecognized construct
9d1f0f55
AM
10037 ;; A preceding comma at the top level means that a
10038 ;; new variable declaration starts here. Use
10039 ;; topmost-intro-cont for it, for consistency with
10040 ;; the first variable declaration. C.f. case 5N.
10041 ((eq char-before-ip ?,) 'topmost-intro-cont)
10042 (t 'statement-cont))
10043 nil nil containing-sexp paren-state))
10044 ))
0386b551 10045
9d1f0f55
AM
10046 ;; CASE 5F: Close of a non-class declaration level block.
10047 ((and (eq char-after-ip ?})
10048 (c-keyword-member containing-decl-kwd
10049 'c-other-block-decl-kwds))
10050 ;; This is inconsistent: Should use `containing-decl-open'
10051 ;; here if it's at boi, like in case 5J.
10052 (goto-char containing-decl-start)
10053 (c-add-stmt-syntax
10054 (if (string-equal (symbol-name containing-decl-kwd) "extern")
10055 ;; Special case for compatibility with the
10056 ;; extern-lang syntactic symbols.
10057 'extern-lang-close
10058 (intern (concat (symbol-name containing-decl-kwd)
10059 "-close")))
10060 nil t
10061 (c-most-enclosing-brace paren-state (point))
10062 paren-state))
10063
10064 ;; CASE 5G: we are looking at the brace which closes the
10065 ;; enclosing nested class decl
10066 ((and containing-sexp
10067 (eq char-after-ip ?})
10068 (eq containing-decl-open containing-sexp))
10069 (c-add-class-syntax 'class-close
10070 containing-decl-open
10071 containing-decl-start
10072 containing-decl-kwd
10073 paren-state))
10074
10075 ;; CASE 5H: we could be looking at subsequent knr-argdecls
10076 ((and c-recognize-knr-p
10077 (not containing-sexp) ; can't be knr inside braces.
10078 (not (eq char-before-ip ?}))
10079 (save-excursion
10080 (setq placeholder (cdr (c-beginning-of-decl-1 lim)))
10081 (and placeholder
10082 ;; Do an extra check to avoid tripping up on
10083 ;; statements that occur in invalid contexts
10084 ;; (e.g. in macro bodies where we don't really
10085 ;; know the context of what we're looking at).
10086 (not (and c-opt-block-stmt-key
10087 (looking-at c-opt-block-stmt-key)))))
10088 (< placeholder indent-point))
10089 (goto-char placeholder)
10090 (c-add-syntax 'knr-argdecl (point)))
10091
10092 ;; CASE 5I: ObjC method definition.
10093 ((and c-opt-method-key
10094 (looking-at c-opt-method-key))
10095 (c-beginning-of-statement-1 nil t)
10096 (if (= (point) indent-point)
10097 ;; Handle the case when it's the first (non-comment)
10098 ;; thing in the buffer. Can't look for a 'same return
10099 ;; value from cbos1 since ObjC directives currently
10100 ;; aren't recognized fully, so that we get 'same
10101 ;; instead of 'previous if it moved over a preceding
10102 ;; directive.
10103 (goto-char (point-min)))
10104 (c-add-syntax 'objc-method-intro (c-point 'boi)))
10105
9d1f0f55
AM
10106 ;; CASE 5N: At a variable declaration that follows a class
10107 ;; definition or some other block declaration that doesn't
10108 ;; end at the closing '}'. C.f. case 5D.5.
10109 ((progn
10110 (c-backward-syntactic-ws lim)
10111 (and (eq (char-before) ?})
10112 (save-excursion
10113 (let ((start (point)))
10114 (if (and c-state-cache
10115 (consp (car c-state-cache))
10116 (eq (cdar c-state-cache) (point)))
10117 ;; Speed up the backward search a bit.
10118 (goto-char (caar c-state-cache)))
2e492df3 10119 (c-beginning-of-decl-1 containing-sexp) ; Can't use `lim' here.
9d1f0f55
AM
10120 (setq placeholder (point))
10121 (if (= start (point))
10122 ;; The '}' is unbalanced.
10123 nil
10124 (c-end-of-decl-1)
10125 (>= (point) indent-point))))))
10126 (goto-char placeholder)
10127 (c-add-stmt-syntax 'topmost-intro-cont nil nil
10128 containing-sexp paren-state))
10129
10130 ;; NOTE: The point is at the end of the previous token here.
10131
10132 ;; CASE 5J: we are at the topmost level, make
10133 ;; sure we skip back past any access specifiers
10134 ((and
10135 ;; A macro continuation line is never at top level.
10136 (not (and macro-start
10137 (> indent-point macro-start)))
10138 (save-excursion
10139 (setq placeholder (point))
10140 (or (memq char-before-ip '(?\; ?{ ?} nil))
10141 (c-at-vsemi-p before-ws-ip)
10142 (when (and (eq char-before-ip ?:)
10143 (eq (c-beginning-of-statement-1 lim)
10144 'label))
10145 (c-backward-syntactic-ws lim)
10146 (setq placeholder (point)))
10147 (and (c-major-mode-is 'objc-mode)
10148 (catch 'not-in-directive
10149 (c-beginning-of-statement-1 lim)
10150 (setq placeholder (point))
10151 (while (and (c-forward-objc-directive)
10152 (< (point) indent-point))
10153 (c-forward-syntactic-ws)
10154 (if (>= (point) indent-point)
10155 (throw 'not-in-directive t))
10156 (setq placeholder (point)))
10157 nil)))))
10158 ;; For historic reasons we anchor at bol of the last
10159 ;; line of the previous declaration. That's clearly
10160 ;; highly bogus and useless, and it makes our lives hard
10161 ;; to remain compatible. :P
10162 (goto-char placeholder)
10163 (c-add-syntax 'topmost-intro (c-point 'bol))
10164 (if containing-decl-open
10165 (if (c-keyword-member containing-decl-kwd
10166 'c-other-block-decl-kwds)
10167 (progn
10168 (goto-char (c-brace-anchor-point containing-decl-open))
10169 (c-add-stmt-syntax
10170 (if (string-equal (symbol-name containing-decl-kwd)
10171 "extern")
10172 ;; Special case for compatibility with the
10173 ;; extern-lang syntactic symbols.
10174 'inextern-lang
10175 (intern (concat "in"
10176 (symbol-name containing-decl-kwd))))
10177 nil t
10178 (c-most-enclosing-brace paren-state (point))
10179 paren-state))
10180 (c-add-class-syntax 'inclass
10181 containing-decl-open
10182 containing-decl-start
10183 containing-decl-kwd
10184 paren-state)))
10185 (when (and c-syntactic-indentation-in-macros
10186 macro-start
10187 (/= macro-start (c-point 'boi indent-point)))
10188 (c-add-syntax 'cpp-define-intro)
10189 (setq macro-start nil)))
10190
10191 ;; CASE 5K: we are at an ObjC method definition
10192 ;; continuation line.
10193 ((and c-opt-method-key
10194 (save-excursion
10195 (c-beginning-of-statement-1 lim)
10196 (beginning-of-line)
10197 (when (looking-at c-opt-method-key)
10198 (setq placeholder (point)))))
10199 (c-add-syntax 'objc-method-args-cont placeholder))
10200
10201 ;; CASE 5L: we are at the first argument of a template
452ea855
AM
10202 ;; arglist that begins on the previous line.
10203 ((and c-recognize-<>-arglists
10204 (eq (char-before) ?<)
10205 (not (and c-overloadable-operators-regexp
10206 (c-after-special-operator-id lim))))
10207 (c-beginning-of-statement-1 (c-safe-position (point) paren-state))
10208 (c-add-syntax 'template-args-cont (c-point 'boi)))
10209
10210 ;; CASE 5Q: we are at a statement within a macro.
10211 (macro-start
10212 (c-beginning-of-statement-1 containing-sexp)
10213 (c-add-stmt-syntax 'statement nil t containing-sexp paren-state))
10214
40ba43b4 10215 ;;CASE 5N: We are at a topmost continuation line and the only
9d1f0f55 10216 ;;preceding items are annotations.
452ea855
AM
10217 ((and (c-major-mode-is 'java-mode)
10218 (setq placeholder (point))
9d1f0f55
AM
10219 (c-beginning-of-statement-1)
10220 (progn
452ea855 10221 (while (and (c-forward-annotation))
9d1f0f55
AM
10222 (c-forward-syntactic-ws))
10223 t)
10224 (prog1
10225 (>= (point) placeholder)
10226 (goto-char placeholder)))
10227 (c-add-syntax 'annotation-top-cont (c-point 'boi)))
452ea855
AM
10228
10229 ;; CASE 5M: we are at a topmost continuation line
10230 (t
10231 (c-beginning-of-statement-1 (c-safe-position (point) paren-state))
9d1f0f55
AM
10232 (when (c-major-mode-is 'objc-mode)
10233 (setq placeholder (point))
10234 (while (and (c-forward-objc-directive)
10235 (< (point) indent-point))
10236 (c-forward-syntactic-ws)
10237 (setq placeholder (point)))
10238 (goto-char placeholder))
10239 (c-add-syntax 'topmost-intro-cont (c-point 'boi)))
10240 ))
0386b551 10241
0386b551 10242
452ea855 10243 ;; (CASE 6 has been removed.)
d0fcee66 10244
452ea855
AM
10245 ;; CASE 7: line is an expression, not a statement. Most
10246 ;; likely we are either in a function prototype or a function
10247 ;; call argument list
10248 ((not (or (and c-special-brace-lists
10249 (save-excursion
10250 (goto-char containing-sexp)
10251 (c-looking-at-special-brace-list)))
10252 (eq (char-after containing-sexp) ?{)))
10253 (cond
0386b551 10254
452ea855 10255 ;; CASE 7A: we are looking at the arglist closing paren.
9d1f0f55
AM
10256 ;; C.f. case 7F.
10257 ((memq char-after-ip '(?\) ?\]))
10258 (goto-char containing-sexp)
10259 (setq placeholder (c-point 'boi))
10260 (if (and (c-safe (backward-up-list 1) t)
10261 (>= (point) placeholder))
10262 (progn
10263 (forward-char)
10264 (skip-chars-forward " \t"))
10265 (goto-char placeholder))
10266 (c-add-stmt-syntax 'arglist-close (list containing-sexp) t
452ea855
AM
10267 (c-most-enclosing-brace paren-state (point))
10268 paren-state))
0386b551 10269
452ea855
AM
10270 ;; CASE 7B: Looking at the opening brace of an
10271 ;; in-expression block or brace list. C.f. cases 4, 16A
10272 ;; and 17E.
9d1f0f55
AM
10273 ((and (eq char-after-ip ?{)
10274 (progn
10275 (setq placeholder (c-inside-bracelist-p (point)
10276 paren-state))
10277 (if placeholder
10278 (setq tmpsymbol '(brace-list-open . inexpr-class))
10279 (setq tmpsymbol '(block-open . inexpr-statement)
10280 placeholder
10281 (cdr-safe (c-looking-at-inexpr-block
10282 (c-safe-position containing-sexp
10283 paren-state)
10284 containing-sexp)))
10285 ;; placeholder is nil if it's a block directly in
10286 ;; a function arglist. That makes us skip out of
10287 ;; this case.
10288 )))
10289 (goto-char placeholder)
10290 (back-to-indentation)
10291 (c-add-stmt-syntax (car tmpsymbol) nil t
10292 (c-most-enclosing-brace paren-state (point))
10293 paren-state)
10294 (if (/= (point) placeholder)
10295 (c-add-syntax (cdr tmpsymbol))))
10296
10297 ;; CASE 7C: we are looking at the first argument in an empty
10298 ;; argument list. Use arglist-close if we're actually
10299 ;; looking at a close paren or bracket.
10300 ((memq char-before-ip '(?\( ?\[))
10301 (goto-char containing-sexp)
10302 (setq placeholder (c-point 'boi))
10303 (if (and (c-safe (backward-up-list 1) t)
10304 (>= (point) placeholder))
10305 (progn
10306 (forward-char)
10307 (skip-chars-forward " \t"))
10308 (goto-char placeholder))
10309 (c-add-stmt-syntax 'arglist-intro (list containing-sexp) t
10310 (c-most-enclosing-brace paren-state (point))
10311 paren-state))
0386b551 10312
9d1f0f55
AM
10313 ;; CASE 7D: we are inside a conditional test clause. treat
10314 ;; these things as statements
10315 ((progn
785eecbb 10316 (goto-char containing-sexp)
9d1f0f55
AM
10317 (and (c-safe (c-forward-sexp -1) t)
10318 (looking-at "\\<for\\>[^_]")))
10319 (goto-char (1+ containing-sexp))
10320 (c-forward-syntactic-ws indent-point)
10321 (if (eq char-before-ip ?\;)
10322 (c-add-syntax 'statement (point))
10323 (c-add-syntax 'statement-cont (point))
10324 ))
10325
10326 ;; CASE 7E: maybe a continued ObjC method call. This is the
10327 ;; case when we are inside a [] bracketed exp, and what
10328 ;; precede the opening bracket is not an identifier.
10329 ((and c-opt-method-key
10330 (eq (char-after containing-sexp) ?\[)
10331 (progn
10332 (goto-char (1- containing-sexp))
10333 (c-backward-syntactic-ws (c-point 'bod))
10334 (if (not (looking-at c-symbol-key))
10335 (c-add-syntax 'objc-method-call-cont containing-sexp))
10336 )))
0386b551 10337
9d1f0f55
AM
10338 ;; CASE 7F: we are looking at an arglist continuation line,
10339 ;; but the preceding argument is on the same line as the
10340 ;; opening paren. This case includes multi-line
10341 ;; mathematical paren groupings, but we could be on a
10342 ;; for-list continuation line. C.f. case 7A.
10343 ((progn
785eecbb 10344 (goto-char (1+ containing-sexp))
9d1f0f55
AM
10345 (< (save-excursion
10346 (c-forward-syntactic-ws)
10347 (point))
10348 (c-point 'bonl)))
10349 (goto-char containing-sexp) ; paren opening the arglist
10350 (setq placeholder (c-point 'boi))
10351 (if (and (c-safe (backward-up-list 1) t)
10352 (>= (point) placeholder))
10353 (progn
10354 (forward-char)
10355 (skip-chars-forward " \t"))
10356 (goto-char placeholder))
10357 (c-add-stmt-syntax 'arglist-cont-nonempty (list containing-sexp) t
10358 (c-most-enclosing-brace c-state-cache (point))
10359 paren-state))
0386b551 10360
9d1f0f55
AM
10361 ;; CASE 7G: we are looking at just a normal arglist
10362 ;; continuation line
10363 (t (c-forward-syntactic-ws indent-point)
10364 (c-add-syntax 'arglist-cont (c-point 'boi)))
10365 ))
0386b551 10366
9d1f0f55
AM
10367 ;; CASE 8: func-local multi-inheritance line
10368 ((and (c-major-mode-is 'c++-mode)
10369 (save-excursion
10370 (goto-char indent-point)
10371 (skip-chars-forward " \t")
10372 (looking-at c-opt-postfix-decl-spec-key)))
10373 (goto-char indent-point)
10374 (skip-chars-forward " \t")
10375 (cond
0386b551 10376
9d1f0f55
AM
10377 ;; CASE 8A: non-hanging colon on an inher intro
10378 ((eq char-after-ip ?:)
10379 (c-backward-syntactic-ws lim)
10380 (c-add-syntax 'inher-intro (c-point 'boi)))
0386b551 10381
9d1f0f55
AM
10382 ;; CASE 8B: hanging colon on an inher intro
10383 ((eq char-before-ip ?:)
10384 (c-add-syntax 'inher-intro (c-point 'boi)))
0386b551 10385
9d1f0f55
AM
10386 ;; CASE 8C: a continued inheritance line
10387 (t
10388 (c-beginning-of-inheritance-list lim)
10389 (c-add-syntax 'inher-cont (point))
10390 )))
10391
10392 ;; CASE 9: we are inside a brace-list
10393 ((and (not (c-major-mode-is 'awk-mode)) ; Maybe this isn't needed (ACM, 2002/3/29)
10394 (setq special-brace-list
10395 (or (and c-special-brace-lists ;;;; ALWAYS NIL FOR AWK!!
10396 (save-excursion
10397 (goto-char containing-sexp)
10398 (c-looking-at-special-brace-list)))
10399 (c-inside-bracelist-p containing-sexp paren-state))))
10400 (cond
0386b551 10401
9d1f0f55
AM
10402 ;; CASE 9A: In the middle of a special brace list opener.
10403 ((and (consp special-brace-list)
10404 (save-excursion
10405 (goto-char containing-sexp)
10406 (eq (char-after) ?\())
10407 (eq char-after-ip (car (cdr special-brace-list))))
10408 (goto-char (car (car special-brace-list)))
10409 (skip-chars-backward " \t")
10410 (if (and (bolp)
10411 (assoc 'statement-cont
10412 (setq placeholder (c-guess-basic-syntax))))
10413 (setq c-syntactic-context placeholder)
10414 (c-beginning-of-statement-1
10415 (c-safe-position (1- containing-sexp) paren-state))
10416 (c-forward-token-2 0)
10417 (while (looking-at c-specifier-key)
10418 (goto-char (match-end 1))
10419 (c-forward-syntactic-ws))
10420 (c-add-syntax 'brace-list-open (c-point 'boi))))
10421
10422 ;; CASE 9B: brace-list-close brace
10423 ((if (consp special-brace-list)
10424 ;; Check special brace list closer.
10425 (progn
10426 (goto-char (car (car special-brace-list)))
10427 (save-excursion
10428 (goto-char indent-point)
10429 (back-to-indentation)
10430 (or
10431 ;; We were between the special close char and the `)'.
10432 (and (eq (char-after) ?\))
10433 (eq (1+ (point)) (cdr (car special-brace-list))))
10434 ;; We were before the special close char.
10435 (and (eq (char-after) (cdr (cdr special-brace-list)))
10436 (zerop (c-forward-token-2))
10437 (eq (1+ (point)) (cdr (car special-brace-list)))))))
10438 ;; Normal brace list check.
10439 (and (eq char-after-ip ?})
10440 (c-safe (goto-char (c-up-list-backward (point))) t)
10441 (= (point) containing-sexp)))
10442 (if (eq (point) (c-point 'boi))
10443 (c-add-syntax 'brace-list-close (point))
10444 (setq lim (c-most-enclosing-brace c-state-cache (point)))
10445 (c-beginning-of-statement-1 lim)
10446 (c-add-stmt-syntax 'brace-list-close nil t lim paren-state)))
0386b551 10447
9d1f0f55
AM
10448 (t
10449 ;; Prepare for the rest of the cases below by going to the
10450 ;; token following the opening brace
10451 (if (consp special-brace-list)
10452 (progn
10453 (goto-char (car (car special-brace-list)))
10454 (c-forward-token-2 1 nil indent-point))
10455 (goto-char containing-sexp))
10456 (forward-char)
10457 (let ((start (point)))
10458 (c-forward-syntactic-ws indent-point)
10459 (goto-char (max start (c-point 'bol))))
10460 (c-skip-ws-forward indent-point)
785eecbb 10461 (cond
0386b551 10462
9d1f0f55
AM
10463 ;; CASE 9C: we're looking at the first line in a brace-list
10464 ((= (point) indent-point)
10465 (if (consp special-brace-list)
10466 (goto-char (car (car special-brace-list)))
10467 (goto-char containing-sexp))
a66cd3ee 10468 (if (eq (point) (c-point 'boi))
9d1f0f55 10469 (c-add-syntax 'brace-list-intro (point))
a66cd3ee
MS
10470 (setq lim (c-most-enclosing-brace c-state-cache (point)))
10471 (c-beginning-of-statement-1 lim)
9d1f0f55 10472 (c-add-stmt-syntax 'brace-list-intro nil t lim paren-state)))
0386b551 10473
9d1f0f55
AM
10474 ;; CASE 9D: this is just a later brace-list-entry or
10475 ;; brace-entry-open
10476 (t (if (or (eq char-after-ip ?{)
10477 (and c-special-brace-lists
10478 (save-excursion
10479 (goto-char indent-point)
10480 (c-forward-syntactic-ws (c-point 'eol))
10481 (c-looking-at-special-brace-list (point)))))
10482 (c-add-syntax 'brace-entry-open (point))
10483 (c-add-syntax 'brace-list-entry (point))
10484 ))
10485 ))))
10486
10487 ;; CASE 10: A continued statement or top level construct.
10488 ((and (not (memq char-before-ip '(?\; ?:)))
10489 (not (c-at-vsemi-p before-ws-ip))
10490 (or (not (eq char-before-ip ?}))
10491 (c-looking-at-inexpr-block-backward c-state-cache))
10492 (> (point)
10493 (save-excursion
10494 (c-beginning-of-statement-1 containing-sexp)
10495 (setq placeholder (point))))
10496 (/= placeholder containing-sexp))
10497 ;; This is shared with case 18.
10498 (c-guess-continued-construct indent-point
10499 char-after-ip
10500 placeholder
10501 containing-sexp
10502 paren-state))
10503
10504 ;; CASE 16: block close brace, possibly closing the defun or
10505 ;; the class
10506 ((eq char-after-ip ?})
10507 ;; From here on we have the next containing sexp in lim.
10508 (setq lim (c-most-enclosing-brace paren-state))
10509 (goto-char containing-sexp)
10510 (cond
0386b551 10511
9d1f0f55
AM
10512 ;; CASE 16E: Closing a statement block? This catches
10513 ;; cases where it's preceded by a statement keyword,
10514 ;; which works even when used in an "invalid" context,
10515 ;; e.g. a macro argument.
10516 ((c-after-conditional)
10517 (c-backward-to-block-anchor lim)
10518 (c-add-stmt-syntax 'block-close nil t lim paren-state))
10519
10520 ;; CASE 16A: closing a lambda defun or an in-expression
10521 ;; block? C.f. cases 4, 7B and 17E.
10522 ((setq placeholder (c-looking-at-inexpr-block
10523 (c-safe-position containing-sexp paren-state)
10524 nil))
10525 (setq tmpsymbol (if (eq (car placeholder) 'inlambda)
10526 'inline-close
10527 'block-close))
a66cd3ee 10528 (goto-char containing-sexp)
9d1f0f55
AM
10529 (back-to-indentation)
10530 (if (= containing-sexp (point))
10531 (c-add-syntax tmpsymbol (point))
10532 (goto-char (cdr placeholder))
10533 (back-to-indentation)
10534 (c-add-stmt-syntax tmpsymbol nil t
10535 (c-most-enclosing-brace paren-state (point))
10536 paren-state)
10537 (if (/= (point) (cdr placeholder))
10538 (c-add-syntax (car placeholder)))))
0386b551 10539
9d1f0f55
AM
10540 ;; CASE 16B: does this close an inline or a function in
10541 ;; a non-class declaration level block?
10542 ((save-excursion
10543 (and lim
10544 (progn
10545 (goto-char lim)
10546 (c-looking-at-decl-block
10547 (c-most-enclosing-brace paren-state lim)
10548 nil))
10549 (setq placeholder (point))))
10550 (c-backward-to-decl-anchor lim)
10551 (back-to-indentation)
10552 (if (save-excursion
10553 (goto-char placeholder)
10554 (looking-at c-other-decl-block-key))
10555 (c-add-syntax 'defun-close (point))
10556 (c-add-syntax 'inline-close (point))))
10557
10558 ;; CASE 16F: Can be a defun-close of a function declared
10559 ;; in a statement block, e.g. in Pike or when using gcc
10560 ;; extensions, but watch out for macros followed by
10561 ;; blocks. Let it through to be handled below.
10562 ;; C.f. cases B.3 and 17G.
10563 ((save-excursion
10564 (and (not (c-at-statement-start-p))
10565 (eq (c-beginning-of-statement-1 lim nil nil t) 'same)
10566 (setq placeholder (point))
10567 (let ((c-recognize-typeless-decls nil))
10568 ;; Turn off recognition of constructs that
10569 ;; lacks a type in this case, since that's more
10570 ;; likely to be a macro followed by a block.
10571 (c-forward-decl-or-cast-1 (c-point 'bosws) nil nil))))
10572 (back-to-indentation)
10573 (if (/= (point) containing-sexp)
10574 (goto-char placeholder))
10575 (c-add-stmt-syntax 'defun-close nil t lim paren-state))
10576
10577 ;; CASE 16C: If there is an enclosing brace then this is
10578 ;; a block close since defun closes inside declaration
10579 ;; level blocks have been handled above.
10580 (lim
10581 ;; If the block is preceded by a case/switch label on
10582 ;; the same line, we anchor at the first preceding label
10583 ;; at boi. The default handling in c-add-stmt-syntax
10584 ;; really fixes it better, but we do like this to keep
10585 ;; the indentation compatible with version 5.28 and
10586 ;; earlier. C.f. case 17H.
10587 (while (and (/= (setq placeholder (point)) (c-point 'boi))
10588 (eq (c-beginning-of-statement-1 lim) 'label)))
10589 (goto-char placeholder)
10590 (if (looking-at c-label-kwds-regexp)
10591 (c-add-syntax 'block-close (point))
10592 (goto-char containing-sexp)
10593 ;; c-backward-to-block-anchor not necessary here; those
10594 ;; situations are handled in case 16E above.
10595 (c-add-stmt-syntax 'block-close nil t lim paren-state)))
0386b551 10596
9d1f0f55 10597 ;; CASE 16D: Only top level defun close left.
785eecbb 10598 (t
9d1f0f55
AM
10599 (goto-char containing-sexp)
10600 (c-backward-to-decl-anchor lim)
10601 (c-add-stmt-syntax 'defun-close nil nil
10602 (c-most-enclosing-brace paren-state)
10603 paren-state))
10604 ))
0386b551 10605
804a116c
AM
10606 ;; CASE 19: line is an expression, not a statement, and is directly
10607 ;; contained by a template delimiter. Most likely, we are in a
10608 ;; template arglist within a statement. This case is based on CASE
10609 ;; 7. At some point in the future, we may wish to create more
10610 ;; syntactic symbols such as `template-intro',
10611 ;; `template-cont-nonempty', etc., and distinguish between them as we
10612 ;; do for `arglist-intro' etc. (2009-12-07).
10613 ((and c-recognize-<>-arglists
10614 (setq containing-< (c-up-list-backward indent-point containing-sexp))
10615 (eq (char-after containing-<) ?\<))
10616 (setq placeholder (c-point 'boi containing-<))
10617 (goto-char containing-sexp) ; Most nested Lbrace/Lparen (but not
10618 ; '<') before indent-point.
10619 (if (>= (point) placeholder)
10620 (progn
10621 (forward-char)
10622 (skip-chars-forward " \t"))
10623 (goto-char placeholder))
10624 (c-add-stmt-syntax 'template-args-cont (list containing-<) t
10625 (c-most-enclosing-brace c-state-cache (point))
10626 paren-state))
10627
9d1f0f55
AM
10628 ;; CASE 17: Statement or defun catchall.
10629 (t
10630 (goto-char indent-point)
10631 ;; Back up statements until we find one that starts at boi.
10632 (while (let* ((prev-point (point))
10633 (last-step-type (c-beginning-of-statement-1
10634 containing-sexp)))
10635 (if (= (point) prev-point)
10636 (progn
10637 (setq step-type (or step-type last-step-type))
10638 nil)
10639 (setq step-type last-step-type)
10640 (/= (point) (c-point 'boi)))))
10641 (cond
0386b551 10642
9d1f0f55
AM
10643 ;; CASE 17B: continued statement
10644 ((and (eq step-type 'same)
10645 (/= (point) indent-point))
10646 (c-add-stmt-syntax 'statement-cont nil nil
10647 containing-sexp paren-state))
10648
10649 ;; CASE 17A: After a case/default label?
10650 ((progn
10651 (while (and (eq step-type 'label)
10652 (not (looking-at c-label-kwds-regexp)))
10653 (setq step-type
10654 (c-beginning-of-statement-1 containing-sexp)))
10655 (eq step-type 'label))
10656 (c-add-stmt-syntax (if (eq char-after-ip ?{)
10657 'statement-case-open
10658 'statement-case-intro)
10659 nil t containing-sexp paren-state))
10660
10661 ;; CASE 17D: any old statement
10662 ((progn
10663 (while (eq step-type 'label)
10664 (setq step-type
10665 (c-beginning-of-statement-1 containing-sexp)))
10666 (eq step-type 'previous))
10667 (c-add-stmt-syntax 'statement nil t
10668 containing-sexp paren-state)
10669 (if (eq char-after-ip ?{)
10670 (c-add-syntax 'block-open)))
10671
10672 ;; CASE 17I: Inside a substatement block.
10673 ((progn
10674 ;; The following tests are all based on containing-sexp.
10675 (goto-char containing-sexp)
10676 ;; From here on we have the next containing sexp in lim.
10677 (setq lim (c-most-enclosing-brace paren-state containing-sexp))
10678 (c-after-conditional))
10679 (c-backward-to-block-anchor lim)
10680 (c-add-stmt-syntax 'statement-block-intro nil t
10681 lim paren-state)
10682 (if (eq char-after-ip ?{)
10683 (c-add-syntax 'block-open)))
10684
10685 ;; CASE 17E: first statement in an in-expression block.
10686 ;; C.f. cases 4, 7B and 16A.
10687 ((setq placeholder (c-looking-at-inexpr-block
10688 (c-safe-position containing-sexp paren-state)
10689 nil))
10690 (setq tmpsymbol (if (eq (car placeholder) 'inlambda)
10691 'defun-block-intro
10692 'statement-block-intro))
10693 (back-to-indentation)
10694 (if (= containing-sexp (point))
10695 (c-add-syntax tmpsymbol (point))
10696 (goto-char (cdr placeholder))
a66cd3ee 10697 (back-to-indentation)
9d1f0f55
AM
10698 (c-add-stmt-syntax tmpsymbol nil t
10699 (c-most-enclosing-brace c-state-cache (point))
10700 paren-state)
10701 (if (/= (point) (cdr placeholder))
10702 (c-add-syntax (car placeholder))))
10703 (if (eq char-after-ip ?{)
10704 (c-add-syntax 'block-open)))
10705
10706 ;; CASE 17F: first statement in an inline, or first
10707 ;; statement in a top-level defun. we can tell this is it
10708 ;; if there are no enclosing braces that haven't been
10709 ;; narrowed out by a class (i.e. don't use bod here).
10710 ((save-excursion
10711 (or (not (setq placeholder (c-most-enclosing-brace
10712 paren-state)))
10713 (and (progn
10714 (goto-char placeholder)
10715 (eq (char-after) ?{))
10716 (c-looking-at-decl-block (c-most-enclosing-brace
10717 paren-state (point))
10718 nil))))
10719 (c-backward-to-decl-anchor lim)
10720 (back-to-indentation)
10721 (c-add-syntax 'defun-block-intro (point)))
0386b551 10722
9d1f0f55
AM
10723 ;; CASE 17G: First statement in a function declared inside
10724 ;; a normal block. This can occur in Pike and with
10725 ;; e.g. the gcc extensions, but watch out for macros
10726 ;; followed by blocks. C.f. cases B.3 and 16F.
10727 ((save-excursion
10728 (and (not (c-at-statement-start-p))
10729 (eq (c-beginning-of-statement-1 lim nil nil t) 'same)
10730 (setq placeholder (point))
10731 (let ((c-recognize-typeless-decls nil))
10732 ;; Turn off recognition of constructs that lacks
10733 ;; a type in this case, since that's more likely
10734 ;; to be a macro followed by a block.
10735 (c-forward-decl-or-cast-1 (c-point 'bosws) nil nil))))
10736 (back-to-indentation)
10737 (if (/= (point) containing-sexp)
10738 (goto-char placeholder))
10739 (c-add-stmt-syntax 'defun-block-intro nil t
10740 lim paren-state))
0386b551 10741
9d1f0f55
AM
10742 ;; CASE 17H: First statement in a block.
10743 (t
10744 ;; If the block is preceded by a case/switch label on the
10745 ;; same line, we anchor at the first preceding label at
10746 ;; boi. The default handling in c-add-stmt-syntax is
10747 ;; really fixes it better, but we do like this to keep the
10748 ;; indentation compatible with version 5.28 and earlier.
10749 ;; C.f. case 16C.
10750 (while (and (/= (setq placeholder (point)) (c-point 'boi))
10751 (eq (c-beginning-of-statement-1 lim) 'label)))
10752 (goto-char placeholder)
10753 (if (looking-at c-label-kwds-regexp)
10754 (c-add-syntax 'statement-block-intro (point))
10755 (goto-char containing-sexp)
10756 ;; c-backward-to-block-anchor not necessary here; those
10757 ;; situations are handled in case 17I above.
10758 (c-add-stmt-syntax 'statement-block-intro nil t
10759 lim paren-state))
10760 (if (eq char-after-ip ?{)
10761 (c-add-syntax 'block-open)))
10762 ))
10763 )
10764
10765 ;; now we need to look at any modifiers
10766 (goto-char indent-point)
10767 (skip-chars-forward " \t")
10768
10769 ;; are we looking at a comment only line?
10770 (when (and (looking-at c-comment-start-regexp)
10771 (/= (c-forward-token-2 0 nil (c-point 'eol)) 0))
10772 (c-append-syntax 'comment-intro))
10773
10774 ;; we might want to give additional offset to friends (in C++).
10775 (when (and c-opt-friend-key
10776 (looking-at c-opt-friend-key))
10777 (c-append-syntax 'friend))
10778
10779 ;; Set syntactic-relpos.
10780 (let ((p c-syntactic-context))
10781 (while (and p
10782 (if (integerp (c-langelem-pos (car p)))
10783 (progn
10784 (setq syntactic-relpos (c-langelem-pos (car p)))
10785 nil)
10786 t))
10787 (setq p (cdr p))))
0386b551 10788
9d1f0f55
AM
10789 ;; Start of or a continuation of a preprocessor directive?
10790 (if (and macro-start
10791 (eq macro-start (c-point 'boi))
10792 (not (and (c-major-mode-is 'pike-mode)
10793 (eq (char-after (1+ macro-start)) ?\"))))
10794 (c-append-syntax 'cpp-macro)
10795 (when (and c-syntactic-indentation-in-macros macro-start)
10796 (if in-macro-expr
10797 (when (or
10798 (< syntactic-relpos macro-start)
10799 (not (or
10800 (assq 'arglist-intro c-syntactic-context)
10801 (assq 'arglist-cont c-syntactic-context)
10802 (assq 'arglist-cont-nonempty c-syntactic-context)
10803 (assq 'arglist-close c-syntactic-context))))
10804 ;; If inside a cpp expression, i.e. anywhere in a
10805 ;; cpp directive except a #define body, we only let
10806 ;; through the syntactic analysis that is internal
10807 ;; in the expression. That means the arglist
10808 ;; elements, if they are anchored inside the cpp
10809 ;; expression.
10810 (setq c-syntactic-context nil)
10811 (c-add-syntax 'cpp-macro-cont macro-start))
10812 (when (and (eq macro-start syntactic-relpos)
10813 (not (assq 'cpp-define-intro c-syntactic-context))
10814 (save-excursion
10815 (goto-char macro-start)
10816 (or (not (c-forward-to-cpp-define-body))
10817 (<= (point) (c-point 'boi indent-point)))))
10818 ;; Inside a #define body and the syntactic analysis is
10819 ;; anchored on the start of the #define. In this case
10820 ;; we add cpp-define-intro to get the extra
10821 ;; indentation of the #define body.
10822 (c-add-syntax 'cpp-define-intro)))))
10823
10824 ;; return the syntax
10825 c-syntactic-context)))
785eecbb
RS
10826
10827\f
d9e94c22 10828;; Indentation calculation.
785eecbb 10829
130c507e
GM
10830(defun c-evaluate-offset (offset langelem symbol)
10831 ;; offset can be a number, a function, a variable, a list, or one of
10832 ;; the symbols + or -
0386b551
AM
10833 ;;
10834 ;; This function might do hidden buffer changes.
10835 (let ((res
10836 (cond
10837 ((numberp offset) offset)
10838 ((vectorp offset) offset)
10839 ((null offset) nil)
10840
10841 ((eq offset '+) c-basic-offset)
10842 ((eq offset '-) (- c-basic-offset))
10843 ((eq offset '++) (* 2 c-basic-offset))
10844 ((eq offset '--) (* 2 (- c-basic-offset)))
10845 ((eq offset '*) (/ c-basic-offset 2))
10846 ((eq offset '/) (/ (- c-basic-offset) 2))
10847
10848 ((functionp offset)
10849 (c-evaluate-offset
10850 (funcall offset
10851 (cons (c-langelem-sym langelem)
10852 (c-langelem-pos langelem)))
10853 langelem symbol))
10854
10855 ((listp offset)
10856 (cond
10857 ((eq (car offset) 'quote)
10858 (c-benign-error "The offset %S for %s was mistakenly quoted"
10859 offset symbol)
10860 nil)
10861
10862 ((memq (car offset) '(min max))
10863 (let (res val (method (car offset)))
10864 (setq offset (cdr offset))
10865 (while offset
10866 (setq val (c-evaluate-offset (car offset) langelem symbol))
10867 (cond
10868 ((not val))
10869 ((not res)
10870 (setq res val))
10871 ((integerp val)
10872 (if (vectorp res)
10873 (c-benign-error "\
10874Error evaluating offset %S for %s: \
10875Cannot combine absolute offset %S with relative %S in `%s' method"
10876 (car offset) symbol res val method)
10877 (setq res (funcall method res val))))
10878 (t
10879 (if (integerp res)
10880 (c-benign-error "\
10881Error evaluating offset %S for %s: \
10882Cannot combine relative offset %S with absolute %S in `%s' method"
10883 (car offset) symbol res val method)
10884 (setq res (vector (funcall method (aref res 0)
10885 (aref val 0)))))))
10886 (setq offset (cdr offset)))
10887 res))
10888
10889 ((eq (car offset) 'add)
10890 (let (res val)
10891 (setq offset (cdr offset))
10892 (while offset
10893 (setq val (c-evaluate-offset (car offset) langelem symbol))
10894 (cond
10895 ((not val))
10896 ((not res)
10897 (setq res val))
10898 ((integerp val)
10899 (if (vectorp res)
10900 (setq res (vector (+ (aref res 0) val)))
10901 (setq res (+ res val))))
10902 (t
10903 (if (vectorp res)
10904 (c-benign-error "\
10905Error evaluating offset %S for %s: \
10906Cannot combine absolute offsets %S and %S in `add' method"
10907 (car offset) symbol res val)
10908 (setq res val)))) ; Override.
10909 (setq offset (cdr offset)))
10910 res))
10911
10912 (t
10913 (let (res)
10914 (when (eq (car offset) 'first)
10915 (setq offset (cdr offset)))
10916 (while (and (not res) offset)
10917 (setq res (c-evaluate-offset (car offset) langelem symbol)
10918 offset (cdr offset)))
10919 res))))
10920
10921 ((and (symbolp offset) (boundp offset))
10922 (symbol-value offset))
10923
10924 (t
10925 (c-benign-error "Unknown offset format %S for %s" offset symbol)
10926 nil))))
10927
10928 (if (or (null res) (integerp res)
10929 (and (vectorp res) (= (length res) 1) (integerp (aref res 0))))
10930 res
10931 (c-benign-error "Error evaluating offset %S for %s: Got invalid value %S"
10932 offset symbol res)
10933 nil)))
130c507e 10934
d9e94c22
MS
10935(defun c-calc-offset (langelem)
10936 ;; Get offset from LANGELEM which is a list beginning with the
10937 ;; syntactic symbol and followed by any analysis data it provides.
10938 ;; That data may be zero or more elements, but if at least one is
0386b551
AM
10939 ;; given then the first is the anchor position (or nil). The symbol
10940 ;; is matched against `c-offsets-alist' and the offset calculated
10941 ;; from that is returned.
10942 ;;
10943 ;; This function might do hidden buffer changes.
10944 (let* ((symbol (c-langelem-sym langelem))
130c507e
GM
10945 (match (assq symbol c-offsets-alist))
10946 (offset (cdr-safe match)))
a66cd3ee
MS
10947 (if match
10948 (setq offset (c-evaluate-offset offset langelem symbol))
10949 (if c-strict-syntax-p
10950 (c-benign-error "No offset found for syntactic symbol %s" symbol))
10951 (setq offset 0))
130c507e
GM
10952 (if (vectorp offset)
10953 offset
a66cd3ee
MS
10954 (or (and (numberp offset) offset)
10955 (and (symbolp offset) (symbol-value offset))
10956 0))
130c507e
GM
10957 ))
10958
d9e94c22
MS
10959(defun c-get-offset (langelem)
10960 ;; This is a compatibility wrapper for `c-calc-offset' in case
10961 ;; someone is calling it directly. It takes an old style syntactic
0386b551
AM
10962 ;; element on the form (SYMBOL . ANCHOR-POS) and converts it to the
10963 ;; new list form.
10964 ;;
10965 ;; This function might do hidden buffer changes.
10966 (if (c-langelem-pos langelem)
10967 (c-calc-offset (list (c-langelem-sym langelem)
10968 (c-langelem-pos langelem)))
d9e94c22
MS
10969 (c-calc-offset langelem)))
10970
130c507e 10971(defun c-get-syntactic-indentation (langelems)
d9e94c22
MS
10972 ;; Calculate the syntactic indentation from a syntactic description
10973 ;; as returned by `c-guess-syntax'.
10974 ;;
0386b551 10975 ;; Note that topmost-intro always has an anchor position at bol, for
a66cd3ee
MS
10976 ;; historical reasons. It's often used together with other symbols
10977 ;; that has more sane positions. Since we always use the first
0386b551
AM
10978 ;; found anchor position, we rely on that these other symbols always
10979 ;; precede topmost-intro in the LANGELEMS list.
10980 ;;
10981 ;; This function might do hidden buffer changes.
a66cd3ee 10982 (let ((indent 0) anchor)
d9e94c22
MS
10983
10984 (while langelems
10985 (let* ((c-syntactic-element (car langelems))
10986 (res (c-calc-offset c-syntactic-element)))
10987
10988 (if (vectorp res)
10989 ;; Got an absolute column that overrides any indentation
10990 ;; we've collected so far, but not the relative
10991 ;; indentation we might get for the nested structures
10992 ;; further down the langelems list.
10993 (setq indent (elt res 0)
10994 anchor (point-min)) ; A position at column 0.
10995
10996 ;; Got a relative change of the current calculated
10997 ;; indentation.
10998 (setq indent (+ indent res))
10999
11000 ;; Use the anchor position from the first syntactic
11001 ;; element with one.
11002 (unless anchor
0386b551 11003 (setq anchor (c-langelem-pos (car langelems)))))
d9e94c22
MS
11004
11005 (setq langelems (cdr langelems))))
11006
11007 (if anchor
11008 (+ indent (save-excursion
11009 (goto-char anchor)
11010 (current-column)))
11011 indent)))
51f606de 11012
785eecbb 11013\f
130c507e 11014(cc-provide 'cc-engine)
3afbc435 11015
785eecbb 11016;;; cc-engine.el ends here