* dired.el (dired-revert): If DIRED-DIRECTORY is a cons cell, call
[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 16211.
CommitLineData
785eecbb
RS
1;;; cc-engine.el --- core syntax guessing engine for CC mode
2
92ab3834 3;; Copyright (C) 1985, 1987, 1992, 1993, 1994, 1995, 1996, 1997, 1998,
114f9c96 4;; 1999, 2000, 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010
d7a0267c 5;; Free Software Foundation, Inc.
785eecbb 6
e309f66c
AM
7;; Authors: 2001- Alan Mackenzie
8;; 1998- Martin Stjernholm
d9e94c22 9;; 1992-1999 Barry A. Warsaw
5858f68c
GM
10;; 1987 Dave Detlefs
11;; 1987 Stewart Clamen
785eecbb 12;; 1985 Richard M. Stallman
0ec8351b 13;; Maintainer: bug-cc-mode@gnu.org
785eecbb 14;; Created: 22-Apr-1997 (split from cc-mode.el)
6430c434 15;; Version: See cc-mode.el
785eecbb
RS
16;; Keywords: c languages oop
17
18;; This file is part of GNU Emacs.
19
b1fc2b50 20;; GNU Emacs is free software: you can redistribute it and/or modify
785eecbb 21;; it under the terms of the GNU General Public License as published by
b1fc2b50
GM
22;; the Free Software Foundation, either version 3 of the License, or
23;; (at your option) any later version.
785eecbb
RS
24
25;; GNU Emacs is distributed in the hope that it will be useful,
26;; but WITHOUT ANY WARRANTY; without even the implied warranty of
27;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
28;; GNU General Public License for more details.
29
30;; You should have received a copy of the GNU General Public License
b1fc2b50 31;; along with GNU Emacs. If not, see <http://www.gnu.org/licenses/>.
785eecbb 32
3afbc435
PJ
33;;; Commentary:
34
a66cd3ee
MS
35;; The functions which have docstring documentation can be considered
36;; part of an API which other packages can use in CC Mode buffers.
37;; Otoh, undocumented functions and functions with the documentation
38;; in comments are considered purely internal and can change semantics
39;; or even disappear in the future.
40;;
41;; (This policy applies to CC Mode as a whole, not just this file. It
42;; probably also applies to many other Emacs packages, but here it's
43;; clearly spelled out.)
44
d9e94c22
MS
45;; Hidden buffer changes
46;;
47;; Various functions in CC Mode use text properties for caching and
48;; syntactic markup purposes, and those of them that might modify such
0386b551
AM
49;; properties but still don't modify the buffer in a visible way are
50;; said to do "hidden buffer changes". They should be used within
51;; `c-save-buffer-state' or a similar function that saves and restores
52;; buffer modifiedness, disables buffer change hooks, etc.
d9e94c22 53;;
0386b551
AM
54;; Interactive functions are assumed to not do hidden buffer changes,
55;; except in the specific parts of them that do real changes.
d9e94c22 56;;
0386b551
AM
57;; Lineup functions are assumed to do hidden buffer changes. They
58;; must not do real changes, though.
d9e94c22 59;;
0386b551
AM
60;; All other functions that do hidden buffer changes have that noted
61;; in their doc string or comment.
62;;
63;; The intention with this system is to avoid wrapping every leaf
64;; function that do hidden buffer changes inside
65;; `c-save-buffer-state'. It should be used as near the top of the
66;; interactive functions as possible.
67;;
68;; Functions called during font locking are allowed to do hidden
69;; buffer changes since the font-lock package run them in a context
70;; similar to `c-save-buffer-state' (in fact, that function is heavily
71;; inspired by `save-buffer-state' in the font-lock package).
d9e94c22
MS
72
73;; Use of text properties
74;;
75;; CC Mode uses several text properties internally to mark up various
76;; positions, e.g. to improve speed and to eliminate glitches in
77;; interactive refontification.
78;;
7bfc3fdb
MS
79;; Note: This doc is for internal use only. Other packages should not
80;; assume that these text properties are used as described here.
81;;
0ec1d2c5
AM
82;; 'category
83;; Used for "indirection". With its help, some other property can
84;; be cheaply and easily switched on or off everywhere it occurs.
85;;
d9e94c22 86;; 'syntax-table
c8018ede
AM
87;; Used to modify the syntax of some characters. It is used to
88;; mark the "<" and ">" of angle bracket parens with paren syntax, and
89;; to "hide" obtrusive characters in preprocessor lines.
d9e94c22
MS
90;;
91;; This property is used on single characters and is therefore
92;; always treated as front and rear nonsticky (or start and end open
93;; in XEmacs vocabulary). It's therefore installed on
94;; `text-property-default-nonsticky' if that variable exists (Emacs
95;; >= 21).
96;;
97;; 'c-is-sws and 'c-in-sws
98;; Used by `c-forward-syntactic-ws' and `c-backward-syntactic-ws' to
99;; speed them up. See the comment blurb before `c-put-is-sws'
100;; below for further details.
101;;
102;; 'c-type
103;; This property is used on single characters to mark positions with
0386b551
AM
104;; special syntactic relevance of various sorts. Its primary use is
105;; to avoid glitches when multiline constructs are refontified
d9e94c22
MS
106;; interactively (on font lock decoration level 3). It's cleared in
107;; a region before it's fontified and is then put on relevant chars
108;; in that region as they are encountered during the fontification.
109;; The value specifies the kind of position:
110;;
111;; 'c-decl-arg-start
112;; Put on the last char of the token preceding each declaration
113;; inside a declaration style arglist (typically in a function
114;; prototype).
115;;
116;; 'c-decl-end
117;; Put on the last char of the token preceding a declaration.
118;; This is used in cases where declaration boundaries can't be
119;; recognized simply by looking for a token like ";" or "}".
120;; `c-type-decl-end-used' must be set if this is used (see also
121;; `c-find-decl-spots').
122;;
123;; 'c-<>-arg-sep
124;; Put on the commas that separate arguments in angle bracket
125;; arglists like C++ template arglists.
126;;
127;; 'c-decl-id-start and 'c-decl-type-start
128;; Put on the last char of the token preceding each declarator
129;; in the declarator list of a declaration. They are also used
130;; between the identifiers cases like enum declarations.
131;; 'c-decl-type-start is used when the declarators are types,
132;; 'c-decl-id-start otherwise.
133;;
134;; 'c-awk-NL-prop
135;; Used in AWK mode to mark the various kinds of newlines. See
136;; cc-awk.el.
137
3afbc435
PJ
138;;; Code:
139
0ec8351b 140(eval-when-compile
51f606de 141 (let ((load-path
130c507e
GM
142 (if (and (boundp 'byte-compile-dest-file)
143 (stringp byte-compile-dest-file))
144 (cons (file-name-directory byte-compile-dest-file) load-path)
51f606de 145 load-path)))
d9e94c22 146 (load "cc-bytecomp" nil t)))
130c507e
GM
147
148(cc-require 'cc-defs)
d9e94c22 149(cc-require-when-compile 'cc-langs)
130c507e 150(cc-require 'cc-vars)
d9e94c22 151
130c507e
GM
152;; Silence the compiler.
153(cc-bytecomp-defun buffer-syntactic-context) ; XEmacs
0ec8351b 154
51f606de 155\f
d9e94c22
MS
156;; Make declarations for all the `c-lang-defvar' variables in cc-langs.
157
158(defmacro c-declare-lang-variables ()
159 `(progn
485fe977
RS
160 ,@(apply 'nconc
161 (mapcar (lambda (init)
162 `(,(if (elt init 2)
163 `(defvar ,(car init) nil ,(elt init 2))
164 `(defvar ,(car init) nil))
165 (make-variable-buffer-local ',(car init))))
166 (cdr c-lang-variable-inits)))))
d9e94c22
MS
167(c-declare-lang-variables)
168
169\f
170;;; Internal state variables.
171
172;; Internal state of hungry delete key feature
173(defvar c-hungry-delete-key nil)
174(make-variable-buffer-local 'c-hungry-delete-key)
175
0386b551
AM
176;; The electric flag (toggled by `c-toggle-electric-state').
177;; If t, electric actions (like automatic reindentation, and (if
178;; c-auto-newline is also set) auto newlining) will happen when an electric
179;; key like `{' is pressed (or an electric keyword like `else').
180(defvar c-electric-flag t)
181(make-variable-buffer-local 'c-electric-flag)
182
d9e94c22
MS
183;; Internal state of auto newline feature.
184(defvar c-auto-newline nil)
185(make-variable-buffer-local 'c-auto-newline)
186
0386b551 187;; Included in the mode line to indicate the active submodes.
cb694ab7
AM
188;; (defvar c-submode-indicators nil)
189;; (make-variable-buffer-local 'c-submode-indicators)
d9e94c22 190
a66cd3ee
MS
191(defun c-calculate-state (arg prevstate)
192 ;; Calculate the new state of PREVSTATE, t or nil, based on arg. If
193 ;; arg is nil or zero, toggle the state. If arg is negative, turn
194 ;; the state off, and if arg is positive, turn the state on
195 (if (or (not arg)
196 (zerop (setq arg (prefix-numeric-value arg))))
197 (not prevstate)
198 (> arg 0)))
199
d9e94c22 200;; Dynamically bound cache for `c-in-literal'.
130c507e 201(defvar c-in-literal-cache t)
d9e94c22 202
d9e94c22 203\f
037558bf
MS
204;; Basic handling of preprocessor directives.
205
206;; This is a dynamically bound cache used together with
207;; `c-query-macro-start' and `c-query-and-set-macro-start'. It only
208;; works as long as point doesn't cross a macro boundary.
209(defvar c-macro-start 'unknown)
210
211(defsubst c-query-and-set-macro-start ()
037558bf
MS
212 (if (symbolp c-macro-start)
213 (setq c-macro-start (save-excursion
0386b551
AM
214 (c-save-buffer-state ()
215 (and (c-beginning-of-macro)
216 (point)))))
037558bf
MS
217 c-macro-start))
218
219(defsubst c-query-macro-start ()
037558bf
MS
220 (if (symbolp c-macro-start)
221 (save-excursion
0386b551
AM
222 (c-save-buffer-state ()
223 (and (c-beginning-of-macro)
224 (point))))
037558bf
MS
225 c-macro-start))
226
227(defun c-beginning-of-macro (&optional lim)
228 "Go to the beginning of a preprocessor directive.
229Leave point at the beginning of the directive and return t if in one,
230otherwise return nil and leave point unchanged.
231
0386b551
AM
232Note that this function might do hidden buffer changes. See the
233comment at the start of cc-engine.el for more info."
037558bf
MS
234 (when c-opt-cpp-prefix
235 (let ((here (point)))
236 (save-restriction
237 (if lim (narrow-to-region lim (point-max)))
238 (beginning-of-line)
239 (while (eq (char-before (1- (point))) ?\\)
240 (forward-line -1))
241 (back-to-indentation)
242 (if (and (<= (point) here)
243 (looking-at c-opt-cpp-start))
244 t
245 (goto-char here)
246 nil)))))
247
248(defun c-end-of-macro ()
249 "Go to the end of a preprocessor directive.
0386b551
AM
250More accurately, move the point to the end of the closest following
251line that doesn't end with a line continuation backslash - no check is
252done that the point is inside a cpp directive to begin with.
037558bf 253
0386b551
AM
254Note that this function might do hidden buffer changes. See the
255comment at the start of cc-engine.el for more info."
037558bf
MS
256 (while (progn
257 (end-of-line)
258 (when (and (eq (char-before) ?\\)
259 (not (eobp)))
260 (forward-char)
261 t))))
262
0ec1d2c5
AM
263(defun c-syntactic-end-of-macro ()
264 ;; Go to the end of a CPP directive, or a "safe" pos just before.
265 ;;
266 ;; This is normally the end of the next non-escaped line. A "safe"
267 ;; position is one not within a string or comment. (The EOL on a line
268 ;; comment is NOT "safe").
269 ;;
270 ;; This function must only be called from the beginning of a CPP construct.
271 ;;
272 ;; Note that this function might do hidden buffer changes. See the comment
273 ;; at the start of cc-engine.el for more info.
274 (let* ((here (point))
275 (there (progn (c-end-of-macro) (point)))
276 (s (parse-partial-sexp here there)))
277 (while (and (or (nth 3 s) ; in a string
278 (nth 4 s)) ; in a comment (maybe at end of line comment)
279 (> there here)) ; No infinite loops, please.
280 (setq there (1- (nth 8 s)))
281 (setq s (parse-partial-sexp here there)))
282 (point)))
283
51c9af45
AM
284(defun c-forward-over-cpp-define-id ()
285 ;; Assuming point is at the "#" that introduces a preprocessor
286 ;; directive, it's moved forward to the end of the identifier which is
287 ;; "#define"d (or whatever c-opt-cpp-macro-define specifies). Non-nil
288 ;; is returned in this case, in all other cases nil is returned and
289 ;; point isn't moved.
290 ;;
291 ;; This function might do hidden buffer changes.
292 (when (and c-opt-cpp-macro-define-id
293 (looking-at c-opt-cpp-macro-define-id))
294 (goto-char (match-end 0))))
295
037558bf
MS
296(defun c-forward-to-cpp-define-body ()
297 ;; Assuming point is at the "#" that introduces a preprocessor
298 ;; directive, it's moved forward to the start of the definition body
0386b551
AM
299 ;; if it's a "#define" (or whatever c-opt-cpp-macro-define
300 ;; specifies). Non-nil is returned in this case, in all other cases
301 ;; nil is returned and point isn't moved.
302 ;;
303 ;; This function might do hidden buffer changes.
304 (when (and c-opt-cpp-macro-define-start
305 (looking-at c-opt-cpp-macro-define-start)
037558bf
MS
306 (not (= (match-end 0) (c-point 'eol))))
307 (goto-char (match-end 0))))
308
309\f
d9e94c22
MS
310;;; Basic utility functions.
311
0386b551 312(defun c-syntactic-content (from to paren-level)
d9e94c22
MS
313 ;; Return the given region as a string where all syntactic
314 ;; whitespace is removed or, where necessary, replaced with a single
0386b551
AM
315 ;; space. If PAREN-LEVEL is given then all parens in the region are
316 ;; collapsed to "()", "[]" etc.
317 ;;
318 ;; This function might do hidden buffer changes.
319
d9e94c22 320 (save-excursion
0386b551
AM
321 (save-restriction
322 (narrow-to-region from to)
323 (goto-char from)
324 (let* ((parts (list nil)) (tail parts) pos in-paren)
325
326 (while (re-search-forward c-syntactic-ws-start to t)
327 (goto-char (setq pos (match-beginning 0)))
328 (c-forward-syntactic-ws)
329 (if (= (point) pos)
330 (forward-char)
331
332 (when paren-level
333 (save-excursion
334 (setq in-paren (= (car (parse-partial-sexp from pos 1)) 1)
335 pos (point))))
336
337 (if (and (> pos from)
338 (< (point) to)
339 (looking-at "\\w\\|\\s_")
340 (save-excursion
341 (goto-char (1- pos))
342 (looking-at "\\w\\|\\s_")))
343 (progn
344 (setcdr tail (list (buffer-substring-no-properties from pos)
345 " "))
346 (setq tail (cddr tail)))
347 (setcdr tail (list (buffer-substring-no-properties from pos)))
348 (setq tail (cdr tail)))
349
350 (when in-paren
351 (when (= (car (parse-partial-sexp pos to -1)) -1)
352 (setcdr tail (list (buffer-substring-no-properties
353 (1- (point)) (point))))
354 (setq tail (cdr tail))))
355
356 (setq from (point))))
357
358 (setcdr tail (list (buffer-substring-no-properties from to)))
359 (apply 'concat (cdr parts))))))
360
361(defun c-shift-line-indentation (shift-amt)
362 ;; Shift the indentation of the current line with the specified
363 ;; amount (positive inwards). The buffer is modified only if
364 ;; SHIFT-AMT isn't equal to zero.
365 (let ((pos (- (point-max) (point)))
366 (c-macro-start c-macro-start)
367 tmp-char-inserted)
368 (if (zerop shift-amt)
369 nil
370 ;; If we're on an empty line inside a macro, we take the point
371 ;; to be at the current indentation and shift it to the
372 ;; appropriate column. This way we don't treat the extra
373 ;; whitespace out to the line continuation as indentation.
374 (when (and (c-query-and-set-macro-start)
375 (looking-at "[ \t]*\\\\$")
376 (save-excursion
377 (skip-chars-backward " \t")
378 (bolp)))
379 (insert ?x)
380 (backward-char)
381 (setq tmp-char-inserted t))
382 (unwind-protect
383 (let ((col (current-indentation)))
384 (delete-region (c-point 'bol) (c-point 'boi))
385 (beginning-of-line)
386 (indent-to (+ col shift-amt)))
387 (when tmp-char-inserted
388 (delete-char 1))))
389 ;; If initial point was within line's indentation and we're not on
390 ;; a line with a line continuation in a macro, position after the
391 ;; indentation. Else stay at same point in text.
392 (if (and (< (point) (c-point 'boi))
393 (not tmp-char-inserted))
394 (back-to-indentation)
395 (if (> (- (point-max) pos) (point))
396 (goto-char (- (point-max) pos))))))
d9e94c22
MS
397
398(defsubst c-keyword-sym (keyword)
399 ;; Return non-nil if the string KEYWORD is a known keyword. More
400 ;; precisely, the value is the symbol for the keyword in
401 ;; `c-keywords-obarray'.
402 (intern-soft keyword c-keywords-obarray))
403
404(defsubst c-keyword-member (keyword-sym lang-constant)
405 ;; Return non-nil if the symbol KEYWORD-SYM, as returned by
406 ;; `c-keyword-sym', is a member of LANG-CONSTANT, which is the name
407 ;; of a language constant that ends with "-kwds". If KEYWORD-SYM is
408 ;; nil then the result is nil.
409 (get keyword-sym lang-constant))
410
411;; String syntax chars, suitable for skip-syntax-(forward|backward).
412(defconst c-string-syntax (if (memq 'gen-string-delim c-emacs-features)
413 "\"|"
414 "\""))
415
0386b551 416;; Regexp matching string limit syntax.
d9e94c22
MS
417(defconst c-string-limit-regexp (if (memq 'gen-string-delim c-emacs-features)
418 "\\s\"\\|\\s|"
419 "\\s\""))
420
0386b551
AM
421;; Regexp matching WS followed by string limit syntax.
422(defconst c-ws*-string-limit-regexp
423 (concat "[ \t]*\\(" c-string-limit-regexp "\\)"))
424
d9e94c22
MS
425;; Holds formatted error strings for the few cases where parse errors
426;; are reported.
a66cd3ee 427(defvar c-parsing-error nil)
d9e94c22
MS
428(make-variable-buffer-local 'c-parsing-error)
429
430(defun c-echo-parsing-error (&optional quiet)
d9e94c22
MS
431 (when (and c-report-syntactic-errors c-parsing-error (not quiet))
432 (c-benign-error "%s" c-parsing-error))
433 c-parsing-error)
434
435;; Faces given to comments and string literals. This is used in some
436;; situations to speed up recognition; it isn't mandatory that font
437;; locking is in use. This variable is extended with the face in
438;; `c-doc-face-name' when fontification is activated in cc-fonts.el.
1e330469 439(defvar c-literal-faces
0386b551
AM
440 (append '(font-lock-comment-face font-lock-string-face)
441 (when (facep 'font-lock-comment-delimiter-face)
442 ;; New in Emacs 22.
443 '(font-lock-comment-delimiter-face))))
444
445(defsubst c-put-c-type-property (pos value)
446 ;; Put a c-type property with the given value at POS.
447 (c-put-char-property pos 'c-type value))
448
449(defun c-clear-c-type-property (from to value)
5a89f0a7 450 ;; Remove all occurrences of the c-type property that has the given
0386b551
AM
451 ;; value in the region between FROM and TO. VALUE is assumed to not
452 ;; be nil.
453 ;;
454 ;; Note: This assumes that c-type is put on single chars only; it's
455 ;; very inefficient if matching properties cover large regions.
456 (save-excursion
457 (goto-char from)
458 (while (progn
459 (when (eq (get-text-property (point) 'c-type) value)
460 (c-clear-char-property (point) 'c-type))
461 (goto-char (next-single-property-change (point) 'c-type nil to))
462 (< (point) to)))))
037558bf 463
d9e94c22
MS
464\f
465;; Some debug tools to visualize various special positions. This
466;; debug code isn't as portable as the rest of CC Mode.
467
468(cc-bytecomp-defun overlays-in)
469(cc-bytecomp-defun overlay-get)
470(cc-bytecomp-defun overlay-start)
471(cc-bytecomp-defun overlay-end)
472(cc-bytecomp-defun delete-overlay)
473(cc-bytecomp-defun overlay-put)
474(cc-bytecomp-defun make-overlay)
475
476(defun c-debug-add-face (beg end face)
477 (c-save-buffer-state ((overlays (overlays-in beg end)) overlay)
478 (while overlays
479 (setq overlay (car overlays)
480 overlays (cdr overlays))
481 (when (eq (overlay-get overlay 'face) face)
482 (setq beg (min beg (overlay-start overlay))
483 end (max end (overlay-end overlay)))
484 (delete-overlay overlay)))
485 (overlay-put (make-overlay beg end) 'face face)))
486
487(defun c-debug-remove-face (beg end face)
488 (c-save-buffer-state ((overlays (overlays-in beg end)) overlay
489 (ol-beg beg) (ol-end end))
490 (while overlays
491 (setq overlay (car overlays)
492 overlays (cdr overlays))
493 (when (eq (overlay-get overlay 'face) face)
494 (setq ol-beg (min ol-beg (overlay-start overlay))
495 ol-end (max ol-end (overlay-end overlay)))
496 (delete-overlay overlay)))
497 (when (< ol-beg beg)
498 (overlay-put (make-overlay ol-beg beg) 'face face))
499 (when (> ol-end end)
500 (overlay-put (make-overlay end ol-end) 'face face))))
501
502\f
503;; `c-beginning-of-statement-1' and accompanying stuff.
130c507e 504
64001211
RS
505;; KLUDGE ALERT: c-maybe-labelp is used to pass information between
506;; c-crosses-statement-barrier-p and c-beginning-of-statement-1. A
507;; better way should be implemented, but this will at least shut up
508;; the byte compiler.
0386b551 509(defvar c-maybe-labelp)
64001211 510
d9e94c22
MS
511;; New awk-compatible version of c-beginning-of-statement-1, ACM 2002/6/22
512
a66cd3ee
MS
513;; Macros used internally in c-beginning-of-statement-1 for the
514;; automaton actions.
515(defmacro c-bos-push-state ()
516 '(setq stack (cons (cons state saved-pos)
517 stack)))
518(defmacro c-bos-pop-state (&optional do-if-done)
519 `(if (setq state (car (car stack))
520 saved-pos (cdr (car stack))
521 stack (cdr stack))
522 t
523 ,do-if-done
524 (throw 'loop nil)))
525(defmacro c-bos-pop-state-and-retry ()
526 '(throw 'loop (setq state (car (car stack))
527 saved-pos (cdr (car stack))
528 ;; Throw nil if stack is empty, else throw non-nil.
529 stack (cdr stack))))
530(defmacro c-bos-save-pos ()
531 '(setq saved-pos (vector pos tok ptok pptok)))
532(defmacro c-bos-restore-pos ()
533 '(unless (eq (elt saved-pos 0) start)
534 (setq pos (elt saved-pos 0)
535 tok (elt saved-pos 1)
536 ptok (elt saved-pos 2)
537 pptok (elt saved-pos 3))
538 (goto-char pos)
539 (setq sym nil)))
540(defmacro c-bos-save-error-info (missing got)
541 `(setq saved-pos (vector pos ,missing ,got)))
542(defmacro c-bos-report-error ()
543 '(unless noerror
544 (setq c-parsing-error
545 (format "No matching `%s' found for `%s' on line %d"
546 (elt saved-pos 1)
547 (elt saved-pos 2)
548 (1+ (count-lines (point-min)
549 (c-point 'bol (elt saved-pos 0))))))))
550
551(defun c-beginning-of-statement-1 (&optional lim ignore-labels
552 noerror comma-delim)
553 "Move to the start of the current statement or declaration, or to
554the previous one if already at the beginning of one. Only
555statements/declarations on the same level are considered, i.e. don't
556move into or out of sexps (not even normal expression parentheses).
557
5a89f0a7 558If point is already at the earliest statement within braces or parens,
a85fd6da
AM
559this function doesn't move back into any whitespace preceding it; it
560returns 'same in this case.
561
d9e94c22
MS
562Stop at statement continuation tokens like \"else\", \"catch\",
563\"finally\" and the \"while\" in \"do ... while\" if the start point
564is within the continuation. If starting at such a token, move to the
565corresponding statement start. If at the beginning of a statement,
566move to the closest containing statement if there is any. This might
567also stop at a continuation clause.
a66cd3ee 568
0386b551
AM
569Labels are treated as part of the following statements if
570IGNORE-LABELS is non-nil. (FIXME: Doesn't work if we stop at a known
cb694ab7
AM
571statement start keyword.) Otherwise, each label is treated as a
572separate statement.
a66cd3ee 573
cb694ab7
AM
574Macros are ignored \(i.e. skipped over) unless point is within one, in
575which case the content of the macro is treated as normal code. Aside
576from any normal statement starts found in it, stop at the first token
577of the content in the macro, i.e. the expression of an \"#if\" or the
578start of the definition in a \"#define\". Also stop at start of
579macros before leaving them.
a66cd3ee 580
a85fd6da 581Return:
d28e7f28 582'label if stopped at a label or \"case...:\" or \"default:\";
a85fd6da
AM
583'same if stopped at the beginning of the current statement;
584'up if stepped to a containing statement;
585'previous if stepped to a preceding statement;
586'beginning if stepped from a statement continuation clause to
587 its start clause; or
588'macro if stepped to a macro start.
589Note that 'same and not 'label is returned if stopped at the same
590label without crossing the colon character.
a66cd3ee
MS
591
592LIM may be given to limit the search. If the search hits the limit,
593point will be left at the closest following token, or at the start
594position if that is less ('same is returned in this case).
595
596NOERROR turns off error logging to `c-parsing-error'.
597
cb694ab7
AM
598Normally only ';' and virtual semicolons are considered to delimit
599statements, but if COMMA-DELIM is non-nil then ',' is treated
600as a delimiter too.
0386b551
AM
601
602Note that this function might do hidden buffer changes. See the
603comment at the start of cc-engine.el for more info."
a66cd3ee 604
d9e94c22
MS
605 ;; The bulk of this function is a pushdown automaton that looks at statement
606 ;; boundaries and the tokens (such as "while") in c-opt-block-stmt-key. Its
607 ;; purpose is to keep track of nested statements, ensuring that such
5a89f0a7 608 ;; statements are skipped over in their entirety (somewhat akin to what C-M-p
d9e94c22 609 ;; does with nested braces/brackets/parentheses).
a66cd3ee
MS
610 ;;
611 ;; Note: The position of a boundary is the following token.
612 ;;
d9e94c22
MS
613 ;; Beginning with the current token (the one following point), move back one
614 ;; sexp at a time (where a sexp is, more or less, either a token or the
615 ;; entire contents of a brace/bracket/paren pair). Each time a statement
616 ;; boundary is crossed or a "while"-like token is found, update the state of
617 ;; the PDA. Stop at the beginning of a statement when the stack (holding
618 ;; nested statement info) is empty and the position has been moved.
619 ;;
620 ;; The following variables constitute the PDA:
621 ;;
622 ;; sym: This is either the "while"-like token (e.g. 'for) we've just
623 ;; scanned back over, 'boundary if we've just gone back over a
624 ;; statement boundary, or nil otherwise.
625 ;; state: takes one of the values (nil else else-boundary while
626 ;; while-boundary catch catch-boundary).
627 ;; nil means "no "while"-like token yet scanned".
628 ;; 'else, for example, means "just gone back over an else".
629 ;; 'else-boundary means "just gone back over a statement boundary
630 ;; immediately after having gone back over an else".
631 ;; saved-pos: A vector of either saved positions (tok ptok pptok, etc.) or
632 ;; of error reporting information.
633 ;; stack: The stack onto which the PDA pushes its state. Each entry
634 ;; consists of a saved value of state and saved-pos. An entry is
635 ;; pushed when we move back over a "continuation" token (e.g. else)
636 ;; and popped when we encounter the corresponding opening token
637 ;; (e.g. if).
638 ;;
639 ;;
b414f371 640 ;; The following diagram briefly outlines the PDA.
a66cd3ee
MS
641 ;;
642 ;; Common state:
d9e94c22
MS
643 ;; "else": Push state, goto state `else'.
644 ;; "while": Push state, goto state `while'.
645 ;; "catch" or "finally": Push state, goto state `catch'.
646 ;; boundary: Pop state.
a66cd3ee
MS
647 ;; other: Do nothing special.
648 ;;
d9e94c22
MS
649 ;; State `else':
650 ;; boundary: Goto state `else-boundary'.
651 ;; other: Error, pop state, retry token.
652 ;;
653 ;; State `else-boundary':
654 ;; "if": Pop state.
655 ;; boundary: Error, pop state.
656 ;; other: See common state.
657 ;;
658 ;; State `while':
659 ;; boundary: Save position, goto state `while-boundary'.
660 ;; other: Pop state, retry token.
661 ;;
662 ;; State `while-boundary':
663 ;; "do": Pop state.
664 ;; boundary: Restore position if it's not at start, pop state. [*see below]
665 ;; other: See common state.
666 ;;
667 ;; State `catch':
668 ;; boundary: Goto state `catch-boundary'.
669 ;; other: Error, pop state, retry token.
670 ;;
671 ;; State `catch-boundary':
672 ;; "try": Pop state.
673 ;; "catch": Goto state `catch'.
674 ;; boundary: Error, pop state.
675 ;; other: See common state.
676 ;;
677 ;; [*] In the `while-boundary' state, we had pushed a 'while state, and were
678 ;; searching for a "do" which would have opened a do-while. If we didn't
679 ;; find it, we discard the analysis done since the "while", go back to this
680 ;; token in the buffer and restart the scanning there, this time WITHOUT
681 ;; pushing the 'while state onto the stack.
682 ;;
a66cd3ee
MS
683 ;; In addition to the above there is some special handling of labels
684 ;; and macros.
685
686 (let ((case-fold-search nil)
687 (start (point))
688 macro-start
689 (delims (if comma-delim '(?\; ?,) '(?\;)))
690 (c-stmt-delim-chars (if comma-delim
691 c-stmt-delim-chars-with-comma
692 c-stmt-delim-chars))
d28e7f28 693 c-in-literal-cache c-maybe-labelp after-case:-pos saved
0386b551
AM
694 ;; Current position.
695 pos
696 ;; Position of last stmt boundary character (e.g. ;).
697 boundary-pos
698 ;; The position of the last sexp or bound that follows the
699 ;; first found colon, i.e. the start of the nonlabel part of
700 ;; the statement. It's `start' if a colon is found just after
701 ;; the start.
702 after-labels-pos
703 ;; Like `after-labels-pos', but the first such position inside
704 ;; a label, i.e. the start of the last label before the start
705 ;; of the nonlabel part of the statement.
706 last-label-pos
707 ;; The last position where a label is possible provided the
708 ;; statement started there. It's nil as long as no invalid
709 ;; label content has been found (according to
710 ;; `c-nonlabel-token-key'. It's `start' if no valid label
711 ;; content was found in the label. Note that we might still
712 ;; regard it a label if it starts with `c-label-kwds'.
713 label-good-pos
714 ;; Symbol just scanned back over (e.g. 'while or 'boundary).
715 ;; See above.
716 sym
717 ;; Current state in the automaton. See above.
718 state
719 ;; Current saved positions. See above.
720 saved-pos
721 ;; Stack of conses (state . saved-pos).
722 stack
723 ;; Regexp which matches "for", "if", etc.
724 (cond-key (or c-opt-block-stmt-key
a66cd3ee 725 "\\<\\>")) ; Matches nothing.
0386b551
AM
726 ;; Return value.
727 (ret 'same)
728 ;; Positions of the last three sexps or bounds we've stopped at.
729 tok ptok pptok)
a66cd3ee
MS
730
731 (save-restriction
732 (if lim (narrow-to-region lim (point-max)))
733
734 (if (save-excursion
735 (and (c-beginning-of-macro)
736 (/= (point) start)))
737 (setq macro-start (point)))
738
d9e94c22 739 ;; Try to skip back over unary operator characters, to register
a66cd3ee
MS
740 ;; that we've moved.
741 (while (progn
742 (setq pos (point))
0386b551
AM
743 (c-backward-syntactic-ws)
744 ;; Protect post-++/-- operators just before a virtual semicolon.
745 (and (not (c-at-vsemi-p))
746 (/= (skip-chars-backward "-+!*&~@`#") 0))))
d9e94c22
MS
747
748 ;; Skip back over any semicolon here. If it was a bare semicolon, we're
0386b551 749 ;; done. Later on we ignore the boundaries for statements that don't
d9e94c22
MS
750 ;; contain any sexp. The only thing that is affected is that the error
751 ;; checking is a little less strict, and we really don't bother.
a66cd3ee
MS
752 (if (and (memq (char-before) delims)
753 (progn (forward-char -1)
754 (setq saved (point))
0386b551 755 (c-backward-syntactic-ws)
a66cd3ee
MS
756 (or (memq (char-before) delims)
757 (memq (char-before) '(?: nil))
d9e94c22 758 (eq (char-syntax (char-before)) ?\()
0386b551 759 (c-at-vsemi-p))))
a66cd3ee
MS
760 (setq ret 'previous
761 pos saved)
762
763 ;; Begin at start and not pos to detect macros if we stand
764 ;; directly after the #.
765 (goto-char start)
766 (if (looking-at "\\<\\|\\W")
767 ;; Record this as the first token if not starting inside it.
768 (setq tok start))
769
d9e94c22
MS
770 ;; The following while loop goes back one sexp (balanced parens,
771 ;; etc. with contents, or symbol or suchlike) each iteration. This
772 ;; movement is accomplished with a call to scan-sexps approx 130 lines
773 ;; below.
a66cd3ee
MS
774 (while
775 (catch 'loop ;; Throw nil to break, non-nil to continue.
776 (cond
a66cd3ee 777 ((save-excursion
0386b551 778 (and macro-start ; Always NIL for AWK.
a66cd3ee
MS
779 (progn (skip-chars-backward " \t")
780 (eq (char-before) ?#))
781 (progn (setq saved (1- (point)))
782 (beginning-of-line)
783 (not (eq (char-before (1- (point))) ?\\)))
d9e94c22 784 (looking-at c-opt-cpp-start)
a66cd3ee
MS
785 (progn (skip-chars-forward " \t")
786 (eq (point) saved))))
787 (goto-char saved)
788 (if (and (c-forward-to-cpp-define-body)
789 (progn (c-forward-syntactic-ws start)
790 (< (point) start)))
791 ;; Stop at the first token in the content of the macro.
792 (setq pos (point)
793 ignore-labels t) ; Avoid the label check on exit.
794 (setq pos saved
795 ret 'macro
796 ignore-labels t))
797 (throw 'loop nil))
798
d9e94c22
MS
799 ;; Do a round through the automaton if we've just passed a
800 ;; statement boundary or passed a "while"-like token.
a66cd3ee
MS
801 ((or sym
802 (and (looking-at cond-key)
803 (setq sym (intern (match-string 1)))))
804
805 (when (and (< pos start) (null stack))
806 (throw 'loop nil))
807
d9e94c22
MS
808 ;; The PDA state handling.
809 ;;
037558bf 810 ;; Refer to the description of the PDA in the opening
d9e94c22
MS
811 ;; comments. In the following OR form, the first leaf
812 ;; attempts to handles one of the specific actions detailed
813 ;; (e.g., finding token "if" whilst in state `else-boundary').
814 ;; We drop through to the second leaf (which handles common
815 ;; state) if no specific handler is found in the first cond.
816 ;; If a parsing error is detected (e.g. an "else" with no
817 ;; preceding "if"), we throw to the enclosing catch.
818 ;;
819 ;; Note that the (eq state 'else) means
820 ;; "we've just passed an else", NOT "we're looking for an
821 ;; else".
a66cd3ee
MS
822 (or (cond
823 ((eq state 'else)
824 (if (eq sym 'boundary)
825 (setq state 'else-boundary)
826 (c-bos-report-error)
827 (c-bos-pop-state-and-retry)))
828
829 ((eq state 'else-boundary)
830 (cond ((eq sym 'if)
831 (c-bos-pop-state (setq ret 'beginning)))
832 ((eq sym 'boundary)
833 (c-bos-report-error)
834 (c-bos-pop-state))))
835
836 ((eq state 'while)
837 (if (and (eq sym 'boundary)
838 ;; Since this can cause backtracking we do a
839 ;; little more careful analysis to avoid it:
840 ;; If there's a label in front of the while
841 ;; it can't be part of a do-while.
842 (not after-labels-pos))
843 (progn (c-bos-save-pos)
844 (setq state 'while-boundary))
d9e94c22 845 (c-bos-pop-state-and-retry))) ; Can't be a do-while
a66cd3ee
MS
846
847 ((eq state 'while-boundary)
848 (cond ((eq sym 'do)
849 (c-bos-pop-state (setq ret 'beginning)))
d9e94c22
MS
850 ((eq sym 'boundary) ; isn't a do-while
851 (c-bos-restore-pos) ; the position of the while
852 (c-bos-pop-state)))) ; no longer searching for do.
a66cd3ee
MS
853
854 ((eq state 'catch)
855 (if (eq sym 'boundary)
856 (setq state 'catch-boundary)
857 (c-bos-report-error)
858 (c-bos-pop-state-and-retry)))
859
860 ((eq state 'catch-boundary)
861 (cond
862 ((eq sym 'try)
863 (c-bos-pop-state (setq ret 'beginning)))
864 ((eq sym 'catch)
865 (setq state 'catch))
866 ((eq sym 'boundary)
867 (c-bos-report-error)
868 (c-bos-pop-state)))))
869
d9e94c22
MS
870 ;; This is state common. We get here when the previous
871 ;; cond statement found no particular state handler.
a66cd3ee 872 (cond ((eq sym 'boundary)
d9e94c22
MS
873 ;; If we have a boundary at the start
874 ;; position we push a frame to go to the
875 ;; previous statement.
876 (if (>= pos start)
877 (c-bos-push-state)
878 (c-bos-pop-state)))
a66cd3ee
MS
879 ((eq sym 'else)
880 (c-bos-push-state)
881 (c-bos-save-error-info 'if 'else)
882 (setq state 'else))
883 ((eq sym 'while)
0386b551
AM
884 ;; Is this a real while, or a do-while?
885 ;; The next `when' triggers unless we are SURE that
886 ;; the `while' is not the tailend of a `do-while'.
a66cd3ee 887 (when (or (not pptok)
d9e94c22 888 (memq (char-after pptok) delims)
0386b551
AM
889 ;; The following kludge is to prevent
890 ;; infinite recursion when called from
891 ;; c-awk-after-if-for-while-condition-p,
892 ;; or the like.
893 (and (eq (point) start)
894 (c-vsemi-status-unknown-p))
895 (c-at-vsemi-p pptok))
a66cd3ee
MS
896 ;; Since this can cause backtracking we do a
897 ;; little more careful analysis to avoid it: If
0386b551
AM
898 ;; the while isn't followed by a (possibly
899 ;; virtual) semicolon it can't be a do-while.
a66cd3ee
MS
900 (c-bos-push-state)
901 (setq state 'while)))
902 ((memq sym '(catch finally))
903 (c-bos-push-state)
904 (c-bos-save-error-info 'try sym)
905 (setq state 'catch))))
906
907 (when c-maybe-labelp
908 ;; We're either past a statement boundary or at the
909 ;; start of a statement, so throw away any label data
910 ;; for the previous one.
911 (setq after-labels-pos nil
912 last-label-pos nil
913 c-maybe-labelp nil))))
914
d9e94c22
MS
915 ;; Step to the previous sexp, but not if we crossed a
916 ;; boundary, since that doesn't consume an sexp.
a66cd3ee
MS
917 (if (eq sym 'boundary)
918 (setq ret 'previous)
d9e94c22
MS
919
920 ;; HERE IS THE SINGLE PLACE INSIDE THE PDA LOOP WHERE WE MOVE
0386b551
AM
921 ;; BACKWARDS THROUGH THE SOURCE.
922
923 ;; This is typically fast with the caching done by
924 ;; c-(backward|forward)-sws.
925 (c-backward-syntactic-ws)
926
927 (let ((before-sws-pos (point))
928 ;; Set as long as we have to continue jumping by sexps.
929 ;; It's the position to use as end in the next round.
930 sexp-loop-continue-pos
931 ;; The end position of the area to search for statement
932 ;; barriers in this round.
933 (sexp-loop-end-pos pos))
934
cb694ab7 935 ;; The following while goes back one sexp per iteration.
0386b551
AM
936 (while
937 (progn
938 (unless (c-safe (c-backward-sexp) t)
939 ;; Give up if we hit an unbalanced block. Since the
940 ;; stack won't be empty the code below will report a
941 ;; suitable error.
a66cd3ee 942 (throw 'loop nil))
0386b551
AM
943
944 ;; Check if the sexp movement crossed a statement or
945 ;; declaration boundary. But first modify the point
946 ;; so that `c-crosses-statement-barrier-p' only looks
947 ;; at the non-sexp chars following the sexp.
948 (save-excursion
949 (when (setq
950 boundary-pos
951 (cond
952 ((if macro-start
953 nil
954 (save-excursion
955 (when (c-beginning-of-macro)
956 ;; Set continuation position in case
957 ;; `c-crosses-statement-barrier-p'
958 ;; doesn't detect anything below.
959 (setq sexp-loop-continue-pos (point)))))
960 ;; If the sexp movement took us into a
961 ;; macro then there were only some non-sexp
962 ;; chars after it. Skip out of the macro
963 ;; to analyze them but not the non-sexp
964 ;; chars that might be inside the macro.
965 (c-end-of-macro)
966 (c-crosses-statement-barrier-p
967 (point) sexp-loop-end-pos))
968
969 ((and
970 (eq (char-after) ?{)
971 (not (c-looking-at-inexpr-block lim nil t)))
972 ;; Passed a block sexp. That's a boundary
973 ;; alright.
974 (point))
975
976 ((looking-at "\\s\(")
977 ;; Passed some other paren. Only analyze
978 ;; the non-sexp chars after it.
979 (goto-char (1+ (c-down-list-backward
980 before-sws-pos)))
981 ;; We're at a valid token start position
982 ;; (outside the `save-excursion') if
983 ;; `c-crosses-statement-barrier-p' failed.
984 (c-crosses-statement-barrier-p
985 (point) sexp-loop-end-pos))
986
987 (t
988 ;; Passed a symbol sexp or line
989 ;; continuation. It doesn't matter that
990 ;; it's included in the analyzed region.
991 (if (c-crosses-statement-barrier-p
992 (point) sexp-loop-end-pos)
993 t
994 ;; If it was a line continuation then we
995 ;; have to continue looping.
996 (if (looking-at "\\\\$")
997 (setq sexp-loop-continue-pos (point)))
998 nil))))
999
1000 (setq pptok ptok
1001 ptok tok
1002 tok boundary-pos
1003 sym 'boundary)
1004 ;; Like a C "continue". Analyze the next sexp.
1005 (throw 'loop t)))
1006
d28e7f28 1007 sexp-loop-continue-pos) ; End of "go back a sexp" loop condition.
0386b551
AM
1008 (goto-char sexp-loop-continue-pos)
1009 (setq sexp-loop-end-pos sexp-loop-continue-pos
1010 sexp-loop-continue-pos nil))))
a66cd3ee
MS
1011
1012 ;; ObjC method def?
1013 (when (and c-opt-method-key
1014 (setq saved (c-in-method-def-p)))
1015 (setq pos saved
1016 ignore-labels t) ; Avoid the label check on exit.
1017 (throw 'loop nil))
1018
0386b551
AM
1019 ;; Handle labels.
1020 (unless (eq ignore-labels t)
1021 (when (numberp c-maybe-labelp)
cb694ab7
AM
1022 ;; `c-crosses-statement-barrier-p' has found a colon, so we
1023 ;; might be in a label now. Have we got a real label
1024 ;; (including a case label) or something like C++'s "public:"?
d28e7f28
AM
1025 ;; A case label might use an expression rather than a token.
1026 (setq after-case:-pos (or tok start))
1027 (if (looking-at c-nonlabel-token-key) ; e.g. "while" or "'a'"
1028 (setq c-maybe-labelp nil)
1029 (if after-labels-pos ; Have we already encountered a label?
1030 (if (not last-label-pos)
1031 (setq last-label-pos (or tok start)))
1032 (setq after-labels-pos (or tok start)))
1033 (setq c-maybe-labelp t
1034 label-good-pos nil))) ; bogus "label"
cb694ab7
AM
1035
1036 (when (and (not label-good-pos) ; i.e. no invalid "label"'s yet
1037 ; been found.
1038 (looking-at c-nonlabel-token-key)) ; e.g. "while :"
0386b551
AM
1039 ;; We're in a potential label and it's the first
1040 ;; time we've found something that isn't allowed in
1041 ;; one.
1042 (setq label-good-pos (or tok start))))
1043
1044 ;; We've moved back by a sexp, so update the token positions.
a66cd3ee
MS
1045 (setq sym nil
1046 pptok ptok
1047 ptok tok
1048 tok (point)
d9e94c22 1049 pos tok))) ; Not nil (for the while loop).
a66cd3ee
MS
1050
1051 ;; If the stack isn't empty there might be errors to report.
1052 (while stack
1053 (if (and (vectorp saved-pos) (eq (length saved-pos) 3))
1054 (c-bos-report-error))
1055 (setq saved-pos (cdr (car stack))
1056 stack (cdr stack)))
1057
1058 (when (and (eq ret 'same)
1059 (not (memq sym '(boundary ignore nil))))
1060 ;; Need to investigate closer whether we've crossed
1061 ;; between a substatement and its containing statement.
1062 (if (setq saved (if (looking-at c-block-stmt-1-key)
1063 ptok
1064 pptok))
1065 (cond ((> start saved) (setq pos saved))
1066 ((= start saved) (setq ret 'up)))))
1067
0386b551
AM
1068 (when (and (not ignore-labels)
1069 (eq c-maybe-labelp t)
d9e94c22 1070 (not (eq ret 'beginning))
0386b551
AM
1071 after-labels-pos
1072 (or (not label-good-pos)
1073 (<= label-good-pos pos)
1074 (progn
1075 (goto-char (if (and last-label-pos
1076 (< last-label-pos start))
1077 last-label-pos
1078 pos))
1079 (looking-at c-label-kwds-regexp))))
a66cd3ee
MS
1080 ;; We're in a label. Maybe we should step to the statement
1081 ;; after it.
1082 (if (< after-labels-pos start)
1083 (setq pos after-labels-pos)
1084 (setq ret 'label)
0386b551
AM
1085 (if (and last-label-pos (< last-label-pos start))
1086 ;; Might have jumped over several labels. Go to the last one.
a66cd3ee
MS
1087 (setq pos last-label-pos)))))
1088
d28e7f28 1089 ;; Have we got "case <expression>:"?
a66cd3ee 1090 (goto-char pos)
d28e7f28
AM
1091 (when (and after-case:-pos
1092 (not (eq ret 'beginning))
1093 (looking-at c-case-kwds-regexp))
1094 (if (< after-case:-pos start)
0000ee90
AM
1095 (setq pos after-case:-pos))
1096 (if (eq ret 'same)
1097 (setq ret 'label)))
d28e7f28
AM
1098
1099 ;; Skip over the unary operators that can start the statement.
a66cd3ee 1100 (while (progn
0386b551
AM
1101 (c-backward-syntactic-ws)
1102 ;; protect AWK post-inc/decrement operators, etc.
1103 (and (not (c-at-vsemi-p (point)))
1104 (/= (skip-chars-backward "-+!*&~@`#") 0)))
a66cd3ee
MS
1105 (setq pos (point)))
1106 (goto-char pos)
1107 ret)))
785eecbb 1108
785eecbb 1109(defun c-crosses-statement-barrier-p (from to)
a66cd3ee
MS
1110 "Return non-nil if buffer positions FROM to TO cross one or more
1111statement or declaration boundaries. The returned value is actually
d9e94c22
MS
1112the position of the earliest boundary char. FROM must not be within
1113a string or comment.
a66cd3ee
MS
1114
1115The variable `c-maybe-labelp' is set to the position of the first `:' that
1116might start a label (i.e. not part of `::' and not preceded by `?'). If a
0386b551
AM
1117single `?' is found, then `c-maybe-labelp' is cleared.
1118
1119For AWK, a statement which is terminated by an EOL (not a \; or a }) is
1120regarded as having a \"virtual semicolon\" immediately after the last token on
b414f371 1121the line. If this virtual semicolon is _at_ from, the function recognizes it.
0386b551
AM
1122
1123Note that this function might do hidden buffer changes. See the
1124comment at the start of cc-engine.el for more info."
a66cd3ee
MS
1125 (let ((skip-chars c-stmt-delim-chars)
1126 lit-range)
1127 (save-excursion
1128 (catch 'done
1129 (goto-char from)
1130 (while (progn (skip-chars-forward skip-chars to)
785eecbb 1131 (< (point) to))
0386b551
AM
1132 (cond
1133 ((setq lit-range (c-literal-limits from)) ; Have we landed in a string/comment?
1134 (goto-char (cdr lit-range)))
1135 ((eq (char-after) ?:)
1136 (forward-char)
1137 (if (and (eq (char-after) ?:)
1138 (< (point) to))
1139 ;; Ignore scope operators.
1140 (forward-char)
1141 (setq c-maybe-labelp (1- (point)))))
1142 ((eq (char-after) ??)
1143 ;; A question mark. Can't be a label, so stop
1144 ;; looking for more : and ?.
1145 (setq c-maybe-labelp nil
1146 skip-chars (substring c-stmt-delim-chars 0 -2)))
1147 ((memq (char-after) '(?# ?\n ?\r)) ; A virtual semicolon?
1148 (if (and (eq (char-before) ?\\) (memq (char-after) '(?\n ?\r)))
1149 (backward-char))
1150 (skip-chars-backward " \t" from)
1151 (if (c-at-vsemi-p)
1152 (throw 'done (point))
1153 (forward-line)))
1154 (t (throw 'done (point)))))
1155 ;; In trailing space after an as yet undetected virtual semicolon?
1156 (c-backward-syntactic-ws from)
1157 (if (and (< (point) to)
1158 (c-at-vsemi-p))
1159 (point)
1160 nil)))))
1161
1162(defun c-at-statement-start-p ()
1163 "Return non-nil if the point is at the first token in a statement
1164or somewhere in the syntactic whitespace before it.
1165
1166A \"statement\" here is not restricted to those inside code blocks.
1167Any kind of declaration-like construct that occur outside function
1168bodies is also considered a \"statement\".
1169
1170Note that this function might do hidden buffer changes. See the
1171comment at the start of cc-engine.el for more info."
1172
1173 (save-excursion
1174 (let ((end (point))
1175 c-maybe-labelp)
1176 (c-syntactic-skip-backward (substring c-stmt-delim-chars 1) nil t)
1177 (or (bobp)
1178 (eq (char-before) ?})
1179 (and (eq (char-before) ?{)
1180 (not (and c-special-brace-lists
1181 (progn (backward-char)
1182 (c-looking-at-special-brace-list)))))
1183 (c-crosses-statement-barrier-p (point) end)))))
1184
1185(defun c-at-expression-start-p ()
1186 "Return non-nil if the point is at the first token in an expression or
1187statement, or somewhere in the syntactic whitespace before it.
1188
1189An \"expression\" here is a bit different from the normal language
1190grammar sense: It's any sequence of expression tokens except commas,
1191unless they are enclosed inside parentheses of some kind. Also, an
1192expression never continues past an enclosing parenthesis, but it might
1193contain parenthesis pairs of any sort except braces.
1194
1195Since expressions never cross statement boundaries, this function also
1196recognizes statement beginnings, just like `c-at-statement-start-p'.
1197
1198Note that this function might do hidden buffer changes. See the
1199comment at the start of cc-engine.el for more info."
1200
1201 (save-excursion
1202 (let ((end (point))
1203 (c-stmt-delim-chars c-stmt-delim-chars-with-comma)
1204 c-maybe-labelp)
1205 (c-syntactic-skip-backward (substring c-stmt-delim-chars 1) nil t)
1206 (or (bobp)
1207 (memq (char-before) '(?{ ?}))
1208 (save-excursion (backward-char)
1209 (looking-at "\\s("))
1210 (c-crosses-statement-barrier-p (point) end)))))
785eecbb
RS
1211
1212\f
d9e94c22
MS
1213;; A set of functions that covers various idiosyncrasies in
1214;; implementations of `forward-comment'.
1215
1216;; Note: Some emacsen considers incorrectly that any line comment
1217;; ending with a backslash continues to the next line. I can't think
1218;; of any way to work around that in a reliable way without changing
1219;; the buffer, though. Suggestions welcome. ;) (No, temporarily
1220;; changing the syntax for backslash doesn't work since we must treat
1221;; escapes in string literals correctly.)
1222
1223(defun c-forward-single-comment ()
1224 "Move forward past whitespace and the closest following comment, if any.
1225Return t if a comment was found, nil otherwise. In either case, the
1226point is moved past the following whitespace. Line continuations,
1227i.e. a backslashes followed by line breaks, are treated as whitespace.
1228The line breaks that end line comments are considered to be the
1229comment enders, so the point will be put on the beginning of the next
1230line if it moved past a line comment.
1231
1232This function does not do any hidden buffer changes."
1233
1234 (let ((start (point)))
1235 (when (looking-at "\\([ \t\n\r\f\v]\\|\\\\[\n\r]\\)+")
1236 (goto-char (match-end 0)))
1237
1238 (when (forward-comment 1)
1239 (if (eobp)
1240 ;; Some emacsen (e.g. XEmacs 21) return t when moving
1241 ;; forwards at eob.
1242 nil
1243
1244 ;; Emacs includes the ending newline in a b-style (c++)
1245 ;; comment, but XEmacs doesn't. We depend on the Emacs
1246 ;; behavior (which also is symmetric).
1247 (if (and (eolp) (elt (parse-partial-sexp start (point)) 7))
1248 (condition-case nil (forward-char 1)))
1249
1250 t))))
1251
1252(defsubst c-forward-comments ()
1253 "Move forward past all following whitespace and comments.
1254Line continuations, i.e. a backslashes followed by line breaks, are
1255treated as whitespace.
1256
0386b551
AM
1257Note that this function might do hidden buffer changes. See the
1258comment at the start of cc-engine.el for more info."
d9e94c22
MS
1259
1260 (while (or
1261 ;; If forward-comment in at least XEmacs 21 is given a large
1262 ;; positive value, it'll loop all the way through if it hits
1263 ;; eob.
1264 (and (forward-comment 5)
1265 ;; Some emacsen (e.g. XEmacs 21) return t when moving
1266 ;; forwards at eob.
1267 (not (eobp)))
1268
1269 (when (looking-at "\\\\[\n\r]")
1270 (forward-char 2)
1271 t))))
1272
1273(defun c-backward-single-comment ()
1274 "Move backward past whitespace and the closest preceding comment, if any.
1275Return t if a comment was found, nil otherwise. In either case, the
1276point is moved past the preceding whitespace. Line continuations,
1277i.e. a backslashes followed by line breaks, are treated as whitespace.
1278The line breaks that end line comments are considered to be the
1279comment enders, so the point cannot be at the end of the same line to
1280move over a line comment.
1281
1282This function does not do any hidden buffer changes."
1283
1284 (let ((start (point)))
1285 ;; When we got newline terminated comments, forward-comment in all
1286 ;; supported emacsen so far will stop at eol of each line not
1287 ;; ending with a comment when moving backwards. This corrects for
1288 ;; that, and at the same time handles line continuations.
1289 (while (progn
1290 (skip-chars-backward " \t\n\r\f\v")
1291 (and (looking-at "[\n\r]")
0386b551 1292 (eq (char-before) ?\\)))
d9e94c22
MS
1293 (backward-char))
1294
1295 (if (bobp)
1296 ;; Some emacsen (e.g. Emacs 19.34) return t when moving
1297 ;; backwards at bob.
1298 nil
1299
1300 ;; Leave point after the closest following newline if we've
1301 ;; backed up over any above, since forward-comment won't move
1302 ;; backward over a line comment if point is at the end of the
1303 ;; same line.
1304 (re-search-forward "\\=\\s *[\n\r]" start t)
1305
1306 (if (if (forward-comment -1)
1307 (if (eolp)
1308 ;; If forward-comment above succeeded and we're at eol
1309 ;; then the newline we moved over above didn't end a
1310 ;; line comment, so we give it another go.
1311 (forward-comment -1)
1312 t))
1313
1314 ;; Emacs <= 20 and XEmacs move back over the closer of a
1315 ;; block comment that lacks an opener.
1316 (if (looking-at "\\*/")
1317 (progn (forward-char 2) nil)
1318 t)))))
1319
1320(defsubst c-backward-comments ()
1321 "Move backward past all preceding whitespace and comments.
1322Line continuations, i.e. a backslashes followed by line breaks, are
1323treated as whitespace. The line breaks that end line comments are
1324considered to be the comment enders, so the point cannot be at the end
0386b551
AM
1325of the same line to move over a line comment. Unlike
1326c-backward-syntactic-ws, this function doesn't move back over
1327preprocessor directives.
d9e94c22 1328
0386b551
AM
1329Note that this function might do hidden buffer changes. See the
1330comment at the start of cc-engine.el for more info."
d9e94c22
MS
1331
1332 (let ((start (point)))
1333 (while (and
0386b551 1334 ;; `forward-comment' in some emacsen (e.g. XEmacs 21.4)
d9e94c22
MS
1335 ;; return t when moving backwards at bob.
1336 (not (bobp))
1337
1338 (if (forward-comment -1)
1339 (if (looking-at "\\*/")
1340 ;; Emacs <= 20 and XEmacs move back over the
1341 ;; closer of a block comment that lacks an opener.
1342 (progn (forward-char 2) nil)
1343 t)
1344
1345 ;; XEmacs treats line continuations as whitespace but
1346 ;; only in the backward direction, which seems a bit
1347 ;; odd. Anyway, this is necessary for Emacs.
1348 (when (and (looking-at "[\n\r]")
1349 (eq (char-before) ?\\)
1350 (< (point) start))
1351 (backward-char)
1352 t))))))
1353
1354\f
d9e94c22 1355;; Tools for skipping over syntactic whitespace.
a66cd3ee 1356
d9e94c22
MS
1357;; The following functions use text properties to cache searches over
1358;; large regions of syntactic whitespace. It works as follows:
1359;;
1360;; o If a syntactic whitespace region contains anything but simple
1361;; whitespace (i.e. space, tab and line breaks), the text property
1362;; `c-in-sws' is put over it. At places where we have stopped
1363;; within that region there's also a `c-is-sws' text property.
1364;; That since there typically are nested whitespace inside that
1365;; must be handled separately, e.g. whitespace inside a comment or
1366;; cpp directive. Thus, from one point with `c-is-sws' it's safe
1367;; to jump to another point with that property within the same
1368;; `c-in-sws' region. It can be likened to a ladder where
1369;; `c-in-sws' marks the bars and `c-is-sws' the rungs.
1370;;
1371;; o The `c-is-sws' property is put on the simple whitespace chars at
1372;; a "rung position" and also maybe on the first following char.
1373;; As many characters as can be conveniently found in this range
1374;; are marked, but no assumption can be made that the whole range
1375;; is marked (it could be clobbered by later changes, for
1376;; instance).
1377;;
1378;; Note that some part of the beginning of a sequence of simple
1379;; whitespace might be part of the end of a preceding line comment
1380;; or cpp directive and must not be considered part of the "rung".
1381;; Such whitespace is some amount of horizontal whitespace followed
1382;; by a newline. In the case of cpp directives it could also be
1383;; two newlines with horizontal whitespace between them.
1384;;
1385;; The reason to include the first following char is to cope with
1386;; "rung positions" that doesn't have any ordinary whitespace. If
1387;; `c-is-sws' is put on a token character it does not have
1388;; `c-in-sws' set simultaneously. That's the only case when that
1389;; can occur, and the reason for not extending the `c-in-sws'
1390;; region to cover it is that the `c-in-sws' region could then be
1391;; accidentally merged with a following one if the token is only
1392;; one character long.
1393;;
1394;; o On buffer changes the `c-in-sws' and `c-is-sws' properties are
1395;; removed in the changed region. If the change was inside
1396;; syntactic whitespace that means that the "ladder" is broken, but
1397;; a later call to `c-forward-sws' or `c-backward-sws' will use the
1398;; parts on either side and use an ordinary search only to "repair"
1399;; the gap.
1400;;
1401;; Special care needs to be taken if a region is removed: If there
1402;; are `c-in-sws' on both sides of it which do not connect inside
1403;; the region then they can't be joined. If e.g. a marked macro is
1404;; broken, syntactic whitespace inside the new text might be
1405;; marked. If those marks would become connected with the old
1406;; `c-in-sws' range around the macro then we could get a ladder
1407;; with one end outside the macro and the other at some whitespace
1408;; within it.
1409;;
1410;; The main motivation for this system is to increase the speed in
1411;; skipping over the large whitespace regions that can occur at the
1412;; top level in e.g. header files that contain a lot of comments and
1413;; cpp directives. For small comments inside code it's probably
1414;; slower than using `forward-comment' straightforwardly, but speed is
1415;; not a significant factor there anyway.
1416
1417; (defface c-debug-is-sws-face
1418; '((t (:background "GreenYellow")))
1419; "Debug face to mark the `c-is-sws' property.")
1420; (defface c-debug-in-sws-face
1421; '((t (:underline t)))
1422; "Debug face to mark the `c-in-sws' property.")
1423
1424; (defun c-debug-put-sws-faces ()
1425; ;; Put the sws debug faces on all the `c-is-sws' and `c-in-sws'
1426; ;; properties in the buffer.
1427; (interactive)
1428; (save-excursion
0386b551 1429; (c-save-buffer-state (in-face)
d9e94c22
MS
1430; (goto-char (point-min))
1431; (setq in-face (if (get-text-property (point) 'c-is-sws)
1432; (point)))
1433; (while (progn
1434; (goto-char (next-single-property-change
1435; (point) 'c-is-sws nil (point-max)))
1436; (if in-face
1437; (progn
1438; (c-debug-add-face in-face (point) 'c-debug-is-sws-face)
1439; (setq in-face nil))
1440; (setq in-face (point)))
1441; (not (eobp))))
1442; (goto-char (point-min))
1443; (setq in-face (if (get-text-property (point) 'c-in-sws)
1444; (point)))
1445; (while (progn
1446; (goto-char (next-single-property-change
1447; (point) 'c-in-sws nil (point-max)))
1448; (if in-face
1449; (progn
1450; (c-debug-add-face in-face (point) 'c-debug-in-sws-face)
1451; (setq in-face nil))
1452; (setq in-face (point)))
1453; (not (eobp)))))))
1454
1455(defmacro c-debug-sws-msg (&rest args)
1456 ;;`(message ,@args)
1457 )
1458
1459(defmacro c-put-is-sws (beg end)
0386b551 1460 ;; This macro does a hidden buffer change.
d9e94c22
MS
1461 `(let ((beg ,beg) (end ,end))
1462 (put-text-property beg end 'c-is-sws t)
1463 ,@(when (facep 'c-debug-is-sws-face)
1464 `((c-debug-add-face beg end 'c-debug-is-sws-face)))))
1465
1466(defmacro c-put-in-sws (beg end)
0386b551 1467 ;; This macro does a hidden buffer change.
d9e94c22
MS
1468 `(let ((beg ,beg) (end ,end))
1469 (put-text-property beg end 'c-in-sws t)
1470 ,@(when (facep 'c-debug-is-sws-face)
1471 `((c-debug-add-face beg end 'c-debug-in-sws-face)))))
1472
1473(defmacro c-remove-is-sws (beg end)
0386b551 1474 ;; This macro does a hidden buffer change.
d9e94c22
MS
1475 `(let ((beg ,beg) (end ,end))
1476 (remove-text-properties beg end '(c-is-sws nil))
1477 ,@(when (facep 'c-debug-is-sws-face)
1478 `((c-debug-remove-face beg end 'c-debug-is-sws-face)))))
1479
1480(defmacro c-remove-in-sws (beg end)
0386b551 1481 ;; This macro does a hidden buffer change.
d9e94c22
MS
1482 `(let ((beg ,beg) (end ,end))
1483 (remove-text-properties beg end '(c-in-sws nil))
1484 ,@(when (facep 'c-debug-is-sws-face)
1485 `((c-debug-remove-face beg end 'c-debug-in-sws-face)))))
1486
1487(defmacro c-remove-is-and-in-sws (beg end)
0386b551 1488 ;; This macro does a hidden buffer change.
d9e94c22
MS
1489 `(let ((beg ,beg) (end ,end))
1490 (remove-text-properties beg end '(c-is-sws nil c-in-sws nil))
1491 ,@(when (facep 'c-debug-is-sws-face)
1492 `((c-debug-remove-face beg end 'c-debug-is-sws-face)
1493 (c-debug-remove-face beg end 'c-debug-in-sws-face)))))
1494
1495(defsubst c-invalidate-sws-region-after (beg end)
1496 ;; Called from `after-change-functions'. Note that if
1497 ;; `c-forward-sws' or `c-backward-sws' are used outside
1498 ;; `c-save-buffer-state' or similar then this will remove the cache
1499 ;; properties right after they're added.
0386b551
AM
1500 ;;
1501 ;; This function does hidden buffer changes.
d9e94c22
MS
1502
1503 (save-excursion
1504 ;; Adjust the end to remove the properties in any following simple
1505 ;; ws up to and including the next line break, if there is any
1506 ;; after the changed region. This is necessary e.g. when a rung
1507 ;; marked empty line is converted to a line comment by inserting
1508 ;; "//" before the line break. In that case the line break would
1509 ;; keep the rung mark which could make a later `c-backward-sws'
1510 ;; move into the line comment instead of over it.
1511 (goto-char end)
1512 (skip-chars-forward " \t\f\v")
1513 (when (and (eolp) (not (eobp)))
1514 (setq end (1+ (point)))))
1515
1516 (when (and (= beg end)
1517 (get-text-property beg 'c-in-sws)
2a15eb73 1518 (> beg (point-min))
d9e94c22
MS
1519 (get-text-property (1- beg) 'c-in-sws))
1520 ;; Ensure that an `c-in-sws' range gets broken. Note that it isn't
1521 ;; safe to keep a range that was continuous before the change. E.g:
1522 ;;
1523 ;; #define foo
1524 ;; \
1525 ;; bar
1526 ;;
1527 ;; There can be a "ladder" between "#" and "b". Now, if the newline
1528 ;; after "foo" is removed then "bar" will become part of the cpp
1529 ;; directive instead of a syntactically relevant token. In that
1530 ;; case there's no longer syntactic ws from "#" to "b".
1531 (setq beg (1- beg)))
1532
1533 (c-debug-sws-msg "c-invalidate-sws-region-after [%s..%s]" beg end)
1534 (c-remove-is-and-in-sws beg end))
1535
1536(defun c-forward-sws ()
1537 ;; Used by `c-forward-syntactic-ws' to implement the unbounded search.
0386b551
AM
1538 ;;
1539 ;; This function might do hidden buffer changes.
d9e94c22
MS
1540
1541 (let (;; `rung-pos' is set to a position as early as possible in the
1542 ;; unmarked part of the simple ws region.
1543 (rung-pos (point)) next-rung-pos rung-end-pos last-put-in-sws-pos
1544 rung-is-marked next-rung-is-marked simple-ws-end
1545 ;; `safe-start' is set when it's safe to cache the start position.
1546 ;; It's not set if we've initially skipped over comments and line
1547 ;; continuations since we might have gone out through the end of a
1548 ;; macro then. This provision makes `c-forward-sws' not populate the
1549 ;; cache in the majority of cases, but otoh is `c-backward-sws' by far
1550 ;; more common.
1551 safe-start)
1552
1553 ;; Skip simple ws and do a quick check on the following character to see
1554 ;; if it's anything that can't start syntactic ws, so we can bail out
1555 ;; early in the majority of cases when there just are a few ws chars.
1556 (skip-chars-forward " \t\n\r\f\v")
1557 (when (looking-at c-syntactic-ws-start)
1558
1559 (setq rung-end-pos (min (1+ (point)) (point-max)))
1560 (if (setq rung-is-marked (text-property-any rung-pos rung-end-pos
1561 'c-is-sws t))
1562 ;; Find the last rung position to avoid setting properties in all
1563 ;; the cases when the marked rung is complete.
1564 ;; (`next-single-property-change' is certain to move at least one
1565 ;; step forward.)
1566 (setq rung-pos (1- (next-single-property-change
1567 rung-is-marked 'c-is-sws nil rung-end-pos)))
1568 ;; Got no marked rung here. Since the simple ws might have started
1569 ;; inside a line comment or cpp directive we must set `rung-pos' as
1570 ;; high as possible.
1571 (setq rung-pos (point)))
1572
1573 (while
1574 (progn
1575 (while
1576 (when (and rung-is-marked
1577 (get-text-property (point) 'c-in-sws))
1578
1579 ;; The following search is the main reason that `c-in-sws'
1580 ;; and `c-is-sws' aren't combined to one property.
1581 (goto-char (next-single-property-change
1582 (point) 'c-in-sws nil (point-max)))
1583 (unless (get-text-property (point) 'c-is-sws)
1584 ;; If the `c-in-sws' region extended past the last
1585 ;; `c-is-sws' char we have to go back a bit.
1586 (or (get-text-property (1- (point)) 'c-is-sws)
1587 (goto-char (previous-single-property-change
1588 (point) 'c-is-sws)))
1589 (backward-char))
1590
1591 (c-debug-sws-msg
1592 "c-forward-sws cached move %s -> %s (max %s)"
1593 rung-pos (point) (point-max))
1594
1595 (setq rung-pos (point))
1596 (and (> (skip-chars-forward " \t\n\r\f\v") 0)
1597 (not (eobp))))
1598
1599 ;; We'll loop here if there is simple ws after the last rung.
1600 ;; That means that there's been some change in it and it's
1601 ;; possible that we've stepped into another ladder, so extend
1602 ;; the previous one to join with it if there is one, and try to
1603 ;; use the cache again.
1604 (c-debug-sws-msg
1605 "c-forward-sws extending rung with [%s..%s] (max %s)"
1606 (1+ rung-pos) (1+ (point)) (point-max))
1607 (unless (get-text-property (point) 'c-is-sws)
1608 ;; Remove any `c-in-sws' property from the last char of
1609 ;; the rung before we mark it with `c-is-sws', so that we
1610 ;; won't connect with the remains of a broken "ladder".
1611 (c-remove-in-sws (point) (1+ (point))))
1612 (c-put-is-sws (1+ rung-pos)
1613 (1+ (point)))
1614 (c-put-in-sws rung-pos
1615 (setq rung-pos (point)
1616 last-put-in-sws-pos rung-pos)))
1617
1618 (setq simple-ws-end (point))
1619 (c-forward-comments)
1620
1621 (cond
1622 ((/= (point) simple-ws-end)
1623 ;; Skipped over comments. Don't cache at eob in case the buffer
1624 ;; is narrowed.
1625 (not (eobp)))
1626
1627 ((save-excursion
1628 (and c-opt-cpp-prefix
1629 (looking-at c-opt-cpp-start)
1630 (progn (skip-chars-backward " \t")
1631 (bolp))
1632 (or (bobp)
1633 (progn (backward-char)
1634 (not (eq (char-before) ?\\))))))
1635 ;; Skip a preprocessor directive.
1636 (end-of-line)
1637 (while (and (eq (char-before) ?\\)
1638 (= (forward-line 1) 0))
1639 (end-of-line))
1640 (forward-line 1)
1641 (setq safe-start t)
1642 ;; Don't cache at eob in case the buffer is narrowed.
1643 (not (eobp)))))
1644
1645 ;; We've searched over a piece of non-white syntactic ws. See if this
1646 ;; can be cached.
1647 (setq next-rung-pos (point))
1648 (skip-chars-forward " \t\n\r\f\v")
1649 (setq rung-end-pos (min (1+ (point)) (point-max)))
1650
1651 (if (or
1652 ;; Cache if we haven't skipped comments only, and if we started
1653 ;; either from a marked rung or from a completely uncached
1654 ;; position.
1655 (and safe-start
1656 (or rung-is-marked
1657 (not (get-text-property simple-ws-end 'c-in-sws))))
1658
1659 ;; See if there's a marked rung in the encountered simple ws. If
1660 ;; so then we can cache, unless `safe-start' is nil. Even then
1661 ;; we need to do this to check if the cache can be used for the
1662 ;; next step.
1663 (and (setq next-rung-is-marked
1664 (text-property-any next-rung-pos rung-end-pos
1665 'c-is-sws t))
1666 safe-start))
b2acd789 1667
0ec8351b 1668 (progn
d9e94c22
MS
1669 (c-debug-sws-msg
1670 "c-forward-sws caching [%s..%s] - [%s..%s] (max %s)"
1671 rung-pos (1+ simple-ws-end) next-rung-pos rung-end-pos
1672 (point-max))
1673
1674 ;; Remove the properties for any nested ws that might be cached.
1675 ;; Only necessary for `c-is-sws' since `c-in-sws' will be set
1676 ;; anyway.
1677 (c-remove-is-sws (1+ simple-ws-end) next-rung-pos)
1678 (unless (and rung-is-marked (= rung-pos simple-ws-end))
1679 (c-put-is-sws rung-pos
1680 (1+ simple-ws-end))
1681 (setq rung-is-marked t))
1682 (c-put-in-sws rung-pos
1683 (setq rung-pos (point)
1684 last-put-in-sws-pos rung-pos))
1685 (unless (get-text-property (1- rung-end-pos) 'c-is-sws)
1686 ;; Remove any `c-in-sws' property from the last char of
1687 ;; the rung before we mark it with `c-is-sws', so that we
1688 ;; won't connect with the remains of a broken "ladder".
1689 (c-remove-in-sws (1- rung-end-pos) rung-end-pos))
1690 (c-put-is-sws next-rung-pos
1691 rung-end-pos))
1692
1693 (c-debug-sws-msg
1694 "c-forward-sws not caching [%s..%s] - [%s..%s] (max %s)"
1695 rung-pos (1+ simple-ws-end) next-rung-pos rung-end-pos
1696 (point-max))
1697
1698 ;; Set `rung-pos' for the next rung. It's the same thing here as
1699 ;; initially, except that the rung position is set as early as
1700 ;; possible since we can't be in the ending ws of a line comment or
1701 ;; cpp directive now.
1702 (if (setq rung-is-marked next-rung-is-marked)
1703 (setq rung-pos (1- (next-single-property-change
1704 rung-is-marked 'c-is-sws nil rung-end-pos)))
1705 (setq rung-pos next-rung-pos))
1706 (setq safe-start t)))
1707
1708 ;; Make sure that the newly marked `c-in-sws' region doesn't connect to
1709 ;; another one after the point (which might occur when editing inside a
1710 ;; comment or macro).
1711 (when (eq last-put-in-sws-pos (point))
1712 (cond ((< last-put-in-sws-pos (point-max))
1713 (c-debug-sws-msg
1714 "c-forward-sws clearing at %s for cache separation"
1715 last-put-in-sws-pos)
1716 (c-remove-in-sws last-put-in-sws-pos
1717 (1+ last-put-in-sws-pos)))
1718 (t
1719 ;; If at eob we have to clear the last character before the end
1720 ;; instead since the buffer might be narrowed and there might
1721 ;; be a `c-in-sws' after (point-max). In this case it's
1722 ;; necessary to clear both properties.
1723 (c-debug-sws-msg
1724 "c-forward-sws clearing thoroughly at %s for cache separation"
1725 (1- last-put-in-sws-pos))
1726 (c-remove-is-and-in-sws (1- last-put-in-sws-pos)
1727 last-put-in-sws-pos))))
1728 )))
b2acd789 1729
d9e94c22
MS
1730(defun c-backward-sws ()
1731 ;; Used by `c-backward-syntactic-ws' to implement the unbounded search.
0386b551
AM
1732 ;;
1733 ;; This function might do hidden buffer changes.
d9e94c22
MS
1734
1735 (let (;; `rung-pos' is set to a position as late as possible in the unmarked
1736 ;; part of the simple ws region.
1737 (rung-pos (point)) next-rung-pos last-put-in-sws-pos
1738 rung-is-marked simple-ws-beg cmt-skip-pos)
1739
1740 ;; Skip simple horizontal ws and do a quick check on the preceding
1741 ;; character to see if it's anying that can't end syntactic ws, so we can
1742 ;; bail out early in the majority of cases when there just are a few ws
1743 ;; chars. Newlines are complicated in the backward direction, so we can't
1744 ;; skip over them.
1745 (skip-chars-backward " \t\f")
1746 (when (and (not (bobp))
1747 (save-excursion
1748 (backward-char)
1749 (looking-at c-syntactic-ws-end)))
1750
1751 ;; Try to find a rung position in the simple ws preceding point, so that
1752 ;; we can get a cache hit even if the last bit of the simple ws has
1753 ;; changed recently.
1754 (setq simple-ws-beg (point))
1755 (skip-chars-backward " \t\n\r\f\v")
1756 (if (setq rung-is-marked (text-property-any
1757 (point) (min (1+ rung-pos) (point-max))
1758 'c-is-sws t))
1759 ;; `rung-pos' will be the earliest marked position, which means that
1760 ;; there might be later unmarked parts in the simple ws region.
1761 ;; It's not worth the effort to fix that; the last part of the
1762 ;; simple ws is also typically edited often, so it could be wasted.
1763 (goto-char (setq rung-pos rung-is-marked))
1764 (goto-char simple-ws-beg))
1765
1766 (while
1767 (progn
1768 (while
1769 (when (and rung-is-marked
1770 (not (bobp))
1771 (get-text-property (1- (point)) 'c-in-sws))
1772
1773 ;; The following search is the main reason that `c-in-sws'
1774 ;; and `c-is-sws' aren't combined to one property.
1775 (goto-char (previous-single-property-change
1776 (point) 'c-in-sws nil (point-min)))
1777 (unless (get-text-property (point) 'c-is-sws)
1778 ;; If the `c-in-sws' region extended past the first
1779 ;; `c-is-sws' char we have to go forward a bit.
1780 (goto-char (next-single-property-change
1781 (point) 'c-is-sws)))
1782
1783 (c-debug-sws-msg
1784 "c-backward-sws cached move %s <- %s (min %s)"
1785 (point) rung-pos (point-min))
1786
1787 (setq rung-pos (point))
1788 (if (and (< (min (skip-chars-backward " \t\f\v")
1789 (progn
1790 (setq simple-ws-beg (point))
1791 (skip-chars-backward " \t\n\r\f\v")))
1792 0)
1793 (setq rung-is-marked
1794 (text-property-any (point) rung-pos
1795 'c-is-sws t)))
1796 t
1797 (goto-char simple-ws-beg)
1798 nil))
1799
1800 ;; We'll loop here if there is simple ws before the first rung.
1801 ;; That means that there's been some change in it and it's
1802 ;; possible that we've stepped into another ladder, so extend
1803 ;; the previous one to join with it if there is one, and try to
1804 ;; use the cache again.
1805 (c-debug-sws-msg
1806 "c-backward-sws extending rung with [%s..%s] (min %s)"
1807 rung-is-marked rung-pos (point-min))
1808 (unless (get-text-property (1- rung-pos) 'c-is-sws)
1809 ;; Remove any `c-in-sws' property from the last char of
1810 ;; the rung before we mark it with `c-is-sws', so that we
1811 ;; won't connect with the remains of a broken "ladder".
1812 (c-remove-in-sws (1- rung-pos) rung-pos))
1813 (c-put-is-sws rung-is-marked
1814 rung-pos)
1815 (c-put-in-sws rung-is-marked
1816 (1- rung-pos))
1817 (setq rung-pos rung-is-marked
1818 last-put-in-sws-pos rung-pos))
1819
1820 (c-backward-comments)
1821 (setq cmt-skip-pos (point))
a66cd3ee 1822
d9e94c22
MS
1823 (cond
1824 ((and c-opt-cpp-prefix
1825 (/= cmt-skip-pos simple-ws-beg)
1826 (c-beginning-of-macro))
1827 ;; Inside a cpp directive. See if it should be skipped over.
1828 (let ((cpp-beg (point)))
1829
1830 ;; Move back over all line continuations in the region skipped
1831 ;; over by `c-backward-comments'. If we go past it then we
1832 ;; started inside the cpp directive.
1833 (goto-char simple-ws-beg)
1834 (beginning-of-line)
1835 (while (and (> (point) cmt-skip-pos)
1836 (progn (backward-char)
1837 (eq (char-before) ?\\)))
1838 (beginning-of-line))
1839
1840 (if (< (point) cmt-skip-pos)
1841 ;; Don't move past the cpp directive if we began inside
1842 ;; it. Note that the position at the end of the last line
1843 ;; of the macro is also considered to be within it.
1844 (progn (goto-char cmt-skip-pos)
1845 nil)
1846
1847 ;; It's worthwhile to spend a little bit of effort on finding
1848 ;; the end of the macro, to get a good `simple-ws-beg'
1849 ;; position for the cache. Note that `c-backward-comments'
1850 ;; could have stepped over some comments before going into
1851 ;; the macro, and then `simple-ws-beg' must be kept on the
1852 ;; same side of those comments.
1853 (goto-char simple-ws-beg)
1854 (skip-chars-backward " \t\n\r\f\v")
1855 (if (eq (char-before) ?\\)
1856 (forward-char))
1857 (forward-line 1)
1858 (if (< (point) simple-ws-beg)
1859 ;; Might happen if comments after the macro were skipped
1860 ;; over.
1861 (setq simple-ws-beg (point)))
1862
1863 (goto-char cpp-beg)
1864 t)))
1865
1866 ((/= (save-excursion
1867 (skip-chars-forward " \t\n\r\f\v" simple-ws-beg)
1868 (setq next-rung-pos (point)))
1869 simple-ws-beg)
1870 ;; Skipped over comments. Must put point at the end of
1871 ;; the simple ws at point since we might be after a line
1872 ;; comment or cpp directive that's been partially
1873 ;; narrowed out, and we can't risk marking the simple ws
1874 ;; at the end of it.
1875 (goto-char next-rung-pos)
1876 t)))
1877
1878 ;; We've searched over a piece of non-white syntactic ws. See if this
1879 ;; can be cached.
1880 (setq next-rung-pos (point))
1881 (skip-chars-backward " \t\f\v")
1882
1883 (if (or
1884 ;; Cache if we started either from a marked rung or from a
1885 ;; completely uncached position.
1886 rung-is-marked
1887 (not (get-text-property (1- simple-ws-beg) 'c-in-sws))
1888
1889 ;; Cache if there's a marked rung in the encountered simple ws.
1890 (save-excursion
1891 (skip-chars-backward " \t\n\r\f\v")
1892 (text-property-any (point) (min (1+ next-rung-pos) (point-max))
1893 'c-is-sws t)))
a66cd3ee 1894
d9e94c22
MS
1895 (progn
1896 (c-debug-sws-msg
1897 "c-backward-sws caching [%s..%s] - [%s..%s] (min %s)"
1898 (point) (1+ next-rung-pos)
1899 simple-ws-beg (min (1+ rung-pos) (point-max))
1900 (point-min))
1901
1902 ;; Remove the properties for any nested ws that might be cached.
1903 ;; Only necessary for `c-is-sws' since `c-in-sws' will be set
1904 ;; anyway.
1905 (c-remove-is-sws (1+ next-rung-pos) simple-ws-beg)
1906 (unless (and rung-is-marked (= simple-ws-beg rung-pos))
1907 (let ((rung-end-pos (min (1+ rung-pos) (point-max))))
1908 (unless (get-text-property (1- rung-end-pos) 'c-is-sws)
1909 ;; Remove any `c-in-sws' property from the last char of
1910 ;; the rung before we mark it with `c-is-sws', so that we
1911 ;; won't connect with the remains of a broken "ladder".
1912 (c-remove-in-sws (1- rung-end-pos) rung-end-pos))
1913 (c-put-is-sws simple-ws-beg
1914 rung-end-pos)
1915 (setq rung-is-marked t)))
1916 (c-put-in-sws (setq simple-ws-beg (point)
1917 last-put-in-sws-pos simple-ws-beg)
1918 rung-pos)
1919 (c-put-is-sws (setq rung-pos simple-ws-beg)
1920 (1+ next-rung-pos)))
1921
1922 (c-debug-sws-msg
1923 "c-backward-sws not caching [%s..%s] - [%s..%s] (min %s)"
1924 (point) (1+ next-rung-pos)
1925 simple-ws-beg (min (1+ rung-pos) (point-max))
1926 (point-min))
1927 (setq rung-pos next-rung-pos
1928 simple-ws-beg (point))
1929 ))
1930
1931 ;; Make sure that the newly marked `c-in-sws' region doesn't connect to
1932 ;; another one before the point (which might occur when editing inside a
1933 ;; comment or macro).
1934 (when (eq last-put-in-sws-pos (point))
1935 (cond ((< (point-min) last-put-in-sws-pos)
1936 (c-debug-sws-msg
1937 "c-backward-sws clearing at %s for cache separation"
1938 (1- last-put-in-sws-pos))
1939 (c-remove-in-sws (1- last-put-in-sws-pos)
1940 last-put-in-sws-pos))
1941 ((> (point-min) 1)
1942 ;; If at bob and the buffer is narrowed, we have to clear the
1943 ;; character we're standing on instead since there might be a
1944 ;; `c-in-sws' before (point-min). In this case it's necessary
1945 ;; to clear both properties.
1946 (c-debug-sws-msg
1947 "c-backward-sws clearing thoroughly at %s for cache separation"
1948 last-put-in-sws-pos)
1949 (c-remove-is-and-in-sws last-put-in-sws-pos
1950 (1+ last-put-in-sws-pos)))))
1951 )))
785eecbb 1952
d9e94c22 1953\f
580fba94
AM
1954;; Other whitespace tools
1955(defun c-partial-ws-p (beg end)
1956 ;; Is the region (beg end) WS, and is there WS (or BOB/EOB) next to the
1957 ;; region? This is a "heuristic" function. .....
b414f371 1958 ;;
c421028a
AM
1959 ;; The motivation for the second bit is to check whether removing this
1960 ;; region would coalesce two symbols.
580fba94
AM
1961 ;;
1962 ;; FIXME!!! This function doesn't check virtual semicolons in any way. Be
1963 ;; careful about using this function for, e.g. AWK. (2007/3/7)
1964 (save-excursion
1965 (let ((end+1 (min (1+ end) (point-max))))
1966 (or (progn (goto-char (max (point-min) (1- beg)))
1967 (c-skip-ws-forward end)
1968 (eq (point) end))
1969 (progn (goto-char beg)
1970 (c-skip-ws-forward end+1)
1971 (eq (point) end+1))))))
1972\f
0386b551 1973;; A system for finding noteworthy parens before the point.
e1c458ae 1974
0ec1d2c5
AM
1975(defconst c-state-cache-too-far 5000)
1976;; A maximum comfortable scanning distance, e.g. between
1977;; `c-state-cache-good-pos' and "HERE" (where we call c-parse-state). When
1978;; this distance is exceeded, we take "emergency meausures", e.g. by clearing
1979;; the cache and starting again from point-min or a beginning of defun. This
1980;; value can be tuned for efficiency or set to a lower value for testing.
1981
d9e94c22
MS
1982(defvar c-state-cache nil)
1983(make-variable-buffer-local 'c-state-cache)
1984;; The state cache used by `c-parse-state' to cut down the amount of
0ec1d2c5
AM
1985;; searching. It's the result from some earlier `c-parse-state' call. See
1986;; `c-parse-state''s doc string for details of its structure.
0386b551 1987;;
d9e94c22
MS
1988;; The use of the cached info is more effective if the next
1989;; `c-parse-state' call is on a line close by the one the cached state
1990;; was made at; the cache can actually slow down a little if the
1991;; cached state was made very far back in the buffer. The cache is
1992;; most effective if `c-parse-state' is used on each line while moving
1993;; forward.
e1c458ae 1994
0386b551
AM
1995(defvar c-state-cache-good-pos 1)
1996(make-variable-buffer-local 'c-state-cache-good-pos)
0ec1d2c5
AM
1997;; This is a position where `c-state-cache' is known to be correct, or
1998;; nil (see below). It's a position inside one of the recorded unclosed
1999;; parens or the top level, but not further nested inside any literal or
2000;; subparen that is closed before the last recorded position.
0386b551
AM
2001;;
2002;; The exact position is chosen to try to be close to yet earlier than
2003;; the position where `c-state-cache' will be called next. Right now
2004;; the heuristic is to set it to the position after the last found
2005;; closing paren (of any type) before the line on which
2006;; `c-parse-state' was called. That is chosen primarily to work well
2007;; with refontification of the current line.
0ec1d2c5
AM
2008;;
2009;; 2009-07-28: When `c-state-point-min' and the last position where
2010;; `c-parse-state' or for which `c-invalidate-state-cache' was called, are
2011;; both in the same literal, there is no such "good position", and
2012;; c-state-cache-good-pos is then nil. This is the ONLY circumstance in which
2013;; it can be nil. In this case, `c-state-point-min-literal' will be non-nil.
2014;;
2015;; 2009-06-12: In a brace desert, c-state-cache-good-pos may also be in
2016;; the middle of the desert, as long as it is not within a brace pair
2017;; recorded in `c-state-cache' or a paren/bracket pair.
2018
2019
2020;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2021;; We maintain a simple cache of positions which aren't in a literal, so as to
2022;; speed up testing for non-literality.
2023(defconst c-state-nonlit-pos-interval 10000)
2024;; The approximate interval between entries in `c-state-nonlit-pos-cache'.
2025
2026(defvar c-state-nonlit-pos-cache nil)
2027(make-variable-buffer-local 'c-state-nonlit-pos-cache)
2028;; A list of buffer positions which are known not to be in a literal. This is
2029;; ordered with higher positions at the front of the list. Only those which
2030;; are less than `c-state-nonlit-pos-cache-limit' are valid.
2031
2032(defvar c-state-nonlit-pos-cache-limit 1)
2033(make-variable-buffer-local 'c-state-nonlit-pos-cache-limit)
2034;; An upper limit on valid entries in `c-state-nonlit-pos-cache'. This is
2035;; reduced by buffer changes, and increased by invocations of
2036;; `c-state-literal-at'.
2037
2038(defsubst c-state-pp-to-literal (from to)
2039 ;; Do a parse-partial-sexp from FROM to TO, returning the bounds of any
2040 ;; literal at TO as a cons, otherwise NIL.
2041 ;; FROM must not be in a literal, and the buffer should already be wide
2042 ;; enough.
2043 (save-excursion
2044 (let ((s (parse-partial-sexp from to)))
2045 (when (or (nth 3 s) (nth 4 s)) ; in a string or comment
2046 (parse-partial-sexp (point) (point-max)
2047 nil ; TARGETDEPTH
2048 nil ; STOPBEFORE
2049 s ; OLDSTATE
2050 'syntax-table) ; stop at end of literal
2051 (cons (nth 8 s) (point))))))
2052
2053(defun c-state-literal-at (here)
2054 ;; If position HERE is inside a literal, return (START . END), the
2055 ;; boundaries of the literal (which may be outside the accessible bit of the
2056 ;; buffer). Otherwise, return nil.
2057 ;;
2058 ;; This function is almost the same as `c-literal-limits'. It differs in
2059 ;; that it is a lower level function, and that it rigourously follows the
2060 ;; syntax from BOB, whereas `c-literal-limits' uses a "local" safe position.
2061 (save-restriction
2062 (widen)
2063 (save-excursion
2064 (let ((c c-state-nonlit-pos-cache)
2065 pos npos lit)
2066 ;; Trim the cache to take account of buffer changes.
2067 (while (and c (> (car c) c-state-nonlit-pos-cache-limit))
2068 (setq c (cdr c)))
2069 (setq c-state-nonlit-pos-cache c)
2070
2071 (while (and c (> (car c) here))
2072 (setq c (cdr c)))
2073 (setq pos (or (car c) (point-min)))
2074
2075 (while (<= (setq npos (+ pos c-state-nonlit-pos-interval))
2076 here)
2077 (setq lit (c-state-pp-to-literal pos npos))
2078 (setq pos (or (cdr lit) npos)) ; end of literal containing npos.
2079 (setq c-state-nonlit-pos-cache (cons pos c-state-nonlit-pos-cache)))
2080
2081 (if (> pos c-state-nonlit-pos-cache-limit)
2082 (setq c-state-nonlit-pos-cache-limit pos))
2083 (if (< pos here)
2084 (setq lit (c-state-pp-to-literal pos here)))
2085 lit))))
2086
2087(defsubst c-state-lit-beg (pos)
2088 ;; Return the start of the literal containing POS, or POS itself.
2089 (or (car (c-state-literal-at pos))
2090 pos))
2091
2092(defsubst c-state-cache-non-literal-place (pos state)
2093 ;; Return a position outside of a string/comment at or before POS.
2094 ;; STATE is the parse-partial-sexp state at POS.
2095 (if (or (nth 3 state) ; in a string?
2096 (nth 4 state)) ; in a comment?
2097 (nth 8 state)
2098 pos))
2099
2100
2101;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2102;; Stuff to do with point-min, and coping with any literal there.
2103(defvar c-state-point-min 1)
2104(make-variable-buffer-local 'c-state-point-min)
2105;; This is (point-min) when `c-state-cache' was last calculated. A change of
2106;; narrowing is likely to affect the parens that are visible before the point.
2107
2108(defvar c-state-point-min-lit-type nil)
2109(make-variable-buffer-local 'c-state-point-min-lit-type)
2110(defvar c-state-point-min-lit-start nil)
2111(make-variable-buffer-local 'c-state-point-min-lit-start)
2112;; These two variables define the literal, if any, containing point-min.
2113;; Their values are, respectively, 'string, c, or c++, and the start of the
2114;; literal. If there's no literal there, they're both nil.
2115
2116(defvar c-state-min-scan-pos 1)
2117(make-variable-buffer-local 'c-state-min-scan-pos)
2118;; This is the earliest buffer-pos from which scanning can be done. It is
2119;; either the end of the literal containing point-min, or point-min itself.
2120;; It becomes nil if the buffer is changed earlier than this point.
2121(defun c-state-get-min-scan-pos ()
2122 ;; Return the lowest valid scanning pos. This will be the end of the
2123 ;; literal enclosing point-min, or point-min itself.
2124 (or c-state-min-scan-pos
2125 (save-restriction
2126 (save-excursion
2127 (widen)
2128 (goto-char c-state-point-min-lit-start)
2129 (if (eq c-state-point-min-lit-type 'string)
2130 (forward-sexp)
2131 (forward-comment 1))
2132 (setq c-state-min-scan-pos (point))))))
2133
2134(defun c-state-mark-point-min-literal ()
2135 ;; Determine the properties of any literal containing POINT-MIN, setting the
2136 ;; variables `c-state-point-min-lit-type', `c-state-point-min-lit-start',
2137 ;; and `c-state-min-scan-pos' accordingly. The return value is meaningless.
2138 (let ((p-min (point-min))
2139 lit)
2140 (save-restriction
2141 (widen)
2142 (setq lit (c-state-literal-at p-min))
2143 (if lit
2144 (setq c-state-point-min-lit-type
2145 (save-excursion
2146 (goto-char (car lit))
2147 (cond
2148 ((looking-at c-block-comment-start-regexp) 'c)
2149 ((looking-at c-line-comment-starter) 'c++)
2150 (t 'string)))
2151 c-state-point-min-lit-start (car lit)
2152 c-state-min-scan-pos (cdr lit))
2153 (setq c-state-point-min-lit-type nil
2154 c-state-point-min-lit-start nil
2155 c-state-min-scan-pos p-min)))))
2156
2157
2158;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2159;; A variable which signals a brace dessert - helpful for reducing the number
2160;; of fruitless backward scans.
2161(defvar c-state-brace-pair-desert nil)
2162(make-variable-buffer-local 'c-state-brace-pair-desert)
2163;; Used only in `c-append-lower-brace-pair-to-state-cache'. It is set when an
2164;; that defun has searched backwards for a brace pair and not found one. Its
2165;; value is either nil or a cons (PA . FROM), where PA is the position of the
2166;; enclosing opening paren/brace/bracket which bounds the backwards search (or
2167;; nil when at top level) and FROM is where the backward search started. It
2168;; is reset to nil in `c-invalidate-state-cache'.
2169
2170
2171;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2172;; Lowish level functions/macros which work directly on `c-state-cache', or a
2173;; list of like structure.
2174(defmacro c-state-cache-top-lparen (&optional cache)
2175 ;; Return the address of the top left brace/bracket/paren recorded in CACHE
2176 ;; (default `c-state-cache') (or nil).
2177 (let ((cash (or cache 'c-state-cache)))
2178 `(if (consp (car ,cash))
2179 (caar ,cash)
2180 (car ,cash))))
2181
2182(defmacro c-state-cache-top-paren (&optional cache)
2183 ;; Return the address of the latest brace/bracket/paren (whether left or
2184 ;; right) recorded in CACHE (default `c-state-cache') or nil.
2185 (let ((cash (or cache 'c-state-cache)))
2186 `(if (consp (car ,cash))
2187 (cdar ,cash)
2188 (car ,cash))))
2189
2190(defmacro c-state-cache-after-top-paren (&optional cache)
2191 ;; Return the position just after the latest brace/bracket/paren (whether
2192 ;; left or right) recorded in CACHE (default `c-state-cache') or nil.
2193 (let ((cash (or cache 'c-state-cache)))
2194 `(if (consp (car ,cash))
2195 (cdar ,cash)
2196 (and (car ,cash)
2197 (1+ (car ,cash))))))
2198
2199(defun c-get-cache-scan-pos (here)
2200 ;; From the state-cache, determine the buffer position from which we might
2201 ;; scan forward to HERE to update this cache. This position will be just
2202 ;; after a paren/brace/bracket recorded in the cache, if possible, otherwise
2203 ;; return the earliest position in the accessible region which isn't within
2204 ;; a literal. If the visible portion of the buffer is entirely within a
2205 ;; literal, return NIL.
2206 (let ((c c-state-cache) elt)
2207 ;(while (>= (or (c-state-cache-top-lparen c) 1) here)
2208 (while (and c
2209 (>= (c-state-cache-top-lparen c) here))
2210 (setq c (cdr c)))
2211
2212 (setq elt (car c))
2213 (cond
2214 ((consp elt)
2215 (if (> (cdr elt) here)
2216 (1+ (car elt))
2217 (cdr elt)))
2218 (elt (1+ elt))
2219 ((<= (c-state-get-min-scan-pos) here)
2220 (c-state-get-min-scan-pos))
2221 (t nil))))
2222
2223;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2224;; Variables which keep track of preprocessor constructs.
2225(defvar c-state-old-cpp-beg nil)
2226(make-variable-buffer-local 'c-state-old-cpp-beg)
2227(defvar c-state-old-cpp-end nil)
2228(make-variable-buffer-local 'c-state-old-cpp-end)
2229;; These are the limits of the macro containing point at the previous call of
2230;; `c-parse-state', or nil.
2231
2232;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2233;; Defuns which analyse the buffer, yet don't change `c-state-cache'.
2234(defun c-get-fallback-scan-pos (here)
2235 ;; Return a start position for building `c-state-cache' from
2236 ;; scratch. This will be at the top level, 2 defuns back.
0386b551
AM
2237 (save-excursion
2238 ;; Go back 2 bods, but ignore any bogus positions returned by
2239 ;; beginning-of-defun (i.e. open paren in column zero).
2240 (goto-char here)
2241 (let ((cnt 2))
2242 (while (not (or (bobp) (zerop cnt)))
0ec1d2c5 2243 (c-beginning-of-defun-1) ; Pure elisp BOD.
0386b551
AM
2244 (if (eq (char-after) ?\{)
2245 (setq cnt (1- cnt)))))
2246 (point)))
2247
0ec1d2c5
AM
2248(defun c-state-balance-parens-backwards (here top)
2249 ;; Return the position of the opening paren/brace/bracket before HERE which
2250 ;; matches the outermost close p/b/b between HERE and TOP, like this:
2251 ;;
2252 ;; ......................................
2253 ;; | |
2254 ;; ( [ ( ........... ) ( ) ] )
2255 ;; ^ ^ ^
2256 ;; | | |
2257 ;; return HERE TOP
2258 ;;
2259 ;; If there aren't enough opening paren/brace/brackets, return the position
2260 ;; of the outermost one found, or HERE it there are none. If there are no
2261 ;; closeing p/b/bs between HERE and TOP, return HERE. HERE and TOP must not
2262 ;; be inside literals. Only the accessible portion of the buffer will be
2263 ;; scanned.
2264
2265 ;; PART 1: scan from `here' up to `top', accumulating ")"s which enclose
2266 ;; `here'. Go round the next loop each time we pass over such a ")". These
2267 ;; probably match "("s before `here'.
2268 (let (pos pa ren+1 lonely-rens)
2269 (save-excursion
2270 (save-restriction
2271 (narrow-to-region (point-min) top) ; This can move point, sometimes.
2272 (setq pos here)
2273 (c-safe
2274 (while
2275 (setq ren+1 (scan-lists pos 1 1)) ; might signal
2276 (setq lonely-rens (cons ren+1 lonely-rens)
2277 pos ren+1)))))
2278
2279 ;; PART 2: Scan back before `here' searching for the "("s
2280 ;; matching/mismatching the ")"s found above. We only need to direct the
2281 ;; caller to scan when we've encountered unmatched right parens.
2282 (when lonely-rens
2283 (setq pos here)
2284 (c-safe
2285 (while
2286 (and lonely-rens ; actual values aren't used.
2287 (setq pa (scan-lists pos -1 1)))
2288 (setq pos pa)
2289 (setq lonely-rens (cdr lonely-rens)))) ;)
2290 )
2291 pos))
2292
2293(defun c-parse-state-get-strategy (here good-pos)
2294 ;; Determine the scanning strategy for adjusting `c-parse-state', attempting
2295 ;; to minimise the amount of scanning. HERE is the pertinent position in
2296 ;; the buffer, GOOD-POS is a position where `c-state-cache' (possibly with
2297 ;; its head trimmed) is known to be good, or nil if there is no such
2298 ;; position.
2299 ;;
2300 ;; The return value is a list, one of the following:
2301 ;;
2302 ;; o - ('forward CACHE-POS START-POINT) - scan forward from START-POINT,
2303 ;; which is not less than CACHE-POS.
2304 ;; o - ('backward CACHE-POS nil) - scan backwards (from HERE).
2305 ;; o - ('BOD nil START-POINT) - scan forwards from START-POINT, which is at the
2306 ;; top level.
2307 ;; o - ('IN-LIT nil nil) - point is inside the literal containing point-min.
2308 ;; , where CACHE-POS is the highest position recorded in `c-state-cache' at
2309 ;; or below HERE.
2310 (let ((cache-pos (c-get-cache-scan-pos here)) ; highest position below HERE in cache (or 1)
2311 BOD-pos ; position of 2nd BOD before HERE.
2312 strategy ; 'forward, 'backward, 'BOD, or 'IN-LIT.
2313 start-point
2314 how-far) ; putative scanning distance.
2315 (setq good-pos (or good-pos (c-state-get-min-scan-pos)))
2316 (cond
2317 ((< here (c-state-get-min-scan-pos))
2318 (setq strategy 'IN-LIT
2319 start-point nil
2320 cache-pos nil
2321 how-far 0))
2322 ((<= good-pos here)
2323 (setq strategy 'forward
2324 start-point (max good-pos cache-pos)
2325 how-far (- here start-point)))
2326 ((< (- good-pos here) (- here cache-pos)) ; FIXME!!! ; apply some sort of weighting.
2327 (setq strategy 'backward
2328 how-far (- good-pos here)))
2329 (t
2330 (setq strategy 'forward
2331 how-far (- here cache-pos)
2332 start-point cache-pos)))
2333
2334 ;; Might we be better off starting from the top level, two defuns back,
2335 ;; instead?
2336 (when (> how-far c-state-cache-too-far)
2337 (setq BOD-pos (c-get-fallback-scan-pos here)) ; somewhat EXPENSIVE!!!
2338 (if (< (- here BOD-pos) how-far)
2339 (setq strategy 'BOD
2340 start-point BOD-pos)))
2341
2342 (list
2343 strategy
2344 (and (memq strategy '(forward backward)) cache-pos)
2345 (and (memq strategy '(forward BOD)) start-point))))
2346
2347
2348;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2349;; Routines which change `c-state-cache' and associated values.
2350(defun c-renarrow-state-cache ()
2351 ;; The region (more precisely, point-min) has changed since we
2352 ;; calculated `c-state-cache'. Amend `c-state-cache' accordingly.
2353 (if (< (point-min) c-state-point-min)
2354 ;; If point-min has MOVED BACKWARDS then we drop the state completely.
2355 ;; It would be possible to do a better job here and recalculate the top
2356 ;; only.
2357 (progn
2358 (c-state-mark-point-min-literal)
2359 (setq c-state-cache nil
2360 c-state-cache-good-pos c-state-min-scan-pos
2361 c-state-brace-pair-desert nil))
2362
2363 ;; point-min has MOVED FORWARD.
2364
2365 ;; Is the new point-min inside a (different) literal?
2366 (unless (and c-state-point-min-lit-start ; at prev. point-min
2367 (< (point-min) (c-state-get-min-scan-pos)))
2368 (c-state-mark-point-min-literal))
2369
2370 ;; Cut off a bit of the tail from `c-state-cache'.
2371 (let ((ptr (cons nil c-state-cache))
2372 pa)
2373 (while (and (setq pa (c-state-cache-top-lparen (cdr ptr)))
2374 (>= pa (point-min)))
2375 (setq ptr (cdr ptr)))
2376
2377 (when (consp ptr)
2378 (if (eq (cdr ptr) c-state-cache)
2379 (setq c-state-cache nil
2380 c-state-cache-good-pos c-state-min-scan-pos)
2381 (setcdr ptr nil)
2382 (setq c-state-cache-good-pos (1+ (c-state-cache-top-lparen))))
2383 )))
2384
2385 (setq c-state-point-min (point-min)))
2386
2387(defun c-append-lower-brace-pair-to-state-cache (from &optional upper-lim)
2388 ;; If there is a brace pair preceding FROM in the buffer (not necessarily
2389 ;; immediately preceding), push a cons onto `c-state-cache' to represent it.
2390 ;; FROM must not be inside a literal. If UPPER-LIM is non-nil, we append
2391 ;; the highest brace pair whose "}" is below UPPER-LIM.
2392 ;;
2393 ;; Return non-nil when this has been done.
2394 ;;
2395 ;; This routine should be fast. Since it can get called a LOT, we maintain
2396 ;; `c-state-brace-pair-desert', a small cache of "failures", such that we
2397 ;; reduce the time wasted in repeated fruitless searches in brace deserts.
2398 (save-excursion
2399 (save-restriction
2400 (let ((bra from) ce ; Positions of "{" and "}".
2401 new-cons
2402 (cache-pos (c-state-cache-top-lparen)) ; might be nil.
2403 (macro-start-or-from
2404 (progn (goto-char from)
2405 (c-beginning-of-macro)
2406 (point))))
2407 (or upper-lim (setq upper-lim from))
2408
2409 ;; If we're essentially repeating a fruitless search, just give up.
2410 (unless (and c-state-brace-pair-desert
2411 (eq cache-pos (car c-state-brace-pair-desert))
2412 (<= from (cdr c-state-brace-pair-desert)))
2413 ;; Only search what we absolutely need to:
2414 (if (and c-state-brace-pair-desert
2415 (> from (cdr c-state-brace-pair-desert)))
2416 (narrow-to-region (cdr c-state-brace-pair-desert) (point-max)))
2417
2418 ;; In the next pair of nested loops, the inner one moves back past a
2419 ;; pair of (mis-)matching parens or brackets; the outer one moves
2420 ;; back over a sequence of unmatched close brace/paren/bracket each
2421 ;; time round.
2422 (while
2423 (progn
2424 (c-safe
2425 (while
2426 (and (setq ce (scan-lists bra -1 -1)) ; back past )/]/}; might signal
2427 (setq bra (scan-lists ce -1 1)) ; back past (/[/{; might signal
2428 (or (> ce upper-lim)
2429 (not (eq (char-after bra) ?\{))
2430 (and (goto-char bra)
2431 (c-beginning-of-macro)
2432 (< (point) macro-start-or-from))))))
2433 (and ce (< ce bra)))
2434 (setq bra ce)) ; If we just backed over an unbalanced closing
2435 ; brace, ignore it.
2436
2437 (if (and ce (< bra ce) (eq (char-after bra) ?\{))
2438 ;; We've found the desired brace-pair.
2439 (progn
2440 (setq new-cons (cons bra (1+ ce)))
2441 (cond
2442 ((consp (car c-state-cache))
2443 (setcar c-state-cache new-cons))
2444 ((and (numberp (car c-state-cache)) ; probably never happens
2445 (< ce (car c-state-cache)))
2446 (setcdr c-state-cache
2447 (cons new-cons (cdr c-state-cache))))
2448 (t (setq c-state-cache (cons new-cons c-state-cache)))))
2449
2450 ;; We haven't found a brace pair. Record this.
2451 (setq c-state-brace-pair-desert (cons cache-pos from))))))))
2452
2453(defsubst c-state-push-any-brace-pair (bra+1 macro-start-or-here)
2454 ;; If BRA+1 is nil, do nothing. Otherwise, BRA+1 is the buffer position
2455 ;; following a {, and that brace has a (mis-)matching } (or ]), and we
2456 ;; "push" "a" brace pair onto `c-state-cache'.
2457 ;;
2458 ;; Here "push" means overwrite the top element if it's itself a brace-pair,
2459 ;; otherwise push it normally.
2460 ;;
2461 ;; The brace pair we push is normally the one surrounding BRA+1, but if the
2462 ;; latter is inside a macro, not being a macro containing
2463 ;; MACRO-START-OR-HERE, we scan backwards through the buffer for a non-macro
2464 ;; base pair. This latter case is assumed to be rare.
2465 ;;
2466 ;; Note: POINT is not preserved in this routine.
2467 (if bra+1
2468 (if (or (> bra+1 macro-start-or-here)
2469 (progn (goto-char bra+1)
2470 (not (c-beginning-of-macro))))
2471 (setq c-state-cache
2472 (cons (cons (1- bra+1)
2473 (scan-lists bra+1 1 1))
2474 (if (consp (car c-state-cache))
2475 (cdr c-state-cache)
2476 c-state-cache)))
2477 ;; N.B. This defsubst codes one method for the simple, normal case,
2478 ;; and a more sophisticated, slower way for the general case. Don't
2479 ;; eliminate this defsubst - it's a speed optimisation.
2480 (c-append-lower-brace-pair-to-state-cache (1- bra+1)))))
2481
2482(defun c-append-to-state-cache (from)
2483 ;; Scan the buffer from FROM to (point-max), adding elements into
2484 ;; `c-state-cache' for braces etc. Return a candidate for
2485 ;; `c-state-cache-good-pos'.
2486 ;;
2487 ;; FROM must be after the latest brace/paren/bracket in `c-state-cache', if
2488 ;; any. Typically, it is immediately after it. It must not be inside a
2489 ;; literal.
2490 (let ((here-bol (c-point 'bol (point-max)))
2491 (macro-start-or-here
2492 (save-excursion (goto-char (point-max))
2493 (if (c-beginning-of-macro)
2494 (point)
2495 (point-max))))
2496 pa+1 ; pos just after an opening PAren (or brace).
2497 (ren+1 from) ; usually a pos just after an closing paREN etc.
2498 ; Is actually the pos. to scan for a (/{/[ from,
2499 ; which sometimes is after a silly )/}/].
2500 paren+1 ; Pos after some opening or closing paren.
2501 paren+1s ; A list of `paren+1's; used to determine a
2502 ; good-pos.
2503 bra+1 ce+1 ; just after L/R bra-ces.
2504 bra+1s ; list of OLD values of bra+1.
2505 mstart) ; start of a macro.
2506
2507 (save-excursion
2508 ;; Each time round the following loop, we enter a succesively deeper
2509 ;; level of brace/paren nesting. (Except sometimes we "continue at
2510 ;; the existing level".) `pa+1' is a pos inside an opening
2511 ;; brace/paren/bracket, usually just after it.
2512 (while
2513 (progn
2514 ;; Each time round the next loop moves forward over an opening then
2515 ;; a closing brace/bracket/paren. This loop is white hot, so it
2516 ;; plays ugly tricks to go fast. DON'T PUT ANYTHING INTO THIS
2517 ;; LOOP WHICH ISN'T ABSOLUTELY NECESSARY!!! It terminates when a
2518 ;; call of `scan-lists' signals an error, which happens when there
2519 ;; are no more b/b/p's to scan.
2520 (c-safe
2521 (while t
2522 (setq pa+1 (scan-lists ren+1 1 -1) ; Into (/{/[; might signal
2523 paren+1s (cons pa+1 paren+1s))
2524 (setq ren+1 (scan-lists pa+1 1 1)) ; Out of )/}/]; might signal
2525 (if (and (eq (char-before pa+1) ?{)) ; Check for a macro later.
2526 (setq bra+1 pa+1))
2527 (setcar paren+1s ren+1)))
2528
2529 (if (and pa+1 (> pa+1 ren+1))
2530 ;; We've just entered a deeper nesting level.
2531 (progn
2532 ;; Insert the brace pair (if present) and the single open
2533 ;; paren/brace/bracket into `c-state-cache' It cannot be
2534 ;; inside a macro, except one around point, because of what
2535 ;; `c-neutralize-syntax-in-CPP' has done.
2536 (c-state-push-any-brace-pair bra+1 macro-start-or-here)
2537 ;; Insert the opening brace/bracket/paren position.
2538 (setq c-state-cache (cons (1- pa+1) c-state-cache))
2539 ;; Clear admin stuff for the next more nested part of the scan.
2540 (setq ren+1 pa+1 pa+1 nil bra+1 nil bra+1s nil)
2541 t) ; Carry on the loop
2542
2543 ;; All open p/b/b's at this nesting level, if any, have probably
2544 ;; been closed by matching/mismatching ones. We're probably
2545 ;; finished - we just need to check for having found an
2546 ;; unmatched )/}/], which we ignore. Such a )/}/] can't be in a
2547 ;; macro, due the action of `c-neutralize-syntax-in-CPP'.
2548 (c-safe (setq ren+1 (scan-lists ren+1 1 1)))))) ; acts as loop control.
2549
2550 ;; Record the final, innermost, brace-pair if there is one.
2551 (c-state-push-any-brace-pair bra+1 macro-start-or-here)
2552
2553 ;; Determine a good pos
2554 (while (and (setq paren+1 (car paren+1s))
2555 (> (if (> paren+1 macro-start-or-here)
2556 paren+1
2557 (goto-char paren+1)
2558 (setq mstart (and (c-beginning-of-macro)
2559 (point)))
2560 (or mstart paren+1))
2561 here-bol))
2562 (setq paren+1s (cdr paren+1s)))
2563 (cond
2564 ((and paren+1 mstart)
2565 (min paren+1 mstart))
2566 (paren+1)
2567 (t from)))))
2568
2569(defun c-remove-stale-state-cache (good-pos pps-point)
2570 ;; Remove stale entries from the `c-cache-state', i.e. those which will
2571 ;; not be in it when it is amended for position (point-max).
2572 ;; Additionally, the "outermost" open-brace entry before (point-max)
2573 ;; will be converted to a cons if the matching close-brace is scanned.
2574 ;;
2575 ;; GOOD-POS is a "maximal" "safe position" - there must be no open
2576 ;; parens/braces/brackets between GOOD-POS and (point-max).
2577 ;;
2578 ;; As a second thing, calculate the result of parse-partial-sexp at
2579 ;; PPS-POINT, w.r.t. GOOD-POS. The motivation here is that
2580 ;; `c-state-cache-good-pos' may become PPS-POINT, but the caller may need to
2581 ;; adjust it to get outside a string/comment. (Sorry about this! The code
2582 ;; needs to be FAST).
2583 ;;
2584 ;; Return a list (GOOD-POS SCAN-BACK-POS PPS-STATE), where
2585 ;; o - GOOD-POS is a position where the new value `c-state-cache' is known
2586 ;; to be good (we aim for this to be as high as possible);
2587 ;; o - SCAN-BACK-POS, if not nil, indicates there may be a brace pair
2588 ;; preceding POS which needs to be recorded in `c-state-cache'. It is a
2589 ;; position to scan backwards from.
2590 ;; o - PPS-STATE is the parse-partial-sexp state at PPS-POINT.
2591 (save-restriction
2592 (narrow-to-region 1 (point-max))
2593 (save-excursion
2594 (let* ((in-macro-start ; point-max or beginning of macro containing it
2595 (save-excursion
2596 (goto-char (point-max))
2597 (and (c-beginning-of-macro)
2598 (point))))
2599 (good-pos-actual-macro-start ; Start of macro containing good-pos
2600 ; or nil
2601 (and (< good-pos (point-max))
2602 (save-excursion
2603 (goto-char good-pos)
2604 (and (c-beginning-of-macro)
2605 (point)))))
2606 (good-pos-actual-macro-end ; End of this macro, (maybe
2607 ; (point-max)), or nil.
2608 (and good-pos-actual-macro-start
2609 (save-excursion
2610 (goto-char good-pos-actual-macro-start)
2611 (c-end-of-macro)
2612 (point))))
2613 pps-state ; Will be 9 or 10 elements long.
2614 pos
2615 upper-lim ; ,beyond which `c-state-cache' entries are removed
2616 scan-back-pos
2617 pair-beg pps-point-state target-depth)
2618
2619 ;; Remove entries beyond (point-max). Also remove any entries inside
2620 ;; a macro, unless (point-max) is in the same macro.
2621 (setq upper-lim
2622 (if (or (null c-state-old-cpp-beg)
2623 (and (> (point-max) c-state-old-cpp-beg)
2624 (< (point-max) c-state-old-cpp-end)))
2625 (point-max)
2626 (min (point-max) c-state-old-cpp-beg)))
2627 (while (and c-state-cache (> (c-state-cache-top-lparen) upper-lim))
2628 (setq c-state-cache (cdr c-state-cache)))
2629 ;; If `upper-lim' is inside the last recorded brace pair, remove its
2630 ;; RBrace and indicate we'll need to search backwards for a previous
2631 ;; brace pair.
2632 (when (and c-state-cache
2633 (consp (car c-state-cache))
2634 (> (cdar c-state-cache) upper-lim))
2635 (setcar c-state-cache (caar c-state-cache))
2636 (setq scan-back-pos (car c-state-cache)))
2637
2638 ;; The next loop jumps forward out of a nested level of parens each
2639 ;; time round; the corresponding elements in `c-state-cache' are
2640 ;; removed. `pos' is just after the brace-pair or the open paren at
2641 ;; (car c-state-cache). There can be no open parens/braces/brackets
2642 ;; between `good-pos'/`good-pos-actual-macro-start' and (point-max),
2643 ;; due to the interface spec to this function.
2644 (setq pos (if good-pos-actual-macro-end
2645 (1+ good-pos-actual-macro-end) ; get outside the macro as
2646 ; marked by a `category' text property.
2647 good-pos))
2648 (goto-char pos)
2649 (while (and c-state-cache
2650 (< (point) (point-max)))
2651 (cond
2652 ((null pps-state) ; first time through
2653 (setq target-depth -1))
2654 ((eq (car pps-state) target-depth) ; found closing ),},]
2655 (setq target-depth (1- (car pps-state))))
2656 ;; Do nothing when we've merely reached pps-point.
2657 )
2658
2659 ;; Scan!
2660 (setq pps-state
2661 (parse-partial-sexp
2662 (point) (if (< (point) pps-point) pps-point (point-max))
2663 target-depth
2664 nil pps-state))
2665
2666 (if (= (point) pps-point)
2667 (setq pps-point-state pps-state))
2668
2669 (when (eq (car pps-state) target-depth)
2670 (setq pos (point)) ; POS is now just after an R-paren/brace.
2671 (cond
2672 ((and (consp (car c-state-cache))
2673 (eq (point) (cdar c-state-cache)))
2674 ;; We've just moved out of the paren pair containing the brace-pair
2675 ;; at (car c-state-cache). `pair-beg' is where the open paren is,
2676 ;; and is potentially where the open brace of a cons in
2677 ;; c-state-cache will be.
2678 (setq pair-beg (car-safe (cdr c-state-cache))
2679 c-state-cache (cdr-safe (cdr c-state-cache)))) ; remove {}pair + containing Lparen.
2680 ((numberp (car c-state-cache))
2681 (setq pair-beg (car c-state-cache)
2682 c-state-cache (cdr c-state-cache))) ; remove this
2683 ; containing Lparen
2684 ((numberp (cadr c-state-cache))
2685 (setq pair-beg (cadr c-state-cache)
2686 c-state-cache (cddr c-state-cache))) ; Remove a paren pair
2687 ; together with enclosed brace pair.
2688 ;; (t nil) ; Ignore an unmated Rparen.
2689 )))
2690
2691 (if (< (point) pps-point)
2692 (setq pps-state (parse-partial-sexp (point) pps-point
2693 nil nil ; TARGETDEPTH, STOPBEFORE
2694 pps-state)))
2695
2696 ;; If the last paren pair we moved out of was actually a brace pair,
2697 ;; insert it into `c-state-cache'.
2698 (when (and pair-beg (eq (char-after pair-beg) ?{))
2699 (if (consp (car-safe c-state-cache))
2700 (setq c-state-cache (cdr c-state-cache)))
2701 (setq c-state-cache (cons (cons pair-beg pos)
2702 c-state-cache)))
2703
2704 (list pos scan-back-pos pps-state)))))
2705
2706(defun c-remove-stale-state-cache-backwards (here cache-pos)
2707 ;; Strip stale elements of `c-state-cache' by moving backwards through the
2708 ;; buffer, and inform the caller of the scenario detected.
2709 ;;
2710 ;; HERE is the position we're setting `c-state-cache' for.
2711 ;; CACHE-POS is just after the latest recorded position in `c-state-cache'
2712 ;; before HERE, or a position at or near point-min which isn't in a
2713 ;; literal.
2714 ;;
2715 ;; This function must only be called only when (> `c-state-cache-good-pos'
2716 ;; HERE). Usually the gap between CACHE-POS and HERE is large. It is thus
2717 ;; optimised to eliminate (or minimise) scanning between these two
2718 ;; positions.
2719 ;;
2720 ;; Return a three element list (GOOD-POS SCAN-BACK-POS FWD-FLAG), where:
2721 ;; o - GOOD-POS is a "good position", where `c-state-cache' is valid, or
2722 ;; could become so after missing elements are inserted into
2723 ;; `c-state-cache'. This is JUST AFTER an opening or closing
2724 ;; brace/paren/bracket which is already in `c-state-cache' or just before
2725 ;; one otherwise. exceptionally (when there's no such b/p/b handy) the BOL
2726 ;; before `here''s line, or the start of the literal containing it.
2727 ;; o - SCAN-BACK-POS, if non-nil, indicates there may be a brace pair
2728 ;; preceding POS which isn't recorded in `c-state-cache'. It is a position
2729 ;; to scan backwards from.
2730 ;; o - FWD-FLAG, if non-nil, indicates there may be parens/braces between
2731 ;; POS and HERE which aren't recorded in `c-state-cache'.
2732 ;;
2733 ;; The comments in this defun use "paren" to mean parenthesis or square
2734 ;; bracket (as contrasted with a brace), and "(" and ")" likewise.
2735 ;;
2736 ;; . {..} (..) (..) ( .. { } ) (...) ( .... . ..)
2737 ;; | | | | | |
2738 ;; CP E here D C good
2739 (let ((pos c-state-cache-good-pos)
2740 pa ren ; positions of "(" and ")"
2741 dropped-cons ; whether the last element dropped from `c-state-cache'
2742 ; was a cons (representing a brace-pair)
2743 good-pos ; see above.
2744 lit ; (START . END) of a literal containing some point.
2745 here-lit-start here-lit-end ; bounds of literal containing `here'
2746 ; or `here' itself.
2747 (here-bol (c-point 'bol here))
2748 (too-far-back (max (- here c-state-cache-too-far) 1)))
2749
2750 ;; Remove completely irrelevant entries from `c-state-cache'.
2751 (while (and c-state-cache
2752 (>= (setq pa (c-state-cache-top-lparen)) here))
2753 (setq dropped-cons (consp (car c-state-cache)))
2754 (setq c-state-cache (cdr c-state-cache))
2755 (setq pos pa))
2756 ;; At this stage, (> pos here);
2757 ;; (< (c-state-cache-top-lparen) here) (or is nil).
2758
2759 ;; CASE 1: The top of the cache is a brace pair which now encloses `here'.
2760 ;; As good-pos, return the address. of the "{".
2761 (if (and (consp (car c-state-cache))
2762 (> (cdar c-state-cache) here))
2763 ;; Since we've no knowledge of what's inside these braces, we have no
2764 ;; alternative but to direct the caller to scan the buffer from the
2765 ;; opening brace.
2766 (progn
2767 (setq pos (caar c-state-cache))
2768 (setcar c-state-cache pos)
2769 (list (1+ pos) pos t)) ; return value. We've just converted a brace
2770 ; pair entry into a { entry, so the caller
2771 ; needs to search for a brace pair before the
2772 ; {.
2773
2774 ;; ;; `here' might be inside a literal. Check for this.
2775 (setq lit (c-state-literal-at here)
2776 here-lit-start (or (car lit) here)
2777 here-lit-end (or (cdr lit) here))
2778
2779 ;; `here' might be nested inside any depth of parens (or brackets but
2780 ;; not braces). Scan backwards to find the outermost such opening
2781 ;; paren, if there is one. This will be the scan position to return.
2782 (save-restriction
2783 (narrow-to-region cache-pos (point-max))
2784 (setq pos (c-state-balance-parens-backwards here-lit-end pos)))
2785
2786 (if (< pos here-lit-start)
2787 ;; CASE 2: Address of outermost ( or [ which now encloses `here',
2788 ;; but didn't enclose the (previous) `c-state-cache-good-pos'. If
2789 ;; there is a brace pair preceding this, it will already be in
2790 ;; `c-state-cache', unless there was a brace pair after it,
2791 ;; i.e. there'll only be one to scan for if we've just deleted one.
2792 (list pos (and dropped-cons pos) t) ; Return value.
2793
2794 ;; `here' isn't enclosed in a (previously unrecorded) bracket/paren.
2795 ;; Further forward scanning isn't needed, but we still need to find a
2796 ;; GOOD-POS. Step out of all enclosing "("s on HERE's line.
2797 (save-restriction
2798 (narrow-to-region here-bol (point-max))
2799 (setq pos here-lit-start)
2800 (c-safe (while (setq pa (scan-lists pos -1 1))
2801 (setq pos pa)))) ; might signal
2802 (if (setq ren (c-safe-scan-lists pos -1 -1 too-far-back))
2803 ;; CASE 3: After a }/)/] before `here''s BOL.
2804 (list (1+ ren) (and dropped-cons pos) nil) ; Return value
2805
2806 ;; CASE 4; Best of a bad job: BOL before `here-bol', or beginning of
2807 ;; literal containing it.
2808 (setq good-pos (c-state-lit-beg (c-point 'bopl here-bol)))
2809 (list good-pos (and dropped-cons good-pos) nil))))))
2810
2811
2812;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2813;; Externally visible routines.
2814
2815(defun c-state-cache-init ()
2816 (setq c-state-cache nil
2817 c-state-cache-good-pos 1
2818 c-state-nonlit-pos-cache nil
2819 c-state-nonlit-pos-cache-limit 1
2820 c-state-brace-pair-desert nil
2821 c-state-point-min 1
2822 c-state-point-min-lit-type nil
2823 c-state-point-min-lit-start nil
2824 c-state-min-scan-pos 1
2825 c-state-old-cpp-beg nil
2826 c-state-old-cpp-end nil)
2827 (c-state-mark-point-min-literal))
2828
2829(defun c-invalidate-state-cache-1 (here)
2830 ;; Invalidate all info on `c-state-cache' that applies to the buffer at HERE
2831 ;; or higher and set `c-state-cache-good-pos' accordingly. The cache is
2832 ;; left in a consistent state.
2833 ;;
2834 ;; This is much like `c-whack-state-after', but it never changes a paren
2835 ;; pair element into an open paren element. Doing that would mean that the
2836 ;; new open paren wouldn't have the required preceding paren pair element.
785eecbb 2837 ;;
0ec1d2c5
AM
2838 ;; This function is called from c-after-change.
2839
2840 ;; The cache of non-literals:
2841 (if (< here c-state-nonlit-pos-cache-limit)
2842 (setq c-state-nonlit-pos-cache-limit here))
2843
2844 ;; `c-state-cache':
2845 ;; Case 1: if `here' is in a literal containing point-min, everything
2846 ;; becomes (or is already) nil.
2847 (if (or (null c-state-cache-good-pos)
2848 (< here (c-state-get-min-scan-pos)))
2849 (setq c-state-cache nil
2850 c-state-cache-good-pos nil
2851 c-state-min-scan-pos nil)
2852
2853;;; Truncate `c-state-cache' and set `c-state-cache-good-pos' to a value below
2854;;; `here'. To maintain its consistency, we may need to insert a new brace
2855;;; pair.
2856 (let ((here-bol (c-point 'bol here))
2857 too-high-pa ; recorded {/(/[ next above here, or nil.
2858 dropped-cons ; was the last removed element a brace pair?
2859 pa)
2860 ;; The easy bit - knock over-the-top bits off `c-state-cache'.
2861 (while (and c-state-cache
2862 (>= (setq pa (c-state-cache-top-paren)) here))
2863 (setq dropped-cons (consp (car c-state-cache))
2864 too-high-pa (c-state-cache-top-lparen)
2865 c-state-cache (cdr c-state-cache)))
2866
2867 ;; Do we need to add in an earlier brace pair, having lopped one off?
2868 (if (and dropped-cons
2869 (< too-high-pa (+ here c-state-cache-too-far)))
2870 (c-append-lower-brace-pair-to-state-cache too-high-pa here-bol))
2871 (setq c-state-cache-good-pos (or (c-state-cache-after-top-paren)
2872 (c-state-get-min-scan-pos)))))
2873
2874 ;; The brace-pair desert marker:
2875 (when (car c-state-brace-pair-desert)
2876 (if (< here (car c-state-brace-pair-desert))
2877 (setq c-state-brace-pair-desert nil)
2878 (if (< here (cdr c-state-brace-pair-desert))
2879 (setcdr c-state-brace-pair-desert here)))))
2880
2881(defun c-parse-state-1 ()
2882 ;; Find and record all noteworthy parens between some good point earlier in
2883 ;; the file and point. That good point is at least the beginning of the
2884 ;; top-level construct we are in, or the beginning of the preceding
2885 ;; top-level construct if we aren't in one.
2886 ;;
2887 ;; The returned value is a list of the noteworthy parens with the last one
2888 ;; first. If an element in the list is an integer, it's the position of an
2889 ;; open paren (of any type) which has not been closed before the point. If
2890 ;; an element is a cons, it gives the position of a closed BRACE paren
2891 ;; pair[*]; the car is the start brace position and the cdr is the position
2892 ;; following the closing brace. Only the last closed brace paren pair
2893 ;; before each open paren and before the point is recorded, and thus the
2894 ;; state never contains two cons elements in succession. When a close brace
2895 ;; has no matching open brace (e.g., the matching brace is outside the
2896 ;; visible region), it is not represented in the returned value.
2897 ;;
2898 ;; [*] N.B. The close "brace" might be a mismatching close bracket or paren.
2899 ;; This defun explicitly treats mismatching parens/braces/brackets as
2900 ;; matching. It is the open brace which makes it a "brace" pair.
2901 ;;
2902 ;; If POINT is within a macro, open parens and brace pairs within
2903 ;; THIS macro MIGHT be recorded. This depends on whether their
2904 ;; syntactic properties have been suppressed by
2905 ;; `c-neutralize-syntax-in-CPP'. This might need fixing (2008-12-11).
d9e94c22
MS
2906 ;;
2907 ;; Currently no characters which are given paren syntax with the
2908 ;; syntax-table property are recorded, i.e. angle bracket arglist
2909 ;; parens are never present here. Note that this might change.
2910 ;;
0386b551 2911 ;; BUG: This function doesn't cope entirely well with unbalanced
0ec1d2c5
AM
2912 ;; parens in macros. (2008-12-11: this has probably been resolved
2913 ;; by the function `c-neutralize-syntax-in-CPP'.) E.g. in the
2914 ;; following case the brace before the macro isn't balanced with the
2915 ;; one after it:
0386b551
AM
2916 ;;
2917 ;; {
2918 ;; #define X {
2919 ;; }
2920 ;;
0ec1d2c5
AM
2921 ;; Note to maintainers: this function DOES get called with point
2922 ;; within comments and strings, so don't assume it doesn't!
2923 ;;
0386b551 2924 ;; This function might do hidden buffer changes.
0ec1d2c5
AM
2925 (let* ((here (point))
2926 (here-bopl (c-point 'bopl))
2927 strategy ; 'forward, 'backward etc..
2928 ;; Candidate positions to start scanning from:
2929 cache-pos ; highest position below HERE already existing in
2930 ; cache (or 1).
2931 good-pos
2932 start-point
2933 bopl-state
2934 res
2935 scan-backward-pos scan-forward-p) ; used for 'backward.
2936 ;; If POINT-MIN has changed, adjust the cache
2937 (unless (= (point-min) c-state-point-min)
2938 (c-renarrow-state-cache))
2939
2940 ;; Strategy?
2941 (setq res (c-parse-state-get-strategy here c-state-cache-good-pos)
2942 strategy (car res)
2943 cache-pos (cadr res)
2944 start-point (nth 2 res))
2945
2946 (when (eq strategy 'BOD)
2947 (setq c-state-cache nil
2948 c-state-cache-good-pos start-point))
2949
2950 ;; SCAN!
2951 (save-restriction
2952 (cond
2953 ((memq strategy '(forward BOD))
0386b551 2954 (narrow-to-region (point-min) here)
0ec1d2c5
AM
2955 (setq res (c-remove-stale-state-cache start-point here-bopl))
2956 (setq cache-pos (car res)
2957 scan-backward-pos (cadr res)
2958 bopl-state (car (cddr res))) ; will be nil if (< here-bopl
2959 ; start-point)
2960 (if scan-backward-pos
2961 (c-append-lower-brace-pair-to-state-cache scan-backward-pos))
2962 (setq good-pos
2963 (c-append-to-state-cache cache-pos))
2964 (setq c-state-cache-good-pos
2965 (if (and bopl-state
2966 (< good-pos (- here c-state-cache-too-far)))
2967 (c-state-cache-non-literal-place here-bopl bopl-state)
2968 good-pos)))
2969
2970 ((eq strategy 'backward)
2971 (setq res (c-remove-stale-state-cache-backwards here cache-pos)
2972 good-pos (car res)
2973 scan-backward-pos (cadr res)
2974 scan-forward-p (car (cddr res)))
2975 (if scan-backward-pos
2976 (c-append-lower-brace-pair-to-state-cache
2977 scan-backward-pos))
2978 (setq c-state-cache-good-pos
2979 (if scan-forward-p
2980 (progn (narrow-to-region (point-min) here)
2981 (c-append-to-state-cache good-pos))
2982
2983 (c-get-cache-scan-pos good-pos))))
2984
2985 (t ; (eq strategy 'IN-LIT)
2986 (setq c-state-cache nil
2987 c-state-cache-good-pos nil)))))
2988
2989 c-state-cache)
2990
2991(defun c-invalidate-state-cache (here)
2992 ;; This is a wrapper over `c-invalidate-state-cache-1'.
2993 ;;
2994 ;; It suppresses the syntactic effect of the < and > (template) brackets and
2995 ;; of all parens in preprocessor constructs, except for any such construct
2996 ;; containing point. We can then call `c-invalidate-state-cache-1' without
2997 ;; worrying further about macros and template delimiters.
2998 (c-with-<->-as-parens-suppressed
3e8f7d91
AM
2999 (if (and c-state-old-cpp-beg
3000 (< c-state-old-cpp-beg here))
0ec1d2c5 3001 (c-with-all-but-one-cpps-commented-out
3e8f7d91
AM
3002 c-state-old-cpp-beg
3003 (min c-state-old-cpp-end here)
0ec1d2c5
AM
3004 (c-invalidate-state-cache-1 here))
3005 (c-with-cpps-commented-out
3006 (c-invalidate-state-cache-1 here)))))
3007
3008(defun c-parse-state ()
3009 ;; This is a wrapper over `c-parse-state-1'. See that function for a
3010 ;; description of the functionality and return value.
3011 ;;
3012 ;; It suppresses the syntactic effect of the < and > (template) brackets and
3013 ;; of all parens in preprocessor constructs, except for any such construct
3014 ;; containing point. We can then call `c-parse-state-1' without worrying
3015 ;; further about macros and template delimiters.
3016 (let (here-cpp-beg here-cpp-end)
3017 (save-excursion
3018 (when (c-beginning-of-macro)
3019 (setq here-cpp-beg (point))
3020 (unless
3021 (> (setq here-cpp-end (c-syntactic-end-of-macro))
3022 here-cpp-beg)
3023 (setq here-cpp-beg nil here-cpp-end nil))))
3024 ;; FIXME!!! Put in a `condition-case' here to protect the integrity of the
3025 ;; subsystem.
3026 (prog1
3027 (c-with-<->-as-parens-suppressed
3028 (if (and here-cpp-beg (> here-cpp-end here-cpp-beg))
3029 (c-with-all-but-one-cpps-commented-out
3030 here-cpp-beg here-cpp-end
3031 (c-parse-state-1))
3032 (c-with-cpps-commented-out
3033 (c-parse-state-1))))
3e8f7d91
AM
3034 (setq c-state-old-cpp-beg (and here-cpp-beg (copy-marker here-cpp-beg t))
3035 c-state-old-cpp-end (and here-cpp-end (copy-marker here-cpp-end t)))
3036 )))
0ec1d2c5
AM
3037
3038;; Debug tool to catch cache inconsistencies. This is called from
3039;; 000tests.el.
a66cd3ee
MS
3040(defvar c-debug-parse-state nil)
3041(unless (fboundp 'c-real-parse-state)
3042 (fset 'c-real-parse-state (symbol-function 'c-parse-state)))
3043(cc-bytecomp-defun c-real-parse-state)
3044(defun c-debug-parse-state ()
0ec1d2c5 3045 (let ((here (point)) (res1 (c-real-parse-state)) res2)
0386b551 3046 (let ((c-state-cache nil)
0ec1d2c5
AM
3047 (c-state-cache-good-pos 1)
3048 (c-state-nonlit-pos-cache nil)
3049 (c-state-nonlit-pos-cache-limit 1)
3050 (c-state-brace-pair-desert nil)
3051 (c-state-point-min 1)
3052 (c-state-point-min-lit-type nil)
3053 (c-state-point-min-lit-start nil)
3054 (c-state-min-scan-pos 1)
3055 (c-state-old-cpp-beg nil)
3056 (c-state-old-cpp-end nil))
a66cd3ee
MS
3057 (setq res2 (c-real-parse-state)))
3058 (unless (equal res1 res2)
0386b551
AM
3059 ;; The cache can actually go further back due to the ad-hoc way
3060 ;; the first paren is found, so try to whack off a bit of its
3061 ;; start before complaining.
3062 (save-excursion
3063 (goto-char (or (c-least-enclosing-brace res2) (point)))
3064 (c-beginning-of-defun-1)
3065 (while (not (or (bobp) (eq (char-after) ?{)))
3066 (c-beginning-of-defun-1))
3067 (unless (equal (c-whack-state-before (point) res1) res2)
0ec1d2c5 3068 (message (concat "c-parse-state inconsistency at %s: "
0386b551 3069 "using cache: %s, from scratch: %s")
0ec1d2c5 3070 here res1 res2))))
a66cd3ee 3071 res1))
0ec1d2c5 3072
a66cd3ee
MS
3073(defun c-toggle-parse-state-debug (&optional arg)
3074 (interactive "P")
3075 (setq c-debug-parse-state (c-calculate-state arg c-debug-parse-state))
3076 (fset 'c-parse-state (symbol-function (if c-debug-parse-state
3077 'c-debug-parse-state
3078 'c-real-parse-state)))
3079 (c-keep-region-active))
0386b551
AM
3080(when c-debug-parse-state
3081 (c-toggle-parse-state-debug 1))
a66cd3ee 3082
0ec1d2c5 3083\f
d9e94c22
MS
3084(defun c-whack-state-before (bufpos paren-state)
3085 ;; Whack off any state information from PAREN-STATE which lies
3086 ;; before BUFPOS. Not destructive on PAREN-STATE.
d9e94c22
MS
3087 (let* ((newstate (list nil))
3088 (ptr newstate)
3089 car)
3090 (while paren-state
3091 (setq car (car paren-state)
3092 paren-state (cdr paren-state))
3093 (if (< (if (consp car) (car car) car) bufpos)
3094 (setq paren-state nil)
3095 (setcdr ptr (list car))
3096 (setq ptr (cdr ptr))))
3097 (cdr newstate)))
3098
3099(defun c-whack-state-after (bufpos paren-state)
3100 ;; Whack off any state information from PAREN-STATE which lies at or
3101 ;; after BUFPOS. Not destructive on PAREN-STATE.
d9e94c22
MS
3102 (catch 'done
3103 (while paren-state
3104 (let ((car (car paren-state)))
3105 (if (consp car)
3106 ;; just check the car, because in a balanced brace
3107 ;; expression, it must be impossible for the corresponding
3108 ;; close brace to be before point, but the open brace to
3109 ;; be after.
3110 (if (<= bufpos (car car))
3111 nil ; whack it off
3112 (if (< bufpos (cdr car))
3113 ;; its possible that the open brace is before
3114 ;; bufpos, but the close brace is after. In that
3115 ;; case, convert this to a non-cons element. The
3116 ;; rest of the state is before bufpos, so we're
3117 ;; done.
3118 (throw 'done (cons (car car) (cdr paren-state)))
3119 ;; we know that both the open and close braces are
3120 ;; before bufpos, so we also know that everything else
3121 ;; on state is before bufpos.
3122 (throw 'done paren-state)))
3123 (if (<= bufpos car)
3124 nil ; whack it off
3125 ;; it's before bufpos, so everything else should too.
3126 (throw 'done paren-state)))
3127 (setq paren-state (cdr paren-state)))
3128 nil)))
3129
3130(defun c-most-enclosing-brace (paren-state &optional bufpos)
3131 ;; Return the bufpos of the innermost enclosing open paren before
0386b551 3132 ;; bufpos, or nil if none was found.
d9e94c22
MS
3133 (let (enclosingp)
3134 (or bufpos (setq bufpos 134217727))
3135 (while paren-state
3136 (setq enclosingp (car paren-state)
3137 paren-state (cdr paren-state))
3138 (if (or (consp enclosingp)
3139 (>= enclosingp bufpos))
3140 (setq enclosingp nil)
d9e94c22
MS
3141 (setq paren-state nil)))
3142 enclosingp))
3143
0386b551
AM
3144(defun c-least-enclosing-brace (paren-state)
3145 ;; Return the bufpos of the outermost enclosing open paren, or nil
3146 ;; if none was found.
d9e94c22 3147 (let (pos elem)
d9e94c22
MS
3148 (while paren-state
3149 (setq elem (car paren-state)
3150 paren-state (cdr paren-state))
0386b551
AM
3151 (if (integerp elem)
3152 (setq pos elem)))
d9e94c22
MS
3153 pos))
3154
3155(defun c-safe-position (bufpos paren-state)
0386b551
AM
3156 ;; Return the closest "safe" position recorded on PAREN-STATE that
3157 ;; is higher up than BUFPOS. Return nil if PAREN-STATE doesn't
3158 ;; contain any. Return nil if BUFPOS is nil, which is useful to
3159 ;; find the closest limit before a given limit that might be nil.
d9e94c22 3160 ;;
0386b551
AM
3161 ;; A "safe" position is a position at or after a recorded open
3162 ;; paren, or after a recorded close paren. The returned position is
3163 ;; thus either the first position after a close brace, or the first
3164 ;; position after an enclosing paren, or at the enclosing paren in
3165 ;; case BUFPOS is immediately after it.
d9e94c22
MS
3166 (when bufpos
3167 (let (elem)
3168 (catch 'done
3169 (while paren-state
3170 (setq elem (car paren-state))
3171 (if (consp elem)
3172 (cond ((< (cdr elem) bufpos)
3173 (throw 'done (cdr elem)))
3174 ((< (car elem) bufpos)
3175 ;; See below.
3176 (throw 'done (min (1+ (car elem)) bufpos))))
3177 (if (< elem bufpos)
3178 ;; elem is the position at and not after the opening paren, so
3179 ;; we can go forward one more step unless it's equal to
3180 ;; bufpos. This is useful in some cases avoid an extra paren
3181 ;; level between the safe position and bufpos.
3182 (throw 'done (min (1+ elem) bufpos))))
3183 (setq paren-state (cdr paren-state)))))))
3184
3185(defun c-beginning-of-syntax ()
3186 ;; This is used for `font-lock-beginning-of-syntax-function'. It
3187 ;; goes to the closest previous point that is known to be outside
3188 ;; any string literal or comment. `c-state-cache' is used if it has
3189 ;; a position in the vicinity.
3190 (let* ((paren-state c-state-cache)
3191 elem
3192
3193 (pos (catch 'done
3194 ;; Note: Similar code in `c-safe-position'. The
3195 ;; difference is that we accept a safe position at
3196 ;; the point and don't bother to go forward past open
3197 ;; parens.
3198 (while paren-state
3199 (setq elem (car paren-state))
3200 (if (consp elem)
3201 (cond ((<= (cdr elem) (point))
3202 (throw 'done (cdr elem)))
3203 ((<= (car elem) (point))
3204 (throw 'done (car elem))))
3205 (if (<= elem (point))
3206 (throw 'done elem)))
3207 (setq paren-state (cdr paren-state)))
3208 (point-min))))
3209
3210 (if (> pos (- (point) 4000))
3211 (goto-char pos)
3212 ;; The position is far back. Try `c-beginning-of-defun-1'
3213 ;; (although we can't be entirely sure it will go to a position
3214 ;; outside a comment or string in current emacsen). FIXME:
3215 ;; Consult `syntax-ppss' here.
3216 (c-beginning-of-defun-1)
3217 (if (< (point) pos)
3218 (goto-char pos)))))
3219
3220\f
3221;; Tools for scanning identifiers and other tokens.
3222
3223(defun c-on-identifier ()
3224 "Return non-nil if the point is on or directly after an identifier.
3225Keywords are recognized and not considered identifiers. If an
3226identifier is detected, the returned value is its starting position.
0386b551
AM
3227If an identifier ends at the point and another begins at it \(can only
3228happen in Pike) then the point for the preceding one is returned.
d9e94c22 3229
0386b551
AM
3230Note that this function might do hidden buffer changes. See the
3231comment at the start of cc-engine.el for more info."
3232
3233 ;; FIXME: Shouldn't this function handle "operator" in C++?
d9e94c22
MS
3234
3235 (save-excursion
0386b551
AM
3236 (skip-syntax-backward "w_")
3237
3238 (or
3239
3240 ;; Check for a normal (non-keyword) identifier.
3241 (and (looking-at c-symbol-start)
3242 (not (looking-at c-keywords-regexp))
3243 (point))
3244
3245 (when (c-major-mode-is 'pike-mode)
3246 ;; Handle the `<operator> syntax in Pike.
3247 (let ((pos (point)))
3248 (skip-chars-backward "-!%&*+/<=>^|~[]()")
3249 (and (if (< (skip-chars-backward "`") 0)
3250 t
3251 (goto-char pos)
3252 (eq (char-after) ?\`))
3253 (looking-at c-symbol-key)
3254 (>= (match-end 0) pos)
3255 (point))))
3256
3257 ;; Handle the "operator +" syntax in C++.
3258 (when (and c-overloadable-operators-regexp
3259 (= (c-backward-token-2 0) 0))
3260
3261 (cond ((and (looking-at c-overloadable-operators-regexp)
51c9af45 3262 (or (not c-opt-op-identifier-prefix)
0386b551 3263 (and (= (c-backward-token-2 1) 0)
51c9af45 3264 (looking-at c-opt-op-identifier-prefix))))
0386b551
AM
3265 (point))
3266
3267 ((save-excursion
51c9af45
AM
3268 (and c-opt-op-identifier-prefix
3269 (looking-at c-opt-op-identifier-prefix)
0386b551
AM
3270 (= (c-forward-token-2 1) 0)
3271 (looking-at c-overloadable-operators-regexp)))
3272 (point))))
3273
3274 )))
d9e94c22
MS
3275
3276(defsubst c-simple-skip-symbol-backward ()
3277 ;; If the point is at the end of a symbol then skip backward to the
3278 ;; beginning of it. Don't move otherwise. Return non-nil if point
3279 ;; moved.
0386b551
AM
3280 ;;
3281 ;; This function might do hidden buffer changes.
d9e94c22
MS
3282 (or (< (skip-syntax-backward "w_") 0)
3283 (and (c-major-mode-is 'pike-mode)
3284 ;; Handle the `<operator> syntax in Pike.
3285 (let ((pos (point)))
2a15eb73 3286 (if (and (< (skip-chars-backward "-!%&*+/<=>^|~[]()") 0)
d9e94c22
MS
3287 (< (skip-chars-backward "`") 0)
3288 (looking-at c-symbol-key)
3289 (>= (match-end 0) pos))
3290 t
3291 (goto-char pos)
3292 nil)))))
3293
0386b551 3294(defun c-beginning-of-current-token (&optional back-limit)
d9e94c22
MS
3295 ;; Move to the beginning of the current token. Do not move if not
3296 ;; in the middle of one. BACK-LIMIT may be used to bound the
3297 ;; backward search; if given it's assumed to be at the boundary
a85fd6da 3298 ;; between two tokens. Return non-nil if the point is moved, nil
580fba94 3299 ;; otherwise.
0386b551
AM
3300 ;;
3301 ;; This function might do hidden buffer changes.
d9e94c22 3302 (let ((start (point)))
580fba94
AM
3303 (if (looking-at "\\w\\|\\s_")
3304 (skip-syntax-backward "w_" back-limit)
3305 (when (< (skip-syntax-backward ".()" back-limit) 0)
3306 (while (let ((pos (or (and (looking-at c-nonsymbol-token-regexp)
3307 (match-end 0))
3308 ;; `c-nonsymbol-token-regexp' should always match
3309 ;; since we've skipped backward over punctuator
3310 ;; or paren syntax, but consume one char in case
3311 ;; it doesn't so that we don't leave point before
3312 ;; some earlier incorrect token.
3313 (1+ (point)))))
3314 (if (<= pos start)
3315 (goto-char pos))))))
3316 (< (point) start)))
d9e94c22 3317
ff959bab 3318(defun c-end-of-current-token (&optional back-limit)
d9e94c22
MS
3319 ;; Move to the end of the current token. Do not move if not in the
3320 ;; middle of one. BACK-LIMIT may be used to bound the backward
3321 ;; search; if given it's assumed to be at the boundary between two
ff959bab 3322 ;; tokens. Return non-nil if the point is moved, nil otherwise.
0386b551
AM
3323 ;;
3324 ;; This function might do hidden buffer changes.
d9e94c22
MS
3325 (let ((start (point)))
3326 (cond ((< (skip-syntax-backward "w_" (1- start)) 0)
3327 (skip-syntax-forward "w_"))
3328 ((< (skip-syntax-backward ".()" back-limit) 0)
3329 (while (progn
3330 (if (looking-at c-nonsymbol-token-regexp)
3331 (goto-char (match-end 0))
3332 ;; `c-nonsymbol-token-regexp' should always match since
3333 ;; we've skipped backward over punctuator or paren
3334 ;; syntax, but move forward in case it doesn't so that
3335 ;; we don't leave point earlier than we started with.
3336 (forward-char))
ff959bab
MS
3337 (< (point) start)))))
3338 (> (point) start)))
d9e94c22
MS
3339
3340(defconst c-jump-syntax-balanced
3341 (if (memq 'gen-string-delim c-emacs-features)
3342 "\\w\\|\\s_\\|\\s\(\\|\\s\)\\|\\s\"\\|\\s|"
3343 "\\w\\|\\s_\\|\\s\(\\|\\s\)\\|\\s\""))
3344
3345(defconst c-jump-syntax-unbalanced
3346 (if (memq 'gen-string-delim c-emacs-features)
3347 "\\w\\|\\s_\\|\\s\"\\|\\s|"
3348 "\\w\\|\\s_\\|\\s\""))
3349
3350(defun c-forward-token-2 (&optional count balanced limit)
3351 "Move forward by tokens.
3352A token is defined as all symbols and identifiers which aren't
3353syntactic whitespace \(note that multicharacter tokens like \"==\" are
3354treated properly). Point is always either left at the beginning of a
3355token or not moved at all. COUNT specifies the number of tokens to
3356move; a negative COUNT moves in the opposite direction. A COUNT of 0
3357moves to the next token beginning only if not already at one. If
3358BALANCED is true, move over balanced parens, otherwise move into them.
3359Also, if BALANCED is true, never move out of an enclosing paren.
3360
3361LIMIT sets the limit for the movement and defaults to the point limit.
3362The case when LIMIT is set in the middle of a token, comment or macro
3363is handled correctly, i.e. the point won't be left there.
3364
3365Return the number of tokens left to move \(positive or negative). If
3366BALANCED is true, a move over a balanced paren counts as one. Note
3367that if COUNT is 0 and no appropriate token beginning is found, 1 will
3368be returned. Thus, a return value of 0 guarantees that point is at
3369the requested position and a return value less \(without signs) than
0386b551
AM
3370COUNT guarantees that point is at the beginning of some token.
3371
3372Note that this function might do hidden buffer changes. See the
3373comment at the start of cc-engine.el for more info."
d9e94c22
MS
3374
3375 (or count (setq count 1))
3376 (if (< count 0)
3377 (- (c-backward-token-2 (- count) balanced limit))
3378
3379 (let ((jump-syntax (if balanced
3380 c-jump-syntax-balanced
3381 c-jump-syntax-unbalanced))
3382 (last (point))
3383 (prev (point)))
3384
3385 (if (zerop count)
3386 ;; If count is zero we should jump if in the middle of a token.
3387 (c-end-of-current-token))
3388
3389 (save-restriction
3390 (if limit (narrow-to-region (point-min) limit))
3391 (if (/= (point)
3392 (progn (c-forward-syntactic-ws) (point)))
3393 ;; Skip whitespace. Count this as a move if we did in
3394 ;; fact move.
3395 (setq count (max (1- count) 0)))
3396
3397 (if (eobp)
3398 ;; Moved out of bounds. Make sure the returned count isn't zero.
3399 (progn
3400 (if (zerop count) (setq count 1))
3401 (goto-char last))
3402
3403 ;; Use `condition-case' to avoid having the limit tests
3404 ;; inside the loop.
3405 (condition-case nil
3406 (while (and
3407 (> count 0)
3408 (progn
3409 (setq last (point))
3410 (cond ((looking-at jump-syntax)
3411 (goto-char (scan-sexps (point) 1))
3412 t)
3413 ((looking-at c-nonsymbol-token-regexp)
3414 (goto-char (match-end 0))
3415 t)
3416 ;; `c-nonsymbol-token-regexp' above should always
3417 ;; match if there are correct tokens. Try to
3418 ;; widen to see if the limit was set in the
3419 ;; middle of one, else fall back to treating
3420 ;; the offending thing as a one character token.
3421 ((and limit
3422 (save-restriction
3423 (widen)
3424 (looking-at c-nonsymbol-token-regexp)))
3425 nil)
3426 (t
3427 (forward-char)
3428 t))))
3429 (c-forward-syntactic-ws)
3430 (setq prev last
3431 count (1- count)))
3432 (error (goto-char last)))
3433
3434 (when (eobp)
3435 (goto-char prev)
3436 (setq count (1+ count)))))
3437
3438 count)))
3439
3440(defun c-backward-token-2 (&optional count balanced limit)
3441 "Move backward by tokens.
3442See `c-forward-token-2' for details."
3443
3444 (or count (setq count 1))
3445 (if (< count 0)
3446 (- (c-forward-token-2 (- count) balanced limit))
3447
3448 (or limit (setq limit (point-min)))
3449 (let ((jump-syntax (if balanced
3450 c-jump-syntax-balanced
3451 c-jump-syntax-unbalanced))
3452 (last (point)))
3453
3454 (if (zerop count)
3455 ;; The count is zero so try to skip to the beginning of the
3456 ;; current token.
3457 (if (> (point)
3458 (progn (c-beginning-of-current-token) (point)))
3459 (if (< (point) limit)
3460 ;; The limit is inside the same token, so return 1.
3461 (setq count 1))
3462
3463 ;; We're not in the middle of a token. If there's
3464 ;; whitespace after the point then we must move backward,
3465 ;; so set count to 1 in that case.
3466 (and (looking-at c-syntactic-ws-start)
3467 ;; If we're looking at a '#' that might start a cpp
3468 ;; directive then we have to do a more elaborate check.
3469 (or (/= (char-after) ?#)
3470 (not c-opt-cpp-prefix)
3471 (save-excursion
3472 (and (= (point)
3473 (progn (beginning-of-line)
3474 (looking-at "[ \t]*")
3475 (match-end 0)))
3476 (or (bobp)
3477 (progn (backward-char)
3478 (not (eq (char-before) ?\\)))))))
3479 (setq count 1))))
3480
3481 ;; Use `condition-case' to avoid having to check for buffer
3482 ;; limits in `backward-char', `scan-sexps' and `goto-char' below.
3483 (condition-case nil
3484 (while (and
3485 (> count 0)
3486 (progn
3487 (c-backward-syntactic-ws)
3488 (backward-char)
3489 (if (looking-at jump-syntax)
3490 (goto-char (scan-sexps (1+ (point)) -1))
3491 ;; This can be very inefficient if there's a long
3492 ;; sequence of operator tokens without any separation.
3493 ;; That doesn't happen in practice, anyway.
3494 (c-beginning-of-current-token))
3495 (>= (point) limit)))
3496 (setq last (point)
3497 count (1- count)))
3498 (error (goto-char last)))
3499
3500 (if (< (point) limit)
3501 (goto-char last))
3502
3503 count)))
3504
3505(defun c-forward-token-1 (&optional count balanced limit)
3506 "Like `c-forward-token-2' but doesn't treat multicharacter operator
3507tokens like \"==\" as single tokens, i.e. all sequences of symbol
3508characters are jumped over character by character. This function is
3509for compatibility only; it's only a wrapper over `c-forward-token-2'."
3510 (let ((c-nonsymbol-token-regexp "\\s.\\|\\s\(\\|\\s\)"))
3511 (c-forward-token-2 count balanced limit)))
3512
3513(defun c-backward-token-1 (&optional count balanced limit)
3514 "Like `c-backward-token-2' but doesn't treat multicharacter operator
3515tokens like \"==\" as single tokens, i.e. all sequences of symbol
3516characters are jumped over character by character. This function is
3517for compatibility only; it's only a wrapper over `c-backward-token-2'."
3518 (let ((c-nonsymbol-token-regexp "\\s.\\|\\s\(\\|\\s\)"))
3519 (c-backward-token-2 count balanced limit)))
3520
3521\f
3522;; Tools for doing searches restricted to syntactically relevant text.
3523
3524(defun c-syntactic-re-search-forward (regexp &optional bound noerror
3525 paren-level not-inside-token
3526 lookbehind-submatch)
3527 "Like `re-search-forward', but only report matches that are found
3528in syntactically significant text. I.e. matches in comments, macros
3529or string literals are ignored. The start point is assumed to be
3530outside any comment, macro or string literal, or else the content of
3531that region is taken as syntactically significant text.
3532
3533If PAREN-LEVEL is non-nil, an additional restriction is added to
2a15eb73
MS
3534ignore matches in nested paren sexps. The search will also not go
3535outside the current list sexp, which has the effect that if the point
3536should be moved to BOUND when no match is found \(i.e. NOERROR is
3537neither nil nor t), then it will be at the closing paren if the end of
3538the current list sexp is encountered first.
d9e94c22
MS
3539
3540If NOT-INSIDE-TOKEN is non-nil, matches in the middle of tokens are
3541ignored. Things like multicharacter operators and special symbols
3542\(e.g. \"`()\" in Pike) are handled but currently not floating point
3543constants.
3544
3545If LOOKBEHIND-SUBMATCH is non-nil, it's taken as a number of a
3546subexpression in REGEXP. The end of that submatch is used as the
3547position to check for syntactic significance. If LOOKBEHIND-SUBMATCH
3548isn't used or if that subexpression didn't match then the start
3549position of the whole match is used instead. The \"look behind\"
3550subexpression is never tested before the starting position, so it
3551might be a good idea to include \\=\\= as a match alternative in it.
3552
3553Optimization note: Matches might be missed if the \"look behind\"
2a15eb73 3554subexpression can match the end of nonwhite syntactic whitespace,
d9e94c22 3555i.e. the end of comments or cpp directives. This since the function
2a15eb73
MS
3556skips over such things before resuming the search. It's on the other
3557hand not safe to assume that the \"look behind\" subexpression never
3558matches syntactic whitespace.
3559
3560Bug: Unbalanced parens inside cpp directives are currently not handled
3561correctly \(i.e. they don't get ignored as they should) when
0386b551
AM
3562PAREN-LEVEL is set.
3563
3564Note that this function might do hidden buffer changes. See the
3565comment at the start of cc-engine.el for more info."
d9e94c22
MS
3566
3567 (or bound (setq bound (point-max)))
3568 (if paren-level (setq paren-level -1))
3569
3570 ;;(message "c-syntactic-re-search-forward %s %s %S" (point) bound regexp)
3571
3572 (let ((start (point))
2a15eb73
MS
3573 tmp
3574 ;; Start position for the last search.
3575 search-pos
3576 ;; The `parse-partial-sexp' state between the start position
3577 ;; and the point.
3578 state
3579 ;; The current position after the last state update. The next
3580 ;; `parse-partial-sexp' continues from here.
3581 (state-pos (point))
3582 ;; The position at which to check the state and the state
3583 ;; there. This is separate from `state-pos' since we might
3584 ;; need to back up before doing the next search round.
3585 check-pos check-state
3586 ;; Last position known to end a token.
d9e94c22 3587 (last-token-end-pos (point-min))
2a15eb73
MS
3588 ;; Set when a valid match is found.
3589 found)
d9e94c22
MS
3590
3591 (condition-case err
3592 (while
3593 (and
2a15eb73
MS
3594 (progn
3595 (setq search-pos (point))
3596 (re-search-forward regexp bound noerror))
d9e94c22
MS
3597
3598 (progn
2a15eb73
MS
3599 (setq state (parse-partial-sexp
3600 state-pos (match-beginning 0) paren-level nil state)
3601 state-pos (point))
d9e94c22 3602 (if (setq check-pos (and lookbehind-submatch
2a15eb73
MS
3603 (or (not paren-level)
3604 (>= (car state) 0))
d9e94c22
MS
3605 (match-end lookbehind-submatch)))
3606 (setq check-state (parse-partial-sexp
2a15eb73
MS
3607 state-pos check-pos paren-level nil state))
3608 (setq check-pos state-pos
d9e94c22
MS
3609 check-state state))
3610
2a15eb73
MS
3611 ;; NOTE: If we got a look behind subexpression and get
3612 ;; an insignificant match in something that isn't
d9e94c22
MS
3613 ;; syntactic whitespace (i.e. strings or in nested
3614 ;; parentheses), then we can never skip more than a
2a15eb73
MS
3615 ;; single character from the match start position
3616 ;; (i.e. `state-pos' here) before continuing the
3617 ;; search. That since the look behind subexpression
3618 ;; might match the end of the insignificant region in
3619 ;; the next search.
d9e94c22
MS
3620
3621 (cond
d9e94c22
MS
3622 ((elt check-state 7)
3623 ;; Match inside a line comment. Skip to eol. Use
3624 ;; `re-search-forward' instead of `skip-chars-forward' to get
3625 ;; the right bound behavior.
3626 (re-search-forward "[\n\r]" bound noerror))
3627
3628 ((elt check-state 4)
3629 ;; Match inside a block comment. Skip to the '*/'.
3630 (search-forward "*/" bound noerror))
3631
3632 ((and (not (elt check-state 5))
3633 (eq (char-before check-pos) ?/)
2a15eb73 3634 (not (c-get-char-property (1- check-pos) 'syntax-table))
d9e94c22
MS
3635 (memq (char-after check-pos) '(?/ ?*)))
3636 ;; Match in the middle of the opener of a block or line
3637 ;; comment.
3638 (if (= (char-after check-pos) ?/)
3639 (re-search-forward "[\n\r]" bound noerror)
3640 (search-forward "*/" bound noerror)))
3641
2a15eb73
MS
3642 ;; The last `parse-partial-sexp' above might have
3643 ;; stopped short of the real check position if the end
3644 ;; of the current sexp was encountered in paren-level
3645 ;; mode. The checks above are always false in that
3646 ;; case, and since they can do better skipping in
3647 ;; lookbehind-submatch mode, we do them before
3648 ;; checking the paren level.
3649
3650 ((and paren-level
3651 (/= (setq tmp (car check-state)) 0))
3652 ;; Check the paren level first since we're short of the
3653 ;; syntactic checking position if the end of the
3654 ;; current sexp was encountered by `parse-partial-sexp'.
3655 (if (> tmp 0)
3656
3657 ;; Inside a nested paren sexp.
3658 (if lookbehind-submatch
3659 ;; See the NOTE above.
3660 (progn (goto-char state-pos) t)
3661 ;; Skip out of the paren quickly.
3662 (setq state (parse-partial-sexp state-pos bound 0 nil state)
3663 state-pos (point)))
3664
3665 ;; Have exited the current paren sexp.
3666 (if noerror
3667 (progn
3668 ;; The last `parse-partial-sexp' call above
3669 ;; has left us just after the closing paren
3670 ;; in this case, so we can modify the bound
3671 ;; to leave the point at the right position
3672 ;; upon return.
3673 (setq bound (1- (point)))
3674 nil)
3675 (signal 'search-failed (list regexp)))))
3676
3677 ((setq tmp (elt check-state 3))
3678 ;; Match inside a string.
3679 (if (or lookbehind-submatch
3680 (not (integerp tmp)))
3681 ;; See the NOTE above.
3682 (progn (goto-char state-pos) t)
3683 ;; Skip to the end of the string before continuing.
3684 (let ((ender (make-string 1 tmp)) (continue t))
3685 (while (if (search-forward ender bound noerror)
3686 (progn
3687 (setq state (parse-partial-sexp
3688 state-pos (point) nil nil state)
3689 state-pos (point))
3690 (elt state 3))
3691 (setq continue nil)))
3692 continue)))
d9e94c22
MS
3693
3694 ((save-excursion
3695 (save-match-data
3696 (c-beginning-of-macro start)))
3697 ;; Match inside a macro. Skip to the end of it.
3698 (c-end-of-macro)
3699 (cond ((<= (point) bound) t)
3700 (noerror nil)
2a15eb73 3701 (t (signal 'search-failed (list regexp)))))
d9e94c22 3702
2a15eb73
MS
3703 ((and not-inside-token
3704 (or (< check-pos last-token-end-pos)
3705 (< check-pos
3706 (save-excursion
3707 (goto-char check-pos)
3708 (save-match-data
3709 (c-end-of-current-token last-token-end-pos))
3710 (setq last-token-end-pos (point))))))
3711 ;; Inside a token.
3712 (if lookbehind-submatch
3713 ;; See the NOTE above.
3714 (goto-char state-pos)
3715 (goto-char (min last-token-end-pos bound))))
d9e94c22
MS
3716
3717 (t
3718 ;; A real match.
3719 (setq found t)
2a15eb73
MS
3720 nil)))
3721
3722 ;; Should loop to search again, but take care to avoid
3723 ;; looping on the same spot.
3724 (or (/= search-pos (point))
3725 (if (= (point) bound)
3726 (if noerror
3727 nil
3728 (signal 'search-failed (list regexp)))
3729 (forward-char)
3730 t))))
d9e94c22
MS
3731
3732 (error
3733 (goto-char start)
3734 (signal (car err) (cdr err))))
3735
2a15eb73 3736 ;;(message "c-syntactic-re-search-forward done %s" (or (match-end 0) (point)))
d9e94c22
MS
3737
3738 (if found
3739 (progn
2a15eb73
MS
3740 (goto-char (match-end 0))
3741 (match-end 0))
d9e94c22
MS
3742
3743 ;; Search failed. Set point as appropriate.
2a15eb73
MS
3744 (if (eq noerror t)
3745 (goto-char start)
3746 (goto-char bound))
d9e94c22
MS
3747 nil)))
3748
47641aac
GM
3749(defvar safe-pos-list) ; bound in c-syntactic-skip-backward
3750
d0fcee66
AM
3751(defsubst c-ssb-lit-begin ()
3752 ;; Return the start of the literal point is in, or nil.
3753 ;; We read and write the variables `safe-pos', `safe-pos-list', `state'
3754 ;; bound in the caller.
3755
3756 ;; Use `parse-partial-sexp' from a safe position down to the point to check
3757 ;; if it's outside comments and strings.
3758 (save-excursion
3759 (let ((pos (point)) safe-pos state pps-end-pos)
3760 ;; Pick a safe position as close to the point as possible.
3761 ;;
3762 ;; FIXME: Consult `syntax-ppss' here if our cache doesn't give a good
3763 ;; position.
47641aac 3764
d0fcee66
AM
3765 (while (and safe-pos-list
3766 (> (car safe-pos-list) (point)))
3767 (setq safe-pos-list (cdr safe-pos-list)))
3768 (unless (setq safe-pos (car-safe safe-pos-list))
3769 (setq safe-pos (max (or (c-safe-position
3770 (point) (or c-state-cache
3771 (c-parse-state)))
3772 0)
3773 (point-min))
3774 safe-pos-list (list safe-pos)))
3775
3776 ;; Cache positions along the way to use if we have to back up more. We
3777 ;; cache every closing paren on the same level. If the paren cache is
3778 ;; relevant in this region then we're typically already on the same
3779 ;; level as the target position. Note that we might cache positions
3780 ;; after opening parens in case safe-pos is in a nested list. That's
3781 ;; both uncommon and harmless.
3782 (while (progn
3783 (setq state (parse-partial-sexp
3784 safe-pos pos 0))
3785 (< (point) pos))
3786 (setq safe-pos (point)
3787 safe-pos-list (cons safe-pos safe-pos-list)))
3788
3789 ;; If the state contains the start of the containing sexp we cache that
3790 ;; position too, so that parse-partial-sexp in the next run has a bigger
3791 ;; chance of starting at the same level as the target position and thus
3792 ;; will get more good safe positions into the list.
3793 (if (elt state 1)
3794 (setq safe-pos (1+ (elt state 1))
3795 safe-pos-list (cons safe-pos safe-pos-list)))
3796
3797 (if (or (elt state 3) (elt state 4))
3798 ;; Inside string or comment. Continue search at the
3799 ;; beginning of it.
3800 (elt state 8)))))
3801
0386b551 3802(defun c-syntactic-skip-backward (skip-chars &optional limit paren-level)
d9e94c22
MS
3803 "Like `skip-chars-backward' but only look at syntactically relevant chars,
3804i.e. don't stop at positions inside syntactic whitespace or string
3805literals. Preprocessor directives are also ignored, with the exception
3806of the one that the point starts within, if any. If LIMIT is given,
0386b551
AM
3807it's assumed to be at a syntactically relevant position.
3808
3809If PAREN-LEVEL is non-nil, the function won't stop in nested paren
3810sexps, and the search will also not go outside the current paren sexp.
3811However, if LIMIT or the buffer limit is reached inside a nested paren
3812then the point will be left at the limit.
3813
3814Non-nil is returned if the point moved, nil otherwise.
3815
3816Note that this function might do hidden buffer changes. See the
3817comment at the start of cc-engine.el for more info."
d9e94c22
MS
3818
3819 (let ((start (point))
d0fcee66 3820 state-2
d9e94c22
MS
3821 ;; A list of syntactically relevant positions in descending
3822 ;; order. It's used to avoid scanning repeatedly over
3823 ;; potentially large regions with `parse-partial-sexp' to verify
d0fcee66 3824 ;; each position. Used in `c-ssb-lit-begin'
d9e94c22
MS
3825 safe-pos-list
3826 ;; The result from `c-beginning-of-macro' at the start position or the
3827 ;; start position itself if it isn't within a macro. Evaluated on
3828 ;; demand.
0386b551
AM
3829 start-macro-beg
3830 ;; The earliest position after the current one with the same paren
3831 ;; level. Used only when `paren-level' is set.
d0fcee66 3832 lit-beg
0386b551 3833 (paren-level-pos (point)))
d9e94c22 3834
d0fcee66
AM
3835 (while
3836 (progn
3837 ;; The next loop "tries" to find the end point each time round,
3838 ;; loops when it hasn't succeeded.
3839 (while
3840 (and
3841 (< (skip-chars-backward skip-chars limit) 0)
3842
3843 (let ((pos (point)) state-2 pps-end-pos)
3844
3845 (cond
3846 ;; Don't stop inside a literal
3847 ((setq lit-beg (c-ssb-lit-begin))
3848 (goto-char lit-beg)
3849 t)
3850
3851 ((and paren-level
3852 (save-excursion
3853 (setq state-2 (parse-partial-sexp
3854 pos paren-level-pos -1)
3855 pps-end-pos (point))
3856 (/= (car state-2) 0)))
3857 ;; Not at the right level.
3858
3859 (if (and (< (car state-2) 0)
3860 ;; We stop above if we go out of a paren.
3861 ;; Now check whether it precedes or is
3862 ;; nested in the starting sexp.
3863 (save-excursion
3864 (setq state-2
3865 (parse-partial-sexp
3866 pps-end-pos paren-level-pos
3867 nil nil state-2))
3868 (< (car state-2) 0)))
3869
3870 ;; We've stopped short of the starting position
3871 ;; so the hit was inside a nested list. Go up
3872 ;; until we are at the right level.
3873 (condition-case nil
3874 (progn
3875 (goto-char (scan-lists pos -1
3876 (- (car state-2))))
3877 (setq paren-level-pos (point))
3878 (if (and limit (>= limit paren-level-pos))
0386b551 3879 (progn
d0fcee66
AM
3880 (goto-char limit)
3881 nil)
3882 t))
3883 (error
3884 (goto-char (or limit (point-min)))
3885 nil))
3886
3887 ;; The hit was outside the list at the start
3888 ;; position. Go to the start of the list and exit.
3889 (goto-char (1+ (elt state-2 1)))
3890 nil))
3891
3892 ((c-beginning-of-macro limit)
3893 ;; Inside a macro.
3894 (if (< (point)
3895 (or start-macro-beg
3896 (setq start-macro-beg
3897 (save-excursion
3898 (goto-char start)
3899 (c-beginning-of-macro limit)
3900 (point)))))
3901 t
3902
3903 ;; It's inside the same macro we started in so it's
3904 ;; a relevant match.
3905 (goto-char pos)
3906 nil))))))
3907
3908 (> (point)
3909 (progn
3910 ;; Skip syntactic ws afterwards so that we don't stop at the
3911 ;; end of a comment if `skip-chars' is something like "^/".
3912 (c-backward-syntactic-ws)
3913 (point)))))
d9e94c22 3914
0386b551
AM
3915 ;; We might want to extend this with more useful return values in
3916 ;; the future.
3917 (/= (point) start)))
3918
3919;; The following is an alternative implementation of
3920;; `c-syntactic-skip-backward' that uses backward movement to keep
3921;; track of the syntactic context. It turned out to be generally
3922;; slower than the one above which uses forward checks from earlier
3923;; safe positions.
3924;;
3925;;(defconst c-ssb-stop-re
3926;; ;; The regexp matching chars `c-syntactic-skip-backward' needs to
3927;; ;; stop at to avoid going into comments and literals.
3928;; (concat
3929;; ;; Match comment end syntax and string literal syntax. Also match
3930;; ;; '/' for block comment endings (not covered by comment end
3931;; ;; syntax).
3932;; "\\s>\\|/\\|\\s\""
3933;; (if (memq 'gen-string-delim c-emacs-features)
3934;; "\\|\\s|"
3935;; "")
3936;; (if (memq 'gen-comment-delim c-emacs-features)
3937;; "\\|\\s!"
3938;; "")))
3939;;
3940;;(defconst c-ssb-stop-paren-re
3941;; ;; Like `c-ssb-stop-re' but also stops at paren chars.
3942;; (concat c-ssb-stop-re "\\|\\s(\\|\\s)"))
3943;;
3944;;(defconst c-ssb-sexp-end-re
3945;; ;; Regexp matching the ending syntax of a complex sexp.
3946;; (concat c-string-limit-regexp "\\|\\s)"))
3947;;
3948;;(defun c-syntactic-skip-backward (skip-chars &optional limit paren-level)
3949;; "Like `skip-chars-backward' but only look at syntactically relevant chars,
3950;;i.e. don't stop at positions inside syntactic whitespace or string
3951;;literals. Preprocessor directives are also ignored. However, if the
3952;;point is within a comment, string literal or preprocessor directory to
3953;;begin with, its contents is treated as syntactically relevant chars.
3954;;If LIMIT is given, it limits the backward search and the point will be
3955;;left there if no earlier position is found.
3956;;
3957;;If PAREN-LEVEL is non-nil, the function won't stop in nested paren
3958;;sexps, and the search will also not go outside the current paren sexp.
3959;;However, if LIMIT or the buffer limit is reached inside a nested paren
3960;;then the point will be left at the limit.
3961;;
3962;;Non-nil is returned if the point moved, nil otherwise.
3963;;
3964;;Note that this function might do hidden buffer changes. See the
3965;;comment at the start of cc-engine.el for more info."
3966;;
3967;; (save-restriction
3968;; (when limit
3969;; (narrow-to-region limit (point-max)))
3970;;
3971;; (let ((start (point)))
3972;; (catch 'done
3973;; (while (let ((last-pos (point))
3974;; (stop-pos (progn
3975;; (skip-chars-backward skip-chars)
3976;; (point))))
3977;;
3978;; ;; Skip back over the same region as
3979;; ;; `skip-chars-backward' above, but keep to
3980;; ;; syntactically relevant positions.
3981;; (goto-char last-pos)
3982;; (while (and
3983;; ;; `re-search-backward' with a single char regexp
3984;; ;; should be fast.
3985;; (re-search-backward
3986;; (if paren-level c-ssb-stop-paren-re c-ssb-stop-re)
3987;; stop-pos 'move)
3988;;
3989;; (progn
3990;; (cond
3991;; ((looking-at "\\s(")
3992;; ;; `paren-level' is set and we've found the
3993;; ;; start of the containing paren.
3994;; (forward-char)
3995;; (throw 'done t))
3996;;
3997;; ((looking-at c-ssb-sexp-end-re)
3998;; ;; We're at the end of a string literal or paren
3999;; ;; sexp (if `paren-level' is set).
4000;; (forward-char)
4001;; (condition-case nil
4002;; (c-backward-sexp)
4003;; (error
4004;; (goto-char limit)
4005;; (throw 'done t))))
4006;;
4007;; (t
4008;; (forward-char)
4009;; ;; At the end of some syntactic ws or possibly
4010;; ;; after a plain '/' operator.
4011;; (let ((pos (point)))
4012;; (c-backward-syntactic-ws)
4013;; (if (= pos (point))
4014;; ;; Was a plain '/' operator. Go past it.
4015;; (backward-char)))))
4016;;
4017;; (> (point) stop-pos))))
4018;;
4019;; ;; Now the point is either at `stop-pos' or at some
4020;; ;; position further back if `stop-pos' was at a
4021;; ;; syntactically irrelevant place.
4022;;
4023;; ;; Skip additional syntactic ws so that we don't stop
4024;; ;; at the end of a comment if `skip-chars' is
4025;; ;; something like "^/".
4026;; (c-backward-syntactic-ws)
4027;;
4028;; (< (point) stop-pos))))
4029;;
4030;; ;; We might want to extend this with more useful return values
4031;; ;; in the future.
4032;; (/= (point) start))))
d9e94c22
MS
4033
4034\f
4035;; Tools for handling comments and string literals.
4036
4037(defun c-slow-in-literal (&optional lim detect-cpp)
4038 "Return the type of literal point is in, if any.
4039The return value is `c' if in a C-style comment, `c++' if in a C++
4040style comment, `string' if in a string literal, `pound' if DETECT-CPP
4041is non-nil and in a preprocessor line, or nil if somewhere else.
4042Optional LIM is used as the backward limit of the search. If omitted,
4043or nil, `c-beginning-of-defun' is used.
4044
4045The last point calculated is cached if the cache is enabled, i.e. if
4046`c-in-literal-cache' is bound to a two element vector.
4047
0386b551
AM
4048Note that this function might do hidden buffer changes. See the
4049comment at the start of cc-engine.el for more info."
4050
d9e94c22
MS
4051 (if (and (vectorp c-in-literal-cache)
4052 (= (point) (aref c-in-literal-cache 0)))
4053 (aref c-in-literal-cache 1)
4054 (let ((rtn (save-excursion
4055 (let* ((pos (point))
4056 (lim (or lim (progn
4057 (c-beginning-of-syntax)
4058 (point))))
4059 (state (parse-partial-sexp lim pos)))
4060 (cond
4061 ((elt state 3) 'string)
4062 ((elt state 4) (if (elt state 7) 'c++ 'c))
4063 ((and detect-cpp (c-beginning-of-macro lim)) 'pound)
4064 (t nil))))))
4065 ;; cache this result if the cache is enabled
4066 (if (not c-in-literal-cache)
4067 (setq c-in-literal-cache (vector (point) rtn)))
4068 rtn)))
4069
4070;; XEmacs has a built-in function that should make this much quicker.
4071;; I don't think we even need the cache, which makes our lives more
4072;; complicated anyway. In this case, lim is only used to detect
4073;; cpp directives.
4074;;
4075;; Note that there is a bug in Xemacs's buffer-syntactic-context when used in
4076;; conjunction with syntax-table-properties. The bug is present in, e.g.,
4077;; Xemacs 21.4.4. It manifested itself thus:
4078;;
4079;; Starting with an empty AWK Mode buffer, type
4080;; /regexp/ {<C-j>
4081;; Point gets wrongly left at column 0, rather than being indented to tab-width.
4082;;
4083;; AWK Mode is designed such that when the first / is typed, it gets the
4084;; syntax-table property "string fence". When the second / is typed, BOTH /s
4085;; are given the s-t property "string". However, buffer-syntactic-context
4086;; fails to take account of the change of the s-t property on the opening / to
4087;; "string", and reports that the { is within a string started by the second /.
4088;;
4089;; The workaround for this is for the AWK Mode initialisation to switch the
4090;; defalias for c-in-literal to c-slow-in-literal. This will slow down other
4091;; cc-modes in Xemacs whenever an awk-buffer has been initialised.
b414f371 4092;;
d9e94c22
MS
4093;; (Alan Mackenzie, 2003/4/30).
4094
4095(defun c-fast-in-literal (&optional lim detect-cpp)
0386b551 4096 ;; This function might do hidden buffer changes.
d9e94c22
MS
4097 (let ((context (buffer-syntactic-context)))
4098 (cond
4099 ((eq context 'string) 'string)
4100 ((eq context 'comment) 'c++)
4101 ((eq context 'block-comment) 'c)
4102 ((and detect-cpp (save-excursion (c-beginning-of-macro lim))) 'pound))))
4103
4104(defalias 'c-in-literal
4105 (if (fboundp 'buffer-syntactic-context)
7bfc3fdb 4106 'c-fast-in-literal ; XEmacs
d9e94c22
MS
4107 'c-slow-in-literal)) ; GNU Emacs
4108
4109;; The defalias above isn't enough to shut up the byte compiler.
4110(cc-bytecomp-defun c-in-literal)
4111
4112(defun c-literal-limits (&optional lim near not-in-delimiter)
4113 "Return a cons of the beginning and end positions of the comment or
4114string surrounding point (including both delimiters), or nil if point
4115isn't in one. If LIM is non-nil, it's used as the \"safe\" position
4116to start parsing from. If NEAR is non-nil, then the limits of any
4117literal next to point is returned. \"Next to\" means there's only
4118spaces and tabs between point and the literal. The search for such a
4119literal is done first in forward direction. If NOT-IN-DELIMITER is
4120non-nil, the case when point is inside a starting delimiter won't be
a85fd6da 4121recognized. This only has effect for comments which have starting
d9e94c22
MS
4122delimiters with more than one character.
4123
0386b551
AM
4124Note that this function might do hidden buffer changes. See the
4125comment at the start of cc-engine.el for more info."
d9e94c22
MS
4126
4127 (save-excursion
4128 (let* ((pos (point))
4129 (lim (or lim (progn
4130 (c-beginning-of-syntax)
4131 (point))))
4132 (state (parse-partial-sexp lim pos)))
4133
0386b551
AM
4134 (cond ((elt state 3) ; String.
4135 (goto-char (elt state 8))
d9e94c22
MS
4136 (cons (point) (or (c-safe (c-forward-sexp 1) (point))
4137 (point-max))))
4138
0386b551
AM
4139 ((elt state 4) ; Comment.
4140 (goto-char (elt state 8))
d9e94c22
MS
4141 (cons (point) (progn (c-forward-single-comment) (point))))
4142
4143 ((and (not not-in-delimiter)
4144 (not (elt state 5))
4145 (eq (char-before) ?/)
4146 (looking-at "[/*]"))
4147 ;; We're standing in a comment starter.
4148 (backward-char 1)
4149 (cons (point) (progn (c-forward-single-comment) (point))))
4150
4151 (near
4152 (goto-char pos)
4153
4154 ;; Search forward for a literal.
4155 (skip-chars-forward " \t")
4156
4157 (cond
4158 ((looking-at c-string-limit-regexp) ; String.
4159 (cons (point) (or (c-safe (c-forward-sexp 1) (point))
4160 (point-max))))
4161
4162 ((looking-at c-comment-start-regexp) ; Line or block comment.
4163 (cons (point) (progn (c-forward-single-comment) (point))))
4164
4165 (t
4166 ;; Search backward.
4167 (skip-chars-backward " \t")
4168
4169 (let ((end (point)) beg)
4170 (cond
4171 ((save-excursion
4172 (< (skip-syntax-backward c-string-syntax) 0)) ; String.
4173 (setq beg (c-safe (c-backward-sexp 1) (point))))
4174
4175 ((and (c-safe (forward-char -2) t)
4176 (looking-at "*/"))
4177 ;; Block comment. Due to the nature of line
4178 ;; comments, they will always be covered by the
4179 ;; normal case above.
4180 (goto-char end)
4181 (c-backward-single-comment)
4182 ;; If LIM is bogus, beg will be bogus.
4183 (setq beg (point))))
4184
4185 (if beg (cons beg end))))))
4186 ))))
4187
0386b551
AM
4188;; In case external callers use this; it did have a docstring.
4189(defalias 'c-literal-limits-fast 'c-literal-limits)
d9e94c22 4190
0386b551
AM
4191(defun c-collect-line-comments (range)
4192 "If the argument is a cons of two buffer positions (such as returned by
4193`c-literal-limits'), and that range contains a C++ style line comment,
4194then an extended range is returned that contains all adjacent line
4195comments (i.e. all comments that starts in the same column with no
4196empty lines or non-whitespace characters between them). Otherwise the
4197argument is returned.
d9e94c22 4198
0386b551
AM
4199Note that this function might do hidden buffer changes. See the
4200comment at the start of cc-engine.el for more info."
d9e94c22
MS
4201
4202 (save-excursion
0386b551
AM
4203 (condition-case nil
4204 (if (and (consp range) (progn
4205 (goto-char (car range))
4206 (looking-at c-line-comment-starter)))
b414f371 4207 (let ((col (current-column))
0386b551
AM
4208 (beg (point))
4209 (bopl (c-point 'bopl))
4210 (end (cdr range)))
4211 ;; Got to take care in the backward direction to handle
4212 ;; comments which are preceded by code.
4213 (while (and (c-backward-single-comment)
4214 (>= (point) bopl)
4215 (looking-at c-line-comment-starter)
4216 (= col (current-column)))
4217 (setq beg (point)
4218 bopl (c-point 'bopl)))
4219 (goto-char end)
4220 (while (and (progn (skip-chars-forward " \t")
4221 (looking-at c-line-comment-starter))
4222 (= col (current-column))
4223 (prog1 (zerop (forward-line 1))
4224 (setq end (point)))))
4225 (cons beg end))
4226 range)
4227 (error range))))
d9e94c22
MS
4228
4229(defun c-literal-type (range)
4230 "Convenience function that given the result of `c-literal-limits',
a85fd6da
AM
4231returns nil or the type of literal that the range surrounds, one
4232of the symbols 'c, 'c++ or 'string. It's much faster than using
4233`c-in-literal' and is intended to be used when you need both the
4234type of a literal and its limits.
d9e94c22 4235
0386b551
AM
4236Note that this function might do hidden buffer changes. See the
4237comment at the start of cc-engine.el for more info."
4238
d9e94c22
MS
4239 (if (consp range)
4240 (save-excursion
4241 (goto-char (car range))
4242 (cond ((looking-at c-string-limit-regexp) 'string)
4243 ((or (looking-at "//") ; c++ line comment
4244 (and (looking-at "\\s<") ; comment starter
4245 (looking-at "#"))) ; awk comment.
4246 'c++)
4247 (t 'c))) ; Assuming the range is valid.
4248 range))
4249
4250\f
4251;; `c-find-decl-spots' and accompanying stuff.
4252
4253;; Variables used in `c-find-decl-spots' to cache the search done for
4254;; the first declaration in the last call. When that function starts,
4255;; it needs to back up over syntactic whitespace to look at the last
4256;; token before the region being searched. That can sometimes cause
4257;; moves back and forth over a quite large region of comments and
4258;; macros, which would be repeated for each changed character when
4259;; we're called during fontification, since font-lock refontifies the
4260;; current line for each change. Thus it's worthwhile to cache the
4261;; first match.
4262;;
4263;; `c-find-decl-syntactic-pos' is a syntactically relevant position in
4264;; the syntactic whitespace less or equal to some start position.
4265;; There's no cached value if it's nil.
4266;;
4267;; `c-find-decl-match-pos' is the match position if
4268;; `c-find-decl-prefix-search' matched before the syntactic whitespace
4269;; at `c-find-decl-syntactic-pos', or nil if there's no such match.
4270(defvar c-find-decl-syntactic-pos nil)
4271(make-variable-buffer-local 'c-find-decl-syntactic-pos)
4272(defvar c-find-decl-match-pos nil)
4273(make-variable-buffer-local 'c-find-decl-match-pos)
4274
4275(defsubst c-invalidate-find-decl-cache (change-min-pos)
4276 (and c-find-decl-syntactic-pos
4277 (< change-min-pos c-find-decl-syntactic-pos)
4278 (setq c-find-decl-syntactic-pos nil)))
4279
4280; (defface c-debug-decl-spot-face
4281; '((t (:background "Turquoise")))
4282; "Debug face to mark the spots where `c-find-decl-spots' stopped.")
4283; (defface c-debug-decl-sws-face
4284; '((t (:background "Khaki")))
4285; "Debug face to mark the syntactic whitespace between the declaration
4286; spots and the preceding token end.")
4287
4288(defmacro c-debug-put-decl-spot-faces (match-pos decl-pos)
4289 (when (facep 'c-debug-decl-spot-face)
0386b551 4290 `(c-save-buffer-state ((match-pos ,match-pos) (decl-pos ,decl-pos))
d9e94c22
MS
4291 (c-debug-add-face (max match-pos (point-min)) decl-pos
4292 'c-debug-decl-sws-face)
4293 (c-debug-add-face decl-pos (min (1+ decl-pos) (point-max))
4294 'c-debug-decl-spot-face))))
4295(defmacro c-debug-remove-decl-spot-faces (beg end)
4296 (when (facep 'c-debug-decl-spot-face)
0386b551 4297 `(c-save-buffer-state ()
d9e94c22
MS
4298 (c-debug-remove-face ,beg ,end 'c-debug-decl-spot-face)
4299 (c-debug-remove-face ,beg ,end 'c-debug-decl-sws-face))))
4300
4301(defmacro c-find-decl-prefix-search ()
4302 ;; Macro used inside `c-find-decl-spots'. It ought to be a defun,
4303 ;; but it contains lots of free variables that refer to things
4304 ;; inside `c-find-decl-spots'. The point is left at `cfd-match-pos'
4305 ;; if there is a match, otherwise at `cfd-limit'.
0386b551
AM
4306 ;;
4307 ;; This macro might do hidden buffer changes.
d9e94c22
MS
4308
4309 '(progn
4310 ;; Find the next property match position if we haven't got one already.
4311 (unless cfd-prop-match
4312 (save-excursion
4313 (while (progn
4314 (goto-char (next-single-property-change
4315 (point) 'c-type nil cfd-limit))
4316 (and (< (point) cfd-limit)
4317 (not (eq (c-get-char-property (1- (point)) 'c-type)
4318 'c-decl-end)))))
4319 (setq cfd-prop-match (point))))
4320
0386b551
AM
4321 ;; Find the next `c-decl-prefix-or-start-re' match if we haven't
4322 ;; got one already.
d9e94c22 4323 (unless cfd-re-match
0386b551
AM
4324
4325 (if (> cfd-re-match-end (point))
4326 (goto-char cfd-re-match-end))
4327
4328 (while (if (setq cfd-re-match-end
4329 (re-search-forward c-decl-prefix-or-start-re
4330 cfd-limit 'move))
4331
4332 ;; Match. Check if it's inside a comment or string literal.
4333 (c-got-face-at
4334 (if (setq cfd-re-match (match-end 1))
4335 ;; Matched the end of a token preceding a decl spot.
4336 (progn
4337 (goto-char cfd-re-match)
4338 (1- cfd-re-match))
4339 ;; Matched a token that start a decl spot.
4340 (goto-char (match-beginning 0))
4341 (point))
4342 c-literal-faces)
4343
4344 ;; No match. Finish up and exit the loop.
4345 (setq cfd-re-match cfd-limit)
4346 nil)
4347
4348 ;; Skip out of comments and string literals.
d9e94c22
MS
4349 (while (progn
4350 (goto-char (next-single-property-change
0386b551 4351 (point) 'face nil cfd-limit))
d9e94c22 4352 (and (< (point) cfd-limit)
0386b551
AM
4353 (c-got-face-at (point) c-literal-faces)))))
4354
4355 ;; If we matched at the decl start, we have to back up over the
4356 ;; preceding syntactic ws to set `cfd-match-pos' and to catch
4357 ;; any decl spots in the syntactic ws.
d9e94c22 4358 (unless cfd-re-match
0386b551
AM
4359 (c-backward-syntactic-ws)
4360 (setq cfd-re-match (point))))
d9e94c22
MS
4361
4362 ;; Choose whichever match is closer to the start.
4363 (if (< cfd-re-match cfd-prop-match)
4364 (setq cfd-match-pos cfd-re-match
4365 cfd-re-match nil)
4366 (setq cfd-match-pos cfd-prop-match
4367 cfd-prop-match nil))
4368
4369 (goto-char cfd-match-pos)
4370
4371 (when (< cfd-match-pos cfd-limit)
4372 ;; Skip forward past comments only so we don't skip macros.
4373 (c-forward-comments)
4374 ;; Set the position to continue at. We can avoid going over
4375 ;; the comments skipped above a second time, but it's possible
4376 ;; that the comment skipping has taken us past `cfd-prop-match'
4377 ;; since the property might be used inside comments.
4378 (setq cfd-continue-pos (if cfd-prop-match
4379 (min cfd-prop-match (point))
4380 (point))))))
4381
4382(defun c-find-decl-spots (cfd-limit cfd-decl-re cfd-face-checklist cfd-fun)
0386b551 4383 ;; Call CFD-FUN for each possible spot for a declaration, cast or
a85fd6da
AM
4384 ;; label from the point to CFD-LIMIT.
4385 ;;
4386 ;; CFD-FUN is called with point at the start of the spot. It's
4387 ;; passed two arguments: The first is the end position of the token
4388 ;; preceding the spot, or 0 for the implicit match at bob. The
4389 ;; second is a flag that is t when the match is inside a macro. If
4390 ;; CFD-FUN adds `c-decl-end' properties somewhere below the current
4391 ;; spot, it should return non-nil to ensure that the next search
4392 ;; will find them.
0386b551 4393 ;;
a85fd6da 4394 ;; Such a spot is:
0386b551
AM
4395 ;; o The first token after bob.
4396 ;; o The first token after the end of submatch 1 in
4397 ;; `c-decl-prefix-or-start-re' when that submatch matches.
4398 ;; o The start of each `c-decl-prefix-or-start-re' match when
4399 ;; submatch 1 doesn't match.
5a89f0a7 4400 ;; o The first token after the end of each occurrence of the
0386b551
AM
4401 ;; `c-type' text property with the value `c-decl-end', provided
4402 ;; `c-type-decl-end-used' is set.
4403 ;;
4404 ;; Only a spot that match CFD-DECL-RE and whose face is in the
4405 ;; CFD-FACE-CHECKLIST list causes CFD-FUN to be called. The face
4406 ;; check is disabled if CFD-FACE-CHECKLIST is nil.
d9e94c22
MS
4407 ;;
4408 ;; If the match is inside a macro then the buffer is narrowed to the
4409 ;; end of it, so that CFD-FUN can investigate the following tokens
4410 ;; without matching something that begins inside a macro and ends
4411 ;; outside it. It's to avoid this work that the CFD-DECL-RE and
4412 ;; CFD-FACE-CHECKLIST checks exist.
4413 ;;
0386b551
AM
4414 ;; The spots are visited approximately in order from top to bottom.
4415 ;; It's however the positions where `c-decl-prefix-or-start-re'
4416 ;; matches and where `c-decl-end' properties are found that are in
4417 ;; order. Since the spots often are at the following token, they
4418 ;; might be visited out of order insofar as more spots are reported
4419 ;; later on within the syntactic whitespace between the match
4420 ;; positions and their spots.
4421 ;;
4422 ;; It's assumed that comments and strings are fontified in the
d9e94c22
MS
4423 ;; searched range.
4424 ;;
4425 ;; This is mainly used in fontification, and so has an elaborate
4426 ;; cache to handle repeated calls from the same start position; see
4427 ;; the variables above.
4428 ;;
4429 ;; All variables in this function begin with `cfd-' to avoid name
4430 ;; collision with the (dynamically bound) variables used in CFD-FUN.
0386b551
AM
4431 ;;
4432 ;; This function might do hidden buffer changes.
d9e94c22 4433
0386b551
AM
4434 (let ((cfd-start-pos (point))
4435 (cfd-buffer-end (point-max))
4436 ;; The end of the token preceding the decl spot last found
4437 ;; with `c-decl-prefix-or-start-re'. `cfd-limit' if there's
4438 ;; no match.
d9e94c22 4439 cfd-re-match
0386b551
AM
4440 ;; The end position of the last `c-decl-prefix-or-start-re'
4441 ;; match. If this is greater than `cfd-continue-pos', the
4442 ;; next regexp search is started here instead.
4443 (cfd-re-match-end (point-min))
4444 ;; The end of the last `c-decl-end' found by
4445 ;; `c-find-decl-prefix-search'. `cfd-limit' if there's no
4446 ;; match. If searching for the property isn't needed then we
4447 ;; disable it by setting it to `cfd-limit' directly.
d9e94c22 4448 (cfd-prop-match (unless c-type-decl-end-used cfd-limit))
0386b551
AM
4449 ;; The end of the token preceding the decl spot last found by
4450 ;; `c-find-decl-prefix-search'. 0 for the implicit match at
4451 ;; bob. `cfd-limit' if there's no match. In other words,
4452 ;; this is the minimum of `cfd-re-match' and `cfd-prop-match'.
d9e94c22
MS
4453 (cfd-match-pos cfd-limit)
4454 ;; The position to continue searching at.
4455 cfd-continue-pos
4456 ;; The position of the last "real" token we've stopped at.
4457 ;; This can be greater than `cfd-continue-pos' when we get
4458 ;; hits inside macros or at `c-decl-end' positions inside
4459 ;; comments.
4460 (cfd-token-pos 0)
4461 ;; The end position of the last entered macro.
4462 (cfd-macro-end 0))
4463
4464 ;; Initialize by finding a syntactically relevant start position
0386b551
AM
4465 ;; before the point, and do the first `c-decl-prefix-or-start-re'
4466 ;; search unless we're at bob.
d9e94c22 4467
0386b551 4468 (let (start-in-literal start-in-macro syntactic-pos)
d9e94c22
MS
4469 ;; Must back up a bit since we look for the end of the previous
4470 ;; statement or declaration, which is earlier than the first
4471 ;; returned match.
4472
0386b551
AM
4473 (cond
4474 ;; First we need to move to a syntactically relevant position.
4475 ;; Begin by backing out of comment or string literals.
4476 ((and
4477 (when (c-got-face-at (point) c-literal-faces)
4478 ;; Try to use the faces to back up to the start of the
4479 ;; literal. FIXME: What if the point is on a declaration
4480 ;; inside a comment?
4481 (while (and (not (bobp))
4482 (c-got-face-at (1- (point)) c-literal-faces))
4483 (goto-char (previous-single-property-change
4484 (point) 'face nil (point-min))))
4485
4486 ;; XEmacs doesn't fontify the quotes surrounding string
4487 ;; literals.
4488 (and (featurep 'xemacs)
4489 (eq (get-text-property (point) 'face)
4490 'font-lock-string-face)
4491 (not (bobp))
4492 (progn (backward-char)
4493 (not (looking-at c-string-limit-regexp)))
4494 (forward-char))
4495
4496 ;; Don't trust the literal to contain only literal faces
4497 ;; (the font lock package might not have fontified the
4498 ;; start of it at all, for instance) so check that we have
4499 ;; arrived at something that looks like a start or else
4500 ;; resort to `c-literal-limits'.
4501 (unless (looking-at c-literal-start-regexp)
4502 (let ((range (c-literal-limits)))
4503 (if range (goto-char (car range)))))
4504
4505 (setq start-in-literal (point)))
4506
4507 ;; The start is in a literal. If the limit is in the same
4508 ;; one we don't have to find a syntactic position etc. We
4509 ;; only check that if the limit is at or before bonl to save
4510 ;; time; it covers the by far most common case when font-lock
4511 ;; refontifies the current line only.
4512 (<= cfd-limit (c-point 'bonl cfd-start-pos))
4513 (save-excursion
4514 (goto-char cfd-start-pos)
4515 (while (progn
4516 (goto-char (next-single-property-change
4517 (point) 'face nil cfd-limit))
4518 (and (< (point) cfd-limit)
4519 (c-got-face-at (point) c-literal-faces))))
4520 (= (point) cfd-limit)))
4521
4522 ;; Completely inside a literal. Set up variables to trig the
4523 ;; (< cfd-continue-pos cfd-start-pos) case below and it'll
4524 ;; find a suitable start position.
4525 (setq cfd-continue-pos start-in-literal))
4526
4527 ;; Check if the region might be completely inside a macro, to
4528 ;; optimize that like the completely-inside-literal above.
4529 ((save-excursion
4530 (and (= (forward-line 1) 0)
4531 (bolp) ; forward-line has funny behavior at eob.
4532 (>= (point) cfd-limit)
4533 (progn (backward-char)
4534 (eq (char-before) ?\\))))
4535 ;; (Maybe) completely inside a macro. Only need to trig the
4536 ;; (< cfd-continue-pos cfd-start-pos) case below to make it
4537 ;; set things up.
4538 (setq cfd-continue-pos (1- cfd-start-pos)
4539 start-in-macro t))
d9e94c22 4540
0386b551
AM
4541 (t
4542 ;; Back out of any macro so we don't miss any declaration
4543 ;; that could follow after it.
4544 (when (c-beginning-of-macro)
4545 (setq start-in-macro t))
4546
4547 ;; Now we're at a proper syntactically relevant position so we
4548 ;; can use the cache. But first clear it if it applied
4549 ;; further down.
4550 (c-invalidate-find-decl-cache cfd-start-pos)
4551
4552 (setq syntactic-pos (point))
4553 (unless (eq syntactic-pos c-find-decl-syntactic-pos)
4554 ;; Don't have to do this if the cache is relevant here,
4555 ;; typically if the same line is refontified again. If
4556 ;; we're just some syntactic whitespace further down we can
4557 ;; still use the cache to limit the skipping.
4558 (c-backward-syntactic-ws c-find-decl-syntactic-pos))
4559
4560 ;; If we hit `c-find-decl-syntactic-pos' and
4561 ;; `c-find-decl-match-pos' is set then we install the cached
4562 ;; values. If we hit `c-find-decl-syntactic-pos' and
4563 ;; `c-find-decl-match-pos' is nil then we know there's no decl
4564 ;; prefix in the whitespace before `c-find-decl-syntactic-pos'
4565 ;; and so we can continue the search from this point. If we
4566 ;; didn't hit `c-find-decl-syntactic-pos' then we're now in
4567 ;; the right spot to begin searching anyway.
4568 (if (and (eq (point) c-find-decl-syntactic-pos)
4569 c-find-decl-match-pos)
d9e94c22
MS
4570 (setq cfd-match-pos c-find-decl-match-pos
4571 cfd-continue-pos syntactic-pos)
0386b551
AM
4572
4573 (setq c-find-decl-syntactic-pos syntactic-pos)
4574
4575 (when (if (bobp)
4576 ;; Always consider bob a match to get the first
4577 ;; declaration in the file. Do this separately instead of
4578 ;; letting `c-decl-prefix-or-start-re' match bob, so that
4579 ;; regexp always can consume at least one character to
4580 ;; ensure that we won't get stuck in an infinite loop.
4581 (setq cfd-re-match 0)
4582 (backward-char)
4583 (c-beginning-of-current-token)
4584 (< (point) cfd-limit))
4585 ;; Do an initial search now. In the bob case above it's
4586 ;; only done to search for a `c-decl-end' spot.
4587 (c-find-decl-prefix-search))
4588
4589 (setq c-find-decl-match-pos (and (< cfd-match-pos cfd-start-pos)
4590 cfd-match-pos)))))
4591
4592 ;; Advance `cfd-continue-pos' if it's before the start position.
4593 ;; The closest continue position that might have effect at or
4594 ;; after the start depends on what we started in. This also
4595 ;; finds a suitable start position in the special cases when the
4596 ;; region is completely within a literal or macro.
4597 (when (and cfd-continue-pos (< cfd-continue-pos cfd-start-pos))
4598
4599 (cond
4600 (start-in-macro
4601 ;; If we're in a macro then it's the closest preceding token
4602 ;; in the macro. Check this before `start-in-literal',
4603 ;; since if we're inside a literal in a macro, the preceding
4604 ;; token is earlier than any `c-decl-end' spot inside the
4605 ;; literal (comment).
4606 (goto-char (or start-in-literal cfd-start-pos))
4607 ;; The only syntactic ws in macros are comments.
d9e94c22 4608 (c-backward-comments)
0386b551
AM
4609 (backward-char)
4610 (c-beginning-of-current-token))
4611
4612 (start-in-literal
4613 ;; If we're in a comment it can only be the closest
4614 ;; preceding `c-decl-end' position within that comment, if
4615 ;; any. Go back to the beginning of such a property so that
4616 ;; `c-find-decl-prefix-search' will find the end of it.
4617 ;; (Can't stop at the end and install it directly on
4618 ;; `cfd-prop-match' since that variable might be cleared
4619 ;; after `cfd-fun' below.)
4620 ;;
4621 ;; Note that if the literal is a string then the property
4622 ;; search will simply skip to the beginning of it right
4623 ;; away.
4624 (if (not c-type-decl-end-used)
4625 (goto-char start-in-literal)
4626 (goto-char cfd-start-pos)
4627 (while (progn
4628 (goto-char (previous-single-property-change
4629 (point) 'c-type nil start-in-literal))
4630 (and (> (point) start-in-literal)
4631 (not (eq (c-get-char-property (point) 'c-type)
4632 'c-decl-end))))))
4633
4634 (when (= (point) start-in-literal)
4635 ;; Didn't find any property inside the comment, so we can
4636 ;; skip it entirely. (This won't skip past a string, but
4637 ;; that'll be handled quickly by the next
4638 ;; `c-find-decl-prefix-search' anyway.)
4639 (c-forward-single-comment)
4640 (if (> (point) cfd-limit)
4641 (goto-char cfd-limit))))
d9e94c22 4642
0386b551
AM
4643 (t
4644 ;; If we started in normal code, the only match that might
4645 ;; apply before the start is what we already got in
4646 ;; `cfd-match-pos' so we can continue at the start position.
4647 ;; (Note that we don't get here if the first match is below
4648 ;; it.)
4649 (goto-char cfd-start-pos)))
4650
4651 ;; Delete found matches if they are before our new continue
4652 ;; position, so that `c-find-decl-prefix-search' won't back up
4653 ;; to them later on.
4654 (setq cfd-continue-pos (point))
4655 (when (and cfd-re-match (< cfd-re-match cfd-continue-pos))
4656 (setq cfd-re-match nil))
4657 (when (and cfd-prop-match (< cfd-prop-match cfd-continue-pos))
4658 (setq cfd-prop-match nil)))
4659
4660 (if syntactic-pos
4661 ;; This is the normal case and we got a proper syntactic
4662 ;; position. If there's a match then it's always outside
4663 ;; macros and comments, so advance to the next token and set
4664 ;; `cfd-token-pos'. The loop below will later go back using
4665 ;; `cfd-continue-pos' to fix declarations inside the
4666 ;; syntactic ws.
4667 (when (and cfd-match-pos (< cfd-match-pos syntactic-pos))
4668 (goto-char syntactic-pos)
4669 (c-forward-syntactic-ws)
4670 (and cfd-continue-pos
4671 (< cfd-continue-pos (point))
4672 (setq cfd-token-pos (point))))
4673
4674 ;; Have one of the special cases when the region is completely
4675 ;; within a literal or macro. `cfd-continue-pos' is set to a
4676 ;; good start position for the search, so do it.
4677 (c-find-decl-prefix-search)))
d9e94c22 4678
51c9af45 4679 ;; Now loop. Round what? (ACM, 2006/7/5). We already got the first match.
d9e94c22
MS
4680
4681 (while (progn
4682 (while (and
4683 (< cfd-match-pos cfd-limit)
4684
4685 (or
4686 ;; Kludge to filter out matches on the "<" that
4687 ;; aren't open parens, for the sake of languages
4688 ;; that got `c-recognize-<>-arglists' set.
4689 (and (eq (char-before cfd-match-pos) ?<)
4690 (not (c-get-char-property (1- cfd-match-pos)
4691 'syntax-table)))
4692
4693 ;; If `cfd-continue-pos' is less or equal to
4694 ;; `cfd-token-pos', we've got a hit inside a macro
4695 ;; that's in the syntactic whitespace before the last
4696 ;; "real" declaration we've checked. If they're equal
4697 ;; we've arrived at the declaration a second time, so
4698 ;; there's nothing to do.
4699 (= cfd-continue-pos cfd-token-pos)
4700
4701 (progn
4702 ;; If `cfd-continue-pos' is less than `cfd-token-pos'
4703 ;; we're still searching for declarations embedded in
4704 ;; the syntactic whitespace. In that case we need
4705 ;; only to skip comments and not macros, since they
4706 ;; can't be nested, and that's already been done in
4707 ;; `c-find-decl-prefix-search'.
4708 (when (> cfd-continue-pos cfd-token-pos)
4709 (c-forward-syntactic-ws)
4710 (setq cfd-token-pos (point)))
4711
4712 ;; Continue if the following token fails the
4713 ;; CFD-DECL-RE and CFD-FACE-CHECKLIST checks.
4714 (when (or (>= (point) cfd-limit)
4715 (not (looking-at cfd-decl-re))
4716 (and cfd-face-checklist
4717 (not (c-got-face-at
4718 (point) cfd-face-checklist))))
4719 (goto-char cfd-continue-pos)
4720 t)))
4721
4722 (< (point) cfd-limit))
4723 (c-find-decl-prefix-search))
4724
4725 (< (point) cfd-limit))
4726
0386b551
AM
4727 (when (and
4728 (>= (point) cfd-start-pos)
d9e94c22 4729
0386b551
AM
4730 (progn
4731 ;; Narrow to the end of the macro if we got a hit inside
4732 ;; one, to avoid recognizing things that start inside the
4733 ;; macro and end outside it.
4734 (when (> cfd-match-pos cfd-macro-end)
4735 ;; Not in the same macro as in the previous round.
4736 (save-excursion
4737 (goto-char cfd-match-pos)
4738 (setq cfd-macro-end
4739 (if (save-excursion (and (c-beginning-of-macro)
4740 (< (point) cfd-match-pos)))
4741 (progn (c-end-of-macro)
4742 (point))
4743 0))))
4744
4745 (if (zerop cfd-macro-end)
4746 t
4747 (if (> cfd-macro-end (point))
4748 (progn (narrow-to-region (point-min) cfd-macro-end)
4749 t)
4750 ;; The matched token was the last thing in the macro,
4751 ;; so the whole match is bogus.
4752 (setq cfd-macro-end 0)
4753 nil))))
d9e94c22
MS
4754
4755 (c-debug-put-decl-spot-faces cfd-match-pos (point))
0386b551
AM
4756 (if (funcall cfd-fun cfd-match-pos (/= cfd-macro-end 0))
4757 (setq cfd-prop-match nil))
d9e94c22
MS
4758
4759 (when (/= cfd-macro-end 0)
4760 ;; Restore limits if we did macro narrowment above.
4761 (narrow-to-region (point-min) cfd-buffer-end)))
4762
4763 (goto-char cfd-continue-pos)
4764 (if (= cfd-continue-pos cfd-limit)
4765 (setq cfd-match-pos cfd-limit)
4766 (c-find-decl-prefix-search)))))
4767
4768\f
4769;; A cache for found types.
4770
4771;; Buffer local variable that contains an obarray with the types we've
4772;; found. If a declaration is recognized somewhere we record the
4773;; fully qualified identifier in it to recognize it as a type
4774;; elsewhere in the file too. This is not accurate since we do not
4775;; bother with the scoping rules of the languages, but in practice the
4776;; same name is seldom used as both a type and something else in a
4777;; file, and we only use this as a last resort in ambiguous cases (see
0386b551
AM
4778;; `c-forward-decl-or-cast-1').
4779;;
580fba94
AM
4780;; Not every type need be in this cache. However, things which have
4781;; ceased to be types must be removed from it.
4782;;
0386b551
AM
4783;; Template types in C++ are added here too but with the template
4784;; arglist replaced with "<>" in references or "<" for the one in the
4785;; primary type. E.g. the type "Foo<A,B>::Bar<C>" is stored as
4786;; "Foo<>::Bar<". This avoids storing very long strings (since C++
4787;; template specs can be fairly sized programs in themselves) and
4788;; improves the hit ratio (it's a type regardless of the template
4789;; args; it's just not the same type, but we're only interested in
4790;; recognizing types, not telling distinct types apart). Note that
4791;; template types in references are added here too; from the example
4792;; above there will also be an entry "Foo<".
d9e94c22
MS
4793(defvar c-found-types nil)
4794(make-variable-buffer-local 'c-found-types)
4795
4796(defsubst c-clear-found-types ()
4797 ;; Clears `c-found-types'.
d9e94c22
MS
4798 (setq c-found-types (make-vector 53 0)))
4799
4800(defun c-add-type (from to)
4801 ;; Add the given region as a type in `c-found-types'. If the region
4802 ;; doesn't match an existing type but there is a type which is equal
4803 ;; to the given one except that the last character is missing, then
4804 ;; the shorter type is removed. That's done to avoid adding all
4805 ;; prefixes of a type as it's being entered and font locked. This
4806 ;; doesn't cover cases like when characters are removed from a type
4807 ;; or added in the middle. We'd need the position of point when the
4808 ;; font locking is invoked to solve this well.
0386b551
AM
4809 ;;
4810 ;; This function might do hidden buffer changes.
4811 (let ((type (c-syntactic-content from to c-recognize-<>-arglists)))
4812 (unless (intern-soft type c-found-types)
4813 (unintern (substring type 0 -1) c-found-types)
4814 (intern type c-found-types))))
d9e94c22 4815
580fba94
AM
4816(defun c-unfind-type (name)
4817 ;; Remove the "NAME" from c-found-types, if present.
4818 (unintern name c-found-types))
4819
d9e94c22
MS
4820(defsubst c-check-type (from to)
4821 ;; Return non-nil if the given region contains a type in
4822 ;; `c-found-types'.
0386b551
AM
4823 ;;
4824 ;; This function might do hidden buffer changes.
4825 (intern-soft (c-syntactic-content from to c-recognize-<>-arglists)
4826 c-found-types))
d9e94c22
MS
4827
4828(defun c-list-found-types ()
4829 ;; Return all the types in `c-found-types' as a sorted list of
4830 ;; strings.
4831 (let (type-list)
4832 (mapatoms (lambda (type)
4833 (setq type-list (cons (symbol-name type)
4834 type-list)))
4835 c-found-types)
4836 (sort type-list 'string-lessp)))
a66cd3ee 4837
2f42c75f
DN
4838;; Shut up the byte compiler.
4839(defvar c-maybe-stale-found-type)
4840
580fba94
AM
4841(defun c-trim-found-types (beg end old-len)
4842 ;; An after change function which, in conjunction with the info in
4843 ;; c-maybe-stale-found-type (set in c-before-change), removes a type
4844 ;; from `c-found-types', should this type have become stale. For
4845 ;; example, this happens to "foo" when "foo \n bar();" becomes
4846 ;; "foo(); \n bar();". Such stale types, if not removed, foul up
4847 ;; the fontification.
b414f371 4848 ;;
580fba94
AM
4849 ;; Have we, perhaps, added non-ws characters to the front/back of a found
4850 ;; type?
4851 (when (> end beg)
4852 (save-excursion
4853 (when (< end (point-max))
4854 (goto-char end)
4855 (if (and (c-beginning-of-current-token) ; only moves when we started in the middle
4856 (progn (goto-char end)
4857 (c-end-of-current-token)))
4858 (c-unfind-type (buffer-substring-no-properties
4859 end (point)))))
4860 (when (> beg (point-min))
4861 (goto-char beg)
4862 (if (and (c-end-of-current-token) ; only moves when we started in the middle
4863 (progn (goto-char beg)
4864 (c-beginning-of-current-token)))
4865 (c-unfind-type (buffer-substring-no-properties
4866 (point) beg))))))
b414f371 4867
580fba94
AM
4868 (if c-maybe-stale-found-type ; e.g. (c-decl-id-start "foo" 97 107 " (* ooka) " "o")
4869 (cond
4870 ;; Changing the amount of (already existing) whitespace - don't do anything.
4871 ((and (c-partial-ws-p beg end)
4872 (or (= beg end) ; removal of WS
580fba94
AM
4873 (string-match "^[ \t\n\r\f\v]*$" (nth 5 c-maybe-stale-found-type)))))
4874
4875 ;; The syntactic relationship which defined a "found type" has been
4876 ;; destroyed.
4877 ((eq (car c-maybe-stale-found-type) 'c-decl-id-start)
4878 (c-unfind-type (cadr c-maybe-stale-found-type)))
4879;; ((eq (car c-maybe-stale-found-type) 'c-decl-type-start) FIXME!!!
4880 )))
4881
d9e94c22
MS
4882\f
4883;; Handling of small scale constructs like types and names.
4884
0386b551
AM
4885(defun c-after-change-check-<>-operators (beg end)
4886 ;; This is called from `after-change-functions' when
4887 ;; c-recognize-<>-arglists' is set. It ensures that no "<" or ">"
4888 ;; chars with paren syntax become part of another operator like "<<"
4889 ;; or ">=".
4890 ;;
4891 ;; This function might do hidden buffer changes.
4892
cb694ab7
AM
4893 (save-excursion
4894 (goto-char beg)
4895 (when (or (looking-at "[<>]")
4896 (< (skip-chars-backward "<>") 0))
4897
0386b551 4898 (goto-char beg)
cb694ab7
AM
4899 (c-beginning-of-current-token)
4900 (when (and (< (point) beg)
4901 (looking-at c-<>-multichar-token-regexp)
4902 (< beg (setq beg (match-end 0))))
4903 (while (progn (skip-chars-forward "^<>" beg)
4904 (< (point) beg))
4905 (c-clear-char-property (point) 'syntax-table)
4906 (forward-char))))
4907
4908 (when (< beg end)
4909 (goto-char end)
0386b551
AM
4910 (when (or (looking-at "[<>]")
4911 (< (skip-chars-backward "<>") 0))
4912
cb694ab7 4913 (goto-char end)
0386b551 4914 (c-beginning-of-current-token)
cb694ab7 4915 (when (and (< (point) end)
0386b551 4916 (looking-at c-<>-multichar-token-regexp)
cb694ab7
AM
4917 (< end (setq end (match-end 0))))
4918 (while (progn (skip-chars-forward "^<>" end)
4919 (< (point) end))
0386b551 4920 (c-clear-char-property (point) 'syntax-table)
cb694ab7 4921 (forward-char)))))))
d9e94c22
MS
4922
4923;; Dynamically bound variable that instructs `c-forward-type' to also
4924;; treat possible types (i.e. those that it normally returns 'maybe or
4925;; 'found for) as actual types (and always return 'found for them).
4926;; This means that it records them in `c-record-type-identifiers' if
4927;; that is set, and that it adds them to `c-found-types'.
4928(defvar c-promote-possible-types nil)
4929
0386b551
AM
4930;; Dynamically bound variable that instructs `c-forward-<>-arglist' to
4931;; mark up successfully parsed arglists with paren syntax properties on
4932;; the surrounding angle brackets and with `c-<>-arg-sep' in the
4933;; `c-type' property of each argument separating comma.
4934;;
4935;; Setting this variable also makes `c-forward-<>-arglist' recurse into
4936;; all arglists for side effects (i.e. recording types), otherwise it
4937;; exploits any existing paren syntax properties to quickly jump to the
4938;; end of already parsed arglists.
4939;;
4940;; Marking up the arglists is not the default since doing that correctly
4941;; depends on a proper value for `c-restricted-<>-arglists'.
4942(defvar c-parse-and-markup-<>-arglists nil)
4943
d9e94c22 4944;; Dynamically bound variable that instructs `c-forward-<>-arglist' to
037558bf
MS
4945;; not accept arglists that contain binary operators.
4946;;
4947;; This is primarily used to handle C++ template arglists. C++
4948;; disambiguates them by checking whether the preceding name is a
4949;; template or not. We can't do that, so we assume it is a template
4950;; if it can be parsed as one. That usually works well since
4951;; comparison expressions on the forms "a < b > c" or "a < b, c > d"
4952;; in almost all cases would be pointless.
4953;;
4954;; However, in function arglists, e.g. in "foo (a < b, c > d)", we
4955;; should let the comma separate the function arguments instead. And
4956;; in a context where the value of the expression is taken, e.g. in
4957;; "if (a < b || c > d)", it's probably not a template.
4958(defvar c-restricted-<>-arglists nil)
d9e94c22 4959
0386b551
AM
4960;; Dynamically bound variables that instructs
4961;; `c-forward-keyword-clause', `c-forward-<>-arglist',
4962;; `c-forward-name', `c-forward-type', `c-forward-decl-or-cast-1', and
4963;; `c-forward-label' to record the ranges of all the type and
4964;; reference identifiers they encounter. They will build lists on
4965;; these variables where each element is a cons of the buffer
4966;; positions surrounding each identifier. This recording is only
4967;; activated when `c-record-type-identifiers' is non-nil.
d9e94c22
MS
4968;;
4969;; All known types that can't be identifiers are recorded, and also
4970;; other possible types if `c-promote-possible-types' is set.
4971;; Recording is however disabled inside angle bracket arglists that
4972;; are encountered inside names and other angle bracket arglists.
0386b551 4973;; Such occurrences are taken care of by `c-font-lock-<>-arglists'
d9e94c22
MS
4974;; instead.
4975;;
4976;; Only the names in C++ template style references (e.g. "tmpl" in
4977;; "tmpl<a,b>::foo") are recorded as references, other references
4978;; aren't handled here.
0386b551
AM
4979;;
4980;; `c-forward-label' records the label identifier(s) on
4981;; `c-record-ref-identifiers'.
d9e94c22
MS
4982(defvar c-record-type-identifiers nil)
4983(defvar c-record-ref-identifiers nil)
4984
0386b551
AM
4985;; This variable will receive a cons cell of the range of the last
4986;; single identifier symbol stepped over by `c-forward-name' if it's
4987;; successful. This is the range that should be put on one of the
4988;; record lists above by the caller. It's assigned nil if there's no
4989;; such symbol in the name.
d9e94c22
MS
4990(defvar c-last-identifier-range nil)
4991
4992(defmacro c-record-type-id (range)
4993 (if (eq (car-safe range) 'cons)
4994 ;; Always true.
4995 `(setq c-record-type-identifiers
4996 (cons ,range c-record-type-identifiers))
4997 `(let ((range ,range))
4998 (if range
4999 (setq c-record-type-identifiers
5000 (cons range c-record-type-identifiers))))))
5001
5002(defmacro c-record-ref-id (range)
5003 (if (eq (car-safe range) 'cons)
5004 ;; Always true.
5005 `(setq c-record-ref-identifiers
5006 (cons ,range c-record-ref-identifiers))
5007 `(let ((range ,range))
5008 (if range
5009 (setq c-record-ref-identifiers
5010 (cons range c-record-ref-identifiers))))))
5011
5012;; Dynamically bound variable that instructs `c-forward-type' to
5013;; record the ranges of types that only are found. Behaves otherwise
5014;; like `c-record-type-identifiers'.
5015(defvar c-record-found-types nil)
5016
5017(defmacro c-forward-keyword-prefixed-id (type)
5018 ;; Used internally in `c-forward-keyword-clause' to move forward
5019 ;; over a type (if TYPE is 'type) or a name (otherwise) which
5020 ;; possibly is prefixed by keywords and their associated clauses.
5021 ;; Try with a type/name first to not trip up on those that begin
5022 ;; with a keyword. Return t if a known or found type is moved
5023 ;; over. The point is clobbered if nil is returned. If range
5024 ;; recording is enabled, the identifier is recorded on as a type
5025 ;; if TYPE is 'type or as a reference if TYPE is 'ref.
0386b551
AM
5026 ;;
5027 ;; This macro might do hidden buffer changes.
d9e94c22
MS
5028 `(let (res)
5029 (while (if (setq res ,(if (eq type 'type)
5030 `(c-forward-type)
5031 `(c-forward-name)))
5032 nil
5033 (and (looking-at c-keywords-regexp)
0386b551 5034 (c-forward-keyword-clause 1))))
d9e94c22
MS
5035 (when (memq res '(t known found prefix))
5036 ,(when (eq type 'ref)
5037 `(when c-record-type-identifiers
5038 (c-record-ref-id c-last-identifier-range)))
5039 t)))
5040
0386b551 5041(defmacro c-forward-id-comma-list (type update-safe-pos)
d9e94c22
MS
5042 ;; Used internally in `c-forward-keyword-clause' to move forward
5043 ;; over a comma separated list of types or names using
5044 ;; `c-forward-keyword-prefixed-id'.
0386b551
AM
5045 ;;
5046 ;; This macro might do hidden buffer changes.
d9e94c22 5047 `(while (and (progn
0386b551
AM
5048 ,(when update-safe-pos
5049 `(setq safe-pos (point)))
d9e94c22
MS
5050 (eq (char-after) ?,))
5051 (progn
5052 (forward-char)
5053 (c-forward-syntactic-ws)
5054 (c-forward-keyword-prefixed-id ,type)))))
5055
0386b551
AM
5056(defun c-forward-keyword-clause (match)
5057 ;; Submatch MATCH in the current match data is assumed to surround a
5058 ;; token. If it's a keyword, move over it and any immediately
5059 ;; following clauses associated with it, stopping at the start of
5060 ;; the next token. t is returned in that case, otherwise the point
d9e94c22
MS
5061 ;; stays and nil is returned. The kind of clauses that are
5062 ;; recognized are those specified by `c-type-list-kwds',
5063 ;; `c-ref-list-kwds', `c-colon-type-list-kwds',
5064 ;; `c-paren-nontype-kwds', `c-paren-type-kwds', `c-<>-type-kwds',
5065 ;; and `c-<>-arglist-kwds'.
0386b551
AM
5066 ;;
5067 ;; This function records identifier ranges on
5068 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
5069 ;; `c-record-type-identifiers' is non-nil.
5070 ;;
5071 ;; Note that for `c-colon-type-list-kwds', which doesn't necessary
5072 ;; apply directly after the keyword, the type list is moved over
5073 ;; only when there is no unaccounted token before it (i.e. a token
5074 ;; that isn't moved over due to some other keyword list). The
5075 ;; identifier ranges in the list are still recorded if that should
5076 ;; be done, though.
5077 ;;
5078 ;; This function might do hidden buffer changes.
5079
5080 (let ((kwd-sym (c-keyword-sym (match-string match))) safe-pos pos
5081 ;; The call to `c-forward-<>-arglist' below is made after
5082 ;; `c-<>-sexp-kwds' keywords, so we're certain they actually
5083 ;; are angle bracket arglists and `c-restricted-<>-arglists'
5084 ;; should therefore be nil.
5085 (c-parse-and-markup-<>-arglists t)
5086 c-restricted-<>-arglists)
d9e94c22 5087
d9e94c22 5088 (when kwd-sym
0386b551 5089 (goto-char (match-end match))
d9e94c22
MS
5090 (c-forward-syntactic-ws)
5091 (setq safe-pos (point))
a66cd3ee 5092
d9e94c22
MS
5093 (cond
5094 ((and (c-keyword-member kwd-sym 'c-type-list-kwds)
5095 (c-forward-keyword-prefixed-id type))
5096 ;; There's a type directly after a keyword in `c-type-list-kwds'.
0386b551 5097 (c-forward-id-comma-list type t))
d9e94c22
MS
5098
5099 ((and (c-keyword-member kwd-sym 'c-ref-list-kwds)
5100 (c-forward-keyword-prefixed-id ref))
5101 ;; There's a name directly after a keyword in `c-ref-list-kwds'.
0386b551 5102 (c-forward-id-comma-list ref t))
d9e94c22
MS
5103
5104 ((and (c-keyword-member kwd-sym 'c-paren-any-kwds)
5105 (eq (char-after) ?\())
5106 ;; There's an open paren after a keyword in `c-paren-any-kwds'.
5107
5108 (forward-char)
5109 (when (and (setq pos (c-up-list-forward))
5110 (eq (char-before pos) ?\)))
5111 (when (and c-record-type-identifiers
5112 (c-keyword-member kwd-sym 'c-paren-type-kwds))
5113 ;; Use `c-forward-type' on every identifier we can find
5114 ;; inside the paren, to record the types.
5115 (while (c-syntactic-re-search-forward c-symbol-start pos t)
5116 (goto-char (match-beginning 0))
5117 (unless (c-forward-type)
5118 (looking-at c-symbol-key) ; Always matches.
5119 (goto-char (match-end 0)))))
5120
5121 (goto-char pos)
5122 (c-forward-syntactic-ws)
5123 (setq safe-pos (point))))
5124
5125 ((and (c-keyword-member kwd-sym 'c-<>-sexp-kwds)
5126 (eq (char-after) ?<)
0386b551 5127 (c-forward-<>-arglist (c-keyword-member kwd-sym 'c-<>-type-kwds)))
d9e94c22
MS
5128 (c-forward-syntactic-ws)
5129 (setq safe-pos (point)))
5130
5131 ((and (c-keyword-member kwd-sym 'c-nonsymbol-sexp-kwds)
449a2b0d
MS
5132 (not (looking-at c-symbol-start))
5133 (c-safe (c-forward-sexp) t))
d9e94c22
MS
5134 (c-forward-syntactic-ws)
5135 (setq safe-pos (point))))
5136
0386b551
AM
5137 (when (c-keyword-member kwd-sym 'c-colon-type-list-kwds)
5138 (if (eq (char-after) ?:)
5139 ;; If we are at the colon already, we move over the type
5140 ;; list after it.
5141 (progn
5142 (forward-char)
5143 (c-forward-syntactic-ws)
5144 (when (c-forward-keyword-prefixed-id type)
5145 (c-forward-id-comma-list type t)))
5146 ;; Not at the colon, so stop here. But the identifier
5147 ;; ranges in the type list later on should still be
5148 ;; recorded.
5149 (and c-record-type-identifiers
5150 (progn
5151 ;; If a keyword matched both one of the types above and
5152 ;; this one, we match `c-colon-type-list-re' after the
5153 ;; clause matched above.
5154 (goto-char safe-pos)
5155 (looking-at c-colon-type-list-re))
5156 (progn
5157 (goto-char (match-end 0))
5158 (c-forward-syntactic-ws)
5159 (c-forward-keyword-prefixed-id type))
5160 ;; There's a type after the `c-colon-type-list-re' match
5161 ;; after a keyword in `c-colon-type-list-kwds'.
5162 (c-forward-id-comma-list type nil))))
d9e94c22
MS
5163
5164 (goto-char safe-pos)
5165 t)))
5166
0386b551
AM
5167(defun c-forward-<>-arglist (all-types)
5168 ;; The point is assumed to be at a "<". Try to treat it as the open
b4dc7d98 5169 ;; paren of an angle bracket arglist and move forward to the
0386b551
AM
5170 ;; corresponding ">". If successful, the point is left after the
5171 ;; ">" and t is returned, otherwise the point isn't moved and nil is
d9e94c22
MS
5172 ;; returned. If ALL-TYPES is t then all encountered arguments in
5173 ;; the arglist that might be types are treated as found types.
5174 ;;
0386b551
AM
5175 ;; The variable `c-parse-and-markup-<>-arglists' controls how this
5176 ;; function handles text properties on the angle brackets and argument
5177 ;; separating commas.
d9e94c22 5178 ;;
0386b551
AM
5179 ;; `c-restricted-<>-arglists' controls how lenient the template
5180 ;; arglist recognition should be.
5181 ;;
5182 ;; This function records identifier ranges on
5183 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
5184 ;; `c-record-type-identifiers' is non-nil.
5185 ;;
5186 ;; This function might do hidden buffer changes.
d9e94c22
MS
5187
5188 (let ((start (point))
5189 ;; If `c-record-type-identifiers' is set then activate
5190 ;; recording of any found types that constitute an argument in
5191 ;; the arglist.
5192 (c-record-found-types (if c-record-type-identifiers t)))
5193 (if (catch 'angle-bracket-arglist-escape
5194 (setq c-record-found-types
0386b551 5195 (c-forward-<>-arglist-recur all-types)))
d9e94c22
MS
5196 (progn
5197 (when (consp c-record-found-types)
5198 (setq c-record-type-identifiers
5199 ;; `nconc' doesn't mind that the tail of
5200 ;; `c-record-found-types' is t.
5201 (nconc c-record-found-types c-record-type-identifiers)))
5202 t)
5203
5204 (goto-char start)
a66cd3ee 5205 nil)))
785eecbb 5206
0386b551 5207(defun c-forward-<>-arglist-recur (all-types)
d9e94c22 5208 ;; Recursive part of `c-forward-<>-arglist'.
0386b551
AM
5209 ;;
5210 ;; This function might do hidden buffer changes.
d9e94c22
MS
5211
5212 (let ((start (point)) res pos tmp
5213 ;; Cover this so that any recorded found type ranges are
5214 ;; automatically lost if it turns out to not be an angle
5215 ;; bracket arglist. It's propagated through the return value
5216 ;; on successful completion.
5217 (c-record-found-types c-record-found-types)
5218 ;; List that collects the positions after the argument
5219 ;; separating ',' in the arglist.
5220 arg-start-pos)
5221
0386b551
AM
5222 ;; If the '<' has paren open syntax then we've marked it as an angle
5223 ;; bracket arglist before, so skip to the end.
5224 (if (and (not c-parse-and-markup-<>-arglists)
5225 (c-get-char-property (point) 'syntax-table))
5226
5227 (progn
5228 (forward-char)
5229 (if (and (c-go-up-list-forward)
5230 (eq (char-before) ?>))
5231 t
5232
5233 ;; Got unmatched paren angle brackets. We don't clear the paren
5234 ;; syntax properties and retry, on the basis that it's very
5235 ;; unlikely that paren angle brackets become operators by code
5236 ;; manipulation. It's far more likely that it doesn't match due
5237 ;; to narrowing or some temporary change.
5238 (goto-char start)
5239 nil))
d9e94c22
MS
5240
5241 (forward-char)
5242 (unless (looking-at c-<-op-cont-regexp)
5243 (while (and
5244 (progn
5245
5246 (when c-record-type-identifiers
5247 (if all-types
5248
5249 ;; All encountered identifiers are types, so set the
5250 ;; promote flag and parse the type.
5251 (progn
5252 (c-forward-syntactic-ws)
5253 (when (looking-at c-identifier-start)
5254 (let ((c-promote-possible-types t))
5255 (c-forward-type))))
5256
5257 ;; Check if this arglist argument is a sole type. If
5258 ;; it's known then it's recorded in
5259 ;; `c-record-type-identifiers'. If it only is found
5260 ;; then it's recorded in `c-record-found-types' which we
5261 ;; might roll back if it turns out that this isn't an
5262 ;; angle bracket arglist afterall.
5263 (when (memq (char-before) '(?, ?<))
5264 (let ((orig-record-found-types c-record-found-types))
5265 (c-forward-syntactic-ws)
5266 (and (memq (c-forward-type) '(known found))
5267 (not (looking-at "[,>]"))
5268 ;; A found type was recorded but it's not the
5269 ;; only thing in the arglist argument, so reset
5270 ;; `c-record-found-types'.
5271 (setq c-record-found-types
5272 orig-record-found-types))))))
5273
5274 (setq pos (point))
5275 (or (when (eq (char-after) ?>)
5276 ;; Must check for '>' at the very start separately,
5277 ;; since the regexp below has to avoid ">>" without
5278 ;; using \\=.
5279 (forward-char)
5280 t)
5281
037558bf
MS
5282 ;; Note: These regexps exploit the match order in \| so
5283 ;; that "<>" is matched by "<" rather than "[^>:-]>".
d9e94c22 5284 (c-syntactic-re-search-forward
037558bf
MS
5285 (if c-restricted-<>-arglists
5286 ;; Stop on ',', '|', '&', '+' and '-' to catch
5287 ;; common binary operators that could be between
5288 ;; two comparison expressions "a<b" and "c>d".
5289 "[<;{},|&+-]\\|\\([^>:-]>\\)"
5290 ;; Otherwise we still stop on ',' to find the
5291 ;; argument start positions.
5292 "[<;{},]\\|\\([^>:-]>\\)")
5293 nil 'move t t 1)
d9e94c22
MS
5294
5295 ;; If the arglist starter has lost its open paren
5296 ;; syntax but not the closer, we won't find the
5297 ;; closer above since we only search in the
5298 ;; balanced sexp. In that case we stop just short
5299 ;; of it so check if the following char is the closer.
5300 (when (eq (char-after) ?>)
d9e94c22
MS
5301 (forward-char)
5302 t)))
5303
5304 (cond
5305 ((eq (char-before) ?>)
5306 ;; Either an operator starting with '>' or the end of
5307 ;; the angle bracket arglist.
5308
0386b551 5309 (if (looking-at c->-op-cont-regexp)
d9e94c22 5310 (progn
0386b551
AM
5311 (goto-char (match-end 0))
5312 t) ; Continue the loop.
d9e94c22 5313
0386b551
AM
5314 ;; The angle bracket arglist is finished.
5315 (when c-parse-and-markup-<>-arglists
d9e94c22 5316 (while arg-start-pos
0386b551
AM
5317 (c-put-c-type-property (1- (car arg-start-pos))
5318 'c-<>-arg-sep)
d9e94c22
MS
5319 (setq arg-start-pos (cdr arg-start-pos)))
5320 (c-mark-<-as-paren start)
0386b551
AM
5321 (c-mark->-as-paren (1- (point))))
5322 (setq res t)
5323 nil)) ; Exit the loop.
d9e94c22
MS
5324
5325 ((eq (char-before) ?<)
5326 ;; Either an operator starting with '<' or a nested arglist.
5327
5328 (setq pos (point))
5329 (let (id-start id-end subres keyword-match)
5330 (if (if (looking-at c-<-op-cont-regexp)
5331 (setq tmp (match-end 0))
5332 (setq tmp pos)
5333 (backward-char)
5334 (not
5335 (and
5336
5337 (save-excursion
0386b551 5338 ;; There's always an identifier before an angle
d9e94c22
MS
5339 ;; bracket arglist, or a keyword in
5340 ;; `c-<>-type-kwds' or `c-<>-arglist-kwds'.
5341 (c-backward-syntactic-ws)
5342 (setq id-end (point))
5343 (c-simple-skip-symbol-backward)
5344 (when (or (setq keyword-match
5345 (looking-at c-opt-<>-sexp-key))
5346 (not (looking-at c-keywords-regexp)))
5347 (setq id-start (point))))
5348
5349 (setq subres
5350 (let ((c-record-type-identifiers nil)
5351 (c-record-found-types nil))
5352 (c-forward-<>-arglist-recur
5353 (and keyword-match
5354 (c-keyword-member
5355 (c-keyword-sym (match-string 1))
0386b551 5356 'c-<>-type-kwds)))))
d9e94c22
MS
5357 )))
5358
5359 ;; It was not an angle bracket arglist.
0386b551 5360 (goto-char tmp)
d9e94c22
MS
5361
5362 ;; It was an angle bracket arglist.
5363 (setq c-record-found-types subres)
5364
5365 ;; Record the identifier before the template as a type
5366 ;; or reference depending on whether the arglist is last
5367 ;; in a qualified identifier.
5368 (when (and c-record-type-identifiers
5369 (not keyword-match))
5370 (if (and c-opt-identifier-concat-key
5371 (progn
5372 (c-forward-syntactic-ws)
5373 (looking-at c-opt-identifier-concat-key)))
5374 (c-record-ref-id (cons id-start id-end))
5375 (c-record-type-id (cons id-start id-end))))))
5376 t)
5377
5378 ((and (eq (char-before) ?,)
037558bf 5379 (not c-restricted-<>-arglists))
d9e94c22
MS
5380 ;; Just another argument. Record the position. The
5381 ;; type check stuff that made us stop at it is at
5382 ;; the top of the loop.
5383 (setq arg-start-pos (cons (point) arg-start-pos)))
5384
5385 (t
5386 ;; Got a character that can't be in an angle bracket
5387 ;; arglist argument. Abort using `throw', since
5388 ;; it's useless to try to find a surrounding arglist
5389 ;; if we're nested.
5390 (throw 'angle-bracket-arglist-escape nil))))))
5391
5392 (if res
5393 (or c-record-found-types t)))))
5394
0386b551
AM
5395(defun c-backward-<>-arglist (all-types &optional limit)
5396 ;; The point is assumed to be directly after a ">". Try to treat it
5397 ;; as the close paren of an angle bracket arglist and move back to
5398 ;; the corresponding "<". If successful, the point is left at
5399 ;; the "<" and t is returned, otherwise the point isn't moved and
5400 ;; nil is returned. ALL-TYPES is passed on to
5401 ;; `c-forward-<>-arglist'.
5402 ;;
5403 ;; If the optional LIMIT is given, it bounds the backward search.
5404 ;; It's then assumed to be at a syntactically relevant position.
5405 ;;
5406 ;; This is a wrapper around `c-forward-<>-arglist'. See that
5407 ;; function for more details.
5408
5409 (let ((start (point)))
5410 (backward-char)
5411 (if (and (not c-parse-and-markup-<>-arglists)
5412 (c-get-char-property (point) 'syntax-table))
5413
5414 (if (and (c-go-up-list-backward)
5415 (eq (char-after) ?<))
5416 t
5417 ;; See corresponding note in `c-forward-<>-arglist'.
5418 (goto-char start)
5419 nil)
5420
51c9af45 5421 (while (progn
0386b551
AM
5422 (c-syntactic-skip-backward "^<;{}" limit t)
5423
51c9af45
AM
5424 (and
5425 (if (eq (char-before) ?<)
5426 t
5427 ;; Stopped at bob or a char that isn't allowed in an
5428 ;; arglist, so we've failed.
5429 (goto-char start)
5430 nil)
0386b551 5431
51c9af45
AM
5432 (if (> (point)
5433 (progn (c-beginning-of-current-token)
5434 (point)))
5435 ;; If we moved then the "<" was part of some
5436 ;; multicharacter token.
5437 t
0386b551 5438
51c9af45
AM
5439 (backward-char)
5440 (let ((beg-pos (point)))
5441 (if (c-forward-<>-arglist all-types)
5442 (cond ((= (point) start)
5443 ;; Matched the arglist. Break the while.
5444 (goto-char beg-pos)
5445 nil)
5446 ((> (point) start)
5447 ;; We started from a non-paren ">" inside an
5448 ;; arglist.
5449 (goto-char start)
5450 nil)
5451 (t
5452 ;; Matched a shorter arglist. Can be a nested
5453 ;; one so continue looking.
5454 (goto-char beg-pos)
5455 t))
5456 t))))))
0386b551
AM
5457
5458 (/= (point) start))))
5459
d9e94c22
MS
5460(defun c-forward-name ()
5461 ;; Move forward over a complete name if at the beginning of one,
5462 ;; stopping at the next following token. If the point is not at
5463 ;; something that are recognized as name then it stays put. A name
5464 ;; could be something as simple as "foo" in C or something as
5465 ;; complex as "X<Y<class A<int>::B, BIT_MAX >> b>, ::operator<> ::
5466 ;; Z<(a>b)> :: operator const X<&foo>::T Q::G<unsigned short
5467 ;; int>::*volatile const" in C++ (this function is actually little
5468 ;; more than a `looking-at' call in all modes except those that,
5469 ;; like C++, have `c-recognize-<>-arglists' set). Return nil if no
5470 ;; name is found, 'template if it's an identifier ending with an
5471 ;; angle bracket arglist, 'operator of it's an operator identifier,
5472 ;; or t if it's some other kind of name.
0386b551
AM
5473 ;;
5474 ;; This function records identifier ranges on
5475 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
5476 ;; `c-record-type-identifiers' is non-nil.
5477 ;;
5478 ;; This function might do hidden buffer changes.
d9e94c22 5479
0386b551 5480 (let ((pos (point)) (start (point)) res id-start id-end
d9e94c22
MS
5481 ;; Turn off `c-promote-possible-types' here since we might
5482 ;; call `c-forward-<>-arglist' and we don't want it to promote
5483 ;; every suspect thing in the arglist to a type. We're
5484 ;; typically called from `c-forward-type' in this case, and
5485 ;; the caller only wants the top level type that it finds to
5486 ;; be promoted.
5487 c-promote-possible-types)
5488 (while
5489 (and
5490 (looking-at c-identifier-key)
5491
5492 (progn
5493 ;; Check for keyword. We go to the last symbol in
5494 ;; `c-identifier-key' first.
0386b551
AM
5495 (goto-char (setq id-end (match-end 0)))
5496 (c-simple-skip-symbol-backward)
5497 (setq id-start (point))
d9e94c22
MS
5498
5499 (if (looking-at c-keywords-regexp)
5500 (when (and (c-major-mode-is 'c++-mode)
5501 (looking-at
5502 (cc-eval-when-compile
5503 (concat "\\(operator\\|\\(template\\)\\)"
5504 "\\(" (c-lang-const c-nonsymbol-key c++)
5505 "\\|$\\)")))
5506 (if (match-beginning 2)
5507 ;; "template" is only valid inside an
5508 ;; identifier if preceded by "::".
5509 (save-excursion
5510 (c-backward-syntactic-ws)
5511 (and (c-safe (backward-char 2) t)
5512 (looking-at "::")))
5513 t))
5514
5515 ;; Handle a C++ operator or template identifier.
5516 (goto-char id-end)
5517 (c-forward-syntactic-ws)
5518 (cond ((eq (char-before id-end) ?e)
5519 ;; Got "... ::template".
5520 (let ((subres (c-forward-name)))
5521 (when subres
5522 (setq pos (point)
5523 res subres))))
5524
5525 ((looking-at c-identifier-start)
5526 ;; Got a cast operator.
5527 (when (c-forward-type)
5528 (setq pos (point)
5529 res 'operator)
5530 ;; Now we should match a sequence of either
5531 ;; '*', '&' or a name followed by ":: *",
5532 ;; where each can be followed by a sequence
5533 ;; of `c-opt-type-modifier-key'.
5534 (while (cond ((looking-at "[*&]")
5535 (goto-char (match-end 0))
5536 t)
5537 ((looking-at c-identifier-start)
5538 (and (c-forward-name)
5539 (looking-at "::")
5540 (progn
5541 (goto-char (match-end 0))
5542 (c-forward-syntactic-ws)
5543 (eq (char-after) ?*))
5544 (progn
5545 (forward-char)
5546 t))))
5547 (while (progn
5548 (c-forward-syntactic-ws)
5549 (setq pos (point))
5550 (looking-at c-opt-type-modifier-key))
5551 (goto-char (match-end 1))))))
5552
5553 ((looking-at c-overloadable-operators-regexp)
5554 ;; Got some other operator.
0386b551
AM
5555 (setq c-last-identifier-range
5556 (cons (point) (match-end 0)))
d9e94c22
MS
5557 (goto-char (match-end 0))
5558 (c-forward-syntactic-ws)
5559 (setq pos (point)
5560 res 'operator)))
5561
5562 nil)
5563
0386b551
AM
5564 ;; `id-start' is equal to `id-end' if we've jumped over
5565 ;; an identifier that doesn't end with a symbol token.
5566 ;; That can occur e.g. for Java import directives on the
5567 ;; form "foo.bar.*".
5568 (when (and id-start (/= id-start id-end))
d9e94c22
MS
5569 (setq c-last-identifier-range
5570 (cons id-start id-end)))
5571 (goto-char id-end)
5572 (c-forward-syntactic-ws)
5573 (setq pos (point)
5574 res t)))
5575
5576 (progn
5577 (goto-char pos)
5578 (when (or c-opt-identifier-concat-key
5579 c-recognize-<>-arglists)
5580
5581 (cond
5582 ((and c-opt-identifier-concat-key
5583 (looking-at c-opt-identifier-concat-key))
5584 ;; Got a concatenated identifier. This handles the
5585 ;; cases with tricky syntactic whitespace that aren't
5586 ;; covered in `c-identifier-key'.
5587 (goto-char (match-end 0))
5588 (c-forward-syntactic-ws)
5589 t)
5590
5591 ((and c-recognize-<>-arglists
5592 (eq (char-after) ?<))
5593 ;; Maybe an angle bracket arglist.
0386b551
AM
5594
5595 (when (let (c-record-type-identifiers
5596 c-record-found-types)
5597 (c-forward-<>-arglist nil))
5598
5599 (c-add-type start (1+ pos))
d9e94c22 5600 (c-forward-syntactic-ws)
0386b551
AM
5601 (setq pos (point)
5602 c-last-identifier-range nil)
5603
d9e94c22
MS
5604 (if (and c-opt-identifier-concat-key
5605 (looking-at c-opt-identifier-concat-key))
0386b551 5606
d9e94c22
MS
5607 ;; Continue if there's an identifier concatenation
5608 ;; operator after the template argument.
5609 (progn
0386b551
AM
5610 (when (and c-record-type-identifiers id-start)
5611 (c-record-ref-id (cons id-start id-end)))
d9e94c22
MS
5612 (forward-char 2)
5613 (c-forward-syntactic-ws)
5614 t)
0386b551
AM
5615
5616 (when (and c-record-type-identifiers id-start)
5617 (c-record-type-id (cons id-start id-end)))
d9e94c22
MS
5618 (setq res 'template)
5619 nil)))
5620 )))))
5621
5622 (goto-char pos)
5623 res))
5624
5625(defun c-forward-type ()
5626 ;; Move forward over a type spec if at the beginning of one,
5627 ;; stopping at the next following token. Return t if it's a known
2a15eb73
MS
5628 ;; type that can't be a name or other expression, 'known if it's an
5629 ;; otherwise known type (according to `*-font-lock-extra-types'),
5630 ;; 'prefix if it's a known prefix of a type, 'found if it's a type
5631 ;; that matches one in `c-found-types', 'maybe if it's an identfier
5632 ;; that might be a type, or nil if it can't be a type (the point
5633 ;; isn't moved then). The point is assumed to be at the beginning
5634 ;; of a token.
d9e94c22
MS
5635 ;;
5636 ;; Note that this function doesn't skip past the brace definition
5637 ;; that might be considered part of the type, e.g.
5638 ;; "enum {a, b, c} foo".
0386b551
AM
5639 ;;
5640 ;; This function records identifier ranges on
5641 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
5642 ;; `c-record-type-identifiers' is non-nil.
5643 ;;
5644 ;; This function might do hidden buffer changes.
5645
5646 (let ((start (point)) pos res name-res id-start id-end id-range)
d9e94c22
MS
5647
5648 ;; Skip leading type modifiers. If any are found we know it's a
5649 ;; prefix of a type.
5650 (when c-opt-type-modifier-key
5651 (while (looking-at c-opt-type-modifier-key)
5652 (goto-char (match-end 1))
5653 (c-forward-syntactic-ws)
5654 (setq res 'prefix)))
5655
5656 (cond
5657 ((looking-at c-type-prefix-key)
5658 ;; Looking at a keyword that prefixes a type identifier,
5659 ;; e.g. "class".
5660 (goto-char (match-end 1))
5661 (c-forward-syntactic-ws)
5662 (setq pos (point))
0386b551 5663 (if (memq (setq name-res (c-forward-name)) '(t template))
d9e94c22 5664 (progn
0386b551 5665 (when (eq name-res t)
d9e94c22
MS
5666 ;; In many languages the name can be used without the
5667 ;; prefix, so we add it to `c-found-types'.
5668 (c-add-type pos (point))
0386b551
AM
5669 (when (and c-record-type-identifiers
5670 c-last-identifier-range)
d9e94c22
MS
5671 (c-record-type-id c-last-identifier-range)))
5672 (setq res t))
5673 ;; Invalid syntax.
5674 (goto-char start)
5675 (setq res nil)))
5676
5677 ((progn
5678 (setq pos nil)
5679 (if (looking-at c-identifier-start)
5680 (save-excursion
5681 (setq id-start (point)
0386b551
AM
5682 name-res (c-forward-name))
5683 (when name-res
d9e94c22
MS
5684 (setq id-end (point)
5685 id-range c-last-identifier-range))))
5686 (and (cond ((looking-at c-primitive-type-key)
5687 (setq res t))
5688 ((c-with-syntax-table c-identifier-syntax-table
5689 (looking-at c-known-type-key))
5690 (setq res 'known)))
5691 (or (not id-end)
5692 (>= (save-excursion
5693 (save-match-data
5694 (goto-char (match-end 1))
5695 (c-forward-syntactic-ws)
5696 (setq pos (point))))
5697 id-end)
5698 (setq res nil))))
5699 ;; Looking at a primitive or known type identifier. We've
5700 ;; checked for a name first so that we don't go here if the
5701 ;; known type match only is a prefix of another name.
5702
5703 (setq id-end (match-end 1))
5704
5705 (when (and c-record-type-identifiers
5706 (or c-promote-possible-types (eq res t)))
5707 (c-record-type-id (cons (match-beginning 1) (match-end 1))))
5708
5709 (if (and c-opt-type-component-key
5710 (save-match-data
5711 (looking-at c-opt-type-component-key)))
5712 ;; There might be more keywords for the type.
5713 (let (safe-pos)
0386b551 5714 (c-forward-keyword-clause 1)
d9e94c22
MS
5715 (while (progn
5716 (setq safe-pos (point))
5717 (looking-at c-opt-type-component-key))
5718 (when (and c-record-type-identifiers
5719 (looking-at c-primitive-type-key))
5720 (c-record-type-id (cons (match-beginning 1)
5721 (match-end 1))))
0386b551 5722 (c-forward-keyword-clause 1))
d9e94c22
MS
5723 (if (looking-at c-primitive-type-key)
5724 (progn
5725 (when c-record-type-identifiers
5726 (c-record-type-id (cons (match-beginning 1)
5727 (match-end 1))))
0386b551 5728 (c-forward-keyword-clause 1)
d9e94c22
MS
5729 (setq res t))
5730 (goto-char safe-pos)
5731 (setq res 'prefix)))
0386b551 5732 (unless (save-match-data (c-forward-keyword-clause 1))
d9e94c22
MS
5733 (if pos
5734 (goto-char pos)
5735 (goto-char (match-end 1))
5736 (c-forward-syntactic-ws)))))
5737
0386b551
AM
5738 (name-res
5739 (cond ((eq name-res t)
d9e94c22
MS
5740 ;; A normal identifier.
5741 (goto-char id-end)
5742 (if (or res c-promote-possible-types)
5743 (progn
5744 (c-add-type id-start id-end)
0386b551 5745 (when (and c-record-type-identifiers id-range)
d9e94c22
MS
5746 (c-record-type-id id-range))
5747 (unless res
5748 (setq res 'found)))
5749 (setq res (if (c-check-type id-start id-end)
5750 ;; It's an identifier that has been used as
5751 ;; a type somewhere else.
5752 'found
5753 ;; It's an identifier that might be a type.
5754 'maybe))))
0386b551 5755 ((eq name-res 'template)
d9e94c22
MS
5756 ;; A template is a type.
5757 (goto-char id-end)
5758 (setq res t))
5759 (t
5760 ;; Otherwise it's an operator identifier, which is not a type.
5761 (goto-char start)
5762 (setq res nil)))))
5763
5764 (when res
5765 ;; Skip trailing type modifiers. If any are found we know it's
5766 ;; a type.
5767 (when c-opt-type-modifier-key
5768 (while (looking-at c-opt-type-modifier-key)
5769 (goto-char (match-end 1))
5770 (c-forward-syntactic-ws)
5771 (setq res t)))
5772
5773 ;; Step over any type suffix operator. Do not let the existence
5774 ;; of these alter the classification of the found type, since
5775 ;; these operators typically are allowed in normal expressions
5776 ;; too.
5777 (when c-opt-type-suffix-key
5778 (while (looking-at c-opt-type-suffix-key)
5779 (goto-char (match-end 1))
5780 (c-forward-syntactic-ws)))
5781
5782 (when c-opt-type-concat-key
0386b551
AM
5783 ;; Look for a trailing operator that concatenates the type
5784 ;; with a following one, and if so step past that one through
5785 ;; a recursive call. Note that we don't record concatenated
5786 ;; types in `c-found-types' - it's the component types that
5787 ;; are recorded when appropriate.
d9e94c22
MS
5788 (setq pos (point))
5789 (let* ((c-promote-possible-types (or (memq res '(t known))
5790 c-promote-possible-types))
5791 ;; If we can't promote then set `c-record-found-types' so that
5792 ;; we can merge in the types from the second part afterwards if
5793 ;; it turns out to be a known type there.
5794 (c-record-found-types (and c-record-type-identifiers
0386b551
AM
5795 (not c-promote-possible-types)))
5796 subres)
d9e94c22
MS
5797 (if (and (looking-at c-opt-type-concat-key)
5798
5799 (progn
5800 (goto-char (match-end 1))
5801 (c-forward-syntactic-ws)
0386b551 5802 (setq subres (c-forward-type))))
d9e94c22
MS
5803
5804 (progn
5805 ;; If either operand certainly is a type then both are, but we
5806 ;; don't let the existence of the operator itself promote two
5807 ;; uncertain types to a certain one.
5808 (cond ((eq res t))
0386b551
AM
5809 ((eq subres t)
5810 (unless (eq name-res 'template)
5811 (c-add-type id-start id-end))
5812 (when (and c-record-type-identifiers id-range)
d9e94c22
MS
5813 (c-record-type-id id-range))
5814 (setq res t))
2a15eb73 5815 ((eq res 'known))
0386b551 5816 ((eq subres 'known)
2a15eb73 5817 (setq res 'known))
d9e94c22 5818 ((eq res 'found))
0386b551 5819 ((eq subres 'found)
d9e94c22
MS
5820 (setq res 'found))
5821 (t
5822 (setq res 'maybe)))
5823
5824 (when (and (eq res t)
5825 (consp c-record-found-types))
5826 ;; Merge in the ranges of any types found by the second
5827 ;; `c-forward-type'.
5828 (setq c-record-type-identifiers
5829 ;; `nconc' doesn't mind that the tail of
5830 ;; `c-record-found-types' is t.
5831 (nconc c-record-found-types
5832 c-record-type-identifiers))))
5833
5834 (goto-char pos))))
5835
5836 (when (and c-record-found-types (memq res '(known found)) id-range)
5837 (setq c-record-found-types
5838 (cons id-range c-record-found-types))))
5839
5840 ;;(message "c-forward-type %s -> %s: %s" start (point) res)
5841
5842 res))
5843
785eecbb 5844\f
d9e94c22
MS
5845;; Handling of large scale constructs like statements and declarations.
5846
0386b551
AM
5847;; Macro used inside `c-forward-decl-or-cast-1'. It ought to be a
5848;; defsubst or perhaps even a defun, but it contains lots of free
5849;; variables that refer to things inside `c-forward-decl-or-cast-1'.
5850(defmacro c-fdoc-shift-type-backward (&optional short)
5851 ;; `c-forward-decl-or-cast-1' can consume an arbitrary length list
5852 ;; of types when parsing a declaration, which means that it
5853 ;; sometimes consumes the identifier in the declaration as a type.
5854 ;; This is used to "backtrack" and make the last type be treated as
5855 ;; an identifier instead.
5856 `(progn
5857 ,(unless short
5858 ;; These identifiers are bound only in the inner let.
5859 '(setq identifier-type at-type
5860 identifier-start type-start
5861 got-parens nil
5862 got-identifier t
5863 got-suffix t
5864 got-suffix-after-parens id-start
5865 paren-depth 0))
5866
5867 (if (setq at-type (if (eq backup-at-type 'prefix)
5868 t
5869 backup-at-type))
5870 (setq type-start backup-type-start
5871 id-start backup-id-start)
5872 (setq type-start start-pos
5873 id-start start-pos))
5874
5875 ;; When these flags already are set we've found specifiers that
5876 ;; unconditionally signal these attributes - backtracking doesn't
5877 ;; change that. So keep them set in that case.
5878 (or at-type-decl
5879 (setq at-type-decl backup-at-type-decl))
5880 (or maybe-typeless
5881 (setq maybe-typeless backup-maybe-typeless))
5882
5883 ,(unless short
5884 ;; This identifier is bound only in the inner let.
5885 '(setq start id-start))))
5886
5887(defun c-forward-decl-or-cast-1 (preceding-token-end context last-cast-end)
5888 ;; Move forward over a declaration or a cast if at the start of one.
5889 ;; The point is assumed to be at the start of some token. Nil is
5890 ;; returned if no declaration or cast is recognized, and the point
5891 ;; is clobbered in that case.
5892 ;;
5893 ;; If a declaration is parsed:
5894 ;;
5895 ;; The point is left at the first token after the first complete
5896 ;; declarator, if there is one. The return value is a cons where
022d0cf4
AM
5897 ;; the car is the position of the first token in the declarator. (See
5898 ;; below for the cdr.)
0386b551
AM
5899 ;; Some examples:
5900 ;;
5901 ;; void foo (int a, char *b) stuff ...
5902 ;; car ^ ^ point
5903 ;; float (*a)[], b;
5904 ;; car ^ ^ point
5905 ;; unsigned int a = c_style_initializer, b;
5906 ;; car ^ ^ point
5907 ;; unsigned int a (cplusplus_style_initializer), b;
5908 ;; car ^ ^ point (might change)
5909 ;; class Foo : public Bar {}
5910 ;; car ^ ^ point
5911 ;; class PikeClass (int a, string b) stuff ...
5912 ;; car ^ ^ point
5913 ;; enum bool;
5914 ;; car ^ ^ point
5915 ;; enum bool flag;
5916 ;; car ^ ^ point
5917 ;; void cplusplus_function (int x) throw (Bad);
5918 ;; car ^ ^ point
5919 ;; Foo::Foo (int b) : Base (b) {}
5920 ;; car ^ ^ point
5921 ;;
022d0cf4
AM
5922 ;; The cdr of the return value is non-nil iff a `c-typedef-decl-kwds'
5923 ;; specifier (e.g. class, struct, enum, typedef) is found in the
5924 ;; declaration, i.e. the declared identifier(s) are types.
0386b551
AM
5925 ;;
5926 ;; If a cast is parsed:
5927 ;;
5928 ;; The point is left at the first token after the closing paren of
5929 ;; the cast. The return value is `cast'. Note that the start
5930 ;; position must be at the first token inside the cast parenthesis
5931 ;; to recognize it.
5932 ;;
5933 ;; PRECEDING-TOKEN-END is the first position after the preceding
5934 ;; token, i.e. on the other side of the syntactic ws from the point.
5935 ;; Use a value less than or equal to (point-min) if the point is at
5936 ;; the first token in (the visible part of) the buffer.
5937 ;;
5938 ;; CONTEXT is a symbol that describes the context at the point:
022d0cf4 5939 ;; 'decl In a comma-separated declaration context (typically
0386b551
AM
5940 ;; inside a function declaration arglist).
5941 ;; '<> In an angle bracket arglist.
5942 ;; 'arglist Some other type of arglist.
a85fd6da
AM
5943 ;; nil Some other context or unknown context. Includes
5944 ;; within the parens of an if, for, ... construct.
0386b551
AM
5945 ;;
5946 ;; LAST-CAST-END is the first token after the closing paren of a
5947 ;; preceding cast, or nil if none is known. If
5948 ;; `c-forward-decl-or-cast-1' is used in succession, it should be
5949 ;; the position after the closest preceding call where a cast was
5950 ;; matched. In that case it's used to discover chains of casts like
5951 ;; "(a) (b) c".
5952 ;;
5953 ;; This function records identifier ranges on
5954 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
5955 ;; `c-record-type-identifiers' is non-nil.
5956 ;;
5957 ;; This function might do hidden buffer changes.
5958
5959 (let (;; `start-pos' is used below to point to the start of the
5960 ;; first type, i.e. after any leading specifiers. It might
5961 ;; also point at the beginning of the preceding syntactic
5962 ;; whitespace.
5963 (start-pos (point))
5964 ;; Set to the result of `c-forward-type'.
5965 at-type
5966 ;; The position of the first token in what we currently
5967 ;; believe is the type in the declaration or cast, after any
5968 ;; specifiers and their associated clauses.
5969 type-start
5970 ;; The position of the first token in what we currently
5971 ;; believe is the declarator for the first identifier. Set
5972 ;; when the type is found, and moved forward over any
5973 ;; `c-decl-hangon-kwds' and their associated clauses that
5974 ;; occurs after the type.
5975 id-start
5976 ;; These store `at-type', `type-start' and `id-start' of the
5977 ;; identifier before the one in those variables. The previous
5978 ;; identifier might turn out to be the real type in a
5979 ;; declaration if the last one has to be the declarator in it.
5980 ;; If `backup-at-type' is nil then the other variables have
5981 ;; undefined values.
5982 backup-at-type backup-type-start backup-id-start
5983 ;; Set if we've found a specifier that makes the defined
5984 ;; identifier(s) types.
5985 at-type-decl
5986 ;; Set if we've found a specifier that can start a declaration
5987 ;; where there's no type.
5988 maybe-typeless
5989 ;; If a specifier is found that also can be a type prefix,
5990 ;; these flags are set instead of those above. If we need to
5991 ;; back up an identifier, they are copied to the real flag
5992 ;; variables. Thus they only take effect if we fail to
5993 ;; interpret it as a type.
5994 backup-at-type-decl backup-maybe-typeless
5995 ;; Whether we've found a declaration or a cast. We might know
5996 ;; this before we've found the type in it. It's 'ids if we've
5997 ;; found two consecutive identifiers (usually a sure sign, but
5998 ;; we should allow that in labels too), and t if we've found a
5999 ;; specifier keyword (a 100% sure sign).
6000 at-decl-or-cast
6001 ;; Set when we need to back up to parse this as a declaration
6002 ;; but not as a cast.
6003 backup-if-not-cast
6004 ;; For casts, the return position.
6005 cast-end
6006 ;; Save `c-record-type-identifiers' and
6007 ;; `c-record-ref-identifiers' since ranges are recorded
6008 ;; speculatively and should be thrown away if it turns out
6009 ;; that it isn't a declaration or cast.
6010 (save-rec-type-ids c-record-type-identifiers)
6011 (save-rec-ref-ids c-record-ref-identifiers))
6012
6013 ;; Check for a type. Unknown symbols are treated as possible
6014 ;; types, but they could also be specifiers disguised through
6015 ;; macros like __INLINE__, so we recognize both types and known
6016 ;; specifiers after them too.
6017 (while
6018 (let* ((start (point)) kwd-sym kwd-clause-end found-type)
6019
6020 ;; Look for a specifier keyword clause.
6021 (when (looking-at c-prefix-spec-kwds-re)
6022 (setq kwd-sym (c-keyword-sym (match-string 1)))
6023 (save-excursion
6024 (c-forward-keyword-clause 1)
6025 (setq kwd-clause-end (point))))
6026
6027 (when (setq found-type (c-forward-type))
6028 ;; Found a known or possible type or a prefix of a known type.
6029
6030 (when at-type
6031 ;; Got two identifiers with nothing but whitespace
6032 ;; between them. That can only happen in declarations.
6033 (setq at-decl-or-cast 'ids)
6034
6035 (when (eq at-type 'found)
6036 ;; If the previous identifier is a found type we
6037 ;; record it as a real one; it might be some sort of
6038 ;; alias for a prefix like "unsigned".
6039 (save-excursion
6040 (goto-char type-start)
6041 (let ((c-promote-possible-types t))
6042 (c-forward-type)))))
6043
6044 (setq backup-at-type at-type
6045 backup-type-start type-start
6046 backup-id-start id-start
6047 at-type found-type
6048 type-start start
6049 id-start (point)
6050 ;; The previous ambiguous specifier/type turned out
6051 ;; to be a type since we've parsed another one after
6052 ;; it, so clear these backup flags.
6053 backup-at-type-decl nil
6054 backup-maybe-typeless nil))
6055
6056 (if kwd-sym
6057 (progn
6058 ;; Handle known specifier keywords and
6059 ;; `c-decl-hangon-kwds' which can occur after known
6060 ;; types.
6061
6062 (if (c-keyword-member kwd-sym 'c-decl-hangon-kwds)
6063 ;; It's a hang-on keyword that can occur anywhere.
6064 (progn
6065 (setq at-decl-or-cast t)
6066 (if at-type
6067 ;; Move the identifier start position if
6068 ;; we've passed a type.
6069 (setq id-start kwd-clause-end)
6070 ;; Otherwise treat this as a specifier and
6071 ;; move the fallback position.
6072 (setq start-pos kwd-clause-end))
6073 (goto-char kwd-clause-end))
6074
6075 ;; It's an ordinary specifier so we know that
6076 ;; anything before this can't be the type.
6077 (setq backup-at-type nil
6078 start-pos kwd-clause-end)
6079
6080 (if found-type
6081 ;; It's ambiguous whether this keyword is a
6082 ;; specifier or a type prefix, so set the backup
6083 ;; flags. (It's assumed that `c-forward-type'
6084 ;; moved further than `c-forward-keyword-clause'.)
6085 (progn
6086 (when (c-keyword-member kwd-sym 'c-typedef-decl-kwds)
6087 (setq backup-at-type-decl t))
6088 (when (c-keyword-member kwd-sym 'c-typeless-decl-kwds)
6089 (setq backup-maybe-typeless t)))
6090
6091 (when (c-keyword-member kwd-sym 'c-typedef-decl-kwds)
6092 (setq at-type-decl t))
6093 (when (c-keyword-member kwd-sym 'c-typeless-decl-kwds)
6094 (setq maybe-typeless t))
6095
6096 ;; Haven't matched a type so it's an umambiguous
6097 ;; specifier keyword and we know we're in a
6098 ;; declaration.
6099 (setq at-decl-or-cast t)
6100
6101 (goto-char kwd-clause-end))))
6102
6103 ;; If the type isn't known we continue so that we'll jump
6104 ;; over all specifiers and type identifiers. The reason
6105 ;; to do this for a known type prefix is to make things
6106 ;; like "unsigned INT16" work.
6107 (and found-type (not (eq found-type t))))))
6108
6109 (cond
6110 ((eq at-type t)
6111 ;; If a known type was found, we still need to skip over any
6112 ;; hangon keyword clauses after it. Otherwise it has already
6113 ;; been done in the loop above.
6114 (while (looking-at c-decl-hangon-key)
6115 (c-forward-keyword-clause 1))
6116 (setq id-start (point)))
6117
6118 ((eq at-type 'prefix)
6119 ;; A prefix type is itself a primitive type when it's not
6120 ;; followed by another type.
6121 (setq at-type t))
6122
6123 ((not at-type)
6124 ;; Got no type but set things up to continue anyway to handle
6125 ;; the various cases when a declaration doesn't start with a
6126 ;; type.
6127 (setq id-start start-pos))
6128
6129 ((and (eq at-type 'maybe)
6130 (c-major-mode-is 'c++-mode))
6131 ;; If it's C++ then check if the last "type" ends on the form
6132 ;; "foo::foo" or "foo::~foo", i.e. if it's the name of a
6133 ;; (con|de)structor.
6134 (save-excursion
6135 (let (name end-2 end-1)
6136 (goto-char id-start)
6137 (c-backward-syntactic-ws)
6138 (setq end-2 (point))
6139 (when (and
6140 (c-simple-skip-symbol-backward)
6141 (progn
6142 (setq name
6143 (buffer-substring-no-properties (point) end-2))
6144 ;; Cheating in the handling of syntactic ws below.
6145 (< (skip-chars-backward ":~ \t\n\r\v\f") 0))
6146 (progn
6147 (setq end-1 (point))
6148 (c-simple-skip-symbol-backward))
6149 (>= (point) type-start)
6150 (equal (buffer-substring-no-properties (point) end-1)
6151 name))
6152 ;; It is a (con|de)structor name. In that case the
6153 ;; declaration is typeless so zap out any preceding
6154 ;; identifier(s) that we might have taken as types.
6155 (goto-char type-start)
6156 (setq at-type nil
6157 backup-at-type nil
6158 id-start type-start))))))
6159
6160 ;; Check for and step over a type decl expression after the thing
6161 ;; that is or might be a type. This can't be skipped since we
6162 ;; need the correct end position of the declarator for
6163 ;; `max-type-decl-end-*'.
6164 (let ((start (point)) (paren-depth 0) pos
6165 ;; True if there's a non-open-paren match of
6166 ;; `c-type-decl-prefix-key'.
6167 got-prefix
6168 ;; True if the declarator is surrounded by a parenthesis pair.
6169 got-parens
6170 ;; True if there is an identifier in the declarator.
6171 got-identifier
6172 ;; True if there's a non-close-paren match of
6173 ;; `c-type-decl-suffix-key'.
6174 got-suffix
6175 ;; True if there's a prefix match outside the outermost
6176 ;; paren pair that surrounds the declarator.
6177 got-prefix-before-parens
1379f2c5 6178 ;; True if there's a suffix match outside the outermost
0386b551
AM
6179 ;; paren pair that surrounds the declarator. The value is
6180 ;; the position of the first suffix match.
6181 got-suffix-after-parens
6182 ;; True if we've parsed the type decl to a token that is
6183 ;; known to end declarations in this context.
6184 at-decl-end
6185 ;; The earlier values of `at-type' and `type-start' if we've
6186 ;; shifted the type backwards.
6187 identifier-type identifier-start
6188 ;; If `c-parse-and-markup-<>-arglists' is set we need to
6189 ;; turn it off during the name skipping below to avoid
6190 ;; getting `c-type' properties that might be bogus. That
6191 ;; can happen since we don't know if
6192 ;; `c-restricted-<>-arglists' will be correct inside the
6193 ;; arglist paren that gets entered.
6194 c-parse-and-markup-<>-arglists)
6195
6196 (goto-char id-start)
6197
6198 ;; Skip over type decl prefix operators. (Note similar code in
6199 ;; `c-font-lock-declarators'.)
6200 (while (and (looking-at c-type-decl-prefix-key)
6201 (if (and (c-major-mode-is 'c++-mode)
6202 (match-beginning 2))
6203 ;; If the second submatch matches in C++ then
6204 ;; we're looking at an identifier that's a
6205 ;; prefix only if it specifies a member pointer.
6206 (when (setq got-identifier (c-forward-name))
6207 (if (looking-at "\\(::\\)")
6208 ;; We only check for a trailing "::" and
6209 ;; let the "*" that should follow be
6210 ;; matched in the next round.
6211 (progn (setq got-identifier nil) t)
6212 ;; It turned out to be the real identifier,
6213 ;; so stop.
6214 nil))
6215 t))
6216
6217 (if (eq (char-after) ?\()
6218 (progn
6219 (setq paren-depth (1+ paren-depth))
6220 (forward-char))
6221 (unless got-prefix-before-parens
6222 (setq got-prefix-before-parens (= paren-depth 0)))
6223 (setq got-prefix t)
6224 (goto-char (match-end 1)))
6225 (c-forward-syntactic-ws))
6226
6227 (setq got-parens (> paren-depth 0))
6228
6229 ;; Skip over an identifier.
6230 (or got-identifier
6231 (and (looking-at c-identifier-start)
6232 (setq got-identifier (c-forward-name))))
6233
6234 ;; Skip over type decl suffix operators.
6235 (while (if (looking-at c-type-decl-suffix-key)
6236
6237 (if (eq (char-after) ?\))
6238 (when (> paren-depth 0)
6239 (setq paren-depth (1- paren-depth))
6240 (forward-char)
6241 t)
6242 (when (if (save-match-data (looking-at "\\s\("))
6243 (c-safe (c-forward-sexp 1) t)
6244 (goto-char (match-end 1))
6245 t)
6246 (when (and (not got-suffix-after-parens)
6247 (= paren-depth 0))
6248 (setq got-suffix-after-parens (match-beginning 0)))
6249 (setq got-suffix t)))
6250
6251 ;; No suffix matched. We might have matched the
6252 ;; identifier as a type and the open paren of a
6253 ;; function arglist as a type decl prefix. In that
6254 ;; case we should "backtrack": Reinterpret the last
6255 ;; type as the identifier, move out of the arglist and
6256 ;; continue searching for suffix operators.
6257 ;;
6258 ;; Do this even if there's no preceding type, to cope
6259 ;; with old style function declarations in K&R C,
6260 ;; (con|de)structors in C++ and `c-typeless-decl-kwds'
6261 ;; style declarations. That isn't applicable in an
6262 ;; arglist context, though.
6263 (when (and (= paren-depth 1)
6264 (not got-prefix-before-parens)
6265 (not (eq at-type t))
6266 (or backup-at-type
6267 maybe-typeless
6268 backup-maybe-typeless
6269 (when c-recognize-typeless-decls
6270 (not context)))
6271 (setq pos (c-up-list-forward (point)))
6272 (eq (char-before pos) ?\)))
6273 (c-fdoc-shift-type-backward)
6274 (goto-char pos)
6275 t))
6276
6277 (c-forward-syntactic-ws))
6278
6279 (when (and (or maybe-typeless backup-maybe-typeless)
6280 (not got-identifier)
6281 (not got-prefix)
6282 at-type)
6283 ;; Have found no identifier but `c-typeless-decl-kwds' has
6284 ;; matched so we know we're inside a declaration. The
6285 ;; preceding type must be the identifier instead.
6286 (c-fdoc-shift-type-backward))
6287
6288 (setq
6289 at-decl-or-cast
6290 (catch 'at-decl-or-cast
6291
a85fd6da 6292 ;; CASE 1
0386b551
AM
6293 (when (> paren-depth 0)
6294 ;; Encountered something inside parens that isn't matched by
6295 ;; the `c-type-decl-*' regexps, so it's not a type decl
6296 ;; expression. Try to skip out to the same paren depth to
6297 ;; not confuse the cast check below.
6298 (c-safe (goto-char (scan-lists (point) 1 paren-depth)))
6299 ;; If we've found a specifier keyword then it's a
6300 ;; declaration regardless.
6301 (throw 'at-decl-or-cast (eq at-decl-or-cast t)))
6302
6303 (setq at-decl-end
6304 (looking-at (cond ((eq context '<>) "[,>]")
6305 (context "[,\)]")
6306 (t "[,;]"))))
6307
6308 ;; Now we've collected info about various characteristics of
6309 ;; the construct we're looking at. Below follows a decision
6310 ;; tree based on that. It's ordered to check more certain
6311 ;; signs before less certain ones.
6312
6313 (if got-identifier
6314 (progn
6315
a85fd6da 6316 ;; CASE 2
0386b551
AM
6317 (when (and (or at-type maybe-typeless)
6318 (not (or got-prefix got-parens)))
6319 ;; Got another identifier directly after the type, so it's a
6320 ;; declaration.
6321 (throw 'at-decl-or-cast t))
6322
6323 (when (and got-parens
6324 (not got-prefix)
6325 (not got-suffix-after-parens)
6326 (or backup-at-type
6327 maybe-typeless
6328 backup-maybe-typeless))
6329 ;; Got a declaration of the form "foo bar (gnu);" where we've
6330 ;; recognized "bar" as the type and "gnu" as the declarator.
6331 ;; In this case it's however more likely that "bar" is the
6332 ;; declarator and "gnu" a function argument or initializer (if
6333 ;; `c-recognize-paren-inits' is set), since the parens around
6334 ;; "gnu" would be superfluous if it's a declarator. Shift the
6335 ;; type one step backward.
6336 (c-fdoc-shift-type-backward)))
6337
6338 ;; Found no identifier.
6339
6340 (if backup-at-type
6341 (progn
6342
a85fd6da 6343 ;; CASE 3
0386b551
AM
6344 (when (= (point) start)
6345 ;; Got a plain list of identifiers. If a colon follows it's
6346 ;; a valid label. Otherwise the last one probably is the
6347 ;; declared identifier and we should back up to the previous
6348 ;; type, providing it isn't a cast.
6349 (if (eq (char-after) ?:)
6350 ;; If we've found a specifier keyword then it's a
6351 ;; declaration regardless.
6352 (throw 'at-decl-or-cast (eq at-decl-or-cast t))
6353 (setq backup-if-not-cast t)
6354 (throw 'at-decl-or-cast t)))
6355
a85fd6da 6356 ;; CASE 4
0386b551
AM
6357 (when (and got-suffix
6358 (not got-prefix)
6359 (not got-parens))
6360 ;; Got a plain list of identifiers followed by some suffix.
6361 ;; If this isn't a cast then the last identifier probably is
6362 ;; the declared one and we should back up to the previous
6363 ;; type.
6364 (setq backup-if-not-cast t)
6365 (throw 'at-decl-or-cast t)))
6366
a85fd6da 6367 ;; CASE 5
0386b551
AM
6368 (when (eq at-type t)
6369 ;; If the type is known we know that there can't be any
6370 ;; identifier somewhere else, and it's only in declarations in
6371 ;; e.g. function prototypes and in casts that the identifier may
6372 ;; be left out.
6373 (throw 'at-decl-or-cast t))
6374
6375 (when (= (point) start)
6376 ;; Only got a single identifier (parsed as a type so far).
a85fd6da 6377 ;; CASE 6
0386b551
AM
6378 (if (and
6379 ;; Check that the identifier isn't at the start of an
6380 ;; expression.
6381 at-decl-end
6382 (cond
6383 ((eq context 'decl)
6384 ;; Inside an arglist that contains declarations. If K&R
6385 ;; style declarations and parenthesis style initializers
6386 ;; aren't allowed then the single identifier must be a
6387 ;; type, else we require that it's known or found
6388 ;; (primitive types are handled above).
6389 (or (and (not c-recognize-knr-p)
6390 (not c-recognize-paren-inits))
6391 (memq at-type '(known found))))
6392 ((eq context '<>)
6393 ;; Inside a template arglist. Accept known and found
6394 ;; types; other identifiers could just as well be
6395 ;; constants in C++.
6396 (memq at-type '(known found)))))
6397 (throw 'at-decl-or-cast t)
a85fd6da 6398 ;; CASE 7
0386b551
AM
6399 ;; Can't be a valid declaration or cast, but if we've found a
6400 ;; specifier it can't be anything else either, so treat it as
6401 ;; an invalid/unfinished declaration or cast.
6402 (throw 'at-decl-or-cast at-decl-or-cast))))
6403
6404 (if (and got-parens
6405 (not got-prefix)
6406 (not context)
6407 (not (eq at-type t))
6408 (or backup-at-type
6409 maybe-typeless
6410 backup-maybe-typeless
6411 (when c-recognize-typeless-decls
6412 (or (not got-suffix)
6413 (not (looking-at
6414 c-after-suffixed-type-maybe-decl-key))))))
6415 ;; Got an empty paren pair and a preceding type that probably
6416 ;; really is the identifier. Shift the type backwards to make
6417 ;; the last one the identifier. This is analogous to the
6418 ;; "backtracking" done inside the `c-type-decl-suffix-key' loop
6419 ;; above.
6420 ;;
6421 ;; Exception: In addition to the conditions in that
6422 ;; "backtracking" code, do not shift backward if we're not
6423 ;; looking at either `c-after-suffixed-type-decl-key' or "[;,]".
6424 ;; Since there's no preceding type, the shift would mean that
6425 ;; the declaration is typeless. But if the regexp doesn't match
6426 ;; then we will simply fall through in the tests below and not
6427 ;; recognize it at all, so it's better to try it as an abstract
6428 ;; declarator instead.
6429 (c-fdoc-shift-type-backward)
6430
6431 ;; Still no identifier.
a85fd6da 6432 ;; CASE 8
0386b551
AM
6433 (when (and got-prefix (or got-parens got-suffix))
6434 ;; Require `got-prefix' together with either `got-parens' or
6435 ;; `got-suffix' to recognize it as an abstract declarator:
6436 ;; `got-parens' only is probably an empty function call.
6437 ;; `got-suffix' only can build an ordinary expression together
6438 ;; with the preceding identifier which we've taken as a type.
6439 ;; We could actually accept on `got-prefix' only, but that can
6440 ;; easily occur temporarily while writing an expression so we
6441 ;; avoid that case anyway. We could do a better job if we knew
6442 ;; the point when the fontification was invoked.
6443 (throw 'at-decl-or-cast t))
6444
a85fd6da 6445 ;; CASE 9
0386b551
AM
6446 (when (and at-type
6447 (not got-prefix)
6448 (not got-parens)
6449 got-suffix-after-parens
6450 (eq (char-after got-suffix-after-parens) ?\())
6451 ;; Got a type, no declarator but a paren suffix. I.e. it's a
6452 ;; normal function call afterall (or perhaps a C++ style object
6453 ;; instantiation expression).
6454 (throw 'at-decl-or-cast nil))))
6455
a85fd6da 6456 ;; CASE 10
0386b551
AM
6457 (when at-decl-or-cast
6458 ;; By now we've located the type in the declaration that we know
6459 ;; we're in.
6460 (throw 'at-decl-or-cast t))
6461
a85fd6da 6462 ;; CASE 11
0386b551
AM
6463 (when (and got-identifier
6464 (not context)
6465 (looking-at c-after-suffixed-type-decl-key)
6466 (if (and got-parens
6467 (not got-prefix)
6468 (not got-suffix)
6469 (not (eq at-type t)))
6470 ;; Shift the type backward in the case that there's a
6471 ;; single identifier inside parens. That can only
6472 ;; occur in K&R style function declarations so it's
6473 ;; more likely that it really is a function call.
6474 ;; Therefore we only do this after
6475 ;; `c-after-suffixed-type-decl-key' has matched.
6476 (progn (c-fdoc-shift-type-backward) t)
6477 got-suffix-after-parens))
6478 ;; A declaration according to `c-after-suffixed-type-decl-key'.
6479 (throw 'at-decl-or-cast t))
6480
a85fd6da 6481 ;; CASE 12
0386b551
AM
6482 (when (and (or got-prefix (not got-parens))
6483 (memq at-type '(t known)))
6484 ;; It's a declaration if a known type precedes it and it can't be a
6485 ;; function call.
6486 (throw 'at-decl-or-cast t))
6487
6488 ;; If we get here we can't tell if this is a type decl or a normal
6489 ;; expression by looking at it alone. (That's under the assumption
6490 ;; that normal expressions always can look like type decl expressions,
6491 ;; which isn't really true but the cases where it doesn't hold are so
6492 ;; uncommon (e.g. some placements of "const" in C++) it's not worth
6493 ;; the effort to look for them.)
6494
6495 (unless (or at-decl-end (looking-at "=[^=]"))
6496 ;; If this is a declaration it should end here or its initializer(*)
6497 ;; should start here, so check for allowed separation tokens. Note
6498 ;; that this rule doesn't work e.g. with a K&R arglist after a
6499 ;; function header.
6500 ;;
6501 ;; *) Don't check for C++ style initializers using parens
6502 ;; since those already have been matched as suffixes.
6503 ;;
6504 ;; If `at-decl-or-cast' is then we've found some other sign that
6505 ;; it's a declaration or cast, so then it's probably an
6506 ;; invalid/unfinished one.
6507 (throw 'at-decl-or-cast at-decl-or-cast))
6508
6509 ;; Below are tests that only should be applied when we're certain to
6510 ;; not have parsed halfway through an expression.
6511
a85fd6da 6512 ;; CASE 14
0386b551
AM
6513 (when (memq at-type '(t known))
6514 ;; The expression starts with a known type so treat it as a
6515 ;; declaration.
6516 (throw 'at-decl-or-cast t))
6517
a85fd6da 6518 ;; CASE 15
0386b551
AM
6519 (when (and (c-major-mode-is 'c++-mode)
6520 ;; In C++ we check if the identifier is a known type, since
6521 ;; (con|de)structors use the class name as identifier.
6522 ;; We've always shifted over the identifier as a type and
6523 ;; then backed up again in this case.
6524 identifier-type
6525 (or (memq identifier-type '(found known))
6526 (and (eq (char-after identifier-start) ?~)
6527 ;; `at-type' probably won't be 'found for
6528 ;; destructors since the "~" is then part of the
6529 ;; type name being checked against the list of
6530 ;; known types, so do a check without that
6531 ;; operator.
6532 (or (save-excursion
6533 (goto-char (1+ identifier-start))
6534 (c-forward-syntactic-ws)
6535 (c-with-syntax-table
6536 c-identifier-syntax-table
6537 (looking-at c-known-type-key)))
6538 (save-excursion
6539 (goto-char (1+ identifier-start))
6540 ;; We have already parsed the type earlier,
6541 ;; so it'd be possible to cache the end
6542 ;; position instead of redoing it here, but
6543 ;; then we'd need to keep track of another
6544 ;; position everywhere.
6545 (c-check-type (point)
6546 (progn (c-forward-type)
6547 (point))))))))
6548 (throw 'at-decl-or-cast t))
6549
6550 (if got-identifier
6551 (progn
a85fd6da 6552 ;; CASE 16
0386b551
AM
6553 (when (and got-prefix-before-parens
6554 at-type
6555 (or at-decl-end (looking-at "=[^=]"))
6556 (not context)
6557 (not got-suffix))
6558 ;; Got something like "foo * bar;". Since we're not inside an
6559 ;; arglist it would be a meaningless expression because the
6560 ;; result isn't used. We therefore choose to recognize it as
6561 ;; a declaration. Do not allow a suffix since it could then
6562 ;; be a function call.
6563 (throw 'at-decl-or-cast t))
6564
a85fd6da 6565 ;; CASE 17
0386b551
AM
6566 (when (and (or got-suffix-after-parens
6567 (looking-at "=[^=]"))
6568 (eq at-type 'found)
6569 (not (eq context 'arglist)))
6570 ;; Got something like "a (*b) (c);" or "a (b) = c;". It could
6571 ;; be an odd expression or it could be a declaration. Treat
6572 ;; it as a declaration if "a" has been used as a type
6573 ;; somewhere else (if it's a known type we won't get here).
6574 (throw 'at-decl-or-cast t)))
6575
a85fd6da 6576 ;; CASE 18
0386b551
AM
6577 (when (and context
6578 (or got-prefix
6579 (and (eq context 'decl)
6580 (not c-recognize-paren-inits)
6581 (or got-parens got-suffix))))
6582 ;; Got a type followed by an abstract declarator. If `got-prefix'
6583 ;; is set it's something like "a *" without anything after it. If
6584 ;; `got-parens' or `got-suffix' is set it's "a()", "a[]", "a()[]",
6585 ;; or similar, which we accept only if the context rules out
6586 ;; expressions.
6587 (throw 'at-decl-or-cast t)))
6588
6589 ;; If we had a complete symbol table here (which rules out
6590 ;; `c-found-types') we should return t due to the disambiguation rule
6591 ;; (in at least C++) that anything that can be parsed as a declaration
6592 ;; is a declaration. Now we're being more defensive and prefer to
6593 ;; highlight things like "foo (bar);" as a declaration only if we're
6594 ;; inside an arglist that contains declarations.
6595 (eq context 'decl))))
6596
6597 ;; The point is now after the type decl expression.
6598
6599 (cond
6600 ;; Check for a cast.
6601 ((save-excursion
6602 (and
6603 c-cast-parens
6604
6605 ;; Should be the first type/identifier in a cast paren.
6606 (> preceding-token-end (point-min))
6607 (memq (char-before preceding-token-end) c-cast-parens)
6608
6609 ;; The closing paren should follow.
6610 (progn
6611 (c-forward-syntactic-ws)
6612 (looking-at "\\s\)"))
6613
6614 ;; There should be a primary expression after it.
6615 (let (pos)
6616 (forward-char)
6617 (c-forward-syntactic-ws)
6618 (setq cast-end (point))
6619 (and (looking-at c-primary-expr-regexp)
6620 (progn
6621 (setq pos (match-end 0))
6622 (or
6623 ;; Check if the expression begins with a prefix keyword.
6624 (match-beginning 2)
6625 (if (match-beginning 1)
6626 ;; Expression begins with an ambiguous operator. Treat
6627 ;; it as a cast if it's a type decl or if we've
6628 ;; recognized the type somewhere else.
6629 (or at-decl-or-cast
6630 (memq at-type '(t known found)))
6631 ;; Unless it's a keyword, it's the beginning of a primary
6632 ;; expression.
6633 (not (looking-at c-keywords-regexp)))))
6634 ;; If `c-primary-expr-regexp' matched a nonsymbol token, check
6635 ;; that it matched a whole one so that we don't e.g. confuse
6636 ;; the operator '-' with '->'. It's ok if it matches further,
6637 ;; though, since it e.g. can match the float '.5' while the
6638 ;; operator regexp only matches '.'.
6639 (or (not (looking-at c-nonsymbol-token-regexp))
6640 (<= (match-end 0) pos))))
6641
6642 ;; There should either be a cast before it or something that isn't an
6643 ;; identifier or close paren.
6644 (> preceding-token-end (point-min))
6645 (progn
6646 (goto-char (1- preceding-token-end))
6647 (or (eq (point) last-cast-end)
6648 (progn
6649 (c-backward-syntactic-ws)
6650 (if (< (skip-syntax-backward "w_") 0)
6651 ;; It's a symbol. Accept it only if it's one of the
6652 ;; keywords that can precede an expression (without
6653 ;; surrounding parens).
6654 (looking-at c-simple-stmt-key)
6655 (and
6656 ;; Check that it isn't a close paren (block close is ok,
6657 ;; though).
6658 (not (memq (char-before) '(?\) ?\])))
6659 ;; Check that it isn't a nonsymbol identifier.
6660 (not (c-on-identifier)))))))))
6661
6662 ;; Handle the cast.
6663 (when (and c-record-type-identifiers at-type (not (eq at-type t)))
6664 (let ((c-promote-possible-types t))
6665 (goto-char type-start)
6666 (c-forward-type)))
6667
6668 (goto-char cast-end)
6669 'cast)
6670
6671 (at-decl-or-cast
6672 ;; We're at a declaration. Highlight the type and the following
6673 ;; declarators.
6674
6675 (when backup-if-not-cast
6676 (c-fdoc-shift-type-backward t))
6677
6678 (when (and (eq context 'decl) (looking-at ","))
6679 ;; Make sure to propagate the `c-decl-arg-start' property to
6680 ;; the next argument if it's set in this one, to cope with
6681 ;; interactive refontification.
6682 (c-put-c-type-property (point) 'c-decl-arg-start))
6683
6684 (when (and c-record-type-identifiers at-type (not (eq at-type t)))
6685 (let ((c-promote-possible-types t))
6686 (save-excursion
6687 (goto-char type-start)
6688 (c-forward-type))))
6689
6690 (cons id-start at-type-decl))
6691
6692 (t
6693 ;; False alarm. Restore the recorded ranges.
6694 (setq c-record-type-identifiers save-rec-type-ids
6695 c-record-ref-identifiers save-rec-ref-ids)
6696 nil))))
6697
6698(defun c-forward-label (&optional assume-markup preceding-token-end limit)
51c9af45 6699 ;; Assuming that point is at the beginning of a token, check if it starts a
1379f2c5
AM
6700 ;; label and if so move over it and return non-nil (t in default situations,
6701 ;; specific symbols (see below) for interesting situations), otherwise don't
6702 ;; move and return nil. "Label" here means "most things with a colon".
51c9af45
AM
6703 ;;
6704 ;; More precisely, a "label" is regarded as one of:
1379f2c5
AM
6705 ;; (i) a goto target like "foo:" - returns the symbol `goto-target';
6706 ;; (ii) A case label - either the entire construct "case FOO:", or just the
6707 ;; bare "case", should the colon be missing. We return t;
6708 ;; (iii) a keyword which needs a colon, like "default:" or "private:"; We
6709 ;; return t;
51c9af45 6710 ;; (iv) One of QT's "extended" C++ variants of
1379f2c5
AM
6711 ;; "private:"/"protected:"/"public:"/"more:" looking like "public slots:".
6712 ;; Returns the symbol `qt-2kwds-colon'.
6713 ;; (v) QT's construct "signals:". Returns the symbol `qt-1kwd-colon'.
82ba65cf 6714 ;; (vi) One of the keywords matched by `c-opt-extra-label-key' (without any
51c9af45 6715 ;; colon). Currently (2006-03), this applies only to Objective C's
1379f2c5 6716 ;; keywords "@private", "@protected", and "@public". Returns t.
51c9af45
AM
6717 ;;
6718 ;; One of the things which will NOT be recognised as a label is a bit-field
6719 ;; element of a struct, something like "int foo:5".
6720 ;;
6721 ;; The end of the label is taken to be just after the colon, or the end of
6722 ;; the first submatch in `c-opt-extra-label-key'. The point is directly
6723 ;; after the end on return. The terminating char gets marked with
6724 ;; `c-decl-end' to improve recognition of the following declaration or
6725 ;; statement.
0386b551
AM
6726 ;;
6727 ;; If ASSUME-MARKUP is non-nil, it's assumed that the preceding
51c9af45 6728 ;; label, if any, has already been marked up like that.
0386b551
AM
6729 ;;
6730 ;; If PRECEDING-TOKEN-END is given, it should be the first position
6731 ;; after the preceding token, i.e. on the other side of the
6732 ;; syntactic ws from the point. Use a value less than or equal to
6733 ;; (point-min) if the point is at the first token in (the visible
6734 ;; part of) the buffer.
6735 ;;
6736 ;; The optional LIMIT limits the forward scan for the colon.
6737 ;;
6738 ;; This function records the ranges of the label symbols on
6739 ;; `c-record-ref-identifiers' if `c-record-type-identifiers' (!) is
6740 ;; non-nil.
6741 ;;
6742 ;; This function might do hidden buffer changes.
6743
51c9af45 6744 (let ((start (point))
1379f2c5 6745 label-end
51c9af45 6746 qt-symbol-idx
1379f2c5 6747 macro-start ; if we're in one.
f412a567
AM
6748 label-type
6749 kwd)
0386b551 6750 (cond
b414f371 6751 ;; "case" or "default" (Doesn't apply to AWK).
0386b551
AM
6752 ((looking-at c-label-kwds-regexp)
6753 (let ((kwd-end (match-end 1)))
6754 ;; Record only the keyword itself for fontification, since in
6755 ;; case labels the following is a constant expression and not
6756 ;; a label.
6757 (when c-record-type-identifiers
6758 (c-record-ref-id (cons (match-beginning 1) kwd-end)))
6759
6760 ;; Find the label end.
6761 (goto-char kwd-end)
1379f2c5
AM
6762 (setq label-type
6763 (if (and (c-syntactic-re-search-forward
6764 ;; Stop on chars that aren't allowed in expressions,
6765 ;; and on operator chars that would be meaningless
6766 ;; there. FIXME: This doesn't cope with ?: operators.
6767 "[;{=,@]\\|\\(\\=\\|[^:]\\):\\([^:]\\|\\'\\)"
6768 limit t t nil 1)
6769 (match-beginning 2))
6770
6771 (progn ; there's a proper :
6772 (goto-char (match-beginning 2)) ; just after the :
6773 (c-put-c-type-property (1- (point)) 'c-decl-end)
6774 t)
6775
6776 ;; It's an unfinished label. We consider the keyword enough
6777 ;; to recognize it as a label, so that it gets fontified.
6778 ;; Leave the point at the end of it, but don't put any
6779 ;; `c-decl-end' marker.
6780 (goto-char kwd-end)
6781 t))))
0386b551 6782
51c9af45 6783 ;; @private, @protected, @public, in Objective C, or similar.
0386b551
AM
6784 ((and c-opt-extra-label-key
6785 (looking-at c-opt-extra-label-key))
6786 ;; For a `c-opt-extra-label-key' match, we record the whole
6787 ;; thing for fontification. That's to get the leading '@' in
6788 ;; Objective-C protection labels fontified.
6789 (goto-char (match-end 1))
6790 (when c-record-type-identifiers
6791 (c-record-ref-id (cons (match-beginning 1) (point))))
6792 (c-put-c-type-property (1- (point)) 'c-decl-end)
1379f2c5 6793 (setq label-type t))
0386b551 6794
51c9af45
AM
6795 ;; All other cases of labels.
6796 ((and c-recognize-colon-labels ; nil for AWK and IDL, otherwise t.
0386b551
AM
6797
6798 ;; A colon label must have something before the colon.
6799 (not (eq (char-after) ?:))
6800
6801 ;; Check that we're not after a token that can't precede a label.
6802 (or
6803 ;; Trivially succeeds when there's no preceding token.
6804 (if preceding-token-end
6805 (<= preceding-token-end (point-min))
6806 (save-excursion
6807 (c-backward-syntactic-ws)
6808 (setq preceding-token-end (point))
6809 (bobp)))
6810
6811 ;; Check if we're after a label, if we're after a closing
6812 ;; paren that belong to statement, and with
6813 ;; `c-label-prefix-re'. It's done in different order
6814 ;; depending on `assume-markup' since the checks have
6815 ;; different expensiveness.
6816 (if assume-markup
6817 (or
6818 (eq (c-get-char-property (1- preceding-token-end) 'c-type)
6819 'c-decl-end)
6820
6821 (save-excursion
6822 (goto-char (1- preceding-token-end))
6823 (c-beginning-of-current-token)
51c9af45
AM
6824 (or (looking-at c-label-prefix-re)
6825 (looking-at c-block-stmt-1-key)))
0386b551
AM
6826
6827 (and (eq (char-before preceding-token-end) ?\))
6828 (c-after-conditional)))
6829
6830 (or
6831 (save-excursion
6832 (goto-char (1- preceding-token-end))
6833 (c-beginning-of-current-token)
51c9af45
AM
6834 (or (looking-at c-label-prefix-re)
6835 (looking-at c-block-stmt-1-key)))
0386b551
AM
6836
6837 (cond
6838 ((eq (char-before preceding-token-end) ?\))
6839 (c-after-conditional))
6840
6841 ((eq (char-before preceding-token-end) ?:)
6842 ;; Might be after another label, so check it recursively.
51c9af45
AM
6843 (save-restriction
6844 (save-excursion
6845 (goto-char (1- preceding-token-end))
6846 ;; Essentially the same as the
6847 ;; `c-syntactic-re-search-forward' regexp below.
6848 (setq macro-start
6849 (save-excursion (and (c-beginning-of-macro)
6850 (point))))
6851 (if macro-start (narrow-to-region macro-start (point-max)))
6852 (c-syntactic-skip-backward "^-]:?;}=*/%&|,<>!@+" nil t)
6853 ;; Note: the following should work instead of the
6854 ;; narrow-to-region above. Investigate why not,
6855 ;; sometime. ACM, 2006-03-31.
6856 ;; (c-syntactic-skip-backward "^-]:?;}=*/%&|,<>!@+"
6857 ;; macro-start t)
6858 (let ((pte (point))
6859 ;; If the caller turned on recording for us,
6860 ;; it shouldn't apply when we check the
6861 ;; preceding label.
6862 c-record-type-identifiers)
6863 ;; A label can't start at a cpp directive. Check for
6864 ;; this, since c-forward-syntactic-ws would foul up on it.
6865 (unless (and c-opt-cpp-prefix (looking-at c-opt-cpp-prefix))
6866 (c-forward-syntactic-ws)
6867 (c-forward-label nil pte start))))))))))
6868
1379f2c5 6869 ;; Point is still at the beginning of the possible label construct.
b414f371 6870 ;;
51c9af45
AM
6871 ;; Check that the next nonsymbol token is ":", or that we're in one
6872 ;; of QT's "slots" declarations. Allow '(' for the sake of macro
6873 ;; arguments. FIXME: Should build this regexp from the language
6874 ;; constants.
1379f2c5
AM
6875 (cond
6876 ;; public: protected: private:
6877 ((and
6878 (c-major-mode-is 'c++-mode)
6879 (search-forward-regexp
6880 "\\=p\\(r\\(ivate\\|otected\\)\\|ublic\\)\\>[^_]" nil t)
6881 (progn (backward-char)
6882 (c-forward-syntactic-ws limit)
6883 (looking-at ":\\([^:]\\|\\'\\)"))) ; A single colon.
6884 (forward-char)
6885 (setq label-type t))
6886 ;; QT double keyword like "protected slots:" or goto target.
6887 ((progn (goto-char start) nil))
6888 ((when (c-syntactic-re-search-forward
6889 "[ \t\n[:?;{=*/%&|,<>!@+-]" limit t t) ; not at EOB
6890 (backward-char)
6891 (setq label-end (point))
6892 (setq qt-symbol-idx
6893 (and (c-major-mode-is 'c++-mode)
6894 (string-match
6895 "\\(p\\(r\\(ivate\\|otected\\)\\|ublic\\)\\|more\\)\\>"
b414f371 6896 (buffer-substring start (point)))))
1379f2c5
AM
6897 (c-forward-syntactic-ws limit)
6898 (cond
6899 ((looking-at ":\\([^:]\\|\\'\\)") ; A single colon.
6900 (forward-char)
6901 (setq label-type
f412a567
AM
6902 (if (or (string= "signals" ; Special QT macro
6903 (setq kwd (buffer-substring-no-properties start label-end)))
6904 (string= "Q_SIGNALS" kwd))
1379f2c5
AM
6905 'qt-1kwd-colon
6906 'goto-target)))
6907 ((and qt-symbol-idx
f412a567 6908 (search-forward-regexp "\\=\\(slots\\|Q_SLOTS\\)\\>" limit t)
1379f2c5
AM
6909 (progn (c-forward-syntactic-ws limit)
6910 (looking-at ":\\([^:]\\|\\'\\)"))) ; A single colon
6911 (forward-char)
6912 (setq label-type 'qt-2kwds-colon)))))))
0386b551
AM
6913
6914 (save-restriction
6915 (narrow-to-region start (point))
6916
6917 ;; Check that `c-nonlabel-token-key' doesn't match anywhere.
6918 (catch 'check-label
6919 (goto-char start)
6920 (while (progn
6921 (when (looking-at c-nonlabel-token-key)
6922 (goto-char start)
1379f2c5 6923 (setq label-type nil)
0386b551
AM
6924 (throw 'check-label nil))
6925 (and (c-safe (c-forward-sexp)
6926 (c-forward-syntactic-ws)
6927 t)
6928 (not (eobp)))))
6929
6930 ;; Record the identifiers in the label for fontification, unless
6931 ;; it begins with `c-label-kwds' in which case the following
6932 ;; identifiers are part of a (constant) expression that
6933 ;; shouldn't be fontified.
6934 (when (and c-record-type-identifiers
6935 (progn (goto-char start)
6936 (not (looking-at c-label-kwds-regexp))))
6937 (while (c-syntactic-re-search-forward c-symbol-key nil t)
6938 (c-record-ref-id (cons (match-beginning 0)
6939 (match-end 0)))))
6940
6941 (c-put-c-type-property (1- (point-max)) 'c-decl-end)
82ba65cf 6942 (goto-char (point-max)))))
0386b551
AM
6943
6944 (t
6945 ;; Not a label.
1379f2c5
AM
6946 (goto-char start)))
6947 label-type))
0386b551
AM
6948
6949(defun c-forward-objc-directive ()
6950 ;; Assuming the point is at the beginning of a token, try to move
6951 ;; forward to the end of the Objective-C directive that starts
6952 ;; there. Return t if a directive was fully recognized, otherwise
6953 ;; the point is moved as far as one could be successfully parsed and
6954 ;; nil is returned.
6955 ;;
6956 ;; This function records identifier ranges on
6957 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
6958 ;; `c-record-type-identifiers' is non-nil.
6959 ;;
6960 ;; This function might do hidden buffer changes.
6961
6962 (let ((start (point))
6963 start-char
6964 (c-promote-possible-types t)
6965 ;; Turn off recognition of angle bracket arglists while parsing
6966 ;; types here since the protocol reference list might then be
6967 ;; considered part of the preceding name or superclass-name.
6968 c-recognize-<>-arglists)
6969
6970 (if (or
6971 (when (looking-at
6972 (eval-when-compile
6973 (c-make-keywords-re t
6974 (append (c-lang-const c-protection-kwds objc)
6975 '("@end"))
6976 'objc-mode)))
6977 (goto-char (match-end 1))
6978 t)
6979
6980 (and
6981 (looking-at
6982 (eval-when-compile
6983 (c-make-keywords-re t
6984 '("@interface" "@implementation" "@protocol")
6985 'objc-mode)))
6986
6987 ;; Handle the name of the class itself.
6988 (progn
cb694ab7
AM
6989; (c-forward-token-2) ; 2006/1/13 This doesn't move if the token's
6990; at EOB.
6991 (goto-char (match-end 0))
6992 (c-skip-ws-forward)
0386b551
AM
6993 (c-forward-type))
6994
6995 (catch 'break
6996 ;; Look for ": superclass-name" or "( category-name )".
6997 (when (looking-at "[:\(]")
6998 (setq start-char (char-after))
6999 (forward-char)
7000 (c-forward-syntactic-ws)
7001 (unless (c-forward-type) (throw 'break nil))
7002 (when (eq start-char ?\()
7003 (unless (eq (char-after) ?\)) (throw 'break nil))
7004 (forward-char)
7005 (c-forward-syntactic-ws)))
7006
7007 ;; Look for a protocol reference list.
7008 (if (eq (char-after) ?<)
7009 (let ((c-recognize-<>-arglists t)
7010 (c-parse-and-markup-<>-arglists t)
7011 c-restricted-<>-arglists)
7012 (c-forward-<>-arglist t))
7013 t))))
7014
7015 (progn
7016 (c-backward-syntactic-ws)
7017 (c-clear-c-type-property start (1- (point)) 'c-decl-end)
7018 (c-put-c-type-property (1- (point)) 'c-decl-end)
7019 t)
7020
7021 (c-clear-c-type-property start (point) 'c-decl-end)
7022 nil)))
7023
785eecbb
RS
7024(defun c-beginning-of-inheritance-list (&optional lim)
7025 ;; Go to the first non-whitespace after the colon that starts a
7026 ;; multiple inheritance introduction. Optional LIM is the farthest
7027 ;; back we should search.
0386b551
AM
7028 ;;
7029 ;; This function might do hidden buffer changes.
7030 (c-with-syntax-table c++-template-syntax-table
7031 (c-backward-token-2 0 t lim)
7032 (while (and (or (looking-at c-symbol-start)
7033 (looking-at "[<,]\\|::"))
7034 (zerop (c-backward-token-2 1 t lim))))))
785eecbb 7035
785eecbb
RS
7036(defun c-in-method-def-p ()
7037 ;; Return nil if we aren't in a method definition, otherwise the
7038 ;; position of the initial [+-].
0386b551
AM
7039 ;;
7040 ;; This function might do hidden buffer changes.
785eecbb
RS
7041 (save-excursion
7042 (beginning-of-line)
a66cd3ee
MS
7043 (and c-opt-method-key
7044 (looking-at c-opt-method-key)
785eecbb
RS
7045 (point))
7046 ))
7047
a66cd3ee
MS
7048;; Contributed by Kevin Ryde <user42@zip.com.au>.
7049(defun c-in-gcc-asm-p ()
7050 ;; Return non-nil if point is within a gcc \"asm\" block.
7051 ;;
7052 ;; This should be called with point inside an argument list.
7053 ;;
7054 ;; Only one level of enclosing parentheses is considered, so for
7055 ;; instance `nil' is returned when in a function call within an asm
7056 ;; operand.
0386b551
AM
7057 ;;
7058 ;; This function might do hidden buffer changes.
a66cd3ee
MS
7059
7060 (and c-opt-asm-stmt-key
7061 (save-excursion
7062 (beginning-of-line)
7063 (backward-up-list 1)
7064 (c-beginning-of-statement-1 (point-min) nil t)
7065 (looking-at c-opt-asm-stmt-key))))
7066
abb7e5cf 7067(defun c-at-toplevel-p ()
a85fd6da
AM
7068 "Return a determination as to whether point is \"at the top level\".
7069Informally, \"at the top level\" is anywhere where you can write
7070a function.
7071
7072More precisely, being at the top-level means that point is either
7073outside any enclosing block (such as a function definition), or
7074directly inside a class, namespace or other block that contains
7075another declaration level.
abb7e5cf
SM
7076
7077If point is not at the top-level (e.g. it is inside a method
7078definition), then nil is returned. Otherwise, if point is at a
7079top-level not enclosed within a class definition, t is returned.
7080Otherwise, a 2-vector is returned where the zeroth element is the
7081buffer position of the start of the class declaration, and the first
7082element is the buffer position of the enclosing class's opening
0386b551
AM
7083brace.
7084
7085Note that this function might do hidden buffer changes. See the
7086comment at the start of cc-engine.el for more info."
a66cd3ee
MS
7087 (let ((paren-state (c-parse-state)))
7088 (or (not (c-most-enclosing-brace paren-state))
7089 (c-search-uplist-for-classkey paren-state))))
7090
d9e94c22 7091(defun c-just-after-func-arglist-p (&optional lim)
0386b551
AM
7092 ;; Return non-nil if the point is in the region after the argument
7093 ;; list of a function and its opening brace (or semicolon in case it
7094 ;; got no body). If there are K&R style argument declarations in
7095 ;; that region, the point has to be inside the first one for this
7096 ;; function to recognize it.
a66cd3ee 7097 ;;
0386b551
AM
7098 ;; If successful, the point is moved to the first token after the
7099 ;; function header (see `c-forward-decl-or-cast-1' for details) and
7100 ;; the position of the opening paren of the function arglist is
7101 ;; returned.
7102 ;;
7103 ;; The point is clobbered if not successful.
7104 ;;
7105 ;; LIM is used as bound for backward buffer searches.
7106 ;;
7107 ;; This function might do hidden buffer changes.
7108
7109 (let ((beg (point)) end id-start)
7110 (and
7111 (eq (c-beginning-of-statement-1 lim) 'same)
7112
7113 (not (or (c-major-mode-is 'objc-mode)
7114 (c-forward-objc-directive)))
7115
7116 (setq id-start
7117 (car-safe (c-forward-decl-or-cast-1 (c-point 'bosws) nil nil)))
7118 (< id-start beg)
7119
7120 ;; There should not be a '=' or ',' between beg and the
7121 ;; start of the declaration since that means we were in the
7122 ;; "expression part" of the declaration.
7123 (or (> (point) beg)
7124 (not (looking-at "[=,]")))
7125
7126 (save-excursion
7127 ;; Check that there's an arglist paren in the
7128 ;; declaration.
7129 (goto-char id-start)
7130 (cond ((eq (char-after) ?\()
7131 ;; The declarator is a paren expression, so skip past it
7132 ;; so that we don't get stuck on that instead of the
7133 ;; function arglist.
7134 (c-forward-sexp))
51c9af45
AM
7135 ((and c-opt-op-identifier-prefix
7136 (looking-at c-opt-op-identifier-prefix))
0386b551
AM
7137 ;; Don't trip up on "operator ()".
7138 (c-forward-token-2 2 t)))
7139 (and (< (point) beg)
7140 (c-syntactic-re-search-forward "(" beg t t)
7141 (1- (point)))))))
785eecbb 7142
a66cd3ee
MS
7143(defun c-in-knr-argdecl (&optional lim)
7144 ;; Return the position of the first argument declaration if point is
7145 ;; inside a K&R style argument declaration list, nil otherwise.
7146 ;; `c-recognize-knr-p' is not checked. If LIM is non-nil, it's a
7147 ;; position that bounds the backward search for the argument list.
7148 ;;
9cf17ef1
AM
7149 ;; Point must be within a possible K&R region, e.g. just before a top-level
7150 ;; "{". It must be outside of parens and brackets. The test can return
7151 ;; false positives otherwise.
0386b551
AM
7152 ;;
7153 ;; This function might do hidden buffer changes.
d9e94c22 7154
a66cd3ee
MS
7155 (save-excursion
7156 (save-restriction
9cf17ef1
AM
7157 ;; If we're in a macro, our search range is restricted to it. Narrow to
7158 ;; the searchable range.
7159 (let* ((macro-start (c-query-macro-start))
7160 (lim (max (or lim (point-min)) (or macro-start (point-min))))
6b6481ed
AM
7161 before-lparen after-rparen
7162 (pp-count-out 20)) ; Max number of paren/brace constructs before we give up
9cf17ef1
AM
7163 (narrow-to-region lim (c-point 'eol))
7164
7165 ;; Search backwards for the defun's argument list. We give up if we
7166 ;; encounter a "}" (end of a previous defun) or BOB.
7167 ;;
7168 ;; The criterion for a paren structure being the arg list is:
7169 ;; o - there is non-WS stuff after it but before any "{"; AND
7170 ;; o - the token after it isn't a ";" AND
7171 ;; o - it is preceded by either an identifier (the function name) or
7172 ;; a macro expansion like "DEFUN (...)"; AND
7173 ;; o - its content is a non-empty comma-separated list of identifiers
7174 ;; (an empty arg list won't have a knr region).
7175 ;;
7176 ;; The following snippet illustrates these rules:
7177 ;; int foo (bar, baz, yuk)
7178 ;; int bar [] ;
7179 ;; int (*baz) (my_type) ;
7180 ;; int (*) (void) (*yuk) (void) ;
7181 ;; {
7182
7183 (catch 'knr
6b6481ed
AM
7184 (while (> pp-count-out 0) ; go back one paren/bracket pair each time.
7185 (setq pp-count-out (1- pp-count-out))
10489fcb 7186 (c-syntactic-skip-backward "^)]}")
9cf17ef1
AM
7187 (cond ((eq (char-before) ?\))
7188 (setq after-rparen (point)))
10489fcb
AM
7189 ((eq (char-before) ?\])
7190 (setq after-rparen nil))
7191 (t ; either } (hit previous defun) or no more parens/brackets
7192 (throw 'knr nil)))
9cf17ef1
AM
7193
7194 (if after-rparen
7195 ;; We're inside a paren. Could it be our argument list....?
7196 (if
7197 (and
7198 (progn
7199 (goto-char after-rparen)
7200 (unless (c-go-list-backward) (throw 'knr nil)) ;
7201 ;; FIXME!!! What about macros between the parens? 2007/01/20
7202 (setq before-lparen (point)))
d9e94c22 7203
9cf17ef1
AM
7204 ;; It can't be the arg list if next token is ; or {
7205 (progn (goto-char after-rparen)
7206 (c-forward-syntactic-ws)
7207 (not (memq (char-after) '(?\; ?\{))))
d9e94c22 7208
9cf17ef1
AM
7209 ;; Is the thing preceding the list an identifier (the
7210 ;; function name), or a macro expansion?
7211 (progn
7212 (goto-char before-lparen)
7213 (eq (c-backward-token-2) 0)
7214 (or (c-on-identifier)
7215 (and (eq (char-after) ?\))
7216 (c-go-up-list-backward)
7217 (eq (c-backward-token-2) 0)
7218 (c-on-identifier))))
7219
7220 ;; Have we got a non-empty list of comma-separated
7221 ;; identifiers?
7222 (progn
7223 (goto-char before-lparen)
7224 (c-forward-token-2) ; to first token inside parens
7225 (and
7226 (c-on-identifier)
7227 (c-forward-token-2)
7228 (catch 'id-list
7229 (while (eq (char-after) ?\,)
7230 (c-forward-token-2)
7231 (unless (c-on-identifier) (throw 'id-list nil))
7232 (c-forward-token-2))
7233 (eq (char-after) ?\))))))
7234
7235 ;; ...Yes. We've identified the function's argument list.
7236 (throw 'knr
7237 (progn (goto-char after-rparen)
7238 (c-forward-syntactic-ws)
7239 (point)))
7240
7241 ;; ...No. The current parens aren't the function's arg list.
7242 (goto-char before-lparen))
7243
7244 (or (c-go-list-backward) ; backwards over [ .... ]
7245 (throw 'knr nil)))))))))
785eecbb
RS
7246
7247(defun c-skip-conditional ()
7248 ;; skip forward over conditional at point, including any predicate
7249 ;; statements in parentheses. No error checking is performed.
0386b551
AM
7250 ;;
7251 ;; This function might do hidden buffer changes.
0ec8351b
BW
7252 (c-forward-sexp (cond
7253 ;; else if()
a66cd3ee
MS
7254 ((looking-at (concat "\\<else"
7255 "\\([ \t\n]\\|\\\\\n\\)+"
7256 "if\\>\\([^_]\\|$\\)"))
7257 3)
0ec8351b 7258 ;; do, else, try, finally
a66cd3ee
MS
7259 ((looking-at (concat "\\<\\("
7260 "do\\|else\\|try\\|finally"
7261 "\\)\\>\\([^_]\\|$\\)"))
130c507e 7262 1)
ce8c7486 7263 ;; for, if, while, switch, catch, synchronized, foreach
0ec8351b 7264 (t 2))))
785eecbb 7265
a66cd3ee
MS
7266(defun c-after-conditional (&optional lim)
7267 ;; If looking at the token after a conditional then return the
7268 ;; position of its start, otherwise return nil.
0386b551
AM
7269 ;;
7270 ;; This function might do hidden buffer changes.
a66cd3ee 7271 (save-excursion
d9e94c22 7272 (and (zerop (c-backward-token-2 1 t lim))
a66cd3ee
MS
7273 (or (looking-at c-block-stmt-1-key)
7274 (and (eq (char-after) ?\()
d9e94c22 7275 (zerop (c-backward-token-2 1 t lim))
a66cd3ee
MS
7276 (looking-at c-block-stmt-2-key)))
7277 (point))))
7278
0386b551
AM
7279(defun c-after-special-operator-id (&optional lim)
7280 ;; If the point is after an operator identifier that isn't handled
7281 ;; like an ordinary symbol (i.e. like "operator =" in C++) then the
7282 ;; position of the start of that identifier is returned. nil is
7283 ;; returned otherwise. The point may be anywhere in the syntactic
7284 ;; whitespace after the last token of the operator identifier.
7285 ;;
7286 ;; This function might do hidden buffer changes.
7287 (save-excursion
7288 (and c-overloadable-operators-regexp
7289 (zerop (c-backward-token-2 1 nil lim))
7290 (looking-at c-overloadable-operators-regexp)
51c9af45 7291 (or (not c-opt-op-identifier-prefix)
0386b551
AM
7292 (and
7293 (zerop (c-backward-token-2 1 nil lim))
51c9af45 7294 (looking-at c-opt-op-identifier-prefix)))
0386b551
AM
7295 (point))))
7296
a66cd3ee
MS
7297(defsubst c-backward-to-block-anchor (&optional lim)
7298 ;; Assuming point is at a brace that opens a statement block of some
7299 ;; kind, move to the proper anchor point for that block. It might
7300 ;; need to be adjusted further by c-add-stmt-syntax, but the
7301 ;; position at return is suitable as start position for that
7302 ;; function.
0386b551
AM
7303 ;;
7304 ;; This function might do hidden buffer changes.
a66cd3ee
MS
7305 (unless (= (point) (c-point 'boi))
7306 (let ((start (c-after-conditional lim)))
7307 (if start
7308 (goto-char start)))))
7309
037558bf 7310(defsubst c-backward-to-decl-anchor (&optional lim)
a66cd3ee
MS
7311 ;; Assuming point is at a brace that opens the block of a top level
7312 ;; declaration of some kind, move to the proper anchor point for
7313 ;; that block.
0386b551
AM
7314 ;;
7315 ;; This function might do hidden buffer changes.
a66cd3ee 7316 (unless (= (point) (c-point 'boi))
037558bf 7317 (c-beginning-of-statement-1 lim)))
a66cd3ee 7318
ff959bab 7319(defun c-search-decl-header-end ()
a66cd3ee
MS
7320 ;; Search forward for the end of the "header" of the current
7321 ;; declaration. That's the position where the definition body
7322 ;; starts, or the first variable initializer, or the ending
7323 ;; semicolon. I.e. search forward for the closest following
7324 ;; (syntactically relevant) '{', '=' or ';' token. Point is left
7325 ;; _after_ the first found token, or at point-max if none is found.
0386b551
AM
7326 ;;
7327 ;; This function might do hidden buffer changes.
ff959bab
MS
7328
7329 (let ((base (point)))
7330 (if (c-major-mode-is 'c++-mode)
7331
7332 ;; In C++ we need to take special care to handle operator
7333 ;; tokens and those pesky template brackets.
7334 (while (and
7335 (c-syntactic-re-search-forward "[;{<=]" nil 'move t t)
7336 (or
7337 (c-end-of-current-token base)
7338 ;; Handle operator identifiers, i.e. ignore any
7339 ;; operator token preceded by "operator".
7340 (save-excursion
7341 (and (c-safe (c-backward-sexp) t)
51c9af45 7342 (looking-at c-opt-op-identifier-prefix)))
ff959bab
MS
7343 (and (eq (char-before) ?<)
7344 (c-with-syntax-table c++-template-syntax-table
7345 (if (c-safe (goto-char (c-up-list-forward (point))))
7346 t
7347 (goto-char (point-max))
7348 nil)))))
7349 (setq base (point)))
7350
7351 (while (and
7352 (c-syntactic-re-search-forward "[;{=]" nil 'move t t)
7353 (c-end-of-current-token base))
7354 (setq base (point))))))
a66cd3ee
MS
7355
7356(defun c-beginning-of-decl-1 (&optional lim)
7357 ;; Go to the beginning of the current declaration, or the beginning
7358 ;; of the previous one if already at the start of it. Point won't
0386b551 7359 ;; be moved out of any surrounding paren. Return a cons cell of the
a66cd3ee
MS
7360 ;; form (MOVE . KNR-POS). MOVE is like the return value from
7361 ;; `c-beginning-of-statement-1'. If point skipped over some K&R
7362 ;; style argument declarations (and they are to be recognized) then
7363 ;; KNR-POS is set to the start of the first such argument
7364 ;; declaration, otherwise KNR-POS is nil. If LIM is non-nil, it's a
7365 ;; position that bounds the backward search.
7366 ;;
7367 ;; NB: Cases where the declaration continues after the block, as in
7368 ;; "struct foo { ... } bar;", are currently recognized as two
7369 ;; declarations, e.g. "struct foo { ... }" and "bar;" in this case.
0386b551
AM
7370 ;;
7371 ;; This function might do hidden buffer changes.
a66cd3ee
MS
7372 (catch 'return
7373 (let* ((start (point))
d9e94c22 7374 (last-stmt-start (point))
0386b551 7375 (move (c-beginning-of-statement-1 lim nil t)))
a66cd3ee 7376
a66cd3ee
MS
7377 ;; `c-beginning-of-statement-1' stops at a block start, but we
7378 ;; want to continue if the block doesn't begin a top level
2a15eb73
MS
7379 ;; construct, i.e. if it isn't preceded by ';', '}', ':', bob,
7380 ;; or an open paren.
d9e94c22 7381 (let ((beg (point)) tentative-move)
51c9af45
AM
7382 ;; Go back one "statement" each time round the loop until we're just
7383 ;; after a ;, }, or :, or at BOB or the start of a macro or start of
7384 ;; an ObjC method. This will move over a multiple declaration whose
7385 ;; components are comma separated.
d9e94c22
MS
7386 (while (and
7387 ;; Must check with c-opt-method-key in ObjC mode.
7388 (not (and c-opt-method-key
7389 (looking-at c-opt-method-key)))
7390 (/= last-stmt-start (point))
7391 (progn
7392 (c-backward-syntactic-ws lim)
7393 (not (memq (char-before) '(?\; ?} ?: nil))))
2a15eb73
MS
7394 (save-excursion
7395 (backward-char)
7396 (not (looking-at "\\s(")))
d9e94c22
MS
7397 ;; Check that we don't move from the first thing in a
7398 ;; macro to its header.
7399 (not (eq (setq tentative-move
0386b551 7400 (c-beginning-of-statement-1 lim nil t))
d9e94c22
MS
7401 'macro)))
7402 (setq last-stmt-start beg
7403 beg (point)
7404 move tentative-move))
7405 (goto-char beg))
7406
7407 (when c-recognize-knr-p
7408 (let ((fallback-pos (point)) knr-argdecl-start)
7409 ;; Handle K&R argdecls. Back up after the "statement" jumped
7410 ;; over by `c-beginning-of-statement-1', unless it was the
7411 ;; function body, in which case we're sitting on the opening
7412 ;; brace now. Then test if we're in a K&R argdecl region and
7413 ;; that we started at the other side of the first argdecl in
7414 ;; it.
7415 (unless (eq (char-after) ?{)
7416 (goto-char last-stmt-start))
7417 (if (and (setq knr-argdecl-start (c-in-knr-argdecl lim))
7418 (< knr-argdecl-start start)
7419 (progn
7420 (goto-char knr-argdecl-start)
0386b551 7421 (not (eq (c-beginning-of-statement-1 lim nil t) 'macro))))
d9e94c22
MS
7422 (throw 'return
7423 (cons (if (eq (char-after fallback-pos) ?{)
7424 'previous
7425 'same)
7426 knr-argdecl-start))
7427 (goto-char fallback-pos))))
7428
51c9af45
AM
7429 ;; `c-beginning-of-statement-1' counts each brace block as a separate
7430 ;; statement, so the result will be 'previous if we've moved over any.
7431 ;; So change our result back to 'same if necessary.
7432 ;;
7433 ;; If they were brace list initializers we might not have moved over a
7434 ;; declaration boundary though, so change it to 'same if we've moved
7435 ;; past a '=' before '{', but not ';'. (This ought to be integrated
7436 ;; into `c-beginning-of-statement-1', so we avoid this extra pass which
7437 ;; potentially can search over a large amount of text.). Take special
7438 ;; pains not to get mislead by C++'s "operator=", and the like.
d9e94c22
MS
7439 (if (and (eq move 'previous)
7440 (c-with-syntax-table (if (c-major-mode-is 'c++-mode)
7441 c++-template-syntax-table
7442 (syntax-table))
7443 (save-excursion
51c9af45
AM
7444 (and
7445 (progn
7446 (while ; keep going back to "[;={"s until we either find
7447 ; no more, or get to one which isn't an "operator ="
7448 (and (c-syntactic-re-search-forward "[;={]" start t t t)
7449 (eq (char-before) ?=)
7450 c-overloadable-operators-regexp
7451 c-opt-op-identifier-prefix
7452 (save-excursion
7453 (eq (c-backward-token-2) 0)
7454 (looking-at c-overloadable-operators-regexp)
7455 (eq (c-backward-token-2) 0)
7456 (looking-at c-opt-op-identifier-prefix))))
7457 (eq (char-before) ?=))
7458 (c-syntactic-re-search-forward "[;{]" start t t)
7459 (eq (char-before) ?{)
7460 (c-safe (goto-char (c-up-list-forward (point))) t)
7461 (not (c-syntactic-re-search-forward ";" start t t))))))
d9e94c22
MS
7462 (cons 'same nil)
7463 (cons move nil)))))
a66cd3ee
MS
7464
7465(defun c-end-of-decl-1 ()
7466 ;; Assuming point is at the start of a declaration (as detected by
7467 ;; e.g. `c-beginning-of-decl-1'), go to the end of it. Unlike
7468 ;; `c-beginning-of-decl-1', this function handles the case when a
7469 ;; block is followed by identifiers in e.g. struct declarations in C
7470 ;; or C++. If a proper end was found then t is returned, otherwise
7471 ;; point is moved as far as possible within the current sexp and nil
7472 ;; is returned. This function doesn't handle macros; use
7473 ;; `c-end-of-macro' instead in those cases.
0386b551
AM
7474 ;;
7475 ;; This function might do hidden buffer changes.
ce8c7486 7476 (let ((start (point))
a66cd3ee
MS
7477 (decl-syntax-table (if (c-major-mode-is 'c++-mode)
7478 c++-template-syntax-table
7479 (syntax-table))))
7480 (catch 'return
7481 (c-search-decl-header-end)
7482
7483 (when (and c-recognize-knr-p
7484 (eq (char-before) ?\;)
7485 (c-in-knr-argdecl start))
7486 ;; Stopped at the ';' in a K&R argdecl section which is
7487 ;; detected using the same criteria as in
7488 ;; `c-beginning-of-decl-1'. Move to the following block
7489 ;; start.
d9e94c22 7490 (c-syntactic-re-search-forward "{" nil 'move t))
a66cd3ee
MS
7491
7492 (when (eq (char-before) ?{)
7493 ;; Encountered a block in the declaration. Jump over it.
7494 (condition-case nil
7495 (goto-char (c-up-list-forward (point)))
d9e94c22
MS
7496 (error (goto-char (point-max))
7497 (throw 'return nil)))
a66cd3ee
MS
7498 (if (or (not c-opt-block-decls-with-vars-key)
7499 (save-excursion
7500 (c-with-syntax-table decl-syntax-table
7501 (let ((lim (point)))
7502 (goto-char start)
b3cf7e18
MS
7503 (not (and
7504 ;; Check for `c-opt-block-decls-with-vars-key'
7505 ;; before the first paren.
7506 (c-syntactic-re-search-forward
d9e94c22 7507 (concat "[;=\(\[{]\\|\\("
b3cf7e18
MS
7508 c-opt-block-decls-with-vars-key
7509 "\\)")
d9e94c22 7510 lim t t t)
b3cf7e18
MS
7511 (match-beginning 1)
7512 (not (eq (char-before) ?_))
d9e94c22
MS
7513 ;; Check that the first following paren is
7514 ;; the block.
7515 (c-syntactic-re-search-forward "[;=\(\[{]"
7516 lim t t t)
b3cf7e18 7517 (eq (char-before) ?{)))))))
a66cd3ee
MS
7518 ;; The declaration doesn't have any of the
7519 ;; `c-opt-block-decls-with-vars' keywords in the
7520 ;; beginning, so it ends here at the end of the block.
7521 (throw 'return t)))
7522
7523 (c-with-syntax-table decl-syntax-table
7524 (while (progn
7525 (if (eq (char-before) ?\;)
7526 (throw 'return t))
d9e94c22 7527 (c-syntactic-re-search-forward ";" nil 'move t))))
a66cd3ee 7528 nil)))
ce8c7486 7529
0386b551
AM
7530(defun c-looking-at-decl-block (containing-sexp goto-start &optional limit)
7531 ;; Assuming the point is at an open brace, check if it starts a
7532 ;; block that contains another declaration level, i.e. that isn't a
7533 ;; statement block or a brace list, and if so return non-nil.
7534 ;;
7535 ;; If the check is successful, the return value is the start of the
7536 ;; keyword that tells what kind of construct it is, i.e. typically
7537 ;; what `c-decl-block-key' matched. Also, if GOTO-START is set then
7538 ;; the point will be at the start of the construct, before any
7539 ;; leading specifiers, otherwise it's at the returned position.
7540 ;;
7541 ;; The point is clobbered if the check is unsuccessful.
7542 ;;
7543 ;; CONTAINING-SEXP is the position of the open of the surrounding
7544 ;; paren, or nil if none.
7545 ;;
7546 ;; The optional LIMIT limits the backward search for the start of
7547 ;; the construct. It's assumed to be at a syntactically relevant
7548 ;; position.
7549 ;;
7550 ;; If any template arglists are found in the searched region before
7551 ;; the open brace, they get marked with paren syntax.
7552 ;;
7553 ;; This function might do hidden buffer changes.
7554
7555 (let ((open-brace (point)) kwd-start first-specifier-pos)
7556 (c-syntactic-skip-backward c-block-prefix-charset limit t)
7557
7558 (when (and c-recognize-<>-arglists
7559 (eq (char-before) ?>))
7560 ;; Could be at the end of a template arglist.
7561 (let ((c-parse-and-markup-<>-arglists t)
7562 (c-disallow-comma-in-<>-arglists
7563 (and containing-sexp
7564 (not (eq (char-after containing-sexp) ?{)))))
7565 (while (and
7566 (c-backward-<>-arglist nil limit)
7567 (progn
7568 (c-syntactic-skip-backward c-block-prefix-charset limit t)
7569 (eq (char-before) ?>))))))
7570
7571 ;; Note: Can't get bogus hits inside template arglists below since they
7572 ;; have gotten paren syntax above.
7573 (when (and
7574 ;; If `goto-start' is set we begin by searching for the
7575 ;; first possible position of a leading specifier list.
7576 ;; The `c-decl-block-key' search continues from there since
7577 ;; we know it can't match earlier.
7578 (if goto-start
7579 (when (c-syntactic-re-search-forward c-symbol-start
7580 open-brace t t)
7581 (goto-char (setq first-specifier-pos (match-beginning 0)))
7582 t)
7583 t)
7584
7585 (cond
7586 ((c-syntactic-re-search-forward c-decl-block-key open-brace t t t)
7587 (goto-char (setq kwd-start (match-beginning 0)))
7588 (or
7589
7590 ;; Found a keyword that can't be a type?
7591 (match-beginning 1)
7592
7593 ;; Can be a type too, in which case it's the return type of a
7594 ;; function (under the assumption that no declaration level
7595 ;; block construct starts with a type).
7596 (not (c-forward-type))
7597
7598 ;; Jumped over a type, but it could be a declaration keyword
7599 ;; followed by the declared identifier that we've jumped over
7600 ;; instead (e.g. in "class Foo {"). If it indeed is a type
7601 ;; then we should be at the declarator now, so check for a
7602 ;; valid declarator start.
7603 ;;
7604 ;; Note: This doesn't cope with the case when a declared
7605 ;; identifier is followed by e.g. '(' in a language where '('
7606 ;; also might be part of a declarator expression. Currently
7607 ;; there's no such language.
7608 (not (or (looking-at c-symbol-start)
7609 (looking-at c-type-decl-prefix-key)))))
7610
7611 ;; In Pike a list of modifiers may be followed by a brace
7612 ;; to make them apply to many identifiers. Note that the
7613 ;; match data will be empty on return in this case.
7614 ((and (c-major-mode-is 'pike-mode)
7615 (progn
7616 (goto-char open-brace)
7617 (= (c-backward-token-2) 0))
7618 (looking-at c-specifier-key)
7619 ;; Use this variant to avoid yet another special regexp.
7620 (c-keyword-member (c-keyword-sym (match-string 1))
7621 'c-modifier-kwds))
7622 (setq kwd-start (point))
7623 t)))
7624
7625 ;; Got a match.
7626
7627 (if goto-start
7628 ;; Back up over any preceding specifiers and their clauses
7629 ;; by going forward from `first-specifier-pos', which is the
7630 ;; earliest possible position where the specifier list can
7631 ;; start.
7632 (progn
7633 (goto-char first-specifier-pos)
7634
7635 (while (< (point) kwd-start)
7636 (if (looking-at c-symbol-key)
7637 ;; Accept any plain symbol token on the ground that
7638 ;; it's a specifier masked through a macro (just
7639 ;; like `c-forward-decl-or-cast-1' skip forward over
7640 ;; such tokens).
7641 ;;
7642 ;; Could be more restrictive wrt invalid keywords,
7643 ;; but that'd only occur in invalid code so there's
7644 ;; no use spending effort on it.
7645 (let ((end (match-end 0)))
7646 (unless (c-forward-keyword-clause 0)
7647 (goto-char end)
7648 (c-forward-syntactic-ws)))
7649
7650 ;; Can't parse a declaration preamble and is still
7651 ;; before `kwd-start'. That means `first-specifier-pos'
7652 ;; was in some earlier construct. Search again.
7653 (if (c-syntactic-re-search-forward c-symbol-start
7654 kwd-start 'move t)
7655 (goto-char (setq first-specifier-pos (match-beginning 0)))
7656 ;; Got no preamble before the block declaration keyword.
7657 (setq first-specifier-pos kwd-start))))
7658
7659 (goto-char first-specifier-pos))
7660 (goto-char kwd-start))
7661
7662 kwd-start)))
ce8c7486 7663
a66cd3ee 7664(defun c-search-uplist-for-classkey (paren-state)
0386b551
AM
7665 ;; Check if the closest containing paren sexp is a declaration
7666 ;; block, returning a 2 element vector in that case. Aref 0
7667 ;; contains the bufpos at boi of the class key line, and aref 1
7668 ;; contains the bufpos of the open brace. This function is an
7669 ;; obsolete wrapper for `c-looking-at-decl-block'.
7670 ;;
7671 ;; This function might do hidden buffer changes.
7672 (let ((open-paren-pos (c-most-enclosing-brace paren-state)))
7673 (when open-paren-pos
7674 (save-excursion
7675 (goto-char open-paren-pos)
7676 (when (and (eq (char-after) ?{)
7677 (c-looking-at-decl-block
7678 (c-safe-position open-paren-pos paren-state)
7679 nil))
7680 (back-to-indentation)
7681 (vector (point) open-paren-pos))))))
785eecbb 7682
a66cd3ee 7683(defun c-inside-bracelist-p (containing-sexp paren-state)
785eecbb
RS
7684 ;; return the buffer position of the beginning of the brace list
7685 ;; statement if we're inside a brace list, otherwise return nil.
7686 ;; CONTAINING-SEXP is the buffer pos of the innermost containing
0386b551 7687 ;; paren. PAREN-STATE is the remainder of the state of enclosing
130c507e 7688 ;; braces
785eecbb
RS
7689 ;;
7690 ;; N.B.: This algorithm can potentially get confused by cpp macros
0386b551 7691 ;; placed in inconvenient locations. It's a trade-off we make for
785eecbb 7692 ;; speed.
0386b551
AM
7693 ;;
7694 ;; This function might do hidden buffer changes.
785eecbb 7695 (or
d9e94c22 7696 ;; This will pick up brace list declarations.
b2acd789
RS
7697 (c-safe
7698 (save-excursion
7699 (goto-char containing-sexp)
0ec8351b 7700 (c-forward-sexp -1)
b2acd789 7701 (let (bracepos)
d9e94c22 7702 (if (and (or (looking-at c-brace-list-key)
0ec8351b 7703 (progn (c-forward-sexp -1)
d9e94c22 7704 (looking-at c-brace-list-key)))
a66cd3ee 7705 (setq bracepos (c-down-list-forward (point)))
b2acd789
RS
7706 (not (c-crosses-statement-barrier-p (point)
7707 (- bracepos 2))))
7708 (point)))))
785eecbb
RS
7709 ;; this will pick up array/aggregate init lists, even if they are nested.
7710 (save-excursion
0ec8351b
BW
7711 (let ((class-key
7712 ;; Pike can have class definitions anywhere, so we must
7713 ;; check for the class key here.
7714 (and (c-major-mode-is 'pike-mode)
a66cd3ee
MS
7715 c-decl-block-key))
7716 bufpos braceassignp lim next-containing)
785eecbb
RS
7717 (while (and (not bufpos)
7718 containing-sexp)
a66cd3ee
MS
7719 (when paren-state
7720 (if (consp (car paren-state))
7721 (setq lim (cdr (car paren-state))
7722 paren-state (cdr paren-state))
7723 (setq lim (car paren-state)))
7724 (when paren-state
7725 (setq next-containing (car paren-state)
7726 paren-state (cdr paren-state))))
785eecbb 7727 (goto-char containing-sexp)
a66cd3ee
MS
7728 (if (c-looking-at-inexpr-block next-containing next-containing)
7729 ;; We're in an in-expression block of some kind. Do not
7730 ;; check nesting. We deliberately set the limit to the
7731 ;; containing sexp, so that c-looking-at-inexpr-block
7732 ;; doesn't check for an identifier before it.
0ec8351b
BW
7733 (setq containing-sexp nil)
7734 ;; see if the open brace is preceded by = or [...] in
7735 ;; this statement, but watch out for operator=
a66cd3ee 7736 (setq braceassignp 'dontknow)
d9e94c22 7737 (c-backward-token-2 1 t lim)
6393fef2 7738 ;; Checks to do only on the first sexp before the brace.
d9e94c22 7739 (when (and c-opt-inexpr-brace-list-key
6393fef2
RS
7740 (eq (char-after) ?\[))
7741 ;; In Java, an initialization brace list may follow
7742 ;; directly after "new Foo[]", so check for a "new"
7743 ;; earlier.
7744 (while (eq braceassignp 'dontknow)
7745 (setq braceassignp
d9e94c22
MS
7746 (cond ((/= (c-backward-token-2 1 t lim) 0) nil)
7747 ((looking-at c-opt-inexpr-brace-list-key) t)
6393fef2
RS
7748 ((looking-at "\\sw\\|\\s_\\|[.[]")
7749 ;; Carry on looking if this is an
7750 ;; identifier (may contain "." in Java)
7751 ;; or another "[]" sexp.
7752 'dontknow)
7753 (t nil)))))
7754 ;; Checks to do on all sexps before the brace, up to the
7755 ;; beginning of the statement.
7756 (while (eq braceassignp 'dontknow)
0ec8351b
BW
7757 (cond ((eq (char-after) ?\;)
7758 (setq braceassignp nil))
7759 ((and class-key
7760 (looking-at class-key))
7761 (setq braceassignp nil))
7762 ((eq (char-after) ?=)
7763 ;; We've seen a =, but must check earlier tokens so
7764 ;; that it isn't something that should be ignored.
7765 (setq braceassignp 'maybe)
7766 (while (and (eq braceassignp 'maybe)
d9e94c22 7767 (zerop (c-backward-token-2 1 t lim)))
0ec8351b
BW
7768 (setq braceassignp
7769 (cond
7770 ;; Check for operator =
51c9af45
AM
7771 ((and c-opt-op-identifier-prefix
7772 (looking-at c-opt-op-identifier-prefix))
0386b551 7773 nil)
130c507e
GM
7774 ;; Check for `<opchar>= in Pike.
7775 ((and (c-major-mode-is 'pike-mode)
7776 (or (eq (char-after) ?`)
7777 ;; Special case for Pikes
7778 ;; `[]=, since '[' is not in
7779 ;; the punctuation class.
7780 (and (eq (char-after) ?\[)
7781 (eq (char-before) ?`))))
7782 nil)
0ec8351b
BW
7783 ((looking-at "\\s.") 'maybe)
7784 ;; make sure we're not in a C++ template
7785 ;; argument assignment
a66cd3ee
MS
7786 ((and
7787 (c-major-mode-is 'c++-mode)
7788 (save-excursion
7789 (let ((here (point))
7790 (pos< (progn
7791 (skip-chars-backward "^<>")
7792 (point))))
7793 (and (eq (char-before) ?<)
7794 (not (c-crosses-statement-barrier-p
7795 pos< here))
7796 (not (c-in-literal))
7797 ))))
0ec8351b 7798 nil)
6393fef2
RS
7799 (t t))))))
7800 (if (and (eq braceassignp 'dontknow)
d9e94c22 7801 (/= (c-backward-token-2 1 t lim) 0))
6393fef2
RS
7802 (setq braceassignp nil)))
7803 (if (not braceassignp)
0ec8351b
BW
7804 (if (eq (char-after) ?\;)
7805 ;; Brace lists can't contain a semicolon, so we're done.
7806 (setq containing-sexp nil)
a66cd3ee
MS
7807 ;; Go up one level.
7808 (setq containing-sexp next-containing
7809 lim nil
7810 next-containing nil))
0ec8351b
BW
7811 ;; we've hit the beginning of the aggregate list
7812 (c-beginning-of-statement-1
a66cd3ee 7813 (c-most-enclosing-brace paren-state))
0ec8351b 7814 (setq bufpos (point))))
a66cd3ee 7815 )
785eecbb
RS
7816 bufpos))
7817 ))
7818
0ec8351b
BW
7819(defun c-looking-at-special-brace-list (&optional lim)
7820