Merge from emacs-24; up to 2012-12-06T01:39:03Z!monnier@iro.umontreal.ca
[bpt/emacs.git] / lisp / progmodes / cc-engine.el
Content-type: text/html HCoop Git - bpt/emacs.git/blame - lisp/progmodes/cc-engine.el


500 - Internal Server Error

Malformed UTF-8 character (fatal) at (eval 8) line 1, <$fd> line 18152.
CommitLineData
785eecbb
RS
1;;; cc-engine.el --- core syntax guessing engine for CC mode
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
130c507e
GM
150;; Silence the compiler.
151(cc-bytecomp-defun buffer-syntactic-context) ; XEmacs
0ec8351b 152
51f606de 153\f
d9e94c22
MS
154;; Make declarations for all the `c-lang-defvar' variables in cc-langs.
155
156(defmacro c-declare-lang-variables ()
157 `(progn
485fe977
RS
158 ,@(apply 'nconc
159 (mapcar (lambda (init)
160 `(,(if (elt init 2)
161 `(defvar ,(car init) nil ,(elt init 2))
162 `(defvar ,(car init) nil))
163 (make-variable-buffer-local ',(car init))))
164 (cdr c-lang-variable-inits)))))
d9e94c22
MS
165(c-declare-lang-variables)
166
167\f
168;;; Internal state variables.
169
170;; Internal state of hungry delete key feature
171(defvar c-hungry-delete-key nil)
172(make-variable-buffer-local 'c-hungry-delete-key)
173
0386b551
AM
174;; The electric flag (toggled by `c-toggle-electric-state').
175;; If t, electric actions (like automatic reindentation, and (if
176;; c-auto-newline is also set) auto newlining) will happen when an electric
177;; key like `{' is pressed (or an electric keyword like `else').
178(defvar c-electric-flag t)
179(make-variable-buffer-local 'c-electric-flag)
180
d9e94c22
MS
181;; Internal state of auto newline feature.
182(defvar c-auto-newline nil)
183(make-variable-buffer-local 'c-auto-newline)
184
0386b551 185;; Included in the mode line to indicate the active submodes.
cb694ab7
AM
186;; (defvar c-submode-indicators nil)
187;; (make-variable-buffer-local 'c-submode-indicators)
d9e94c22 188
a66cd3ee
MS
189(defun c-calculate-state (arg prevstate)
190 ;; Calculate the new state of PREVSTATE, t or nil, based on arg. If
191 ;; arg is nil or zero, toggle the state. If arg is negative, turn
192 ;; the state off, and if arg is positive, turn the state on
193 (if (or (not arg)
194 (zerop (setq arg (prefix-numeric-value arg))))
195 (not prevstate)
196 (> arg 0)))
197
d9e94c22 198\f
037558bf
MS
199;; Basic handling of preprocessor directives.
200
201;; This is a dynamically bound cache used together with
202;; `c-query-macro-start' and `c-query-and-set-macro-start'. It only
203;; works as long as point doesn't cross a macro boundary.
204(defvar c-macro-start 'unknown)
205
206(defsubst c-query-and-set-macro-start ()
037558bf
MS
207 (if (symbolp c-macro-start)
208 (setq c-macro-start (save-excursion
0386b551
AM
209 (c-save-buffer-state ()
210 (and (c-beginning-of-macro)
211 (point)))))
037558bf
MS
212 c-macro-start))
213
214(defsubst c-query-macro-start ()
037558bf
MS
215 (if (symbolp c-macro-start)
216 (save-excursion
0386b551
AM
217 (c-save-buffer-state ()
218 (and (c-beginning-of-macro)
219 (point))))
037558bf
MS
220 c-macro-start))
221
56d093a9
AM
222;; One element macro cache to cope with continual movement within very large
223;; CPP macros.
224(defvar c-macro-cache nil)
225(make-variable-buffer-local 'c-macro-cache)
226;; Nil or cons of the bounds of the most recent CPP form probed by
227;; `c-beginning-of-macro', `c-end-of-macro' or `c-syntactic-end-of-macro'.
228;; The cdr will be nil if we know only the start of the CPP form.
229(defvar c-macro-cache-start-pos nil)
230(make-variable-buffer-local 'c-macro-cache-start-pos)
231;; The starting position from where we determined `c-macro-cache'.
232(defvar c-macro-cache-syntactic nil)
233(make-variable-buffer-local 'c-macro-cache-syntactic)
234;; non-nil iff `c-macro-cache' has both elements set AND the cdr is at a
235;; syntactic end of macro, not merely an apparent one.
236
237(defun c-invalidate-macro-cache (beg end)
238 ;; Called from a before-change function. If the change region is before or
f003f294 239 ;; in the macro characterized by `c-macro-cache' etc., nullify it
56d093a9
AM
240 ;; appropriately. BEG and END are the standard before-change-functions
241 ;; parameters. END isn't used.
242 (cond
243 ((null c-macro-cache))
244 ((< beg (car c-macro-cache))
245 (setq c-macro-cache nil
246 c-macro-cache-start-pos nil
247 c-macro-cache-syntactic nil))
248 ((and (cdr c-macro-cache)
249 (< beg (cdr c-macro-cache)))
250 (setcdr c-macro-cache nil)
251 (setq c-macro-cache-start-pos beg
252 c-macro-cache-syntactic nil))))
253
037558bf
MS
254(defun c-beginning-of-macro (&optional lim)
255 "Go to the beginning of a preprocessor directive.
256Leave point at the beginning of the directive and return t if in one,
257otherwise return nil and leave point unchanged.
258
0386b551
AM
259Note that this function might do hidden buffer changes. See the
260comment at the start of cc-engine.el for more info."
56d093a9
AM
261 (let ((here (point)))
262 (when c-opt-cpp-prefix
263 (if (and (car c-macro-cache)
264 (>= (point) (car c-macro-cache))
265 (or (and (cdr c-macro-cache)
266 (<= (point) (cdr c-macro-cache)))
267 (<= (point) c-macro-cache-start-pos)))
268 (unless (< (car c-macro-cache) (or lim (point-min)))
269 (progn (goto-char (max (or lim (point-min)) (car c-macro-cache)))
270 (setq c-macro-cache-start-pos
271 (max c-macro-cache-start-pos here))
272 t))
273 (setq c-macro-cache nil
274 c-macro-cache-start-pos nil
275 c-macro-cache-syntactic nil)
276
277 (save-restriction
278 (if lim (narrow-to-region lim (point-max)))
279 (beginning-of-line)
280 (while (eq (char-before (1- (point))) ?\\)
281 (forward-line -1))
282 (back-to-indentation)
283 (if (and (<= (point) here)
284 (looking-at c-opt-cpp-start))
285 (progn
286 (setq c-macro-cache (cons (point) nil)
287 c-macro-cache-start-pos here)
288 t)
289 (goto-char here)
290 nil))))))
037558bf
MS
291
292(defun c-end-of-macro ()
293 "Go to the end of a preprocessor directive.
0386b551
AM
294More accurately, move the point to the end of the closest following
295line that doesn't end with a line continuation backslash - no check is
296done that the point is inside a cpp directive to begin with.
037558bf 297
0386b551
AM
298Note that this function might do hidden buffer changes. See the
299comment at the start of cc-engine.el for more info."
56d093a9
AM
300 (if (and (cdr c-macro-cache)
301 (<= (point) (cdr c-macro-cache))
302 (>= (point) (car c-macro-cache)))
303 (goto-char (cdr c-macro-cache))
304 (unless (and (car c-macro-cache)
305 (<= (point) c-macro-cache-start-pos)
306 (>= (point) (car c-macro-cache)))
307 (setq c-macro-cache nil
308 c-macro-cache-start-pos nil
309 c-macro-cache-syntactic nil))
310 (while (progn
311 (end-of-line)
312 (when (and (eq (char-before) ?\\)
313 (not (eobp)))
314 (forward-char)
315 t)))
316 (when (car c-macro-cache)
317 (setcdr c-macro-cache (point)))))
037558bf 318
0ec1d2c5
AM
319(defun c-syntactic-end-of-macro ()
320 ;; Go to the end of a CPP directive, or a "safe" pos just before.
321 ;;
322 ;; This is normally the end of the next non-escaped line. A "safe"
323 ;; position is one not within a string or comment. (The EOL on a line
324 ;; comment is NOT "safe").
325 ;;
326 ;; This function must only be called from the beginning of a CPP construct.
327 ;;
328 ;; Note that this function might do hidden buffer changes. See the comment
329 ;; at the start of cc-engine.el for more info.
330 (let* ((here (point))
331 (there (progn (c-end-of-macro) (point)))
56d093a9
AM
332 s)
333 (unless c-macro-cache-syntactic
334 (setq s (parse-partial-sexp here there))
335 (while (and (or (nth 3 s) ; in a string
336 (nth 4 s)) ; in a comment (maybe at end of line comment)
337 (> there here)) ; No infinite loops, please.
338 (setq there (1- (nth 8 s)))
339 (setq s (parse-partial-sexp here there)))
340 (setq c-macro-cache-syntactic (car c-macro-cache)))
0ec1d2c5
AM
341 (point)))
342
51c9af45
AM
343(defun c-forward-over-cpp-define-id ()
344 ;; Assuming point is at the "#" that introduces a preprocessor
345 ;; directive, it's moved forward to the end of the identifier which is
346 ;; "#define"d (or whatever c-opt-cpp-macro-define specifies). Non-nil
347 ;; is returned in this case, in all other cases nil is returned and
348 ;; point isn't moved.
349 ;;
350 ;; This function might do hidden buffer changes.
351 (when (and c-opt-cpp-macro-define-id
352 (looking-at c-opt-cpp-macro-define-id))
353 (goto-char (match-end 0))))
354
037558bf
MS
355(defun c-forward-to-cpp-define-body ()
356 ;; Assuming point is at the "#" that introduces a preprocessor
357 ;; directive, it's moved forward to the start of the definition body
0386b551
AM
358 ;; if it's a "#define" (or whatever c-opt-cpp-macro-define
359 ;; specifies). Non-nil is returned in this case, in all other cases
360 ;; nil is returned and point isn't moved.
361 ;;
362 ;; This function might do hidden buffer changes.
363 (when (and c-opt-cpp-macro-define-start
364 (looking-at c-opt-cpp-macro-define-start)
037558bf
MS
365 (not (= (match-end 0) (c-point 'eol))))
366 (goto-char (match-end 0))))
367
368\f
d9e94c22
MS
369;;; Basic utility functions.
370
0386b551 371(defun c-syntactic-content (from to paren-level)
d9e94c22
MS
372 ;; Return the given region as a string where all syntactic
373 ;; whitespace is removed or, where necessary, replaced with a single
0386b551
AM
374 ;; space. If PAREN-LEVEL is given then all parens in the region are
375 ;; collapsed to "()", "[]" etc.
376 ;;
377 ;; This function might do hidden buffer changes.
378
d9e94c22 379 (save-excursion
0386b551
AM
380 (save-restriction
381 (narrow-to-region from to)
382 (goto-char from)
383 (let* ((parts (list nil)) (tail parts) pos in-paren)
384
385 (while (re-search-forward c-syntactic-ws-start to t)
386 (goto-char (setq pos (match-beginning 0)))
387 (c-forward-syntactic-ws)
388 (if (= (point) pos)
389 (forward-char)
390
391 (when paren-level
392 (save-excursion
393 (setq in-paren (= (car (parse-partial-sexp from pos 1)) 1)
394 pos (point))))
395
396 (if (and (> pos from)
397 (< (point) to)
398 (looking-at "\\w\\|\\s_")
399 (save-excursion
400 (goto-char (1- pos))
401 (looking-at "\\w\\|\\s_")))
402 (progn
403 (setcdr tail (list (buffer-substring-no-properties from pos)
404 " "))
405 (setq tail (cddr tail)))
406 (setcdr tail (list (buffer-substring-no-properties from pos)))
407 (setq tail (cdr tail)))
408
409 (when in-paren
410 (when (= (car (parse-partial-sexp pos to -1)) -1)
411 (setcdr tail (list (buffer-substring-no-properties
412 (1- (point)) (point))))
413 (setq tail (cdr tail))))
414
415 (setq from (point))))
416
417 (setcdr tail (list (buffer-substring-no-properties from to)))
418 (apply 'concat (cdr parts))))))
419
420(defun c-shift-line-indentation (shift-amt)
421 ;; Shift the indentation of the current line with the specified
422 ;; amount (positive inwards). The buffer is modified only if
423 ;; SHIFT-AMT isn't equal to zero.
424 (let ((pos (- (point-max) (point)))
425 (c-macro-start c-macro-start)
426 tmp-char-inserted)
427 (if (zerop shift-amt)
428 nil
429 ;; If we're on an empty line inside a macro, we take the point
430 ;; to be at the current indentation and shift it to the
431 ;; appropriate column. This way we don't treat the extra
432 ;; whitespace out to the line continuation as indentation.
433 (when (and (c-query-and-set-macro-start)
434 (looking-at "[ \t]*\\\\$")
435 (save-excursion
436 (skip-chars-backward " \t")
437 (bolp)))
438 (insert ?x)
439 (backward-char)
440 (setq tmp-char-inserted t))
441 (unwind-protect
442 (let ((col (current-indentation)))
443 (delete-region (c-point 'bol) (c-point 'boi))
444 (beginning-of-line)
445 (indent-to (+ col shift-amt)))
446 (when tmp-char-inserted
447 (delete-char 1))))
448 ;; If initial point was within line's indentation and we're not on
449 ;; a line with a line continuation in a macro, position after the
450 ;; indentation. Else stay at same point in text.
451 (if (and (< (point) (c-point 'boi))
452 (not tmp-char-inserted))
453 (back-to-indentation)
454 (if (> (- (point-max) pos) (point))
455 (goto-char (- (point-max) pos))))))
d9e94c22
MS
456
457(defsubst c-keyword-sym (keyword)
458 ;; Return non-nil if the string KEYWORD is a known keyword. More
459 ;; precisely, the value is the symbol for the keyword in
460 ;; `c-keywords-obarray'.
461 (intern-soft keyword c-keywords-obarray))
462
463(defsubst c-keyword-member (keyword-sym lang-constant)
464 ;; Return non-nil if the symbol KEYWORD-SYM, as returned by
465 ;; `c-keyword-sym', is a member of LANG-CONSTANT, which is the name
466 ;; of a language constant that ends with "-kwds". If KEYWORD-SYM is
467 ;; nil then the result is nil.
468 (get keyword-sym lang-constant))
469
470;; String syntax chars, suitable for skip-syntax-(forward|backward).
471(defconst c-string-syntax (if (memq 'gen-string-delim c-emacs-features)
472 "\"|"
473 "\""))
474
0386b551 475;; Regexp matching string limit syntax.
d9e94c22
MS
476(defconst c-string-limit-regexp (if (memq 'gen-string-delim c-emacs-features)
477 "\\s\"\\|\\s|"
478 "\\s\""))
479
0386b551
AM
480;; Regexp matching WS followed by string limit syntax.
481(defconst c-ws*-string-limit-regexp
482 (concat "[ \t]*\\(" c-string-limit-regexp "\\)"))
483
d9e94c22
MS
484;; Holds formatted error strings for the few cases where parse errors
485;; are reported.
a66cd3ee 486(defvar c-parsing-error nil)
d9e94c22
MS
487(make-variable-buffer-local 'c-parsing-error)
488
489(defun c-echo-parsing-error (&optional quiet)
d9e94c22
MS
490 (when (and c-report-syntactic-errors c-parsing-error (not quiet))
491 (c-benign-error "%s" c-parsing-error))
492 c-parsing-error)
493
494;; Faces given to comments and string literals. This is used in some
495;; situations to speed up recognition; it isn't mandatory that font
496;; locking is in use. This variable is extended with the face in
497;; `c-doc-face-name' when fontification is activated in cc-fonts.el.
1e330469 498(defvar c-literal-faces
0386b551
AM
499 (append '(font-lock-comment-face font-lock-string-face)
500 (when (facep 'font-lock-comment-delimiter-face)
501 ;; New in Emacs 22.
502 '(font-lock-comment-delimiter-face))))
503
504(defsubst c-put-c-type-property (pos value)
505 ;; Put a c-type property with the given value at POS.
506 (c-put-char-property pos 'c-type value))
507
508(defun c-clear-c-type-property (from to value)
5a89f0a7 509 ;; Remove all occurrences of the c-type property that has the given
0386b551
AM
510 ;; value in the region between FROM and TO. VALUE is assumed to not
511 ;; be nil.
512 ;;
513 ;; Note: This assumes that c-type is put on single chars only; it's
514 ;; very inefficient if matching properties cover large regions.
515 (save-excursion
516 (goto-char from)
517 (while (progn
518 (when (eq (get-text-property (point) 'c-type) value)
519 (c-clear-char-property (point) 'c-type))
520 (goto-char (next-single-property-change (point) 'c-type nil to))
521 (< (point) to)))))
037558bf 522
d9e94c22
MS
523\f
524;; Some debug tools to visualize various special positions. This
525;; debug code isn't as portable as the rest of CC Mode.
526
527(cc-bytecomp-defun overlays-in)
528(cc-bytecomp-defun overlay-get)
529(cc-bytecomp-defun overlay-start)
530(cc-bytecomp-defun overlay-end)
531(cc-bytecomp-defun delete-overlay)
532(cc-bytecomp-defun overlay-put)
533(cc-bytecomp-defun make-overlay)
534
535(defun c-debug-add-face (beg end face)
536 (c-save-buffer-state ((overlays (overlays-in beg end)) overlay)
537 (while overlays
538 (setq overlay (car overlays)
539 overlays (cdr overlays))
540 (when (eq (overlay-get overlay 'face) face)
541 (setq beg (min beg (overlay-start overlay))
542 end (max end (overlay-end overlay)))
543 (delete-overlay overlay)))
544 (overlay-put (make-overlay beg end) 'face face)))
545
546(defun c-debug-remove-face (beg end face)
547 (c-save-buffer-state ((overlays (overlays-in beg end)) overlay
548 (ol-beg beg) (ol-end end))
549 (while overlays
550 (setq overlay (car overlays)
551 overlays (cdr overlays))
552 (when (eq (overlay-get overlay 'face) face)
553 (setq ol-beg (min ol-beg (overlay-start overlay))
554 ol-end (max ol-end (overlay-end overlay)))
555 (delete-overlay overlay)))
556 (when (< ol-beg beg)
557 (overlay-put (make-overlay ol-beg beg) 'face face))
558 (when (> ol-end end)
559 (overlay-put (make-overlay end ol-end) 'face face))))
560
561\f
562;; `c-beginning-of-statement-1' and accompanying stuff.
130c507e 563
64001211
RS
564;; KLUDGE ALERT: c-maybe-labelp is used to pass information between
565;; c-crosses-statement-barrier-p and c-beginning-of-statement-1. A
566;; better way should be implemented, but this will at least shut up
567;; the byte compiler.
0386b551 568(defvar c-maybe-labelp)
64001211 569
d9e94c22
MS
570;; New awk-compatible version of c-beginning-of-statement-1, ACM 2002/6/22
571
a66cd3ee
MS
572;; Macros used internally in c-beginning-of-statement-1 for the
573;; automaton actions.
574(defmacro c-bos-push-state ()
575 '(setq stack (cons (cons state saved-pos)
576 stack)))
577(defmacro c-bos-pop-state (&optional do-if-done)
578 `(if (setq state (car (car stack))
579 saved-pos (cdr (car stack))
580 stack (cdr stack))
581 t
582 ,do-if-done
583 (throw 'loop nil)))
584(defmacro c-bos-pop-state-and-retry ()
585 '(throw 'loop (setq state (car (car stack))
586 saved-pos (cdr (car stack))
587 ;; Throw nil if stack is empty, else throw non-nil.
588 stack (cdr stack))))
589(defmacro c-bos-save-pos ()
590 '(setq saved-pos (vector pos tok ptok pptok)))
591(defmacro c-bos-restore-pos ()
592 '(unless (eq (elt saved-pos 0) start)
593 (setq pos (elt saved-pos 0)
594 tok (elt saved-pos 1)
595 ptok (elt saved-pos 2)
596 pptok (elt saved-pos 3))
597 (goto-char pos)
598 (setq sym nil)))
599(defmacro c-bos-save-error-info (missing got)
600 `(setq saved-pos (vector pos ,missing ,got)))
601(defmacro c-bos-report-error ()
602 '(unless noerror
603 (setq c-parsing-error
604 (format "No matching `%s' found for `%s' on line %d"
605 (elt saved-pos 1)
606 (elt saved-pos 2)
607 (1+ (count-lines (point-min)
608 (c-point 'bol (elt saved-pos 0))))))))
609
610(defun c-beginning-of-statement-1 (&optional lim ignore-labels
611 noerror comma-delim)
612 "Move to the start of the current statement or declaration, or to
613the previous one if already at the beginning of one. Only
614statements/declarations on the same level are considered, i.e. don't
615move into or out of sexps (not even normal expression parentheses).
616
5a89f0a7 617If point is already at the earliest statement within braces or parens,
a85fd6da
AM
618this function doesn't move back into any whitespace preceding it; it
619returns 'same in this case.
620
d9e94c22
MS
621Stop at statement continuation tokens like \"else\", \"catch\",
622\"finally\" and the \"while\" in \"do ... while\" if the start point
623is within the continuation. If starting at such a token, move to the
624corresponding statement start. If at the beginning of a statement,
625move to the closest containing statement if there is any. This might
626also stop at a continuation clause.
a66cd3ee 627
0386b551
AM
628Labels are treated as part of the following statements if
629IGNORE-LABELS is non-nil. (FIXME: Doesn't work if we stop at a known
cb694ab7
AM
630statement start keyword.) Otherwise, each label is treated as a
631separate statement.
a66cd3ee 632
cb694ab7
AM
633Macros are ignored \(i.e. skipped over) unless point is within one, in
634which case the content of the macro is treated as normal code. Aside
635from any normal statement starts found in it, stop at the first token
636of the content in the macro, i.e. the expression of an \"#if\" or the
637start of the definition in a \"#define\". Also stop at start of
638macros before leaving them.
a66cd3ee 639
a85fd6da 640Return:
d28e7f28 641'label if stopped at a label or \"case...:\" or \"default:\";
a85fd6da
AM
642'same if stopped at the beginning of the current statement;
643'up if stepped to a containing statement;
644'previous if stepped to a preceding statement;
645'beginning if stepped from a statement continuation clause to
646 its start clause; or
647'macro if stepped to a macro start.
648Note that 'same and not 'label is returned if stopped at the same
649label without crossing the colon character.
a66cd3ee
MS
650
651LIM may be given to limit the search. If the search hits the limit,
652point will be left at the closest following token, or at the start
653position if that is less ('same is returned in this case).
654
655NOERROR turns off error logging to `c-parsing-error'.
656
cb694ab7
AM
657Normally only ';' and virtual semicolons are considered to delimit
658statements, but if COMMA-DELIM is non-nil then ',' is treated
659as a delimiter too.
0386b551
AM
660
661Note that this function might do hidden buffer changes. See the
662comment at the start of cc-engine.el for more info."
a66cd3ee 663
d9e94c22
MS
664 ;; The bulk of this function is a pushdown automaton that looks at statement
665 ;; boundaries and the tokens (such as "while") in c-opt-block-stmt-key. Its
666 ;; purpose is to keep track of nested statements, ensuring that such
5a89f0a7 667 ;; statements are skipped over in their entirety (somewhat akin to what C-M-p
d9e94c22 668 ;; does with nested braces/brackets/parentheses).
a66cd3ee
MS
669 ;;
670 ;; Note: The position of a boundary is the following token.
671 ;;
d9e94c22
MS
672 ;; Beginning with the current token (the one following point), move back one
673 ;; sexp at a time (where a sexp is, more or less, either a token or the
674 ;; entire contents of a brace/bracket/paren pair). Each time a statement
675 ;; boundary is crossed or a "while"-like token is found, update the state of
676 ;; the PDA. Stop at the beginning of a statement when the stack (holding
677 ;; nested statement info) is empty and the position has been moved.
678 ;;
679 ;; The following variables constitute the PDA:
680 ;;
681 ;; sym: This is either the "while"-like token (e.g. 'for) we've just
682 ;; scanned back over, 'boundary if we've just gone back over a
683 ;; statement boundary, or nil otherwise.
684 ;; state: takes one of the values (nil else else-boundary while
685 ;; while-boundary catch catch-boundary).
686 ;; nil means "no "while"-like token yet scanned".
687 ;; 'else, for example, means "just gone back over an else".
688 ;; 'else-boundary means "just gone back over a statement boundary
689 ;; immediately after having gone back over an else".
690 ;; saved-pos: A vector of either saved positions (tok ptok pptok, etc.) or
691 ;; of error reporting information.
692 ;; stack: The stack onto which the PDA pushes its state. Each entry
693 ;; consists of a saved value of state and saved-pos. An entry is
694 ;; pushed when we move back over a "continuation" token (e.g. else)
695 ;; and popped when we encounter the corresponding opening token
696 ;; (e.g. if).
697 ;;
698 ;;
b414f371 699 ;; The following diagram briefly outlines the PDA.
a66cd3ee
MS
700 ;;
701 ;; Common state:
d9e94c22
MS
702 ;; "else": Push state, goto state `else'.
703 ;; "while": Push state, goto state `while'.
704 ;; "catch" or "finally": Push state, goto state `catch'.
705 ;; boundary: Pop state.
a66cd3ee
MS
706 ;; other: Do nothing special.
707 ;;
d9e94c22
MS
708 ;; State `else':
709 ;; boundary: Goto state `else-boundary'.
710 ;; other: Error, pop state, retry token.
711 ;;
712 ;; State `else-boundary':
713 ;; "if": Pop state.
714 ;; boundary: Error, pop state.
715 ;; other: See common state.
716 ;;
717 ;; State `while':
718 ;; boundary: Save position, goto state `while-boundary'.
719 ;; other: Pop state, retry token.
720 ;;
721 ;; State `while-boundary':
722 ;; "do": Pop state.
723 ;; boundary: Restore position if it's not at start, pop state. [*see below]
724 ;; other: See common state.
725 ;;
726 ;; State `catch':
727 ;; boundary: Goto state `catch-boundary'.
728 ;; other: Error, pop state, retry token.
729 ;;
730 ;; State `catch-boundary':
731 ;; "try": Pop state.
732 ;; "catch": Goto state `catch'.
733 ;; boundary: Error, pop state.
734 ;; other: See common state.
735 ;;
736 ;; [*] In the `while-boundary' state, we had pushed a 'while state, and were
737 ;; searching for a "do" which would have opened a do-while. If we didn't
738 ;; find it, we discard the analysis done since the "while", go back to this
739 ;; token in the buffer and restart the scanning there, this time WITHOUT
740 ;; pushing the 'while state onto the stack.
741 ;;
a66cd3ee
MS
742 ;; In addition to the above there is some special handling of labels
743 ;; and macros.
744
745 (let ((case-fold-search nil)
746 (start (point))
747 macro-start
748 (delims (if comma-delim '(?\; ?,) '(?\;)))
749 (c-stmt-delim-chars (if comma-delim
750 c-stmt-delim-chars-with-comma
751 c-stmt-delim-chars))
d28e7f28 752 c-in-literal-cache c-maybe-labelp after-case:-pos saved
0386b551
AM
753 ;; Current position.
754 pos
755 ;; Position of last stmt boundary character (e.g. ;).
756 boundary-pos
757 ;; The position of the last sexp or bound that follows the
758 ;; first found colon, i.e. the start of the nonlabel part of
759 ;; the statement. It's `start' if a colon is found just after
760 ;; the start.
761 after-labels-pos
762 ;; Like `after-labels-pos', but the first such position inside
763 ;; a label, i.e. the start of the last label before the start
764 ;; of the nonlabel part of the statement.
765 last-label-pos
766 ;; The last position where a label is possible provided the
767 ;; statement started there. It's nil as long as no invalid
768 ;; label content has been found (according to
020716e1 769 ;; `c-nonlabel-token-key'). It's `start' if no valid label
0386b551
AM
770 ;; content was found in the label. Note that we might still
771 ;; regard it a label if it starts with `c-label-kwds'.
772 label-good-pos
b248a85d
AM
773 ;; Putative positions of the components of a bitfield declaration,
774 ;; e.g. "int foo : NUM_FOO_BITS ;"
775 bitfield-type-pos bitfield-id-pos bitfield-size-pos
0386b551
AM
776 ;; Symbol just scanned back over (e.g. 'while or 'boundary).
777 ;; See above.
778 sym
779 ;; Current state in the automaton. See above.
780 state
781 ;; Current saved positions. See above.
782 saved-pos
783 ;; Stack of conses (state . saved-pos).
784 stack
785 ;; Regexp which matches "for", "if", etc.
786 (cond-key (or c-opt-block-stmt-key
a66cd3ee 787 "\\<\\>")) ; Matches nothing.
0386b551
AM
788 ;; Return value.
789 (ret 'same)
790 ;; Positions of the last three sexps or bounds we've stopped at.
791 tok ptok pptok)
a66cd3ee
MS
792
793 (save-restriction
794 (if lim (narrow-to-region lim (point-max)))
795
796 (if (save-excursion
797 (and (c-beginning-of-macro)
798 (/= (point) start)))
799 (setq macro-start (point)))
800
d9e94c22 801 ;; Try to skip back over unary operator characters, to register
a66cd3ee
MS
802 ;; that we've moved.
803 (while (progn
804 (setq pos (point))
0386b551
AM
805 (c-backward-syntactic-ws)
806 ;; Protect post-++/-- operators just before a virtual semicolon.
807 (and (not (c-at-vsemi-p))
808 (/= (skip-chars-backward "-+!*&~@`#") 0))))
d9e94c22
MS
809
810 ;; Skip back over any semicolon here. If it was a bare semicolon, we're
0386b551 811 ;; done. Later on we ignore the boundaries for statements that don't
d9e94c22
MS
812 ;; contain any sexp. The only thing that is affected is that the error
813 ;; checking is a little less strict, and we really don't bother.
a66cd3ee
MS
814 (if (and (memq (char-before) delims)
815 (progn (forward-char -1)
816 (setq saved (point))
0386b551 817 (c-backward-syntactic-ws)
a66cd3ee
MS
818 (or (memq (char-before) delims)
819 (memq (char-before) '(?: nil))
d9e94c22 820 (eq (char-syntax (char-before)) ?\()
0386b551 821 (c-at-vsemi-p))))
a66cd3ee
MS
822 (setq ret 'previous
823 pos saved)
824
825 ;; Begin at start and not pos to detect macros if we stand
826 ;; directly after the #.
827 (goto-char start)
828 (if (looking-at "\\<\\|\\W")
829 ;; Record this as the first token if not starting inside it.
830 (setq tok start))
831
b248a85d
AM
832
833 ;; The following while loop goes back one sexp (balanced parens,
834 ;; etc. with contents, or symbol or suchlike) each iteration. This
835 ;; movement is accomplished with a call to c-backward-sexp approx 170
836 ;; lines below.
837 ;;
838 ;; The loop is exited only by throwing nil to the (catch 'loop ...):
839 ;; 1. On reaching the start of a macro;
840 ;; 2. On having passed a stmt boundary with the PDA stack empty;
841 ;; 3. On reaching the start of an Objective C method def;
842 ;; 4. From macro `c-bos-pop-state'; when the stack is empty;
843 ;; 5. From macro `c-bos-pop-state-and-retry' when the stack is empty.
a66cd3ee
MS
844 (while
845 (catch 'loop ;; Throw nil to break, non-nil to continue.
846 (cond
b248a85d 847 ;; Are we in a macro, just after the opening #?
a66cd3ee 848 ((save-excursion
0386b551 849 (and macro-start ; Always NIL for AWK.
a66cd3ee
MS
850 (progn (skip-chars-backward " \t")
851 (eq (char-before) ?#))
852 (progn (setq saved (1- (point)))
853 (beginning-of-line)
854 (not (eq (char-before (1- (point))) ?\\)))
d9e94c22 855 (looking-at c-opt-cpp-start)
a66cd3ee
MS
856 (progn (skip-chars-forward " \t")
857 (eq (point) saved))))
858 (goto-char saved)
859 (if (and (c-forward-to-cpp-define-body)
860 (progn (c-forward-syntactic-ws start)
861 (< (point) start)))
862 ;; Stop at the first token in the content of the macro.
863 (setq pos (point)
864 ignore-labels t) ; Avoid the label check on exit.
865 (setq pos saved
866 ret 'macro
867 ignore-labels t))
b248a85d 868 (throw 'loop nil)) ; 1. Start of macro.
a66cd3ee 869
d9e94c22
MS
870 ;; Do a round through the automaton if we've just passed a
871 ;; statement boundary or passed a "while"-like token.
a66cd3ee
MS
872 ((or sym
873 (and (looking-at cond-key)
874 (setq sym (intern (match-string 1)))))
875
876 (when (and (< pos start) (null stack))
b248a85d 877 (throw 'loop nil)) ; 2. Statement boundary.
a66cd3ee 878
d9e94c22
MS
879 ;; The PDA state handling.
880 ;;
037558bf 881 ;; Refer to the description of the PDA in the opening
d9e94c22
MS
882 ;; comments. In the following OR form, the first leaf
883 ;; attempts to handles one of the specific actions detailed
884 ;; (e.g., finding token "if" whilst in state `else-boundary').
885 ;; We drop through to the second leaf (which handles common
886 ;; state) if no specific handler is found in the first cond.
887 ;; If a parsing error is detected (e.g. an "else" with no
888 ;; preceding "if"), we throw to the enclosing catch.
889 ;;
890 ;; Note that the (eq state 'else) means
891 ;; "we've just passed an else", NOT "we're looking for an
892 ;; else".
a66cd3ee
MS
893 (or (cond
894 ((eq state 'else)
895 (if (eq sym 'boundary)
896 (setq state 'else-boundary)
897 (c-bos-report-error)
898 (c-bos-pop-state-and-retry)))
899
900 ((eq state 'else-boundary)
901 (cond ((eq sym 'if)
902 (c-bos-pop-state (setq ret 'beginning)))
903 ((eq sym 'boundary)
904 (c-bos-report-error)
905 (c-bos-pop-state))))
906
907 ((eq state 'while)
908 (if (and (eq sym 'boundary)
909 ;; Since this can cause backtracking we do a
910 ;; little more careful analysis to avoid it:
911 ;; If there's a label in front of the while
912 ;; it can't be part of a do-while.
913 (not after-labels-pos))
914 (progn (c-bos-save-pos)
915 (setq state 'while-boundary))
d9e94c22 916 (c-bos-pop-state-and-retry))) ; Can't be a do-while
a66cd3ee
MS
917
918 ((eq state 'while-boundary)
919 (cond ((eq sym 'do)
920 (c-bos-pop-state (setq ret 'beginning)))
d9e94c22
MS
921 ((eq sym 'boundary) ; isn't a do-while
922 (c-bos-restore-pos) ; the position of the while
923 (c-bos-pop-state)))) ; no longer searching for do.
a66cd3ee
MS
924
925 ((eq state 'catch)
926 (if (eq sym 'boundary)
927 (setq state 'catch-boundary)
928 (c-bos-report-error)
929 (c-bos-pop-state-and-retry)))
930
931 ((eq state 'catch-boundary)
932 (cond
933 ((eq sym 'try)
934 (c-bos-pop-state (setq ret 'beginning)))
935 ((eq sym 'catch)
936 (setq state 'catch))
937 ((eq sym 'boundary)
938 (c-bos-report-error)
939 (c-bos-pop-state)))))
940
d9e94c22
MS
941 ;; This is state common. We get here when the previous
942 ;; cond statement found no particular state handler.
a66cd3ee 943 (cond ((eq sym 'boundary)
d9e94c22
MS
944 ;; If we have a boundary at the start
945 ;; position we push a frame to go to the
946 ;; previous statement.
947 (if (>= pos start)
948 (c-bos-push-state)
949 (c-bos-pop-state)))
a66cd3ee
MS
950 ((eq sym 'else)
951 (c-bos-push-state)
952 (c-bos-save-error-info 'if 'else)
953 (setq state 'else))
954 ((eq sym 'while)
0386b551
AM
955 ;; Is this a real while, or a do-while?
956 ;; The next `when' triggers unless we are SURE that
40ba43b4 957 ;; the `while' is not the tail end of a `do-while'.
a66cd3ee 958 (when (or (not pptok)
d9e94c22 959 (memq (char-after pptok) delims)
0386b551
AM
960 ;; The following kludge is to prevent
961 ;; infinite recursion when called from
962 ;; c-awk-after-if-for-while-condition-p,
963 ;; or the like.
964 (and (eq (point) start)
965 (c-vsemi-status-unknown-p))
966 (c-at-vsemi-p pptok))
a66cd3ee
MS
967 ;; Since this can cause backtracking we do a
968 ;; little more careful analysis to avoid it: If
0386b551
AM
969 ;; the while isn't followed by a (possibly
970 ;; virtual) semicolon it can't be a do-while.
a66cd3ee
MS
971 (c-bos-push-state)
972 (setq state 'while)))
973 ((memq sym '(catch finally))
974 (c-bos-push-state)
975 (c-bos-save-error-info 'try sym)
976 (setq state 'catch))))
977
978 (when c-maybe-labelp
979 ;; We're either past a statement boundary or at the
980 ;; start of a statement, so throw away any label data
981 ;; for the previous one.
982 (setq after-labels-pos nil
983 last-label-pos nil
984 c-maybe-labelp nil))))
985
d9e94c22
MS
986 ;; Step to the previous sexp, but not if we crossed a
987 ;; boundary, since that doesn't consume an sexp.
a66cd3ee
MS
988 (if (eq sym 'boundary)
989 (setq ret 'previous)
d9e94c22
MS
990
991 ;; HERE IS THE SINGLE PLACE INSIDE THE PDA LOOP WHERE WE MOVE
0386b551
AM
992 ;; BACKWARDS THROUGH THE SOURCE.
993
0386b551 994 (c-backward-syntactic-ws)
0386b551 995 (let ((before-sws-pos (point))
0386b551
AM
996 ;; The end position of the area to search for statement
997 ;; barriers in this round.
b248a85d 998 (maybe-after-boundary-pos pos))
0386b551 999
b248a85d
AM
1000 ;; Go back over exactly one logical sexp, taking proper
1001 ;; account of macros and escaped EOLs.
0386b551
AM
1002 (while
1003 (progn
1004 (unless (c-safe (c-backward-sexp) t)
1005 ;; Give up if we hit an unbalanced block. Since the
1006 ;; stack won't be empty the code below will report a
1007 ;; suitable error.
a66cd3ee 1008 (throw 'loop nil))
b248a85d
AM
1009 (cond
1010 ;; Have we moved into a macro?
1011 ((and (not macro-start)
1012 (c-beginning-of-macro))
1013 ;; Have we crossed a statement boundary? If not,
1014 ;; keep going back until we find one or a "real" sexp.
1015 (and
1016 (save-excursion
1017 (c-end-of-macro)
1018 (not (c-crosses-statement-barrier-p
1019 (point) maybe-after-boundary-pos)))
1020 (setq maybe-after-boundary-pos (point))))
1021 ;; Have we just gone back over an escaped NL? This
1022 ;; doesn't count as a sexp.
1023 ((looking-at "\\\\$")))))
1024
1025 ;; Have we crossed a statement boundary?
1026 (setq boundary-pos
1027 (cond
1028 ;; Are we at a macro beginning?
1029 ((and (not macro-start)
1030 c-opt-cpp-prefix
1031 (looking-at c-opt-cpp-prefix))
1032 (save-excursion
1033 (c-end-of-macro)
1034 (c-crosses-statement-barrier-p
1035 (point) maybe-after-boundary-pos)))
1036 ;; Just gone back over a brace block?
1037 ((and
1038 (eq (char-after) ?{)
1039 (not (c-looking-at-inexpr-block lim nil t)))
1040 (save-excursion
1041 (c-forward-sexp) (point)))
1042 ;; Just gone back over some paren block?
1043 ((looking-at "\\s\(")
1044 (save-excursion
1045 (goto-char (1+ (c-down-list-backward
1046 before-sws-pos)))
1047 (c-crosses-statement-barrier-p
1048 (point) maybe-after-boundary-pos)))
1049 ;; Just gone back over an ordinary symbol of some sort?
1050 (t (c-crosses-statement-barrier-p
1051 (point) maybe-after-boundary-pos))))
91af3942 1052
b248a85d
AM
1053 (when boundary-pos
1054 (setq pptok ptok
1055 ptok tok
1056 tok boundary-pos
1057 sym 'boundary)
1058 ;; Like a C "continue". Analyze the next sexp.
1059 (throw 'loop t))))
a66cd3ee
MS
1060
1061 ;; ObjC method def?
1062 (when (and c-opt-method-key
1063 (setq saved (c-in-method-def-p)))
1064 (setq pos saved
1065 ignore-labels t) ; Avoid the label check on exit.
b248a85d
AM
1066 (throw 'loop nil)) ; 3. ObjC method def.
1067
1068 ;; Might we have a bitfield declaration, "<type> <id> : <size>"?
1069 (if c-has-bitfields
1070 (cond
1071 ;; The : <size> and <id> fields?
1072 ((and (numberp c-maybe-labelp)
1073 (not bitfield-size-pos)
1074 (save-excursion
1075 (goto-char (or tok start))
1076 (not (looking-at c-keywords-regexp)))
1077 (not (looking-at c-keywords-regexp))
1078 (not (c-punctuation-in (point) c-maybe-labelp)))
1079 (setq bitfield-size-pos (or tok start)
1080 bitfield-id-pos (point)))
1081 ;; The <type> field?
1082 ((and bitfield-id-pos
1083 (not bitfield-type-pos))
1084 (if (and (looking-at c-symbol-key) ; Can only be an integer type. :-)
1085 (not (looking-at c-not-primitive-type-keywords-regexp))
1086 (not (c-punctuation-in (point) tok)))
1087 (setq bitfield-type-pos (point))
1088 (setq bitfield-size-pos nil
1089 bitfield-id-pos nil)))))
a66cd3ee 1090
0386b551
AM
1091 ;; Handle labels.
1092 (unless (eq ignore-labels t)
1093 (when (numberp c-maybe-labelp)
cb694ab7
AM
1094 ;; `c-crosses-statement-barrier-p' has found a colon, so we
1095 ;; might be in a label now. Have we got a real label
1096 ;; (including a case label) or something like C++'s "public:"?
d28e7f28
AM
1097 ;; A case label might use an expression rather than a token.
1098 (setq after-case:-pos (or tok start))
020716e1
AM
1099 (if (or (looking-at c-nonlabel-token-key) ; e.g. "while" or "'a'"
1100 ;; Catch C++'s inheritance construct "class foo : bar".
1101 (save-excursion
1102 (and
1103 (c-safe (c-backward-sexp) t)
1104 (looking-at c-nonlabel-token-2-key))))
d28e7f28
AM
1105 (setq c-maybe-labelp nil)
1106 (if after-labels-pos ; Have we already encountered a label?
1107 (if (not last-label-pos)
1108 (setq last-label-pos (or tok start)))
1109 (setq after-labels-pos (or tok start)))
1110 (setq c-maybe-labelp t
1111 label-good-pos nil))) ; bogus "label"
cb694ab7
AM
1112
1113 (when (and (not label-good-pos) ; i.e. no invalid "label"'s yet
1114 ; been found.
1115 (looking-at c-nonlabel-token-key)) ; e.g. "while :"
0386b551
AM
1116 ;; We're in a potential label and it's the first
1117 ;; time we've found something that isn't allowed in
1118 ;; one.
1119 (setq label-good-pos (or tok start))))
1120
1121 ;; We've moved back by a sexp, so update the token positions.
a66cd3ee
MS
1122 (setq sym nil
1123 pptok ptok
1124 ptok tok
1125 tok (point)
b248a85d
AM
1126 pos tok) ; always non-nil
1127 ) ; end of (catch loop ....)
1128 ) ; end of sexp-at-a-time (while ....)
91af3942 1129
a66cd3ee
MS
1130 ;; If the stack isn't empty there might be errors to report.
1131 (while stack
1132 (if (and (vectorp saved-pos) (eq (length saved-pos) 3))
1133 (c-bos-report-error))
1134 (setq saved-pos (cdr (car stack))
1135 stack (cdr stack)))
1136
1137 (when (and (eq ret 'same)
1138 (not (memq sym '(boundary ignore nil))))
1139 ;; Need to investigate closer whether we've crossed
1140 ;; between a substatement and its containing statement.
1141 (if (setq saved (if (looking-at c-block-stmt-1-key)
1142 ptok
1143 pptok))
1144 (cond ((> start saved) (setq pos saved))
1145 ((= start saved) (setq ret 'up)))))
1146
0386b551
AM
1147 (when (and (not ignore-labels)
1148 (eq c-maybe-labelp t)
d9e94c22 1149 (not (eq ret 'beginning))
0386b551 1150 after-labels-pos
b248a85d 1151 (not bitfield-type-pos) ; Bitfields take precedence over labels.
0386b551
AM
1152 (or (not label-good-pos)
1153 (<= label-good-pos pos)
1154 (progn
1155 (goto-char (if (and last-label-pos
1156 (< last-label-pos start))
1157 last-label-pos
1158 pos))
1159 (looking-at c-label-kwds-regexp))))
a66cd3ee
MS
1160 ;; We're in a label. Maybe we should step to the statement
1161 ;; after it.
1162 (if (< after-labels-pos start)
1163 (setq pos after-labels-pos)
1164 (setq ret 'label)
0386b551
AM
1165 (if (and last-label-pos (< last-label-pos start))
1166 ;; Might have jumped over several labels. Go to the last one.
a66cd3ee
MS
1167 (setq pos last-label-pos)))))
1168
d28e7f28 1169 ;; Have we got "case <expression>:"?
a66cd3ee 1170 (goto-char pos)
d28e7f28
AM
1171 (when (and after-case:-pos
1172 (not (eq ret 'beginning))
1173 (looking-at c-case-kwds-regexp))
1174 (if (< after-case:-pos start)
0000ee90
AM
1175 (setq pos after-case:-pos))
1176 (if (eq ret 'same)
1177 (setq ret 'label)))
d28e7f28
AM
1178
1179 ;; Skip over the unary operators that can start the statement.
a66cd3ee 1180 (while (progn
0386b551
AM
1181 (c-backward-syntactic-ws)
1182 ;; protect AWK post-inc/decrement operators, etc.
1183 (and (not (c-at-vsemi-p (point)))
1184 (/= (skip-chars-backward "-+!*&~@`#") 0)))
a66cd3ee
MS
1185 (setq pos (point)))
1186 (goto-char pos)
1187 ret)))
785eecbb 1188
b248a85d
AM
1189(defun c-punctuation-in (from to)
1190 "Return non-nil if there is a non-comment non-macro punctuation character
1191between FROM and TO. FROM must not be in a string or comment. The returned
1192value is the position of the first such character."
1193 (save-excursion
1194 (goto-char from)
1195 (let ((pos (point)))
1196 (while (progn (skip-chars-forward c-symbol-chars to)
1197 (c-forward-syntactic-ws to)
1198 (> (point) pos))
1199 (setq pos (point))))
1200 (and (< (point) to) (point))))
1201
785eecbb 1202(defun c-crosses-statement-barrier-p (from to)
a66cd3ee
MS
1203 "Return non-nil if buffer positions FROM to TO cross one or more
1204statement or declaration boundaries. The returned value is actually
d9e94c22
MS
1205the position of the earliest boundary char. FROM must not be within
1206a string or comment.
a66cd3ee
MS
1207
1208The variable `c-maybe-labelp' is set to the position of the first `:' that
1209might start a label (i.e. not part of `::' and not preceded by `?'). If a
0386b551
AM
1210single `?' is found, then `c-maybe-labelp' is cleared.
1211
1212For AWK, a statement which is terminated by an EOL (not a \; or a }) is
1213regarded as having a \"virtual semicolon\" immediately after the last token on
b414f371 1214the line. If this virtual semicolon is _at_ from, the function recognizes it.
0386b551
AM
1215
1216Note that this function might do hidden buffer changes. See the
1217comment at the start of cc-engine.el for more info."
536610a4
AM
1218 (let* ((skip-chars
1219 ;; If the current language has CPP macros, insert # into skip-chars.
1220 (if c-opt-cpp-symbol
1221 (concat (substring c-stmt-delim-chars 0 1) ; "^"
1222 c-opt-cpp-symbol ; usually "#"
1223 (substring c-stmt-delim-chars 1)) ; e.g. ";{}?:"
1224 c-stmt-delim-chars))
1225 (non-skip-list
1226 (append (substring skip-chars 1) nil)) ; e.g. (?# ?\; ?{ ?} ?? ?:)
1227 lit-range vsemi-pos)
1228 (save-restriction
1229 (widen)
1230 (save-excursion
1231 (catch 'done
1232 (goto-char from)
1233 (while (progn (skip-chars-forward
1234 skip-chars
1235 (min to (c-point 'bonl)))
1236 (< (point) to))
1237 (cond
1238 ;; Virtual semicolon?
1239 ((and (bolp)
1240 (save-excursion
1241 (progn
1242 (if (setq lit-range (c-literal-limits from)) ; Have we landed in a string/comment?
1243 (goto-char (car lit-range)))
1244 (c-backward-syntactic-ws) ; ? put a limit here, maybe?
1245 (setq vsemi-pos (point))
1246 (c-at-vsemi-p))))
1247 (throw 'done vsemi-pos))
1248 ;; In a string/comment?
2e492df3 1249 ((setq lit-range (c-literal-limits from))
536610a4
AM
1250 (goto-char (cdr lit-range)))
1251 ((eq (char-after) ?:)
1252 (forward-char)
1253 (if (and (eq (char-after) ?:)
1254 (< (point) to))
1255 ;; Ignore scope operators.
1256 (forward-char)
1257 (setq c-maybe-labelp (1- (point)))))
1258 ((eq (char-after) ??)
1259 ;; A question mark. Can't be a label, so stop
1260 ;; looking for more : and ?.
1261 (setq c-maybe-labelp nil
1262 skip-chars (substring c-stmt-delim-chars 0 -2)))
1263 ;; At a CPP construct?
1264 ((and c-opt-cpp-symbol (looking-at c-opt-cpp-symbol)
1265 (save-excursion
1266 (forward-line 0)
1267 (looking-at c-opt-cpp-prefix)))
1268 (c-end-of-macro))
1269 ((memq (char-after) non-skip-list)
1270 (throw 'done (point)))))
1271 ;; In trailing space after an as yet undetected virtual semicolon?
1272 (c-backward-syntactic-ws from)
1273 (if (and (< (point) to)
1274 (c-at-vsemi-p))
1275 (point)
1276 nil))))))
0386b551
AM
1277
1278(defun c-at-statement-start-p ()
1279 "Return non-nil if the point is at the first token in a statement
1280or somewhere in the syntactic whitespace before it.
1281
1282A \"statement\" here is not restricted to those inside code blocks.
1283Any kind of declaration-like construct that occur outside function
1284bodies is also considered a \"statement\".
1285
1286Note that this function might do hidden buffer changes. See the
1287comment at the start of cc-engine.el for more info."
1288
1289 (save-excursion
1290 (let ((end (point))
1291 c-maybe-labelp)
1292 (c-syntactic-skip-backward (substring c-stmt-delim-chars 1) nil t)
1293 (or (bobp)
1294 (eq (char-before) ?})
1295 (and (eq (char-before) ?{)
1296 (not (and c-special-brace-lists
1297 (progn (backward-char)
1298 (c-looking-at-special-brace-list)))))
1299 (c-crosses-statement-barrier-p (point) end)))))
1300
1301(defun c-at-expression-start-p ()
1302 "Return non-nil if the point is at the first token in an expression or
1303statement, or somewhere in the syntactic whitespace before it.
1304
1305An \"expression\" here is a bit different from the normal language
1306grammar sense: It's any sequence of expression tokens except commas,
1307unless they are enclosed inside parentheses of some kind. Also, an
1308expression never continues past an enclosing parenthesis, but it might
1309contain parenthesis pairs of any sort except braces.
1310
1311Since expressions never cross statement boundaries, this function also
1312recognizes statement beginnings, just like `c-at-statement-start-p'.
1313
1314Note that this function might do hidden buffer changes. See the
1315comment at the start of cc-engine.el for more info."
1316
1317 (save-excursion
1318 (let ((end (point))
1319 (c-stmt-delim-chars c-stmt-delim-chars-with-comma)
1320 c-maybe-labelp)
1321 (c-syntactic-skip-backward (substring c-stmt-delim-chars 1) nil t)
1322 (or (bobp)
1323 (memq (char-before) '(?{ ?}))
1324 (save-excursion (backward-char)
1325 (looking-at "\\s("))
1326 (c-crosses-statement-barrier-p (point) end)))))
785eecbb
RS
1327
1328\f
d9e94c22
MS
1329;; A set of functions that covers various idiosyncrasies in
1330;; implementations of `forward-comment'.
1331
1332;; Note: Some emacsen considers incorrectly that any line comment
1333;; ending with a backslash continues to the next line. I can't think
1334;; of any way to work around that in a reliable way without changing
1335;; the buffer, though. Suggestions welcome. ;) (No, temporarily
1336;; changing the syntax for backslash doesn't work since we must treat
1337;; escapes in string literals correctly.)
1338
1339(defun c-forward-single-comment ()
1340 "Move forward past whitespace and the closest following comment, if any.
1341Return t if a comment was found, nil otherwise. In either case, the
1342point is moved past the following whitespace. Line continuations,
1343i.e. a backslashes followed by line breaks, are treated as whitespace.
1344The line breaks that end line comments are considered to be the
1345comment enders, so the point will be put on the beginning of the next
1346line if it moved past a line comment.
1347
1348This function does not do any hidden buffer changes."
1349
1350 (let ((start (point)))
1351 (when (looking-at "\\([ \t\n\r\f\v]\\|\\\\[\n\r]\\)+")
1352 (goto-char (match-end 0)))
1353
1354 (when (forward-comment 1)
1355 (if (eobp)
1356 ;; Some emacsen (e.g. XEmacs 21) return t when moving
1357 ;; forwards at eob.
1358 nil
1359
1360 ;; Emacs includes the ending newline in a b-style (c++)
1361 ;; comment, but XEmacs doesn't. We depend on the Emacs
1362 ;; behavior (which also is symmetric).
1363 (if (and (eolp) (elt (parse-partial-sexp start (point)) 7))
1364 (condition-case nil (forward-char 1)))
1365
1366 t))))
1367
1368(defsubst c-forward-comments ()
1369 "Move forward past all following whitespace and comments.
1370Line continuations, i.e. a backslashes followed by line breaks, are
1371treated as whitespace.
1372
0386b551
AM
1373Note that this function might do hidden buffer changes. See the
1374comment at the start of cc-engine.el for more info."
d9e94c22
MS
1375
1376 (while (or
1377 ;; If forward-comment in at least XEmacs 21 is given a large
1378 ;; positive value, it'll loop all the way through if it hits
1379 ;; eob.
1380 (and (forward-comment 5)
1381 ;; Some emacsen (e.g. XEmacs 21) return t when moving
1382 ;; forwards at eob.
1383 (not (eobp)))
1384
1385 (when (looking-at "\\\\[\n\r]")
1386 (forward-char 2)
1387 t))))
1388
1389(defun c-backward-single-comment ()
1390 "Move backward past whitespace and the closest preceding comment, if any.
1391Return t if a comment was found, nil otherwise. In either case, the
1392point is moved past the preceding whitespace. Line continuations,
1393i.e. a backslashes followed by line breaks, are treated as whitespace.
1394The line breaks that end line comments are considered to be the
1395comment enders, so the point cannot be at the end of the same line to
1396move over a line comment.
1397
1398This function does not do any hidden buffer changes."
1399
1400 (let ((start (point)))
1401 ;; When we got newline terminated comments, forward-comment in all
1402 ;; supported emacsen so far will stop at eol of each line not
1403 ;; ending with a comment when moving backwards. This corrects for
1404 ;; that, and at the same time handles line continuations.
1405 (while (progn
1406 (skip-chars-backward " \t\n\r\f\v")
1407 (and (looking-at "[\n\r]")
0386b551 1408 (eq (char-before) ?\\)))
d9e94c22
MS
1409 (backward-char))
1410
1411 (if (bobp)
1412 ;; Some emacsen (e.g. Emacs 19.34) return t when moving
1413 ;; backwards at bob.
1414 nil
1415
1416 ;; Leave point after the closest following newline if we've
1417 ;; backed up over any above, since forward-comment won't move
1418 ;; backward over a line comment if point is at the end of the
1419 ;; same line.
1420 (re-search-forward "\\=\\s *[\n\r]" start t)
1421
bf2c1571 1422 (if (if (let (open-paren-in-column-0-is-defun-start) (forward-comment -1))
d9e94c22
MS
1423 (if (eolp)
1424 ;; If forward-comment above succeeded and we're at eol
1425 ;; then the newline we moved over above didn't end a
1426 ;; line comment, so we give it another go.
bf2c1571
AM
1427 (let (open-paren-in-column-0-is-defun-start)
1428 (forward-comment -1))
d9e94c22
MS
1429 t))
1430
1431 ;; Emacs <= 20 and XEmacs move back over the closer of a
1432 ;; block comment that lacks an opener.
1433 (if (looking-at "\\*/")
1434 (progn (forward-char 2) nil)
1435 t)))))
1436
1437(defsubst c-backward-comments ()
1438 "Move backward past all preceding whitespace and comments.
1439Line continuations, i.e. a backslashes followed by line breaks, are
1440treated as whitespace. The line breaks that end line comments are
1441considered to be the comment enders, so the point cannot be at the end
0386b551
AM
1442of the same line to move over a line comment. Unlike
1443c-backward-syntactic-ws, this function doesn't move back over
1444preprocessor directives.
d9e94c22 1445
0386b551
AM
1446Note that this function might do hidden buffer changes. See the
1447comment at the start of cc-engine.el for more info."
d9e94c22
MS
1448
1449 (let ((start (point)))
1450 (while (and
0386b551 1451 ;; `forward-comment' in some emacsen (e.g. XEmacs 21.4)
d9e94c22
MS
1452 ;; return t when moving backwards at bob.
1453 (not (bobp))
1454
14b8c3d9
AM
1455 (if (let (open-paren-in-column-0-is-defun-start moved-comment)
1456 (while
1457 (and (not (setq moved-comment (forward-comment -1)))
1458 ;; Cope specifically with ^M^J here -
1459 ;; forward-comment sometimes gets stuck after ^Ms,
1460 ;; sometimes after ^M^J.
1461 (or
1462 (when (eq (char-before) ?\r)
1463 (backward-char)
1464 t)
1465 (when (and (eq (char-before) ?\n)
1466 (eq (char-before (1- (point))) ?\r))
1467 (backward-char 2)
1468 t))))
1469 moved-comment)
d9e94c22
MS
1470 (if (looking-at "\\*/")
1471 ;; Emacs <= 20 and XEmacs move back over the
1472 ;; closer of a block comment that lacks an opener.
1473 (progn (forward-char 2) nil)
1474 t)
1475
1476 ;; XEmacs treats line continuations as whitespace but
1477 ;; only in the backward direction, which seems a bit
1478 ;; odd. Anyway, this is necessary for Emacs.
1479 (when (and (looking-at "[\n\r]")
1480 (eq (char-before) ?\\)
1481 (< (point) start))
1482 (backward-char)
1483 t))))))
1484
1485\f
d9e94c22 1486;; Tools for skipping over syntactic whitespace.
a66cd3ee 1487
d9e94c22
MS
1488;; The following functions use text properties to cache searches over
1489;; large regions of syntactic whitespace. It works as follows:
1490;;
1491;; o If a syntactic whitespace region contains anything but simple
1492;; whitespace (i.e. space, tab and line breaks), the text property
1493;; `c-in-sws' is put over it. At places where we have stopped
1494;; within that region there's also a `c-is-sws' text property.
1495;; That since there typically are nested whitespace inside that
1496;; must be handled separately, e.g. whitespace inside a comment or
1497;; cpp directive. Thus, from one point with `c-is-sws' it's safe
1498;; to jump to another point with that property within the same
1499;; `c-in-sws' region. It can be likened to a ladder where
1500;; `c-in-sws' marks the bars and `c-is-sws' the rungs.
1501;;
1502;; o The `c-is-sws' property is put on the simple whitespace chars at
1503;; a "rung position" and also maybe on the first following char.
1504;; As many characters as can be conveniently found in this range
1505;; are marked, but no assumption can be made that the whole range
1506;; is marked (it could be clobbered by later changes, for
1507;; instance).
1508;;
1509;; Note that some part of the beginning of a sequence of simple
1510;; whitespace might be part of the end of a preceding line comment
1511;; or cpp directive and must not be considered part of the "rung".
1512;; Such whitespace is some amount of horizontal whitespace followed
1513;; by a newline. In the case of cpp directives it could also be
1514;; two newlines with horizontal whitespace between them.
1515;;
1516;; The reason to include the first following char is to cope with
1517;; "rung positions" that doesn't have any ordinary whitespace. If
1518;; `c-is-sws' is put on a token character it does not have
1519;; `c-in-sws' set simultaneously. That's the only case when that
1520;; can occur, and the reason for not extending the `c-in-sws'
1521;; region to cover it is that the `c-in-sws' region could then be
1522;; accidentally merged with a following one if the token is only
1523;; one character long.
1524;;
1525;; o On buffer changes the `c-in-sws' and `c-is-sws' properties are
1526;; removed in the changed region. If the change was inside
1527;; syntactic whitespace that means that the "ladder" is broken, but
1528;; a later call to `c-forward-sws' or `c-backward-sws' will use the
1529;; parts on either side and use an ordinary search only to "repair"
1530;; the gap.
1531;;
1532;; Special care needs to be taken if a region is removed: If there
1533;; are `c-in-sws' on both sides of it which do not connect inside
1534;; the region then they can't be joined. If e.g. a marked macro is
1535;; broken, syntactic whitespace inside the new text might be
1536;; marked. If those marks would become connected with the old
1537;; `c-in-sws' range around the macro then we could get a ladder
1538;; with one end outside the macro and the other at some whitespace
1539;; within it.
1540;;
1541;; The main motivation for this system is to increase the speed in
1542;; skipping over the large whitespace regions that can occur at the
1543;; top level in e.g. header files that contain a lot of comments and
1544;; cpp directives. For small comments inside code it's probably
1545;; slower than using `forward-comment' straightforwardly, but speed is
1546;; not a significant factor there anyway.
1547
1548; (defface c-debug-is-sws-face
1549; '((t (:background "GreenYellow")))
1550; "Debug face to mark the `c-is-sws' property.")
1551; (defface c-debug-in-sws-face
1552; '((t (:underline t)))
1553; "Debug face to mark the `c-in-sws' property.")
1554
1555; (defun c-debug-put-sws-faces ()
1556; ;; Put the sws debug faces on all the `c-is-sws' and `c-in-sws'
1557; ;; properties in the buffer.
1558; (interactive)
1559; (save-excursion
0386b551 1560; (c-save-buffer-state (in-face)
d9e94c22
MS
1561; (goto-char (point-min))
1562; (setq in-face (if (get-text-property (point) 'c-is-sws)
1563; (point)))
1564; (while (progn
1565; (goto-char (next-single-property-change
1566; (point) 'c-is-sws nil (point-max)))
1567; (if in-face
1568; (progn
1569; (c-debug-add-face in-face (point) 'c-debug-is-sws-face)
1570; (setq in-face nil))
1571; (setq in-face (point)))
1572; (not (eobp))))
1573; (goto-char (point-min))
1574; (setq in-face (if (get-text-property (point) 'c-in-sws)
1575; (point)))
1576; (while (progn
1577; (goto-char (next-single-property-change
1578; (point) 'c-in-sws nil (point-max)))
1579; (if in-face
1580; (progn
1581; (c-debug-add-face in-face (point) 'c-debug-in-sws-face)
1582; (setq in-face nil))
1583; (setq in-face (point)))
1584; (not (eobp)))))))
1585
1586(defmacro c-debug-sws-msg (&rest args)
1587 ;;`(message ,@args)
1588 )
1589
1590(defmacro c-put-is-sws (beg end)
0386b551 1591 ;; This macro does a hidden buffer change.
d9e94c22
MS
1592 `(let ((beg ,beg) (end ,end))
1593 (put-text-property beg end 'c-is-sws t)
1594 ,@(when (facep 'c-debug-is-sws-face)
1595 `((c-debug-add-face beg end 'c-debug-is-sws-face)))))
1596
1597(defmacro c-put-in-sws (beg end)
0386b551 1598 ;; This macro does a hidden buffer change.
d9e94c22
MS
1599 `(let ((beg ,beg) (end ,end))
1600 (put-text-property beg end 'c-in-sws t)
1601 ,@(when (facep 'c-debug-is-sws-face)
1602 `((c-debug-add-face beg end 'c-debug-in-sws-face)))))
1603
1604(defmacro c-remove-is-sws (beg end)
0386b551 1605 ;; This macro does a hidden buffer change.
d9e94c22
MS
1606 `(let ((beg ,beg) (end ,end))
1607 (remove-text-properties beg end '(c-is-sws nil))
1608 ,@(when (facep 'c-debug-is-sws-face)
1609 `((c-debug-remove-face beg end 'c-debug-is-sws-face)))))
1610
1611(defmacro c-remove-in-sws (beg end)
0386b551 1612 ;; This macro does a hidden buffer change.
d9e94c22
MS
1613 `(let ((beg ,beg) (end ,end))
1614 (remove-text-properties beg end '(c-in-sws nil))
1615 ,@(when (facep 'c-debug-is-sws-face)
1616 `((c-debug-remove-face beg end 'c-debug-in-sws-face)))))
1617
1618(defmacro c-remove-is-and-in-sws (beg end)
0386b551 1619 ;; This macro does a hidden buffer change.
d9e94c22
MS
1620 `(let ((beg ,beg) (end ,end))
1621 (remove-text-properties beg end '(c-is-sws nil c-in-sws nil))
1622 ,@(when (facep 'c-debug-is-sws-face)
1623 `((c-debug-remove-face beg end 'c-debug-is-sws-face)
1624 (c-debug-remove-face beg end 'c-debug-in-sws-face)))))
1625
1626(defsubst c-invalidate-sws-region-after (beg end)
1627 ;; Called from `after-change-functions'. Note that if
1628 ;; `c-forward-sws' or `c-backward-sws' are used outside
1629 ;; `c-save-buffer-state' or similar then this will remove the cache
1630 ;; properties right after they're added.
0386b551
AM
1631 ;;
1632 ;; This function does hidden buffer changes.
d9e94c22
MS
1633
1634 (save-excursion
1635 ;; Adjust the end to remove the properties in any following simple
1636 ;; ws up to and including the next line break, if there is any
1637 ;; after the changed region. This is necessary e.g. when a rung
1638 ;; marked empty line is converted to a line comment by inserting
1639 ;; "//" before the line break. In that case the line break would
1640 ;; keep the rung mark which could make a later `c-backward-sws'
1641 ;; move into the line comment instead of over it.
1642 (goto-char end)
1643 (skip-chars-forward " \t\f\v")
1644 (when (and (eolp) (not (eobp)))
1645 (setq end (1+ (point)))))
1646
1647 (when (and (= beg end)
1648 (get-text-property beg 'c-in-sws)
2a15eb73 1649 (> beg (point-min))
d9e94c22
MS
1650 (get-text-property (1- beg) 'c-in-sws))
1651 ;; Ensure that an `c-in-sws' range gets broken. Note that it isn't
1652 ;; safe to keep a range that was continuous before the change. E.g:
1653 ;;
1654 ;; #define foo
1655 ;; \
1656 ;; bar
1657 ;;
1658 ;; There can be a "ladder" between "#" and "b". Now, if the newline
1659 ;; after "foo" is removed then "bar" will become part of the cpp
1660 ;; directive instead of a syntactically relevant token. In that
1661 ;; case there's no longer syntactic ws from "#" to "b".
1662 (setq beg (1- beg)))
1663
1664 (c-debug-sws-msg "c-invalidate-sws-region-after [%s..%s]" beg end)
1665 (c-remove-is-and-in-sws beg end))
1666
1667(defun c-forward-sws ()
1668 ;; Used by `c-forward-syntactic-ws' to implement the unbounded search.
0386b551
AM
1669 ;;
1670 ;; This function might do hidden buffer changes.
d9e94c22
MS
1671
1672 (let (;; `rung-pos' is set to a position as early as possible in the
1673 ;; unmarked part of the simple ws region.
1674 (rung-pos (point)) next-rung-pos rung-end-pos last-put-in-sws-pos
1675 rung-is-marked next-rung-is-marked simple-ws-end
1676 ;; `safe-start' is set when it's safe to cache the start position.
1677 ;; It's not set if we've initially skipped over comments and line
1678 ;; continuations since we might have gone out through the end of a
1679 ;; macro then. This provision makes `c-forward-sws' not populate the
1680 ;; cache in the majority of cases, but otoh is `c-backward-sws' by far
1681 ;; more common.
1682 safe-start)
1683
1684 ;; Skip simple ws and do a quick check on the following character to see
1685 ;; if it's anything that can't start syntactic ws, so we can bail out
1686 ;; early in the majority of cases when there just are a few ws chars.
1687 (skip-chars-forward " \t\n\r\f\v")
1688 (when (looking-at c-syntactic-ws-start)
1689
1690 (setq rung-end-pos (min (1+ (point)) (point-max)))
1691 (if (setq rung-is-marked (text-property-any rung-pos rung-end-pos
1692 'c-is-sws t))
1693 ;; Find the last rung position to avoid setting properties in all
1694 ;; the cases when the marked rung is complete.
1695 ;; (`next-single-property-change' is certain to move at least one
1696 ;; step forward.)
1697 (setq rung-pos (1- (next-single-property-change
1698 rung-is-marked 'c-is-sws nil rung-end-pos)))
1699 ;; Got no marked rung here. Since the simple ws might have started
1700 ;; inside a line comment or cpp directive we must set `rung-pos' as
1701 ;; high as possible.
1702 (setq rung-pos (point)))
1703
2143fa32 1704 (with-silent-modifications
d9e94c22
MS
1705 (while
1706 (progn
1707 (while
1708 (when (and rung-is-marked
1709 (get-text-property (point) 'c-in-sws))
1710
1711 ;; The following search is the main reason that `c-in-sws'
1712 ;; and `c-is-sws' aren't combined to one property.
1713 (goto-char (next-single-property-change
1714 (point) 'c-in-sws nil (point-max)))
1715 (unless (get-text-property (point) 'c-is-sws)
1716 ;; If the `c-in-sws' region extended past the last
1717 ;; `c-is-sws' char we have to go back a bit.
1718 (or (get-text-property (1- (point)) 'c-is-sws)
1719 (goto-char (previous-single-property-change
1720 (point) 'c-is-sws)))
1721 (backward-char))
1722
1723 (c-debug-sws-msg
1724 "c-forward-sws cached move %s -> %s (max %s)"
1725 rung-pos (point) (point-max))
1726
1727 (setq rung-pos (point))
1728 (and (> (skip-chars-forward " \t\n\r\f\v") 0)
1729 (not (eobp))))
1730
1731 ;; We'll loop here if there is simple ws after the last rung.
1732 ;; That means that there's been some change in it and it's
1733 ;; possible that we've stepped into another ladder, so extend
1734 ;; the previous one to join with it if there is one, and try to
1735 ;; use the cache again.
1736 (c-debug-sws-msg
1737 "c-forward-sws extending rung with [%s..%s] (max %s)"
1738 (1+ rung-pos) (1+ (point)) (point-max))
1739 (unless (get-text-property (point) 'c-is-sws)
1740 ;; Remove any `c-in-sws' property from the last char of
1741 ;; the rung before we mark it with `c-is-sws', so that we
1742 ;; won't connect with the remains of a broken "ladder".
1743 (c-remove-in-sws (point) (1+ (point))))
1744 (c-put-is-sws (1+ rung-pos)
1745 (1+ (point)))
1746 (c-put-in-sws rung-pos
1747 (setq rung-pos (point)
1748 last-put-in-sws-pos rung-pos)))
1749
1750 (setq simple-ws-end (point))
1751 (c-forward-comments)
1752
1753 (cond
1754 ((/= (point) simple-ws-end)
1755 ;; Skipped over comments. Don't cache at eob in case the buffer
1756 ;; is narrowed.
1757 (not (eobp)))
1758
1759 ((save-excursion
1760 (and c-opt-cpp-prefix
1761 (looking-at c-opt-cpp-start)
1762 (progn (skip-chars-backward " \t")
1763 (bolp))
1764 (or (bobp)
1765 (progn (backward-char)
1766 (not (eq (char-before) ?\\))))))
1767 ;; Skip a preprocessor directive.
1768 (end-of-line)
1769 (while (and (eq (char-before) ?\\)
1770 (= (forward-line 1) 0))
1771 (end-of-line))
1772 (forward-line 1)
1773 (setq safe-start t)
1774 ;; Don't cache at eob in case the buffer is narrowed.
1775 (not (eobp)))))
1776
1777 ;; We've searched over a piece of non-white syntactic ws. See if this
1778 ;; can be cached.
1779 (setq next-rung-pos (point))
1780 (skip-chars-forward " \t\n\r\f\v")
1781 (setq rung-end-pos (min (1+ (point)) (point-max)))
1782
1783 (if (or
1784 ;; Cache if we haven't skipped comments only, and if we started
1785 ;; either from a marked rung or from a completely uncached
1786 ;; position.
1787 (and safe-start
1788 (or rung-is-marked
1789 (not (get-text-property simple-ws-end 'c-in-sws))))
1790
1791 ;; See if there's a marked rung in the encountered simple ws. If
1792 ;; so then we can cache, unless `safe-start' is nil. Even then
1793 ;; we need to do this to check if the cache can be used for the
1794 ;; next step.
1795 (and (setq next-rung-is-marked
1796 (text-property-any next-rung-pos rung-end-pos
1797 'c-is-sws t))
1798 safe-start))
b2acd789 1799
0ec8351b 1800 (progn
d9e94c22
MS
1801 (c-debug-sws-msg
1802 "c-forward-sws caching [%s..%s] - [%s..%s] (max %s)"
1803 rung-pos (1+ simple-ws-end) next-rung-pos rung-end-pos
1804 (point-max))
1805
1806 ;; Remove the properties for any nested ws that might be cached.
1807 ;; Only necessary for `c-is-sws' since `c-in-sws' will be set
1808 ;; anyway.
1809 (c-remove-is-sws (1+ simple-ws-end) next-rung-pos)
1810 (unless (and rung-is-marked (= rung-pos simple-ws-end))
1811 (c-put-is-sws rung-pos
1812 (1+ simple-ws-end))
1813 (setq rung-is-marked t))
1814 (c-put-in-sws rung-pos
1815 (setq rung-pos (point)
1816 last-put-in-sws-pos rung-pos))
1817 (unless (get-text-property (1- rung-end-pos) 'c-is-sws)
1818 ;; Remove any `c-in-sws' property from the last char of
1819 ;; the rung before we mark it with `c-is-sws', so that we
1820 ;; won't connect with the remains of a broken "ladder".
1821 (c-remove-in-sws (1- rung-end-pos) rung-end-pos))
1822 (c-put-is-sws next-rung-pos
1823 rung-end-pos))
1824
1825 (c-debug-sws-msg
1826 "c-forward-sws not caching [%s..%s] - [%s..%s] (max %s)"
1827 rung-pos (1+ simple-ws-end) next-rung-pos rung-end-pos
1828 (point-max))
1829
1830 ;; Set `rung-pos' for the next rung. It's the same thing here as
1831 ;; initially, except that the rung position is set as early as
1832 ;; possible since we can't be in the ending ws of a line comment or
1833 ;; cpp directive now.
1834 (if (setq rung-is-marked next-rung-is-marked)
1835 (setq rung-pos (1- (next-single-property-change
1836 rung-is-marked 'c-is-sws nil rung-end-pos)))
1837 (setq rung-pos next-rung-pos))
1838 (setq safe-start t)))
1839
1840 ;; Make sure that the newly marked `c-in-sws' region doesn't connect to
1841 ;; another one after the point (which might occur when editing inside a
1842 ;; comment or macro).
1843 (when (eq last-put-in-sws-pos (point))
1844 (cond ((< last-put-in-sws-pos (point-max))
1845 (c-debug-sws-msg
1846 "c-forward-sws clearing at %s for cache separation"
1847 last-put-in-sws-pos)
1848 (c-remove-in-sws last-put-in-sws-pos
1849 (1+ last-put-in-sws-pos)))
1850 (t
1851 ;; If at eob we have to clear the last character before the end
1852 ;; instead since the buffer might be narrowed and there might
1853 ;; be a `c-in-sws' after (point-max). In this case it's
1854 ;; necessary to clear both properties.
1855 (c-debug-sws-msg
1856 "c-forward-sws clearing thoroughly at %s for cache separation"
1857 (1- last-put-in-sws-pos))
1858 (c-remove-is-and-in-sws (1- last-put-in-sws-pos)
1859 last-put-in-sws-pos))))
2143fa32 1860 ))))
b2acd789 1861
d9e94c22
MS
1862(defun c-backward-sws ()
1863 ;; Used by `c-backward-syntactic-ws' to implement the unbounded search.
0386b551
AM
1864 ;;
1865 ;; This function might do hidden buffer changes.
d9e94c22
MS
1866
1867 (let (;; `rung-pos' is set to a position as late as possible in the unmarked
1868 ;; part of the simple ws region.
1869 (rung-pos (point)) next-rung-pos last-put-in-sws-pos
1870 rung-is-marked simple-ws-beg cmt-skip-pos)
1871
1872 ;; Skip simple horizontal ws and do a quick check on the preceding
91af3942 1873 ;; character to see if it's anything that can't end syntactic ws, so we can
d9e94c22
MS
1874 ;; bail out early in the majority of cases when there just are a few ws
1875 ;; chars. Newlines are complicated in the backward direction, so we can't
1876 ;; skip over them.
1877 (skip-chars-backward " \t\f")
1878 (when (and (not (bobp))
1879 (save-excursion
1880 (backward-char)
1881 (looking-at c-syntactic-ws-end)))
1882
1883 ;; Try to find a rung position in the simple ws preceding point, so that
1884 ;; we can get a cache hit even if the last bit of the simple ws has
1885 ;; changed recently.
1886 (setq simple-ws-beg (point))
1887 (skip-chars-backward " \t\n\r\f\v")
1888 (if (setq rung-is-marked (text-property-any
1889 (point) (min (1+ rung-pos) (point-max))
1890 'c-is-sws t))
1891 ;; `rung-pos' will be the earliest marked position, which means that
1892 ;; there might be later unmarked parts in the simple ws region.
1893 ;; It's not worth the effort to fix that; the last part of the
1894 ;; simple ws is also typically edited often, so it could be wasted.
1895 (goto-char (setq rung-pos rung-is-marked))
1896 (goto-char simple-ws-beg))
1897
2143fa32 1898 (with-silent-modifications
d9e94c22
MS
1899 (while
1900 (progn
1901 (while
1902 (when (and rung-is-marked
1903 (not (bobp))
1904 (get-text-property (1- (point)) 'c-in-sws))
1905
1906 ;; The following search is the main reason that `c-in-sws'
1907 ;; and `c-is-sws' aren't combined to one property.
1908 (goto-char (previous-single-property-change
1909 (point) 'c-in-sws nil (point-min)))
1910 (unless (get-text-property (point) 'c-is-sws)
1911 ;; If the `c-in-sws' region extended past the first
1912 ;; `c-is-sws' char we have to go forward a bit.
1913 (goto-char (next-single-property-change
1914 (point) 'c-is-sws)))
1915
1916 (c-debug-sws-msg
1917 "c-backward-sws cached move %s <- %s (min %s)"
1918 (point) rung-pos (point-min))
1919
1920 (setq rung-pos (point))
1921 (if (and (< (min (skip-chars-backward " \t\f\v")
1922 (progn
1923 (setq simple-ws-beg (point))
1924 (skip-chars-backward " \t\n\r\f\v")))
1925 0)
1926 (setq rung-is-marked
1927 (text-property-any (point) rung-pos
1928 'c-is-sws t)))
1929 t
1930 (goto-char simple-ws-beg)
1931 nil))
1932
1933 ;; We'll loop here if there is simple ws before the first rung.
1934 ;; That means that there's been some change in it and it's
1935 ;; possible that we've stepped into another ladder, so extend
1936 ;; the previous one to join with it if there is one, and try to
1937 ;; use the cache again.
1938 (c-debug-sws-msg
1939 "c-backward-sws extending rung with [%s..%s] (min %s)"
1940 rung-is-marked rung-pos (point-min))
1941 (unless (get-text-property (1- rung-pos) 'c-is-sws)
1942 ;; Remove any `c-in-sws' property from the last char of
1943 ;; the rung before we mark it with `c-is-sws', so that we
1944 ;; won't connect with the remains of a broken "ladder".
1945 (c-remove-in-sws (1- rung-pos) rung-pos))
1946 (c-put-is-sws rung-is-marked
1947 rung-pos)
1948 (c-put-in-sws rung-is-marked
1949 (1- rung-pos))
1950 (setq rung-pos rung-is-marked
1951 last-put-in-sws-pos rung-pos))
1952
1953 (c-backward-comments)
1954 (setq cmt-skip-pos (point))
a66cd3ee 1955
d9e94c22
MS
1956 (cond
1957 ((and c-opt-cpp-prefix
1958 (/= cmt-skip-pos simple-ws-beg)
1959 (c-beginning-of-macro))
1960 ;; Inside a cpp directive. See if it should be skipped over.
1961 (let ((cpp-beg (point)))
1962
1963 ;; Move back over all line continuations in the region skipped
1964 ;; over by `c-backward-comments'. If we go past it then we
1965 ;; started inside the cpp directive.
1966 (goto-char simple-ws-beg)
1967 (beginning-of-line)
1968 (while (and (> (point) cmt-skip-pos)
1969 (progn (backward-char)
1970 (eq (char-before) ?\\)))
1971 (beginning-of-line))
1972
1973 (if (< (point) cmt-skip-pos)
1974 ;; Don't move past the cpp directive if we began inside
1975 ;; it. Note that the position at the end of the last line
1976 ;; of the macro is also considered to be within it.
1977 (progn (goto-char cmt-skip-pos)
1978 nil)
1979
1980 ;; It's worthwhile to spend a little bit of effort on finding
1981 ;; the end of the macro, to get a good `simple-ws-beg'
1982 ;; position for the cache. Note that `c-backward-comments'
1983 ;; could have stepped over some comments before going into
1984 ;; the macro, and then `simple-ws-beg' must be kept on the
1985 ;; same side of those comments.
1986 (goto-char simple-ws-beg)
1987 (skip-chars-backward " \t\n\r\f\v")
1988 (if (eq (char-before) ?\\)
1989 (forward-char))
1990 (forward-line 1)
1991 (if (< (point) simple-ws-beg)
1992 ;; Might happen if comments after the macro were skipped
1993 ;; over.
1994 (setq simple-ws-beg (point)))
1995
1996 (goto-char cpp-beg)
1997 t)))
1998
1999 ((/= (save-excursion
2000 (skip-chars-forward " \t\n\r\f\v" simple-ws-beg)
2001 (setq next-rung-pos (point)))
2002 simple-ws-beg)
2003 ;; Skipped over comments. Must put point at the end of
2004 ;; the simple ws at point since we might be after a line
2005 ;; comment or cpp directive that's been partially
2006 ;; narrowed out, and we can't risk marking the simple ws
2007 ;; at the end of it.
2008 (goto-char next-rung-pos)
2009 t)))
2010
2011 ;; We've searched over a piece of non-white syntactic ws. See if this
2012 ;; can be cached.
2013 (setq next-rung-pos (point))
2014 (skip-chars-backward " \t\f\v")
2015
2016 (if (or
2017 ;; Cache if we started either from a marked rung or from a
2018 ;; completely uncached position.
2019 rung-is-marked
2020 (not (get-text-property (1- simple-ws-beg) 'c-in-sws))
2021
2022 ;; Cache if there's a marked rung in the encountered simple ws.
2023 (save-excursion
2024 (skip-chars-backward " \t\n\r\f\v")
2025 (text-property-any (point) (min (1+ next-rung-pos) (point-max))
2026 'c-is-sws t)))
a66cd3ee 2027
d9e94c22
MS
2028 (progn
2029 (c-debug-sws-msg
2030 "c-backward-sws caching [%s..%s] - [%s..%s] (min %s)"
2031 (point) (1+ next-rung-pos)
2032 simple-ws-beg (min (1+ rung-pos) (point-max))
2033 (point-min))
2034
2035 ;; Remove the properties for any nested ws that might be cached.
2036 ;; Only necessary for `c-is-sws' since `c-in-sws' will be set
2037 ;; anyway.
2038 (c-remove-is-sws (1+ next-rung-pos) simple-ws-beg)
2039 (unless (and rung-is-marked (= simple-ws-beg rung-pos))
2040 (let ((rung-end-pos (min (1+ rung-pos) (point-max))))
2041 (unless (get-text-property (1- rung-end-pos) 'c-is-sws)
2042 ;; Remove any `c-in-sws' property from the last char of
2043 ;; the rung before we mark it with `c-is-sws', so that we
2044 ;; won't connect with the remains of a broken "ladder".
2045 (c-remove-in-sws (1- rung-end-pos) rung-end-pos))
2046 (c-put-is-sws simple-ws-beg
2047 rung-end-pos)
2048 (setq rung-is-marked t)))
2049 (c-put-in-sws (setq simple-ws-beg (point)
2050 last-put-in-sws-pos simple-ws-beg)
2051 rung-pos)
2052 (c-put-is-sws (setq rung-pos simple-ws-beg)
2053 (1+ next-rung-pos)))
2054
2055 (c-debug-sws-msg
2056 "c-backward-sws not caching [%s..%s] - [%s..%s] (min %s)"
2057 (point) (1+ next-rung-pos)
2058 simple-ws-beg (min (1+ rung-pos) (point-max))
2059 (point-min))
2060 (setq rung-pos next-rung-pos
2061 simple-ws-beg (point))
2062 ))
2063
2064 ;; Make sure that the newly marked `c-in-sws' region doesn't connect to
2065 ;; another one before the point (which might occur when editing inside a
2066 ;; comment or macro).
2067 (when (eq last-put-in-sws-pos (point))
2068 (cond ((< (point-min) last-put-in-sws-pos)
2069 (c-debug-sws-msg
2070 "c-backward-sws clearing at %s for cache separation"
2071 (1- last-put-in-sws-pos))
2072 (c-remove-in-sws (1- last-put-in-sws-pos)
2073 last-put-in-sws-pos))
2074 ((> (point-min) 1)
2075 ;; If at bob and the buffer is narrowed, we have to clear the
2076 ;; character we're standing on instead since there might be a
2077 ;; `c-in-sws' before (point-min). In this case it's necessary
2078 ;; to clear both properties.
2079 (c-debug-sws-msg
2080 "c-backward-sws clearing thoroughly at %s for cache separation"
2081 last-put-in-sws-pos)
2082 (c-remove-is-and-in-sws last-put-in-sws-pos
2083 (1+ last-put-in-sws-pos)))))
2143fa32 2084 ))))
785eecbb 2085
d9e94c22 2086\f
580fba94
AM
2087;; Other whitespace tools
2088(defun c-partial-ws-p (beg end)
2089 ;; Is the region (beg end) WS, and is there WS (or BOB/EOB) next to the
2090 ;; region? This is a "heuristic" function. .....
b414f371 2091 ;;
c421028a
AM
2092 ;; The motivation for the second bit is to check whether removing this
2093 ;; region would coalesce two symbols.
580fba94
AM
2094 ;;
2095 ;; FIXME!!! This function doesn't check virtual semicolons in any way. Be
2096 ;; careful about using this function for, e.g. AWK. (2007/3/7)
2097 (save-excursion
2098 (let ((end+1 (min (1+ end) (point-max))))
2099 (or (progn (goto-char (max (point-min) (1- beg)))
2100 (c-skip-ws-forward end)
2101 (eq (point) end))
2102 (progn (goto-char beg)
2103 (c-skip-ws-forward end+1)
2104 (eq (point) end+1))))))
2105\f
0386b551 2106;; A system for finding noteworthy parens before the point.
e1c458ae 2107
0ec1d2c5
AM
2108(defconst c-state-cache-too-far 5000)
2109;; A maximum comfortable scanning distance, e.g. between
2110;; `c-state-cache-good-pos' and "HERE" (where we call c-parse-state). When
0d26e0b6 2111;; this distance is exceeded, we take "emergency measures", e.g. by clearing
0ec1d2c5
AM
2112;; the cache and starting again from point-min or a beginning of defun. This
2113;; value can be tuned for efficiency or set to a lower value for testing.
2114
d9e94c22
MS
2115(defvar c-state-cache nil)
2116(make-variable-buffer-local 'c-state-cache)
2117;; The state cache used by `c-parse-state' to cut down the amount of
0ec1d2c5
AM
2118;; searching. It's the result from some earlier `c-parse-state' call. See
2119;; `c-parse-state''s doc string for details of its structure.
0386b551 2120;;
d9e94c22
MS
2121;; The use of the cached info is more effective if the next
2122;; `c-parse-state' call is on a line close by the one the cached state
2123;; was made at; the cache can actually slow down a little if the
2124;; cached state was made very far back in the buffer. The cache is
2125;; most effective if `c-parse-state' is used on each line while moving
2126;; forward.
e1c458ae 2127
0386b551
AM
2128(defvar c-state-cache-good-pos 1)
2129(make-variable-buffer-local 'c-state-cache-good-pos)
0ec1d2c5
AM
2130;; This is a position where `c-state-cache' is known to be correct, or
2131;; nil (see below). It's a position inside one of the recorded unclosed
2132;; parens or the top level, but not further nested inside any literal or
2133;; subparen that is closed before the last recorded position.
0386b551
AM
2134;;
2135;; The exact position is chosen to try to be close to yet earlier than
2136;; the position where `c-state-cache' will be called next. Right now
2137;; the heuristic is to set it to the position after the last found
2138;; closing paren (of any type) before the line on which
2139;; `c-parse-state' was called. That is chosen primarily to work well
2140;; with refontification of the current line.
0ec1d2c5
AM
2141;;
2142;; 2009-07-28: When `c-state-point-min' and the last position where
2143;; `c-parse-state' or for which `c-invalidate-state-cache' was called, are
2144;; both in the same literal, there is no such "good position", and
2145;; c-state-cache-good-pos is then nil. This is the ONLY circumstance in which
2146;; it can be nil. In this case, `c-state-point-min-literal' will be non-nil.
2147;;
2148;; 2009-06-12: In a brace desert, c-state-cache-good-pos may also be in
2149;; the middle of the desert, as long as it is not within a brace pair
2150;; recorded in `c-state-cache' or a paren/bracket pair.
2151
2152
2153;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2154;; We maintain a simple cache of positions which aren't in a literal, so as to
2155;; speed up testing for non-literality.
9657183b 2156(defconst c-state-nonlit-pos-interval 3000)
0ec1d2c5
AM
2157;; The approximate interval between entries in `c-state-nonlit-pos-cache'.
2158
2159(defvar c-state-nonlit-pos-cache nil)
2160(make-variable-buffer-local 'c-state-nonlit-pos-cache)
e84efb70
AM
2161;; A list of buffer positions which are known not to be in a literal or a cpp
2162;; construct. This is ordered with higher positions at the front of the list.
2163;; Only those which are less than `c-state-nonlit-pos-cache-limit' are valid.
0ec1d2c5
AM
2164
2165(defvar c-state-nonlit-pos-cache-limit 1)
2166(make-variable-buffer-local 'c-state-nonlit-pos-cache-limit)
2167;; An upper limit on valid entries in `c-state-nonlit-pos-cache'. This is
2168;; reduced by buffer changes, and increased by invocations of
2169;; `c-state-literal-at'.
2170
56d093a9
AM
2171(defvar c-state-semi-nonlit-pos-cache nil)
2172(make-variable-buffer-local 'c-state-semi-nonlit-pos-cache)
2173;; A list of buffer positions which are known not to be in a literal. This is
2174;; ordered with higher positions at the front of the list. Only those which
2175;; are less than `c-state-semi-nonlit-pos-cache-limit' are valid.
2176
2177(defvar c-state-semi-nonlit-pos-cache-limit 1)
2178(make-variable-buffer-local 'c-state-semi-nonlit-pos-cache-limit)
2179;; An upper limit on valid entries in `c-state-semi-nonlit-pos-cache'. This is
2180;; reduced by buffer changes, and increased by invocations of
f003f294 2181;; `c-state-literal-at'. FIXME!!!
56d093a9 2182
0ec1d2c5 2183(defsubst c-state-pp-to-literal (from to)
13d49cbb
AM
2184 ;; Do a parse-partial-sexp from FROM to TO, returning either
2185 ;; (STATE TYPE (BEG . END)) if TO is in a literal; or
2186 ;; (STATE) otherwise,
2187 ;; where STATE is the parsing state at TO, TYPE is the type of the literal
2188 ;; (one of 'c, 'c++, 'string) and (BEG . END) is the boundaries of the literal.
2189 ;;
2190 ;; Only elements 3 (in a string), 4 (in a comment), 5 (following a quote),
2191 ;; 7 (comment type) and 8 (start of comment/string) (and possibly 9) of
2192 ;; STATE are valid.
0ec1d2c5 2193 (save-excursion
13d49cbb
AM
2194 (let ((s (parse-partial-sexp from to))
2195 ty)
0ec1d2c5 2196 (when (or (nth 3 s) (nth 4 s)) ; in a string or comment
13d49cbb
AM
2197 (setq ty (cond
2198 ((nth 3 s) 'string)
2199 ((eq (nth 7 s) t) 'c++)
2200 (t 'c)))
0ec1d2c5
AM
2201 (parse-partial-sexp (point) (point-max)
2202 nil ; TARGETDEPTH
2203 nil ; STOPBEFORE
2204 s ; OLDSTATE
13d49cbb
AM
2205 'syntax-table)) ; stop at end of literal
2206 (if ty
2207 `(,s ,ty (,(nth 8 s) . ,(point)))
2208 `(,s)))))
2209
2210(defun c-state-safe-place (here)
2211 ;; Return a buffer position before HERE which is "safe", i.e. outside any
2212 ;; string, comment, or macro.
e84efb70
AM
2213 ;;
2214 ;; NOTE: This function manipulates `c-state-nonlit-pos-cache'. This cache
2215 ;; MAY NOT contain any positions within macros, since macros are frequently
2216 ;; turned into comments by use of the `c-cpp-delimiter' category properties.
2217 ;; We cannot rely on this mechanism whilst determining a cache pos since
2218 ;; this function is also called from outwith `c-parse-state'.
0ec1d2c5
AM
2219 (save-restriction
2220 (widen)
2221 (save-excursion
2222 (let ((c c-state-nonlit-pos-cache)
56d093a9 2223 pos npos high-pos lit macro-beg macro-end)
0ec1d2c5
AM
2224 ;; Trim the cache to take account of buffer changes.
2225 (while (and c (> (car c) c-state-nonlit-pos-cache-limit))
2226 (setq c (cdr c)))
2227 (setq c-state-nonlit-pos-cache c)
2228
2229 (while (and c (> (car c) here))
56d093a9 2230 (setq high-pos (car c))
0ec1d2c5
AM
2231 (setq c (cdr c)))
2232 (setq pos (or (car c) (point-min)))
2233
56d093a9
AM
2234 (unless high-pos
2235 (while
2236 ;; Add an element to `c-state-nonlit-pos-cache' each iteration.
2237 (and
2238 (<= (setq npos (+ pos c-state-nonlit-pos-interval)) here)
2239
2240 ;; Test for being in a literal. If so, go to after it.
2241 (progn
2242 (setq lit (car (cddr (c-state-pp-to-literal pos npos))))
2243 (or (null lit)
2244 (prog1 (<= (cdr lit) here)
2245 (setq npos (cdr lit)))))
2246
2247 ;; Test for being in a macro. If so, go to after it.
2248 (progn
2249 (goto-char npos)
2250 (setq macro-beg
2251 (and (c-beginning-of-macro) (/= (point) npos) (point)))
2252 (when macro-beg
2253 (c-syntactic-end-of-macro)
2254 (or (eobp) (forward-char))
2255 (setq macro-end (point)))
2256 (or (null macro-beg)
2257 (prog1 (<= macro-end here)
2258 (setq npos macro-end)))))
2259
2260 (setq pos npos)
2261 (setq c-state-nonlit-pos-cache (cons pos c-state-nonlit-pos-cache)))
2262 ;; Add one extra element above HERE so as to to avoid the previous
2263 ;; expensive calculation when the next call is close to the current
2264 ;; one. This is especially useful when inside a large macro.
2265 (setq c-state-nonlit-pos-cache (cons npos c-state-nonlit-pos-cache)))
0ec1d2c5
AM
2266
2267 (if (> pos c-state-nonlit-pos-cache-limit)
2268 (setq c-state-nonlit-pos-cache-limit pos))
13d49cbb 2269 pos))))
9858f6c3 2270
56d093a9
AM
2271(defun c-state-semi-safe-place (here)
2272 ;; Return a buffer position before HERE which is "safe", i.e. outside any
2273 ;; string or comment. It may be in a macro.
2274 (save-restriction
2275 (widen)
2276 (save-excursion
2277 (let ((c c-state-semi-nonlit-pos-cache)
2278 pos npos high-pos lit macro-beg macro-end)
2279 ;; Trim the cache to take account of buffer changes.
2280 (while (and c (> (car c) c-state-semi-nonlit-pos-cache-limit))
2281 (setq c (cdr c)))
2282 (setq c-state-semi-nonlit-pos-cache c)
2283
2284 (while (and c (> (car c) here))
2285 (setq high-pos (car c))
2286 (setq c (cdr c)))
2287 (setq pos (or (car c) (point-min)))
2288
2289 (unless high-pos
2290 (while
2291 ;; Add an element to `c-state-semi-nonlit-pos-cache' each iteration.
2292 (and
2293 (<= (setq npos (+ pos c-state-nonlit-pos-interval)) here)
2294
2295 ;; Test for being in a literal. If so, go to after it.
2296 (progn
2297 (setq lit (car (cddr (c-state-pp-to-literal pos npos))))
2298 (or (null lit)
2299 (prog1 (<= (cdr lit) here)
2300 (setq npos (cdr lit))))))
2301
2302 (setq pos npos)
2303 (setq c-state-semi-nonlit-pos-cache
2304 (cons pos c-state-semi-nonlit-pos-cache))))
2305
2306 (if (> pos c-state-semi-nonlit-pos-cache-limit)
2307 (setq c-state-semi-nonlit-pos-cache-limit pos))
2308 pos))))
2309
13d49cbb
AM
2310(defun c-state-literal-at (here)
2311 ;; If position HERE is inside a literal, return (START . END), the
2312 ;; boundaries of the literal (which may be outside the accessible bit of the
2313 ;; buffer). Otherwise, return nil.
2314 ;;
2315 ;; This function is almost the same as `c-literal-limits'. Previously, it
9858f6c3 2316 ;; differed in that it was a lower level function, and that it rigorously
13d49cbb
AM
2317 ;; followed the syntax from BOB. `c-literal-limits' is now (2011-12)
2318 ;; virtually identical to this function.
2319 (save-restriction
2320 (widen)
2321 (save-excursion
2322 (let ((pos (c-state-safe-place here)))
2323 (car (cddr (c-state-pp-to-literal pos here)))))))
0ec1d2c5
AM
2324
2325(defsubst c-state-lit-beg (pos)
2326 ;; Return the start of the literal containing POS, or POS itself.
2327 (or (car (c-state-literal-at pos))
2328 pos))
2329
2330(defsubst c-state-cache-non-literal-place (pos state)
9c4aeabf 2331 ;; Return a position outside of a string/comment/macro at or before POS.
0ec1d2c5 2332 ;; STATE is the parse-partial-sexp state at POS.
9c4aeabf
AM
2333 (let ((res (if (or (nth 3 state) ; in a string?
2334 (nth 4 state)) ; in a comment?
2335 (nth 8 state)
2336 pos)))
2337 (save-excursion
2338 (goto-char res)
2339 (if (c-beginning-of-macro)
2340 (point)
2341 res))))
0ec1d2c5
AM
2342
2343;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2344;; Stuff to do with point-min, and coping with any literal there.
2345(defvar c-state-point-min 1)
2346(make-variable-buffer-local 'c-state-point-min)
2347;; This is (point-min) when `c-state-cache' was last calculated. A change of
2348;; narrowing is likely to affect the parens that are visible before the point.
2349
2350(defvar c-state-point-min-lit-type nil)
2351(make-variable-buffer-local 'c-state-point-min-lit-type)
2352(defvar c-state-point-min-lit-start nil)
2353(make-variable-buffer-local 'c-state-point-min-lit-start)
2354;; These two variables define the literal, if any, containing point-min.
2355;; Their values are, respectively, 'string, c, or c++, and the start of the
2356;; literal. If there's no literal there, they're both nil.
2357
2358(defvar c-state-min-scan-pos 1)
2359(make-variable-buffer-local 'c-state-min-scan-pos)
2360;; This is the earliest buffer-pos from which scanning can be done. It is
2361;; either the end of the literal containing point-min, or point-min itself.
2362;; It becomes nil if the buffer is changed earlier than this point.
2363(defun c-state-get-min-scan-pos ()
2364 ;; Return the lowest valid scanning pos. This will be the end of the
2365 ;; literal enclosing point-min, or point-min itself.
2366 (or c-state-min-scan-pos
2367 (save-restriction
2368 (save-excursion
2369 (widen)
2370 (goto-char c-state-point-min-lit-start)
2371 (if (eq c-state-point-min-lit-type 'string)
2372 (forward-sexp)
2373 (forward-comment 1))
2374 (setq c-state-min-scan-pos (point))))))
2375
2376(defun c-state-mark-point-min-literal ()
2377 ;; Determine the properties of any literal containing POINT-MIN, setting the
2378 ;; variables `c-state-point-min-lit-type', `c-state-point-min-lit-start',
2379 ;; and `c-state-min-scan-pos' accordingly. The return value is meaningless.
2380 (let ((p-min (point-min))
2381 lit)
2382 (save-restriction
2383 (widen)
2384 (setq lit (c-state-literal-at p-min))
2385 (if lit
2386 (setq c-state-point-min-lit-type
2387 (save-excursion
2388 (goto-char (car lit))
2389 (cond
2390 ((looking-at c-block-comment-start-regexp) 'c)
2391 ((looking-at c-line-comment-starter) 'c++)
2392 (t 'string)))
2393 c-state-point-min-lit-start (car lit)
2394 c-state-min-scan-pos (cdr lit))
2395 (setq c-state-point-min-lit-type nil
2396 c-state-point-min-lit-start nil
2397 c-state-min-scan-pos p-min)))))
2398
2399
2400;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2401;; A variable which signals a brace dessert - helpful for reducing the number
2402;; of fruitless backward scans.
2403(defvar c-state-brace-pair-desert nil)
2404(make-variable-buffer-local 'c-state-brace-pair-desert)
e84efb70 2405;; Used only in `c-append-lower-brace-pair-to-state-cache'. It is set when
0ec1d2c5
AM
2406;; that defun has searched backwards for a brace pair and not found one. Its
2407;; value is either nil or a cons (PA . FROM), where PA is the position of the
2408;; enclosing opening paren/brace/bracket which bounds the backwards search (or
2409;; nil when at top level) and FROM is where the backward search started. It
2410;; is reset to nil in `c-invalidate-state-cache'.
2411
2412
2413;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2414;; Lowish level functions/macros which work directly on `c-state-cache', or a
2415;; list of like structure.
2416(defmacro c-state-cache-top-lparen (&optional cache)
2417 ;; Return the address of the top left brace/bracket/paren recorded in CACHE
2418 ;; (default `c-state-cache') (or nil).
2419 (let ((cash (or cache 'c-state-cache)))
2420 `(if (consp (car ,cash))
2421 (caar ,cash)
2422 (car ,cash))))
2423
2424(defmacro c-state-cache-top-paren (&optional cache)
2425 ;; Return the address of the latest brace/bracket/paren (whether left or
2426 ;; right) recorded in CACHE (default `c-state-cache') or nil.
2427 (let ((cash (or cache 'c-state-cache)))
2428 `(if (consp (car ,cash))
2429 (cdar ,cash)
2430 (car ,cash))))
2431
2432(defmacro c-state-cache-after-top-paren (&optional cache)
2433 ;; Return the position just after the latest brace/bracket/paren (whether
2434 ;; left or right) recorded in CACHE (default `c-state-cache') or nil.
2435 (let ((cash (or cache 'c-state-cache)))
2436 `(if (consp (car ,cash))
2437 (cdar ,cash)
2438 (and (car ,cash)
2439 (1+ (car ,cash))))))
2440
2441(defun c-get-cache-scan-pos (here)
2442 ;; From the state-cache, determine the buffer position from which we might
2443 ;; scan forward to HERE to update this cache. This position will be just
2444 ;; after a paren/brace/bracket recorded in the cache, if possible, otherwise
2445 ;; return the earliest position in the accessible region which isn't within
2446 ;; a literal. If the visible portion of the buffer is entirely within a
2447 ;; literal, return NIL.
2448 (let ((c c-state-cache) elt)
2449 ;(while (>= (or (c-state-cache-top-lparen c) 1) here)
2450 (while (and c
2451 (>= (c-state-cache-top-lparen c) here))
2452 (setq c (cdr c)))
2453
2454 (setq elt (car c))
2455 (cond
2456 ((consp elt)
2457 (if (> (cdr elt) here)
2458 (1+ (car elt))
2459 (cdr elt)))
2460 (elt (1+ elt))
2461 ((<= (c-state-get-min-scan-pos) here)
2462 (c-state-get-min-scan-pos))
2463 (t nil))))
2464
2465;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2466;; Variables which keep track of preprocessor constructs.
2467(defvar c-state-old-cpp-beg nil)
2468(make-variable-buffer-local 'c-state-old-cpp-beg)
2469(defvar c-state-old-cpp-end nil)
2470(make-variable-buffer-local 'c-state-old-cpp-end)
2471;; These are the limits of the macro containing point at the previous call of
2472;; `c-parse-state', or nil.
2473
2474;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
c7015153 2475;; Defuns which analyze the buffer, yet don't change `c-state-cache'.
0ec1d2c5
AM
2476(defun c-get-fallback-scan-pos (here)
2477 ;; Return a start position for building `c-state-cache' from
2478 ;; scratch. This will be at the top level, 2 defuns back.
0386b551
AM
2479 (save-excursion
2480 ;; Go back 2 bods, but ignore any bogus positions returned by
2481 ;; beginning-of-defun (i.e. open paren in column zero).
2482 (goto-char here)
2483 (let ((cnt 2))
2484 (while (not (or (bobp) (zerop cnt)))
0ec1d2c5 2485 (c-beginning-of-defun-1) ; Pure elisp BOD.
0386b551
AM
2486 (if (eq (char-after) ?\{)
2487 (setq cnt (1- cnt)))))
2488 (point)))
2489
c0209c2c
AM
2490(defun c-state-balance-parens-backwards (here- here+ top)
2491 ;; Return the position of the opening paren/brace/bracket before HERE- which
2492 ;; matches the outermost close p/b/b between HERE+ and TOP. Except when
2493 ;; there's a macro, HERE- and HERE+ are the same. Like this:
0ec1d2c5 2494 ;;
c0209c2c
AM
2495 ;; ............................................
2496 ;; | |
2497 ;; ( [ ( .........#macro.. ) ( ) ] )
2498 ;; ^ ^ ^ ^
2499 ;; | | | |
2500 ;; return HERE- HERE+ TOP
0ec1d2c5
AM
2501 ;;
2502 ;; If there aren't enough opening paren/brace/brackets, return the position
c0209c2c 2503 ;; of the outermost one found, or HERE- if there are none. If there are no
da6062e6 2504 ;; closing p/b/bs between HERE+ and TOP, return HERE-. HERE-/+ and TOP
c0209c2c
AM
2505 ;; must not be inside literals. Only the accessible portion of the buffer
2506 ;; will be scanned.
2507
2508 ;; PART 1: scan from `here+' up to `top', accumulating ")"s which enclose
2509 ;; `here'. Go round the next loop each time we pass over such a ")". These
2510 ;; probably match "("s before `here-'.
0ec1d2c5
AM
2511 (let (pos pa ren+1 lonely-rens)
2512 (save-excursion
2513 (save-restriction
2514 (narrow-to-region (point-min) top) ; This can move point, sometimes.
c0209c2c 2515 (setq pos here+)
0ec1d2c5
AM
2516 (c-safe
2517 (while
2518 (setq ren+1 (scan-lists pos 1 1)) ; might signal
2519 (setq lonely-rens (cons ren+1 lonely-rens)
2520 pos ren+1)))))
2521
c0209c2c 2522 ;; PART 2: Scan back before `here-' searching for the "("s
0ec1d2c5
AM
2523 ;; matching/mismatching the ")"s found above. We only need to direct the
2524 ;; caller to scan when we've encountered unmatched right parens.
c0209c2c
AM
2525 (setq pos here-)
2526 (when lonely-rens
2527 (c-safe
2528 (while
2529 (and lonely-rens ; actual values aren't used.
2530 (setq pa (scan-lists pos -1 1)))
2531 (setq pos pa)
2532 (setq lonely-rens (cdr lonely-rens)))))
2533 pos))
0ec1d2c5
AM
2534
2535(defun c-parse-state-get-strategy (here good-pos)
2536 ;; Determine the scanning strategy for adjusting `c-parse-state', attempting
53964682 2537 ;; to minimize the amount of scanning. HERE is the pertinent position in
0ec1d2c5
AM
2538 ;; the buffer, GOOD-POS is a position where `c-state-cache' (possibly with
2539 ;; its head trimmed) is known to be good, or nil if there is no such
2540 ;; position.
2541 ;;
2542 ;; The return value is a list, one of the following:
2543 ;;
2544 ;; o - ('forward CACHE-POS START-POINT) - scan forward from START-POINT,
2545 ;; which is not less than CACHE-POS.
2546 ;; o - ('backward CACHE-POS nil) - scan backwards (from HERE).
2547 ;; o - ('BOD nil START-POINT) - scan forwards from START-POINT, which is at the
2548 ;; top level.
2549 ;; o - ('IN-LIT nil nil) - point is inside the literal containing point-min.
2550 ;; , where CACHE-POS is the highest position recorded in `c-state-cache' at
2551 ;; or below HERE.
2552 (let ((cache-pos (c-get-cache-scan-pos here)) ; highest position below HERE in cache (or 1)
2553 BOD-pos ; position of 2nd BOD before HERE.
2554 strategy ; 'forward, 'backward, 'BOD, or 'IN-LIT.
2555 start-point
2556 how-far) ; putative scanning distance.
2557 (setq good-pos (or good-pos (c-state-get-min-scan-pos)))
2558 (cond
2559 ((< here (c-state-get-min-scan-pos))
2560 (setq strategy 'IN-LIT
2561 start-point nil
2562 cache-pos nil
2563 how-far 0))
2564 ((<= good-pos here)
2565 (setq strategy 'forward
2566 start-point (max good-pos cache-pos)
2567 how-far (- here start-point)))
2568 ((< (- good-pos here) (- here cache-pos)) ; FIXME!!! ; apply some sort of weighting.
2569 (setq strategy 'backward
2570 how-far (- good-pos here)))
2571 (t
2572 (setq strategy 'forward
2573 how-far (- here cache-pos)
2574 start-point cache-pos)))
2575
2576 ;; Might we be better off starting from the top level, two defuns back,
b03aabda
AM
2577 ;; instead? This heuristic no longer works well in C++, where
2578 ;; declarations inside namespace brace blocks are frequently placed at
2579 ;; column zero.
2580 (when (and (not (c-major-mode-is 'c++-mode))
2581 (> how-far c-state-cache-too-far))
0ec1d2c5
AM
2582 (setq BOD-pos (c-get-fallback-scan-pos here)) ; somewhat EXPENSIVE!!!
2583 (if (< (- here BOD-pos) how-far)
2584 (setq strategy 'BOD
2585 start-point BOD-pos)))
2586
2587 (list
2588 strategy
2589 (and (memq strategy '(forward backward)) cache-pos)
2590 (and (memq strategy '(forward BOD)) start-point))))
2591
2592
2593;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2594;; Routines which change `c-state-cache' and associated values.
2595(defun c-renarrow-state-cache ()
2596 ;; The region (more precisely, point-min) has changed since we
2597 ;; calculated `c-state-cache'. Amend `c-state-cache' accordingly.
2598 (if (< (point-min) c-state-point-min)
2599 ;; If point-min has MOVED BACKWARDS then we drop the state completely.
2600 ;; It would be possible to do a better job here and recalculate the top
2601 ;; only.
2602 (progn
2603 (c-state-mark-point-min-literal)
2604 (setq c-state-cache nil
2605 c-state-cache-good-pos c-state-min-scan-pos
2606 c-state-brace-pair-desert nil))
2607
2608 ;; point-min has MOVED FORWARD.
2609
2610 ;; Is the new point-min inside a (different) literal?
2611 (unless (and c-state-point-min-lit-start ; at prev. point-min
2612 (< (point-min) (c-state-get-min-scan-pos)))
2613 (c-state-mark-point-min-literal))
2614
2615 ;; Cut off a bit of the tail from `c-state-cache'.
2616 (let ((ptr (cons nil c-state-cache))
2617 pa)
2618 (while (and (setq pa (c-state-cache-top-lparen (cdr ptr)))
2619 (>= pa (point-min)))
2620 (setq ptr (cdr ptr)))
2621
2622 (when (consp ptr)
2623 (if (eq (cdr ptr) c-state-cache)
2624 (setq c-state-cache nil
2625 c-state-cache-good-pos c-state-min-scan-pos)
2626 (setcdr ptr nil)
2627 (setq c-state-cache-good-pos (1+ (c-state-cache-top-lparen))))
2628 )))
2629
2630 (setq c-state-point-min (point-min)))
2631
2632(defun c-append-lower-brace-pair-to-state-cache (from &optional upper-lim)
583e23bd
AM
2633 ;; If there is a brace pair preceding FROM in the buffer, at the same level
2634 ;; of nesting (not necessarily immediately preceding), push a cons onto
2635 ;; `c-state-cache' to represent it. FROM must not be inside a literal. If
2636 ;; UPPER-LIM is non-nil, we append the highest brace pair whose "}" is below
2637 ;; UPPER-LIM.
0ec1d2c5
AM
2638 ;;
2639 ;; Return non-nil when this has been done.
2640 ;;
583e23bd
AM
2641 ;; The situation it copes with is this transformation:
2642 ;;
2643 ;; OLD: { (.) {...........}
2644 ;; ^ ^
2645 ;; FROM HERE
2646 ;;
2647 ;; NEW: { {....} (.) {.........
2648 ;; ^ ^ ^
2649 ;; LOWER BRACE PAIR HERE or HERE
2650 ;;
0ec1d2c5
AM
2651 ;; This routine should be fast. Since it can get called a LOT, we maintain
2652 ;; `c-state-brace-pair-desert', a small cache of "failures", such that we
2653 ;; reduce the time wasted in repeated fruitless searches in brace deserts.
2654 (save-excursion
2655 (save-restriction
2656 (let ((bra from) ce ; Positions of "{" and "}".
2657 new-cons
2658 (cache-pos (c-state-cache-top-lparen)) ; might be nil.
2659 (macro-start-or-from
2660 (progn (goto-char from)
2661 (c-beginning-of-macro)
2662 (point))))
2663 (or upper-lim (setq upper-lim from))
2664
2665 ;; If we're essentially repeating a fruitless search, just give up.
2666 (unless (and c-state-brace-pair-desert
2667 (eq cache-pos (car c-state-brace-pair-desert))
b03aabda
AM
2668 (or (null (car c-state-brace-pair-desert))
2669 (> from (car c-state-brace-pair-desert)))
0ec1d2c5 2670 (<= from (cdr c-state-brace-pair-desert)))
b03aabda 2671 ;; DESERT-LIM. Avoid repeated searching through the cached desert.
583e23bd
AM
2672 (let ((desert-lim
2673 (and c-state-brace-pair-desert
2674 (eq cache-pos (car c-state-brace-pair-desert))
b03aabda 2675 (>= from (cdr c-state-brace-pair-desert))
583e23bd
AM
2676 (cdr c-state-brace-pair-desert)))
2677 ;; CACHE-LIM. This limit will be necessary when an opening
2678 ;; paren at `cache-pos' has just had its matching close paren
b03aabda
AM
2679 ;; inserted into the buffer. `cache-pos' continues to be a
2680 ;; search bound, even though the algorithm below would skip
2681 ;; over the new paren pair.
583e23bd
AM
2682 (cache-lim (and cache-pos (< cache-pos from) cache-pos)))
2683 (narrow-to-region
2684 (cond
2685 ((and desert-lim cache-lim)
2686 (max desert-lim cache-lim))
2687 (desert-lim)
2688 (cache-lim)
2689 ((point-min)))
2690 (point-max)))
0ec1d2c5
AM
2691
2692 ;; In the next pair of nested loops, the inner one moves back past a
2693 ;; pair of (mis-)matching parens or brackets; the outer one moves
2694 ;; back over a sequence of unmatched close brace/paren/bracket each
2695 ;; time round.
2696 (while
2697 (progn
2698 (c-safe
2699 (while
2700 (and (setq ce (scan-lists bra -1 -1)) ; back past )/]/}; might signal
2701 (setq bra (scan-lists ce -1 1)) ; back past (/[/{; might signal
2702 (or (> ce upper-lim)
2703 (not (eq (char-after bra) ?\{))
2704 (and (goto-char bra)
2705 (c-beginning-of-macro)
2706 (< (point) macro-start-or-from))))))
2707 (and ce (< ce bra)))
2708 (setq bra ce)) ; If we just backed over an unbalanced closing
2709 ; brace, ignore it.
2710
2711 (if (and ce (< bra ce) (eq (char-after bra) ?\{))
2712 ;; We've found the desired brace-pair.
2713 (progn
2714 (setq new-cons (cons bra (1+ ce)))
2715 (cond
2716 ((consp (car c-state-cache))
2717 (setcar c-state-cache new-cons))
2718 ((and (numberp (car c-state-cache)) ; probably never happens
2719 (< ce (car c-state-cache)))
2720 (setcdr c-state-cache
2721 (cons new-cons (cdr c-state-cache))))
2722 (t (setq c-state-cache (cons new-cons c-state-cache)))))
2723
583e23bd 2724 ;; We haven't found a brace pair. Record this in the cache.
0ec1d2c5
AM
2725 (setq c-state-brace-pair-desert (cons cache-pos from))))))))
2726
2727(defsubst c-state-push-any-brace-pair (bra+1 macro-start-or-here)
2728 ;; If BRA+1 is nil, do nothing. Otherwise, BRA+1 is the buffer position
2729 ;; following a {, and that brace has a (mis-)matching } (or ]), and we
2730 ;; "push" "a" brace pair onto `c-state-cache'.
2731 ;;
2732 ;; Here "push" means overwrite the top element if it's itself a brace-pair,
2733 ;; otherwise push it normally.
2734 ;;
2735 ;; The brace pair we push is normally the one surrounding BRA+1, but if the
2736 ;; latter is inside a macro, not being a macro containing
2737 ;; MACRO-START-OR-HERE, we scan backwards through the buffer for a non-macro
2738 ;; base pair. This latter case is assumed to be rare.
2739 ;;
2740 ;; Note: POINT is not preserved in this routine.
2741 (if bra+1
2742 (if (or (> bra+1 macro-start-or-here)
2743 (progn (goto-char bra+1)
2744 (not (c-beginning-of-macro))))
2745 (setq c-state-cache
2746 (cons (cons (1- bra+1)
2747 (scan-lists bra+1 1 1))
2748 (if (consp (car c-state-cache))
2749 (cdr c-state-cache)
2750 c-state-cache)))
2751 ;; N.B. This defsubst codes one method for the simple, normal case,
2752 ;; and a more sophisticated, slower way for the general case. Don't
e1dbe924 2753 ;; eliminate this defsubst - it's a speed optimization.
0ec1d2c5
AM
2754 (c-append-lower-brace-pair-to-state-cache (1- bra+1)))))
2755
2756(defun c-append-to-state-cache (from)
2757 ;; Scan the buffer from FROM to (point-max), adding elements into
2758 ;; `c-state-cache' for braces etc. Return a candidate for
2759 ;; `c-state-cache-good-pos'.
2760 ;;
2761 ;; FROM must be after the latest brace/paren/bracket in `c-state-cache', if
2762 ;; any. Typically, it is immediately after it. It must not be inside a
2763 ;; literal.
2764 (let ((here-bol (c-point 'bol (point-max)))
2765 (macro-start-or-here
2766 (save-excursion (goto-char (point-max))
2767 (if (c-beginning-of-macro)
2768 (point)
2769 (point-max))))
2770 pa+1 ; pos just after an opening PAren (or brace).
2771 (ren+1 from) ; usually a pos just after an closing paREN etc.
2772 ; Is actually the pos. to scan for a (/{/[ from,
2773 ; which sometimes is after a silly )/}/].
2774 paren+1 ; Pos after some opening or closing paren.
2775 paren+1s ; A list of `paren+1's; used to determine a
2776 ; good-pos.
2777 bra+1 ce+1 ; just after L/R bra-ces.
2778 bra+1s ; list of OLD values of bra+1.
2779 mstart) ; start of a macro.
2780
2781 (save-excursion
22bcf204 2782 ;; Each time round the following loop, we enter a successively deeper
0ec1d2c5
AM
2783 ;; level of brace/paren nesting. (Except sometimes we "continue at
2784 ;; the existing level".) `pa+1' is a pos inside an opening
2785 ;; brace/paren/bracket, usually just after it.
2786 (while
2787 (progn
2788 ;; Each time round the next loop moves forward over an opening then
2789 ;; a closing brace/bracket/paren. This loop is white hot, so it
2790 ;; plays ugly tricks to go fast. DON'T PUT ANYTHING INTO THIS
2791 ;; LOOP WHICH ISN'T ABSOLUTELY NECESSARY!!! It terminates when a
2792 ;; call of `scan-lists' signals an error, which happens when there
2793 ;; are no more b/b/p's to scan.
2794 (c-safe
2795 (while t
2796 (setq pa+1 (scan-lists ren+1 1 -1) ; Into (/{/[; might signal
2797 paren+1s (cons pa+1 paren+1s))
2798 (setq ren+1 (scan-lists pa+1 1 1)) ; Out of )/}/]; might signal
2799 (if (and (eq (char-before pa+1) ?{)) ; Check for a macro later.
2800 (setq bra+1 pa+1))
2801 (setcar paren+1s ren+1)))
2802
2803 (if (and pa+1 (> pa+1 ren+1))
2804 ;; We've just entered a deeper nesting level.
2805 (progn
2806 ;; Insert the brace pair (if present) and the single open
2807 ;; paren/brace/bracket into `c-state-cache' It cannot be
2808 ;; inside a macro, except one around point, because of what
2809 ;; `c-neutralize-syntax-in-CPP' has done.
2810 (c-state-push-any-brace-pair bra+1 macro-start-or-here)
2811 ;; Insert the opening brace/bracket/paren position.
2812 (setq c-state-cache (cons (1- pa+1) c-state-cache))
2813 ;; Clear admin stuff for the next more nested part of the scan.
2814 (setq ren+1 pa+1 pa+1 nil bra+1 nil bra+1s nil)
2815 t) ; Carry on the loop
2816
2817 ;; All open p/b/b's at this nesting level, if any, have probably
2818 ;; been closed by matching/mismatching ones. We're probably
2819 ;; finished - we just need to check for having found an
2820 ;; unmatched )/}/], which we ignore. Such a )/}/] can't be in a
2821 ;; macro, due the action of `c-neutralize-syntax-in-CPP'.
2822 (c-safe (setq ren+1 (scan-lists ren+1 1 1)))))) ; acts as loop control.
2823
2824 ;; Record the final, innermost, brace-pair if there is one.
2825 (c-state-push-any-brace-pair bra+1 macro-start-or-here)
2826
2827 ;; Determine a good pos
2828 (while (and (setq paren+1 (car paren+1s))
2829 (> (if (> paren+1 macro-start-or-here)
2830 paren+1
2831 (goto-char paren+1)
2832 (setq mstart (and (c-beginning-of-macro)
2833 (point)))
2834 (or mstart paren+1))
2835 here-bol))
2836 (setq paren+1s (cdr paren+1s)))
2837 (cond
2838 ((and paren+1 mstart)
2839 (min paren+1 mstart))
2840 (paren+1)
2841 (t from)))))
2842
2843(defun c-remove-stale-state-cache (good-pos pps-point)
2844 ;; Remove stale entries from the `c-cache-state', i.e. those which will
2845 ;; not be in it when it is amended for position (point-max).
2846 ;; Additionally, the "outermost" open-brace entry before (point-max)
2847 ;; will be converted to a cons if the matching close-brace is scanned.
2848 ;;
2849 ;; GOOD-POS is a "maximal" "safe position" - there must be no open
2850 ;; parens/braces/brackets between GOOD-POS and (point-max).
2851 ;;
2852 ;; As a second thing, calculate the result of parse-partial-sexp at
2853 ;; PPS-POINT, w.r.t. GOOD-POS. The motivation here is that
2854 ;; `c-state-cache-good-pos' may become PPS-POINT, but the caller may need to
2855 ;; adjust it to get outside a string/comment. (Sorry about this! The code
2856 ;; needs to be FAST).
2857 ;;
2858 ;; Return a list (GOOD-POS SCAN-BACK-POS PPS-STATE), where
2859 ;; o - GOOD-POS is a position where the new value `c-state-cache' is known
2860 ;; to be good (we aim for this to be as high as possible);
2861 ;; o - SCAN-BACK-POS, if not nil, indicates there may be a brace pair
2862 ;; preceding POS which needs to be recorded in `c-state-cache'. It is a
2863 ;; position to scan backwards from.
2864 ;; o - PPS-STATE is the parse-partial-sexp state at PPS-POINT.
2865 (save-restriction
2866 (narrow-to-region 1 (point-max))
2867 (save-excursion
eb2f6eeb 2868 (let* ((in-macro-start ; start of macro containing (point-max) or nil.
0ec1d2c5
AM
2869 (save-excursion
2870 (goto-char (point-max))
2871 (and (c-beginning-of-macro)
2872 (point))))
2873 (good-pos-actual-macro-start ; Start of macro containing good-pos
2874 ; or nil
2875 (and (< good-pos (point-max))
2876 (save-excursion
2877 (goto-char good-pos)
2878 (and (c-beginning-of-macro)
2879 (point)))))
2880 (good-pos-actual-macro-end ; End of this macro, (maybe
2881 ; (point-max)), or nil.
2882 (and good-pos-actual-macro-start
2883 (save-excursion
2884 (goto-char good-pos-actual-macro-start)
2885 (c-end-of-macro)
2886 (point))))
2887 pps-state ; Will be 9 or 10 elements long.
2888 pos
2889 upper-lim ; ,beyond which `c-state-cache' entries are removed
2890 scan-back-pos
2891 pair-beg pps-point-state target-depth)
2892
2893 ;; Remove entries beyond (point-max). Also remove any entries inside
2894 ;; a macro, unless (point-max) is in the same macro.
2895 (setq upper-lim
2896 (if (or (null c-state-old-cpp-beg)
2897 (and (> (point-max) c-state-old-cpp-beg)
2898 (< (point-max) c-state-old-cpp-end)))
2899 (point-max)
2900 (min (point-max) c-state-old-cpp-beg)))
657071fc 2901 (while (and c-state-cache (>= (c-state-cache-top-lparen) upper-lim))
0ec1d2c5
AM
2902 (setq c-state-cache (cdr c-state-cache)))
2903 ;; If `upper-lim' is inside the last recorded brace pair, remove its
2904 ;; RBrace and indicate we'll need to search backwards for a previous
2905 ;; brace pair.
2906 (when (and c-state-cache
2907 (consp (car c-state-cache))
2908 (> (cdar c-state-cache) upper-lim))
2909 (setcar c-state-cache (caar c-state-cache))
2910 (setq scan-back-pos (car c-state-cache)))
2911
2912 ;; The next loop jumps forward out of a nested level of parens each
2913 ;; time round; the corresponding elements in `c-state-cache' are
2914 ;; removed. `pos' is just after the brace-pair or the open paren at
2915 ;; (car c-state-cache). There can be no open parens/braces/brackets
2916 ;; between `good-pos'/`good-pos-actual-macro-start' and (point-max),
2917 ;; due to the interface spec to this function.
dd21b621 2918 (setq pos (if (and good-pos-actual-macro-end
eb2f6eeb
AM
2919 (not (eq good-pos-actual-macro-start
2920 in-macro-start)))
0ec1d2c5
AM
2921 (1+ good-pos-actual-macro-end) ; get outside the macro as
2922 ; marked by a `category' text property.
2923 good-pos))
2924 (goto-char pos)
2925 (while (and c-state-cache
2926 (< (point) (point-max)))
2927 (cond
2928 ((null pps-state) ; first time through
2929 (setq target-depth -1))
2930 ((eq (car pps-state) target-depth) ; found closing ),},]
2931 (setq target-depth (1- (car pps-state))))
2932 ;; Do nothing when we've merely reached pps-point.
2933 )
2934
2935 ;; Scan!
2936 (setq pps-state
2937 (parse-partial-sexp
2938 (point) (if (< (point) pps-point) pps-point (point-max))
2939 target-depth
2940 nil pps-state))
2941
2942 (if (= (point) pps-point)
2943 (setq pps-point-state pps-state))
2944
2945 (when (eq (car pps-state) target-depth)
2946 (setq pos (point)) ; POS is now just after an R-paren/brace.
2947 (cond
2948 ((and (consp (car c-state-cache))
2949 (eq (point) (cdar c-state-cache)))
2950 ;; We've just moved out of the paren pair containing the brace-pair
2951 ;; at (car c-state-cache). `pair-beg' is where the open paren is,
2952 ;; and is potentially where the open brace of a cons in
2953 ;; c-state-cache will be.
2954 (setq pair-beg (car-safe (cdr c-state-cache))
2955 c-state-cache (cdr-safe (cdr c-state-cache)))) ; remove {}pair + containing Lparen.
2956 ((numberp (car c-state-cache))
2957 (setq pair-beg (car c-state-cache)
2958 c-state-cache (cdr c-state-cache))) ; remove this
2959 ; containing Lparen
2960 ((numberp (cadr c-state-cache))
2961 (setq pair-beg (cadr c-state-cache)
2962 c-state-cache (cddr c-state-cache))) ; Remove a paren pair
2963 ; together with enclosed brace pair.
2964 ;; (t nil) ; Ignore an unmated Rparen.
2965 )))
2966
2967 (if (< (point) pps-point)
2968 (setq pps-state (parse-partial-sexp (point) pps-point
2969 nil nil ; TARGETDEPTH, STOPBEFORE
2970 pps-state)))
2971
2972 ;; If the last paren pair we moved out of was actually a brace pair,
2973 ;; insert it into `c-state-cache'.
2974 (when (and pair-beg (eq (char-after pair-beg) ?{))
2975 (if (consp (car-safe c-state-cache))
2976 (setq c-state-cache (cdr c-state-cache)))
2977 (setq c-state-cache (cons (cons pair-beg pos)
2978 c-state-cache)))
2979
2980 (list pos scan-back-pos pps-state)))))
2981
2982(defun c-remove-stale-state-cache-backwards (here cache-pos)
2983 ;; Strip stale elements of `c-state-cache' by moving backwards through the
2984 ;; buffer, and inform the caller of the scenario detected.
2985 ;;
2986 ;; HERE is the position we're setting `c-state-cache' for.
2987 ;; CACHE-POS is just after the latest recorded position in `c-state-cache'
2988 ;; before HERE, or a position at or near point-min which isn't in a
2989 ;; literal.
2990 ;;
2991 ;; This function must only be called only when (> `c-state-cache-good-pos'
2992 ;; HERE). Usually the gap between CACHE-POS and HERE is large. It is thus
e1dbe924 2993 ;; optimized to eliminate (or minimize) scanning between these two
0ec1d2c5
AM
2994 ;; positions.
2995 ;;
2996 ;; Return a three element list (GOOD-POS SCAN-BACK-POS FWD-FLAG), where:
2997 ;; o - GOOD-POS is a "good position", where `c-state-cache' is valid, or
2998 ;; could become so after missing elements are inserted into
2999 ;; `c-state-cache'. This is JUST AFTER an opening or closing
3000 ;; brace/paren/bracket which is already in `c-state-cache' or just before
3001 ;; one otherwise. exceptionally (when there's no such b/p/b handy) the BOL
3002 ;; before `here''s line, or the start of the literal containing it.
3003 ;; o - SCAN-BACK-POS, if non-nil, indicates there may be a brace pair
3004 ;; preceding POS which isn't recorded in `c-state-cache'. It is a position
3005 ;; to scan backwards from.
3006 ;; o - FWD-FLAG, if non-nil, indicates there may be parens/braces between
3007 ;; POS and HERE which aren't recorded in `c-state-cache'.
3008 ;;
3009 ;; The comments in this defun use "paren" to mean parenthesis or square
3010 ;; bracket (as contrasted with a brace), and "(" and ")" likewise.
3011 ;;
3012 ;; . {..} (..) (..) ( .. { } ) (...) ( .... . ..)
3013 ;; | | | | | |
3014 ;; CP E here D C good
3015 (let ((pos c-state-cache-good-pos)
3016 pa ren ; positions of "(" and ")"
3017 dropped-cons ; whether the last element dropped from `c-state-cache'
3018 ; was a cons (representing a brace-pair)
3019 good-pos ; see above.
3020 lit ; (START . END) of a literal containing some point.
3021 here-lit-start here-lit-end ; bounds of literal containing `here'
3022 ; or `here' itself.
c0209c2c 3023 here- here+ ; start/end of macro around HERE, or HERE
0ec1d2c5 3024 (here-bol (c-point 'bol here))
c400c4d7 3025 (too-far-back (max (- here c-state-cache-too-far) (point-min))))
0ec1d2c5
AM
3026
3027 ;; Remove completely irrelevant entries from `c-state-cache'.
3028 (while (and c-state-cache
3029 (>= (setq pa (c-state-cache-top-lparen)) here))
3030 (setq dropped-cons (consp (car c-state-cache)))
3031 (setq c-state-cache (cdr c-state-cache))
3032 (setq pos pa))
3033 ;; At this stage, (> pos here);
3034 ;; (< (c-state-cache-top-lparen) here) (or is nil).
3035
c0209c2c
AM
3036 (cond
3037 ((and (consp (car c-state-cache))
3038 (> (cdar c-state-cache) here))
3039 ;; CASE 1: The top of the cache is a brace pair which now encloses
3040 ;; `here'. As good-pos, return the address. of the "{". Since we've no
3041 ;; knowledge of what's inside these braces, we have no alternative but
3042 ;; to direct the caller to scan the buffer from the opening brace.
3043 (setq pos (caar c-state-cache))
3044 (setcar c-state-cache pos)
3045 (list (1+ pos) pos t)) ; return value. We've just converted a brace pair
3046 ; entry into a { entry, so the caller needs to
3047 ; search for a brace pair before the {.
3048
3049 ;; `here' might be inside a literal. Check for this.
3050 ((progn
3051 (setq lit (c-state-literal-at here)
3052 here-lit-start (or (car lit) here)
3053 here-lit-end (or (cdr lit) here))
3054 ;; Has `here' just "newly entered" a macro?
3055 (save-excursion
3056 (goto-char here-lit-start)
3057 (if (and (c-beginning-of-macro)
3058 (or (null c-state-old-cpp-beg)
3059 (not (= (point) c-state-old-cpp-beg))))
3060 (progn
3061 (setq here- (point))
3062 (c-end-of-macro)
3063 (setq here+ (point)))
3064 (setq here- here-lit-start
3065 here+ here-lit-end)))
3066
3067 ;; `here' might be nested inside any depth of parens (or brackets but
3068 ;; not braces). Scan backwards to find the outermost such opening
3069 ;; paren, if there is one. This will be the scan position to return.
3070 (save-restriction
3071 (narrow-to-region cache-pos (point-max))
3072 (setq pos (c-state-balance-parens-backwards here- here+ pos)))
3073 nil)) ; for the cond
3074
3075 ((< pos here-lit-start)
3076 ;; CASE 2: Address of outermost ( or [ which now encloses `here', but
3077 ;; didn't enclose the (previous) `c-state-cache-good-pos'. If there is
3078 ;; a brace pair preceding this, it will already be in `c-state-cache',
3079 ;; unless there was a brace pair after it, i.e. there'll only be one to
3080 ;; scan for if we've just deleted one.
3081 (list pos (and dropped-cons pos) t)) ; Return value.
3082
3083 ;; `here' isn't enclosed in a (previously unrecorded) bracket/paren.
3084 ;; Further forward scanning isn't needed, but we still need to find a
3085 ;; GOOD-POS. Step out of all enclosing "("s on HERE's line.
3086 ((progn
0ec1d2c5
AM
3087 (save-restriction
3088 (narrow-to-region here-bol (point-max))
3089 (setq pos here-lit-start)
3090 (c-safe (while (setq pa (scan-lists pos -1 1))
3091 (setq pos pa)))) ; might signal
c0209c2c
AM
3092 nil)) ; for the cond
3093
3094 ((setq ren (c-safe-scan-lists pos -1 -1 too-far-back))
3095 ;; CASE 3: After a }/)/] before `here''s BOL.
3096 (list (1+ ren) (and dropped-cons pos) nil)) ; Return value
3097
3098 (t
3099 ;; CASE 4; Best of a bad job: BOL before `here-bol', or beginning of
3100 ;; literal containing it.
3101 (setq good-pos (c-state-lit-beg (c-point 'bopl here-bol)))
3102 (list good-pos (and dropped-cons good-pos) nil)))))
0ec1d2c5
AM
3103
3104
3105;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
3106;; Externally visible routines.
3107
3108(defun c-state-cache-init ()
3109 (setq c-state-cache nil
3110 c-state-cache-good-pos 1
3111 c-state-nonlit-pos-cache nil
3112 c-state-nonlit-pos-cache-limit 1
8ed43f15
AM
3113 c-state-semi-nonlit-pos-cache nil
3114 c-state-semi-nonlit-pos-cache-limit 1
0ec1d2c5
AM
3115 c-state-brace-pair-desert nil
3116 c-state-point-min 1
3117 c-state-point-min-lit-type nil
3118 c-state-point-min-lit-start nil
3119 c-state-min-scan-pos 1
3120 c-state-old-cpp-beg nil
3121 c-state-old-cpp-end nil)
3122 (c-state-mark-point-min-literal))
3123
e84efb70
AM
3124;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
3125;; Debugging routines to dump `c-state-cache' in a "replayable" form.
3126;; (defmacro c-sc-de (elt) ; "c-state-cache-dump-element"
3127;; `(format ,(concat "(setq " (symbol-name elt) " %s) ") ,elt))
3128;; (defmacro c-sc-qde (elt) ; "c-state-cache-quote-dump-element"
3129;; `(format ,(concat "(setq " (symbol-name elt) " '%s) ") ,elt))
3130;; (defun c-state-dump ()
3131;; ;; For debugging.
3132;; ;(message
3133;; (concat
3134;; (c-sc-qde c-state-cache)
3135;; (c-sc-de c-state-cache-good-pos)
3136;; (c-sc-qde c-state-nonlit-pos-cache)
3137;; (c-sc-de c-state-nonlit-pos-cache-limit)
3138;; (c-sc-qde c-state-brace-pair-desert)
3139;; (c-sc-de c-state-point-min)
3140;; (c-sc-de c-state-point-min-lit-type)
3141;; (c-sc-de c-state-point-min-lit-start)
3142;; (c-sc-de c-state-min-scan-pos)
3143;; (c-sc-de c-state-old-cpp-beg)
3144;; (c-sc-de c-state-old-cpp-end)))
3145;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
3146
0ec1d2c5
AM
3147(defun c-invalidate-state-cache-1 (here)
3148 ;; Invalidate all info on `c-state-cache' that applies to the buffer at HERE
3149 ;; or higher and set `c-state-cache-good-pos' accordingly. The cache is
3150 ;; left in a consistent state.
3151 ;;
3152 ;; This is much like `c-whack-state-after', but it never changes a paren
3153 ;; pair element into an open paren element. Doing that would mean that the
3154 ;; new open paren wouldn't have the required preceding paren pair element.
785eecbb 3155 ;;
0ec1d2c5
AM
3156 ;; This function is called from c-after-change.
3157
56d093a9 3158 ;; The caches of non-literals:
0ec1d2c5
AM
3159 (if (< here c-state-nonlit-pos-cache-limit)
3160 (setq c-state-nonlit-pos-cache-limit here))
56d093a9
AM
3161 (if (< here c-state-semi-nonlit-pos-cache-limit)
3162 (setq c-state-semi-nonlit-pos-cache-limit here))
0ec1d2c5
AM
3163
3164 ;; `c-state-cache':
3165 ;; Case 1: if `here' is in a literal containing point-min, everything
3166 ;; becomes (or is already) nil.
3167 (if (or (null c-state-cache-good-pos)
3168 (< here (c-state-get-min-scan-pos)))
3169 (setq c-state-cache nil
3170 c-state-cache-good-pos nil
3171 c-state-min-scan-pos nil)
3172
c400c4d7
SM
3173 ;; Truncate `c-state-cache' and set `c-state-cache-good-pos' to a value
3174 ;; below `here'. To maintain its consistency, we may need to insert a new
3175 ;; brace pair.
0ec1d2c5
AM
3176 (let ((here-bol (c-point 'bol here))
3177 too-high-pa ; recorded {/(/[ next above here, or nil.
3178 dropped-cons ; was the last removed element a brace pair?
3179 pa)
3180 ;; The easy bit - knock over-the-top bits off `c-state-cache'.
3181 (while (and c-state-cache
3182 (>= (setq pa (c-state-cache-top-paren)) here))
3183 (setq dropped-cons (consp (car c-state-cache))
3184 too-high-pa (c-state-cache-top-lparen)
3185 c-state-cache (cdr c-state-cache)))
3186
3187 ;; Do we need to add in an earlier brace pair, having lopped one off?
3188 (if (and dropped-cons
3189 (< too-high-pa (+ here c-state-cache-too-far)))
3190 (c-append-lower-brace-pair-to-state-cache too-high-pa here-bol))
3191 (setq c-state-cache-good-pos (or (c-state-cache-after-top-paren)
3192 (c-state-get-min-scan-pos)))))
3193
3194 ;; The brace-pair desert marker:
3195 (when (car c-state-brace-pair-desert)
3196 (if (< here (car c-state-brace-pair-desert))
3197 (setq c-state-brace-pair-desert nil)
3198 (if (< here (cdr c-state-brace-pair-desert))
3199 (setcdr c-state-brace-pair-desert here)))))
3200
3201(defun c-parse-state-1 ()
3202 ;; Find and record all noteworthy parens between some good point earlier in
3203 ;; the file and point. That good point is at least the beginning of the
3204 ;; top-level construct we are in, or the beginning of the preceding
3205 ;; top-level construct if we aren't in one.
3206 ;;
3207 ;; The returned value is a list of the noteworthy parens with the last one
3208 ;; first. If an element in the list is an integer, it's the position of an
3209 ;; open paren (of any type) which has not been closed before the point. If
3210 ;; an element is a cons, it gives the position of a closed BRACE paren
3211 ;; pair[*]; the car is the start brace position and the cdr is the position
3212 ;; following the closing brace. Only the last closed brace paren pair
3213 ;; before each open paren and before the point is recorded, and thus the
3214 ;; state never contains two cons elements in succession. When a close brace
3215 ;; has no matching open brace (e.g., the matching brace is outside the
3216 ;; visible region), it is not represented in the returned value.
3217 ;;
3218 ;; [*] N.B. The close "brace" might be a mismatching close bracket or paren.
3219 ;; This defun explicitly treats mismatching parens/braces/brackets as
3220 ;; matching. It is the open brace which makes it a "brace" pair.
3221 ;;
3222 ;; If POINT is within a macro, open parens and brace pairs within
3223 ;; THIS macro MIGHT be recorded. This depends on whether their
3224 ;; syntactic properties have been suppressed by
3225 ;; `c-neutralize-syntax-in-CPP'. This might need fixing (2008-12-11).
d9e94c22
MS
3226 ;;
3227 ;; Currently no characters which are given paren syntax with the
3228 ;; syntax-table property are recorded, i.e. angle bracket arglist
3229 ;; parens are never present here. Note that this might change.
3230 ;;
0386b551 3231 ;; BUG: This function doesn't cope entirely well with unbalanced
0ec1d2c5
AM
3232 ;; parens in macros. (2008-12-11: this has probably been resolved
3233 ;; by the function `c-neutralize-syntax-in-CPP'.) E.g. in the
3234 ;; following case the brace before the macro isn't balanced with the
3235 ;; one after it:
0386b551
AM
3236 ;;
3237 ;; {
3238 ;; #define X {
3239 ;; }
3240 ;;
0ec1d2c5
AM
3241 ;; Note to maintainers: this function DOES get called with point
3242 ;; within comments and strings, so don't assume it doesn't!
3243 ;;
0386b551 3244 ;; This function might do hidden buffer changes.
0ec1d2c5
AM
3245 (let* ((here (point))
3246 (here-bopl (c-point 'bopl))
3247 strategy ; 'forward, 'backward etc..
3248 ;; Candidate positions to start scanning from:
3249 cache-pos ; highest position below HERE already existing in
3250 ; cache (or 1).
3251 good-pos
3252 start-point
3253 bopl-state
3254 res
3255 scan-backward-pos scan-forward-p) ; used for 'backward.
3256 ;; If POINT-MIN has changed, adjust the cache
3257 (unless (= (point-min) c-state-point-min)
3258 (c-renarrow-state-cache))
3259
3260 ;; Strategy?
3261 (setq res (c-parse-state-get-strategy here c-state-cache-good-pos)
3262 strategy (car res)
3263 cache-pos (cadr res)
3264 start-point (nth 2 res))
3265
3266 (when (eq strategy 'BOD)
3267 (setq c-state-cache nil
3268 c-state-cache-good-pos start-point))
3269
3270 ;; SCAN!
3271 (save-restriction
3272 (cond
3273 ((memq strategy '(forward BOD))
0386b551 3274 (narrow-to-region (point-min) here)
0ec1d2c5
AM
3275 (setq res (c-remove-stale-state-cache start-point here-bopl))
3276 (setq cache-pos (car res)
3277 scan-backward-pos (cadr res)
3278 bopl-state (car (cddr res))) ; will be nil if (< here-bopl
3279 ; start-point)
3280 (if scan-backward-pos
3281 (c-append-lower-brace-pair-to-state-cache scan-backward-pos))
3282 (setq good-pos
3283 (c-append-to-state-cache cache-pos))
3284 (setq c-state-cache-good-pos
3285 (if (and bopl-state
3286 (< good-pos (- here c-state-cache-too-far)))
3287 (c-state-cache-non-literal-place here-bopl bopl-state)
3288 good-pos)))
3289
3290 ((eq strategy 'backward)
3291 (setq res (c-remove-stale-state-cache-backwards here cache-pos)
3292 good-pos (car res)
3293 scan-backward-pos (cadr res)
3294 scan-forward-p (car (cddr res)))
3295 (if scan-backward-pos
3296 (c-append-lower-brace-pair-to-state-cache
3297 scan-backward-pos))
3298 (setq c-state-cache-good-pos
3299 (if scan-forward-p
3300 (progn (narrow-to-region (point-min) here)
3301 (c-append-to-state-cache good-pos))
2e492df3 3302 good-pos)))
0ec1d2c5
AM
3303
3304 (t ; (eq strategy 'IN-LIT)
3305 (setq c-state-cache nil
3306 c-state-cache-good-pos nil)))))
3307
3308 c-state-cache)
3309
3310(defun c-invalidate-state-cache (here)
3311 ;; This is a wrapper over `c-invalidate-state-cache-1'.
3312 ;;
3313 ;; It suppresses the syntactic effect of the < and > (template) brackets and
3314 ;; of all parens in preprocessor constructs, except for any such construct
3315 ;; containing point. We can then call `c-invalidate-state-cache-1' without
3316 ;; worrying further about macros and template delimiters.
3317 (c-with-<->-as-parens-suppressed
3e8f7d91
AM
3318 (if (and c-state-old-cpp-beg
3319 (< c-state-old-cpp-beg here))
0ec1d2c5 3320 (c-with-all-but-one-cpps-commented-out
3e8f7d91
AM
3321 c-state-old-cpp-beg
3322 (min c-state-old-cpp-end here)
0ec1d2c5
AM
3323 (c-invalidate-state-cache-1 here))
3324 (c-with-cpps-commented-out
3325 (c-invalidate-state-cache-1 here)))))
3326
3327(defun c-parse-state ()
3328 ;; This is a wrapper over `c-parse-state-1'. See that function for a
3329 ;; description of the functionality and return value.
3330 ;;
3331 ;; It suppresses the syntactic effect of the < and > (template) brackets and
3332 ;; of all parens in preprocessor constructs, except for any such construct
3333 ;; containing point. We can then call `c-parse-state-1' without worrying
3334 ;; further about macros and template delimiters.
3335 (let (here-cpp-beg here-cpp-end)
3336 (save-excursion
3337 (when (c-beginning-of-macro)
3338 (setq here-cpp-beg (point))
3339 (unless
3340 (> (setq here-cpp-end (c-syntactic-end-of-macro))
3341 here-cpp-beg)
3342 (setq here-cpp-beg nil here-cpp-end nil))))
3343 ;; FIXME!!! Put in a `condition-case' here to protect the integrity of the
3344 ;; subsystem.
3345 (prog1
3346 (c-with-<->-as-parens-suppressed
3347 (if (and here-cpp-beg (> here-cpp-end here-cpp-beg))
3348 (c-with-all-but-one-cpps-commented-out
3349 here-cpp-beg here-cpp-end
3350 (c-parse-state-1))
3351 (c-with-cpps-commented-out
3352 (c-parse-state-1))))
3e8f7d91
AM
3353 (setq c-state-old-cpp-beg (and here-cpp-beg (copy-marker here-cpp-beg t))
3354 c-state-old-cpp-end (and here-cpp-end (copy-marker here-cpp-end t)))
3355 )))
0ec1d2c5
AM
3356
3357;; Debug tool to catch cache inconsistencies. This is called from
3358;; 000tests.el.
a66cd3ee
MS
3359(defvar c-debug-parse-state nil)
3360(unless (fboundp 'c-real-parse-state)
3361 (fset 'c-real-parse-state (symbol-function 'c-parse-state)))
3362(cc-bytecomp-defun c-real-parse-state)
cd5495ff 3363
b03aabda 3364(defvar c-parse-state-point nil)
cd5495ff
AM
3365(defvar c-parse-state-state nil)
3366(defun c-record-parse-state-state ()
b03aabda 3367 (setq c-parse-state-point (point))
cd5495ff
AM
3368 (setq c-parse-state-state
3369 (mapcar
3370 (lambda (arg)
b03aabda
AM
3371 (let ((val (symbol-value arg)))
3372 (cons arg
3373 (if (consp val)
3374 (copy-tree val)
3375 val))))
cd5495ff
AM
3376 '(c-state-cache
3377 c-state-cache-good-pos
3378 c-state-nonlit-pos-cache
3379 c-state-nonlit-pos-cache-limit
8ed43f15
AM
3380 c-state-semi-nonlit-pos-cache
3381 c-state-semi-nonlit-pos-cache-limit
cd5495ff
AM
3382 c-state-brace-pair-desert
3383 c-state-point-min
3384 c-state-point-min-lit-type
3385 c-state-point-min-lit-start
3386 c-state-min-scan-pos
3387 c-state-old-cpp-beg
b03aabda
AM
3388 c-state-old-cpp-end
3389 c-parse-state-point))))
cd5495ff
AM
3390(defun c-replay-parse-state-state ()
3391 (message
3392 (concat "(setq "
3393 (mapconcat
3394 (lambda (arg)
3395 (format "%s %s%s" (car arg) (if (atom (cdr arg)) "" "'") (cdr arg)))
3396 c-parse-state-state " ")
3397 ")")))
3398
b03aabda
AM
3399(defun c-debug-parse-state-double-cons (state)
3400 (let (state-car conses-not-ok)
3401 (while state
3402 (setq state-car (car state)
3403 state (cdr state))
3404 (if (and (consp state-car)
3405 (consp (car state)))
3406 (setq conses-not-ok t)))
3407 conses-not-ok))
3408
a66cd3ee 3409(defun c-debug-parse-state ()
0ec1d2c5 3410 (let ((here (point)) (res1 (c-real-parse-state)) res2)
0386b551 3411 (let ((c-state-cache nil)
0ec1d2c5
AM
3412 (c-state-cache-good-pos 1)
3413 (c-state-nonlit-pos-cache nil)
3414 (c-state-nonlit-pos-cache-limit 1)
3415 (c-state-brace-pair-desert nil)
3416 (c-state-point-min 1)
3417 (c-state-point-min-lit-type nil)
3418 (c-state-point-min-lit-start nil)
3419 (c-state-min-scan-pos 1)
3420 (c-state-old-cpp-beg nil)
3421 (c-state-old-cpp-end nil))
a66cd3ee
MS
3422 (setq res2 (c-real-parse-state)))
3423 (unless (equal res1 res2)
0386b551
AM
3424 ;; The cache can actually go further back due to the ad-hoc way
3425 ;; the first paren is found, so try to whack off a bit of its
3426 ;; start before complaining.
cd5495ff
AM
3427 ;; (save-excursion
3428 ;; (goto-char (or (c-least-enclosing-brace res2) (point)))
3429 ;; (c-beginning-of-defun-1)
3430 ;; (while (not (or (bobp) (eq (char-after) ?{)))
3431 ;; (c-beginning-of-defun-1))
3432 ;; (unless (equal (c-whack-state-before (point) res1) res2)
3433 ;; (message (concat "c-parse-state inconsistency at %s: "
3434 ;; "using cache: %s, from scratch: %s")
3435 ;; here res1 res2)))
3436 (message (concat "c-parse-state inconsistency at %s: "
3437 "using cache: %s, from scratch: %s")
3438 here res1 res2)
3439 (message "Old state:")
3440 (c-replay-parse-state-state))
b03aabda
AM
3441
3442 (when (c-debug-parse-state-double-cons res1)
3443 (message "c-parse-state INVALIDITY at %s: %s"
3444 here res1)
3445 (message "Old state:")
3446 (c-replay-parse-state-state))
3447
cd5495ff 3448 (c-record-parse-state-state)
b03aabda
AM
3449 res2 ; res1 correct a cascading series of errors ASAP
3450 ))
0ec1d2c5 3451
a66cd3ee
MS
3452(defun c-toggle-parse-state-debug (&optional arg)
3453 (interactive "P")
3454 (setq c-debug-parse-state (c-calculate-state arg c-debug-parse-state))
3455 (fset 'c-parse-state (symbol-function (if c-debug-parse-state
3456 'c-debug-parse-state
3457 'c-real-parse-state)))
b03aabda
AM
3458 (c-keep-region-active)
3459 (message "c-debug-parse-state %sabled"
3460 (if c-debug-parse-state "en" "dis")))
0386b551
AM
3461(when c-debug-parse-state
3462 (c-toggle-parse-state-debug 1))
a66cd3ee 3463
0ec1d2c5 3464\f
d9e94c22
MS
3465(defun c-whack-state-before (bufpos paren-state)
3466 ;; Whack off any state information from PAREN-STATE which lies
3467 ;; before BUFPOS. Not destructive on PAREN-STATE.
d9e94c22
MS
3468 (let* ((newstate (list nil))
3469 (ptr newstate)
3470 car)
3471 (while paren-state
3472 (setq car (car paren-state)
3473 paren-state (cdr paren-state))
3474 (if (< (if (consp car) (car car) car) bufpos)
3475 (setq paren-state nil)
3476 (setcdr ptr (list car))
3477 (setq ptr (cdr ptr))))
3478 (cdr newstate)))
3479
3480(defun c-whack-state-after (bufpos paren-state)
3481 ;; Whack off any state information from PAREN-STATE which lies at or
3482 ;; after BUFPOS. Not destructive on PAREN-STATE.
d9e94c22
MS
3483 (catch 'done
3484 (while paren-state
3485 (let ((car (car paren-state)))
3486 (if (consp car)
3487 ;; just check the car, because in a balanced brace
3488 ;; expression, it must be impossible for the corresponding
3489 ;; close brace to be before point, but the open brace to
3490 ;; be after.
3491 (if (<= bufpos (car car))
3492 nil ; whack it off
3493 (if (< bufpos (cdr car))
3494 ;; its possible that the open brace is before
3495 ;; bufpos, but the close brace is after. In that
3496 ;; case, convert this to a non-cons element. The
3497 ;; rest of the state is before bufpos, so we're
3498 ;; done.
3499 (throw 'done (cons (car car) (cdr paren-state)))
3500 ;; we know that both the open and close braces are
3501 ;; before bufpos, so we also know that everything else
3502 ;; on state is before bufpos.
3503 (throw 'done paren-state)))
3504 (if (<= bufpos car)
3505 nil ; whack it off
3506 ;; it's before bufpos, so everything else should too.
3507 (throw 'done paren-state)))
3508 (setq paren-state (cdr paren-state)))
3509 nil)))
3510
3511(defun c-most-enclosing-brace (paren-state &optional bufpos)
3512 ;; Return the bufpos of the innermost enclosing open paren before
0386b551 3513 ;; bufpos, or nil if none was found.
d9e94c22
MS
3514 (let (enclosingp)
3515 (or bufpos (setq bufpos 134217727))
3516 (while paren-state
3517 (setq enclosingp (car paren-state)
3518 paren-state (cdr paren-state))
3519 (if (or (consp enclosingp)
3520 (>= enclosingp bufpos))
3521 (setq enclosingp nil)
d9e94c22
MS
3522 (setq paren-state nil)))
3523 enclosingp))
3524
0386b551
AM
3525(defun c-least-enclosing-brace (paren-state)
3526 ;; Return the bufpos of the outermost enclosing open paren, or nil
3527 ;; if none was found.
d9e94c22 3528 (let (pos elem)
d9e94c22
MS
3529 (while paren-state
3530 (setq elem (car paren-state)
3531 paren-state (cdr paren-state))
0386b551
AM
3532 (if (integerp elem)
3533 (setq pos elem)))
d9e94c22
MS
3534 pos))
3535
3536(defun c-safe-position (bufpos paren-state)
0386b551
AM
3537 ;; Return the closest "safe" position recorded on PAREN-STATE that
3538 ;; is higher up than BUFPOS. Return nil if PAREN-STATE doesn't
3539 ;; contain any. Return nil if BUFPOS is nil, which is useful to
3540 ;; find the closest limit before a given limit that might be nil.
d9e94c22 3541 ;;
0386b551
AM
3542 ;; A "safe" position is a position at or after a recorded open
3543 ;; paren, or after a recorded close paren. The returned position is
3544 ;; thus either the first position after a close brace, or the first
3545 ;; position after an enclosing paren, or at the enclosing paren in
3546 ;; case BUFPOS is immediately after it.
d9e94c22
MS
3547 (when bufpos
3548 (let (elem)
3549 (catch 'done
3550 (while paren-state
3551 (setq elem (car paren-state))
3552 (if (consp elem)
3553 (cond ((< (cdr elem) bufpos)
3554 (throw 'done (cdr elem)))
3555 ((< (car elem) bufpos)
3556 ;; See below.
3557 (throw 'done (min (1+ (car elem)) bufpos))))
3558 (if (< elem bufpos)
3559 ;; elem is the position at and not after the opening paren, so
3560 ;; we can go forward one more step unless it's equal to
3561 ;; bufpos. This is useful in some cases avoid an extra paren
3562 ;; level between the safe position and bufpos.
3563 (throw 'done (min (1+ elem) bufpos))))
3564 (setq paren-state (cdr paren-state)))))))
3565
3566(defun c-beginning-of-syntax ()
3567 ;; This is used for `font-lock-beginning-of-syntax-function'. It
3568 ;; goes to the closest previous point that is known to be outside
3569 ;; any string literal or comment. `c-state-cache' is used if it has
3570 ;; a position in the vicinity.
3571 (let* ((paren-state c-state-cache)
3572 elem
3573
3574 (pos (catch 'done
3575 ;; Note: Similar code in `c-safe-position'. The
3576 ;; difference is that we accept a safe position at
3577 ;; the point and don't bother to go forward past open
3578 ;; parens.
3579 (while paren-state
3580 (setq elem (car paren-state))
3581 (if (consp elem)
3582 (cond ((<= (cdr elem) (point))
3583 (throw 'done (cdr elem)))
3584 ((<= (car elem) (point))
3585 (throw 'done (car elem))))
3586 (if (<= elem (point))
3587 (throw 'done elem)))
3588 (setq paren-state (cdr paren-state)))
3589 (point-min))))
3590
3591 (if (> pos (- (point) 4000))
3592 (goto-char pos)
3593 ;; The position is far back. Try `c-beginning-of-defun-1'
3594 ;; (although we can't be entirely sure it will go to a position
3595 ;; outside a comment or string in current emacsen). FIXME:
3596 ;; Consult `syntax-ppss' here.
3597 (c-beginning-of-defun-1)
3598 (if (< (point) pos)
3599 (goto-char pos)))))
3600
3601\f
3602;; Tools for scanning identifiers and other tokens.
3603
3604(defun c-on-identifier ()
3605 "Return non-nil if the point is on or directly after an identifier.
3606Keywords are recognized and not considered identifiers. If an
3607identifier is detected, the returned value is its starting position.
0386b551
AM
3608If an identifier ends at the point and another begins at it \(can only
3609happen in Pike) then the point for the preceding one is returned.
d9e94c22 3610
0386b551
AM
3611Note that this function might do hidden buffer changes. See the
3612comment at the start of cc-engine.el for more info."
3613
3614 ;; FIXME: Shouldn't this function handle "operator" in C++?
d9e94c22
MS
3615
3616 (save-excursion
0386b551
AM
3617 (skip-syntax-backward "w_")
3618
3619 (or
3620
3621 ;; Check for a normal (non-keyword) identifier.
3622 (and (looking-at c-symbol-start)
3623 (not (looking-at c-keywords-regexp))
3624 (point))
3625
3626 (when (c-major-mode-is 'pike-mode)
3627 ;; Handle the `<operator> syntax in Pike.
3628 (let ((pos (point)))
3629 (skip-chars-backward "-!%&*+/<=>^|~[]()")
3630 (and (if (< (skip-chars-backward "`") 0)
3631 t
3632 (goto-char pos)
3633 (eq (char-after) ?\`))
3634 (looking-at c-symbol-key)
3635 (>= (match-end 0) pos)
3636 (point))))
3637
3638 ;; Handle the "operator +" syntax in C++.
3639 (when (and c-overloadable-operators-regexp
3640 (= (c-backward-token-2 0) 0))
3641
3642 (cond ((and (looking-at c-overloadable-operators-regexp)
51c9af45 3643 (or (not c-opt-op-identifier-prefix)
0386b551 3644 (and (= (c-backward-token-2 1) 0)
51c9af45 3645 (looking-at c-opt-op-identifier-prefix))))
0386b551
AM
3646 (point))
3647
3648 ((save-excursion
51c9af45
AM
3649 (and c-opt-op-identifier-prefix
3650 (looking-at c-opt-op-identifier-prefix)
0386b551
AM
3651 (= (c-forward-token-2 1) 0)
3652 (looking-at c-overloadable-operators-regexp)))
3653 (point))))
3654
3655 )))
d9e94c22
MS
3656
3657(defsubst c-simple-skip-symbol-backward ()
3658 ;; If the point is at the end of a symbol then skip backward to the
3659 ;; beginning of it. Don't move otherwise. Return non-nil if point
3660 ;; moved.
0386b551
AM
3661 ;;
3662 ;; This function might do hidden buffer changes.
d9e94c22
MS
3663 (or (< (skip-syntax-backward "w_") 0)
3664 (and (c-major-mode-is 'pike-mode)
3665 ;; Handle the `<operator> syntax in Pike.
3666 (let ((pos (point)))
2a15eb73 3667 (if (and (< (skip-chars-backward "-!%&*+/<=>^|~[]()") 0)
d9e94c22
MS
3668 (< (skip-chars-backward "`") 0)
3669 (looking-at c-symbol-key)
3670 (>= (match-end 0) pos))
3671 t
3672 (goto-char pos)
3673 nil)))))
3674
0386b551 3675(defun c-beginning-of-current-token (&optional back-limit)
d9e94c22
MS
3676 ;; Move to the beginning of the current token. Do not move if not
3677 ;; in the middle of one. BACK-LIMIT may be used to bound the
3678 ;; backward search; if given it's assumed to be at the boundary
a85fd6da 3679 ;; between two tokens. Return non-nil if the point is moved, nil
580fba94 3680 ;; otherwise.
0386b551
AM
3681 ;;
3682 ;; This function might do hidden buffer changes.
d9e94c22 3683 (let ((start (point)))
580fba94
AM
3684 (if (looking-at "\\w\\|\\s_")
3685 (skip-syntax-backward "w_" back-limit)
3686 (when (< (skip-syntax-backward ".()" back-limit) 0)
3687 (while (let ((pos (or (and (looking-at c-nonsymbol-token-regexp)
3688 (match-end 0))
3689 ;; `c-nonsymbol-token-regexp' should always match
3690 ;; since we've skipped backward over punctuator
3691 ;; or paren syntax, but consume one char in case
3692 ;; it doesn't so that we don't leave point before
3693 ;; some earlier incorrect token.
3694 (1+ (point)))))
3695 (if (<= pos start)
3696 (goto-char pos))))))
3697 (< (point) start)))
d9e94c22 3698
ff959bab 3699(defun c-end-of-current-token (&optional back-limit)
d9e94c22
MS
3700 ;; Move to the end of the current token. Do not move if not in the
3701 ;; middle of one. BACK-LIMIT may be used to bound the backward
3702 ;; search; if given it's assumed to be at the boundary between two
ff959bab 3703 ;; tokens. Return non-nil if the point is moved, nil otherwise.
0386b551
AM
3704 ;;
3705 ;; This function might do hidden buffer changes.
d9e94c22
MS
3706 (let ((start (point)))
3707 (cond ((< (skip-syntax-backward "w_" (1- start)) 0)
3708 (skip-syntax-forward "w_"))
3709 ((< (skip-syntax-backward ".()" back-limit) 0)
3710 (while (progn
3711 (if (looking-at c-nonsymbol-token-regexp)
3712 (goto-char (match-end 0))
3713 ;; `c-nonsymbol-token-regexp' should always match since
3714 ;; we've skipped backward over punctuator or paren
3715 ;; syntax, but move forward in case it doesn't so that
3716 ;; we don't leave point earlier than we started with.
3717 (forward-char))
ff959bab
MS
3718 (< (point) start)))))
3719 (> (point) start)))
d9e94c22
MS
3720
3721(defconst c-jump-syntax-balanced
3722 (if (memq 'gen-string-delim c-emacs-features)
3723 "\\w\\|\\s_\\|\\s\(\\|\\s\)\\|\\s\"\\|\\s|"
3724 "\\w\\|\\s_\\|\\s\(\\|\\s\)\\|\\s\""))
3725
3726(defconst c-jump-syntax-unbalanced
3727 (if (memq 'gen-string-delim c-emacs-features)
3728 "\\w\\|\\s_\\|\\s\"\\|\\s|"
3729 "\\w\\|\\s_\\|\\s\""))
3730
3731(defun c-forward-token-2 (&optional count balanced limit)
3732 "Move forward by tokens.
3733A token is defined as all symbols and identifiers which aren't
3734syntactic whitespace \(note that multicharacter tokens like \"==\" are
3735treated properly). Point is always either left at the beginning of a
3736token or not moved at all. COUNT specifies the number of tokens to
3737move; a negative COUNT moves in the opposite direction. A COUNT of 0
3738moves to the next token beginning only if not already at one. If
3739BALANCED is true, move over balanced parens, otherwise move into them.
3740Also, if BALANCED is true, never move out of an enclosing paren.
3741
3742LIMIT sets the limit for the movement and defaults to the point limit.
3743The case when LIMIT is set in the middle of a token, comment or macro
3744is handled correctly, i.e. the point won't be left there.
3745
3746Return the number of tokens left to move \(positive or negative). If
3747BALANCED is true, a move over a balanced paren counts as one. Note
3748that if COUNT is 0 and no appropriate token beginning is found, 1 will
3749be returned. Thus, a return value of 0 guarantees that point is at
3750the requested position and a return value less \(without signs) than
0386b551
AM
3751COUNT guarantees that point is at the beginning of some token.
3752
3753Note that this function might do hidden buffer changes. See the
3754comment at the start of cc-engine.el for more info."
d9e94c22
MS
3755
3756 (or count (setq count 1))
3757 (if (< count 0)
3758 (- (c-backward-token-2 (- count) balanced limit))
3759
3760 (let ((jump-syntax (if balanced
3761 c-jump-syntax-balanced
3762 c-jump-syntax-unbalanced))
3763 (last (point))
3764 (prev (point)))
3765
3766 (if (zerop count)
3767 ;; If count is zero we should jump if in the middle of a token.
3768 (c-end-of-current-token))
3769
3770 (save-restriction
3771 (if limit (narrow-to-region (point-min) limit))
3772 (if (/= (point)
3773 (progn (c-forward-syntactic-ws) (point)))
3774 ;; Skip whitespace. Count this as a move if we did in
3775 ;; fact move.
3776 (setq count (max (1- count) 0)))
3777
3778 (if (eobp)
3779 ;; Moved out of bounds. Make sure the returned count isn't zero.
3780 (progn
3781 (if (zerop count) (setq count 1))
3782 (goto-char last))
3783
3784 ;; Use `condition-case' to avoid having the limit tests
3785 ;; inside the loop.
3786 (condition-case nil
3787 (while (and
3788 (> count 0)
3789 (progn
3790 (setq last (point))
3791 (cond ((looking-at jump-syntax)
3792 (goto-char (scan-sexps (point) 1))
3793 t)
3794 ((looking-at c-nonsymbol-token-regexp)
3795 (goto-char (match-end 0))
3796 t)
3797 ;; `c-nonsymbol-token-regexp' above should always
3798 ;; match if there are correct tokens. Try to
3799 ;; widen to see if the limit was set in the
3800 ;; middle of one, else fall back to treating
3801 ;; the offending thing as a one character token.
3802 ((and limit
3803 (save-restriction
3804 (widen)
3805 (looking-at c-nonsymbol-token-regexp)))
3806 nil)
3807 (t
3808 (forward-char)
3809 t))))
3810 (c-forward-syntactic-ws)
3811 (setq prev last
3812 count (1- count)))
3813 (error (goto-char last)))
3814
3815 (when (eobp)
3816 (goto-char prev)
3817 (setq count (1+ count)))))
3818
3819 count)))
3820
3821(defun c-backward-token-2 (&optional count balanced limit)
3822 "Move backward by tokens.
3823See `c-forward-token-2' for details."
3824
3825 (or count (setq count 1))
3826 (if (< count 0)
3827 (- (c-forward-token-2 (- count) balanced limit))
3828
3829 (or limit (setq limit (point-min)))
3830 (let ((jump-syntax (if balanced
3831 c-jump-syntax-balanced
3832 c-jump-syntax-unbalanced))
3833 (last (point)))
3834
3835 (if (zerop count)
3836 ;; The count is zero so try to skip to the beginning of the
3837 ;; current token.
3838 (if (> (point)
3839 (progn (c-beginning-of-current-token) (point)))
3840 (if (< (point) limit)
3841 ;; The limit is inside the same token, so return 1.
3842 (setq count 1))
3843
3844 ;; We're not in the middle of a token. If there's
3845 ;; whitespace after the point then we must move backward,
3846 ;; so set count to 1 in that case.
3847 (and (looking-at c-syntactic-ws-start)
3848 ;; If we're looking at a '#' that might start a cpp
3849 ;; directive then we have to do a more elaborate check.
3850 (or (/= (char-after) ?#)
3851 (not c-opt-cpp-prefix)
3852 (save-excursion
3853 (and (= (point)
3854 (progn (beginning-of-line)
3855 (looking-at "[ \t]*")
3856 (match-end 0)))
3857 (or (bobp)
3858 (progn (backward-char)
3859 (not (eq (char-before) ?\\)))))))
3860 (setq count 1))))
3861
3862 ;; Use `condition-case' to avoid having to check for buffer
3863 ;; limits in `backward-char', `scan-sexps' and `goto-char' below.
3864 (condition-case nil
3865 (while (and
3866 (> count 0)
3867 (progn
3868 (c-backward-syntactic-ws)
3869 (backward-char)
3870 (if (looking-at jump-syntax)
3871 (goto-char (scan-sexps (1+ (point)) -1))
3872 ;; This can be very inefficient if there's a long
3873 ;; sequence of operator tokens without any separation.
3874 ;; That doesn't happen in practice, anyway.
3875 (c-beginning-of-current-token))
3876 (>= (point) limit)))
3877 (setq last (point)
3878 count (1- count)))
3879 (error (goto-char last)))
3880
3881 (if (< (point) limit)
3882 (goto-char last))
3883
3884 count)))
3885
3886(defun c-forward-token-1 (&optional count balanced limit)
3887 "Like `c-forward-token-2' but doesn't treat multicharacter operator
3888tokens like \"==\" as single tokens, i.e. all sequences of symbol
3889characters are jumped over character by character. This function is
3890for compatibility only; it's only a wrapper over `c-forward-token-2'."
3891 (let ((c-nonsymbol-token-regexp "\\s.\\|\\s\(\\|\\s\)"))
3892 (c-forward-token-2 count balanced limit)))
3893
3894(defun c-backward-token-1 (&optional count balanced limit)
3895 "Like `c-backward-token-2' but doesn't treat multicharacter operator
3896tokens like \"==\" as single tokens, i.e. all sequences of symbol
3897characters are jumped over character by character. This function is
3898for compatibility only; it's only a wrapper over `c-backward-token-2'."
3899 (let ((c-nonsymbol-token-regexp "\\s.\\|\\s\(\\|\\s\)"))
3900 (c-backward-token-2 count balanced limit)))
3901
3902\f
3903;; Tools for doing searches restricted to syntactically relevant text.
3904
3905(defun c-syntactic-re-search-forward (regexp &optional bound noerror
3906 paren-level not-inside-token
3907 lookbehind-submatch)
3908 "Like `re-search-forward', but only report matches that are found
3909in syntactically significant text. I.e. matches in comments, macros
3910or string literals are ignored. The start point is assumed to be
3911outside any comment, macro or string literal, or else the content of
3912that region is taken as syntactically significant text.
3913
3914If PAREN-LEVEL is non-nil, an additional restriction is added to
2a15eb73
MS
3915ignore matches in nested paren sexps. The search will also not go
3916outside the current list sexp, which has the effect that if the point
3917should be moved to BOUND when no match is found \(i.e. NOERROR is
3918neither nil nor t), then it will be at the closing paren if the end of
3919the current list sexp is encountered first.
d9e94c22
MS
3920
3921If NOT-INSIDE-TOKEN is non-nil, matches in the middle of tokens are
3922ignored. Things like multicharacter operators and special symbols
3923\(e.g. \"`()\" in Pike) are handled but currently not floating point
3924constants.
3925
3926If LOOKBEHIND-SUBMATCH is non-nil, it's taken as a number of a
3927subexpression in REGEXP. The end of that submatch is used as the
3928position to check for syntactic significance. If LOOKBEHIND-SUBMATCH
3929isn't used or if that subexpression didn't match then the start
3930position of the whole match is used instead. The \"look behind\"
3931subexpression is never tested before the starting position, so it
3932might be a good idea to include \\=\\= as a match alternative in it.
3933
3934Optimization note: Matches might be missed if the \"look behind\"
2a15eb73 3935subexpression can match the end of nonwhite syntactic whitespace,
d9e94c22 3936i.e. the end of comments or cpp directives. This since the function
2a15eb73
MS
3937skips over such things before resuming the search. It's on the other
3938hand not safe to assume that the \"look behind\" subexpression never
3939matches syntactic whitespace.
3940
3941Bug: Unbalanced parens inside cpp directives are currently not handled
3942correctly \(i.e. they don't get ignored as they should) when
0386b551
AM
3943PAREN-LEVEL is set.
3944
3945Note that this function might do hidden buffer changes. See the
3946comment at the start of cc-engine.el for more info."
d9e94c22
MS
3947
3948 (or bound (setq bound (point-max)))
3949 (if paren-level (setq paren-level -1))
3950
3951 ;;(message "c-syntactic-re-search-forward %s %s %S" (point) bound regexp)
3952
3953 (let ((start (point))
2a15eb73
MS
3954 tmp
3955 ;; Start position for the last search.
3956 search-pos
3957 ;; The `parse-partial-sexp' state between the start position
3958 ;; and the point.
3959 state
3960 ;; The current position after the last state update. The next
3961 ;; `parse-partial-sexp' continues from here.
3962 (state-pos (point))
3963 ;; The position at which to check the state and the state
3964 ;; there. This is separate from `state-pos' since we might
3965 ;; need to back up before doing the next search round.
3966 check-pos check-state
3967 ;; Last position known to end a token.
d9e94c22 3968 (last-token-end-pos (point-min))
2a15eb73
MS
3969 ;; Set when a valid match is found.
3970 found)
d9e94c22
MS
3971
3972 (condition-case err
3973 (while
3974 (and
2a15eb73
MS
3975 (progn
3976 (setq search-pos (point))
3977 (re-search-forward regexp bound noerror))
d9e94c22
MS
3978
3979 (progn
2a15eb73
MS
3980 (setq state (parse-partial-sexp
3981 state-pos (match-beginning 0) paren-level nil state)
3982 state-pos (point))
d9e94c22 3983 (if (setq check-pos (and lookbehind-submatch
2a15eb73
MS
3984 (or (not paren-level)
3985 (>= (car state) 0))
d9e94c22
MS
3986 (match-end lookbehind-submatch)))
3987 (setq check-state (parse-partial-sexp
2a15eb73
MS
3988 state-pos check-pos paren-level nil state))
3989 (setq check-pos state-pos
d9e94c22
MS
3990 check-state state))
3991
2a15eb73
MS
3992 ;; NOTE: If we got a look behind subexpression and get
3993 ;; an insignificant match in something that isn't
d9e94c22
MS
3994 ;; syntactic whitespace (i.e. strings or in nested
3995 ;; parentheses), then we can never skip more than a
2a15eb73
MS
3996 ;; single character from the match start position
3997 ;; (i.e. `state-pos' here) before continuing the
3998 ;; search. That since the look behind subexpression
3999 ;; might match the end of the insignificant region in
4000 ;; the next search.
d9e94c22
MS
4001
4002 (cond
d9e94c22
MS
4003 ((elt check-state 7)
4004 ;; Match inside a line comment. Skip to eol. Use
4005 ;; `re-search-forward' instead of `skip-chars-forward' to get
4006 ;; the right bound behavior.
4007 (re-search-forward "[\n\r]" bound noerror))
4008
4009 ((elt check-state 4)
4010 ;; Match inside a block comment. Skip to the '*/'.
4011 (search-forward "*/" bound noerror))
4012
4013 ((and (not (elt check-state 5))
4014 (eq (char-before check-pos) ?/)
2a15eb73 4015 (not (c-get-char-property (1- check-pos) 'syntax-table))
d9e94c22
MS
4016 (memq (char-after check-pos) '(?/ ?*)))
4017 ;; Match in the middle of the opener of a block or line
4018 ;; comment.
4019 (if (= (char-after check-pos) ?/)
4020 (re-search-forward "[\n\r]" bound noerror)
4021 (search-forward "*/" bound noerror)))
4022
2a15eb73
MS
4023 ;; The last `parse-partial-sexp' above might have
4024 ;; stopped short of the real check position if the end
4025 ;; of the current sexp was encountered in paren-level
4026 ;; mode. The checks above are always false in that
4027 ;; case, and since they can do better skipping in
4028 ;; lookbehind-submatch mode, we do them before
4029 ;; checking the paren level.
4030
4031 ((and paren-level
4032 (/= (setq tmp (car check-state)) 0))
4033 ;; Check the paren level first since we're short of the
4034 ;; syntactic checking position if the end of the
4035 ;; current sexp was encountered by `parse-partial-sexp'.
4036 (if (> tmp 0)
4037
4038 ;; Inside a nested paren sexp.
4039 (if lookbehind-submatch
4040 ;; See the NOTE above.
4041 (progn (goto-char state-pos) t)
4042 ;; Skip out of the paren quickly.
4043 (setq state (parse-partial-sexp state-pos bound 0 nil state)
4044 state-pos (point)))
4045
4046 ;; Have exited the current paren sexp.
4047 (if noerror
4048 (progn
4049 ;; The last `parse-partial-sexp' call above
4050 ;; has left us just after the closing paren
4051 ;; in this case, so we can modify the bound
4052 ;; to leave the point at the right position
4053 ;; upon return.
4054 (setq bound (1- (point)))
4055 nil)
4056 (signal 'search-failed (list regexp)))))
4057
4058 ((setq tmp (elt check-state 3))
4059 ;; Match inside a string.
4060 (if (or lookbehind-submatch
4061 (not (integerp tmp)))
4062 ;; See the NOTE above.
4063 (progn (goto-char state-pos) t)
4064 ;; Skip to the end of the string before continuing.
4065 (let ((ender (make-string 1 tmp)) (continue t))
4066 (while (if (search-forward ender bound noerror)
4067 (progn
4068 (setq state (parse-partial-sexp
4069 state-pos (point) nil nil state)
4070 state-pos (point))
4071 (elt state 3))
4072 (setq continue nil)))
4073 continue)))
d9e94c22
MS
4074
4075 ((save-excursion
4076 (save-match-data
4077 (c-beginning-of-macro start)))
4078 ;; Match inside a macro. Skip to the end of it.
4079 (c-end-of-macro)
4080 (cond ((<= (point) bound) t)
4081 (noerror nil)
2a15eb73 4082 (t (signal 'search-failed (list regexp)))))
d9e94c22 4083
2a15eb73
MS
4084 ((and not-inside-token
4085 (or (< check-pos last-token-end-pos)
4086 (< check-pos
4087 (save-excursion
4088 (goto-char check-pos)
4089 (save-match-data
4090 (c-end-of-current-token last-token-end-pos))
4091 (setq last-token-end-pos (point))))))
4092 ;; Inside a token.
4093 (if lookbehind-submatch
4094 ;; See the NOTE above.
4095 (goto-char state-pos)
4096 (goto-char (min last-token-end-pos bound))))
d9e94c22
MS
4097
4098 (t
4099 ;; A real match.
4100 (setq found t)
2a15eb73
MS
4101 nil)))
4102
4103 ;; Should loop to search again, but take care to avoid
4104 ;; looping on the same spot.
4105 (or (/= search-pos (point))
4106 (if (= (point) bound)
4107 (if noerror
4108 nil
4109 (signal 'search-failed (list regexp)))
4110 (forward-char)
4111 t))))
d9e94c22
MS
4112
4113 (error
4114 (goto-char start)
4115 (signal (car err) (cdr err))))
4116
2a15eb73 4117 ;;(message "c-syntactic-re-search-forward done %s" (or (match-end 0) (point)))
d9e94c22
MS
4118
4119 (if found
4120 (progn
2a15eb73
MS
4121 (goto-char (match-end 0))
4122 (match-end 0))
d9e94c22
MS
4123
4124 ;; Search failed. Set point as appropriate.
2a15eb73
MS
4125 (if (eq noerror t)
4126 (goto-char start)
4127 (goto-char bound))
d9e94c22
MS
4128 nil)))
4129
47641aac
GM
4130(defvar safe-pos-list) ; bound in c-syntactic-skip-backward
4131
d0fcee66
AM
4132(defsubst c-ssb-lit-begin ()
4133 ;; Return the start of the literal point is in, or nil.
4134 ;; We read and write the variables `safe-pos', `safe-pos-list', `state'
4135 ;; bound in the caller.
4136
4137 ;; Use `parse-partial-sexp' from a safe position down to the point to check
4138 ;; if it's outside comments and strings.
4139 (save-excursion
4140 (let ((pos (point)) safe-pos state pps-end-pos)
4141 ;; Pick a safe position as close to the point as possible.
4142 ;;
4143 ;; FIXME: Consult `syntax-ppss' here if our cache doesn't give a good
4144 ;; position.
47641aac 4145
d0fcee66
AM
4146 (while (and safe-pos-list
4147 (> (car safe-pos-list) (point)))
4148 (setq safe-pos-list (cdr safe-pos-list)))
4149 (unless (setq safe-pos (car-safe safe-pos-list))
4150 (setq safe-pos (max (or (c-safe-position
4151 (point) (or c-state-cache
4152 (c-parse-state)))
4153 0)
4154 (point-min))
4155 safe-pos-list (list safe-pos)))
4156
4157 ;; Cache positions along the way to use if we have to back up more. We
4158 ;; cache every closing paren on the same level. If the paren cache is
4159 ;; relevant in this region then we're typically already on the same
4160 ;; level as the target position. Note that we might cache positions
4161 ;; after opening parens in case safe-pos is in a nested list. That's
4162 ;; both uncommon and harmless.
4163 (while (progn
4164 (setq state (parse-partial-sexp
4165 safe-pos pos 0))
4166 (< (point) pos))
4167 (setq safe-pos (point)
4168 safe-pos-list (cons safe-pos safe-pos-list)))
4169
4170 ;; If the state contains the start of the containing sexp we cache that
4171 ;; position too, so that parse-partial-sexp in the next run has a bigger
4172 ;; chance of starting at the same level as the target position and thus
4173 ;; will get more good safe positions into the list.
4174 (if (elt state 1)
4175 (setq safe-pos (1+ (elt state 1))
4176 safe-pos-list (cons safe-pos safe-pos-list)))
4177
4178 (if (or (elt state 3) (elt state 4))
4179 ;; Inside string or comment. Continue search at the
4180 ;; beginning of it.
4181 (elt state 8)))))
4182
0386b551 4183(defun c-syntactic-skip-backward (skip-chars &optional limit paren-level)
d9e94c22
MS
4184 "Like `skip-chars-backward' but only look at syntactically relevant chars,
4185i.e. don't stop at positions inside syntactic whitespace or string
4186literals. Preprocessor directives are also ignored, with the exception
4187of the one that the point starts within, if any. If LIMIT is given,
0386b551
AM
4188it's assumed to be at a syntactically relevant position.
4189
4190If PAREN-LEVEL is non-nil, the function won't stop in nested paren
4191sexps, and the search will also not go outside the current paren sexp.
4192However, if LIMIT or the buffer limit is reached inside a nested paren
4193then the point will be left at the limit.
4194
4195Non-nil is returned if the point moved, nil otherwise.
4196
4197Note that this function might do hidden buffer changes. See the
4198comment at the start of cc-engine.el for more info."
d9e94c22
MS
4199
4200 (let ((start (point))
d0fcee66 4201 state-2
d9e94c22
MS
4202 ;; A list of syntactically relevant positions in descending
4203 ;; order. It's used to avoid scanning repeatedly over
4204 ;; potentially large regions with `parse-partial-sexp' to verify
d0fcee66 4205 ;; each position. Used in `c-ssb-lit-begin'
d9e94c22
MS
4206 safe-pos-list
4207 ;; The result from `c-beginning-of-macro' at the start position or the
4208 ;; start position itself if it isn't within a macro. Evaluated on
4209 ;; demand.
0386b551
AM
4210 start-macro-beg
4211 ;; The earliest position after the current one with the same paren
4212 ;; level. Used only when `paren-level' is set.
d0fcee66 4213 lit-beg
0386b551 4214 (paren-level-pos (point)))
d9e94c22 4215
d0fcee66
AM
4216 (while
4217 (progn
4218 ;; The next loop "tries" to find the end point each time round,
4219 ;; loops when it hasn't succeeded.
4220 (while
4221 (and
4222 (< (skip-chars-backward skip-chars limit) 0)
4223
4224 (let ((pos (point)) state-2 pps-end-pos)
4225
4226 (cond
4227 ;; Don't stop inside a literal
4228 ((setq lit-beg (c-ssb-lit-begin))
4229 (goto-char lit-beg)
4230 t)
4231
4232 ((and paren-level
4233 (save-excursion
4234 (setq state-2 (parse-partial-sexp
4235 pos paren-level-pos -1)
4236 pps-end-pos (point))
4237 (/= (car state-2) 0)))
4238 ;; Not at the right level.
4239
4240 (if (and (< (car state-2) 0)
4241 ;; We stop above if we go out of a paren.
4242 ;; Now check whether it precedes or is
4243 ;; nested in the starting sexp.
4244 (save-excursion
4245 (setq state-2
4246 (parse-partial-sexp
4247 pps-end-pos paren-level-pos
4248 nil nil state-2))
4249 (< (car state-2) 0)))
4250
4251 ;; We've stopped short of the starting position
4252 ;; so the hit was inside a nested list. Go up
4253 ;; until we are at the right level.
4254 (condition-case nil
4255 (progn
4256 (goto-char (scan-lists pos -1
4257 (- (car state-2))))
4258 (setq paren-level-pos (point))
4259 (if (and limit (>= limit paren-level-pos))
0386b551 4260 (progn
d0fcee66
AM
4261 (goto-char limit)
4262 nil)
4263 t))
4264 (error
4265 (goto-char (or limit (point-min)))
4266 nil))
4267
4268 ;; The hit was outside the list at the start
4269 ;; position. Go to the start of the list and exit.
4270 (goto-char (1+ (elt state-2 1)))
4271 nil))
4272
4273 ((c-beginning-of-macro limit)
4274 ;; Inside a macro.
4275 (if (< (point)
4276 (or start-macro-beg
4277 (setq start-macro-beg
4278 (save-excursion
4279 (goto-char start)
4280 (c-beginning-of-macro limit)
4281 (point)))))
4282 t
4283
4284 ;; It's inside the same macro we started in so it's
4285 ;; a relevant match.
4286 (goto-char pos)
4287 nil))))))
91af3942 4288
d0fcee66
AM
4289 (> (point)
4290 (progn
4291 ;; Skip syntactic ws afterwards so that we don't stop at the
4292 ;; end of a comment if `skip-chars' is something like "^/".
4293 (c-backward-syntactic-ws)
4294 (point)))))
d9e94c22 4295
0386b551
AM
4296 ;; We might want to extend this with more useful return values in
4297 ;; the future.
4298 (/= (point) start)))
4299
4300;; The following is an alternative implementation of
4301;; `c-syntactic-skip-backward' that uses backward movement to keep
4302;; track of the syntactic context. It turned out to be generally
4303;; slower than the one above which uses forward checks from earlier
4304;; safe positions.
4305;;
4306;;(defconst c-ssb-stop-re
4307;; ;; The regexp matching chars `c-syntactic-skip-backward' needs to
4308;; ;; stop at to avoid going into comments and literals.
4309;; (concat
4310;; ;; Match comment end syntax and string literal syntax. Also match
4311;; ;; '/' for block comment endings (not covered by comment end
4312;; ;; syntax).
4313;; "\\s>\\|/\\|\\s\""
4314;; (if (memq 'gen-string-delim c-emacs-features)
4315;; "\\|\\s|"
4316;; "")
4317;; (if (memq 'gen-comment-delim c-emacs-features)
4318;; "\\|\\s!"
4319;; "")))
4320;;
4321;;(defconst c-ssb-stop-paren-re
4322;; ;; Like `c-ssb-stop-re' but also stops at paren chars.
4323;; (concat c-ssb-stop-re "\\|\\s(\\|\\s)"))
4324;;
4325;;(defconst c-ssb-sexp-end-re
4326;; ;; Regexp matching the ending syntax of a complex sexp.
4327;; (concat c-string-limit-regexp "\\|\\s)"))
4328;;
4329;;(defun c-syntactic-skip-backward (skip-chars &optional limit paren-level)
4330;; "Like `skip-chars-backward' but only look at syntactically relevant chars,
4331;;i.e. don't stop at positions inside syntactic whitespace or string
4332;;literals. Preprocessor directives are also ignored. However, if the
4333;;point is within a comment, string literal or preprocessor directory to
4334;;begin with, its contents is treated as syntactically relevant chars.
4335;;If LIMIT is given, it limits the backward search and the point will be
4336;;left there if no earlier position is found.
4337;;
4338;;If PAREN-LEVEL is non-nil, the function won't stop in nested paren
4339;;sexps, and the search will also not go outside the current paren sexp.
4340;;However, if LIMIT or the buffer limit is reached inside a nested paren
4341;;then the point will be left at the limit.
4342;;
4343;;Non-nil is returned if the point moved, nil otherwise.
4344;;
4345;;Note that this function might do hidden buffer changes. See the
4346;;comment at the start of cc-engine.el for more info."
4347;;
4348;; (save-restriction
4349;; (when limit
4350;; (narrow-to-region limit (point-max)))
4351;;
4352;; (let ((start (point)))
4353;; (catch 'done
4354;; (while (let ((last-pos (point))
4355;; (stop-pos (progn
4356;; (skip-chars-backward skip-chars)
4357;; (point))))
4358;;
4359;; ;; Skip back over the same region as
4360;; ;; `skip-chars-backward' above, but keep to
4361;; ;; syntactically relevant positions.
4362;; (goto-char last-pos)
4363;; (while (and
4364;; ;; `re-search-backward' with a single char regexp
4365;; ;; should be fast.
4366;; (re-search-backward
4367;; (if paren-level c-ssb-stop-paren-re c-ssb-stop-re)
4368;; stop-pos 'move)
4369;;
4370;; (progn
4371;; (cond
4372;; ((looking-at "\\s(")
4373;; ;; `paren-level' is set and we've found the
4374;; ;; start of the containing paren.
4375;; (forward-char)
4376;; (throw 'done t))
4377;;
4378;; ((looking-at c-ssb-sexp-end-re)
4379;; ;; We're at the end of a string literal or paren
4380;; ;; sexp (if `paren-level' is set).
4381;; (forward-char)
4382;; (condition-case nil
4383;; (c-backward-sexp)
4384;; (error
4385;; (goto-char limit)
4386;; (throw 'done t))))
4387;;
4388;; (t
4389;; (forward-char)
4390;; ;; At the end of some syntactic ws or possibly
4391;; ;; after a plain '/' operator.
4392;; (let ((pos (point)))
4393;; (c-backward-syntactic-ws)
4394;; (if (= pos (point))
4395;; ;; Was a plain '/' operator. Go past it.
4396;; (backward-char)))))
4397;;
4398;; (> (point) stop-pos))))
4399;;
4400;; ;; Now the point is either at `stop-pos' or at some
4401;; ;; position further back if `stop-pos' was at a
4402;; ;; syntactically irrelevant place.
4403;;
4404;; ;; Skip additional syntactic ws so that we don't stop
4405;; ;; at the end of a comment if `skip-chars' is
4406;; ;; something like "^/".
4407;; (c-backward-syntactic-ws)
4408;;
4409;; (< (point) stop-pos))))
4410;;
4411;; ;; We might want to extend this with more useful return values
4412;; ;; in the future.
4413;; (/= (point) start))))
d9e94c22
MS
4414
4415\f
4416;; Tools for handling comments and string literals.
4417
13d49cbb 4418(defun c-in-literal (&optional lim detect-cpp)
d9e94c22
MS
4419 "Return the type of literal point is in, if any.
4420The return value is `c' if in a C-style comment, `c++' if in a C++
4421style comment, `string' if in a string literal, `pound' if DETECT-CPP
4422is non-nil and in a preprocessor line, or nil if somewhere else.
4423Optional LIM is used as the backward limit of the search. If omitted,
4424or nil, `c-beginning-of-defun' is used.
4425
4426The last point calculated is cached if the cache is enabled, i.e. if
4427`c-in-literal-cache' is bound to a two element vector.
4428
0386b551
AM
4429Note that this function might do hidden buffer changes. See the
4430comment at the start of cc-engine.el for more info."
2cc769a8
AM
4431 (save-restriction
4432 (widen)
56d093a9 4433 (let* ((safe-place (c-state-semi-safe-place (point)))
2cc769a8
AM
4434 (lit (c-state-pp-to-literal safe-place (point))))
4435 (or (cadr lit)
4436 (and detect-cpp
4437 (save-excursion (c-beginning-of-macro))
4438 'pound)))))
d9e94c22
MS
4439
4440(defun c-literal-limits (&optional lim near not-in-delimiter)
4441 "Return a cons of the beginning and end positions of the comment or
4442string surrounding point (including both delimiters), or nil if point
4443isn't in one. If LIM is non-nil, it's used as the \"safe\" position
4444to start parsing from. If NEAR is non-nil, then the limits of any
4445literal next to point is returned. \"Next to\" means there's only
4446spaces and tabs between point and the literal. The search for such a
4447literal is done first in forward direction. If NOT-IN-DELIMITER is
4448non-nil, the case when point is inside a starting delimiter won't be
a85fd6da 4449recognized. This only has effect for comments which have starting
d9e94c22
MS
4450delimiters with more than one character.
4451
0386b551
AM
4452Note that this function might do hidden buffer changes. See the
4453comment at the start of cc-engine.el for more info."
d9e94c22
MS
4454
4455 (save-excursion
4456 (let* ((pos (point))
56d093a9 4457 (lim (or lim (c-state-semi-safe-place pos)))
2cc769a8
AM
4458 (pp-to-lit (save-restriction
4459 (widen)
4460 (c-state-pp-to-literal lim pos)))
13d49cbb 4461 (state (car pp-to-lit))
13d49cbb 4462 (lit-limits (car (cddr pp-to-lit))))
d9e94c22 4463
13d49cbb
AM
4464 (cond
4465 (lit-limits)
4466 ((and (not not-in-delimiter)
4467 (not (elt state 5))
4468 (eq (char-before) ?/)
4469 (looking-at "[/*]")) ; FIXME!!! use c-line/block-comment-starter. 2008-09-28.
4470 ;; We're standing in a comment starter.
4471 (backward-char 1)
4472 (cons (point) (progn (c-forward-single-comment) (point))))
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)))
4626
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
AM
4731 ;;
4732 ;; This macro might do hidden buffer changes.
d9e94c22
MS
4733
4734 '(progn
4735 ;; Find the next property match position if we haven't got one already.
4736 (unless cfd-prop-match
4737 (save-excursion
4738 (while (progn
4739 (goto-char (next-single-property-change
4740 (point) 'c-type nil cfd-limit))
4741 (and (< (point) cfd-limit)
4742 (not (eq (c-get-char-property (1- (point)) 'c-type)
4743 'c-decl-end)))))
4744 (setq cfd-prop-match (point))))
4745
0386b551
AM
4746 ;; Find the next `c-decl-prefix-or-start-re' match if we haven't
4747 ;; got one already.
d9e94c22 4748 (unless cfd-re-match
0386b551
AM
4749
4750 (if (> cfd-re-match-end (point))
4751 (goto-char cfd-re-match-end))
4752
4753 (while (if (setq cfd-re-match-end
4754 (re-search-forward c-decl-prefix-or-start-re
4755 cfd-limit 'move))
4756
4757 ;; Match. Check if it's inside a comment or string literal.
4758 (c-got-face-at
4759 (if (setq cfd-re-match (match-end 1))
4760 ;; Matched the end of a token preceding a decl spot.
4761 (progn
4762 (goto-char cfd-re-match)
4763 (1- cfd-re-match))
4764 ;; Matched a token that start a decl spot.
4765 (goto-char (match-beginning 0))
4766 (point))
4767 c-literal-faces)
4768
4769 ;; No match. Finish up and exit the loop.
4770 (setq cfd-re-match cfd-limit)
4771 nil)
4772
4773 ;; Skip out of comments and string literals.
d9e94c22
MS
4774 (while (progn
4775 (goto-char (next-single-property-change
0386b551 4776 (point) 'face nil cfd-limit))
d9e94c22 4777 (and (< (point) cfd-limit)
0386b551
AM
4778 (c-got-face-at (point) c-literal-faces)))))
4779
4780 ;; If we matched at the decl start, we have to back up over the
4781 ;; preceding syntactic ws to set `cfd-match-pos' and to catch
4782 ;; any decl spots in the syntactic ws.
d9e94c22 4783 (unless cfd-re-match
0386b551
AM
4784 (c-backward-syntactic-ws)
4785 (setq cfd-re-match (point))))
d9e94c22
MS
4786
4787 ;; Choose whichever match is closer to the start.
4788 (if (< cfd-re-match cfd-prop-match)
4789 (setq cfd-match-pos cfd-re-match
4790 cfd-re-match nil)
4791 (setq cfd-match-pos cfd-prop-match
4792 cfd-prop-match nil))
4793
4794 (goto-char cfd-match-pos)
4795
4796 (when (< cfd-match-pos cfd-limit)
4797 ;; Skip forward past comments only so we don't skip macros.
4798 (c-forward-comments)
4799 ;; Set the position to continue at. We can avoid going over
4800 ;; the comments skipped above a second time, but it's possible
4801 ;; that the comment skipping has taken us past `cfd-prop-match'
4802 ;; since the property might be used inside comments.
4803 (setq cfd-continue-pos (if cfd-prop-match
4804 (min cfd-prop-match (point))
4805 (point))))))
4806
4807(defun c-find-decl-spots (cfd-limit cfd-decl-re cfd-face-checklist cfd-fun)
0386b551 4808 ;; Call CFD-FUN for each possible spot for a declaration, cast or
a85fd6da
AM
4809 ;; label from the point to CFD-LIMIT.
4810 ;;
9657183b
AM
4811 ;; CFD-FUN is called with point at the start of the spot. It's passed two
4812 ;; arguments: The first is the end position of the token preceding the spot,
4813 ;; or 0 for the implicit match at bob. The second is a flag that is t when
4814 ;; the match is inside a macro. Point should be moved forward by at least
4815 ;; one token.
4816 ;;
4817 ;; If CFD-FUN adds `c-decl-end' properties somewhere below the current spot,
4818 ;; it should return non-nil to ensure that the next search will find them.
0386b551 4819 ;;
a85fd6da 4820 ;; Such a spot is:
0386b551
AM
4821 ;; o The first token after bob.
4822 ;; o The first token after the end of submatch 1 in
4823 ;; `c-decl-prefix-or-start-re' when that submatch matches.
4824 ;; o The start of each `c-decl-prefix-or-start-re' match when
4825 ;; submatch 1 doesn't match.
5a89f0a7 4826 ;; o The first token after the end of each occurrence of the
0386b551
AM
4827 ;; `c-type' text property with the value `c-decl-end', provided
4828 ;; `c-type-decl-end-used' is set.
4829 ;;
4830 ;; Only a spot that match CFD-DECL-RE and whose face is in the
4831 ;; CFD-FACE-CHECKLIST list causes CFD-FUN to be called. The face
4832 ;; check is disabled if CFD-FACE-CHECKLIST is nil.
d9e94c22
MS
4833 ;;
4834 ;; If the match is inside a macro then the buffer is narrowed to the
4835 ;; end of it, so that CFD-FUN can investigate the following tokens
4836 ;; without matching something that begins inside a macro and ends
4837 ;; outside it. It's to avoid this work that the CFD-DECL-RE and
4838 ;; CFD-FACE-CHECKLIST checks exist.
4839 ;;
0386b551
AM
4840 ;; The spots are visited approximately in order from top to bottom.
4841 ;; It's however the positions where `c-decl-prefix-or-start-re'
4842 ;; matches and where `c-decl-end' properties are found that are in
4843 ;; order. Since the spots often are at the following token, they
4844 ;; might be visited out of order insofar as more spots are reported
4845 ;; later on within the syntactic whitespace between the match
4846 ;; positions and their spots.
4847 ;;
4848 ;; It's assumed that comments and strings are fontified in the
d9e94c22
MS
4849 ;; searched range.
4850 ;;
4851 ;; This is mainly used in fontification, and so has an elaborate
4852 ;; cache to handle repeated calls from the same start position; see
4853 ;; the variables above.
4854 ;;
4855 ;; All variables in this function begin with `cfd-' to avoid name
4856 ;; collision with the (dynamically bound) variables used in CFD-FUN.
0386b551
AM
4857 ;;
4858 ;; This function might do hidden buffer changes.
d9e94c22 4859
0386b551
AM
4860 (let ((cfd-start-pos (point))
4861 (cfd-buffer-end (point-max))
4862 ;; The end of the token preceding the decl spot last found
4863 ;; with `c-decl-prefix-or-start-re'. `cfd-limit' if there's
4864 ;; no match.
d9e94c22 4865 cfd-re-match
0386b551
AM
4866 ;; The end position of the last `c-decl-prefix-or-start-re'
4867 ;; match. If this is greater than `cfd-continue-pos', the
4868 ;; next regexp search is started here instead.
4869 (cfd-re-match-end (point-min))
4870 ;; The end of the last `c-decl-end' found by
4871 ;; `c-find-decl-prefix-search'. `cfd-limit' if there's no
4872 ;; match. If searching for the property isn't needed then we
4873 ;; disable it by setting it to `cfd-limit' directly.
d9e94c22 4874 (cfd-prop-match (unless c-type-decl-end-used cfd-limit))
0386b551
AM
4875 ;; The end of the token preceding the decl spot last found by
4876 ;; `c-find-decl-prefix-search'. 0 for the implicit match at
4877 ;; bob. `cfd-limit' if there's no match. In other words,
4878 ;; this is the minimum of `cfd-re-match' and `cfd-prop-match'.
d9e94c22
MS
4879 (cfd-match-pos cfd-limit)
4880 ;; The position to continue searching at.
4881 cfd-continue-pos
4882 ;; The position of the last "real" token we've stopped at.
4883 ;; This can be greater than `cfd-continue-pos' when we get
4884 ;; hits inside macros or at `c-decl-end' positions inside
4885 ;; comments.
4886 (cfd-token-pos 0)
4887 ;; The end position of the last entered macro.
4888 (cfd-macro-end 0))
4889
4890 ;; Initialize by finding a syntactically relevant start position
0386b551
AM
4891 ;; before the point, and do the first `c-decl-prefix-or-start-re'
4892 ;; search unless we're at bob.
d9e94c22 4893
0386b551 4894 (let (start-in-literal start-in-macro syntactic-pos)
d9e94c22
MS
4895 ;; Must back up a bit since we look for the end of the previous
4896 ;; statement or declaration, which is earlier than the first
4897 ;; returned match.
4898
0386b551
AM
4899 (cond
4900 ;; First we need to move to a syntactically relevant position.
4901 ;; Begin by backing out of comment or string literals.
4902 ((and
4903 (when (c-got-face-at (point) c-literal-faces)
4904 ;; Try to use the faces to back up to the start of the
4905 ;; literal. FIXME: What if the point is on a declaration
4906 ;; inside a comment?
4907 (while (and (not (bobp))
4908 (c-got-face-at (1- (point)) c-literal-faces))
4909 (goto-char (previous-single-property-change
4910 (point) 'face nil (point-min))))
4911
4912 ;; XEmacs doesn't fontify the quotes surrounding string
4913 ;; literals.
4914 (and (featurep 'xemacs)
4915 (eq (get-text-property (point) 'face)
4916 'font-lock-string-face)
4917 (not (bobp))
4918 (progn (backward-char)
4919 (not (looking-at c-string-limit-regexp)))
4920 (forward-char))
4921
4922 ;; Don't trust the literal to contain only literal faces
4923 ;; (the font lock package might not have fontified the
4924 ;; start of it at all, for instance) so check that we have
4925 ;; arrived at something that looks like a start or else
4926 ;; resort to `c-literal-limits'.
4927 (unless (looking-at c-literal-start-regexp)
4928 (let ((range (c-literal-limits)))
4929 (if range (goto-char (car range)))))
4930
4931 (setq start-in-literal (point)))
4932
4933 ;; The start is in a literal. If the limit is in the same
4934 ;; one we don't have to find a syntactic position etc. We
4935 ;; only check that if the limit is at or before bonl to save
4936 ;; time; it covers the by far most common case when font-lock
4937 ;; refontifies the current line only.
4938 (<= cfd-limit (c-point 'bonl cfd-start-pos))
4939 (save-excursion
4940 (goto-char cfd-start-pos)
4941 (while (progn
4942 (goto-char (next-single-property-change
4943 (point) 'face nil cfd-limit))
4944 (and (< (point) cfd-limit)
4945 (c-got-face-at (point) c-literal-faces))))
4946 (= (point) cfd-limit)))
4947
4948 ;; Completely inside a literal. Set up variables to trig the
4949 ;; (< cfd-continue-pos cfd-start-pos) case below and it'll
4950 ;; find a suitable start position.
4951 (setq cfd-continue-pos start-in-literal))
4952
4953 ;; Check if the region might be completely inside a macro, to
4954 ;; optimize that like the completely-inside-literal above.
4955 ((save-excursion
4956 (and (= (forward-line 1) 0)
4957 (bolp) ; forward-line has funny behavior at eob.
4958 (>= (point) cfd-limit)
4959 (progn (backward-char)
4960 (eq (char-before) ?\\))))
4961 ;; (Maybe) completely inside a macro. Only need to trig the
4962 ;; (< cfd-continue-pos cfd-start-pos) case below to make it
4963 ;; set things up.
4964 (setq cfd-continue-pos (1- cfd-start-pos)
4965 start-in-macro t))
d9e94c22 4966
0386b551
AM
4967 (t
4968 ;; Back out of any macro so we don't miss any declaration
4969 ;; that could follow after it.
4970 (when (c-beginning-of-macro)
4971 (setq start-in-macro t))
4972
4973 ;; Now we're at a proper syntactically relevant position so we
4974 ;; can use the cache. But first clear it if it applied
4975 ;; further down.
4976 (c-invalidate-find-decl-cache cfd-start-pos)
4977
4978 (setq syntactic-pos (point))
4979 (unless (eq syntactic-pos c-find-decl-syntactic-pos)
4980 ;; Don't have to do this if the cache is relevant here,
4981 ;; typically if the same line is refontified again. If
4982 ;; we're just some syntactic whitespace further down we can
4983 ;; still use the cache to limit the skipping.
4984 (c-backward-syntactic-ws c-find-decl-syntactic-pos))
4985
4986 ;; If we hit `c-find-decl-syntactic-pos' and
4987 ;; `c-find-decl-match-pos' is set then we install the cached
4988 ;; values. If we hit `c-find-decl-syntactic-pos' and
4989 ;; `c-find-decl-match-pos' is nil then we know there's no decl
4990 ;; prefix in the whitespace before `c-find-decl-syntactic-pos'
4991 ;; and so we can continue the search from this point. If we
4992 ;; didn't hit `c-find-decl-syntactic-pos' then we're now in
4993 ;; the right spot to begin searching anyway.
4994 (if (and (eq (point) c-find-decl-syntactic-pos)
4995 c-find-decl-match-pos)
d9e94c22
MS
4996 (setq cfd-match-pos c-find-decl-match-pos
4997 cfd-continue-pos syntactic-pos)
0386b551
AM
4998
4999 (setq c-find-decl-syntactic-pos syntactic-pos)
5000
5001 (when (if (bobp)
5002 ;; Always consider bob a match to get the first
5003 ;; declaration in the file. Do this separately instead of
5004 ;; letting `c-decl-prefix-or-start-re' match bob, so that
5005 ;; regexp always can consume at least one character to
5006 ;; ensure that we won't get stuck in an infinite loop.
5007 (setq cfd-re-match 0)
5008 (backward-char)
5009 (c-beginning-of-current-token)
5010 (< (point) cfd-limit))
5011 ;; Do an initial search now. In the bob case above it's
5012 ;; only done to search for a `c-decl-end' spot.
5013 (c-find-decl-prefix-search))
5014
5015 (setq c-find-decl-match-pos (and (< cfd-match-pos cfd-start-pos)
5016 cfd-match-pos)))))
5017
5018 ;; Advance `cfd-continue-pos' if it's before the start position.
5019 ;; The closest continue position that might have effect at or
5020 ;; after the start depends on what we started in. This also
5021 ;; finds a suitable start position in the special cases when the
5022 ;; region is completely within a literal or macro.
5023 (when (and cfd-continue-pos (< cfd-continue-pos cfd-start-pos))
5024
5025 (cond
5026 (start-in-macro
5027 ;; If we're in a macro then it's the closest preceding token
5028 ;; in the macro. Check this before `start-in-literal',
5029 ;; since if we're inside a literal in a macro, the preceding
5030 ;; token is earlier than any `c-decl-end' spot inside the
5031 ;; literal (comment).
5032 (goto-char (or start-in-literal cfd-start-pos))
5033 ;; The only syntactic ws in macros are comments.
d9e94c22 5034 (c-backward-comments)
0386b551
AM
5035 (backward-char)
5036 (c-beginning-of-current-token))
5037
5038 (start-in-literal
5039 ;; If we're in a comment it can only be the closest
5040 ;; preceding `c-decl-end' position within that comment, if
5041 ;; any. Go back to the beginning of such a property so that
5042 ;; `c-find-decl-prefix-search' will find the end of it.
5043 ;; (Can't stop at the end and install it directly on
5044 ;; `cfd-prop-match' since that variable might be cleared
5045 ;; after `cfd-fun' below.)
5046 ;;
5047 ;; Note that if the literal is a string then the property
5048 ;; search will simply skip to the beginning of it right
5049 ;; away.
5050 (if (not c-type-decl-end-used)
5051 (goto-char start-in-literal)
5052 (goto-char cfd-start-pos)
5053 (while (progn
5054 (goto-char (previous-single-property-change
5055 (point) 'c-type nil start-in-literal))
5056 (and (> (point) start-in-literal)
5057 (not (eq (c-get-char-property (point) 'c-type)
5058 'c-decl-end))))))
5059
5060 (when (= (point) start-in-literal)
5061 ;; Didn't find any property inside the comment, so we can
5062 ;; skip it entirely. (This won't skip past a string, but
5063 ;; that'll be handled quickly by the next
5064 ;; `c-find-decl-prefix-search' anyway.)
5065 (c-forward-single-comment)
5066 (if (> (point) cfd-limit)
5067 (goto-char cfd-limit))))
d9e94c22 5068
0386b551
AM
5069 (t
5070 ;; If we started in normal code, the only match that might
5071 ;; apply before the start is what we already got in
5072 ;; `cfd-match-pos' so we can continue at the start position.
5073 ;; (Note that we don't get here if the first match is below
5074 ;; it.)
5075 (goto-char cfd-start-pos)))
5076
5077 ;; Delete found matches if they are before our new continue
5078 ;; position, so that `c-find-decl-prefix-search' won't back up
5079 ;; to them later on.
5080 (setq cfd-continue-pos (point))
5081 (when (and cfd-re-match (< cfd-re-match cfd-continue-pos))
5082 (setq cfd-re-match nil))
5083 (when (and cfd-prop-match (< cfd-prop-match cfd-continue-pos))
5084 (setq cfd-prop-match nil)))
5085
5086 (if syntactic-pos
5087 ;; This is the normal case and we got a proper syntactic
5088 ;; position. If there's a match then it's always outside
5089 ;; macros and comments, so advance to the next token and set
5090 ;; `cfd-token-pos'. The loop below will later go back using
5091 ;; `cfd-continue-pos' to fix declarations inside the
5092 ;; syntactic ws.
5093 (when (and cfd-match-pos (< cfd-match-pos syntactic-pos))
5094 (goto-char syntactic-pos)
5095 (c-forward-syntactic-ws)
5096 (and cfd-continue-pos
5097 (< cfd-continue-pos (point))
5098 (setq cfd-token-pos (point))))
5099
5100 ;; Have one of the special cases when the region is completely
5101 ;; within a literal or macro. `cfd-continue-pos' is set to a
5102 ;; good start position for the search, so do it.
5103 (c-find-decl-prefix-search)))
d9e94c22 5104
51c9af45 5105 ;; Now loop. Round what? (ACM, 2006/7/5). We already got the first match.
d9e94c22
MS
5106
5107 (while (progn
5108 (while (and
5109 (< cfd-match-pos cfd-limit)
5110
5111 (or
5112 ;; Kludge to filter out matches on the "<" that
5113 ;; aren't open parens, for the sake of languages
5114 ;; that got `c-recognize-<>-arglists' set.
5115 (and (eq (char-before cfd-match-pos) ?<)
5116 (not (c-get-char-property (1- cfd-match-pos)
5117 'syntax-table)))
5118
5119 ;; If `cfd-continue-pos' is less or equal to
5120 ;; `cfd-token-pos', we've got a hit inside a macro
5121 ;; that's in the syntactic whitespace before the last
5122 ;; "real" declaration we've checked. If they're equal
5123 ;; we've arrived at the declaration a second time, so
5124 ;; there's nothing to do.
5125 (= cfd-continue-pos cfd-token-pos)
5126
5127 (progn
5128 ;; If `cfd-continue-pos' is less than `cfd-token-pos'
5129 ;; we're still searching for declarations embedded in
5130 ;; the syntactic whitespace. In that case we need
5131 ;; only to skip comments and not macros, since they
5132 ;; can't be nested, and that's already been done in
5133 ;; `c-find-decl-prefix-search'.
5134 (when (> cfd-continue-pos cfd-token-pos)
5135 (c-forward-syntactic-ws)
5136 (setq cfd-token-pos (point)))
5137
5138 ;; Continue if the following token fails the
5139 ;; CFD-DECL-RE and CFD-FACE-CHECKLIST checks.
5140 (when (or (>= (point) cfd-limit)
5141 (not (looking-at cfd-decl-re))
5142 (and cfd-face-checklist
5143 (not (c-got-face-at
5144 (point) cfd-face-checklist))))
5145 (goto-char cfd-continue-pos)
5146 t)))
5147
5148 (< (point) cfd-limit))
5149 (c-find-decl-prefix-search))
5150
5151 (< (point) cfd-limit))
5152
0386b551
AM
5153 (when (and
5154 (>= (point) cfd-start-pos)
d9e94c22 5155
0386b551
AM
5156 (progn
5157 ;; Narrow to the end of the macro if we got a hit inside
5158 ;; one, to avoid recognizing things that start inside the
5159 ;; macro and end outside it.
5160 (when (> cfd-match-pos cfd-macro-end)
5161 ;; Not in the same macro as in the previous round.
5162 (save-excursion
5163 (goto-char cfd-match-pos)
5164 (setq cfd-macro-end
5165 (if (save-excursion (and (c-beginning-of-macro)
5166 (< (point) cfd-match-pos)))
5167 (progn (c-end-of-macro)
5168 (point))
5169 0))))
5170
5171 (if (zerop cfd-macro-end)
5172 t
5173 (if (> cfd-macro-end (point))
5174 (progn (narrow-to-region (point-min) cfd-macro-end)
5175 t)
5176 ;; The matched token was the last thing in the macro,
5177 ;; so the whole match is bogus.
5178 (setq cfd-macro-end 0)
5179 nil))))
d9e94c22
MS
5180
5181 (c-debug-put-decl-spot-faces cfd-match-pos (point))
0386b551
AM
5182 (if (funcall cfd-fun cfd-match-pos (/= cfd-macro-end 0))
5183 (setq cfd-prop-match nil))
d9e94c22
MS
5184
5185 (when (/= cfd-macro-end 0)
e1dbe924 5186 ;; Restore limits if we did macro narrowing above.
d9e94c22
MS
5187 (narrow-to-region (point-min) cfd-buffer-end)))
5188
5189 (goto-char cfd-continue-pos)
5190 (if (= cfd-continue-pos cfd-limit)
5191 (setq cfd-match-pos cfd-limit)
9657183b
AM
5192 (c-find-decl-prefix-search))))) ; Moves point, sets cfd-continue-pos,
5193 ; cfd-match-pos, etc.
d9e94c22
MS
5194
5195\f
5196;; A cache for found types.
5197
5198;; Buffer local variable that contains an obarray with the types we've
5199;; found. If a declaration is recognized somewhere we record the
5200;; fully qualified identifier in it to recognize it as a type
5201;; elsewhere in the file too. This is not accurate since we do not
5202;; bother with the scoping rules of the languages, but in practice the
5203;; same name is seldom used as both a type and something else in a
5204;; file, and we only use this as a last resort in ambiguous cases (see
0386b551
AM
5205;; `c-forward-decl-or-cast-1').
5206;;
580fba94
AM
5207;; Not every type need be in this cache. However, things which have
5208;; ceased to be types must be removed from it.
5209;;
0386b551
AM
5210;; Template types in C++ are added here too but with the template
5211;; arglist replaced with "<>" in references or "<" for the one in the
5212;; primary type. E.g. the type "Foo<A,B>::Bar<C>" is stored as
5213;; "Foo<>::Bar<". This avoids storing very long strings (since C++
5214;; template specs can be fairly sized programs in themselves) and
5215;; improves the hit ratio (it's a type regardless of the template
5216;; args; it's just not the same type, but we're only interested in
5217;; recognizing types, not telling distinct types apart). Note that
5218;; template types in references are added here too; from the example
5219;; above there will also be an entry "Foo<".
d9e94c22
MS
5220(defvar c-found-types nil)
5221(make-variable-buffer-local 'c-found-types)
5222
5223(defsubst c-clear-found-types ()
5224 ;; Clears `c-found-types'.
d9e94c22
MS
5225 (setq c-found-types (make-vector 53 0)))
5226
5227(defun c-add-type (from to)
5228 ;; Add the given region as a type in `c-found-types'. If the region
5229 ;; doesn't match an existing type but there is a type which is equal
5230 ;; to the given one except that the last character is missing, then
5231 ;; the shorter type is removed. That's done to avoid adding all
5232 ;; prefixes of a type as it's being entered and font locked. This
5233 ;; doesn't cover cases like when characters are removed from a type
5234 ;; or added in the middle. We'd need the position of point when the
5235 ;; font locking is invoked to solve this well.
0386b551
AM
5236 ;;
5237 ;; This function might do hidden buffer changes.
5238 (let ((type (c-syntactic-content from to c-recognize-<>-arglists)))
5239 (unless (intern-soft type c-found-types)
5240 (unintern (substring type 0 -1) c-found-types)
5241 (intern type c-found-types))))
d9e94c22 5242
580fba94
AM
5243(defun c-unfind-type (name)
5244 ;; Remove the "NAME" from c-found-types, if present.
5245 (unintern name c-found-types))
5246
d9e94c22
MS
5247(defsubst c-check-type (from to)
5248 ;; Return non-nil if the given region contains a type in
5249 ;; `c-found-types'.
0386b551
AM
5250 ;;
5251 ;; This function might do hidden buffer changes.
5252 (intern-soft (c-syntactic-content from to c-recognize-<>-arglists)
5253 c-found-types))
d9e94c22
MS
5254
5255(defun c-list-found-types ()
5256 ;; Return all the types in `c-found-types' as a sorted list of
5257 ;; strings.
5258 (let (type-list)
5259 (mapatoms (lambda (type)
5260 (setq type-list (cons (symbol-name type)
5261 type-list)))
5262 c-found-types)
5263 (sort type-list 'string-lessp)))
a66cd3ee 5264
2f42c75f
DN
5265;; Shut up the byte compiler.
5266(defvar c-maybe-stale-found-type)
5267
580fba94
AM
5268(defun c-trim-found-types (beg end old-len)
5269 ;; An after change function which, in conjunction with the info in
5270 ;; c-maybe-stale-found-type (set in c-before-change), removes a type
5271 ;; from `c-found-types', should this type have become stale. For
5272 ;; example, this happens to "foo" when "foo \n bar();" becomes
5273 ;; "foo(); \n bar();". Such stale types, if not removed, foul up
5274 ;; the fontification.
b414f371 5275 ;;
580fba94
AM
5276 ;; Have we, perhaps, added non-ws characters to the front/back of a found
5277 ;; type?
5278 (when (> end beg)
5279 (save-excursion
5280 (when (< end (point-max))
5281 (goto-char end)
5282 (if (and (c-beginning-of-current-token) ; only moves when we started in the middle
5283 (progn (goto-char end)
5284 (c-end-of-current-token)))
5285 (c-unfind-type (buffer-substring-no-properties
5286 end (point)))))
5287 (when (> beg (point-min))
5288 (goto-char beg)
5289 (if (and (c-end-of-current-token) ; only moves when we started in the middle
5290 (progn (goto-char beg)
5291 (c-beginning-of-current-token)))
5292 (c-unfind-type (buffer-substring-no-properties
5293 (point) beg))))))
b414f371 5294
580fba94
AM
5295 (if c-maybe-stale-found-type ; e.g. (c-decl-id-start "foo" 97 107 " (* ooka) " "o")
5296 (cond
5297 ;; Changing the amount of (already existing) whitespace - don't do anything.
5298 ((and (c-partial-ws-p beg end)
5299 (or (= beg end) ; removal of WS
580fba94
AM
5300 (string-match "^[ \t\n\r\f\v]*$" (nth 5 c-maybe-stale-found-type)))))
5301
5302 ;; The syntactic relationship which defined a "found type" has been
5303 ;; destroyed.
5304 ((eq (car c-maybe-stale-found-type) 'c-decl-id-start)
5305 (c-unfind-type (cadr c-maybe-stale-found-type)))
5306;; ((eq (car c-maybe-stale-found-type) 'c-decl-type-start) FIXME!!!
5307 )))
5308
d9e94c22 5309\f
dd969a56
AM
5310;; Setting and removing syntax properties on < and > in languages (C++
5311;; and Java) where they can be template/generic delimiters as well as
5312;; their normal meaning of "less/greater than".
5313
5314;; Normally, < and > have syntax 'punctuation'. When they are found to
5315;; be delimiters, they are marked as such with the category properties
5316;; c-<-as-paren-syntax, c->-as-paren-syntax respectively.
5317
5318;; STRATEGY:
5319;;
5320;; It is impossible to determine with certainty whether a <..> pair in
5321;; C++ is two comparison operators or is template delimiters, unless
5322;; one duplicates a lot of a C++ compiler. For example, the following
5323;; code fragment:
5324;;
5325;; foo (a < b, c > d) ;
5326;;
5327;; could be a function call with two integer parameters (each a
5328;; relational expression), or it could be a constructor for class foo
5329;; taking one parameter d of templated type "a < b, c >". They are
5330;; somewhat easier to distinguish in Java.
5331;;
5332;; The strategy now (2010-01) adopted is to mark and unmark < and
5333;; > IN MATCHING PAIRS ONLY. [Previously, they were marked
5334;; individually when their context so indicated. This gave rise to
53964682 5335;; intractable problems when one of a matching pair was deleted, or
dd969a56
AM
5336;; pulled into a literal.]
5337;;
5338;; At each buffer change, the syntax-table properties are removed in a
5339;; before-change function and reapplied, when needed, in an
5340;; after-change function. It is far more important that the
5341;; properties get removed when they they are spurious than that they
5342;; be present when wanted.
5343;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
5344(defun c-clear-<-pair-props (&optional pos)
5345 ;; POS (default point) is at a < character. If it is marked with
5346 ;; open paren syntax-table text property, remove the property,
5347 ;; together with the close paren property on the matching > (if
5348 ;; any).
5349 (save-excursion
5350 (if pos
5351 (goto-char pos)
5352 (setq pos (point)))
5353 (when (equal (c-get-char-property (point) 'syntax-table)
5354 c-<-as-paren-syntax)
5355 (with-syntax-table c-no-parens-syntax-table ; ignore unbalanced [,{,(,..
5356 (c-go-list-forward))
5357 (when (equal (c-get-char-property (1- (point)) 'syntax-table)
5358 c->-as-paren-syntax) ; should always be true.
8a249abc
AM
5359 (c-clear-char-property (1- (point)) 'category))
5360 (c-clear-char-property pos 'category))))
dd969a56
AM
5361
5362(defun c-clear->-pair-props (&optional pos)
5363 ;; POS (default point) is at a > character. If it is marked with
5364 ;; close paren syntax-table property, remove the property, together
5365 ;; with the open paren property on the matching < (if any).
5366 (save-excursion
5367 (if pos
5368 (goto-char pos)
5369 (setq pos (point)))
5370 (when (equal (c-get-char-property (point) 'syntax-table)
5371 c->-as-paren-syntax)
5372 (with-syntax-table c-no-parens-syntax-table ; ignore unbalanced [,{,(,..
5373 (c-go-up-list-backward))
5374 (when (equal (c-get-char-property (point) 'syntax-table)
5375 c-<-as-paren-syntax) ; should always be true.
8a249abc
AM
5376 (c-clear-char-property (point) 'category))
5377 (c-clear-char-property pos 'category))))
dd969a56
AM
5378
5379(defun c-clear-<>-pair-props (&optional pos)
5380 ;; POS (default point) is at a < or > character. If it has an
5381 ;; open/close paren syntax-table property, remove this property both
5382 ;; from the current character and its partner (which will also be
5383 ;; thusly marked).
5384 (cond
5385 ((eq (char-after) ?\<)
5386 (c-clear-<-pair-props pos))
5387 ((eq (char-after) ?\>)
5388 (c-clear->-pair-props pos))
5389 (t (c-benign-error
5390 "c-clear-<>-pair-props called from wrong position"))))
5391
5392(defun c-clear-<-pair-props-if-match-after (lim &optional pos)
5393 ;; POS (default point) is at a < character. If it is both marked
5394 ;; with open/close paren syntax-table property, and has a matching >
5395 ;; (also marked) which is after LIM, remove the property both from
43a91810
AM
5396 ;; the current > and its partner. Return t when this happens, nil
5397 ;; when it doesn't.
dd969a56
AM
5398 (save-excursion
5399 (if pos
5400 (goto-char pos)
5401 (setq pos (point)))
5402 (when (equal (c-get-char-property (point) 'syntax-table)
5403 c-<-as-paren-syntax)
5404 (with-syntax-table c-no-parens-syntax-table ; ignore unbalanced [,{,(,..
5405 (c-go-list-forward))
5406 (when (and (>= (point) lim)
5407 (equal (c-get-char-property (1- (point)) 'syntax-table)
5408 c->-as-paren-syntax)) ; should always be true.
5409 (c-unmark-<->-as-paren (1- (point)))
43a91810
AM
5410 (c-unmark-<->-as-paren pos))
5411 t)))
dd969a56
AM
5412
5413(defun c-clear->-pair-props-if-match-before (lim &optional pos)
5414 ;; POS (default point) is at a > character. If it is both marked
5415 ;; with open/close paren syntax-table property, and has a matching <
5416 ;; (also marked) which is before LIM, remove the property both from
43a91810
AM
5417 ;; the current < and its partner. Return t when this happens, nil
5418 ;; when it doesn't.
dd969a56
AM
5419 (save-excursion
5420 (if pos
5421 (goto-char pos)
5422 (setq pos (point)))
5423 (when (equal (c-get-char-property (point) 'syntax-table)
5424 c->-as-paren-syntax)
5425 (with-syntax-table c-no-parens-syntax-table ; ignore unbalanced [,{,(,..
5426 (c-go-up-list-backward))
5427 (when (and (<= (point) lim)
5428 (equal (c-get-char-property (point) 'syntax-table)
5429 c-<-as-paren-syntax)) ; should always be true.
5430 (c-unmark-<->-as-paren (point))
43a91810
AM
5431 (c-unmark-<->-as-paren pos))
5432 t)))
dd969a56 5433
f3b554af
GM
5434;; Set by c-common-init in cc-mode.el.
5435(defvar c-new-BEG)
5436(defvar c-new-END)
5437
dd969a56
AM
5438(defun c-before-change-check-<>-operators (beg end)
5439 ;; Unmark certain pairs of "< .... >" which are currently marked as
5440 ;; template/generic delimiters. (This marking is via syntax-table
5441 ;; text properties).
5442 ;;
5443 ;; These pairs are those which are in the current "statement" (i.e.,
5444 ;; the region between the {, }, or ; before BEG and the one after
5445 ;; END), and which enclose any part of the interval (BEG END).
5446 ;;
5447 ;; Note that in C++ (?and Java), template/generic parens cannot
5448 ;; enclose a brace or semicolon, so we use these as bounds on the
5449 ;; region we must work on.
5450 ;;
5451 ;; This function is called from before-change-functions (via
5452 ;; c-get-state-before-change-functions). Thus the buffer is widened,
5453 ;; and point is undefined, both at entry and exit.
5454 ;;
5455 ;; FIXME!!! This routine ignores the possibility of macros entirely.
5456 ;; 2010-01-29.
5457 (save-excursion
5458 (let ((beg-lit-limits (progn (goto-char beg) (c-literal-limits)))
43a91810
AM
5459 (end-lit-limits (progn (goto-char end) (c-literal-limits)))
5460 new-beg new-end need-new-beg need-new-end)
dd969a56
AM
5461 ;; Locate the barrier before the changed region
5462 (goto-char (if beg-lit-limits (car beg-lit-limits) beg))
0de3da9f 5463 (c-syntactic-skip-backward "^;{}" (c-determine-limit 512))
43a91810 5464 (setq new-beg (point))
dd969a56
AM
5465
5466 ;; Remove the syntax-table properties from each pertinent <...> pair.
5467 ;; Firsly, the ones with the < before beg and > after beg.
5468 (while (c-search-forward-char-property 'category 'c-<-as-paren-syntax beg)
43a91810
AM
5469 (if (c-clear-<-pair-props-if-match-after beg (1- (point)))
5470 (setq need-new-beg t)))
dd969a56
AM
5471
5472 ;; Locate the barrier after END.
5473 (goto-char (if end-lit-limits (cdr end-lit-limits) end))
0de3da9f 5474 (c-syntactic-re-search-forward "[;{}]" (c-determine-+ve-limit 512) 'end)
43a91810 5475 (setq new-end (point))
dd969a56
AM
5476
5477 ;; Remove syntax-table properties from the remaining pertinent <...>
5478 ;; pairs, those with a > after end and < before end.
5479 (while (c-search-backward-char-property 'category 'c->-as-paren-syntax end)
43a91810
AM
5480 (if (c-clear->-pair-props-if-match-before end)
5481 (setq need-new-end t)))
5482
5483 ;; Extend the fontification region, if needed.
5484 (when need-new-beg
5485 (goto-char new-beg)
5486 (c-forward-syntactic-ws)
5487 (and (< (point) c-new-BEG) (setq c-new-BEG (point))))
5488
5489 (when need-new-end
5490 (and (> new-end c-new-END) (setq c-new-END new-end))))))
dd969a56
AM
5491
5492
d9e94c22 5493
0386b551
AM
5494(defun c-after-change-check-<>-operators (beg end)
5495 ;; This is called from `after-change-functions' when
5496 ;; c-recognize-<>-arglists' is set. It ensures that no "<" or ">"
5497 ;; chars with paren syntax become part of another operator like "<<"
5498 ;; or ">=".
5499 ;;
5500 ;; This function might do hidden buffer changes.
5501
cb694ab7
AM
5502 (save-excursion
5503 (goto-char beg)
5504 (when (or (looking-at "[<>]")
5505 (< (skip-chars-backward "<>") 0))
5506
0386b551 5507 (goto-char beg)
cb694ab7
AM
5508 (c-beginning-of-current-token)
5509 (when (and (< (point) beg)
5510 (looking-at c-<>-multichar-token-regexp)
5511 (< beg (setq beg (match-end 0))))
5512 (while (progn (skip-chars-forward "^<>" beg)
5513 (< (point) beg))
dd969a56 5514 (c-clear-<>-pair-props)
cb694ab7
AM
5515 (forward-char))))
5516
5517 (when (< beg end)
5518 (goto-char end)
0386b551
AM
5519 (when (or (looking-at "[<>]")
5520 (< (skip-chars-backward "<>") 0))
5521
cb694ab7 5522 (goto-char end)
0386b551 5523 (c-beginning-of-current-token)
cb694ab7 5524 (when (and (< (point) end)
0386b551 5525 (looking-at c-<>-multichar-token-regexp)
cb694ab7
AM
5526 (< end (setq end (match-end 0))))
5527 (while (progn (skip-chars-forward "^<>" end)
5528 (< (point) end))
dd969a56 5529 (c-clear-<>-pair-props)
cb694ab7 5530 (forward-char)))))))
d9e94c22 5531
dd969a56
AM
5532
5533\f
5534;; Handling of small scale constructs like types and names.
5535
d9e94c22
MS
5536;; Dynamically bound variable that instructs `c-forward-type' to also
5537;; treat possible types (i.e. those that it normally returns 'maybe or
5538;; 'found for) as actual types (and always return 'found for them).
5539;; This means that it records them in `c-record-type-identifiers' if
5540;; that is set, and that it adds them to `c-found-types'.
5541(defvar c-promote-possible-types nil)
5542
0386b551
AM
5543;; Dynamically bound variable that instructs `c-forward-<>-arglist' to
5544;; mark up successfully parsed arglists with paren syntax properties on
5545;; the surrounding angle brackets and with `c-<>-arg-sep' in the
5546;; `c-type' property of each argument separating comma.
5547;;
5548;; Setting this variable also makes `c-forward-<>-arglist' recurse into
5549;; all arglists for side effects (i.e. recording types), otherwise it
5550;; exploits any existing paren syntax properties to quickly jump to the
5551;; end of already parsed arglists.
5552;;
5553;; Marking up the arglists is not the default since doing that correctly
5554;; depends on a proper value for `c-restricted-<>-arglists'.
5555(defvar c-parse-and-markup-<>-arglists nil)
5556
d9e94c22 5557;; Dynamically bound variable that instructs `c-forward-<>-arglist' to
037558bf
MS
5558;; not accept arglists that contain binary operators.
5559;;
5560;; This is primarily used to handle C++ template arglists. C++
5561;; disambiguates them by checking whether the preceding name is a
5562;; template or not. We can't do that, so we assume it is a template
5563;; if it can be parsed as one. That usually works well since
5564;; comparison expressions on the forms "a < b > c" or "a < b, c > d"
5565;; in almost all cases would be pointless.
5566;;
5567;; However, in function arglists, e.g. in "foo (a < b, c > d)", we
5568;; should let the comma separate the function arguments instead. And
5569;; in a context where the value of the expression is taken, e.g. in
5570;; "if (a < b || c > d)", it's probably not a template.
5571(defvar c-restricted-<>-arglists nil)
d9e94c22 5572
0386b551
AM
5573;; Dynamically bound variables that instructs
5574;; `c-forward-keyword-clause', `c-forward-<>-arglist',
5575;; `c-forward-name', `c-forward-type', `c-forward-decl-or-cast-1', and
5576;; `c-forward-label' to record the ranges of all the type and
5577;; reference identifiers they encounter. They will build lists on
5578;; these variables where each element is a cons of the buffer
5579;; positions surrounding each identifier. This recording is only
5580;; activated when `c-record-type-identifiers' is non-nil.
d9e94c22
MS
5581;;
5582;; All known types that can't be identifiers are recorded, and also
5583;; other possible types if `c-promote-possible-types' is set.
5584;; Recording is however disabled inside angle bracket arglists that
5585;; are encountered inside names and other angle bracket arglists.
0386b551 5586;; Such occurrences are taken care of by `c-font-lock-<>-arglists'
d9e94c22
MS
5587;; instead.
5588;;
5589;; Only the names in C++ template style references (e.g. "tmpl" in
5590;; "tmpl<a,b>::foo") are recorded as references, other references
5591;; aren't handled here.
0386b551
AM
5592;;
5593;; `c-forward-label' records the label identifier(s) on
5594;; `c-record-ref-identifiers'.
d9e94c22
MS
5595(defvar c-record-type-identifiers nil)
5596(defvar c-record-ref-identifiers nil)
5597
0386b551
AM
5598;; This variable will receive a cons cell of the range of the last
5599;; single identifier symbol stepped over by `c-forward-name' if it's
5600;; successful. This is the range that should be put on one of the
5601;; record lists above by the caller. It's assigned nil if there's no
5602;; such symbol in the name.
d9e94c22
MS
5603(defvar c-last-identifier-range nil)
5604
5605(defmacro c-record-type-id (range)
5606 (if (eq (car-safe range) 'cons)
5607 ;; Always true.
5608 `(setq c-record-type-identifiers
5609 (cons ,range c-record-type-identifiers))
5610 `(let ((range ,range))
5611 (if range
5612 (setq c-record-type-identifiers
5613 (cons range c-record-type-identifiers))))))
5614
5615(defmacro c-record-ref-id (range)
5616 (if (eq (car-safe range) 'cons)
5617 ;; Always true.
5618 `(setq c-record-ref-identifiers
5619 (cons ,range c-record-ref-identifiers))
5620 `(let ((range ,range))
5621 (if range
5622 (setq c-record-ref-identifiers
5623 (cons range c-record-ref-identifiers))))))
5624
5625;; Dynamically bound variable that instructs `c-forward-type' to
5626;; record the ranges of types that only are found. Behaves otherwise
5627;; like `c-record-type-identifiers'.
5628(defvar c-record-found-types nil)
5629
5630(defmacro c-forward-keyword-prefixed-id (type)
5631 ;; Used internally in `c-forward-keyword-clause' to move forward
5632 ;; over a type (if TYPE is 'type) or a name (otherwise) which
5633 ;; possibly is prefixed by keywords and their associated clauses.
5634 ;; Try with a type/name first to not trip up on those that begin
5635 ;; with a keyword. Return t if a known or found type is moved
5636 ;; over. The point is clobbered if nil is returned. If range
5637 ;; recording is enabled, the identifier is recorded on as a type
5638 ;; if TYPE is 'type or as a reference if TYPE is 'ref.
0386b551
AM
5639 ;;
5640 ;; This macro might do hidden buffer changes.
d9e94c22
MS
5641 `(let (res)
5642 (while (if (setq res ,(if (eq type 'type)
5643 `(c-forward-type)
5644 `(c-forward-name)))
5645 nil
5646 (and (looking-at c-keywords-regexp)
0386b551 5647 (c-forward-keyword-clause 1))))
d9e94c22
MS
5648 (when (memq res '(t known found prefix))
5649 ,(when (eq type 'ref)
5650 `(when c-record-type-identifiers
5651 (c-record-ref-id c-last-identifier-range)))
5652 t)))
5653
0386b551 5654(defmacro c-forward-id-comma-list (type update-safe-pos)
d9e94c22
MS
5655 ;; Used internally in `c-forward-keyword-clause' to move forward
5656 ;; over a comma separated list of types or names using
5657 ;; `c-forward-keyword-prefixed-id'.
0386b551
AM
5658 ;;
5659 ;; This macro might do hidden buffer changes.
d9e94c22 5660 `(while (and (progn
0386b551
AM
5661 ,(when update-safe-pos
5662 `(setq safe-pos (point)))
d9e94c22
MS
5663 (eq (char-after) ?,))
5664 (progn
5665 (forward-char)
5666 (c-forward-syntactic-ws)
5667 (c-forward-keyword-prefixed-id ,type)))))
5668
0386b551
AM
5669(defun c-forward-keyword-clause (match)
5670 ;; Submatch MATCH in the current match data is assumed to surround a
5671 ;; token. If it's a keyword, move over it and any immediately
5672 ;; following clauses associated with it, stopping at the start of
5673 ;; the next token. t is returned in that case, otherwise the point
d9e94c22
MS
5674 ;; stays and nil is returned. The kind of clauses that are
5675 ;; recognized are those specified by `c-type-list-kwds',
5676 ;; `c-ref-list-kwds', `c-colon-type-list-kwds',
5677 ;; `c-paren-nontype-kwds', `c-paren-type-kwds', `c-<>-type-kwds',
5678 ;; and `c-<>-arglist-kwds'.
0386b551
AM
5679 ;;
5680 ;; This function records identifier ranges on
5681 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
5682 ;; `c-record-type-identifiers' is non-nil.
5683 ;;
5684 ;; Note that for `c-colon-type-list-kwds', which doesn't necessary
5685 ;; apply directly after the keyword, the type list is moved over
5686 ;; only when there is no unaccounted token before it (i.e. a token
5687 ;; that isn't moved over due to some other keyword list). The
5688 ;; identifier ranges in the list are still recorded if that should
5689 ;; be done, though.
5690 ;;
5691 ;; This function might do hidden buffer changes.
5692
5693 (let ((kwd-sym (c-keyword-sym (match-string match))) safe-pos pos
5694 ;; The call to `c-forward-<>-arglist' below is made after
5695 ;; `c-<>-sexp-kwds' keywords, so we're certain they actually
5696 ;; are angle bracket arglists and `c-restricted-<>-arglists'
5697 ;; should therefore be nil.
5698 (c-parse-and-markup-<>-arglists t)
5699 c-restricted-<>-arglists)
d9e94c22 5700
d9e94c22 5701 (when kwd-sym
0386b551 5702 (goto-char (match-end match))
d9e94c22
MS
5703 (c-forward-syntactic-ws)
5704 (setq safe-pos (point))
a66cd3ee 5705
d9e94c22
MS
5706 (cond
5707 ((and (c-keyword-member kwd-sym 'c-type-list-kwds)
5708 (c-forward-keyword-prefixed-id type))
5709 ;; There's a type directly after a keyword in `c-type-list-kwds'.
0386b551 5710 (c-forward-id-comma-list type t))
d9e94c22
MS
5711
5712 ((and (c-keyword-member kwd-sym 'c-ref-list-kwds)
5713 (c-forward-keyword-prefixed-id ref))
5714 ;; There's a name directly after a keyword in `c-ref-list-kwds'.
0386b551 5715 (c-forward-id-comma-list ref t))
d9e94c22
MS
5716
5717 ((and (c-keyword-member kwd-sym 'c-paren-any-kwds)
5718 (eq (char-after) ?\())
5719 ;; There's an open paren after a keyword in `c-paren-any-kwds'.
5720
5721 (forward-char)
5722 (when (and (setq pos (c-up-list-forward))
5723 (eq (char-before pos) ?\)))
5724 (when (and c-record-type-identifiers
5725 (c-keyword-member kwd-sym 'c-paren-type-kwds))
5726 ;; Use `c-forward-type' on every identifier we can find
5727 ;; inside the paren, to record the types.
5728 (while (c-syntactic-re-search-forward c-symbol-start pos t)
5729 (goto-char (match-beginning 0))
5730 (unless (c-forward-type)
5731 (looking-at c-symbol-key) ; Always matches.
5732 (goto-char (match-end 0)))))
5733
5734 (goto-char pos)
5735 (c-forward-syntactic-ws)
5736 (setq safe-pos (point))))
5737
5738 ((and (c-keyword-member kwd-sym 'c-<>-sexp-kwds)
5739 (eq (char-after) ?<)
0386b551 5740 (c-forward-<>-arglist (c-keyword-member kwd-sym 'c-<>-type-kwds)))
d9e94c22
MS
5741 (c-forward-syntactic-ws)
5742 (setq safe-pos (point)))
5743
5744 ((and (c-keyword-member kwd-sym 'c-nonsymbol-sexp-kwds)
449a2b0d
MS
5745 (not (looking-at c-symbol-start))
5746 (c-safe (c-forward-sexp) t))
d9e94c22
MS
5747 (c-forward-syntactic-ws)
5748 (setq safe-pos (point))))
5749
0386b551
AM
5750 (when (c-keyword-member kwd-sym 'c-colon-type-list-kwds)
5751 (if (eq (char-after) ?:)
5752 ;; If we are at the colon already, we move over the type
5753 ;; list after it.
5754 (progn
5755 (forward-char)
5756 (c-forward-syntactic-ws)
5757 (when (c-forward-keyword-prefixed-id type)
5758 (c-forward-id-comma-list type t)))
5759 ;; Not at the colon, so stop here. But the identifier
5760 ;; ranges in the type list later on should still be
5761 ;; recorded.
5762 (and c-record-type-identifiers
5763 (progn
5764 ;; If a keyword matched both one of the types above and
5765 ;; this one, we match `c-colon-type-list-re' after the
5766 ;; clause matched above.
5767 (goto-char safe-pos)
5768 (looking-at c-colon-type-list-re))
5769 (progn
5770 (goto-char (match-end 0))
5771 (c-forward-syntactic-ws)
5772 (c-forward-keyword-prefixed-id type))
5773 ;; There's a type after the `c-colon-type-list-re' match
5774 ;; after a keyword in `c-colon-type-list-kwds'.
5775 (c-forward-id-comma-list type nil))))
d9e94c22
MS
5776
5777 (goto-char safe-pos)
5778 t)))
5779
f3b554af
GM
5780;; cc-mode requires cc-fonts.
5781(declare-function c-fontify-recorded-types-and-refs "cc-fonts" ())
5782
0386b551
AM
5783(defun c-forward-<>-arglist (all-types)
5784 ;; The point is assumed to be at a "<". Try to treat it as the open
b4dc7d98 5785 ;; paren of an angle bracket arglist and move forward to the
0386b551
AM
5786 ;; corresponding ">". If successful, the point is left after the
5787 ;; ">" and t is returned, otherwise the point isn't moved and nil is
d9e94c22
MS
5788 ;; returned. If ALL-TYPES is t then all encountered arguments in
5789 ;; the arglist that might be types are treated as found types.
5790 ;;
0386b551
AM
5791 ;; The variable `c-parse-and-markup-<>-arglists' controls how this
5792 ;; function handles text properties on the angle brackets and argument
5793 ;; separating commas.
d9e94c22 5794 ;;
0386b551
AM
5795 ;; `c-restricted-<>-arglists' controls how lenient the template
5796 ;; arglist recognition should be.
5797 ;;
5798 ;; This function records identifier ranges on
5799 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
5800 ;; `c-record-type-identifiers' is non-nil.
5801 ;;
5802 ;; This function might do hidden buffer changes.
d9e94c22
MS
5803
5804 (let ((start (point))
5805 ;; If `c-record-type-identifiers' is set then activate
5806 ;; recording of any found types that constitute an argument in
5807 ;; the arglist.
abfc152b 5808 (c-record-found-types (if c-record-type-identifiers t)))
d9e94c22
MS
5809 (if (catch 'angle-bracket-arglist-escape
5810 (setq c-record-found-types
0386b551 5811 (c-forward-<>-arglist-recur all-types)))
d9e94c22
MS
5812 (progn
5813 (when (consp c-record-found-types)
5814 (setq c-record-type-identifiers
5815 ;; `nconc' doesn't mind that the tail of
5816 ;; `c-record-found-types' is t.
5817 (nconc c-record-found-types c-record-type-identifiers)))
452ea855 5818 (if (c-major-mode-is 'java-mode) (c-fontify-recorded-types-and-refs))
d9e94c22
MS
5819 t)
5820
5821 (goto-char start)
a66cd3ee 5822 nil)))
785eecbb 5823
0386b551 5824(defun c-forward-<>-arglist-recur (all-types)
d9e94c22 5825 ;; Recursive part of `c-forward-<>-arglist'.
0386b551
AM
5826 ;;
5827 ;; This function might do hidden buffer changes.
d9e94c22
MS
5828
5829 (let ((start (point)) res pos tmp
5830 ;; Cover this so that any recorded found type ranges are
5831 ;; automatically lost if it turns out to not be an angle
5832 ;; bracket arglist. It's propagated through the return value
5833 ;; on successful completion.
5834 (c-record-found-types c-record-found-types)
5835 ;; List that collects the positions after the argument
5836 ;; separating ',' in the arglist.
5837 arg-start-pos)
0386b551
AM
5838 ;; If the '<' has paren open syntax then we've marked it as an angle
5839 ;; bracket arglist before, so skip to the end.
5840 (if (and (not c-parse-and-markup-<>-arglists)
5841 (c-get-char-property (point) 'syntax-table))
5842
5843 (progn
5844 (forward-char)
5845 (if (and (c-go-up-list-forward)
5846 (eq (char-before) ?>))
5847 t
0386b551
AM
5848 ;; Got unmatched paren angle brackets. We don't clear the paren
5849 ;; syntax properties and retry, on the basis that it's very
5850 ;; unlikely that paren angle brackets become operators by code
5851 ;; manipulation. It's far more likely that it doesn't match due
5852 ;; to narrowing or some temporary change.
5853 (goto-char start)
5854 nil))
d9e94c22 5855
abfc152b 5856 (forward-char) ; Forward over the opening '<'.
452ea855 5857
d9e94c22 5858 (unless (looking-at c-<-op-cont-regexp)
abfc152b
AM
5859 ;; go forward one non-alphanumeric character (group) per iteration of
5860 ;; this loop.
22c3ce97 5861 (while (and
d9e94c22 5862 (progn
22c3ce97
AM
5863 (c-forward-syntactic-ws)
5864 (let ((orig-record-found-types c-record-found-types))
5865 (when (or (and c-record-type-identifiers all-types)
5866 (c-major-mode-is 'java-mode))
5867 ;; All encountered identifiers are types, so set the
5868 ;; promote flag and parse the type.
5869 (progn
5870 (c-forward-syntactic-ws)
5871 (if (looking-at "\\?")
5872 (forward-char)
5873 (when (looking-at c-identifier-start)
5874 (let ((c-promote-possible-types t)
5875 (c-record-found-types t))
5876 (c-forward-type))))
5877
5878 (c-forward-syntactic-ws)
5879
5880 (when (or (looking-at "extends")
5881 (looking-at "super"))
5882 (forward-word)
5883 (c-forward-syntactic-ws)
5884 (let ((c-promote-possible-types t)
5885 (c-record-found-types t))
5886 (c-forward-type)
5887 (c-forward-syntactic-ws))))))
5888
abfc152b 5889 (setq pos (point)) ; e.g. first token inside the '<'
22c3ce97 5890
a4ee83cc
AM
5891 ;; Note: These regexps exploit the match order in \| so
5892 ;; that "<>" is matched by "<" rather than "[^>:-]>".
5893 (c-syntactic-re-search-forward
5894 ;; Stop on ',', '|', '&', '+' and '-' to catch
5895 ;; common binary operators that could be between
5896 ;; two comparison expressions "a<b" and "c>d".
5897 "[<;{},|+&-]\\|[>)]"
5898 nil t t))
22c3ce97
AM
5899
5900 (cond
5901 ((eq (char-before) ?>)
d9e94c22
MS
5902 ;; Either an operator starting with '>' or the end of
5903 ;; the angle bracket arglist.
5904
0386b551 5905 (if (looking-at c->-op-cont-regexp)
d9e94c22 5906 (progn
0386b551
AM
5907 (goto-char (match-end 0))
5908 t) ; Continue the loop.
d9e94c22 5909
0386b551
AM
5910 ;; The angle bracket arglist is finished.
5911 (when c-parse-and-markup-<>-arglists
d9e94c22 5912 (while arg-start-pos
0386b551
AM
5913 (c-put-c-type-property (1- (car arg-start-pos))
5914 'c-<>-arg-sep)
d9e94c22
MS
5915 (setq arg-start-pos (cdr arg-start-pos)))
5916 (c-mark-<-as-paren start)
0386b551
AM
5917 (c-mark->-as-paren (1- (point))))
5918 (setq res t)
5919 nil)) ; Exit the loop.
d9e94c22
MS
5920
5921 ((eq (char-before) ?<)
5922 ;; Either an operator starting with '<' or a nested arglist.
d9e94c22
MS
5923 (setq pos (point))
5924 (let (id-start id-end subres keyword-match)
abfc152b
AM
5925 (cond
5926 ;; The '<' begins a multi-char operator.
5927 ((looking-at c-<-op-cont-regexp)
5928 (setq tmp (match-end 0))
5929 (goto-char (match-end 0)))
5930 ;; We're at a nested <.....>
5931 ((progn
5932 (setq tmp pos)
5933 (backward-char) ; to the '<'
5934 (and
5935 (save-excursion
5936 ;; There's always an identifier before an angle
5937 ;; bracket arglist, or a keyword in `c-<>-type-kwds'
5938 ;; or `c-<>-arglist-kwds'.
5939 (c-backward-syntactic-ws)
5940 (setq id-end (point))
5941 (c-simple-skip-symbol-backward)
5942 (when (or (setq keyword-match
5943 (looking-at c-opt-<>-sexp-key))
5944 (not (looking-at c-keywords-regexp)))
5945 (setq id-start (point))))
5946 (setq subres
5947 (let ((c-promote-possible-types t)
5948 (c-record-found-types t))
5949 (c-forward-<>-arglist-recur
5950 (and keyword-match
5951 (c-keyword-member
5952 (c-keyword-sym (match-string 1))
5953 'c-<>-type-kwds)))))))
d9e94c22
MS
5954
5955 ;; It was an angle bracket arglist.
5956 (setq c-record-found-types subres)
5957
5958 ;; Record the identifier before the template as a type
5959 ;; or reference depending on whether the arglist is last
5960 ;; in a qualified identifier.
5961 (when (and c-record-type-identifiers
5962 (not keyword-match))
5963 (if (and c-opt-identifier-concat-key
5964 (progn
5965 (c-forward-syntactic-ws)
5966 (looking-at c-opt-identifier-concat-key)))
5967 (c-record-ref-id (cons id-start id-end))
abfc152b
AM
5968 (c-record-type-id (cons id-start id-end)))))
5969
5970 ;; At a "less than" operator.
5971 (t
5972 (forward-char)
5973 )))
5974 t) ; carry on looping.
22c3ce97
AM
5975
5976 ((and (not c-restricted-<>-arglists)
5977 (or (and (eq (char-before) ?&)
5978 (not (eq (char-after) ?&)))
5979 (eq (char-before) ?,)))
5980 ;; Just another argument. Record the position. The
5981 ;; type check stuff that made us stop at it is at
5982 ;; the top of the loop.
d9e94c22
MS
5983 (setq arg-start-pos (cons (point) arg-start-pos)))
5984
5985 (t
5986 ;; Got a character that can't be in an angle bracket
5987 ;; arglist argument. Abort using `throw', since
5988 ;; it's useless to try to find a surrounding arglist
5989 ;; if we're nested.
5990 (throw 'angle-bracket-arglist-escape nil))))))
d9e94c22
MS
5991 (if res
5992 (or c-record-found-types t)))))
5993
0386b551
AM
5994(defun c-backward-<>-arglist (all-types &optional limit)
5995 ;; The point is assumed to be directly after a ">". Try to treat it
5996 ;; as the close paren of an angle bracket arglist and move back to
5997 ;; the corresponding "<". If successful, the point is left at
5998 ;; the "<" and t is returned, otherwise the point isn't moved and
5999 ;; nil is returned. ALL-TYPES is passed on to
6000 ;; `c-forward-<>-arglist'.
6001 ;;
6002 ;; If the optional LIMIT is given, it bounds the backward search.
6003 ;; It's then assumed to be at a syntactically relevant position.
6004 ;;
6005 ;; This is a wrapper around `c-forward-<>-arglist'. See that
6006 ;; function for more details.
6007
6008 (let ((start (point)))
6009 (backward-char)
6010 (if (and (not c-parse-and-markup-<>-arglists)
6011 (c-get-char-property (point) 'syntax-table))
6012
6013 (if (and (c-go-up-list-backward)
6014 (eq (char-after) ?<))
6015 t
6016 ;; See corresponding note in `c-forward-<>-arglist'.
6017 (goto-char start)
6018 nil)
6019
51c9af45 6020 (while (progn
0386b551
AM
6021 (c-syntactic-skip-backward "^<;{}" limit t)
6022
51c9af45
AM
6023 (and
6024 (if (eq (char-before) ?<)
6025 t
6026 ;; Stopped at bob or a char that isn't allowed in an
6027 ;; arglist, so we've failed.
6028 (goto-char start)
6029 nil)
0386b551 6030
51c9af45
AM
6031 (if (> (point)
6032 (progn (c-beginning-of-current-token)
6033 (point)))
6034 ;; If we moved then the "<" was part of some
6035 ;; multicharacter token.
6036 t
0386b551 6037
51c9af45
AM
6038 (backward-char)
6039 (let ((beg-pos (point)))
6040 (if (c-forward-<>-arglist all-types)
6041 (cond ((= (point) start)
6042 ;; Matched the arglist. Break the while.
6043 (goto-char beg-pos)
6044 nil)
6045 ((> (point) start)
6046 ;; We started from a non-paren ">" inside an
6047 ;; arglist.
6048 (goto-char start)
6049 nil)
6050 (t
6051 ;; Matched a shorter arglist. Can be a nested
6052 ;; one so continue looking.
6053 (goto-char beg-pos)
6054 t))
6055 t))))))
0386b551
AM
6056
6057 (/= (point) start))))
6058
d9e94c22
MS
6059(defun c-forward-name ()
6060 ;; Move forward over a complete name if at the beginning of one,
e15f8aaa
AM
6061 ;; stopping at the next following token. A keyword, as such,
6062 ;; doesn't count as a name. If the point is not at something that
6063 ;; is recognized as a name then it stays put.
6064 ;;
6065 ;; A name could be something as simple as "foo" in C or something as
d9e94c22
MS
6066 ;; complex as "X<Y<class A<int>::B, BIT_MAX >> b>, ::operator<> ::
6067 ;; Z<(a>b)> :: operator const X<&foo>::T Q::G<unsigned short
6068 ;; int>::*volatile const" in C++ (this function is actually little
6069 ;; more than a `looking-at' call in all modes except those that,
e15f8aaa
AM
6070 ;; like C++, have `c-recognize-<>-arglists' set).
6071 ;;
6072 ;; Return
6073 ;; o - nil if no name is found;
6074 ;; o - 'template if it's an identifier ending with an angle bracket
6075 ;; arglist;
6076 ;; o - 'operator of it's an operator identifier;
6077 ;; o - t if it's some other kind of name.
0386b551
AM
6078 ;;
6079 ;; This function records identifier ranges on
6080 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
6081 ;; `c-record-type-identifiers' is non-nil.
6082 ;;
6083 ;; This function might do hidden buffer changes.
d9e94c22 6084
0386b551 6085 (let ((pos (point)) (start (point)) res id-start id-end
d9e94c22
MS
6086 ;; Turn off `c-promote-possible-types' here since we might
6087 ;; call `c-forward-<>-arglist' and we don't want it to promote
6088 ;; every suspect thing in the arglist to a type. We're
6089 ;; typically called from `c-forward-type' in this case, and
6090 ;; the caller only wants the top level type that it finds to
6091 ;; be promoted.
6092 c-promote-possible-types)
6093 (while
6094 (and
6095 (looking-at c-identifier-key)
6096
6097 (progn
6098 ;; Check for keyword. We go to the last symbol in
6099 ;; `c-identifier-key' first.
0386b551
AM
6100 (goto-char (setq id-end (match-end 0)))
6101 (c-simple-skip-symbol-backward)
6102 (setq id-start (point))
d9e94c22
MS
6103
6104 (if (looking-at c-keywords-regexp)
6105 (when (and (c-major-mode-is 'c++-mode)
6106 (looking-at
6107 (cc-eval-when-compile
6108 (concat "\\(operator\\|\\(template\\)\\)"
6109 "\\(" (c-lang-const c-nonsymbol-key c++)
6110 "\\|$\\)")))
6111 (if (match-beginning 2)
6112 ;; "template" is only valid inside an
6113 ;; identifier if preceded by "::".
6114 (save-excursion
6115 (c-backward-syntactic-ws)
6116 (and (c-safe (backward-char 2) t)
6117 (looking-at "::")))
6118 t))
6119
6120 ;; Handle a C++ operator or template identifier.
6121 (goto-char id-end)
6122 (c-forward-syntactic-ws)
6123 (cond ((eq (char-before id-end) ?e)
6124 ;; Got "... ::template".
6125 (let ((subres (c-forward-name)))
6126 (when subres
6127 (setq pos (point)
6128 res subres))))
6129
6130 ((looking-at c-identifier-start)
6131 ;; Got a cast operator.
6132 (when (c-forward-type)
6133 (setq pos (point)
6134 res 'operator)
6135 ;; Now we should match a sequence of either
6136 ;; '*', '&' or a name followed by ":: *",
6137 ;; where each can be followed by a sequence
6138 ;; of `c-opt-type-modifier-key'.
6139 (while (cond ((looking-at "[*&]")
6140 (goto-char (match-end 0))
6141 t)
6142 ((looking-at c-identifier-start)
6143 (and (c-forward-name)
6144 (looking-at "::")
6145 (progn
6146 (goto-char (match-end 0))
6147 (c-forward-syntactic-ws)
6148 (eq (char-after) ?*))
6149 (progn
6150 (forward-char)
6151 t))))
6152 (while (progn
6153 (c-forward-syntactic-ws)
6154 (setq pos (point))
6155 (looking-at c-opt-type-modifier-key))
6156 (goto-char (match-end 1))))))
6157
6158 ((looking-at c-overloadable-operators-regexp)
6159 ;; Got some other operator.
0386b551
AM
6160 (setq c-last-identifier-range
6161 (cons (point) (match-end 0)))
d9e94c22
MS
6162 (goto-char (match-end 0))
6163 (c-forward-syntactic-ws)
6164 (setq pos (point)
6165 res 'operator)))
6166
6167 nil)
6168
0386b551
AM
6169 ;; `id-start' is equal to `id-end' if we've jumped over
6170 ;; an identifier that doesn't end with a symbol token.
6171 ;; That can occur e.g. for Java import directives on the
6172 ;; form "foo.bar.*".
6173 (when (and id-start (/= id-start id-end))
d9e94c22
MS
6174 (setq c-last-identifier-range
6175 (cons id-start id-end)))
6176 (goto-char id-end)
6177 (c-forward-syntactic-ws)
6178 (setq pos (point)
6179 res t)))
6180
6181 (progn
6182 (goto-char pos)
6183 (when (or c-opt-identifier-concat-key
6184 c-recognize-<>-arglists)
6185
6186 (cond
6187 ((and c-opt-identifier-concat-key
6188 (looking-at c-opt-identifier-concat-key))
6189 ;; Got a concatenated identifier. This handles the
6190 ;; cases with tricky syntactic whitespace that aren't
6191 ;; covered in `c-identifier-key'.
6192 (goto-char (match-end 0))
6193 (c-forward-syntactic-ws)
6194 t)
6195
6196 ((and c-recognize-<>-arglists
6197 (eq (char-after) ?<))
6198 ;; Maybe an angle bracket arglist.
452ea855
AM
6199 (when (let ((c-record-type-identifiers t)
6200 (c-record-found-types t))
0386b551
AM
6201 (c-forward-<>-arglist nil))
6202
6203 (c-add-type start (1+ pos))
d9e94c22 6204 (c-forward-syntactic-ws)
0386b551
AM
6205 (setq pos (point)
6206 c-last-identifier-range nil)
6207
d9e94c22
MS
6208 (if (and c-opt-identifier-concat-key
6209 (looking-at c-opt-identifier-concat-key))
0386b551 6210
d9e94c22
MS
6211 ;; Continue if there's an identifier concatenation
6212 ;; operator after the template argument.
6213 (progn
0386b551
AM
6214 (when (and c-record-type-identifiers id-start)
6215 (c-record-ref-id (cons id-start id-end)))
d9e94c22
MS
6216 (forward-char 2)
6217 (c-forward-syntactic-ws)
6218 t)
0386b551
AM
6219
6220 (when (and c-record-type-identifiers id-start)
6221 (c-record-type-id (cons id-start id-end)))
d9e94c22
MS
6222 (setq res 'template)
6223 nil)))
6224 )))))
6225
6226 (goto-char pos)
6227 res))
6228
e15f8aaa 6229(defun c-forward-type (&optional brace-block-too)
d9e94c22 6230 ;; Move forward over a type spec if at the beginning of one,
e15f8aaa
AM
6231 ;; stopping at the next following token. The keyword "typedef"
6232 ;; isn't part of a type spec here.
6233 ;;
6234 ;; BRACE-BLOCK-TOO, when non-nil, means move over the brace block in
6235 ;; constructs like "struct foo {...} bar ;" or "struct {...} bar;".
6236 ;; The current (2009-03-10) intention is to convert all uses of
6237 ;; `c-forward-type' to call with this parameter set, then to
6238 ;; eliminate it.
6239 ;;
6240 ;; Return
6241 ;; o - t if it's a known type that can't be a name or other
6242 ;; expression;
6243 ;; o - 'known if it's an otherwise known type (according to
6244 ;; `*-font-lock-extra-types');
6245 ;; o - 'prefix if it's a known prefix of a type;
6246 ;; o - 'found if it's a type that matches one in `c-found-types';
53964682 6247 ;; o - 'maybe if it's an identifier that might be a type; or
e15f8aaa
AM
6248 ;; o - nil if it can't be a type (the point isn't moved then).
6249 ;;
6250 ;; The point is assumed to be at the beginning of a token.
d9e94c22
MS
6251 ;;
6252 ;; Note that this function doesn't skip past the brace definition
6253 ;; that might be considered part of the type, e.g.
6254 ;; "enum {a, b, c} foo".
0386b551
AM
6255 ;;
6256 ;; This function records identifier ranges on
6257 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
6258 ;; `c-record-type-identifiers' is non-nil.
6259 ;;
6260 ;; This function might do hidden buffer changes.
0e5cf2b8
AM
6261 (when (and c-recognize-<>-arglists
6262 (looking-at "<"))
452ea855
AM
6263 (c-forward-<>-arglist t)
6264 (c-forward-syntactic-ws))
0386b551
AM
6265
6266 (let ((start (point)) pos res name-res id-start id-end id-range)
d9e94c22
MS
6267
6268 ;; Skip leading type modifiers. If any are found we know it's a
6269 ;; prefix of a type.
e15f8aaa 6270 (when c-opt-type-modifier-key ; e.g. "const" "volatile", but NOT "typedef"
d9e94c22
MS
6271 (while (looking-at c-opt-type-modifier-key)
6272 (goto-char (match-end 1))
6273 (c-forward-syntactic-ws)
6274 (setq res 'prefix)))
6275
6276 (cond
e15f8aaa
AM
6277 ((looking-at c-type-prefix-key) ; e.g. "struct", "class", but NOT
6278 ; "typedef".
d9e94c22
MS
6279 (goto-char (match-end 1))
6280 (c-forward-syntactic-ws)
6281 (setq pos (point))
e15f8aaa
AM
6282
6283 (setq name-res (c-forward-name))
6284 (setq res (not (null name-res)))
6285 (when (eq name-res t)
6286 ;; In many languages the name can be used without the
6287 ;; prefix, so we add it to `c-found-types'.
6288 (c-add-type pos (point))
6289 (when (and c-record-type-identifiers
6290 c-last-identifier-range)
6291 (c-record-type-id c-last-identifier-range)))
6292 (when (and brace-block-too
6293 (memq res '(t nil))
6294 (eq (char-after) ?\{)
6295 (save-excursion
6296 (c-safe
6297 (progn (c-forward-sexp)
6298 (c-forward-syntactic-ws)
6299 (setq pos (point))))))
6300 (goto-char pos)
6301 (setq res t))
6302 (unless res (goto-char start))) ; invalid syntax
d9e94c22
MS
6303
6304 ((progn
6305 (setq pos nil)
6306 (if (looking-at c-identifier-start)
6307 (save-excursion
6308 (setq id-start (point)
0386b551
AM
6309 name-res (c-forward-name))
6310 (when name-res
d9e94c22
MS
6311 (setq id-end (point)
6312 id-range c-last-identifier-range))))
6313 (and (cond ((looking-at c-primitive-type-key)
6314 (setq res t))
6315 ((c-with-syntax-table c-identifier-syntax-table
6316 (looking-at c-known-type-key))
6317 (setq res 'known)))
6318 (or (not id-end)
6319 (>= (save-excursion
6320 (save-match-data
6321 (goto-char (match-end 1))
6322 (c-forward-syntactic-ws)
6323 (setq pos (point))))
6324 id-end)
6325 (setq res nil))))
6326 ;; Looking at a primitive or known type identifier. We've
6327 ;; checked for a name first so that we don't go here if the
6328 ;; known type match only is a prefix of another name.
6329
6330 (setq id-end (match-end 1))
6331
6332 (when (and c-record-type-identifiers
6333 (or c-promote-possible-types (eq res t)))
6334 (c-record-type-id (cons (match-beginning 1) (match-end 1))))
6335
6336 (if (and c-opt-type-component-key
6337 (save-match-data
6338 (looking-at c-opt-type-component-key)))
6339 ;; There might be more keywords for the type.
6340 (let (safe-pos)
0386b551 6341 (c-forward-keyword-clause 1)
d9e94c22
MS
6342 (while (progn
6343 (setq safe-pos (point))
6344 (looking-at c-opt-type-component-key))
6345 (when (and c-record-type-identifiers
6346 (looking-at c-primitive-type-key))
6347 (c-record-type-id (cons (match-beginning 1)
6348 (match-end 1))))
0386b551 6349 (c-forward-keyword-clause 1))
d9e94c22
MS
6350 (if (looking-at c-primitive-type-key)
6351 (progn
6352 (when c-record-type-identifiers
6353 (c-record-type-id (cons (match-beginning 1)
6354 (match-end 1))))
0386b551 6355 (c-forward-keyword-clause 1)
d9e94c22
MS
6356 (setq res t))
6357 (goto-char safe-pos)
6358 (setq res 'prefix)))
0386b551 6359 (unless (save-match-data (c-forward-keyword-clause 1))
d9e94c22
MS
6360 (if pos
6361 (goto-char pos)
6362 (goto-char (match-end 1))
6363 (c-forward-syntactic-ws)))))
6364
0386b551
AM
6365 (name-res
6366 (cond ((eq name-res t)
d9e94c22
MS
6367 ;; A normal identifier.
6368 (goto-char id-end)
6369 (if (or res c-promote-possible-types)
6370 (progn
6371 (c-add-type id-start id-end)
0386b551 6372 (when (and c-record-type-identifiers id-range)
d9e94c22
MS
6373 (c-record-type-id id-range))
6374 (unless res
6375 (setq res 'found)))
6376 (setq res (if (c-check-type id-start id-end)
6377 ;; It's an identifier that has been used as
6378 ;; a type somewhere else.
6379 'found
6380 ;; It's an identifier that might be a type.
6381 'maybe))))
0386b551 6382 ((eq name-res 'template)
d9e94c22
MS
6383 ;; A template is a type.
6384 (goto-char id-end)
6385 (setq res t))
6386 (t
6387 ;; Otherwise it's an operator identifier, which is not a type.
6388 (goto-char start)
6389 (setq res nil)))))
6390
6391 (when res
e15f8aaa 6392 ;; Skip trailing type modifiers. If any are found we know it's
d9e94c22
MS
6393 ;; a type.
6394 (when c-opt-type-modifier-key
e15f8aaa 6395 (while (looking-at c-opt-type-modifier-key) ; e.g. "const", "volatile"
d9e94c22
MS
6396 (goto-char (match-end 1))
6397 (c-forward-syntactic-ws)
6398 (setq res t)))
d9e94c22
MS
6399 ;; Step over any type suffix operator. Do not let the existence
6400 ;; of these alter the classification of the found type, since
6401 ;; these operators typically are allowed in normal expressions
6402 ;; too.
6403 (when c-opt-type-suffix-key
6404 (while (looking-at c-opt-type-suffix-key)
6405 (goto-char (match-end 1))
6406 (c-forward-syntactic-ws)))
6407
e15f8aaa 6408 (when c-opt-type-concat-key ; Only/mainly for pike.
0386b551
AM
6409 ;; Look for a trailing operator that concatenates the type
6410 ;; with a following one, and if so step past that one through
6411 ;; a recursive call. Note that we don't record concatenated
6412 ;; types in `c-found-types' - it's the component types that
6413 ;; are recorded when appropriate.
d9e94c22
MS
6414 (setq pos (point))
6415 (let* ((c-promote-possible-types (or (memq res '(t known))
6416 c-promote-possible-types))
6417 ;; If we can't promote then set `c-record-found-types' so that
6418 ;; we can merge in the types from the second part afterwards if
6419 ;; it turns out to be a known type there.
6420 (c-record-found-types (and c-record-type-identifiers
0386b551
AM
6421 (not c-promote-possible-types)))
6422 subres)
d9e94c22
MS
6423 (if (and (looking-at c-opt-type-concat-key)
6424
6425 (progn
6426 (goto-char (match-end 1))
6427 (c-forward-syntactic-ws)
0386b551 6428 (setq subres (c-forward-type))))
d9e94c22
MS
6429
6430 (progn
6431 ;; If either operand certainly is a type then both are, but we
6432 ;; don't let the existence of the operator itself promote two
6433 ;; uncertain types to a certain one.
6434 (cond ((eq res t))
0386b551
AM
6435 ((eq subres t)
6436 (unless (eq name-res 'template)
6437 (c-add-type id-start id-end))
6438 (when (and c-record-type-identifiers id-range)
d9e94c22
MS
6439 (c-record-type-id id-range))
6440 (setq res t))
2a15eb73 6441 ((eq res 'known))
0386b551 6442 ((eq subres 'known)
2a15eb73 6443 (setq res 'known))
d9e94c22 6444 ((eq res 'found))
0386b551 6445 ((eq subres 'found)
d9e94c22
MS
6446 (setq res 'found))
6447 (t
6448 (setq res 'maybe)))
6449
6450 (when (and (eq res t)
6451 (consp c-record-found-types))
6452 ;; Merge in the ranges of any types found by the second
6453 ;; `c-forward-type'.
6454 (setq c-record-type-identifiers
6455 ;; `nconc' doesn't mind that the tail of
6456 ;; `c-record-found-types' is t.
6457 (nconc c-record-found-types
6458 c-record-type-identifiers))))
6459
6460 (goto-char pos))))
6461
6462 (when (and c-record-found-types (memq res '(known found)) id-range)
6463 (setq c-record-found-types
6464 (cons id-range c-record-found-types))))
6465
6466 ;;(message "c-forward-type %s -> %s: %s" start (point) res)
6467
6468 res))
6469
452ea855
AM
6470(defun c-forward-annotation ()
6471 ;; Used for Java code only at the moment. Assumes point is on the
6472 ;; @, moves forward an annotation. returns nil if there is no
6473 ;; annotation at point.
6474 (and (looking-at "@")
6475 (progn (forward-char) t)
6476 (c-forward-type)
6477 (progn (c-forward-syntactic-ws) t)
6478 (if (looking-at "(")
6479 (c-go-list-forward)
6480 t)))
6481
785eecbb 6482\f
d9e94c22
MS
6483;; Handling of large scale constructs like statements and declarations.
6484
0386b551
AM
6485;; Macro used inside `c-forward-decl-or-cast-1'. It ought to be a
6486;; defsubst or perhaps even a defun, but it contains lots of free
6487;; variables that refer to things inside `c-forward-decl-or-cast-1'.
6488(defmacro c-fdoc-shift-type-backward (&optional short)
6489 ;; `c-forward-decl-or-cast-1' can consume an arbitrary length list
6490 ;; of types when parsing a declaration, which means that it
6491 ;; sometimes consumes the identifier in the declaration as a type.
6492 ;; This is used to "backtrack" and make the last type be treated as
6493 ;; an identifier instead.
6494 `(progn
6495 ,(unless short
6496 ;; These identifiers are bound only in the inner let.
6497 '(setq identifier-type at-type
6498 identifier-start type-start
6499 got-parens nil
6500 got-identifier t
6501 got-suffix t
6502 got-suffix-after-parens id-start
6503 paren-depth 0))
6504
6505 (if (setq at-type (if (eq backup-at-type 'prefix)
6506 t
6507 backup-at-type))
6508 (setq type-start backup-type-start
6509 id-start backup-id-start)
6510 (setq type-start start-pos
6511 id-start start-pos))
6512
6513 ;; When these flags already are set we've found specifiers that
6514 ;; unconditionally signal these attributes - backtracking doesn't
6515 ;; change that. So keep them set in that case.
6516 (or at-type-decl
6517 (setq at-type-decl backup-at-type-decl))
6518 (or maybe-typeless
6519 (setq maybe-typeless backup-maybe-typeless))
6520
6521 ,(unless short
6522 ;; This identifier is bound only in the inner let.
6523 '(setq start id-start))))
6524
6525(defun c-forward-decl-or-cast-1 (preceding-token-end context last-cast-end)
6526 ;; Move forward over a declaration or a cast if at the start of one.
6527 ;; The point is assumed to be at the start of some token. Nil is
6528 ;; returned if no declaration or cast is recognized, and the point
6529 ;; is clobbered in that case.
6530 ;;
6531 ;; If a declaration is parsed:
6532 ;;
6533 ;; The point is left at the first token after the first complete
6534 ;; declarator, if there is one. The return value is a cons where
022d0cf4
AM
6535 ;; the car is the position of the first token in the declarator. (See
6536 ;; below for the cdr.)
0386b551
AM
6537 ;; Some examples:
6538 ;;
6539 ;; void foo (int a, char *b) stuff ...
6540 ;; car ^ ^ point
6541 ;; float (*a)[], b;
6542 ;; car ^ ^ point
6543 ;; unsigned int a = c_style_initializer, b;
6544 ;; car ^ ^ point
6545 ;; unsigned int a (cplusplus_style_initializer), b;
6546 ;; car ^ ^ point (might change)
6547 ;; class Foo : public Bar {}
6548 ;; car ^ ^ point
6549 ;; class PikeClass (int a, string b) stuff ...
6550 ;; car ^ ^ point
6551 ;; enum bool;
6552 ;; car ^ ^ point
6553 ;; enum bool flag;
6554 ;; car ^ ^ point
6555 ;; void cplusplus_function (int x) throw (Bad);
6556 ;; car ^ ^ point
6557 ;; Foo::Foo (int b) : Base (b) {}
6558 ;; car ^ ^ point
91af3942 6559 ;;
e15f8aaa
AM
6560 ;; The cdr of the return value is non-nil when a
6561 ;; `c-typedef-decl-kwds' specifier is found in the declaration.
6562 ;; Specifically it is a dotted pair (A . B) where B is t when a
6563 ;; `c-typedef-kwds' ("typedef") is present, and A is t when some
6564 ;; other `c-typedef-decl-kwds' (e.g. class, struct, enum)
6565 ;; specifier is present. I.e., (some of) the declared
6566 ;; identifier(s) are types.
91af3942 6567 ;;
0386b551
AM
6568 ;; If a cast is parsed:
6569 ;;
6570 ;; The point is left at the first token after the closing paren of
6571 ;; the cast. The return value is `cast'. Note that the start
6572 ;; position must be at the first token inside the cast parenthesis
6573 ;; to recognize it.
6574 ;;
6575 ;; PRECEDING-TOKEN-END is the first position after the preceding
6576 ;; token, i.e. on the other side of the syntactic ws from the point.
6577 ;; Use a value less than or equal to (point-min) if the point is at
6578 ;; the first token in (the visible part of) the buffer.
6579 ;;
6580 ;; CONTEXT is a symbol that describes the context at the point:
022d0cf4 6581 ;; 'decl In a comma-separated declaration context (typically
0386b551
AM
6582 ;; inside a function declaration arglist).
6583 ;; '<> In an angle bracket arglist.
6584 ;; 'arglist Some other type of arglist.
a85fd6da
AM
6585 ;; nil Some other context or unknown context. Includes
6586 ;; within the parens of an if, for, ... construct.
0386b551
AM
6587 ;;
6588 ;; LAST-CAST-END is the first token after the closing paren of a
6589 ;; preceding cast, or nil if none is known. If
6590 ;; `c-forward-decl-or-cast-1' is used in succession, it should be
6591 ;; the position after the closest preceding call where a cast was
6592 ;; matched. In that case it's used to discover chains of casts like
6593 ;; "(a) (b) c".
6594 ;;
6595 ;; This function records identifier ranges on
6596 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
6597 ;; `c-record-type-identifiers' is non-nil.
6598 ;;
6599 ;; This function might do hidden buffer changes.
6600
6601 (let (;; `start-pos' is used below to point to the start of the
6602 ;; first type, i.e. after any leading specifiers. It might
6603 ;; also point at the beginning of the preceding syntactic
6604 ;; whitespace.
6605 (start-pos (point))
6606 ;; Set to the result of `c-forward-type'.
6607 at-type
6608 ;; The position of the first token in what we currently
6609 ;; believe is the type in the declaration or cast, after any
6610 ;; specifiers and their associated clauses.
6611 type-start
6612 ;; The position of the first token in what we currently
6613 ;; believe is the declarator for the first identifier. Set
6614 ;; when the type is found, and moved forward over any
6615 ;; `c-decl-hangon-kwds' and their associated clauses that
6616 ;; occurs after the type.
6617 id-start
6618 ;; These store `at-type', `type-start' and `id-start' of the
6619 ;; identifier before the one in those variables. The previous
6620 ;; identifier might turn out to be the real type in a
6621 ;; declaration if the last one has to be the declarator in it.
6622 ;; If `backup-at-type' is nil then the other variables have
6623 ;; undefined values.
6624 backup-at-type backup-type-start backup-id-start
e15f8aaa
AM
6625 ;; Set if we've found a specifier (apart from "typedef") that makes
6626 ;; the defined identifier(s) types.
0386b551 6627 at-type-decl
e15f8aaa
AM
6628 ;; Set if we've a "typedef" keyword.
6629 at-typedef
0386b551
AM
6630 ;; Set if we've found a specifier that can start a declaration
6631 ;; where there's no type.
6632 maybe-typeless
6633 ;; If a specifier is found that also can be a type prefix,
6634 ;; these flags are set instead of those above. If we need to
6635 ;; back up an identifier, they are copied to the real flag
6636 ;; variables. Thus they only take effect if we fail to
6637 ;; interpret it as a type.
6638 backup-at-type-decl backup-maybe-typeless
6639 ;; Whether we've found a declaration or a cast. We might know
6640 ;; this before we've found the type in it. It's 'ids if we've
6641 ;; found two consecutive identifiers (usually a sure sign, but
6642 ;; we should allow that in labels too), and t if we've found a
6643 ;; specifier keyword (a 100% sure sign).
6644 at-decl-or-cast
6645 ;; Set when we need to back up to parse this as a declaration
6646 ;; but not as a cast.
6647 backup-if-not-cast
6648 ;; For casts, the return position.
6649 cast-end
6650 ;; Save `c-record-type-identifiers' and
6651 ;; `c-record-ref-identifiers' since ranges are recorded
6652 ;; speculatively and should be thrown away if it turns out
6653 ;; that it isn't a declaration or cast.
6654 (save-rec-type-ids c-record-type-identifiers)
6655 (save-rec-ref-ids c-record-ref-identifiers))
6656
452ea855
AM
6657 (while (c-forward-annotation)
6658 (c-forward-syntactic-ws))
6659
0386b551
AM
6660 ;; Check for a type. Unknown symbols are treated as possible
6661 ;; types, but they could also be specifiers disguised through
6662 ;; macros like __INLINE__, so we recognize both types and known
6663 ;; specifiers after them too.
6664 (while
6665 (let* ((start (point)) kwd-sym kwd-clause-end found-type)
6666
6667 ;; Look for a specifier keyword clause.
e3715033
AM
6668 (when (or (looking-at c-prefix-spec-kwds-re)
6669 (and (c-major-mode-is 'java-mode)
6670 (looking-at "@[A-Za-z0-9]+")))
e15f8aaa
AM
6671 (if (looking-at c-typedef-key)
6672 (setq at-typedef t))
0386b551
AM
6673 (setq kwd-sym (c-keyword-sym (match-string 1)))
6674 (save-excursion
6675 (c-forward-keyword-clause 1)
6676 (setq kwd-clause-end (point))))
6677
e15f8aaa 6678 (when (setq found-type (c-forward-type t)) ; brace-block-too
0386b551
AM
6679 ;; Found a known or possible type or a prefix of a known type.
6680
6681 (when at-type
6682 ;; Got two identifiers with nothing but whitespace
6683 ;; between them. That can only happen in declarations.
6684 (setq at-decl-or-cast 'ids)
6685
6686 (when (eq at-type 'found)
6687 ;; If the previous identifier is a found type we
6688 ;; record it as a real one; it might be some sort of
6689 ;; alias for a prefix like "unsigned".
6690 (save-excursion
6691 (goto-char type-start)
6692 (let ((c-promote-possible-types t))
6693 (c-forward-type)))))
6694
6695 (setq backup-at-type at-type
6696 backup-type-start type-start
6697 backup-id-start id-start
6698 at-type found-type
6699 type-start start
6700 id-start (point)
6701 ;; The previous ambiguous specifier/type turned out
6702 ;; to be a type since we've parsed another one after
6703 ;; it, so clear these backup flags.
6704 backup-at-type-decl nil
6705 backup-maybe-typeless nil))
6706
6707 (if kwd-sym
6708 (progn
6709 ;; Handle known specifier keywords and
6710 ;; `c-decl-hangon-kwds' which can occur after known
6711 ;; types.
6712
6713 (if (c-keyword-member kwd-sym 'c-decl-hangon-kwds)
6714 ;; It's a hang-on keyword that can occur anywhere.
6715 (progn
6716 (setq at-decl-or-cast t)
6717 (if at-type
6718 ;; Move the identifier start position if
6719 ;; we've passed a type.
6720 (setq id-start kwd-clause-end)
6721 ;; Otherwise treat this as a specifier and
6722 ;; move the fallback position.
6723 (setq start-pos kwd-clause-end))
6724 (goto-char kwd-clause-end))
6725
6726 ;; It's an ordinary specifier so we know that
6727 ;; anything before this can't be the type.
6728 (setq backup-at-type nil
6729 start-pos kwd-clause-end)
6730
6731 (if found-type
6732 ;; It's ambiguous whether this keyword is a
6733 ;; specifier or a type prefix, so set the backup
6734 ;; flags. (It's assumed that `c-forward-type'
6735 ;; moved further than `c-forward-keyword-clause'.)
6736 (progn
6737 (when (c-keyword-member kwd-sym 'c-typedef-decl-kwds)
6738 (setq backup-at-type-decl t))
6739 (when (c-keyword-member kwd-sym 'c-typeless-decl-kwds)
6740 (setq backup-maybe-typeless t)))
6741
6742 (when (c-keyword-member kwd-sym 'c-typedef-decl-kwds)
e15f8aaa
AM
6743 ;; This test only happens after we've scanned a type.
6744 ;; So, with valid syntax, kwd-sym can't be 'typedef.
0386b551
AM
6745 (setq at-type-decl t))
6746 (when (c-keyword-member kwd-sym 'c-typeless-decl-kwds)
6747 (setq maybe-typeless t))
6748
ee7683eb 6749 ;; Haven't matched a type so it's an unambiguous
0386b551
AM
6750 ;; specifier keyword and we know we're in a
6751 ;; declaration.
6752 (setq at-decl-or-cast t)
6753
6754 (goto-char kwd-clause-end))))
6755
6756 ;; If the type isn't known we continue so that we'll jump
6757 ;; over all specifiers and type identifiers. The reason
6758 ;; to do this for a known type prefix is to make things
6759 ;; like "unsigned INT16" work.
6760 (and found-type (not (eq found-type t))))))
6761
6762 (cond
6763 ((eq at-type t)
6764 ;; If a known type was found, we still need to skip over any
6765 ;; hangon keyword clauses after it. Otherwise it has already
6766 ;; been done in the loop above.
6767 (while (looking-at c-decl-hangon-key)
6768 (c-forward-keyword-clause 1))
6769 (setq id-start (point)))
6770
6771 ((eq at-type 'prefix)
6772 ;; A prefix type is itself a primitive type when it's not
6773 ;; followed by another type.
6774 (setq at-type t))
6775
6776 ((not at-type)
6777 ;; Got no type but set things up to continue anyway to handle
6778 ;; the various cases when a declaration doesn't start with a
6779 ;; type.
6780 (setq id-start start-pos))
6781
6782 ((and (eq at-type 'maybe)
6783 (c-major-mode-is 'c++-mode))
6784 ;; If it's C++ then check if the last "type" ends on the form
6785 ;; "foo::foo" or "foo::~foo", i.e. if it's the name of a
6786 ;; (con|de)structor.
6787 (save-excursion
6788 (let (name end-2 end-1)
6789 (goto-char id-start)
6790 (c-backward-syntactic-ws)
6791 (setq end-2 (point))
6792 (when (and
6793 (c-simple-skip-symbol-backward)
6794 (progn
6795 (setq name
6796 (buffer-substring-no-properties (point) end-2))
6797 ;; Cheating in the handling of syntactic ws below.
6798 (< (skip-chars-backward ":~ \t\n\r\v\f") 0))
6799 (progn
6800 (setq end-1 (point))
6801 (c-simple-skip-symbol-backward))
6802 (>= (point) type-start)
6803 (equal (buffer-substring-no-properties (point) end-1)
6804 name))
6805 ;; It is a (con|de)structor name. In that case the
6806 ;; declaration is typeless so zap out any preceding
6807 ;; identifier(s) that we might have taken as types.
6808 (goto-char type-start)
6809 (setq at-type nil
6810 backup-at-type nil
6811 id-start type-start))))))
6812
6813 ;; Check for and step over a type decl expression after the thing
6814 ;; that is or might be a type. This can't be skipped since we
6815 ;; need the correct end position of the declarator for
6816 ;; `max-type-decl-end-*'.
6817 (let ((start (point)) (paren-depth 0) pos
6818 ;; True if there's a non-open-paren match of
6819 ;; `c-type-decl-prefix-key'.
6820 got-prefix
6821 ;; True if the declarator is surrounded by a parenthesis pair.
6822 got-parens
6823 ;; True if there is an identifier in the declarator.
6824 got-identifier
6825 ;; True if there's a non-close-paren match of
6826 ;; `c-type-decl-suffix-key'.
6827 got-suffix
6828 ;; True if there's a prefix match outside the outermost
6829 ;; paren pair that surrounds the declarator.
6830 got-prefix-before-parens
1379f2c5 6831 ;; True if there's a suffix match outside the outermost
0386b551
AM
6832 ;; paren pair that surrounds the declarator. The value is
6833 ;; the position of the first suffix match.
6834 got-suffix-after-parens
6835 ;; True if we've parsed the type decl to a token that is
6836 ;; known to end declarations in this context.
6837 at-decl-end
6838 ;; The earlier values of `at-type' and `type-start' if we've
6839 ;; shifted the type backwards.
6840 identifier-type identifier-start
6841 ;; If `c-parse-and-markup-<>-arglists' is set we need to
6842 ;; turn it off during the name skipping below to avoid
6843 ;; getting `c-type' properties that might be bogus. That
6844 ;; can happen since we don't know if
6845 ;; `c-restricted-<>-arglists' will be correct inside the
6846 ;; arglist paren that gets entered.
6847 c-parse-and-markup-<>-arglists)
6848
6849 (goto-char id-start)
6850
6851 ;; Skip over type decl prefix operators. (Note similar code in
6852 ;; `c-font-lock-declarators'.)
6853 (while (and (looking-at c-type-decl-prefix-key)
6854 (if (and (c-major-mode-is 'c++-mode)
05842630 6855 (match-beginning 3))
0386b551
AM
6856 ;; If the second submatch matches in C++ then
6857 ;; we're looking at an identifier that's a
6858 ;; prefix only if it specifies a member pointer.
6859 (when (setq got-identifier (c-forward-name))
6860 (if (looking-at "\\(::\\)")
6861 ;; We only check for a trailing "::" and
6862 ;; let the "*" that should follow be
6863 ;; matched in the next round.
6864 (progn (setq got-identifier nil) t)
6865 ;; It turned out to be the real identifier,
6866 ;; so stop.
6867 nil))
6868 t))
6869
6870 (if (eq (char-after) ?\()
6871 (progn
6872 (setq paren-depth (1+ paren-depth))
6873 (forward-char))
6874 (unless got-prefix-before-parens
6875 (setq got-prefix-before-parens (= paren-depth 0)))
6876 (setq got-prefix t)
6877 (goto-char (match-end 1)))
6878 (c-forward-syntactic-ws))
6879
6880 (setq got-parens (> paren-depth 0))
6881
6882 ;; Skip over an identifier.
6883 (or got-identifier
6884 (and (looking-at c-identifier-start)
6885 (setq got-identifier (c-forward-name))))
6886
6887 ;; Skip over type decl suffix operators.
6888 (while (if (looking-at c-type-decl-suffix-key)
6889
6890 (if (eq (char-after) ?\))
6891 (when (> paren-depth 0)
6892 (setq paren-depth (1- paren-depth))
6893 (forward-char)
6894 t)
6895 (when (if (save-match-data (looking-at "\\s\("))
6896 (c-safe (c-forward-sexp 1) t)
6897 (goto-char (match-end 1))
6898 t)
6899 (when (and (not got-suffix-after-parens)
6900 (= paren-depth 0))
6901 (setq got-suffix-after-parens (match-beginning 0)))
6902 (setq got-suffix t)))
6903
6904 ;; No suffix matched. We might have matched the
6905 ;; identifier as a type and the open paren of a
6906 ;; function arglist as a type decl prefix. In that
6907 ;; case we should "backtrack": Reinterpret the last
6908 ;; type as the identifier, move out of the arglist and
6909 ;; continue searching for suffix operators.
6910 ;;
6911 ;; Do this even if there's no preceding type, to cope
6912 ;; with old style function declarations in K&R C,
6913 ;; (con|de)structors in C++ and `c-typeless-decl-kwds'
6914 ;; style declarations. That isn't applicable in an
6915 ;; arglist context, though.
6916 (when (and (= paren-depth 1)
6917 (not got-prefix-before-parens)
6918 (not (eq at-type t))
6919 (or backup-at-type
6920 maybe-typeless
6921 backup-maybe-typeless
6922 (when c-recognize-typeless-decls
6923 (not context)))
6924 (setq pos (c-up-list-forward (point)))
6925 (eq (char-before pos) ?\)))
6926 (c-fdoc-shift-type-backward)
6927 (goto-char pos)
6928 t))
6929
6930 (c-forward-syntactic-ws))
6931
6932 (when (and (or maybe-typeless backup-maybe-typeless)
6933 (not got-identifier)
6934 (not got-prefix)
6935 at-type)
6936 ;; Have found no identifier but `c-typeless-decl-kwds' has
6937 ;; matched so we know we're inside a declaration. The
6938 ;; preceding type must be the identifier instead.
6939 (c-fdoc-shift-type-backward))
6940
6941 (setq
6942 at-decl-or-cast
6943 (catch 'at-decl-or-cast
6944
a85fd6da 6945 ;; CASE 1
0386b551
AM
6946 (when (> paren-depth 0)
6947 ;; Encountered something inside parens that isn't matched by
6948 ;; the `c-type-decl-*' regexps, so it's not a type decl
6949 ;; expression. Try to skip out to the same paren depth to
6950 ;; not confuse the cast check below.
6951 (c-safe (goto-char (scan-lists (point) 1 paren-depth)))
6952 ;; If we've found a specifier keyword then it's a
6953 ;; declaration regardless.
6954 (throw 'at-decl-or-cast (eq at-decl-or-cast t)))
6955
6956 (setq at-decl-end
6957 (looking-at (cond ((eq context '<>) "[,>]")
6958 (context "[,\)]")
6959 (t "[,;]"))))
6960
6961 ;; Now we've collected info about various characteristics of
6962 ;; the construct we're looking at. Below follows a decision
6963 ;; tree based on that. It's ordered to check more certain
6964 ;; signs before less certain ones.
6965
6966 (if got-identifier
6967 (progn
6968
a85fd6da 6969 ;; CASE 2
0386b551
AM
6970 (when (and (or at-type maybe-typeless)
6971 (not (or got-prefix got-parens)))
6972 ;; Got another identifier directly after the type, so it's a
6973 ;; declaration.
6974 (throw 'at-decl-or-cast t))
6975
6976 (when (and got-parens
6977 (not got-prefix)
6978 (not got-suffix-after-parens)
6979 (or backup-at-type
6980 maybe-typeless
6981 backup-maybe-typeless))
6982 ;; Got a declaration of the form "foo bar (gnu);" where we've
6983 ;; recognized "bar" as the type and "gnu" as the declarator.
6984 ;; In this case it's however more likely that "bar" is the
6985 ;; declarator and "gnu" a function argument or initializer (if
6986 ;; `c-recognize-paren-inits' is set), since the parens around
6987 ;; "gnu" would be superfluous if it's a declarator. Shift the
6988 ;; type one step backward.
6989 (c-fdoc-shift-type-backward)))
6990
6991 ;; Found no identifier.
6992
6993 (if backup-at-type
6994 (progn
6995
b248a85d
AM
6996
6997 ;; CASE 3
6998 (when (= (point) start)
6999 ;; Got a plain list of identifiers. If a colon follows it's
7000 ;; a valid label, or maybe a bitfield. Otherwise the last
7001 ;; one probably is the declared identifier and we should
7002 ;; back up to the previous type, providing it isn't a cast.
452ea855
AM
7003 (if (and (eq (char-after) ?:)
7004 (not (c-major-mode-is 'java-mode)))
b248a85d
AM
7005 (cond
7006 ;; If we've found a specifier keyword then it's a
7007 ;; declaration regardless.
7008 ((eq at-decl-or-cast t)
7009 (throw 'at-decl-or-cast t))
7010 ((and c-has-bitfields
7011 (eq at-decl-or-cast 'ids)) ; bitfield.
7012 (setq backup-if-not-cast t)
7013 (throw 'at-decl-or-cast t)))
7014
7015 (setq backup-if-not-cast t)
7016 (throw 'at-decl-or-cast t)))
0386b551 7017
a85fd6da 7018 ;; CASE 4
0386b551
AM
7019 (when (and got-suffix
7020 (not got-prefix)
7021 (not got-parens))
7022 ;; Got a plain list of identifiers followed by some suffix.
7023 ;; If this isn't a cast then the last identifier probably is
7024 ;; the declared one and we should back up to the previous
7025 ;; type.
7026 (setq backup-if-not-cast t)
7027 (throw 'at-decl-or-cast t)))
7028
a85fd6da 7029 ;; CASE 5
0386b551
AM
7030 (when (eq at-type t)
7031 ;; If the type is known we know that there can't be any
7032 ;; identifier somewhere else, and it's only in declarations in
7033 ;; e.g. function prototypes and in casts that the identifier may
7034 ;; be left out.
7035 (throw 'at-decl-or-cast t))
7036
7037 (when (= (point) start)
7038 ;; Only got a single identifier (parsed as a type so far).
a85fd6da 7039 ;; CASE 6
0386b551
AM
7040 (if (and
7041 ;; Check that the identifier isn't at the start of an
7042 ;; expression.
7043 at-decl-end
7044 (cond
7045 ((eq context 'decl)
7046 ;; Inside an arglist that contains declarations. If K&R
7047 ;; style declarations and parenthesis style initializers
7048 ;; aren't allowed then the single identifier must be a
7049 ;; type, else we require that it's known or found
7050 ;; (primitive types are handled above).
7051 (or (and (not c-recognize-knr-p)
7052 (not c-recognize-paren-inits))
7053 (memq at-type '(known found))))
7054 ((eq context '<>)
7055 ;; Inside a template arglist. Accept known and found
7056 ;; types; other identifiers could just as well be
7057 ;; constants in C++.
7058 (memq at-type '(known found)))))
7059 (throw 'at-decl-or-cast t)
a85fd6da 7060 ;; CASE 7
0386b551
AM
7061 ;; Can't be a valid declaration or cast, but if we've found a
7062 ;; specifier it can't be anything else either, so treat it as
7063 ;; an invalid/unfinished declaration or cast.
7064 (throw 'at-decl-or-cast at-decl-or-cast))))
7065
7066 (if (and got-parens
7067 (not got-prefix)
7068 (not context)
7069 (not (eq at-type t))
7070 (or backup-at-type
7071 maybe-typeless
7072 backup-maybe-typeless
7073 (when c-recognize-typeless-decls
7074 (or (not got-suffix)
7075 (not (looking-at
7076 c-after-suffixed-type-maybe-decl-key))))))
7077 ;; Got an empty paren pair and a preceding type that probably
7078 ;; really is the identifier. Shift the type backwards to make
7079 ;; the last one the identifier. This is analogous to the
7080 ;; "backtracking" done inside the `c-type-decl-suffix-key' loop
7081 ;; above.
7082 ;;
7083 ;; Exception: In addition to the conditions in that
7084 ;; "backtracking" code, do not shift backward if we're not
7085 ;; looking at either `c-after-suffixed-type-decl-key' or "[;,]".
7086 ;; Since there's no preceding type, the shift would mean that
7087 ;; the declaration is typeless. But if the regexp doesn't match
7088 ;; then we will simply fall through in the tests below and not
7089 ;; recognize it at all, so it's better to try it as an abstract
7090 ;; declarator instead.
7091 (c-fdoc-shift-type-backward)
7092
7093 ;; Still no identifier.
a85fd6da 7094 ;; CASE 8
0386b551
AM
7095 (when (and got-prefix (or got-parens got-suffix))
7096 ;; Require `got-prefix' together with either `got-parens' or
7097 ;; `got-suffix' to recognize it as an abstract declarator:
7098 ;; `got-parens' only is probably an empty function call.
7099 ;; `got-suffix' only can build an ordinary expression together
7100 ;; with the preceding identifier which we've taken as a type.
7101 ;; We could actually accept on `got-prefix' only, but that can
7102 ;; easily occur temporarily while writing an expression so we
7103 ;; avoid that case anyway. We could do a better job if we knew
7104 ;; the point when the fontification was invoked.
7105 (throw 'at-decl-or-cast t))
7106
a85fd6da 7107 ;; CASE 9
0386b551
AM
7108 (when (and at-type
7109 (not got-prefix)
7110 (not got-parens)
7111 got-suffix-after-parens
7112 (eq (char-after got-suffix-after-parens) ?\())
7113 ;; Got a type, no declarator but a paren suffix. I.e. it's a
fa463103 7114 ;; normal function call after all (or perhaps a C++ style object
0386b551
AM
7115 ;; instantiation expression).
7116 (throw 'at-decl-or-cast nil))))
7117
a85fd6da 7118 ;; CASE 10
0386b551
AM
7119 (when at-decl-or-cast
7120 ;; By now we've located the type in the declaration that we know
7121 ;; we're in.
7122 (throw 'at-decl-or-cast t))
7123
a85fd6da 7124 ;; CASE 11
0386b551
AM
7125 (when (and got-identifier
7126 (not context)
7127 (looking-at c-after-suffixed-type-decl-key)
7128 (if (and got-parens
7129 (not got-prefix)
7130 (not got-suffix)
7131 (not (eq at-type t)))
7132 ;; Shift the type backward in the case that there's a
7133 ;; single identifier inside parens. That can only
7134 ;; occur in K&R style function declarations so it's
7135 ;; more likely that it really is a function call.
7136 ;; Therefore we only do this after
7137 ;; `c-after-suffixed-type-decl-key' has matched.
7138 (progn (c-fdoc-shift-type-backward) t)
7139 got-suffix-after-parens))
7140 ;; A declaration according to `c-after-suffixed-type-decl-key'.
7141 (throw 'at-decl-or-cast t))
7142
a85fd6da 7143 ;; CASE 12
0386b551
AM
7144 (when (and (or got-prefix (not got-parens))
7145 (memq at-type '(t known)))
7146 ;; It's a declaration if a known type precedes it and it can't be a
7147 ;; function call.
7148 (throw 'at-decl-or-cast t))
7149
7150 ;; If we get here we can't tell if this is a type decl or a normal
7151 ;; expression by looking at it alone. (That's under the assumption
7152 ;; that normal expressions always can look like type decl expressions,
7153 ;; which isn't really true but the cases where it doesn't hold are so
7154 ;; uncommon (e.g. some placements of "const" in C++) it's not worth
7155 ;; the effort to look for them.)
7156
7157 (unless (or at-decl-end (looking-at "=[^=]"))
7158 ;; If this is a declaration it should end here or its initializer(*)
7159 ;; should start here, so check for allowed separation tokens. Note
7160 ;; that this rule doesn't work e.g. with a K&R arglist after a
7161 ;; function header.
7162 ;;
7163 ;; *) Don't check for C++ style initializers using parens
7164 ;; since those already have been matched as suffixes.
7165 ;;
7166 ;; If `at-decl-or-cast' is then we've found some other sign that
7167 ;; it's a declaration or cast, so then it's probably an
7168 ;; invalid/unfinished one.
7169 (throw 'at-decl-or-cast at-decl-or-cast))
7170
7171 ;; Below are tests that only should be applied when we're certain to
7172 ;; not have parsed halfway through an expression.
7173
a85fd6da 7174 ;; CASE 14
0386b551
AM
7175 (when (memq at-type '(t known))
7176 ;; The expression starts with a known type so treat it as a
7177 ;; declaration.
7178 (throw 'at-decl-or-cast t))
7179
a85fd6da 7180 ;; CASE 15
0386b551
AM
7181 (when (and (c-major-mode-is 'c++-mode)
7182 ;; In C++ we check if the identifier is a known type, since
7183 ;; (con|de)structors use the class name as identifier.
7184 ;; We've always shifted over the identifier as a type and
7185 ;; then backed up again in this case.
7186 identifier-type
7187 (or (memq identifier-type '(found known))
7188 (and (eq (char-after identifier-start) ?~)
7189 ;; `at-type' probably won't be 'found for
7190 ;; destructors since the "~" is then part of the
7191 ;; type name being checked against the list of
7192 ;; known types, so do a check without that
7193 ;; operator.
7194 (or (save-excursion
7195 (goto-char (1+ identifier-start))
7196 (c-forward-syntactic-ws)
7197 (c-with-syntax-table
7198 c-identifier-syntax-table
7199 (looking-at c-known-type-key)))
7200 (save-excursion
7201 (goto-char (1+ identifier-start))
7202 ;; We have already parsed the type earlier,
7203 ;; so it'd be possible to cache the end
7204 ;; position instead of redoing it here, but
7205 ;; then we'd need to keep track of another
7206 ;; position everywhere.
7207 (c-check-type (point)
7208 (progn (c-forward-type)
7209 (point))))))))
7210 (throw 'at-decl-or-cast t))
7211
7212 (if got-identifier
7213 (progn
a85fd6da 7214 ;; CASE 16
0386b551
AM
7215 (when (and got-prefix-before-parens
7216 at-type
7217 (or at-decl-end (looking-at "=[^=]"))
7218 (not context)
7219 (not got-suffix))
7220 ;; Got something like "foo * bar;". Since we're not inside an
7221 ;; arglist it would be a meaningless expression because the
7222 ;; result isn't used. We therefore choose to recognize it as
7223 ;; a declaration. Do not allow a suffix since it could then
7224 ;; be a function call.
7225 (throw 'at-decl-or-cast t))
7226
a85fd6da 7227 ;; CASE 17
0386b551
AM
7228 (when (and (or got-suffix-after-parens
7229 (looking-at "=[^=]"))
7230 (eq at-type 'found)
7231 (not (eq context 'arglist)))
7232 ;; Got something like "a (*b) (c);" or "a (b) = c;". It could
7233 ;; be an odd expression or it could be a declaration. Treat
7234 ;; it as a declaration if "a" has been used as a type
7235 ;; somewhere else (if it's a known type we won't get here).
7236 (throw 'at-decl-or-cast t)))
7237
a85fd6da 7238 ;; CASE 18
0386b551
AM
7239 (when (and context
7240 (or got-prefix
7241 (and (eq context 'decl)
7242 (not c-recognize-paren-inits)
7243 (or got-parens got-suffix))))
7244 ;; Got a type followed by an abstract declarator. If `got-prefix'
7245 ;; is set it's something like "a *" without anything after it. If
7246 ;; `got-parens' or `got-suffix' is set it's "a()", "a[]", "a()[]",
7247 ;; or similar, which we accept only if the context rules out
7248 ;; expressions.
7249 (throw 'at-decl-or-cast t)))
7250
7251 ;; If we had a complete symbol table here (which rules out
7252 ;; `c-found-types') we should return t due to the disambiguation rule
7253 ;; (in at least C++) that anything that can be parsed as a declaration
7254 ;; is a declaration. Now we're being more defensive and prefer to
7255 ;; highlight things like "foo (bar);" as a declaration only if we're
7256 ;; inside an arglist that contains declarations.
7257 (eq context 'decl))))
7258
7259 ;; The point is now after the type decl expression.
7260
7261 (cond
7262 ;; Check for a cast.
7263 ((save-excursion
7264 (and
7265 c-cast-parens
7266
7267 ;; Should be the first type/identifier in a cast paren.
7268 (> preceding-token-end (point-min))
7269 (memq (char-before preceding-token-end) c-cast-parens)
7270
7271 ;; The closing paren should follow.
7272 (progn
7273 (c-forward-syntactic-ws)
7274 (looking-at "\\s\)"))
7275
7276 ;; There should be a primary expression after it.
7277 (let (pos)
7278 (forward-char)
7279 (c-forward-syntactic-ws)
7280 (setq cast-end (point))
7281 (and (looking-at c-primary-expr-regexp)
7282 (progn
7283 (setq pos (match-end 0))
7284 (or
7285 ;; Check if the expression begins with a prefix keyword.
7286 (match-beginning 2)
7287 (if (match-beginning 1)
7288 ;; Expression begins with an ambiguous operator. Treat
7289 ;; it as a cast if it's a type decl or if we've
7290 ;; recognized the type somewhere else.
7291 (or at-decl-or-cast
7292 (memq at-type '(t known found)))
7293 ;; Unless it's a keyword, it's the beginning of a primary
7294 ;; expression.
7295 (not (looking-at c-keywords-regexp)))))
7296 ;; If `c-primary-expr-regexp' matched a nonsymbol token, check
7297 ;; that it matched a whole one so that we don't e.g. confuse
7298 ;; the operator '-' with '->'. It's ok if it matches further,
7299 ;; though, since it e.g. can match the float '.5' while the
7300 ;; operator regexp only matches '.'.
7301 (or (not (looking-at c-nonsymbol-token-regexp))
7302 (<= (match-end 0) pos))))
7303
7304 ;; There should either be a cast before it or something that isn't an
7305 ;; identifier or close paren.
7306 (> preceding-token-end (point-min))
7307 (progn
7308 (goto-char (1- preceding-token-end))
7309 (or (eq (point) last-cast-end)
7310 (progn
7311 (c-backward-syntactic-ws)
7312 (if (< (skip-syntax-backward "w_") 0)
7313 ;; It's a symbol. Accept it only if it's one of the
7314 ;; keywords that can precede an expression (without
7315 ;; surrounding parens).
7316 (looking-at c-simple-stmt-key)
7317 (and
7318 ;; Check that it isn't a close paren (block close is ok,
7319 ;; though).
7320 (not (memq (char-before) '(?\) ?\])))
7321 ;; Check that it isn't a nonsymbol identifier.
7322 (not (c-on-identifier)))))))))
7323
7324 ;; Handle the cast.
7325 (when (and c-record-type-identifiers at-type (not (eq at-type t)))
7326 (let ((c-promote-possible-types t))
7327 (goto-char type-start)
7328 (c-forward-type)))
7329
7330 (goto-char cast-end)
7331 'cast)
7332
7333 (at-decl-or-cast
7334 ;; We're at a declaration. Highlight the type and the following
7335 ;; declarators.
7336
7337 (when backup-if-not-cast
7338 (c-fdoc-shift-type-backward t))
7339
7340 (when (and (eq context 'decl) (looking-at ","))
7341 ;; Make sure to propagate the `c-decl-arg-start' property to
7342 ;; the next argument if it's set in this one, to cope with
7343 ;; interactive refontification.
7344 (c-put-c-type-property (point) 'c-decl-arg-start))
7345
7346 (when (and c-record-type-identifiers at-type (not (eq at-type t)))
7347 (let ((c-promote-possible-types t))
7348 (save-excursion
7349 (goto-char type-start)
7350 (c-forward-type))))
7351
e15f8aaa
AM
7352 (cons id-start
7353 (and (or at-type-decl at-typedef)
7354 (cons at-type-decl at-typedef))))
0386b551
AM
7355
7356 (t
7357 ;; False alarm. Restore the recorded ranges.
7358 (setq c-record-type-identifiers save-rec-type-ids
7359 c-record-ref-identifiers save-rec-ref-ids)
7360 nil))))
7361
7362(defun c-forward-label (&optional assume-markup preceding-token-end limit)
51c9af45 7363 ;; Assuming that point is at the beginning of a token, check if it starts a
1379f2c5
AM
7364 ;; label and if so move over it and return non-nil (t in default situations,
7365 ;; specific symbols (see below) for interesting situations), otherwise don't
7366 ;; move and return nil. "Label" here means "most things with a colon".
51c9af45
AM
7367 ;;
7368 ;; More precisely, a "label" is regarded as one of:
1379f2c5
AM
7369 ;; (i) a goto target like "foo:" - returns the symbol `goto-target';
7370 ;; (ii) A case label - either the entire construct "case FOO:", or just the
7371 ;; bare "case", should the colon be missing. We return t;
7372 ;; (iii) a keyword which needs a colon, like "default:" or "private:"; We
7373 ;; return t;
51c9af45 7374 ;; (iv) One of QT's "extended" C++ variants of
1379f2c5
AM
7375 ;; "private:"/"protected:"/"public:"/"more:" looking like "public slots:".
7376 ;; Returns the symbol `qt-2kwds-colon'.
7377 ;; (v) QT's construct "signals:". Returns the symbol `qt-1kwd-colon'.
82ba65cf 7378 ;; (vi) One of the keywords matched by `c-opt-extra-label-key' (without any
51c9af45 7379 ;; colon). Currently (2006-03), this applies only to Objective C's
1379f2c5 7380 ;; keywords "@private", "@protected", and "@public". Returns t.
51c9af45 7381 ;;
e1dbe924 7382 ;; One of the things which will NOT be recognized as a label is a bit-field
51c9af45
AM
7383 ;; element of a struct, something like "int foo:5".
7384 ;;
7385 ;; The end of the label is taken to be just after the colon, or the end of
7386 ;; the first submatch in `c-opt-extra-label-key'. The point is directly
7387 ;; after the end on return. The terminating char gets marked with
7388 ;; `c-decl-end' to improve recognition of the following declaration or
7389 ;; statement.
0386b551
AM
7390 ;;
7391 ;; If ASSUME-MARKUP is non-nil, it's assumed that the preceding
51c9af45 7392 ;; label, if any, has already been marked up like that.
0386b551
AM
7393 ;;
7394 ;; If PRECEDING-TOKEN-END is given, it should be the first position
7395 ;; after the preceding token, i.e. on the other side of the
7396 ;; syntactic ws from the point. Use a value less than or equal to
7397 ;; (point-min) if the point is at the first token in (the visible
7398 ;; part of) the buffer.
7399 ;;
7400 ;; The optional LIMIT limits the forward scan for the colon.
7401 ;;
7402 ;; This function records the ranges of the label symbols on
7403 ;; `c-record-ref-identifiers' if `c-record-type-identifiers' (!) is
7404 ;; non-nil.
7405 ;;
7406 ;; This function might do hidden buffer changes.
7407
51c9af45 7408 (let ((start (point))
1379f2c5 7409 label-end
51c9af45 7410 qt-symbol-idx
1379f2c5 7411 macro-start ; if we're in one.
f412a567
AM
7412 label-type
7413 kwd)
0386b551 7414 (cond
b414f371 7415 ;; "case" or "default" (Doesn't apply to AWK).
0386b551
AM
7416 ((looking-at c-label-kwds-regexp)
7417 (let ((kwd-end (match-end 1)))
7418 ;; Record only the keyword itself for fontification, since in
7419 ;; case labels the following is a constant expression and not
7420 ;; a label.
7421 (when c-record-type-identifiers
7422 (c-record-ref-id (cons (match-beginning 1) kwd-end)))
7423
7424 ;; Find the label end.
7425 (goto-char kwd-end)
1379f2c5
AM
7426 (setq label-type
7427 (if (and (c-syntactic-re-search-forward
7428 ;; Stop on chars that aren't allowed in expressions,
7429 ;; and on operator chars that would be meaningless
7430 ;; there. FIXME: This doesn't cope with ?: operators.
7431 "[;{=,@]\\|\\(\\=\\|[^:]\\):\\([^:]\\|\\'\\)"
7432 limit t t nil 1)
7433 (match-beginning 2))
7434
7435 (progn ; there's a proper :
7436 (goto-char (match-beginning 2)) ; just after the :
7437 (c-put-c-type-property (1- (point)) 'c-decl-end)
7438 t)
7439
7440 ;; It's an unfinished label. We consider the keyword enough
7441 ;; to recognize it as a label, so that it gets fontified.
7442 ;; Leave the point at the end of it, but don't put any
7443 ;; `c-decl-end' marker.
7444 (goto-char kwd-end)
7445 t))))
0386b551 7446
51c9af45 7447 ;; @private, @protected, @public, in Objective C, or similar.
0386b551
AM
7448 ((and c-opt-extra-label-key
7449 (looking-at c-opt-extra-label-key))
7450 ;; For a `c-opt-extra-label-key' match, we record the whole
7451 ;; thing for fontification. That's to get the leading '@' in
7452 ;; Objective-C protection labels fontified.
7453 (goto-char (match-end 1))
7454 (when c-record-type-identifiers
7455 (c-record-ref-id (cons (match-beginning 1) (point))))
7456 (c-put-c-type-property (1- (point)) 'c-decl-end)
1379f2c5 7457 (setq label-type t))
0386b551 7458
51c9af45
AM
7459 ;; All other cases of labels.
7460 ((and c-recognize-colon-labels ; nil for AWK and IDL, otherwise t.
0386b551
AM
7461
7462 ;; A colon label must have something before the colon.
7463 (not (eq (char-after) ?:))
7464
7465 ;; Check that we're not after a token that can't precede a label.
7466 (or
7467 ;; Trivially succeeds when there's no preceding token.
536610a4 7468 ;; Succeeds when we're at a virtual semicolon.
0386b551
AM
7469 (if preceding-token-end
7470 (<= preceding-token-end (point-min))
7471 (save-excursion
7472 (c-backward-syntactic-ws)
7473 (setq preceding-token-end (point))
536610a4
AM
7474 (or (bobp)
7475 (c-at-vsemi-p))))
0386b551
AM
7476
7477 ;; Check if we're after a label, if we're after a closing
7478 ;; paren that belong to statement, and with
7479 ;; `c-label-prefix-re'. It's done in different order
7480 ;; depending on `assume-markup' since the checks have
7481 ;; different expensiveness.
7482 (if assume-markup
7483 (or
7484 (eq (c-get-char-property (1- preceding-token-end) 'c-type)
7485 'c-decl-end)
7486
7487 (save-excursion
7488 (goto-char (1- preceding-token-end))
7489 (c-beginning-of-current-token)
51c9af45
AM
7490 (or (looking-at c-label-prefix-re)
7491 (looking-at c-block-stmt-1-key)))
0386b551
AM
7492
7493 (and (eq (char-before preceding-token-end) ?\))
7494 (c-after-conditional)))
7495
7496 (or
7497 (save-excursion
7498 (goto-char (1- preceding-token-end))
7499 (c-beginning-of-current-token)
51c9af45
AM
7500 (or (looking-at c-label-prefix-re)
7501 (looking-at c-block-stmt-1-key)))
0386b551
AM
7502
7503 (cond
7504 ((eq (char-before preceding-token-end) ?\))
7505 (c-after-conditional))
7506
7507 ((eq (char-before preceding-token-end) ?:)
7508 ;; Might be after another label, so check it recursively.
51c9af45
AM
7509 (save-restriction
7510 (save-excursion
7511 (goto-char (1- preceding-token-end))
7512 ;; Essentially the same as the
7513 ;; `c-syntactic-re-search-forward' regexp below.
7514 (setq macro-start
7515 (save-excursion (and (c-beginning-of-macro)
7516 (point))))
7517 (if macro-start (narrow-to-region macro-start (point-max)))
7518 (c-syntactic-skip-backward "^-]:?;}=*/%&|,<>!@+" nil t)
7519 ;; Note: the following should work instead of the
7520 ;; narrow-to-region above. Investigate why not,
7521 ;; sometime. ACM, 2006-03-31.
7522 ;; (c-syntactic-skip-backward "^-]:?;}=*/%&|,<>!@+"
7523 ;; macro-start t)
7524 (let ((pte (point))
7525 ;; If the caller turned on recording for us,
7526 ;; it shouldn't apply when we check the
7527 ;; preceding label.
7528 c-record-type-identifiers)
7529 ;; A label can't start at a cpp directive. Check for
7530 ;; this, since c-forward-syntactic-ws would foul up on it.
7531 (unless (and c-opt-cpp-prefix (looking-at c-opt-cpp-prefix))
7532 (c-forward-syntactic-ws)
7533 (c-forward-label nil pte start))))))))))
7534
1379f2c5 7535 ;; Point is still at the beginning of the possible label construct.
b414f371 7536 ;;
51c9af45
AM
7537 ;; Check that the next nonsymbol token is ":", or that we're in one
7538 ;; of QT's "slots" declarations. Allow '(' for the sake of macro
7539 ;; arguments. FIXME: Should build this regexp from the language
7540 ;; constants.
1379f2c5
AM
7541 (cond
7542 ;; public: protected: private:
7543 ((and
7544 (c-major-mode-is 'c++-mode)
7545 (search-forward-regexp
7546 "\\=p\\(r\\(ivate\\|otected\\)\\|ublic\\)\\>[^_]" nil t)
7547 (progn (backward-char)
7548 (c-forward-syntactic-ws limit)
7549 (looking-at ":\\([^:]\\|\\'\\)"))) ; A single colon.
7550 (forward-char)
7551 (setq label-type t))
7552 ;; QT double keyword like "protected slots:" or goto target.
7553 ((progn (goto-char start) nil))
7554 ((when (c-syntactic-re-search-forward
7555 "[ \t\n[:?;{=*/%&|,<>!@+-]" limit t t) ; not at EOB
7556 (backward-char)
7557 (setq label-end (point))
7558 (setq qt-symbol-idx
7559 (and (c-major-mode-is 'c++-mode)
7560 (string-match
7561 "\\(p\\(r\\(ivate\\|otected\\)\\|ublic\\)\\|more\\)\\>"
b414f371 7562 (buffer-substring start (point)))))
1379f2c5
AM
7563 (c-forward-syntactic-ws limit)
7564 (cond
7565 ((looking-at ":\\([^:]\\|\\'\\)") ; A single colon.
7566 (forward-char)
7567 (setq label-type
f412a567
AM
7568 (if (or (string= "signals" ; Special QT macro
7569 (setq kwd (buffer-substring-no-properties start label-end)))
7570 (string= "Q_SIGNALS" kwd))
1379f2c5
AM
7571 'qt-1kwd-colon
7572 'goto-target)))
7573 ((and qt-symbol-idx
f412a567 7574 (search-forward-regexp "\\=\\(slots\\|Q_SLOTS\\)\\>" limit t)
1379f2c5
AM
7575 (progn (c-forward-syntactic-ws limit)
7576 (looking-at ":\\([^:]\\|\\'\\)"))) ; A single colon
7577 (forward-char)
7578 (setq label-type 'qt-2kwds-colon)))))))
0386b551
AM
7579
7580 (save-restriction
7581 (narrow-to-region start (point))
7582
7583 ;; Check that `c-nonlabel-token-key' doesn't match anywhere.
7584 (catch 'check-label
7585 (goto-char start)
7586 (while (progn
7587 (when (looking-at c-nonlabel-token-key)
7588 (goto-char start)
1379f2c5 7589 (setq label-type nil)
0386b551
AM
7590 (throw 'check-label nil))
7591 (and (c-safe (c-forward-sexp)
7592 (c-forward-syntactic-ws)
7593 t)
7594 (not (eobp)))))
7595
7596 ;; Record the identifiers in the label for fontification, unless
7597 ;; it begins with `c-label-kwds' in which case the following
7598 ;; identifiers are part of a (constant) expression that
7599 ;; shouldn't be fontified.
7600 (when (and c-record-type-identifiers
7601 (progn (goto-char start)
7602 (not (looking-at c-label-kwds-regexp))))
7603 (while (c-syntactic-re-search-forward c-symbol-key nil t)
7604 (c-record-ref-id (cons (match-beginning 0)
7605 (match-end 0)))))
7606
7607 (c-put-c-type-property (1- (point-max)) 'c-decl-end)
82ba65cf 7608 (goto-char (point-max)))))
0386b551
AM
7609
7610 (t
7611 ;; Not a label.
1379f2c5
AM
7612 (goto-char start)))
7613 label-type))
0386b551
AM
7614
7615(defun c-forward-objc-directive ()
7616 ;; Assuming the point is at the beginning of a token, try to move
7617 ;; forward to the end of the Objective-C directive that starts
7618 ;; there. Return t if a directive was fully recognized, otherwise
7619 ;; the point is moved as far as one could be successfully parsed and
7620 ;; nil is returned.
7621 ;;
7622 ;; This function records identifier ranges on
7623 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
7624 ;; `c-record-type-identifiers' is non-nil.
7625 ;;
7626 ;; This function might do hidden buffer changes.
7627
7628 (let ((start (point))
7629 start-char
7630 (c-promote-possible-types t)
667ced3a 7631 lim
0386b551
AM
7632 ;; Turn off recognition of angle bracket arglists while parsing
7633 ;; types here since the protocol reference list might then be
7634 ;; considered part of the preceding name or superclass-name.
7635 c-recognize-<>-arglists)
7636
7637 (if (or
7638 (when (looking-at
7639 (eval-when-compile
7640 (c-make-keywords-re t
7641 (append (c-lang-const c-protection-kwds objc)
7642 '("@end"))
7643 'objc-mode)))
7644 (goto-char (match-end 1))
7645 t)
7646
7647 (and
7648 (looking-at
7649 (eval-when-compile
7650 (c-make-keywords-re t
7651 '("@interface" "@implementation" "@protocol")
7652 'objc-mode)))
7653
7654 ;; Handle the name of the class itself.
7655 (progn
cb694ab7
AM
7656; (c-forward-token-2) ; 2006/1/13 This doesn't move if the token's
7657; at EOB.
7658 (goto-char (match-end 0))
667ced3a 7659 (setq lim (point))
cb694ab7 7660 (c-skip-ws-forward)
0386b551
AM
7661 (c-forward-type))
7662
7663 (catch 'break
7664 ;; Look for ": superclass-name" or "( category-name )".
7665 (when (looking-at "[:\(]")
7666 (setq start-char (char-after))
7667 (forward-char)
7668 (c-forward-syntactic-ws)
7669 (unless (c-forward-type) (throw 'break nil))
7670 (when (eq start-char ?\()
7671 (unless (eq (char-after) ?\)) (throw 'break nil))
7672 (forward-char)
7673 (c-forward-syntactic-ws)))
7674
7675 ;; Look for a protocol reference list.
7676 (if (eq (char-after) ?<)
7677 (let ((c-recognize-<>-arglists t)
7678 (c-parse-and-markup-<>-arglists t)
7679 c-restricted-<>-arglists)
7680 (c-forward-<>-arglist t))
7681 t))))
7682
7683 (progn
667ced3a 7684 (c-backward-syntactic-ws lim)
0386b551
AM
7685 (c-clear-c-type-property start (1- (point)) 'c-decl-end)
7686 (c-put-c-type-property (1- (point)) 'c-decl-end)
7687 t)
7688
7689 (c-clear-c-type-property start (point) 'c-decl-end)
7690 nil)))
7691
785eecbb
RS
7692(defun c-beginning-of-inheritance-list (&optional lim)
7693 ;; Go to the first non-whitespace after the colon that starts a
7694 ;; multiple inheritance introduction. Optional LIM is the farthest
7695 ;; back we should search.
0386b551
AM
7696 ;;
7697 ;; This function might do hidden buffer changes.
7698 (c-with-syntax-table c++-template-syntax-table
7699 (c-backward-token-2 0 t lim)
7700 (while (and (or (looking-at c-symbol-start)
7701 (looking-at "[<,]\\|::"))
7702 (zerop (c-backward-token-2 1 t lim))))))
785eecbb 7703
785eecbb
RS
7704(defun c-in-method-def-p ()
7705 ;; Return nil if we aren't in a method definition, otherwise the
7706 ;; position of the initial [+-].
0386b551
AM
7707 ;;
7708 ;; This function might do hidden buffer changes.
785eecbb
RS
7709 (save-excursion
7710 (beginning-of-line)
a66cd3ee
MS
7711 (and c-opt-method-key
7712 (looking-at c-opt-method-key)
785eecbb
RS
7713 (point))
7714 ))
7715
a66cd3ee
MS
7716;; Contributed by Kevin Ryde <user42@zip.com.au>.
7717(defun c-in-gcc-asm-p ()
7718 ;; Return non-nil if point is within a gcc \"asm\" block.
7719 ;;
7720 ;; This should be called with point inside an argument list.
7721 ;;
7722 ;; Only one level of enclosing parentheses is considered, so for
7723 ;; instance `nil' is returned when in a function call within an asm
7724 ;; operand.
0386b551
AM
7725 ;;
7726 ;; This function might do hidden buffer changes.
a66cd3ee
MS
7727
7728 (and c-opt-asm-stmt-key
7729 (save-excursion
7730 (beginning-of-line)
7731 (backward-up-list 1)
7732 (c-beginning-of-statement-1 (point-min) nil t)
7733 (looking-at c-opt-asm-stmt-key))))
7734
abb7e5cf 7735(defun c-at-toplevel-p ()
a85fd6da
AM
7736 "Return a determination as to whether point is \"at the top level\".
7737Informally, \"at the top level\" is anywhere where you can write
7738a function.
7739
7740More precisely, being at the top-level means that point is either
7741outside any enclosing block (such as a function definition), or
7742directly inside a class, namespace or other block that contains
7743another declaration level.
abb7e5cf
SM
7744
7745If point is not at the top-level (e.g. it is inside a method
7746definition), then nil is returned. Otherwise, if point is at a
7747top-level not enclosed within a class definition, t is returned.
7748Otherwise, a 2-vector is returned where the zeroth element is the
7749buffer position of the start of the class declaration, and the first
7750element is the buffer position of the enclosing class's opening
0386b551
AM
7751brace.
7752
7753Note that this function might do hidden buffer changes. See the
7754comment at the start of cc-engine.el for more info."
a66cd3ee
MS
7755 (let ((paren-state (c-parse-state)))
7756 (or (not (c-most-enclosing-brace paren-state))
7757 (c-search-uplist-for-classkey paren-state))))
7758
d9e94c22 7759(defun c-just-after-func-arglist-p (&optional lim)
0386b551
AM
7760 ;; Return non-nil if the point is in the region after the argument
7761 ;; list of a function and its opening brace (or semicolon in case it
7762 ;; got no body). If there are K&R style argument declarations in
7763 ;; that region, the point has to be inside the first one for this
7764 ;; function to recognize it.
a66cd3ee 7765 ;;
0386b551
AM
7766 ;; If successful, the point is moved to the first token after the
7767 ;; function header (see `c-forward-decl-or-cast-1' for details) and
7768 ;; the position of the opening paren of the function arglist is
7769 ;; returned.
7770 ;;
7771 ;; The point is clobbered if not successful.
7772 ;;
7773 ;; LIM is used as bound for backward buffer searches.
7774 ;;
7775 ;; This function might do hidden buffer changes.
7776
7777 (let ((beg (point)) end id-start)
7778 (and
7779 (eq (c-beginning-of-statement-1 lim) 'same)
7780
2e492df3
AM
7781 (not (and (c-major-mode-is 'objc-mode)
7782 (c-forward-objc-directive)))
0386b551
AM
7783
7784 (setq id-start
7785 (car-safe (c-forward-decl-or-cast-1 (c-point 'bosws) nil nil)))
7786 (< id-start beg)
7787
7788 ;; There should not be a '=' or ',' between beg and the
7789 ;; start of the declaration since that means we were in the
7790 ;; "expression part" of the declaration.
7791 (or (> (point) beg)
7792 (not (looking-at "[=,]")))
7793
7794 (save-excursion
7795 ;; Check that there's an arglist paren in the
7796 ;; declaration.
7797 (goto-char id-start)
7798 (cond ((eq (char-after) ?\()
7799 ;; The declarator is a paren expression, so skip past it
7800 ;; so that we don't get stuck on that instead of the
7801 ;; function arglist.
7802 (c-forward-sexp))
51c9af45
AM
7803 ((and c-opt-op-identifier-prefix
7804 (looking-at c-opt-op-identifier-prefix))
0386b551
AM
7805 ;; Don't trip up on "operator ()".
7806 (c-forward-token-2 2 t)))
7807 (and (< (point) beg)
7808 (c-syntactic-re-search-forward "(" beg t t)
7809 (1- (point)))))))
785eecbb 7810
a66cd3ee
MS
7811(defun c-in-knr-argdecl (&optional lim)
7812 ;; Return the position of the first argument declaration if point is
7813 ;; inside a K&R style argument declaration list, nil otherwise.
7814 ;; `c-recognize-knr-p' is not checked. If LIM is non-nil, it's a
7815 ;; position that bounds the backward search for the argument list.
7816 ;;
9cf17ef1
AM
7817 ;; Point must be within a possible K&R region, e.g. just before a top-level
7818 ;; "{". It must be outside of parens and brackets. The test can return
7819 ;; false positives otherwise.
0386b551
AM
7820 ;;
7821 ;; This function might do hidden buffer changes.
d9e94c22 7822
a66cd3ee
MS
7823 (save-excursion
7824 (save-restriction
9cf17ef1
AM
7825 ;; If we're in a macro, our search range is restricted to it. Narrow to
7826 ;; the searchable range.
98d7371e
AM
7827 (let* ((macro-start (save-excursion (and (c-beginning-of-macro) (point))))
7828 (macro-end (save-excursion (and macro-start (c-end-of-macro) (point))))
7829 (low-lim (max (or lim (point-min)) (or macro-start (point-min))))
6b6481ed 7830 before-lparen after-rparen
98d7371e
AM
7831 (pp-count-out 20)) ; Max number of paren/brace constructs before
7832 ; we give up
7833 (narrow-to-region low-lim (or macro-end (point-max)))
9cf17ef1
AM
7834
7835 ;; Search backwards for the defun's argument list. We give up if we
98d7371e
AM
7836 ;; encounter a "}" (end of a previous defun) an "=" (which can't be in
7837 ;; a knr region) or BOB.
9cf17ef1
AM
7838 ;;
7839 ;; The criterion for a paren structure being the arg list is:
7840 ;; o - there is non-WS stuff after it but before any "{"; AND
7841 ;; o - the token after it isn't a ";" AND
7842 ;; o - it is preceded by either an identifier (the function name) or
7843 ;; a macro expansion like "DEFUN (...)"; AND
7844 ;; o - its content is a non-empty comma-separated list of identifiers
7845 ;; (an empty arg list won't have a knr region).
7846 ;;
7847 ;; The following snippet illustrates these rules:
7848 ;; int foo (bar, baz, yuk)
7849 ;; int bar [] ;
7850 ;; int (*baz) (my_type) ;
7851 ;; int (*) (void) (*yuk) (void) ;
7852 ;; {
7853
7854 (catch 'knr
6b6481ed
AM
7855 (while (> pp-count-out 0) ; go back one paren/bracket pair each time.
7856 (setq pp-count-out (1- pp-count-out))
98d7371e 7857 (c-syntactic-skip-backward "^)]}=")
9cf17ef1
AM
7858 (cond ((eq (char-before) ?\))
7859 (setq after-rparen (point)))
10489fcb
AM
7860 ((eq (char-before) ?\])
7861 (setq after-rparen nil))
98d7371e
AM
7862 (t ; either } (hit previous defun) or = or no more
7863 ; parens/brackets.
10489fcb 7864 (throw 'knr nil)))
9cf17ef1
AM
7865
7866 (if after-rparen
7867 ;; We're inside a paren. Could it be our argument list....?
7868 (if
7869 (and
7870 (progn
7871 (goto-char after-rparen)
7872 (unless (c-go-list-backward) (throw 'knr nil)) ;
7873 ;; FIXME!!! What about macros between the parens? 2007/01/20
7874 (setq before-lparen (point)))
d9e94c22 7875
9cf17ef1
AM
7876 ;; It can't be the arg list if next token is ; or {
7877 (progn (goto-char after-rparen)
7878 (c-forward-syntactic-ws)
98d7371e 7879 (not (memq (char-after) '(?\; ?\{ ?\=))))
d9e94c22 7880
9cf17ef1
AM
7881 ;; Is the thing preceding the list an identifier (the
7882 ;; function name), or a macro expansion?
7883 (progn
7884 (goto-char before-lparen)
7885 (eq (c-backward-token-2) 0)
98d7371e 7886 (or (eq (c-on-identifier) (point))
9cf17ef1
AM
7887 (and (eq (char-after) ?\))
7888 (c-go-up-list-backward)
7889 (eq (c-backward-token-2) 0)
98d7371e 7890 (eq (c-on-identifier) (point)))))
9cf17ef1
AM
7891
7892 ;; Have we got a non-empty list of comma-separated
7893 ;; identifiers?
7894 (progn
7895 (goto-char before-lparen)
7896 (c-forward-token-2) ; to first token inside parens
7897 (and
7898 (c-on-identifier)
7899 (c-forward-token-2)
7900 (catch 'id-list
7901 (while (eq (char-after) ?\,)
7902 (c-forward-token-2)
7903 (unless (c-on-identifier) (throw 'id-list nil))
7904 (c-forward-token-2))
7905 (eq (char-after) ?\))))))
7906
7907 ;; ...Yes. We've identified the function's argument list.
7908 (throw 'knr
7909 (progn (goto-char after-rparen)
7910 (c-forward-syntactic-ws)
7911 (point)))
7912
7913 ;; ...No. The current parens aren't the function's arg list.
7914 (goto-char before-lparen))
7915
7916 (or (c-go-list-backward) ; backwards over [ .... ]
7917 (throw 'knr nil)))))))))
785eecbb
RS
7918
7919(defun c-skip-conditional ()
7920 ;; skip forward over conditional at point, including any predicate
7921 ;; statements in parentheses. No error checking is performed.
0386b551
AM
7922 ;;
7923 ;; This function might do hidden buffer changes.
0ec8351b
BW
7924 (c-forward-sexp (cond
7925 ;; else if()
a66cd3ee
MS
7926 ((looking-at (concat "\\<else"
7927 "\\([ \t\n]\\|\\\\\n\\)+"
7928 "if\\>\\([^_]\\|$\\)"))
7929 3)
0ec8351b 7930 ;; do, else, try, finally
a66cd3ee
MS
7931 ((looking-at (concat "\\<\\("
7932 "do\\|else\\|try\\|finally"
7933 "\\)\\>\\([^_]\\|$\\)"))
130c507e 7934 1)
ce8c7486 7935 ;; for, if, while, switch, catch, synchronized, foreach
0ec8351b 7936 (t 2))))
785eecbb 7937
a66cd3ee
MS
7938(defun c-after-conditional (&optional lim)
7939 ;; If looking at the token after a conditional then return the
7940 ;; position of its start, otherwise return nil.
0386b551
AM
7941 ;;
7942 ;; This function might do hidden buffer changes.
a66cd3ee 7943 (save-excursion
d9e94c22 7944 (and (zerop (c-backward-token-2 1 t lim))
a66cd3ee
MS
7945 (or (looking-at c-block-stmt-1-key)
7946 (and (eq (char-after) ?\()
d9e94c22 7947 (zerop (c-backward-token-2 1 t lim))
a66cd3ee
MS
7948 (looking-at c-block-stmt-2-key)))
7949 (point))))
7950
0386b551
AM
7951(defun c-after-special-operator-id (&optional lim)
7952 ;; If the point is after an operator identifier that isn't handled
7953 ;; like an ordinary symbol (i.e. like "operator =" in C++) then the
7954 ;; position of the start of that identifier is returned. nil is
7955 ;; returned otherwise. The point may be anywhere in the syntactic
7956 ;; whitespace after the last token of the operator identifier.
7957 ;;
7958 ;; This function might do hidden buffer changes.
7959 (save-excursion
7960 (and c-overloadable-operators-regexp
7961 (zerop (c-backward-token-2 1 nil lim))
7962 (looking-at c-overloadable-operators-regexp)
51c9af45 7963 (or (not c-opt-op-identifier-prefix)
0386b551
AM
7964 (and
7965 (zerop (c-backward-token-2 1 nil lim))
51c9af45 7966 (looking-at c-opt-op-identifier-prefix)))
0386b551
AM
7967 (point))))
7968
a66cd3ee
MS
7969(defsubst c-backward-to-block-anchor (&optional lim)
7970 ;; Assuming point is at a brace that opens a statement block of some
7971 ;; kind, move to the proper anchor point for that block. It might
7972 ;; need to be adjusted further by c-add-stmt-syntax, but the
7973 ;; position at return is suitable as start position for that
7974 ;; function.
0386b551
AM
7975 ;;
7976 ;; This function might do hidden buffer changes.
a66cd3ee
MS
7977 (unless (= (point) (c-point 'boi))
7978 (let ((start (c-after-conditional lim)))
7979 (if start
7980 (goto-char start)))))
7981
037558bf 7982(defsubst c-backward-to-decl-anchor (&optional lim)
a66cd3ee
MS
7983 ;; Assuming point is at a brace that opens the block of a top level
7984 ;; declaration of some kind, move to the proper anchor point for
7985 ;; that block.
0386b551
AM
7986 ;;
7987 ;; This function might do hidden buffer changes.
a66cd3ee 7988 (unless (= (point) (c-point 'boi))
037558bf 7989 (c-beginning-of-statement-1 lim)))
a66cd3ee 7990
ff959bab 7991(defun c-search-decl-header-end ()
a66cd3ee
MS
7992 ;; Search forward for the end of the "header" of the current
7993 ;; declaration. That's the position where the definition body
7994 ;; starts, or the first variable initializer, or the ending
7995 ;; semicolon. I.e. search forward for the closest following
7996 ;; (syntactically relevant) '{', '=' or ';' token. Point is left
7997 ;; _after_ the first found token, or at point-max if none is found.
0386b551
AM
7998 ;;
7999 ;; This function might do hidden buffer changes.
ff959bab
MS
8000
8001 (let ((base (point)))
8002 (if (c-major-mode-is 'c++-mode)
8003
8004 ;; In C++ we need to take special care to handle operator
8005 ;; tokens and those pesky template brackets.
8006 (while (and
8007 (c-syntactic-re-search-forward "[;{<=]" nil 'move t t)
8008 (or
8009 (c-end-of-current-token base)
8010 ;; Handle operator identifiers, i.e. ignore any
8011 ;; operator token preceded by "operator".
8012 (save-excursion
8013 (and (c-safe (c-backward-sexp) t)
51c9af45 8014 (looking-at c-opt-op-identifier-prefix)))
ff959bab
MS
8015 (and (eq (char-before) ?<)
8016 (c-with-syntax-table c++-template-syntax-table
8017 (if (c-safe (goto-char (c-up-list-forward (point))))
8018 t
8019 (goto-char (point-max))
8020 nil)))))
8021 (setq base (point)))
8022
8023 (while (and
8024 (c-syntactic-re-search-forward "[;{=]" nil 'move t t)
8025 (c-end-of-current-token base))
8026 (setq base (point))))))
a66cd3ee
MS
8027
8028(defun c-beginning-of-decl-1 (&optional lim)
8029 ;; Go to the beginning of the current declaration, or the beginning
8030 ;; of the previous one if already at the start of it. Point won't
0386b551 8031 ;; be moved out of any surrounding paren. Return a cons cell of the
a66cd3ee
MS
8032 ;; form (MOVE . KNR-POS). MOVE is like the return value from
8033 ;; `c-beginning-of-statement-1'. If point skipped over some K&R
8034 ;; style argument declarations (and they are to be recognized) then
8035 ;; KNR-POS is set to the start of the first such argument
8036 ;; declaration, otherwise KNR-POS is nil. If LIM is non-nil, it's a
8037 ;; position that bounds the backward search.
8038 ;;
8039 ;; NB: Cases where the declaration continues after the block, as in
8040 ;; "struct foo { ... } bar;", are currently recognized as two
8041 ;; declarations, e.g. "struct foo { ... }" and "bar;" in this case.
0386b551
AM
8042 ;;
8043 ;; This function might do hidden buffer changes.
a66cd3ee
MS
8044 (catch 'return
8045 (let* ((start (point))
d9e94c22 8046 (last-stmt-start (point))
0386b551 8047 (move (c-beginning-of-statement-1 lim nil t)))
a66cd3ee 8048
a66cd3ee
MS
8049 ;; `c-beginning-of-statement-1' stops at a block start, but we
8050 ;; want to continue if the block doesn't begin a top level
2a15eb73
MS
8051 ;; construct, i.e. if it isn't preceded by ';', '}', ':', bob,
8052 ;; or an open paren.
d9e94c22 8053 (let ((beg (point)) tentative-move)
51c9af45
AM
8054 ;; Go back one "statement" each time round the loop until we're just
8055 ;; after a ;, }, or :, or at BOB or the start of a macro or start of
8056 ;; an ObjC method. This will move over a multiple declaration whose
8057 ;; components are comma separated.
d9e94c22
MS
8058 (while (and
8059 ;; Must check with c-opt-method-key in ObjC mode.
8060 (not (and c-opt-method-key
8061 (looking-at c-opt-method-key)))
8062 (/= last-stmt-start (point))
8063 (progn
8064 (c-backward-syntactic-ws lim)
8065 (not (memq (char-before) '(?\; ?} ?: nil))))
2a15eb73
MS
8066 (save-excursion
8067 (backward-char)
8068 (not (looking-at "\\s(")))
d9e94c22
MS
8069 ;; Check that we don't move from the first thing in a
8070 ;; macro to its header.
8071 (not (eq (setq tentative-move
0386b551 8072 (c-beginning-of-statement-1 lim nil t))
d9e94c22
MS
8073 'macro)))
8074 (setq last-stmt-start beg
8075 beg (point)
8076 move tentative-move))
8077 (goto-char beg))
8078
8079 (when c-recognize-knr-p
8080 (let ((fallback-pos (point)) knr-argdecl-start)
8081 ;; Handle K&R argdecls. Back up after the "statement" jumped
8082 ;; over by `c-beginning-of-statement-1', unless it was the
8083 ;; function body, in which case we're sitting on the opening
8084 ;; brace now. Then test if we're in a K&R argdecl region and
8085 ;; that we started at the other side of the first argdecl in
8086 ;; it.
8087 (unless (eq (char-after) ?{)
8088 (goto-char last-stmt-start))
8089 (if (and (setq knr-argdecl-start (c-in-knr-argdecl lim))
8090 (< knr-argdecl-start start)
8091 (progn
8092 (goto-char knr-argdecl-start)
0386b551 8093 (not (eq (c-beginning-of-statement-1 lim nil t) 'macro))))
d9e94c22
MS
8094 (throw 'return
8095 (cons (if (eq (char-after fallback-pos) ?{)
8096 'previous
8097 'same)
8098 knr-argdecl-start))
8099 (goto-char fallback-pos))))
8100
51c9af45
AM
8101 ;; `c-beginning-of-statement-1' counts each brace block as a separate
8102 ;; statement, so the result will be 'previous if we've moved over any.
8103 ;; So change our result back to 'same if necessary.
8104 ;;
8105 ;; If they were brace list initializers we might not have moved over a
8106 ;; declaration boundary though, so change it to 'same if we've moved
8107 ;; past a '=' before '{', but not ';'. (This ought to be integrated
8108 ;; into `c-beginning-of-statement-1', so we avoid this extra pass which
8109 ;; potentially can search over a large amount of text.). Take special
8110 ;; pains not to get mislead by C++'s "operator=", and the like.
d9e94c22
MS
8111 (if (and (eq move 'previous)
8112 (c-with-syntax-table (if (c-major-mode-is 'c++-mode)
8113 c++-template-syntax-table
8114 (syntax-table))
8115 (save-excursion
51c9af45
AM
8116 (and
8117 (progn
8118 (while ; keep going back to "[;={"s until we either find
8119 ; no more, or get to one which isn't an "operator ="
8120 (and (c-syntactic-re-search-forward "[;={]" start t t t)
8121 (eq (char-before) ?=)
8122 c-overloadable-operators-regexp
8123 c-opt-op-identifier-prefix
8124 (save-excursion
8125 (eq (c-backward-token-2) 0)
8126 (looking-at c-overloadable-operators-regexp)
8127 (eq (c-backward-token-2) 0)
8128 (looking-at c-opt-op-identifier-prefix))))
8129 (eq (char-before) ?=))
8130 (c-syntactic-re-search-forward "[;{]" start t t)
8131 (eq (char-before) ?{)
8132 (c-safe (goto-char (c-up-list-forward (point))) t)
8133 (not (c-syntactic-re-search-forward ";" start t t))))))
d9e94c22
MS
8134 (cons 'same nil)
8135 (cons move nil)))))
a66cd3ee
MS
8136
8137(defun c-end-of-decl-1 ()
8138 ;; Assuming point is at the start of a declaration (as detected by
8139 ;; e.g. `c-beginning-of-decl-1'), go to the end of it. Unlike
8140 ;; `c-beginning-of-decl-1', this function handles the case when a
8141 ;; block is followed by identifiers in e.g. struct declarations in C
8142 ;; or C++. If a proper end was found then t is returned, otherwise
8143 ;; point is moved as far as possible within the current sexp and nil
8144 ;; is returned. This function doesn't handle macros; use
8145 ;; `c-end-of-macro' instead in those cases.
0386b551
AM
8146 ;;
8147 ;; This function might do hidden buffer changes.
ce8c7486 8148 (let ((start (point))
a66cd3ee
MS
8149 (decl-syntax-table (if (c-major-mode-is 'c++-mode)
8150 c++-template-syntax-table
8151 (syntax-table))))
8152 (catch 'return
8153 (c-search-decl-header-end)
8154
8155 (when (and c-recognize-knr-p
8156 (eq (char-before) ?\;)
8157 (c-in-knr-argdecl start))
8158 ;; Stopped at the ';' in a K&R argdecl section which is
8159 ;; detected using the same criteria as in
8160 ;; `c-beginning-of-decl-1'. Move to the following block
8161 ;; start.
d9e94c22 8162 (c-syntactic-re-search-forward "{" nil 'move t))
a66cd3ee
MS
8163
8164 (when (eq (char-before) ?{)
8165 ;; Encountered a block in the declaration. Jump over it.
8166 (condition-case nil
8167 (goto-char (c-up-list-forward (point)))
d9e94c22
MS
8168 (error (goto-char (point-max))
8169 (throw 'return nil)))
a66cd3ee
MS
8170 (if (or (not c-opt-block-decls-with-vars-key)
8171 (save-excursion
8172 (c-with-syntax-table decl-syntax-table
8173 (let ((lim (point)))
8174 (goto-char start)
b3cf7e18
MS
8175 (not (and
8176 ;; Check for `c-opt-block-decls-with-vars-key'
8177 ;; before the first paren.
8178 (c-syntactic-re-search-forward
d9e94c22 8179 (concat "[;=\(\[{]\\|\\("
b3cf7e18
MS
8180 c-opt-block-decls-with-vars-key
8181 "\\)")
d9e94c22 8182 lim t t t)
b3cf7e18
MS
8183 (match-beginning 1)
8184 (not (eq (char-before) ?_))
d9e94c22
MS
8185 ;; Check that the first following paren is
8186 ;; the block.
8187 (c-syntactic-re-search-forward "[;=\(\[{]"
8188 lim t t t)
b3cf7e18 8189 (eq (char-before) ?{)))))))
a66cd3ee
MS
8190 ;; The declaration doesn't have any of the
8191 ;; `c-opt-block-decls-with-vars' keywords in the
8192 ;; beginning, so it ends here at the end of the block.
8193 (throw 'return t)))
8194
8195 (c-with-syntax-table decl-syntax-table
8196 (while (progn
8197 (if (eq (char-before) ?\;)
8198 (throw 'return t))
d9e94c22 8199 (c-syntactic-re-search-forward ";" nil 'move t))))
a66cd3ee 8200 nil)))
ce8c7486 8201
0386b551
AM
8202(defun c-looking-at-decl-block (containing-sexp goto-start &optional limit)
8203 ;; Assuming the point is at an open brace, check if it starts a
8204 ;; block that contains another declaration level, i.e. that isn't a
8205 ;; statement block or a brace list, and if so return non-nil.
8206 ;;
8207 ;; If the check is successful, the return value is the start of the
8208 ;; keyword that tells what kind of construct it is, i.e. typically
8209 ;; what `c-decl-block-key' matched. Also, if GOTO-START is set then
8210 ;; the point will be at the start of the construct, before any
8211 ;; leading specifiers, otherwise it's at the returned position.
8212 ;;
8213 ;; The point is clobbered if the check is unsuccessful.
8214 ;;
8215 ;; CONTAINING-SEXP is the position of the open of the surrounding
8216 ;; paren, or nil if none.
8217 ;;
8218 ;; The optional LIMIT limits the backward search for the start of
8219 ;; the construct. It's assumed to be at a syntactically relevant
8220 ;; position.
8221 ;;
8222 ;; If any template arglists are found in the searched region before
8223 ;; the open brace, they get marked with paren syntax.
8224 ;;
8225 ;; This function might do hidden buffer changes.
8226
8227 (let ((open-brace (point)) kwd-start first-specifier-pos)
8228 (c-syntactic-skip-backward c-block-prefix-charset limit t)
8229
8230 (when (and c-recognize-<>-arglists
8231 (eq (char-before) ?>))
8232 ;; Could be at the end of a template arglist.
8233 (let ((c-parse-and-markup-<>-arglists t)
8234 (c-disallow-comma-in-<>-arglists
8235 (and containing-sexp
8236 (not (eq (char-after containing-sexp) ?{)))))
8237 (while (and
8238 (c-backward-<>-arglist nil limit)
8239 (progn
8240 (c-syntactic-skip-backward c-block-prefix-charset limit t)
8241 (eq (char-before) ?>))))))
8242
8243 ;; Note: Can't get bogus hits inside template arglists below since they
8244 ;; have gotten paren syntax above.
8245 (when (and
8246 ;; If `goto-start' is set we begin by searching for the
8247 ;; first possible position of a leading specifier list.
8248 ;; The `c-decl-block-key' search continues from there since
8249 ;; we know it can't match earlier.
8250 (if goto-start
8251 (when (c-syntactic-re-search-forward c-symbol-start
8252 open-brace t t)
8253 (goto-char (setq first-specifier-pos (match-beginning 0)))
8254 t)
8255 t)
8256
8257 (cond
8258 ((c-syntactic-re-search-forward c-decl-block-key open-brace t t t)
8259 (goto-char (setq kwd-start (match-beginning 0)))
8260 (or
8261
8262 ;; Found a keyword that can't be a type?
8263 (match-beginning 1)
8264
8265 ;; Can be a type too, in which case it's the return type of a
8266 ;; function (under the assumption that no declaration level
8267 ;; block construct starts with a type).
8268 (not (c-forward-type))
8269
8270 ;; Jumped over a type, but it could be a declaration keyword
8271 ;; followed by the declared identifier that we've jumped over
8272 ;; instead (e.g. in "class Foo {"). If it indeed is a type
8273 ;; then we should be at the declarator now, so check for a
8274 ;; valid declarator start.
8275 ;;
8276 ;; Note: This doesn't cope with the case when a declared
8277 ;; identifier is followed by e.g. '(' in a language where '('
8278 ;; also might be part of a declarator expression. Currently
8279 ;; there's no such language.
8280 (not (or (looking-at c-symbol-start)
8281 (looking-at c-type-decl-prefix-key)))))
8282
8283 ;; In Pike a list of modifiers may be followed by a brace
8284 ;; to make them apply to many identifiers. Note that the
8285 ;; match data will be empty on return in this case.
8286 ((and (c-major-mode-is 'pike-mode)
8287 (progn
8288 (goto-char open-brace)
8289 (= (c-backward-token-2) 0))
8290 (looking-at c-specifier-key)
8291 ;; Use this variant to avoid yet another special regexp.
8292 (c-keyword-member (c-keyword-sym (match-string 1))
8293 'c-modifier-kwds))
8294 (setq kwd-start (point))
8295 t)))
8296
8297 ;; Got a match.
8298
8299 (if goto-start
8300 ;; Back up over any preceding specifiers and their clauses
8301 ;; by going forward from `first-specifier-pos', which is the
8302 ;; earliest possible position where the specifier list can
8303 ;; start.
8304 (progn
8305 (goto-char first-specifier-pos)
8306
8307 (while (< (point) kwd-start)
8308 (if (looking-at c-symbol-key)
8309 ;; Accept any plain symbol token on the ground that
8310 ;; it's a specifier masked through a macro (just
8311 ;; like `c-forward-decl-or-cast-1' skip forward over
8312 ;; such tokens).
8313 ;;
8314 ;; Could be more restrictive wrt invalid keywords,
8315 ;; but that'd only occur in invalid code so there's
8316 ;; no use spending effort on it.
8317 (let ((end (match-end 0)))
8318 (unless (c-forward-keyword-clause 0)
8319 (goto-char end)
8320 (c-forward-syntactic-ws)))
8321
8322 ;; Can't parse a declaration preamble and is still
8323 ;; before `kwd-start'. That means `first-specifier-pos'
8324 ;; was in some earlier construct. Search again.
8325 (if (c-syntactic-re-search-forward c-symbol-start
8326 kwd-start 'move t)
8327 (goto-char (setq first-specifier-pos (match-beginning 0)))
8328 ;; Got no preamble before the block declaration keyword.
8329 (setq first-specifier-pos kwd-start))))
8330
8331 (goto-char first-specifier-pos))
8332 (goto-char kwd-start))
8333
8334 kwd-start)))
ce8c7486 8335
a66cd3ee 8336(defun c-search-uplist-for-classkey (paren-state)
0386b551
AM
8337 ;; Check if the closest containing paren sexp is a declaration
8338 ;; block, returning a 2 element vector in that case. Aref 0
8339 ;; contains the bufpos at boi of the class key line, and aref 1
8340 ;; contains the bufpos of the open brace. This function is an
8341 ;; obsolete wrapper for `c-looking-at-decl-block'.
8342 ;;
8343 ;; This function might do hidden buffer changes.
8344 (let ((open-paren-pos (c-most-enclosing-brace paren-state)))
8345 (when open-paren-pos
8346 (save-excursion
8347 (goto-char open-paren-pos)
8348 (when (and (eq (char-after) ?{)
8349 (c-looking-at-decl-block
8350 (c-safe-position open-paren-pos paren-state)
8351 nil))
8352 (back-to-indentation)
8353 (vector (point) open-paren-pos))))))
785eecbb 8354
020716e1
AM
8355(defmacro c-pull-open-brace (ps)
8356 ;; Pull the next open brace from PS (which has the form of paren-state),
8357 ;; skipping over any brace pairs. Returns NIL when PS is exhausted.
8358 `(progn
8359 (while (consp (car ,ps))
8360 (setq ,ps (cdr ,ps)))
8361 (prog1 (car ,ps)
8362 (setq ,ps (cdr ,ps)))))
8363
8364(defun c-most-enclosing-decl-block (paren-state)
8365 ;; Return the buffer position of the most enclosing decl-block brace (in the
8366 ;; sense of c-looking-at-decl-block) in the PAREN-STATE structure, or nil if
8367 ;; none was found.
8368 (let* ((open-brace (c-pull-open-brace paren-state))
8369 (next-open-brace (c-pull-open-brace paren-state)))
8370 (while (and open-brace
8371 (save-excursion
8372 (goto-char open-brace)
8373 (not (c-looking-at-decl-block next-open-brace nil))))
8374 (setq open-brace next-open-brace
8375 next-open-brace (c-pull-open-brace paren-state)))
8376 open-brace))
8377
9657183b
AM
8378(defun c-cheap-inside-bracelist-p (paren-state)
8379 ;; Return the position of the L-brace if point is inside a brace list
8380 ;; initialization of an array, etc. This is an approximate function,
8381 ;; designed for speed over accuracy. It will not find every bracelist, but
8382 ;; a non-nil result is reliable. We simply search for "= {" (naturally with
8383 ;; syntactic whitespace allowed). PAREN-STATE is the normal thing that it
8384 ;; is everywhere else.
8385 (let (b-pos)
8386 (save-excursion
8387 (while
8388 (and (setq b-pos (c-pull-open-brace paren-state))
8389 (progn (goto-char b-pos)
8390 (c-backward-sws)
8391 (c-backward-token-2)
8392 (not (looking-at "=")))))
8393 b-pos)))
8394
a66cd3ee 8395(defun c-inside-bracelist-p (containing-sexp paren-state)
785eecbb
RS
8396 ;; return the buffer position of the beginning of the brace list
8397 ;; statement if we're inside a brace list, otherwise return nil.
8398 ;; CONTAINING-SEXP is the buffer pos of the innermost containing
0386b551 8399 ;; paren. PAREN-STATE is the remainder of the state of enclosing
130c507e 8400 ;; braces
785eecbb
RS
8401 ;;
8402 ;; N.B.: This algorithm can potentially get confused by cpp macros
0386b551 8403 ;; placed in inconvenient locations. It's a trade-off we make for
785eecbb 8404 ;; speed.
0386b551
AM
8405 ;;
8406 ;; This function might do hidden buffer changes.
785eecbb 8407 (or
d9e94c22 8408 ;; This will pick up brace list declarations.
b2acd789
RS
8409 (c-safe
8410 (save-excursion
8411 (goto-char containing-sexp)
0ec8351b 8412 (c-forward-sexp -1)
b2acd789 8413 (let (bracepos)
d9e94c22 8414 (if (and (or (looking-at c-brace-list-key)
0ec8351b 8415 (progn (c-forward-sexp -1)
d9e94c22 8416 (looking-at c-brace-list-key)))
a66cd3ee 8417 (setq bracepos (c-down-list-forward (point)))
b2acd789
RS
8418 (not (c-crosses-statement-barrier-p (point)
8419 (- bracepos 2))))
8420 (point)))))
785eecbb
RS
8421 ;; this will pick up array/aggregate init lists, even if they are nested.
8422 (save-excursion
0ec8351b
BW
8423 (let ((class-key
8424 ;; Pike can have class definitions anywhere, so we must
8425 ;; check for the class key here.
8426 (and (c-major-mode-is 'pike-mode)
a66cd3ee
MS
8427 c-decl-block-key))
8428 bufpos braceassignp lim next-containing)
785eecbb
RS
8429 (while (and (not bufpos)
8430 containing-sexp)
a66cd3ee
MS
8431 (when paren-state
8432 (if (consp (car paren-state))
8433 (setq lim (cdr (car paren-state))
8434 paren-state (cdr paren-state))
8435 (setq lim (car paren-state)))
8436 (when paren-state
8437 (setq next-containing (car paren-state)
8438 paren-state (cdr paren-state))))
785eecbb 8439 (goto-char containing-sexp)
a66cd3ee
MS
8440 (if (c-looking-at-inexpr-block next-containing next-containing)
8441 ;; We're in an in-expression block of some kind. Do not
8442 ;; check nesting. We deliberately set the limit to the
8443 ;; containing sexp, so that c-looking-at-inexpr-block
8444 ;; doesn't check for an identifier before it.
0ec8351b
BW
8445 (setq containing-sexp nil)
8446 ;; see if the open brace is preceded by = or [...] in
8447 ;; this statement, but watch out for operator=
a66cd3ee 8448 (setq braceassignp 'dontknow)
d9e94c22 8449 (c-backward-token-2 1 t lim)
6393fef2 8450 ;; Checks to do only on the first sexp before the brace.
d9e94c22 8451 (when (and c-opt-inexpr-brace-list-key
6393fef2
RS
8452 (eq (char-after) ?\[))
8453 ;; In Java, an initialization brace list may follow
8454 ;; directly after "new Foo[]", so check for a "new"
8455 ;; earlier.
8456 (while (eq braceassignp 'dontknow)
8457 (setq braceassignp
d9e94c22
MS
8458 (cond ((/= (c-backward-token-2 1 t lim) 0) nil)
8459 ((looking-at c-opt-inexpr-brace-list-key) t)
6393fef2
RS
8460 ((looking-at "\\sw\\|\\s_\\|[.[]")
8461 ;; Carry on looking if this is an
8462 ;; identifier (may contain "." in Java)
8463 ;; or another "[]" sexp.
8464 'dontknow)
8465 (t nil)))))
8466 ;; Checks to do on all sexps before the brace, up to the
8467 ;; beginning of the statement.
8468 (while (eq braceassignp 'dontknow)
0ec8351b
BW
8469 (cond ((eq (char-after) ?\;)
8470 (setq braceassignp nil))
8471 ((and class-key
8472 (looking-at class-key))
8473 (setq braceassignp nil))
8474 ((eq (char-after) ?=)
8475 ;; We've seen a =, but must check earlier tokens so
8476 ;; that it isn't something that should be ignored.
8477 (setq braceassignp 'maybe)
8478 (while (and (eq braceassignp 'maybe)
d9e94c22 8479 (zerop (c-backward-token-2 1 t lim)))
0ec8351b
BW
8480 (setq braceassignp
8481 (cond
8482 ;; Check for operator =
51c9af45
AM
8483 ((and c-opt-op-identifier-prefix
8484 (looking-at c-opt-op-identifier-prefix))
0386b551 8485 nil)
130c507e
GM
8486 ;; Check for `<opchar>= in Pike.
8487 ((and (c-major-mode-is 'pike-mode)
8488 (or (eq (char-after) ?`)
8489 ;; Special case for Pikes
8490 ;; `[]=, since '[' is not in
8491 ;; the punctuation class.
8492 (and (eq (char-after) ?\[)
8493 (eq (char-before) ?`))))
8494 nil)
0ec8351b
BW
8495 ((looking-at "\\s.") 'maybe)
8496 ;; make sure we're not in a C++ template
8497 ;; argument assignment
a66cd3ee
MS
8498 ((and
8499 (c-major-mode-is 'c++-mode)
8500 (save-excursion
8501 (let ((here (point))
8502 (pos< (progn
8503 (skip-chars-backward "^<>")
8504 (point))))
8505 (and (eq (char-before) ?<)
8506 (not (c-crosses-statement-barrier-p
8507 pos< here))
8508 (not (c-in-literal))
8509 ))))
0ec8351b 8510 nil)
6393fef2
RS
8511 (t t))))))
8512 (if (and (eq braceassignp 'dontknow)
d9e94c22 8513 (/= (c-backward-token-2 1 t lim) 0))
6393fef2
RS
8514 (setq braceassignp nil)))
8515 (if (not braceassignp)
0ec8351b
BW
8516 (if (eq (char-after) ?\;)
8517 ;; Brace lists can't contain a semicolon, so we're done.
8518 (setq containing-sexp nil)
a66cd3ee
MS
8519 ;; Go up one level.
8520 (setq containing-sexp next-containing
8521 lim nil
8522 next-containing nil))
0ec8351b
BW
8523 ;; we've hit the beginning of the aggregate list
8524 (c-beginning-of-statement-1
a66cd3ee 8525 (c-most-enclosing-brace paren-state))
0ec8351b 8526 (setq bufpos (point))))
a66cd3ee 8527 )
785eecbb
RS
8528 bufpos))
8529 ))
8530
0ec8351b
BW
8531(defun c-looking-at-special-brace-list (&optional lim)
8532