from trunk
[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 16576.
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.
dd21b621
AM
2644 (setq pos (if (and good-pos-actual-macro-end
2645 (> in-macro-start good-pos-actual-macro-start))
0ec1d2c5
AM
2646 (1+ good-pos-actual-macro-end) ; get outside the macro as
2647 ; marked by a `category' text property.
2648 good-pos))
2649 (goto-char pos)
2650 (while (and c-state-cache
2651 (< (point) (point-max)))
2652 (cond
2653 ((null pps-state) ; first time through
2654 (setq target-depth -1))
2655 ((eq (car pps-state) target-depth) ; found closing ),},]
2656 (setq target-depth (1- (car pps-state))))
2657 ;; Do nothing when we've merely reached pps-point.
2658 )
2659
2660 ;; Scan!
2661 (setq pps-state
2662 (parse-partial-sexp
2663 (point) (if (< (point) pps-point) pps-point (point-max))
2664 target-depth
2665 nil pps-state))
2666
2667 (if (= (point) pps-point)
2668 (setq pps-point-state pps-state))
2669
2670 (when (eq (car pps-state) target-depth)
2671 (setq pos (point)) ; POS is now just after an R-paren/brace.
2672 (cond
2673 ((and (consp (car c-state-cache))
2674 (eq (point) (cdar c-state-cache)))
2675 ;; We've just moved out of the paren pair containing the brace-pair
2676 ;; at (car c-state-cache). `pair-beg' is where the open paren is,
2677 ;; and is potentially where the open brace of a cons in
2678 ;; c-state-cache will be.
2679 (setq pair-beg (car-safe (cdr c-state-cache))
2680 c-state-cache (cdr-safe (cdr c-state-cache)))) ; remove {}pair + containing Lparen.
2681 ((numberp (car c-state-cache))
2682 (setq pair-beg (car c-state-cache)
2683 c-state-cache (cdr c-state-cache))) ; remove this
2684 ; containing Lparen
2685 ((numberp (cadr c-state-cache))
2686 (setq pair-beg (cadr c-state-cache)
2687 c-state-cache (cddr c-state-cache))) ; Remove a paren pair
2688 ; together with enclosed brace pair.
2689 ;; (t nil) ; Ignore an unmated Rparen.
2690 )))
2691
2692 (if (< (point) pps-point)
2693 (setq pps-state (parse-partial-sexp (point) pps-point
2694 nil nil ; TARGETDEPTH, STOPBEFORE
2695 pps-state)))
2696
2697 ;; If the last paren pair we moved out of was actually a brace pair,
2698 ;; insert it into `c-state-cache'.
2699 (when (and pair-beg (eq (char-after pair-beg) ?{))
2700 (if (consp (car-safe c-state-cache))
2701 (setq c-state-cache (cdr c-state-cache)))
2702 (setq c-state-cache (cons (cons pair-beg pos)
2703 c-state-cache)))
2704
2705 (list pos scan-back-pos pps-state)))))
2706
2707(defun c-remove-stale-state-cache-backwards (here cache-pos)
2708 ;; Strip stale elements of `c-state-cache' by moving backwards through the
2709 ;; buffer, and inform the caller of the scenario detected.
2710 ;;
2711 ;; HERE is the position we're setting `c-state-cache' for.
2712 ;; CACHE-POS is just after the latest recorded position in `c-state-cache'
2713 ;; before HERE, or a position at or near point-min which isn't in a
2714 ;; literal.
2715 ;;
2716 ;; This function must only be called only when (> `c-state-cache-good-pos'
2717 ;; HERE). Usually the gap between CACHE-POS and HERE is large. It is thus
2718 ;; optimised to eliminate (or minimise) scanning between these two
2719 ;; positions.
2720 ;;
2721 ;; Return a three element list (GOOD-POS SCAN-BACK-POS FWD-FLAG), where:
2722 ;; o - GOOD-POS is a "good position", where `c-state-cache' is valid, or
2723 ;; could become so after missing elements are inserted into
2724 ;; `c-state-cache'. This is JUST AFTER an opening or closing
2725 ;; brace/paren/bracket which is already in `c-state-cache' or just before
2726 ;; one otherwise. exceptionally (when there's no such b/p/b handy) the BOL
2727 ;; before `here''s line, or the start of the literal containing it.
2728 ;; o - SCAN-BACK-POS, if non-nil, indicates there may be a brace pair
2729 ;; preceding POS which isn't recorded in `c-state-cache'. It is a position
2730 ;; to scan backwards from.
2731 ;; o - FWD-FLAG, if non-nil, indicates there may be parens/braces between
2732 ;; POS and HERE which aren't recorded in `c-state-cache'.
2733 ;;
2734 ;; The comments in this defun use "paren" to mean parenthesis or square
2735 ;; bracket (as contrasted with a brace), and "(" and ")" likewise.
2736 ;;
2737 ;; . {..} (..) (..) ( .. { } ) (...) ( .... . ..)
2738 ;; | | | | | |
2739 ;; CP E here D C good
2740 (let ((pos c-state-cache-good-pos)
2741 pa ren ; positions of "(" and ")"
2742 dropped-cons ; whether the last element dropped from `c-state-cache'
2743 ; was a cons (representing a brace-pair)
2744 good-pos ; see above.
2745 lit ; (START . END) of a literal containing some point.
2746 here-lit-start here-lit-end ; bounds of literal containing `here'
2747 ; or `here' itself.
2748 (here-bol (c-point 'bol here))
2749 (too-far-back (max (- here c-state-cache-too-far) 1)))
2750
2751 ;; Remove completely irrelevant entries from `c-state-cache'.
2752 (while (and c-state-cache
2753 (>= (setq pa (c-state-cache-top-lparen)) here))
2754 (setq dropped-cons (consp (car c-state-cache)))
2755 (setq c-state-cache (cdr c-state-cache))
2756 (setq pos pa))
2757 ;; At this stage, (> pos here);
2758 ;; (< (c-state-cache-top-lparen) here) (or is nil).
2759
2760 ;; CASE 1: The top of the cache is a brace pair which now encloses `here'.
2761 ;; As good-pos, return the address. of the "{".
2762 (if (and (consp (car c-state-cache))
2763 (> (cdar c-state-cache) here))
2764 ;; Since we've no knowledge of what's inside these braces, we have no
2765 ;; alternative but to direct the caller to scan the buffer from the
2766 ;; opening brace.
2767 (progn
2768 (setq pos (caar c-state-cache))
2769 (setcar c-state-cache pos)
2770 (list (1+ pos) pos t)) ; return value. We've just converted a brace
2771 ; pair entry into a { entry, so the caller
2772 ; needs to search for a brace pair before the
2773 ; {.
2774
2775 ;; ;; `here' might be inside a literal. Check for this.
2776 (setq lit (c-state-literal-at here)
2777 here-lit-start (or (car lit) here)
2778 here-lit-end (or (cdr lit) here))
2779
2780 ;; `here' might be nested inside any depth of parens (or brackets but
2781 ;; not braces). Scan backwards to find the outermost such opening
2782 ;; paren, if there is one. This will be the scan position to return.
2783 (save-restriction
2784 (narrow-to-region cache-pos (point-max))
2785 (setq pos (c-state-balance-parens-backwards here-lit-end pos)))
2786
2787 (if (< pos here-lit-start)
2788 ;; CASE 2: Address of outermost ( or [ which now encloses `here',
2789 ;; but didn't enclose the (previous) `c-state-cache-good-pos'. If
2790 ;; there is a brace pair preceding this, it will already be in
2791 ;; `c-state-cache', unless there was a brace pair after it,
2792 ;; i.e. there'll only be one to scan for if we've just deleted one.
2793 (list pos (and dropped-cons pos) t) ; Return value.
2794
2795 ;; `here' isn't enclosed in a (previously unrecorded) bracket/paren.
2796 ;; Further forward scanning isn't needed, but we still need to find a
2797 ;; GOOD-POS. Step out of all enclosing "("s on HERE's line.
2798 (save-restriction
2799 (narrow-to-region here-bol (point-max))
2800 (setq pos here-lit-start)
2801 (c-safe (while (setq pa (scan-lists pos -1 1))
2802 (setq pos pa)))) ; might signal
2803 (if (setq ren (c-safe-scan-lists pos -1 -1 too-far-back))
2804 ;; CASE 3: After a }/)/] before `here''s BOL.
2805 (list (1+ ren) (and dropped-cons pos) nil) ; Return value
2806
2807 ;; CASE 4; Best of a bad job: BOL before `here-bol', or beginning of
2808 ;; literal containing it.
2809 (setq good-pos (c-state-lit-beg (c-point 'bopl here-bol)))
2810 (list good-pos (and dropped-cons good-pos) nil))))))
2811
2812
2813;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2814;; Externally visible routines.
2815
2816(defun c-state-cache-init ()
2817 (setq c-state-cache nil
2818 c-state-cache-good-pos 1
2819 c-state-nonlit-pos-cache nil
2820 c-state-nonlit-pos-cache-limit 1
2821 c-state-brace-pair-desert nil
2822 c-state-point-min 1
2823 c-state-point-min-lit-type nil
2824 c-state-point-min-lit-start nil
2825 c-state-min-scan-pos 1
2826 c-state-old-cpp-beg nil
2827 c-state-old-cpp-end nil)
2828 (c-state-mark-point-min-literal))
2829
2830(defun c-invalidate-state-cache-1 (here)
2831 ;; Invalidate all info on `c-state-cache' that applies to the buffer at HERE
2832 ;; or higher and set `c-state-cache-good-pos' accordingly. The cache is
2833 ;; left in a consistent state.
2834 ;;
2835 ;; This is much like `c-whack-state-after', but it never changes a paren
2836 ;; pair element into an open paren element. Doing that would mean that the
2837 ;; new open paren wouldn't have the required preceding paren pair element.
785eecbb 2838 ;;
0ec1d2c5
AM
2839 ;; This function is called from c-after-change.
2840
2841 ;; The cache of non-literals:
2842 (if (< here c-state-nonlit-pos-cache-limit)
2843 (setq c-state-nonlit-pos-cache-limit here))
2844
2845 ;; `c-state-cache':
2846 ;; Case 1: if `here' is in a literal containing point-min, everything
2847 ;; becomes (or is already) nil.
2848 (if (or (null c-state-cache-good-pos)
2849 (< here (c-state-get-min-scan-pos)))
2850 (setq c-state-cache nil
2851 c-state-cache-good-pos nil
2852 c-state-min-scan-pos nil)
2853
2854;;; Truncate `c-state-cache' and set `c-state-cache-good-pos' to a value below
2855;;; `here'. To maintain its consistency, we may need to insert a new brace
2856;;; pair.
2857 (let ((here-bol (c-point 'bol here))
2858 too-high-pa ; recorded {/(/[ next above here, or nil.
2859 dropped-cons ; was the last removed element a brace pair?
2860 pa)
2861 ;; The easy bit - knock over-the-top bits off `c-state-cache'.
2862 (while (and c-state-cache
2863 (>= (setq pa (c-state-cache-top-paren)) here))
2864 (setq dropped-cons (consp (car c-state-cache))
2865 too-high-pa (c-state-cache-top-lparen)
2866 c-state-cache (cdr c-state-cache)))
2867
2868 ;; Do we need to add in an earlier brace pair, having lopped one off?
2869 (if (and dropped-cons
2870 (< too-high-pa (+ here c-state-cache-too-far)))
2871 (c-append-lower-brace-pair-to-state-cache too-high-pa here-bol))
2872 (setq c-state-cache-good-pos (or (c-state-cache-after-top-paren)
2873 (c-state-get-min-scan-pos)))))
2874
2875 ;; The brace-pair desert marker:
2876 (when (car c-state-brace-pair-desert)
2877 (if (< here (car c-state-brace-pair-desert))
2878 (setq c-state-brace-pair-desert nil)
2879 (if (< here (cdr c-state-brace-pair-desert))
2880 (setcdr c-state-brace-pair-desert here)))))
2881
2882(defun c-parse-state-1 ()
2883 ;; Find and record all noteworthy parens between some good point earlier in
2884 ;; the file and point. That good point is at least the beginning of the
2885 ;; top-level construct we are in, or the beginning of the preceding
2886 ;; top-level construct if we aren't in one.
2887 ;;
2888 ;; The returned value is a list of the noteworthy parens with the last one
2889 ;; first. If an element in the list is an integer, it's the position of an
2890 ;; open paren (of any type) which has not been closed before the point. If
2891 ;; an element is a cons, it gives the position of a closed BRACE paren
2892 ;; pair[*]; the car is the start brace position and the cdr is the position
2893 ;; following the closing brace. Only the last closed brace paren pair
2894 ;; before each open paren and before the point is recorded, and thus the
2895 ;; state never contains two cons elements in succession. When a close brace
2896 ;; has no matching open brace (e.g., the matching brace is outside the
2897 ;; visible region), it is not represented in the returned value.
2898 ;;
2899 ;; [*] N.B. The close "brace" might be a mismatching close bracket or paren.
2900 ;; This defun explicitly treats mismatching parens/braces/brackets as
2901 ;; matching. It is the open brace which makes it a "brace" pair.
2902 ;;
2903 ;; If POINT is within a macro, open parens and brace pairs within
2904 ;; THIS macro MIGHT be recorded. This depends on whether their
2905 ;; syntactic properties have been suppressed by
2906 ;; `c-neutralize-syntax-in-CPP'. This might need fixing (2008-12-11).
d9e94c22
MS
2907 ;;
2908 ;; Currently no characters which are given paren syntax with the
2909 ;; syntax-table property are recorded, i.e. angle bracket arglist
2910 ;; parens are never present here. Note that this might change.
2911 ;;
0386b551 2912 ;; BUG: This function doesn't cope entirely well with unbalanced
0ec1d2c5
AM
2913 ;; parens in macros. (2008-12-11: this has probably been resolved
2914 ;; by the function `c-neutralize-syntax-in-CPP'.) E.g. in the
2915 ;; following case the brace before the macro isn't balanced with the
2916 ;; one after it:
0386b551
AM
2917 ;;
2918 ;; {
2919 ;; #define X {
2920 ;; }
2921 ;;
0ec1d2c5
AM
2922 ;; Note to maintainers: this function DOES get called with point
2923 ;; within comments and strings, so don't assume it doesn't!
2924 ;;
0386b551 2925 ;; This function might do hidden buffer changes.
0ec1d2c5
AM
2926 (let* ((here (point))
2927 (here-bopl (c-point 'bopl))
2928 strategy ; 'forward, 'backward etc..
2929 ;; Candidate positions to start scanning from:
2930 cache-pos ; highest position below HERE already existing in
2931 ; cache (or 1).
2932 good-pos
2933 start-point
2934 bopl-state
2935 res
2936 scan-backward-pos scan-forward-p) ; used for 'backward.
2937 ;; If POINT-MIN has changed, adjust the cache
2938 (unless (= (point-min) c-state-point-min)
2939 (c-renarrow-state-cache))
2940
2941 ;; Strategy?
2942 (setq res (c-parse-state-get-strategy here c-state-cache-good-pos)
2943 strategy (car res)
2944 cache-pos (cadr res)
2945 start-point (nth 2 res))
2946
2947 (when (eq strategy 'BOD)
2948 (setq c-state-cache nil
2949 c-state-cache-good-pos start-point))
2950
2951 ;; SCAN!
2952 (save-restriction
2953 (cond
2954 ((memq strategy '(forward BOD))
0386b551 2955 (narrow-to-region (point-min) here)
0ec1d2c5
AM
2956 (setq res (c-remove-stale-state-cache start-point here-bopl))
2957 (setq cache-pos (car res)
2958 scan-backward-pos (cadr res)
2959 bopl-state (car (cddr res))) ; will be nil if (< here-bopl
2960 ; start-point)
2961 (if scan-backward-pos
2962 (c-append-lower-brace-pair-to-state-cache scan-backward-pos))
2963 (setq good-pos
2964 (c-append-to-state-cache cache-pos))
2965 (setq c-state-cache-good-pos
2966 (if (and bopl-state
2967 (< good-pos (- here c-state-cache-too-far)))
2968 (c-state-cache-non-literal-place here-bopl bopl-state)
2969 good-pos)))
2970
2971 ((eq strategy 'backward)
2972 (setq res (c-remove-stale-state-cache-backwards here cache-pos)
2973 good-pos (car res)
2974 scan-backward-pos (cadr res)
2975 scan-forward-p (car (cddr res)))
2976 (if scan-backward-pos
2977 (c-append-lower-brace-pair-to-state-cache
2978 scan-backward-pos))
2979 (setq c-state-cache-good-pos
2980 (if scan-forward-p
2981 (progn (narrow-to-region (point-min) here)
2982 (c-append-to-state-cache good-pos))
2983
2984 (c-get-cache-scan-pos good-pos))))
2985
2986 (t ; (eq strategy 'IN-LIT)
2987 (setq c-state-cache nil
2988 c-state-cache-good-pos nil)))))
2989
2990 c-state-cache)
2991
2992(defun c-invalidate-state-cache (here)
2993 ;; This is a wrapper over `c-invalidate-state-cache-1'.
2994 ;;
2995 ;; It suppresses the syntactic effect of the < and > (template) brackets and
2996 ;; of all parens in preprocessor constructs, except for any such construct
2997 ;; containing point. We can then call `c-invalidate-state-cache-1' without
2998 ;; worrying further about macros and template delimiters.
2999 (c-with-<->-as-parens-suppressed
3e8f7d91
AM
3000 (if (and c-state-old-cpp-beg
3001 (< c-state-old-cpp-beg here))
0ec1d2c5 3002 (c-with-all-but-one-cpps-commented-out
3e8f7d91
AM
3003 c-state-old-cpp-beg
3004 (min c-state-old-cpp-end here)
0ec1d2c5
AM
3005 (c-invalidate-state-cache-1 here))
3006 (c-with-cpps-commented-out
3007 (c-invalidate-state-cache-1 here)))))
3008
3009(defun c-parse-state ()
3010 ;; This is a wrapper over `c-parse-state-1'. See that function for a
3011 ;; description of the functionality and return value.
3012 ;;
3013 ;; It suppresses the syntactic effect of the < and > (template) brackets and
3014 ;; of all parens in preprocessor constructs, except for any such construct
3015 ;; containing point. We can then call `c-parse-state-1' without worrying
3016 ;; further about macros and template delimiters.
3017 (let (here-cpp-beg here-cpp-end)
3018 (save-excursion
3019 (when (c-beginning-of-macro)
3020 (setq here-cpp-beg (point))
3021 (unless
3022 (> (setq here-cpp-end (c-syntactic-end-of-macro))
3023 here-cpp-beg)
3024 (setq here-cpp-beg nil here-cpp-end nil))))
3025 ;; FIXME!!! Put in a `condition-case' here to protect the integrity of the
3026 ;; subsystem.
3027 (prog1
3028 (c-with-<->-as-parens-suppressed
3029 (if (and here-cpp-beg (> here-cpp-end here-cpp-beg))
3030 (c-with-all-but-one-cpps-commented-out
3031 here-cpp-beg here-cpp-end
3032 (c-parse-state-1))
3033 (c-with-cpps-commented-out
3034 (c-parse-state-1))))
3e8f7d91
AM
3035 (setq c-state-old-cpp-beg (and here-cpp-beg (copy-marker here-cpp-beg t))
3036 c-state-old-cpp-end (and here-cpp-end (copy-marker here-cpp-end t)))
3037 )))
0ec1d2c5
AM
3038
3039;; Debug tool to catch cache inconsistencies. This is called from
3040;; 000tests.el.
a66cd3ee
MS
3041(defvar c-debug-parse-state nil)
3042(unless (fboundp 'c-real-parse-state)
3043 (fset 'c-real-parse-state (symbol-function 'c-parse-state)))
3044(cc-bytecomp-defun c-real-parse-state)
3045(defun c-debug-parse-state ()
0ec1d2c5 3046 (let ((here (point)) (res1 (c-real-parse-state)) res2)
0386b551 3047 (let ((c-state-cache nil)
0ec1d2c5
AM
3048 (c-state-cache-good-pos 1)
3049 (c-state-nonlit-pos-cache nil)
3050 (c-state-nonlit-pos-cache-limit 1)
3051 (c-state-brace-pair-desert nil)
3052 (c-state-point-min 1)
3053 (c-state-point-min-lit-type nil)
3054 (c-state-point-min-lit-start nil)
3055 (c-state-min-scan-pos 1)
3056 (c-state-old-cpp-beg nil)
3057 (c-state-old-cpp-end nil))
a66cd3ee
MS
3058 (setq res2 (c-real-parse-state)))
3059 (unless (equal res1 res2)
0386b551
AM
3060 ;; The cache can actually go further back due to the ad-hoc way
3061 ;; the first paren is found, so try to whack off a bit of its
3062 ;; start before complaining.
3063 (save-excursion
3064 (goto-char (or (c-least-enclosing-brace res2) (point)))
3065 (c-beginning-of-defun-1)
3066 (while (not (or (bobp) (eq (char-after) ?{)))
3067 (c-beginning-of-defun-1))
3068 (unless (equal (c-whack-state-before (point) res1) res2)
0ec1d2c5 3069 (message (concat "c-parse-state inconsistency at %s: "
0386b551 3070 "using cache: %s, from scratch: %s")
0ec1d2c5 3071 here res1 res2))))
a66cd3ee 3072 res1))
0ec1d2c5 3073
a66cd3ee
MS
3074(defun c-toggle-parse-state-debug (&optional arg)
3075 (interactive "P")
3076 (setq c-debug-parse-state (c-calculate-state arg c-debug-parse-state))
3077 (fset 'c-parse-state (symbol-function (if c-debug-parse-state
3078 'c-debug-parse-state
3079 'c-real-parse-state)))
3080 (c-keep-region-active))
0386b551
AM
3081(when c-debug-parse-state
3082 (c-toggle-parse-state-debug 1))
a66cd3ee 3083
0ec1d2c5 3084\f
d9e94c22
MS
3085(defun c-whack-state-before (bufpos paren-state)
3086 ;; Whack off any state information from PAREN-STATE which lies
3087 ;; before BUFPOS. Not destructive on PAREN-STATE.
d9e94c22
MS
3088 (let* ((newstate (list nil))
3089 (ptr newstate)
3090 car)
3091 (while paren-state
3092 (setq car (car paren-state)
3093 paren-state (cdr paren-state))
3094 (if (< (if (consp car) (car car) car) bufpos)
3095 (setq paren-state nil)
3096 (setcdr ptr (list car))
3097 (setq ptr (cdr ptr))))
3098 (cdr newstate)))
3099
3100(defun c-whack-state-after (bufpos paren-state)
3101 ;; Whack off any state information from PAREN-STATE which lies at or
3102 ;; after BUFPOS. Not destructive on PAREN-STATE.
d9e94c22
MS
3103 (catch 'done
3104 (while paren-state
3105 (let ((car (car paren-state)))
3106 (if (consp car)
3107 ;; just check the car, because in a balanced brace
3108 ;; expression, it must be impossible for the corresponding
3109 ;; close brace to be before point, but the open brace to
3110 ;; be after.
3111 (if (<= bufpos (car car))
3112 nil ; whack it off
3113 (if (< bufpos (cdr car))
3114 ;; its possible that the open brace is before
3115 ;; bufpos, but the close brace is after. In that
3116 ;; case, convert this to a non-cons element. The
3117 ;; rest of the state is before bufpos, so we're
3118 ;; done.
3119 (throw 'done (cons (car car) (cdr paren-state)))
3120 ;; we know that both the open and close braces are
3121 ;; before bufpos, so we also know that everything else
3122 ;; on state is before bufpos.
3123 (throw 'done paren-state)))
3124 (if (<= bufpos car)
3125 nil ; whack it off
3126 ;; it's before bufpos, so everything else should too.
3127 (throw 'done paren-state)))
3128 (setq paren-state (cdr paren-state)))
3129 nil)))
3130
3131(defun c-most-enclosing-brace (paren-state &optional bufpos)
3132 ;; Return the bufpos of the innermost enclosing open paren before
0386b551 3133 ;; bufpos, or nil if none was found.
d9e94c22
MS
3134 (let (enclosingp)
3135 (or bufpos (setq bufpos 134217727))
3136 (while paren-state
3137 (setq enclosingp (car paren-state)
3138 paren-state (cdr paren-state))
3139 (if (or (consp enclosingp)
3140 (>= enclosingp bufpos))
3141 (setq enclosingp nil)
d9e94c22
MS
3142 (setq paren-state nil)))
3143 enclosingp))
3144
0386b551
AM
3145(defun c-least-enclosing-brace (paren-state)
3146 ;; Return the bufpos of the outermost enclosing open paren, or nil
3147 ;; if none was found.
d9e94c22 3148 (let (pos elem)
d9e94c22
MS
3149 (while paren-state
3150 (setq elem (car paren-state)
3151 paren-state (cdr paren-state))
0386b551
AM
3152 (if (integerp elem)
3153 (setq pos elem)))
d9e94c22
MS
3154 pos))
3155
3156(defun c-safe-position (bufpos paren-state)
0386b551
AM
3157 ;; Return the closest "safe" position recorded on PAREN-STATE that
3158 ;; is higher up than BUFPOS. Return nil if PAREN-STATE doesn't
3159 ;; contain any. Return nil if BUFPOS is nil, which is useful to
3160 ;; find the closest limit before a given limit that might be nil.
d9e94c22 3161 ;;
0386b551
AM
3162 ;; A "safe" position is a position at or after a recorded open
3163 ;; paren, or after a recorded close paren. The returned position is
3164 ;; thus either the first position after a close brace, or the first
3165 ;; position after an enclosing paren, or at the enclosing paren in
3166 ;; case BUFPOS is immediately after it.
d9e94c22
MS
3167 (when bufpos
3168 (let (elem)
3169 (catch 'done
3170 (while paren-state
3171 (setq elem (car paren-state))
3172 (if (consp elem)
3173 (cond ((< (cdr elem) bufpos)
3174 (throw 'done (cdr elem)))
3175 ((< (car elem) bufpos)
3176 ;; See below.
3177 (throw 'done (min (1+ (car elem)) bufpos))))
3178 (if (< elem bufpos)
3179 ;; elem is the position at and not after the opening paren, so
3180 ;; we can go forward one more step unless it's equal to
3181 ;; bufpos. This is useful in some cases avoid an extra paren
3182 ;; level between the safe position and bufpos.
3183 (throw 'done (min (1+ elem) bufpos))))
3184 (setq paren-state (cdr paren-state)))))))
3185
3186(defun c-beginning-of-syntax ()
3187 ;; This is used for `font-lock-beginning-of-syntax-function'. It
3188 ;; goes to the closest previous point that is known to be outside
3189 ;; any string literal or comment. `c-state-cache' is used if it has
3190 ;; a position in the vicinity.
3191 (let* ((paren-state c-state-cache)
3192 elem
3193
3194 (pos (catch 'done
3195 ;; Note: Similar code in `c-safe-position'. The
3196 ;; difference is that we accept a safe position at
3197 ;; the point and don't bother to go forward past open
3198 ;; parens.
3199 (while paren-state
3200 (setq elem (car paren-state))
3201 (if (consp elem)
3202 (cond ((<= (cdr elem) (point))
3203 (throw 'done (cdr elem)))
3204 ((<= (car elem) (point))
3205 (throw 'done (car elem))))
3206 (if (<= elem (point))
3207 (throw 'done elem)))
3208 (setq paren-state (cdr paren-state)))
3209 (point-min))))
3210
3211 (if (> pos (- (point) 4000))
3212 (goto-char pos)
3213 ;; The position is far back. Try `c-beginning-of-defun-1'
3214 ;; (although we can't be entirely sure it will go to a position
3215 ;; outside a comment or string in current emacsen). FIXME:
3216 ;; Consult `syntax-ppss' here.
3217 (c-beginning-of-defun-1)
3218 (if (< (point) pos)
3219 (goto-char pos)))))
3220
3221\f
3222;; Tools for scanning identifiers and other tokens.
3223
3224(defun c-on-identifier ()
3225 "Return non-nil if the point is on or directly after an identifier.
3226Keywords are recognized and not considered identifiers. If an
3227identifier is detected, the returned value is its starting position.
0386b551
AM
3228If an identifier ends at the point and another begins at it \(can only
3229happen in Pike) then the point for the preceding one is returned.
d9e94c22 3230
0386b551
AM
3231Note that this function might do hidden buffer changes. See the
3232comment at the start of cc-engine.el for more info."
3233
3234 ;; FIXME: Shouldn't this function handle "operator" in C++?
d9e94c22
MS
3235
3236 (save-excursion
0386b551
AM
3237 (skip-syntax-backward "w_")
3238
3239 (or
3240
3241 ;; Check for a normal (non-keyword) identifier.
3242 (and (looking-at c-symbol-start)
3243 (not (looking-at c-keywords-regexp))
3244 (point))
3245
3246 (when (c-major-mode-is 'pike-mode)
3247 ;; Handle the `<operator> syntax in Pike.
3248 (let ((pos (point)))
3249 (skip-chars-backward "-!%&*+/<=>^|~[]()")
3250 (and (if (< (skip-chars-backward "`") 0)
3251 t
3252 (goto-char pos)
3253 (eq (char-after) ?\`))
3254 (looking-at c-symbol-key)
3255 (>= (match-end 0) pos)
3256 (point))))
3257
3258 ;; Handle the "operator +" syntax in C++.
3259 (when (and c-overloadable-operators-regexp
3260 (= (c-backward-token-2 0) 0))
3261
3262 (cond ((and (looking-at c-overloadable-operators-regexp)
51c9af45 3263 (or (not c-opt-op-identifier-prefix)
0386b551 3264 (and (= (c-backward-token-2 1) 0)
51c9af45 3265 (looking-at c-opt-op-identifier-prefix))))
0386b551
AM
3266 (point))
3267
3268 ((save-excursion
51c9af45
AM
3269 (and c-opt-op-identifier-prefix
3270 (looking-at c-opt-op-identifier-prefix)
0386b551
AM
3271 (= (c-forward-token-2 1) 0)
3272 (looking-at c-overloadable-operators-regexp)))
3273 (point))))
3274
3275 )))
d9e94c22
MS
3276
3277(defsubst c-simple-skip-symbol-backward ()
3278 ;; If the point is at the end of a symbol then skip backward to the
3279 ;; beginning of it. Don't move otherwise. Return non-nil if point
3280 ;; moved.
0386b551
AM
3281 ;;
3282 ;; This function might do hidden buffer changes.
d9e94c22
MS
3283 (or (< (skip-syntax-backward "w_") 0)
3284 (and (c-major-mode-is 'pike-mode)
3285 ;; Handle the `<operator> syntax in Pike.
3286 (let ((pos (point)))
2a15eb73 3287 (if (and (< (skip-chars-backward "-!%&*+/<=>^|~[]()") 0)
d9e94c22
MS
3288 (< (skip-chars-backward "`") 0)
3289 (looking-at c-symbol-key)
3290 (>= (match-end 0) pos))
3291 t
3292 (goto-char pos)
3293 nil)))))
3294
0386b551 3295(defun c-beginning-of-current-token (&optional back-limit)
d9e94c22
MS
3296 ;; Move to the beginning of the current token. Do not move if not
3297 ;; in the middle of one. BACK-LIMIT may be used to bound the
3298 ;; backward search; if given it's assumed to be at the boundary
a85fd6da 3299 ;; between two tokens. Return non-nil if the point is moved, nil
580fba94 3300 ;; otherwise.
0386b551
AM
3301 ;;
3302 ;; This function might do hidden buffer changes.
d9e94c22 3303 (let ((start (point)))
580fba94
AM
3304 (if (looking-at "\\w\\|\\s_")
3305 (skip-syntax-backward "w_" back-limit)
3306 (when (< (skip-syntax-backward ".()" back-limit) 0)
3307 (while (let ((pos (or (and (looking-at c-nonsymbol-token-regexp)
3308 (match-end 0))
3309 ;; `c-nonsymbol-token-regexp' should always match
3310 ;; since we've skipped backward over punctuator
3311 ;; or paren syntax, but consume one char in case
3312 ;; it doesn't so that we don't leave point before
3313 ;; some earlier incorrect token.
3314 (1+ (point)))))
3315 (if (<= pos start)
3316 (goto-char pos))))))
3317 (< (point) start)))
d9e94c22 3318
ff959bab 3319(defun c-end-of-current-token (&optional back-limit)
d9e94c22
MS
3320 ;; Move to the end of the current token. Do not move if not in the
3321 ;; middle of one. BACK-LIMIT may be used to bound the backward
3322 ;; search; if given it's assumed to be at the boundary between two
ff959bab 3323 ;; tokens. Return non-nil if the point is moved, nil otherwise.
0386b551
AM
3324 ;;
3325 ;; This function might do hidden buffer changes.
d9e94c22
MS
3326 (let ((start (point)))
3327 (cond ((< (skip-syntax-backward "w_" (1- start)) 0)
3328 (skip-syntax-forward "w_"))
3329 ((< (skip-syntax-backward ".()" back-limit) 0)
3330 (while (progn
3331 (if (looking-at c-nonsymbol-token-regexp)
3332 (goto-char (match-end 0))
3333 ;; `c-nonsymbol-token-regexp' should always match since
3334 ;; we've skipped backward over punctuator or paren
3335 ;; syntax, but move forward in case it doesn't so that
3336 ;; we don't leave point earlier than we started with.
3337 (forward-char))
ff959bab
MS
3338 (< (point) start)))))
3339 (> (point) start)))
d9e94c22
MS
3340
3341(defconst c-jump-syntax-balanced
3342 (if (memq 'gen-string-delim c-emacs-features)
3343 "\\w\\|\\s_\\|\\s\(\\|\\s\)\\|\\s\"\\|\\s|"
3344 "\\w\\|\\s_\\|\\s\(\\|\\s\)\\|\\s\""))
3345
3346(defconst c-jump-syntax-unbalanced
3347 (if (memq 'gen-string-delim c-emacs-features)
3348 "\\w\\|\\s_\\|\\s\"\\|\\s|"
3349 "\\w\\|\\s_\\|\\s\""))
3350
3351(defun c-forward-token-2 (&optional count balanced limit)
3352 "Move forward by tokens.
3353A token is defined as all symbols and identifiers which aren't
3354syntactic whitespace \(note that multicharacter tokens like \"==\" are
3355treated properly). Point is always either left at the beginning of a
3356token or not moved at all. COUNT specifies the number of tokens to
3357move; a negative COUNT moves in the opposite direction. A COUNT of 0
3358moves to the next token beginning only if not already at one. If
3359BALANCED is true, move over balanced parens, otherwise move into them.
3360Also, if BALANCED is true, never move out of an enclosing paren.
3361
3362LIMIT sets the limit for the movement and defaults to the point limit.
3363The case when LIMIT is set in the middle of a token, comment or macro
3364is handled correctly, i.e. the point won't be left there.
3365
3366Return the number of tokens left to move \(positive or negative). If
3367BALANCED is true, a move over a balanced paren counts as one. Note
3368that if COUNT is 0 and no appropriate token beginning is found, 1 will
3369be returned. Thus, a return value of 0 guarantees that point is at
3370the requested position and a return value less \(without signs) than
0386b551
AM
3371COUNT guarantees that point is at the beginning of some token.
3372
3373Note that this function might do hidden buffer changes. See the
3374comment at the start of cc-engine.el for more info."
d9e94c22
MS
3375
3376 (or count (setq count 1))
3377 (if (< count 0)
3378 (- (c-backward-token-2 (- count) balanced limit))
3379
3380 (let ((jump-syntax (if balanced
3381 c-jump-syntax-balanced
3382 c-jump-syntax-unbalanced))
3383 (last (point))
3384 (prev (point)))
3385
3386 (if (zerop count)
3387 ;; If count is zero we should jump if in the middle of a token.
3388 (c-end-of-current-token))
3389
3390 (save-restriction
3391 (if limit (narrow-to-region (point-min) limit))
3392 (if (/= (point)
3393 (progn (c-forward-syntactic-ws) (point)))
3394 ;; Skip whitespace. Count this as a move if we did in
3395 ;; fact move.
3396 (setq count (max (1- count) 0)))
3397
3398 (if (eobp)
3399 ;; Moved out of bounds. Make sure the returned count isn't zero.
3400 (progn
3401 (if (zerop count) (setq count 1))
3402 (goto-char last))
3403
3404 ;; Use `condition-case' to avoid having the limit tests
3405 ;; inside the loop.
3406 (condition-case nil
3407 (while (and
3408 (> count 0)
3409 (progn
3410 (setq last (point))
3411 (cond ((looking-at jump-syntax)
3412 (goto-char (scan-sexps (point) 1))
3413 t)
3414 ((looking-at c-nonsymbol-token-regexp)
3415 (goto-char (match-end 0))
3416 t)
3417 ;; `c-nonsymbol-token-regexp' above should always
3418 ;; match if there are correct tokens. Try to
3419 ;; widen to see if the limit was set in the
3420 ;; middle of one, else fall back to treating
3421 ;; the offending thing as a one character token.
3422 ((and limit
3423 (save-restriction
3424 (widen)
3425 (looking-at c-nonsymbol-token-regexp)))
3426 nil)
3427 (t
3428 (forward-char)
3429 t))))
3430 (c-forward-syntactic-ws)
3431 (setq prev last
3432 count (1- count)))
3433 (error (goto-char last)))
3434
3435 (when (eobp)
3436 (goto-char prev)
3437 (setq count (1+ count)))))
3438
3439 count)))
3440
3441(defun c-backward-token-2 (&optional count balanced limit)
3442 "Move backward by tokens.
3443See `c-forward-token-2' for details."
3444
3445 (or count (setq count 1))
3446 (if (< count 0)
3447 (- (c-forward-token-2 (- count) balanced limit))
3448
3449 (or limit (setq limit (point-min)))
3450 (let ((jump-syntax (if balanced
3451 c-jump-syntax-balanced
3452 c-jump-syntax-unbalanced))
3453 (last (point)))
3454
3455 (if (zerop count)
3456 ;; The count is zero so try to skip to the beginning of the
3457 ;; current token.
3458 (if (> (point)
3459 (progn (c-beginning-of-current-token) (point)))
3460 (if (< (point) limit)
3461 ;; The limit is inside the same token, so return 1.
3462 (setq count 1))
3463
3464 ;; We're not in the middle of a token. If there's
3465 ;; whitespace after the point then we must move backward,
3466 ;; so set count to 1 in that case.
3467 (and (looking-at c-syntactic-ws-start)
3468 ;; If we're looking at a '#' that might start a cpp
3469 ;; directive then we have to do a more elaborate check.
3470 (or (/= (char-after) ?#)
3471 (not c-opt-cpp-prefix)
3472 (save-excursion
3473 (and (= (point)
3474 (progn (beginning-of-line)
3475 (looking-at "[ \t]*")
3476 (match-end 0)))
3477 (or (bobp)
3478 (progn (backward-char)
3479 (not (eq (char-before) ?\\)))))))
3480 (setq count 1))))
3481
3482 ;; Use `condition-case' to avoid having to check for buffer
3483 ;; limits in `backward-char', `scan-sexps' and `goto-char' below.
3484 (condition-case nil
3485 (while (and
3486 (> count 0)
3487 (progn
3488 (c-backward-syntactic-ws)
3489 (backward-char)
3490 (if (looking-at jump-syntax)
3491 (goto-char (scan-sexps (1+ (point)) -1))
3492 ;; This can be very inefficient if there's a long
3493 ;; sequence of operator tokens without any separation.
3494 ;; That doesn't happen in practice, anyway.
3495 (c-beginning-of-current-token))
3496 (>= (point) limit)))
3497 (setq last (point)
3498 count (1- count)))
3499 (error (goto-char last)))
3500
3501 (if (< (point) limit)
3502 (goto-char last))
3503
3504 count)))
3505
3506(defun c-forward-token-1 (&optional count balanced limit)
3507 "Like `c-forward-token-2' but doesn't treat multicharacter operator
3508tokens like \"==\" as single tokens, i.e. all sequences of symbol
3509characters are jumped over character by character. This function is
3510for compatibility only; it's only a wrapper over `c-forward-token-2'."
3511 (let ((c-nonsymbol-token-regexp "\\s.\\|\\s\(\\|\\s\)"))
3512 (c-forward-token-2 count balanced limit)))
3513
3514(defun c-backward-token-1 (&optional count balanced limit)
3515 "Like `c-backward-token-2' but doesn't treat multicharacter operator
3516tokens like \"==\" as single tokens, i.e. all sequences of symbol
3517characters are jumped over character by character. This function is
3518for compatibility only; it's only a wrapper over `c-backward-token-2'."
3519 (let ((c-nonsymbol-token-regexp "\\s.\\|\\s\(\\|\\s\)"))
3520 (c-backward-token-2 count balanced limit)))
3521
3522\f
3523;; Tools for doing searches restricted to syntactically relevant text.
3524
3525(defun c-syntactic-re-search-forward (regexp &optional bound noerror
3526 paren-level not-inside-token
3527 lookbehind-submatch)
3528 "Like `re-search-forward', but only report matches that are found
3529in syntactically significant text. I.e. matches in comments, macros
3530or string literals are ignored. The start point is assumed to be
3531outside any comment, macro or string literal, or else the content of
3532that region is taken as syntactically significant text.
3533
3534If PAREN-LEVEL is non-nil, an additional restriction is added to
2a15eb73
MS
3535ignore matches in nested paren sexps. The search will also not go
3536outside the current list sexp, which has the effect that if the point
3537should be moved to BOUND when no match is found \(i.e. NOERROR is
3538neither nil nor t), then it will be at the closing paren if the end of
3539the current list sexp is encountered first.
d9e94c22
MS
3540
3541If NOT-INSIDE-TOKEN is non-nil, matches in the middle of tokens are
3542ignored. Things like multicharacter operators and special symbols
3543\(e.g. \"`()\" in Pike) are handled but currently not floating point
3544constants.
3545
3546If LOOKBEHIND-SUBMATCH is non-nil, it's taken as a number of a
3547subexpression in REGEXP. The end of that submatch is used as the
3548position to check for syntactic significance. If LOOKBEHIND-SUBMATCH
3549isn't used or if that subexpression didn't match then the start
3550position of the whole match is used instead. The \"look behind\"
3551subexpression is never tested before the starting position, so it
3552might be a good idea to include \\=\\= as a match alternative in it.
3553
3554Optimization note: Matches might be missed if the \"look behind\"
2a15eb73 3555subexpression can match the end of nonwhite syntactic whitespace,
d9e94c22 3556i.e. the end of comments or cpp directives. This since the function
2a15eb73
MS
3557skips over such things before resuming the search. It's on the other
3558hand not safe to assume that the \"look behind\" subexpression never
3559matches syntactic whitespace.
3560
3561Bug: Unbalanced parens inside cpp directives are currently not handled
3562correctly \(i.e. they don't get ignored as they should) when
0386b551
AM
3563PAREN-LEVEL is set.
3564
3565Note that this function might do hidden buffer changes. See the
3566comment at the start of cc-engine.el for more info."
d9e94c22
MS
3567
3568 (or bound (setq bound (point-max)))
3569 (if paren-level (setq paren-level -1))
3570
3571 ;;(message "c-syntactic-re-search-forward %s %s %S" (point) bound regexp)
3572
3573 (let ((start (point))
2a15eb73
MS
3574 tmp
3575 ;; Start position for the last search.
3576 search-pos
3577 ;; The `parse-partial-sexp' state between the start position
3578 ;; and the point.
3579 state
3580 ;; The current position after the last state update. The next
3581 ;; `parse-partial-sexp' continues from here.
3582 (state-pos (point))
3583 ;; The position at which to check the state and the state
3584 ;; there. This is separate from `state-pos' since we might
3585 ;; need to back up before doing the next search round.
3586 check-pos check-state
3587 ;; Last position known to end a token.
d9e94c22 3588 (last-token-end-pos (point-min))
2a15eb73
MS
3589 ;; Set when a valid match is found.
3590 found)
d9e94c22
MS
3591
3592 (condition-case err
3593 (while
3594 (and
2a15eb73
MS
3595 (progn
3596 (setq search-pos (point))
3597 (re-search-forward regexp bound noerror))
d9e94c22
MS
3598
3599 (progn
2a15eb73
MS
3600 (setq state (parse-partial-sexp
3601 state-pos (match-beginning 0) paren-level nil state)
3602 state-pos (point))
d9e94c22 3603 (if (setq check-pos (and lookbehind-submatch
2a15eb73
MS
3604 (or (not paren-level)
3605 (>= (car state) 0))
d9e94c22
MS
3606 (match-end lookbehind-submatch)))
3607 (setq check-state (parse-partial-sexp
2a15eb73
MS
3608 state-pos check-pos paren-level nil state))
3609 (setq check-pos state-pos
d9e94c22
MS
3610 check-state state))
3611
2a15eb73
MS
3612 ;; NOTE: If we got a look behind subexpression and get
3613 ;; an insignificant match in something that isn't
d9e94c22
MS
3614 ;; syntactic whitespace (i.e. strings or in nested
3615 ;; parentheses), then we can never skip more than a
2a15eb73
MS
3616 ;; single character from the match start position
3617 ;; (i.e. `state-pos' here) before continuing the
3618 ;; search. That since the look behind subexpression
3619 ;; might match the end of the insignificant region in
3620 ;; the next search.
d9e94c22
MS
3621
3622 (cond
d9e94c22
MS
3623 ((elt check-state 7)
3624 ;; Match inside a line comment. Skip to eol. Use
3625 ;; `re-search-forward' instead of `skip-chars-forward' to get
3626 ;; the right bound behavior.
3627 (re-search-forward "[\n\r]" bound noerror))
3628
3629 ((elt check-state 4)
3630 ;; Match inside a block comment. Skip to the '*/'.
3631 (search-forward "*/" bound noerror))
3632
3633 ((and (not (elt check-state 5))
3634 (eq (char-before check-pos) ?/)
2a15eb73 3635 (not (c-get-char-property (1- check-pos) 'syntax-table))
d9e94c22
MS
3636 (memq (char-after check-pos) '(?/ ?*)))
3637 ;; Match in the middle of the opener of a block or line
3638 ;; comment.
3639 (if (= (char-after check-pos) ?/)
3640 (re-search-forward "[\n\r]" bound noerror)
3641 (search-forward "*/" bound noerror)))
3642
2a15eb73
MS
3643 ;; The last `parse-partial-sexp' above might have
3644 ;; stopped short of the real check position if the end
3645 ;; of the current sexp was encountered in paren-level
3646 ;; mode. The checks above are always false in that
3647 ;; case, and since they can do better skipping in
3648 ;; lookbehind-submatch mode, we do them before
3649 ;; checking the paren level.
3650
3651 ((and paren-level
3652 (/= (setq tmp (car check-state)) 0))
3653 ;; Check the paren level first since we're short of the
3654 ;; syntactic checking position if the end of the
3655 ;; current sexp was encountered by `parse-partial-sexp'.
3656 (if (> tmp 0)
3657
3658 ;; Inside a nested paren sexp.
3659 (if lookbehind-submatch
3660 ;; See the NOTE above.
3661 (progn (goto-char state-pos) t)
3662 ;; Skip out of the paren quickly.
3663 (setq state (parse-partial-sexp state-pos bound 0 nil state)
3664 state-pos (point)))
3665
3666 ;; Have exited the current paren sexp.
3667 (if noerror
3668 (progn
3669 ;; The last `parse-partial-sexp' call above
3670 ;; has left us just after the closing paren
3671 ;; in this case, so we can modify the bound
3672 ;; to leave the point at the right position
3673 ;; upon return.
3674 (setq bound (1- (point)))
3675 nil)
3676 (signal 'search-failed (list regexp)))))
3677
3678 ((setq tmp (elt check-state 3))
3679 ;; Match inside a string.
3680 (if (or lookbehind-submatch
3681 (not (integerp tmp)))
3682 ;; See the NOTE above.
3683 (progn (goto-char state-pos) t)
3684 ;; Skip to the end of the string before continuing.
3685 (let ((ender (make-string 1 tmp)) (continue t))
3686 (while (if (search-forward ender bound noerror)
3687 (progn
3688 (setq state (parse-partial-sexp
3689 state-pos (point) nil nil state)
3690 state-pos (point))
3691 (elt state 3))
3692 (setq continue nil)))
3693 continue)))
d9e94c22
MS
3694
3695 ((save-excursion
3696 (save-match-data
3697 (c-beginning-of-macro start)))
3698 ;; Match inside a macro. Skip to the end of it.
3699 (c-end-of-macro)
3700 (cond ((<= (point) bound) t)
3701 (noerror nil)
2a15eb73 3702 (t (signal 'search-failed (list regexp)))))
d9e94c22 3703
2a15eb73
MS
3704 ((and not-inside-token
3705 (or (< check-pos last-token-end-pos)
3706 (< check-pos
3707 (save-excursion
3708 (goto-char check-pos)
3709 (save-match-data
3710 (c-end-of-current-token last-token-end-pos))
3711 (setq last-token-end-pos (point))))))
3712 ;; Inside a token.
3713 (if lookbehind-submatch
3714 ;; See the NOTE above.
3715 (goto-char state-pos)
3716 (goto-char (min last-token-end-pos bound))))
d9e94c22
MS
3717
3718 (t
3719 ;; A real match.
3720 (setq found t)
2a15eb73
MS
3721 nil)))
3722
3723 ;; Should loop to search again, but take care to avoid
3724 ;; looping on the same spot.
3725 (or (/= search-pos (point))
3726 (if (= (point) bound)
3727 (if noerror
3728 nil
3729 (signal 'search-failed (list regexp)))
3730 (forward-char)
3731 t))))
d9e94c22
MS
3732
3733 (error
3734 (goto-char start)
3735 (signal (car err) (cdr err))))
3736
2a15eb73 3737 ;;(message "c-syntactic-re-search-forward done %s" (or (match-end 0) (point)))
d9e94c22
MS
3738
3739 (if found
3740 (progn
2a15eb73
MS
3741 (goto-char (match-end 0))
3742 (match-end 0))
d9e94c22
MS
3743
3744 ;; Search failed. Set point as appropriate.
2a15eb73
MS
3745 (if (eq noerror t)
3746 (goto-char start)
3747 (goto-char bound))
d9e94c22
MS
3748 nil)))
3749
47641aac
GM
3750(defvar safe-pos-list) ; bound in c-syntactic-skip-backward
3751
d0fcee66
AM
3752(defsubst c-ssb-lit-begin ()
3753 ;; Return the start of the literal point is in, or nil.
3754 ;; We read and write the variables `safe-pos', `safe-pos-list', `state'
3755 ;; bound in the caller.
3756
3757 ;; Use `parse-partial-sexp' from a safe position down to the point to check
3758 ;; if it's outside comments and strings.
3759 (save-excursion
3760 (let ((pos (point)) safe-pos state pps-end-pos)
3761 ;; Pick a safe position as close to the point as possible.
3762 ;;
3763 ;; FIXME: Consult `syntax-ppss' here if our cache doesn't give a good
3764 ;; position.
47641aac 3765
d0fcee66
AM
3766 (while (and safe-pos-list
3767 (> (car safe-pos-list) (point)))
3768 (setq safe-pos-list (cdr safe-pos-list)))
3769 (unless (setq safe-pos (car-safe safe-pos-list))
3770 (setq safe-pos (max (or (c-safe-position
3771 (point) (or c-state-cache
3772 (c-parse-state)))
3773 0)
3774 (point-min))
3775 safe-pos-list (list safe-pos)))
3776
3777 ;; Cache positions along the way to use if we have to back up more. We
3778 ;; cache every closing paren on the same level. If the paren cache is
3779 ;; relevant in this region then we're typically already on the same
3780 ;; level as the target position. Note that we might cache positions
3781 ;; after opening parens in case safe-pos is in a nested list. That's
3782 ;; both uncommon and harmless.
3783 (while (progn
3784 (setq state (parse-partial-sexp
3785 safe-pos pos 0))
3786 (< (point) pos))
3787 (setq safe-pos (point)
3788 safe-pos-list (cons safe-pos safe-pos-list)))
3789
3790 ;; If the state contains the start of the containing sexp we cache that
3791 ;; position too, so that parse-partial-sexp in the next run has a bigger
3792 ;; chance of starting at the same level as the target position and thus
3793 ;; will get more good safe positions into the list.
3794 (if (elt state 1)
3795 (setq safe-pos (1+ (elt state 1))
3796 safe-pos-list (cons safe-pos safe-pos-list)))
3797
3798 (if (or (elt state 3) (elt state 4))
3799 ;; Inside string or comment. Continue search at the
3800 ;; beginning of it.
3801 (elt state 8)))))
3802
0386b551 3803(defun c-syntactic-skip-backward (skip-chars &optional limit paren-level)
d9e94c22
MS
3804 "Like `skip-chars-backward' but only look at syntactically relevant chars,
3805i.e. don't stop at positions inside syntactic whitespace or string
3806literals. Preprocessor directives are also ignored, with the exception
3807of the one that the point starts within, if any. If LIMIT is given,
0386b551
AM
3808it's assumed to be at a syntactically relevant position.
3809
3810If PAREN-LEVEL is non-nil, the function won't stop in nested paren
3811sexps, and the search will also not go outside the current paren sexp.
3812However, if LIMIT or the buffer limit is reached inside a nested paren
3813then the point will be left at the limit.
3814
3815Non-nil is returned if the point moved, nil otherwise.
3816
3817Note that this function might do hidden buffer changes. See the
3818comment at the start of cc-engine.el for more info."
d9e94c22
MS
3819
3820 (let ((start (point))
d0fcee66 3821 state-2
d9e94c22
MS
3822 ;; A list of syntactically relevant positions in descending
3823 ;; order. It's used to avoid scanning repeatedly over
3824 ;; potentially large regions with `parse-partial-sexp' to verify
d0fcee66 3825 ;; each position. Used in `c-ssb-lit-begin'
d9e94c22
MS
3826 safe-pos-list
3827 ;; The result from `c-beginning-of-macro' at the start position or the
3828 ;; start position itself if it isn't within a macro. Evaluated on
3829 ;; demand.
0386b551
AM
3830 start-macro-beg
3831 ;; The earliest position after the current one with the same paren
3832 ;; level. Used only when `paren-level' is set.
d0fcee66 3833 lit-beg
0386b551 3834 (paren-level-pos (point)))
d9e94c22 3835
d0fcee66
AM
3836 (while
3837 (progn
3838 ;; The next loop "tries" to find the end point each time round,
3839 ;; loops when it hasn't succeeded.
3840 (while
3841 (and
3842 (< (skip-chars-backward skip-chars limit) 0)
3843
3844 (let ((pos (point)) state-2 pps-end-pos)
3845
3846 (cond
3847 ;; Don't stop inside a literal
3848 ((setq lit-beg (c-ssb-lit-begin))
3849 (goto-char lit-beg)
3850 t)
3851
3852 ((and paren-level
3853 (save-excursion
3854 (setq state-2 (parse-partial-sexp
3855 pos paren-level-pos -1)
3856 pps-end-pos (point))
3857 (/= (car state-2) 0)))
3858 ;; Not at the right level.
3859
3860 (if (and (< (car state-2) 0)
3861 ;; We stop above if we go out of a paren.
3862 ;; Now check whether it precedes or is
3863 ;; nested in the starting sexp.
3864 (save-excursion
3865 (setq state-2
3866 (parse-partial-sexp
3867 pps-end-pos paren-level-pos
3868 nil nil state-2))
3869 (< (car state-2) 0)))
3870
3871 ;; We've stopped short of the starting position
3872 ;; so the hit was inside a nested list. Go up
3873 ;; until we are at the right level.
3874 (condition-case nil
3875 (progn
3876 (goto-char (scan-lists pos -1
3877 (- (car state-2))))
3878 (setq paren-level-pos (point))
3879 (if (and limit (>= limit paren-level-pos))
0386b551 3880 (progn
d0fcee66
AM
3881 (goto-char limit)
3882 nil)
3883 t))
3884 (error
3885 (goto-char (or limit (point-min)))
3886 nil))
3887
3888 ;; The hit was outside the list at the start
3889 ;; position. Go to the start of the list and exit.
3890 (goto-char (1+ (elt state-2 1)))
3891 nil))
3892
3893 ((c-beginning-of-macro limit)
3894 ;; Inside a macro.
3895 (if (< (point)
3896 (or start-macro-beg
3897 (setq start-macro-beg
3898 (save-excursion
3899 (goto-char start)
3900 (c-beginning-of-macro limit)
3901 (point)))))
3902 t
3903
3904 ;; It's inside the same macro we started in so it's
3905 ;; a relevant match.
3906 (goto-char pos)
3907 nil))))))
3908
3909 (> (point)
3910 (progn
3911 ;; Skip syntactic ws afterwards so that we don't stop at the
3912 ;; end of a comment if `skip-chars' is something like "^/".
3913 (c-backward-syntactic-ws)
3914 (point)))))
d9e94c22 3915
0386b551
AM
3916 ;; We might want to extend this with more useful return values in
3917 ;; the future.
3918 (/= (point) start)))
3919
3920;; The following is an alternative implementation of
3921;; `c-syntactic-skip-backward' that uses backward movement to keep
3922;; track of the syntactic context. It turned out to be generally
3923;; slower than the one above which uses forward checks from earlier
3924;; safe positions.
3925;;
3926;;(defconst c-ssb-stop-re
3927;; ;; The regexp matching chars `c-syntactic-skip-backward' needs to
3928;; ;; stop at to avoid going into comments and literals.
3929;; (concat
3930;; ;; Match comment end syntax and string literal syntax. Also match
3931;; ;; '/' for block comment endings (not covered by comment end
3932;; ;; syntax).
3933;; "\\s>\\|/\\|\\s\""
3934;; (if (memq 'gen-string-delim c-emacs-features)
3935;; "\\|\\s|"
3936;; "")
3937;; (if (memq 'gen-comment-delim c-emacs-features)
3938;; "\\|\\s!"
3939;; "")))
3940;;
3941;;(defconst c-ssb-stop-paren-re
3942;; ;; Like `c-ssb-stop-re' but also stops at paren chars.
3943;; (concat c-ssb-stop-re "\\|\\s(\\|\\s)"))
3944;;
3945;;(defconst c-ssb-sexp-end-re
3946;; ;; Regexp matching the ending syntax of a complex sexp.
3947;; (concat c-string-limit-regexp "\\|\\s)"))
3948;;
3949;;(defun c-syntactic-skip-backward (skip-chars &optional limit paren-level)
3950;; "Like `skip-chars-backward' but only look at syntactically relevant chars,
3951;;i.e. don't stop at positions inside syntactic whitespace or string
3952;;literals. Preprocessor directives are also ignored. However, if the
3953;;point is within a comment, string literal or preprocessor directory to
3954;;begin with, its contents is treated as syntactically relevant chars.
3955;;If LIMIT is given, it limits the backward search and the point will be
3956;;left there if no earlier position is found.
3957;;
3958;;If PAREN-LEVEL is non-nil, the function won't stop in nested paren
3959;;sexps, and the search will also not go outside the current paren sexp.
3960;;However, if LIMIT or the buffer limit is reached inside a nested paren
3961;;then the point will be left at the limit.
3962;;
3963;;Non-nil is returned if the point moved, nil otherwise.
3964;;
3965;;Note that this function might do hidden buffer changes. See the
3966;;comment at the start of cc-engine.el for more info."
3967;;
3968;; (save-restriction
3969;; (when limit
3970;; (narrow-to-region limit (point-max)))
3971;;
3972;; (let ((start (point)))
3973;; (catch 'done
3974;; (while (let ((last-pos (point))
3975;; (stop-pos (progn
3976;; (skip-chars-backward skip-chars)
3977;; (point))))
3978;;
3979;; ;; Skip back over the same region as
3980;; ;; `skip-chars-backward' above, but keep to
3981;; ;; syntactically relevant positions.
3982;; (goto-char last-pos)
3983;; (while (and
3984;; ;; `re-search-backward' with a single char regexp
3985;; ;; should be fast.
3986;; (re-search-backward
3987;; (if paren-level c-ssb-stop-paren-re c-ssb-stop-re)
3988;; stop-pos 'move)
3989;;
3990;; (progn
3991;; (cond
3992;; ((looking-at "\\s(")
3993;; ;; `paren-level' is set and we've found the
3994;; ;; start of the containing paren.
3995;; (forward-char)
3996;; (throw 'done t))
3997;;
3998;; ((looking-at c-ssb-sexp-end-re)
3999;; ;; We're at the end of a string literal or paren
4000;; ;; sexp (if `paren-level' is set).
4001;; (forward-char)
4002;; (condition-case nil
4003;; (c-backward-sexp)
4004;; (error
4005;; (goto-char limit)
4006;; (throw 'done t))))
4007;;
4008;; (t
4009;; (forward-char)
4010;; ;; At the end of some syntactic ws or possibly
4011;; ;; after a plain '/' operator.
4012;; (let ((pos (point)))
4013;; (c-backward-syntactic-ws)
4014;; (if (= pos (point))
4015;; ;; Was a plain '/' operator. Go past it.
4016;; (backward-char)))))
4017;;
4018;; (> (point) stop-pos))))
4019;;
4020;; ;; Now the point is either at `stop-pos' or at some
4021;; ;; position further back if `stop-pos' was at a
4022;; ;; syntactically irrelevant place.
4023;;
4024;; ;; Skip additional syntactic ws so that we don't stop
4025;; ;; at the end of a comment if `skip-chars' is
4026;; ;; something like "^/".
4027;; (c-backward-syntactic-ws)
4028;;
4029;; (< (point) stop-pos))))
4030;;
4031;; ;; We might want to extend this with more useful return values
4032;; ;; in the future.
4033;; (/= (point) start))))
d9e94c22
MS
4034
4035\f
4036;; Tools for handling comments and string literals.
4037
4038(defun c-slow-in-literal (&optional lim detect-cpp)
4039 "Return the type of literal point is in, if any.
4040The return value is `c' if in a C-style comment, `c++' if in a C++
4041style comment, `string' if in a string literal, `pound' if DETECT-CPP
4042is non-nil and in a preprocessor line, or nil if somewhere else.
4043Optional LIM is used as the backward limit of the search. If omitted,
4044or nil, `c-beginning-of-defun' is used.
4045
4046The last point calculated is cached if the cache is enabled, i.e. if
4047`c-in-literal-cache' is bound to a two element vector.
4048
0386b551
AM
4049Note that this function might do hidden buffer changes. See the
4050comment at the start of cc-engine.el for more info."
4051
d9e94c22
MS
4052 (if (and (vectorp c-in-literal-cache)
4053 (= (point) (aref c-in-literal-cache 0)))
4054 (aref c-in-literal-cache 1)
4055 (let ((rtn (save-excursion
4056 (let* ((pos (point))
4057 (lim (or lim (progn
4058 (c-beginning-of-syntax)
4059 (point))))
4060 (state (parse-partial-sexp lim pos)))
4061 (cond
4062 ((elt state 3) 'string)
4063 ((elt state 4) (if (elt state 7) 'c++ 'c))
4064 ((and detect-cpp (c-beginning-of-macro lim)) 'pound)
4065 (t nil))))))
4066 ;; cache this result if the cache is enabled
4067 (if (not c-in-literal-cache)
4068 (setq c-in-literal-cache (vector (point) rtn)))
4069 rtn)))
4070
4071;; XEmacs has a built-in function that should make this much quicker.
4072;; I don't think we even need the cache, which makes our lives more
4073;; complicated anyway. In this case, lim is only used to detect
4074;; cpp directives.
4075;;
4076;; Note that there is a bug in Xemacs's buffer-syntactic-context when used in
4077;; conjunction with syntax-table-properties. The bug is present in, e.g.,
4078;; Xemacs 21.4.4. It manifested itself thus:
4079;;
4080;; Starting with an empty AWK Mode buffer, type
4081;; /regexp/ {<C-j>
4082;; Point gets wrongly left at column 0, rather than being indented to tab-width.
4083;;
4084;; AWK Mode is designed such that when the first / is typed, it gets the
4085;; syntax-table property "string fence". When the second / is typed, BOTH /s
4086;; are given the s-t property "string". However, buffer-syntactic-context
4087;; fails to take account of the change of the s-t property on the opening / to
4088;; "string", and reports that the { is within a string started by the second /.
4089;;
4090;; The workaround for this is for the AWK Mode initialisation to switch the
4091;; defalias for c-in-literal to c-slow-in-literal. This will slow down other
4092;; cc-modes in Xemacs whenever an awk-buffer has been initialised.
b414f371 4093;;
d9e94c22
MS
4094;; (Alan Mackenzie, 2003/4/30).
4095
4096(defun c-fast-in-literal (&optional lim detect-cpp)
0386b551 4097 ;; This function might do hidden buffer changes.
d9e94c22
MS
4098 (let ((context (buffer-syntactic-context)))
4099 (cond
4100 ((eq context 'string) 'string)
4101 ((eq context 'comment) 'c++)
4102 ((eq context 'block-comment) 'c)
4103 ((and detect-cpp (save-excursion (c-beginning-of-macro lim))) 'pound))))
4104
4105(defalias 'c-in-literal
4106 (if (fboundp 'buffer-syntactic-context)
7bfc3fdb 4107 'c-fast-in-literal ; XEmacs
d9e94c22
MS
4108 'c-slow-in-literal)) ; GNU Emacs
4109
4110;; The defalias above isn't enough to shut up the byte compiler.
4111(cc-bytecomp-defun c-in-literal)
4112
4113(defun c-literal-limits (&optional lim near not-in-delimiter)
4114 "Return a cons of the beginning and end positions of the comment or
4115string surrounding point (including both delimiters), or nil if point
4116isn't in one. If LIM is non-nil, it's used as the \"safe\" position
4117to start parsing from. If NEAR is non-nil, then the limits of any
4118literal next to point is returned. \"Next to\" means there's only
4119spaces and tabs between point and the literal. The search for such a
4120literal is done first in forward direction. If NOT-IN-DELIMITER is
4121non-nil, the case when point is inside a starting delimiter won't be
a85fd6da 4122recognized. This only has effect for comments which have starting
d9e94c22
MS
4123delimiters with more than one character.
4124
0386b551
AM
4125Note that this function might do hidden buffer changes. See the
4126comment at the start of cc-engine.el for more info."
d9e94c22
MS
4127
4128 (save-excursion
4129 (let* ((pos (point))
4130 (lim (or lim (progn
4131 (c-beginning-of-syntax)
4132 (point))))
4133 (state (parse-partial-sexp lim pos)))
4134
0386b551
AM
4135 (cond ((elt state 3) ; String.
4136 (goto-char (elt state 8))
d9e94c22
MS
4137 (cons (point) (or (c-safe (c-forward-sexp 1) (point))
4138 (point-max))))
4139
0386b551
AM
4140 ((elt state 4) ; Comment.
4141 (goto-char (elt state 8))
d9e94c22
MS
4142 (cons (point) (progn (c-forward-single-comment) (point))))
4143
4144 ((and (not not-in-delimiter)
4145 (not (elt state 5))
4146 (eq (char-before) ?/)
4147 (looking-at "[/*]"))
4148 ;; We're standing in a comment starter.
4149 (backward-char 1)
4150 (cons (point) (progn (c-forward-single-comment) (point))))
4151
4152 (near
4153 (goto-char pos)
4154
4155 ;; Search forward for a literal.
4156 (skip-chars-forward " \t")
4157
4158 (cond
4159 ((looking-at c-string-limit-regexp) ; String.
4160 (cons (point) (or (c-safe (c-forward-sexp 1) (point))
4161 (point-max))))
4162
4163 ((looking-at c-comment-start-regexp) ; Line or block comment.
4164 (cons (point) (progn (c-forward-single-comment) (point))))
4165
4166 (t
4167 ;; Search backward.
4168 (skip-chars-backward " \t")
4169
4170 (let ((end (point)) beg)
4171 (cond
4172 ((save-excursion
4173 (< (skip-syntax-backward c-string-syntax) 0)) ; String.
4174 (setq beg (c-safe (c-backward-sexp 1) (point))))
4175
4176 ((and (c-safe (forward-char -2) t)
4177 (looking-at "*/"))
4178 ;; Block comment. Due to the nature of line
4179 ;; comments, they will always be covered by the
4180 ;; normal case above.
4181 (goto-char end)
4182 (c-backward-single-comment)
4183 ;; If LIM is bogus, beg will be bogus.
4184 (setq beg (point))))
4185
4186 (if beg (cons beg end))))))
4187 ))))
4188
0386b551
AM
4189;; In case external callers use this; it did have a docstring.
4190(defalias 'c-literal-limits-fast 'c-literal-limits)
d9e94c22 4191
0386b551
AM
4192(defun c-collect-line-comments (range)
4193 "If the argument is a cons of two buffer positions (such as returned by
4194`c-literal-limits'), and that range contains a C++ style line comment,
4195then an extended range is returned that contains all adjacent line
4196comments (i.e. all comments that starts in the same column with no
4197empty lines or non-whitespace characters between them). Otherwise the
4198argument is returned.
d9e94c22 4199
0386b551
AM
4200Note that this function might do hidden buffer changes. See the
4201comment at the start of cc-engine.el for more info."
d9e94c22
MS
4202
4203 (save-excursion
0386b551
AM
4204 (condition-case nil
4205 (if (and (consp range) (progn
4206 (goto-char (car range))
4207 (looking-at c-line-comment-starter)))
b414f371 4208 (let ((col (current-column))
0386b551
AM
4209 (beg (point))
4210 (bopl (c-point 'bopl))
4211 (end (cdr range)))
4212 ;; Got to take care in the backward direction to handle
4213 ;; comments which are preceded by code.
4214 (while (and (c-backward-single-comment)
4215 (>= (point) bopl)
4216 (looking-at c-line-comment-starter)
4217 (= col (current-column)))
4218 (setq beg (point)
4219 bopl (c-point 'bopl)))
4220 (goto-char end)
4221 (while (and (progn (skip-chars-forward " \t")
4222 (looking-at c-line-comment-starter))
4223 (= col (current-column))
4224 (prog1 (zerop (forward-line 1))
4225 (setq end (point)))))
4226 (cons beg end))
4227 range)
4228 (error range))))
d9e94c22
MS
4229
4230(defun c-literal-type (range)
4231 "Convenience function that given the result of `c-literal-limits',
a85fd6da
AM
4232returns nil or the type of literal that the range surrounds, one
4233of the symbols 'c, 'c++ or 'string. It's much faster than using
4234`c-in-literal' and is intended to be used when you need both the
4235type of a literal and its limits.
d9e94c22 4236
0386b551
AM
4237Note that this function might do hidden buffer changes. See the
4238comment at the start of cc-engine.el for more info."
4239
d9e94c22
MS
4240 (if (consp range)
4241 (save-excursion
4242 (goto-char (car range))
4243 (cond ((looking-at c-string-limit-regexp) 'string)
4244 ((or (looking-at "//") ; c++ line comment
4245 (and (looking-at "\\s<") ; comment starter
4246 (looking-at "#"))) ; awk comment.
4247 'c++)
4248 (t 'c))) ; Assuming the range is valid.
4249 range))
4250
4251\f
4252;; `c-find-decl-spots' and accompanying stuff.
4253
4254;; Variables used in `c-find-decl-spots' to cache the search done for
4255;; the first declaration in the last call. When that function starts,
4256;; it needs to back up over syntactic whitespace to look at the last
4257;; token before the region being searched. That can sometimes cause
4258;; moves back and forth over a quite large region of comments and
4259;; macros, which would be repeated for each changed character when
4260;; we're called during fontification, since font-lock refontifies the
4261;; current line for each change. Thus it's worthwhile to cache the
4262;; first match.
4263;;
4264;; `c-find-decl-syntactic-pos' is a syntactically relevant position in
4265;; the syntactic whitespace less or equal to some start position.
4266;; There's no cached value if it's nil.
4267;;
4268;; `c-find-decl-match-pos' is the match position if
4269;; `c-find-decl-prefix-search' matched before the syntactic whitespace
4270;; at `c-find-decl-syntactic-pos', or nil if there's no such match.
4271(defvar c-find-decl-syntactic-pos nil)
4272(make-variable-buffer-local 'c-find-decl-syntactic-pos)
4273(defvar c-find-decl-match-pos nil)
4274(make-variable-buffer-local 'c-find-decl-match-pos)
4275
4276(defsubst c-invalidate-find-decl-cache (change-min-pos)
4277 (and c-find-decl-syntactic-pos
4278 (< change-min-pos c-find-decl-syntactic-pos)
4279 (setq c-find-decl-syntactic-pos nil)))
4280
4281; (defface c-debug-decl-spot-face
4282; '((t (:background "Turquoise")))
4283; "Debug face to mark the spots where `c-find-decl-spots' stopped.")
4284; (defface c-debug-decl-sws-face
4285; '((t (:background "Khaki")))
4286; "Debug face to mark the syntactic whitespace between the declaration
4287; spots and the preceding token end.")
4288
4289(defmacro c-debug-put-decl-spot-faces (match-pos decl-pos)
4290 (when (facep 'c-debug-decl-spot-face)
0386b551 4291 `(c-save-buffer-state ((match-pos ,match-pos) (decl-pos ,decl-pos))
d9e94c22
MS
4292 (c-debug-add-face (max match-pos (point-min)) decl-pos
4293 'c-debug-decl-sws-face)
4294 (c-debug-add-face decl-pos (min (1+ decl-pos) (point-max))
4295 'c-debug-decl-spot-face))))
4296(defmacro c-debug-remove-decl-spot-faces (beg end)
4297 (when (facep 'c-debug-decl-spot-face)
0386b551 4298 `(c-save-buffer-state ()
d9e94c22
MS
4299 (c-debug-remove-face ,beg ,end 'c-debug-decl-spot-face)
4300 (c-debug-remove-face ,beg ,end 'c-debug-decl-sws-face))))
4301
4302(defmacro c-find-decl-prefix-search ()
4303 ;; Macro used inside `c-find-decl-spots'. It ought to be a defun,
4304 ;; but it contains lots of free variables that refer to things
4305 ;; inside `c-find-decl-spots'. The point is left at `cfd-match-pos'
4306 ;; if there is a match, otherwise at `cfd-limit'.
0386b551
AM
4307 ;;
4308 ;; This macro might do hidden buffer changes.
d9e94c22
MS
4309
4310 '(progn
4311 ;; Find the next property match position if we haven't got one already.
4312 (unless cfd-prop-match
4313 (save-excursion
4314 (while (progn
4315 (goto-char (next-single-property-change
4316 (point) 'c-type nil cfd-limit))
4317 (and (< (point) cfd-limit)
4318 (not (eq (c-get-char-property (1- (point)) 'c-type)
4319 'c-decl-end)))))
4320 (setq cfd-prop-match (point))))
4321
0386b551
AM
4322 ;; Find the next `c-decl-prefix-or-start-re' match if we haven't
4323 ;; got one already.
d9e94c22 4324 (unless cfd-re-match
0386b551
AM
4325
4326 (if (> cfd-re-match-end (point))
4327 (goto-char cfd-re-match-end))
4328
4329 (while (if (setq cfd-re-match-end
4330 (re-search-forward c-decl-prefix-or-start-re
4331 cfd-limit 'move))
4332
4333 ;; Match. Check if it's inside a comment or string literal.
4334 (c-got-face-at
4335 (if (setq cfd-re-match (match-end 1))
4336 ;; Matched the end of a token preceding a decl spot.
4337 (progn
4338 (goto-char cfd-re-match)
4339 (1- cfd-re-match))
4340 ;; Matched a token that start a decl spot.
4341 (goto-char (match-beginning 0))
4342 (point))
4343 c-literal-faces)
4344
4345 ;; No match. Finish up and exit the loop.
4346 (setq cfd-re-match cfd-limit)
4347 nil)
4348
4349 ;; Skip out of comments and string literals.
d9e94c22
MS
4350 (while (progn
4351 (goto-char (next-single-property-change
0386b551 4352 (point) 'face nil cfd-limit))
d9e94c22 4353 (and (< (point) cfd-limit)
0386b551
AM
4354 (c-got-face-at (point) c-literal-faces)))))
4355
4356 ;; If we matched at the decl start, we have to back up over the
4357 ;; preceding syntactic ws to set `cfd-match-pos' and to catch
4358 ;; any decl spots in the syntactic ws.
d9e94c22 4359 (unless cfd-re-match
0386b551
AM
4360 (c-backward-syntactic-ws)
4361 (setq cfd-re-match (point))))
d9e94c22
MS
4362
4363 ;; Choose whichever match is closer to the start.
4364 (if (< cfd-re-match cfd-prop-match)
4365 (setq cfd-match-pos cfd-re-match
4366 cfd-re-match nil)
4367 (setq cfd-match-pos cfd-prop-match
4368 cfd-prop-match nil))
4369
4370 (goto-char cfd-match-pos)
4371
4372 (when (< cfd-match-pos cfd-limit)
4373 ;; Skip forward past comments only so we don't skip macros.
4374 (c-forward-comments)
4375 ;; Set the position to continue at. We can avoid going over
4376 ;; the comments skipped above a second time, but it's possible
4377 ;; that the comment skipping has taken us past `cfd-prop-match'
4378 ;; since the property might be used inside comments.
4379 (setq cfd-continue-pos (if cfd-prop-match
4380 (min cfd-prop-match (point))
4381 (point))))))
4382
4383(defun c-find-decl-spots (cfd-limit cfd-decl-re cfd-face-checklist cfd-fun)
0386b551 4384 ;; Call CFD-FUN for each possible spot for a declaration, cast or
a85fd6da
AM
4385 ;; label from the point to CFD-LIMIT.
4386 ;;
4387 ;; CFD-FUN is called with point at the start of the spot. It's
4388 ;; passed two arguments: The first is the end position of the token
4389 ;; preceding the spot, or 0 for the implicit match at bob. The
4390 ;; second is a flag that is t when the match is inside a macro. If
4391 ;; CFD-FUN adds `c-decl-end' properties somewhere below the current
4392 ;; spot, it should return non-nil to ensure that the next search
4393 ;; will find them.
0386b551 4394 ;;
a85fd6da 4395 ;; Such a spot is:
0386b551
AM
4396 ;; o The first token after bob.
4397 ;; o The first token after the end of submatch 1 in
4398 ;; `c-decl-prefix-or-start-re' when that submatch matches.
4399 ;; o The start of each `c-decl-prefix-or-start-re' match when
4400 ;; submatch 1 doesn't match.
5a89f0a7 4401 ;; o The first token after the end of each occurrence of the
0386b551
AM
4402 ;; `c-type' text property with the value `c-decl-end', provided
4403 ;; `c-type-decl-end-used' is set.
4404 ;;
4405 ;; Only a spot that match CFD-DECL-RE and whose face is in the
4406 ;; CFD-FACE-CHECKLIST list causes CFD-FUN to be called. The face
4407 ;; check is disabled if CFD-FACE-CHECKLIST is nil.
d9e94c22
MS
4408 ;;
4409 ;; If the match is inside a macro then the buffer is narrowed to the
4410 ;; end of it, so that CFD-FUN can investigate the following tokens
4411 ;; without matching something that begins inside a macro and ends
4412 ;; outside it. It's to avoid this work that the CFD-DECL-RE and
4413 ;; CFD-FACE-CHECKLIST checks exist.
4414 ;;
0386b551
AM
4415 ;; The spots are visited approximately in order from top to bottom.
4416 ;; It's however the positions where `c-decl-prefix-or-start-re'
4417 ;; matches and where `c-decl-end' properties are found that are in
4418 ;; order. Since the spots often are at the following token, they
4419 ;; might be visited out of order insofar as more spots are reported
4420 ;; later on within the syntactic whitespace between the match
4421 ;; positions and their spots.
4422 ;;
4423 ;; It's assumed that comments and strings are fontified in the
d9e94c22
MS
4424 ;; searched range.
4425 ;;
4426 ;; This is mainly used in fontification, and so has an elaborate
4427 ;; cache to handle repeated calls from the same start position; see
4428 ;; the variables above.
4429 ;;
4430 ;; All variables in this function begin with `cfd-' to avoid name
4431 ;; collision with the (dynamically bound) variables used in CFD-FUN.
0386b551
AM
4432 ;;
4433 ;; This function might do hidden buffer changes.
d9e94c22 4434
0386b551
AM
4435 (let ((cfd-start-pos (point))
4436 (cfd-buffer-end (point-max))
4437 ;; The end of the token preceding the decl spot last found
4438 ;; with `c-decl-prefix-or-start-re'. `cfd-limit' if there's
4439 ;; no match.
d9e94c22 4440 cfd-re-match
0386b551
AM
4441 ;; The end position of the last `c-decl-prefix-or-start-re'
4442 ;; match. If this is greater than `cfd-continue-pos', the
4443 ;; next regexp search is started here instead.
4444 (cfd-re-match-end (point-min))
4445 ;; The end of the last `c-decl-end' found by
4446 ;; `c-find-decl-prefix-search'. `cfd-limit' if there's no
4447 ;; match. If searching for the property isn't needed then we
4448 ;; disable it by setting it to `cfd-limit' directly.
d9e94c22 4449 (cfd-prop-match (unless c-type-decl-end-used cfd-limit))
0386b551
AM
4450 ;; The end of the token preceding the decl spot last found by
4451 ;; `c-find-decl-prefix-search'. 0 for the implicit match at
4452 ;; bob. `cfd-limit' if there's no match. In other words,
4453 ;; this is the minimum of `cfd-re-match' and `cfd-prop-match'.
d9e94c22
MS
4454 (cfd-match-pos cfd-limit)
4455 ;; The position to continue searching at.
4456 cfd-continue-pos
4457 ;; The position of the last "real" token we've stopped at.
4458 ;; This can be greater than `cfd-continue-pos' when we get
4459 ;; hits inside macros or at `c-decl-end' positions inside
4460 ;; comments.
4461 (cfd-token-pos 0)
4462 ;; The end position of the last entered macro.
4463 (cfd-macro-end 0))
4464
4465 ;; Initialize by finding a syntactically relevant start position
0386b551
AM
4466 ;; before the point, and do the first `c-decl-prefix-or-start-re'
4467 ;; search unless we're at bob.
d9e94c22 4468
0386b551 4469 (let (start-in-literal start-in-macro syntactic-pos)
d9e94c22
MS
4470 ;; Must back up a bit since we look for the end of the previous
4471 ;; statement or declaration, which is earlier than the first
4472 ;; returned match.
4473
0386b551
AM
4474 (cond
4475 ;; First we need to move to a syntactically relevant position.
4476 ;; Begin by backing out of comment or string literals.
4477 ((and
4478 (when (c-got-face-at (point) c-literal-faces)
4479 ;; Try to use the faces to back up to the start of the
4480 ;; literal. FIXME: What if the point is on a declaration
4481 ;; inside a comment?
4482 (while (and (not (bobp))
4483 (c-got-face-at (1- (point)) c-literal-faces))
4484 (goto-char (previous-single-property-change
4485 (point) 'face nil (point-min))))
4486
4487 ;; XEmacs doesn't fontify the quotes surrounding string
4488 ;; literals.
4489 (and (featurep 'xemacs)
4490 (eq (get-text-property (point) 'face)
4491 'font-lock-string-face)
4492 (not (bobp))
4493 (progn (backward-char)
4494 (not (looking-at c-string-limit-regexp)))
4495 (forward-char))
4496
4497 ;; Don't trust the literal to contain only literal faces
4498 ;; (the font lock package might not have fontified the
4499 ;; start of it at all, for instance) so check that we have
4500 ;; arrived at something that looks like a start or else
4501 ;; resort to `c-literal-limits'.
4502 (unless (looking-at c-literal-start-regexp)
4503 (let ((range (c-literal-limits)))
4504 (if range (goto-char (car range)))))
4505
4506 (setq start-in-literal (point)))
4507
4508 ;; The start is in a literal. If the limit is in the same
4509 ;; one we don't have to find a syntactic position etc. We
4510 ;; only check that if the limit is at or before bonl to save
4511 ;; time; it covers the by far most common case when font-lock
4512 ;; refontifies the current line only.
4513 (<= cfd-limit (c-point 'bonl cfd-start-pos))
4514 (save-excursion
4515 (goto-char cfd-start-pos)
4516 (while (progn
4517 (goto-char (next-single-property-change
4518 (point) 'face nil cfd-limit))
4519 (and (< (point) cfd-limit)
4520 (c-got-face-at (point) c-literal-faces))))
4521 (= (point) cfd-limit)))
4522
4523 ;; Completely inside a literal. Set up variables to trig the
4524 ;; (< cfd-continue-pos cfd-start-pos) case below and it'll
4525 ;; find a suitable start position.
4526 (setq cfd-continue-pos start-in-literal))
4527
4528 ;; Check if the region might be completely inside a macro, to
4529 ;; optimize that like the completely-inside-literal above.
4530 ((save-excursion
4531 (and (= (forward-line 1) 0)
4532 (bolp) ; forward-line has funny behavior at eob.
4533 (>= (point) cfd-limit)
4534 (progn (backward-char)
4535 (eq (char-before) ?\\))))
4536 ;; (Maybe) completely inside a macro. Only need to trig the
4537 ;; (< cfd-continue-pos cfd-start-pos) case below to make it
4538 ;; set things up.
4539 (setq cfd-continue-pos (1- cfd-start-pos)
4540 start-in-macro t))
d9e94c22 4541
0386b551
AM
4542 (t
4543 ;; Back out of any macro so we don't miss any declaration
4544 ;; that could follow after it.
4545 (when (c-beginning-of-macro)
4546 (setq start-in-macro t))
4547
4548 ;; Now we're at a proper syntactically relevant position so we
4549 ;; can use the cache. But first clear it if it applied
4550 ;; further down.
4551 (c-invalidate-find-decl-cache cfd-start-pos)
4552
4553 (setq syntactic-pos (point))
4554 (unless (eq syntactic-pos c-find-decl-syntactic-pos)
4555 ;; Don't have to do this if the cache is relevant here,
4556 ;; typically if the same line is refontified again. If
4557 ;; we're just some syntactic whitespace further down we can
4558 ;; still use the cache to limit the skipping.
4559 (c-backward-syntactic-ws c-find-decl-syntactic-pos))
4560
4561 ;; If we hit `c-find-decl-syntactic-pos' and
4562 ;; `c-find-decl-match-pos' is set then we install the cached
4563 ;; values. If we hit `c-find-decl-syntactic-pos' and
4564 ;; `c-find-decl-match-pos' is nil then we know there's no decl
4565 ;; prefix in the whitespace before `c-find-decl-syntactic-pos'
4566 ;; and so we can continue the search from this point. If we
4567 ;; didn't hit `c-find-decl-syntactic-pos' then we're now in
4568 ;; the right spot to begin searching anyway.
4569 (if (and (eq (point) c-find-decl-syntactic-pos)
4570 c-find-decl-match-pos)
d9e94c22
MS
4571 (setq cfd-match-pos c-find-decl-match-pos
4572 cfd-continue-pos syntactic-pos)
0386b551
AM
4573
4574 (setq c-find-decl-syntactic-pos syntactic-pos)
4575
4576 (when (if (bobp)
4577 ;; Always consider bob a match to get the first
4578 ;; declaration in the file. Do this separately instead of
4579 ;; letting `c-decl-prefix-or-start-re' match bob, so that
4580 ;; regexp always can consume at least one character to
4581 ;; ensure that we won't get stuck in an infinite loop.
4582 (setq cfd-re-match 0)
4583 (backward-char)
4584 (c-beginning-of-current-token)
4585 (< (point) cfd-limit))
4586 ;; Do an initial search now. In the bob case above it's
4587 ;; only done to search for a `c-decl-end' spot.
4588 (c-find-decl-prefix-search))
4589
4590 (setq c-find-decl-match-pos (and (< cfd-match-pos cfd-start-pos)
4591 cfd-match-pos)))))
4592
4593 ;; Advance `cfd-continue-pos' if it's before the start position.
4594 ;; The closest continue position that might have effect at or
4595 ;; after the start depends on what we started in. This also
4596 ;; finds a suitable start position in the special cases when the
4597 ;; region is completely within a literal or macro.
4598 (when (and cfd-continue-pos (< cfd-continue-pos cfd-start-pos))
4599
4600 (cond
4601 (start-in-macro
4602 ;; If we're in a macro then it's the closest preceding token
4603 ;; in the macro. Check this before `start-in-literal',
4604 ;; since if we're inside a literal in a macro, the preceding
4605 ;; token is earlier than any `c-decl-end' spot inside the
4606 ;; literal (comment).
4607 (goto-char (or start-in-literal cfd-start-pos))
4608 ;; The only syntactic ws in macros are comments.
d9e94c22 4609 (c-backward-comments)
0386b551
AM
4610 (backward-char)
4611 (c-beginning-of-current-token))
4612
4613 (start-in-literal
4614 ;; If we're in a comment it can only be the closest
4615 ;; preceding `c-decl-end' position within that comment, if
4616 ;; any. Go back to the beginning of such a property so that
4617 ;; `c-find-decl-prefix-search' will find the end of it.
4618 ;; (Can't stop at the end and install it directly on
4619 ;; `cfd-prop-match' since that variable might be cleared
4620 ;; after `cfd-fun' below.)
4621 ;;
4622 ;; Note that if the literal is a string then the property
4623 ;; search will simply skip to the beginning of it right
4624 ;; away.
4625 (if (not c-type-decl-end-used)
4626 (goto-char start-in-literal)
4627 (goto-char cfd-start-pos)
4628 (while (progn
4629 (goto-char (previous-single-property-change
4630 (point) 'c-type nil start-in-literal))
4631 (and (> (point) start-in-literal)
4632 (not (eq (c-get-char-property (point) 'c-type)
4633 'c-decl-end))))))
4634
4635 (when (= (point) start-in-literal)
4636 ;; Didn't find any property inside the comment, so we can
4637 ;; skip it entirely. (This won't skip past a string, but
4638 ;; that'll be handled quickly by the next
4639 ;; `c-find-decl-prefix-search' anyway.)
4640 (c-forward-single-comment)
4641 (if (> (point) cfd-limit)
4642 (goto-char cfd-limit))))
d9e94c22 4643
0386b551
AM
4644 (t
4645 ;; If we started in normal code, the only match that might
4646 ;; apply before the start is what we already got in
4647 ;; `cfd-match-pos' so we can continue at the start position.
4648 ;; (Note that we don't get here if the first match is below
4649 ;; it.)
4650 (goto-char cfd-start-pos)))
4651
4652 ;; Delete found matches if they are before our new continue
4653 ;; position, so that `c-find-decl-prefix-search' won't back up
4654 ;; to them later on.
4655 (setq cfd-continue-pos (point))
4656 (when (and cfd-re-match (< cfd-re-match cfd-continue-pos))
4657 (setq cfd-re-match nil))
4658 (when (and cfd-prop-match (< cfd-prop-match cfd-continue-pos))
4659 (setq cfd-prop-match nil)))
4660
4661 (if syntactic-pos
4662 ;; This is the normal case and we got a proper syntactic
4663 ;; position. If there's a match then it's always outside
4664 ;; macros and comments, so advance to the next token and set
4665 ;; `cfd-token-pos'. The loop below will later go back using
4666 ;; `cfd-continue-pos' to fix declarations inside the
4667 ;; syntactic ws.
4668 (when (and cfd-match-pos (< cfd-match-pos syntactic-pos))
4669 (goto-char syntactic-pos)
4670 (c-forward-syntactic-ws)
4671 (and cfd-continue-pos
4672 (< cfd-continue-pos (point))
4673 (setq cfd-token-pos (point))))
4674
4675 ;; Have one of the special cases when the region is completely
4676 ;; within a literal or macro. `cfd-continue-pos' is set to a
4677 ;; good start position for the search, so do it.
4678 (c-find-decl-prefix-search)))
d9e94c22 4679
51c9af45 4680 ;; Now loop. Round what? (ACM, 2006/7/5). We already got the first match.
d9e94c22
MS
4681
4682 (while (progn
4683 (while (and
4684 (< cfd-match-pos cfd-limit)
4685
4686 (or
4687 ;; Kludge to filter out matches on the "<" that
4688 ;; aren't open parens, for the sake of languages
4689 ;; that got `c-recognize-<>-arglists' set.
4690 (and (eq (char-before cfd-match-pos) ?<)
4691 (not (c-get-char-property (1- cfd-match-pos)
4692 'syntax-table)))
4693
4694 ;; If `cfd-continue-pos' is less or equal to
4695 ;; `cfd-token-pos', we've got a hit inside a macro
4696 ;; that's in the syntactic whitespace before the last
4697 ;; "real" declaration we've checked. If they're equal
4698 ;; we've arrived at the declaration a second time, so
4699 ;; there's nothing to do.
4700 (= cfd-continue-pos cfd-token-pos)
4701
4702 (progn
4703 ;; If `cfd-continue-pos' is less than `cfd-token-pos'
4704 ;; we're still searching for declarations embedded in
4705 ;; the syntactic whitespace. In that case we need
4706 ;; only to skip comments and not macros, since they
4707 ;; can't be nested, and that's already been done in
4708 ;; `c-find-decl-prefix-search'.
4709 (when (> cfd-continue-pos cfd-token-pos)
4710 (c-forward-syntactic-ws)
4711 (setq cfd-token-pos (point)))
4712
4713 ;; Continue if the following token fails the
4714 ;; CFD-DECL-RE and CFD-FACE-CHECKLIST checks.
4715 (when (or (>= (point) cfd-limit)
4716 (not (looking-at cfd-decl-re))
4717 (and cfd-face-checklist
4718 (not (c-got-face-at
4719 (point) cfd-face-checklist))))
4720 (goto-char cfd-continue-pos)
4721 t)))
4722
4723 (< (point) cfd-limit))
4724 (c-find-decl-prefix-search))
4725
4726 (< (point) cfd-limit))
4727
0386b551
AM
4728 (when (and
4729 (>= (point) cfd-start-pos)
d9e94c22 4730
0386b551
AM
4731 (progn
4732 ;; Narrow to the end of the macro if we got a hit inside
4733 ;; one, to avoid recognizing things that start inside the
4734 ;; macro and end outside it.
4735 (when (> cfd-match-pos cfd-macro-end)
4736 ;; Not in the same macro as in the previous round.
4737 (save-excursion
4738 (goto-char cfd-match-pos)
4739 (setq cfd-macro-end
4740 (if (save-excursion (and (c-beginning-of-macro)
4741 (< (point) cfd-match-pos)))
4742 (progn (c-end-of-macro)
4743 (point))
4744 0))))
4745
4746 (if (zerop cfd-macro-end)
4747 t
4748 (if (> cfd-macro-end (point))
4749 (progn (narrow-to-region (point-min) cfd-macro-end)
4750 t)
4751 ;; The matched token was the last thing in the macro,
4752 ;; so the whole match is bogus.
4753 (setq cfd-macro-end 0)
4754 nil))))
d9e94c22
MS
4755
4756 (c-debug-put-decl-spot-faces cfd-match-pos (point))
0386b551
AM
4757 (if (funcall cfd-fun cfd-match-pos (/= cfd-macro-end 0))
4758 (setq cfd-prop-match nil))
d9e94c22
MS
4759
4760 (when (/= cfd-macro-end 0)
4761 ;; Restore limits if we did macro narrowment above.
4762 (narrow-to-region (point-min) cfd-buffer-end)))
4763
4764 (goto-char cfd-continue-pos)
4765 (if (= cfd-continue-pos cfd-limit)
4766 (setq cfd-match-pos cfd-limit)
4767 (c-find-decl-prefix-search)))))
4768
4769\f
4770;; A cache for found types.
4771
4772;; Buffer local variable that contains an obarray with the types we've
4773;; found. If a declaration is recognized somewhere we record the
4774;; fully qualified identifier in it to recognize it as a type
4775;; elsewhere in the file too. This is not accurate since we do not
4776;; bother with the scoping rules of the languages, but in practice the
4777;; same name is seldom used as both a type and something else in a
4778;; file, and we only use this as a last resort in ambiguous cases (see
0386b551
AM
4779;; `c-forward-decl-or-cast-1').
4780;;
580fba94
AM
4781;; Not every type need be in this cache. However, things which have
4782;; ceased to be types must be removed from it.
4783;;
0386b551
AM
4784;; Template types in C++ are added here too but with the template
4785;; arglist replaced with "<>" in references or "<" for the one in the
4786;; primary type. E.g. the type "Foo<A,B>::Bar<C>" is stored as
4787;; "Foo<>::Bar<". This avoids storing very long strings (since C++
4788;; template specs can be fairly sized programs in themselves) and
4789;; improves the hit ratio (it's a type regardless of the template
4790;; args; it's just not the same type, but we're only interested in
4791;; recognizing types, not telling distinct types apart). Note that
4792;; template types in references are added here too; from the example
4793;; above there will also be an entry "Foo<".
d9e94c22
MS
4794(defvar c-found-types nil)
4795(make-variable-buffer-local 'c-found-types)
4796
4797(defsubst c-clear-found-types ()
4798 ;; Clears `c-found-types'.
d9e94c22
MS
4799 (setq c-found-types (make-vector 53 0)))
4800
4801(defun c-add-type (from to)
4802 ;; Add the given region as a type in `c-found-types'. If the region
4803 ;; doesn't match an existing type but there is a type which is equal
4804 ;; to the given one except that the last character is missing, then
4805 ;; the shorter type is removed. That's done to avoid adding all
4806 ;; prefixes of a type as it's being entered and font locked. This
4807 ;; doesn't cover cases like when characters are removed from a type
4808 ;; or added in the middle. We'd need the position of point when the
4809 ;; font locking is invoked to solve this well.
0386b551
AM
4810 ;;
4811 ;; This function might do hidden buffer changes.
4812 (let ((type (c-syntactic-content from to c-recognize-<>-arglists)))
4813 (unless (intern-soft type c-found-types)
4814 (unintern (substring type 0 -1) c-found-types)
4815 (intern type c-found-types))))
d9e94c22 4816
580fba94
AM
4817(defun c-unfind-type (name)
4818 ;; Remove the "NAME" from c-found-types, if present.
4819 (unintern name c-found-types))
4820
d9e94c22
MS
4821(defsubst c-check-type (from to)
4822 ;; Return non-nil if the given region contains a type in
4823 ;; `c-found-types'.
0386b551
AM
4824 ;;
4825 ;; This function might do hidden buffer changes.
4826 (intern-soft (c-syntactic-content from to c-recognize-<>-arglists)
4827 c-found-types))
d9e94c22
MS
4828
4829(defun c-list-found-types ()
4830 ;; Return all the types in `c-found-types' as a sorted list of
4831 ;; strings.
4832 (let (type-list)
4833 (mapatoms (lambda (type)
4834 (setq type-list (cons (symbol-name type)
4835 type-list)))
4836 c-found-types)
4837 (sort type-list 'string-lessp)))
a66cd3ee 4838
2f42c75f
DN
4839;; Shut up the byte compiler.
4840(defvar c-maybe-stale-found-type)
4841
580fba94
AM
4842(defun c-trim-found-types (beg end old-len)
4843 ;; An after change function which, in conjunction with the info in
4844 ;; c-maybe-stale-found-type (set in c-before-change), removes a type
4845 ;; from `c-found-types', should this type have become stale. For
4846 ;; example, this happens to "foo" when "foo \n bar();" becomes
4847 ;; "foo(); \n bar();". Such stale types, if not removed, foul up
4848 ;; the fontification.
b414f371 4849 ;;
580fba94
AM
4850 ;; Have we, perhaps, added non-ws characters to the front/back of a found
4851 ;; type?
4852 (when (> end beg)
4853 (save-excursion
4854 (when (< end (point-max))
4855 (goto-char end)
4856 (if (and (c-beginning-of-current-token) ; only moves when we started in the middle
4857 (progn (goto-char end)
4858 (c-end-of-current-token)))
4859 (c-unfind-type (buffer-substring-no-properties
4860 end (point)))))
4861 (when (> beg (point-min))
4862 (goto-char beg)
4863 (if (and (c-end-of-current-token) ; only moves when we started in the middle
4864 (progn (goto-char beg)
4865 (c-beginning-of-current-token)))
4866 (c-unfind-type (buffer-substring-no-properties
4867 (point) beg))))))
b414f371 4868
580fba94
AM
4869 (if c-maybe-stale-found-type ; e.g. (c-decl-id-start "foo" 97 107 " (* ooka) " "o")
4870 (cond
4871 ;; Changing the amount of (already existing) whitespace - don't do anything.
4872 ((and (c-partial-ws-p beg end)
4873 (or (= beg end) ; removal of WS
580fba94
AM
4874 (string-match "^[ \t\n\r\f\v]*$" (nth 5 c-maybe-stale-found-type)))))
4875
4876 ;; The syntactic relationship which defined a "found type" has been
4877 ;; destroyed.
4878 ((eq (car c-maybe-stale-found-type) 'c-decl-id-start)
4879 (c-unfind-type (cadr c-maybe-stale-found-type)))
4880;; ((eq (car c-maybe-stale-found-type) 'c-decl-type-start) FIXME!!!
4881 )))
4882
d9e94c22 4883\f
dd969a56
AM
4884;; Setting and removing syntax properties on < and > in languages (C++
4885;; and Java) where they can be template/generic delimiters as well as
4886;; their normal meaning of "less/greater than".
4887
4888;; Normally, < and > have syntax 'punctuation'. When they are found to
4889;; be delimiters, they are marked as such with the category properties
4890;; c-<-as-paren-syntax, c->-as-paren-syntax respectively.
4891
4892;; STRATEGY:
4893;;
4894;; It is impossible to determine with certainty whether a <..> pair in
4895;; C++ is two comparison operators or is template delimiters, unless
4896;; one duplicates a lot of a C++ compiler. For example, the following
4897;; code fragment:
4898;;
4899;; foo (a < b, c > d) ;
4900;;
4901;; could be a function call with two integer parameters (each a
4902;; relational expression), or it could be a constructor for class foo
4903;; taking one parameter d of templated type "a < b, c >". They are
4904;; somewhat easier to distinguish in Java.
4905;;
4906;; The strategy now (2010-01) adopted is to mark and unmark < and
4907;; > IN MATCHING PAIRS ONLY. [Previously, they were marked
4908;; individually when their context so indicated. This gave rise to
4909;; intractible problems when one of a matching pair was deleted, or
4910;; pulled into a literal.]
4911;;
4912;; At each buffer change, the syntax-table properties are removed in a
4913;; before-change function and reapplied, when needed, in an
4914;; after-change function. It is far more important that the
4915;; properties get removed when they they are spurious than that they
4916;; be present when wanted.
4917;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
4918(defun c-clear-<-pair-props (&optional pos)
4919 ;; POS (default point) is at a < character. If it is marked with
4920 ;; open paren syntax-table text property, remove the property,
4921 ;; together with the close paren property on the matching > (if
4922 ;; any).
4923 (save-excursion
4924 (if pos
4925 (goto-char pos)
4926 (setq pos (point)))
4927 (when (equal (c-get-char-property (point) 'syntax-table)
4928 c-<-as-paren-syntax)
4929 (with-syntax-table c-no-parens-syntax-table ; ignore unbalanced [,{,(,..
4930 (c-go-list-forward))
4931 (when (equal (c-get-char-property (1- (point)) 'syntax-table)
4932 c->-as-paren-syntax) ; should always be true.
8a249abc
AM
4933 (c-clear-char-property (1- (point)) 'category))
4934 (c-clear-char-property pos 'category))))
dd969a56
AM
4935
4936(defun c-clear->-pair-props (&optional pos)
4937 ;; POS (default point) is at a > character. If it is marked with
4938 ;; close paren syntax-table property, remove the property, together
4939 ;; with the open paren property on the matching < (if any).
4940 (save-excursion
4941 (if pos
4942 (goto-char pos)
4943 (setq pos (point)))
4944 (when (equal (c-get-char-property (point) 'syntax-table)
4945 c->-as-paren-syntax)
4946 (with-syntax-table c-no-parens-syntax-table ; ignore unbalanced [,{,(,..
4947 (c-go-up-list-backward))
4948 (when (equal (c-get-char-property (point) 'syntax-table)
4949 c-<-as-paren-syntax) ; should always be true.
8a249abc
AM
4950 (c-clear-char-property (point) 'category))
4951 (c-clear-char-property pos 'category))))
dd969a56
AM
4952
4953(defun c-clear-<>-pair-props (&optional pos)
4954 ;; POS (default point) is at a < or > character. If it has an
4955 ;; open/close paren syntax-table property, remove this property both
4956 ;; from the current character and its partner (which will also be
4957 ;; thusly marked).
4958 (cond
4959 ((eq (char-after) ?\<)
4960 (c-clear-<-pair-props pos))
4961 ((eq (char-after) ?\>)
4962 (c-clear->-pair-props pos))
4963 (t (c-benign-error
4964 "c-clear-<>-pair-props called from wrong position"))))
4965
4966(defun c-clear-<-pair-props-if-match-after (lim &optional pos)
4967 ;; POS (default point) is at a < character. If it is both marked
4968 ;; with open/close paren syntax-table property, and has a matching >
4969 ;; (also marked) which is after LIM, remove the property both from
4970 ;; the current > and its partner.
4971 (save-excursion
4972 (if pos
4973 (goto-char pos)
4974 (setq pos (point)))
4975 (when (equal (c-get-char-property (point) 'syntax-table)
4976 c-<-as-paren-syntax)
4977 (with-syntax-table c-no-parens-syntax-table ; ignore unbalanced [,{,(,..
4978 (c-go-list-forward))
4979 (when (and (>= (point) lim)
4980 (equal (c-get-char-property (1- (point)) 'syntax-table)
4981 c->-as-paren-syntax)) ; should always be true.
4982 (c-unmark-<->-as-paren (1- (point)))
4983 (c-unmark-<->-as-paren pos)))))
4984
4985(defun c-clear->-pair-props-if-match-before (lim &optional pos)
4986 ;; POS (default point) is at a > character. If it is both marked
4987 ;; with open/close paren syntax-table property, and has a matching <
4988 ;; (also marked) which is before LIM, remove the property both from
4989 ;; the current < and its partner.
4990 (save-excursion
4991 (if pos
4992 (goto-char pos)
4993 (setq pos (point)))
4994 (when (equal (c-get-char-property (point) 'syntax-table)
4995 c->-as-paren-syntax)
4996 (with-syntax-table c-no-parens-syntax-table ; ignore unbalanced [,{,(,..
4997 (c-go-up-list-backward))
4998 (when (and (<= (point) lim)
4999 (equal (c-get-char-property (point) 'syntax-table)
5000 c-<-as-paren-syntax)) ; should always be true.
5001 (c-unmark-<->-as-paren (point))
5002 (c-unmark-<->-as-paren pos)))))
5003
5004(defun c-before-change-check-<>-operators (beg end)
5005 ;; Unmark certain pairs of "< .... >" which are currently marked as
5006 ;; template/generic delimiters. (This marking is via syntax-table
5007 ;; text properties).
5008 ;;
5009 ;; These pairs are those which are in the current "statement" (i.e.,
5010 ;; the region between the {, }, or ; before BEG and the one after
5011 ;; END), and which enclose any part of the interval (BEG END).
5012 ;;
5013 ;; Note that in C++ (?and Java), template/generic parens cannot
5014 ;; enclose a brace or semicolon, so we use these as bounds on the
5015 ;; region we must work on.
5016 ;;
5017 ;; This function is called from before-change-functions (via
5018 ;; c-get-state-before-change-functions). Thus the buffer is widened,
5019 ;; and point is undefined, both at entry and exit.
5020 ;;
5021 ;; FIXME!!! This routine ignores the possibility of macros entirely.
5022 ;; 2010-01-29.
5023 (save-excursion
5024 (let ((beg-lit-limits (progn (goto-char beg) (c-literal-limits)))
5025 (end-lit-limits (progn (goto-char end) (c-literal-limits))))
5026 ;; Locate the barrier before the changed region
5027 (goto-char (if beg-lit-limits (car beg-lit-limits) beg))
5028 (c-syntactic-skip-backward "^;{}" (max (- beg 2048) (point-min)))
5029
5030 ;; Remove the syntax-table properties from each pertinent <...> pair.
5031 ;; Firsly, the ones with the < before beg and > after beg.
5032 (while (c-search-forward-char-property 'category 'c-<-as-paren-syntax beg)
5033 (c-clear-<-pair-props-if-match-after beg (1- (point))))
5034
5035 ;; Locate the barrier after END.
5036 (goto-char (if end-lit-limits (cdr end-lit-limits) end))
5037 (c-syntactic-re-search-forward "[;{}]"
5038 (min (+ end 2048) (point-max)) 'end)
5039
5040 ;; Remove syntax-table properties from the remaining pertinent <...>
5041 ;; pairs, those with a > after end and < before end.
5042 (while (c-search-backward-char-property 'category 'c->-as-paren-syntax end)
5043 (c-clear->-pair-props-if-match-before end)))))
5044
5045
d9e94c22 5046
0386b551
AM
5047(defun c-after-change-check-<>-operators (beg end)
5048 ;; This is called from `after-change-functions' when
5049 ;; c-recognize-<>-arglists' is set. It ensures that no "<" or ">"
5050 ;; chars with paren syntax become part of another operator like "<<"
5051 ;; or ">=".
5052 ;;
5053 ;; This function might do hidden buffer changes.
5054
cb694ab7
AM
5055 (save-excursion
5056 (goto-char beg)
5057 (when (or (looking-at "[<>]")
5058 (< (skip-chars-backward "<>") 0))
5059
0386b551 5060 (goto-char beg)
cb694ab7
AM
5061 (c-beginning-of-current-token)
5062 (when (and (< (point) beg)
5063 (looking-at c-<>-multichar-token-regexp)
5064 (< beg (setq beg (match-end 0))))
5065 (while (progn (skip-chars-forward "^<>" beg)
5066 (< (point) beg))
dd969a56 5067 (c-clear-<>-pair-props)
cb694ab7
AM
5068 (forward-char))))
5069
5070 (when (< beg end)
5071 (goto-char end)
0386b551
AM
5072 (when (or (looking-at "[<>]")
5073 (< (skip-chars-backward "<>") 0))
5074
cb694ab7 5075 (goto-char end)
0386b551 5076 (c-beginning-of-current-token)
cb694ab7 5077 (when (and (< (point) end)
0386b551 5078 (looking-at c-<>-multichar-token-regexp)
cb694ab7
AM
5079 (< end (setq end (match-end 0))))
5080 (while (progn (skip-chars-forward "^<>" end)
5081 (< (point) end))
dd969a56 5082 (c-clear-<>-pair-props)
cb694ab7 5083 (forward-char)))))))
d9e94c22 5084
dd969a56
AM
5085
5086\f
5087;; Handling of small scale constructs like types and names.
5088
d9e94c22
MS
5089;; Dynamically bound variable that instructs `c-forward-type' to also
5090;; treat possible types (i.e. those that it normally returns 'maybe or
5091;; 'found for) as actual types (and always return 'found for them).
5092;; This means that it records them in `c-record-type-identifiers' if
5093;; that is set, and that it adds them to `c-found-types'.
5094(defvar c-promote-possible-types nil)
5095
0386b551
AM
5096;; Dynamically bound variable that instructs `c-forward-<>-arglist' to
5097;; mark up successfully parsed arglists with paren syntax properties on
5098;; the surrounding angle brackets and with `c-<>-arg-sep' in the
5099;; `c-type' property of each argument separating comma.
5100;;
5101;; Setting this variable also makes `c-forward-<>-arglist' recurse into
5102;; all arglists for side effects (i.e. recording types), otherwise it
5103;; exploits any existing paren syntax properties to quickly jump to the
5104;; end of already parsed arglists.
5105;;
5106;; Marking up the arglists is not the default since doing that correctly
5107;; depends on a proper value for `c-restricted-<>-arglists'.
5108(defvar c-parse-and-markup-<>-arglists nil)
5109
d9e94c22 5110;; Dynamically bound variable that instructs `c-forward-<>-arglist' to
037558bf
MS
5111;; not accept arglists that contain binary operators.
5112;;
5113;; This is primarily used to handle C++ template arglists. C++
5114;; disambiguates them by checking whether the preceding name is a
5115;; template or not. We can't do that, so we assume it is a template
5116;; if it can be parsed as one. That usually works well since
5117;; comparison expressions on the forms "a < b > c" or "a < b, c > d"
5118;; in almost all cases would be pointless.
5119;;
5120;; However, in function arglists, e.g. in "foo (a < b, c > d)", we
5121;; should let the comma separate the function arguments instead. And
5122;; in a context where the value of the expression is taken, e.g. in
5123;; "if (a < b || c > d)", it's probably not a template.
5124(defvar c-restricted-<>-arglists nil)
d9e94c22 5125
0386b551
AM
5126;; Dynamically bound variables that instructs
5127;; `c-forward-keyword-clause', `c-forward-<>-arglist',
5128;; `c-forward-name', `c-forward-type', `c-forward-decl-or-cast-1', and
5129;; `c-forward-label' to record the ranges of all the type and
5130;; reference identifiers they encounter. They will build lists on
5131;; these variables where each element is a cons of the buffer
5132;; positions surrounding each identifier. This recording is only
5133;; activated when `c-record-type-identifiers' is non-nil.
d9e94c22
MS
5134;;
5135;; All known types that can't be identifiers are recorded, and also
5136;; other possible types if `c-promote-possible-types' is set.
5137;; Recording is however disabled inside angle bracket arglists that
5138;; are encountered inside names and other angle bracket arglists.
0386b551 5139;; Such occurrences are taken care of by `c-font-lock-<>-arglists'
d9e94c22
MS
5140;; instead.
5141;;
5142;; Only the names in C++ template style references (e.g. "tmpl" in
5143;; "tmpl<a,b>::foo") are recorded as references, other references
5144;; aren't handled here.
0386b551
AM
5145;;
5146;; `c-forward-label' records the label identifier(s) on
5147;; `c-record-ref-identifiers'.
d9e94c22
MS
5148(defvar c-record-type-identifiers nil)
5149(defvar c-record-ref-identifiers nil)
5150
0386b551
AM
5151;; This variable will receive a cons cell of the range of the last
5152;; single identifier symbol stepped over by `c-forward-name' if it's
5153;; successful. This is the range that should be put on one of the
5154;; record lists above by the caller. It's assigned nil if there's no
5155;; such symbol in the name.
d9e94c22
MS
5156(defvar c-last-identifier-range nil)
5157
5158(defmacro c-record-type-id (range)
5159 (if (eq (car-safe range) 'cons)
5160 ;; Always true.
5161 `(setq c-record-type-identifiers
5162 (cons ,range c-record-type-identifiers))
5163 `(let ((range ,range))
5164 (if range
5165 (setq c-record-type-identifiers
5166 (cons range c-record-type-identifiers))))))
5167
5168(defmacro c-record-ref-id (range)
5169 (if (eq (car-safe range) 'cons)
5170 ;; Always true.
5171 `(setq c-record-ref-identifiers
5172 (cons ,range c-record-ref-identifiers))
5173 `(let ((range ,range))
5174 (if range
5175 (setq c-record-ref-identifiers
5176 (cons range c-record-ref-identifiers))))))
5177
5178;; Dynamically bound variable that instructs `c-forward-type' to
5179;; record the ranges of types that only are found. Behaves otherwise
5180;; like `c-record-type-identifiers'.
5181(defvar c-record-found-types nil)
5182
5183(defmacro c-forward-keyword-prefixed-id (type)
5184 ;; Used internally in `c-forward-keyword-clause' to move forward
5185 ;; over a type (if TYPE is 'type) or a name (otherwise) which
5186 ;; possibly is prefixed by keywords and their associated clauses.
5187 ;; Try with a type/name first to not trip up on those that begin
5188 ;; with a keyword. Return t if a known or found type is moved
5189 ;; over. The point is clobbered if nil is returned. If range
5190 ;; recording is enabled, the identifier is recorded on as a type
5191 ;; if TYPE is 'type or as a reference if TYPE is 'ref.
0386b551
AM
5192 ;;
5193 ;; This macro might do hidden buffer changes.
d9e94c22
MS
5194 `(let (res)
5195 (while (if (setq res ,(if (eq type 'type)
5196 `(c-forward-type)
5197 `(c-forward-name)))
5198 nil
5199 (and (looking-at c-keywords-regexp)
0386b551 5200 (c-forward-keyword-clause 1))))
d9e94c22
MS
5201 (when (memq res '(t known found prefix))
5202 ,(when (eq type 'ref)
5203 `(when c-record-type-identifiers
5204 (c-record-ref-id c-last-identifier-range)))
5205 t)))
5206
0386b551 5207(defmacro c-forward-id-comma-list (type update-safe-pos)
d9e94c22
MS
5208 ;; Used internally in `c-forward-keyword-clause' to move forward
5209 ;; over a comma separated list of types or names using
5210 ;; `c-forward-keyword-prefixed-id'.
0386b551
AM
5211 ;;
5212 ;; This macro might do hidden buffer changes.
d9e94c22 5213 `(while (and (progn
0386b551
AM
5214 ,(when update-safe-pos
5215 `(setq safe-pos (point)))
d9e94c22
MS
5216 (eq (char-after) ?,))
5217 (progn
5218 (forward-char)
5219 (c-forward-syntactic-ws)
5220 (c-forward-keyword-prefixed-id ,type)))))
5221
0386b551
AM
5222(defun c-forward-keyword-clause (match)
5223 ;; Submatch MATCH in the current match data is assumed to surround a
5224 ;; token. If it's a keyword, move over it and any immediately
5225 ;; following clauses associated with it, stopping at the start of
5226 ;; the next token. t is returned in that case, otherwise the point
d9e94c22
MS
5227 ;; stays and nil is returned. The kind of clauses that are
5228 ;; recognized are those specified by `c-type-list-kwds',
5229 ;; `c-ref-list-kwds', `c-colon-type-list-kwds',
5230 ;; `c-paren-nontype-kwds', `c-paren-type-kwds', `c-<>-type-kwds',
5231 ;; and `c-<>-arglist-kwds'.
0386b551
AM
5232 ;;
5233 ;; This function records identifier ranges on
5234 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
5235 ;; `c-record-type-identifiers' is non-nil.
5236 ;;
5237 ;; Note that for `c-colon-type-list-kwds', which doesn't necessary
5238 ;; apply directly after the keyword, the type list is moved over
5239 ;; only when there is no unaccounted token before it (i.e. a token
5240 ;; that isn't moved over due to some other keyword list). The
5241 ;; identifier ranges in the list are still recorded if that should
5242 ;; be done, though.
5243 ;;
5244 ;; This function might do hidden buffer changes.
5245
5246 (let ((kwd-sym (c-keyword-sym (match-string match))) safe-pos pos
5247 ;; The call to `c-forward-<>-arglist' below is made after
5248 ;; `c-<>-sexp-kwds' keywords, so we're certain they actually
5249 ;; are angle bracket arglists and `c-restricted-<>-arglists'
5250 ;; should therefore be nil.
5251 (c-parse-and-markup-<>-arglists t)
5252 c-restricted-<>-arglists)
d9e94c22 5253
d9e94c22 5254 (when kwd-sym
0386b551 5255 (goto-char (match-end match))
d9e94c22
MS
5256 (c-forward-syntactic-ws)
5257 (setq safe-pos (point))
a66cd3ee 5258
d9e94c22
MS
5259 (cond
5260 ((and (c-keyword-member kwd-sym 'c-type-list-kwds)
5261 (c-forward-keyword-prefixed-id type))
5262 ;; There's a type directly after a keyword in `c-type-list-kwds'.
0386b551 5263 (c-forward-id-comma-list type t))
d9e94c22
MS
5264
5265 ((and (c-keyword-member kwd-sym 'c-ref-list-kwds)
5266 (c-forward-keyword-prefixed-id ref))
5267 ;; There's a name directly after a keyword in `c-ref-list-kwds'.
0386b551 5268 (c-forward-id-comma-list ref t))
d9e94c22
MS
5269
5270 ((and (c-keyword-member kwd-sym 'c-paren-any-kwds)
5271 (eq (char-after) ?\())
5272 ;; There's an open paren after a keyword in `c-paren-any-kwds'.
5273
5274 (forward-char)
5275 (when (and (setq pos (c-up-list-forward))
5276 (eq (char-before pos) ?\)))
5277 (when (and c-record-type-identifiers
5278 (c-keyword-member kwd-sym 'c-paren-type-kwds))
5279 ;; Use `c-forward-type' on every identifier we can find
5280 ;; inside the paren, to record the types.
5281 (while (c-syntactic-re-search-forward c-symbol-start pos t)
5282 (goto-char (match-beginning 0))
5283 (unless (c-forward-type)
5284 (looking-at c-symbol-key) ; Always matches.
5285 (goto-char (match-end 0)))))
5286
5287 (goto-char pos)
5288 (c-forward-syntactic-ws)
5289 (setq safe-pos (point))))
5290
5291 ((and (c-keyword-member kwd-sym 'c-<>-sexp-kwds)
5292 (eq (char-after) ?<)
0386b551 5293 (c-forward-<>-arglist (c-keyword-member kwd-sym 'c-<>-type-kwds)))
d9e94c22
MS
5294 (c-forward-syntactic-ws)
5295 (setq safe-pos (point)))
5296
5297 ((and (c-keyword-member kwd-sym 'c-nonsymbol-sexp-kwds)
449a2b0d
MS
5298 (not (looking-at c-symbol-start))
5299 (c-safe (c-forward-sexp) t))
d9e94c22
MS
5300 (c-forward-syntactic-ws)
5301 (setq safe-pos (point))))
5302
0386b551
AM
5303 (when (c-keyword-member kwd-sym 'c-colon-type-list-kwds)
5304 (if (eq (char-after) ?:)
5305 ;; If we are at the colon already, we move over the type
5306 ;; list after it.
5307 (progn
5308 (forward-char)
5309 (c-forward-syntactic-ws)
5310 (when (c-forward-keyword-prefixed-id type)
5311 (c-forward-id-comma-list type t)))
5312 ;; Not at the colon, so stop here. But the identifier
5313 ;; ranges in the type list later on should still be
5314 ;; recorded.
5315 (and c-record-type-identifiers
5316 (progn
5317 ;; If a keyword matched both one of the types above and
5318 ;; this one, we match `c-colon-type-list-re' after the
5319 ;; clause matched above.
5320 (goto-char safe-pos)
5321 (looking-at c-colon-type-list-re))
5322 (progn
5323 (goto-char (match-end 0))
5324 (c-forward-syntactic-ws)
5325 (c-forward-keyword-prefixed-id type))
5326 ;; There's a type after the `c-colon-type-list-re' match
5327 ;; after a keyword in `c-colon-type-list-kwds'.
5328 (c-forward-id-comma-list type nil))))
d9e94c22
MS
5329
5330 (goto-char safe-pos)
5331 t)))
5332
0386b551
AM
5333(defun c-forward-<>-arglist (all-types)
5334 ;; The point is assumed to be at a "<". Try to treat it as the open
b4dc7d98 5335 ;; paren of an angle bracket arglist and move forward to the
0386b551
AM
5336 ;; corresponding ">". If successful, the point is left after the
5337 ;; ">" and t is returned, otherwise the point isn't moved and nil is
d9e94c22
MS
5338 ;; returned. If ALL-TYPES is t then all encountered arguments in
5339 ;; the arglist that might be types are treated as found types.
5340 ;;
0386b551
AM
5341 ;; The variable `c-parse-and-markup-<>-arglists' controls how this
5342 ;; function handles text properties on the angle brackets and argument
5343 ;; separating commas.
d9e94c22 5344 ;;
0386b551
AM
5345 ;; `c-restricted-<>-arglists' controls how lenient the template
5346 ;; arglist recognition should be.
5347 ;;
5348 ;; This function records identifier ranges on
5349 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
5350 ;; `c-record-type-identifiers' is non-nil.
5351 ;;
5352 ;; This function might do hidden buffer changes.
d9e94c22
MS
5353
5354 (let ((start (point))
5355 ;; If `c-record-type-identifiers' is set then activate
5356 ;; recording of any found types that constitute an argument in
5357 ;; the arglist.
5358 (c-record-found-types (if c-record-type-identifiers t)))
5359 (if (catch 'angle-bracket-arglist-escape
5360 (setq c-record-found-types
0386b551 5361 (c-forward-<>-arglist-recur all-types)))
d9e94c22
MS
5362 (progn
5363 (when (consp c-record-found-types)
5364 (setq c-record-type-identifiers
5365 ;; `nconc' doesn't mind that the tail of
5366 ;; `c-record-found-types' is t.
5367 (nconc c-record-found-types c-record-type-identifiers)))
5368 t)
5369
5370 (goto-char start)
a66cd3ee 5371 nil)))
785eecbb 5372
0386b551 5373(defun c-forward-<>-arglist-recur (all-types)
d9e94c22 5374 ;; Recursive part of `c-forward-<>-arglist'.
0386b551
AM
5375 ;;
5376 ;; This function might do hidden buffer changes.
d9e94c22
MS
5377
5378 (let ((start (point)) res pos tmp
5379 ;; Cover this so that any recorded found type ranges are
5380 ;; automatically lost if it turns out to not be an angle
5381 ;; bracket arglist. It's propagated through the return value
5382 ;; on successful completion.
5383 (c-record-found-types c-record-found-types)
5384 ;; List that collects the positions after the argument
5385 ;; separating ',' in the arglist.
5386 arg-start-pos)
5387
0386b551
AM
5388 ;; If the '<' has paren open syntax then we've marked it as an angle
5389 ;; bracket arglist before, so skip to the end.
5390 (if (and (not c-parse-and-markup-<>-arglists)
5391 (c-get-char-property (point) 'syntax-table))
5392
5393 (progn
5394 (forward-char)
5395 (if (and (c-go-up-list-forward)
5396 (eq (char-before) ?>))
5397 t
5398
5399 ;; Got unmatched paren angle brackets. We don't clear the paren
5400 ;; syntax properties and retry, on the basis that it's very
5401 ;; unlikely that paren angle brackets become operators by code
5402 ;; manipulation. It's far more likely that it doesn't match due
5403 ;; to narrowing or some temporary change.
5404 (goto-char start)
5405 nil))
d9e94c22
MS
5406
5407 (forward-char)
5408 (unless (looking-at c-<-op-cont-regexp)
5409 (while (and
5410 (progn
5411
5412 (when c-record-type-identifiers
5413 (if all-types
5414
5415 ;; All encountered identifiers are types, so set the
5416 ;; promote flag and parse the type.
5417 (progn
5418 (c-forward-syntactic-ws)
5419 (when (looking-at c-identifier-start)
5420 (let ((c-promote-possible-types t))
5421 (c-forward-type))))
5422
5423 ;; Check if this arglist argument is a sole type. If
5424 ;; it's known then it's recorded in
5425 ;; `c-record-type-identifiers'. If it only is found
5426 ;; then it's recorded in `c-record-found-types' which we
5427 ;; might roll back if it turns out that this isn't an
5428 ;; angle bracket arglist afterall.
5429 (when (memq (char-before) '(?, ?<))
5430 (let ((orig-record-found-types c-record-found-types))
5431 (c-forward-syntactic-ws)
5432 (and (memq (c-forward-type) '(known found))
5433 (not (looking-at "[,>]"))
5434 ;; A found type was recorded but it's not the
5435 ;; only thing in the arglist argument, so reset
5436 ;; `c-record-found-types'.
5437 (setq c-record-found-types
5438 orig-record-found-types))))))
5439
5440 (setq pos (point))
5441 (or (when (eq (char-after) ?>)
5442 ;; Must check for '>' at the very start separately,
5443 ;; since the regexp below has to avoid ">>" without
5444 ;; using \\=.
5445 (forward-char)
5446 t)
5447
037558bf
MS
5448 ;; Note: These regexps exploit the match order in \| so
5449 ;; that "<>" is matched by "<" rather than "[^>:-]>".
d9e94c22 5450 (c-syntactic-re-search-forward
037558bf
MS
5451 (if c-restricted-<>-arglists
5452 ;; Stop on ',', '|', '&', '+' and '-' to catch
5453 ;; common binary operators that could be between
5454 ;; two comparison expressions "a<b" and "c>d".
5455 "[<;{},|&+-]\\|\\([^>:-]>\\)"
5456 ;; Otherwise we still stop on ',' to find the
5457 ;; argument start positions.
5458 "[<;{},]\\|\\([^>:-]>\\)")
5459 nil 'move t t 1)
d9e94c22
MS
5460
5461 ;; If the arglist starter has lost its open paren
5462 ;; syntax but not the closer, we won't find the
5463 ;; closer above since we only search in the
5464 ;; balanced sexp. In that case we stop just short
5465 ;; of it so check if the following char is the closer.
5466 (when (eq (char-after) ?>)
d9e94c22
MS
5467 (forward-char)
5468 t)))
5469
5470 (cond
5471 ((eq (char-before) ?>)
5472 ;; Either an operator starting with '>' or the end of
5473 ;; the angle bracket arglist.
5474
0386b551 5475 (if (looking-at c->-op-cont-regexp)
d9e94c22 5476 (progn
0386b551
AM
5477 (goto-char (match-end 0))
5478 t) ; Continue the loop.
d9e94c22 5479
0386b551
AM
5480 ;; The angle bracket arglist is finished.
5481 (when c-parse-and-markup-<>-arglists
d9e94c22 5482 (while arg-start-pos
0386b551
AM
5483 (c-put-c-type-property (1- (car arg-start-pos))
5484 'c-<>-arg-sep)
d9e94c22
MS
5485 (setq arg-start-pos (cdr arg-start-pos)))
5486 (c-mark-<-as-paren start)
0386b551
AM
5487 (c-mark->-as-paren (1- (point))))
5488 (setq res t)
5489 nil)) ; Exit the loop.
d9e94c22
MS
5490
5491 ((eq (char-before) ?<)
5492 ;; Either an operator starting with '<' or a nested arglist.
5493
5494 (setq pos (point))
5495 (let (id-start id-end subres keyword-match)
5496 (if (if (looking-at c-<-op-cont-regexp)
5497 (setq tmp (match-end 0))
5498 (setq tmp pos)
5499 (backward-char)
5500 (not
5501 (and
5502
5503 (save-excursion
0386b551 5504 ;; There's always an identifier before an angle
d9e94c22
MS
5505 ;; bracket arglist, or a keyword in
5506 ;; `c-<>-type-kwds' or `c-<>-arglist-kwds'.
5507 (c-backward-syntactic-ws)
5508 (setq id-end (point))
5509 (c-simple-skip-symbol-backward)
5510 (when (or (setq keyword-match
5511 (looking-at c-opt-<>-sexp-key))
5512 (not (looking-at c-keywords-regexp)))
5513 (setq id-start (point))))
5514
5515 (setq subres
5516 (let ((c-record-type-identifiers nil)
5517 (c-record-found-types nil))
5518 (c-forward-<>-arglist-recur
5519 (and keyword-match
5520 (c-keyword-member
5521 (c-keyword-sym (match-string 1))
0386b551 5522 'c-<>-type-kwds)))))
d9e94c22
MS
5523 )))
5524
5525 ;; It was not an angle bracket arglist.
0386b551 5526 (goto-char tmp)
d9e94c22
MS
5527
5528 ;; It was an angle bracket arglist.
5529 (setq c-record-found-types subres)
5530
5531 ;; Record the identifier before the template as a type
5532 ;; or reference depending on whether the arglist is last
5533 ;; in a qualified identifier.
5534 (when (and c-record-type-identifiers
5535 (not keyword-match))
5536 (if (and c-opt-identifier-concat-key
5537 (progn
5538 (c-forward-syntactic-ws)
5539 (looking-at c-opt-identifier-concat-key)))
5540 (c-record-ref-id (cons id-start id-end))
5541 (c-record-type-id (cons id-start id-end))))))
5542 t)
5543
5544 ((and (eq (char-before) ?,)
037558bf 5545 (not c-restricted-<>-arglists))
d9e94c22
MS
5546 ;; Just another argument. Record the position. The
5547 ;; type check stuff that made us stop at it is at
5548 ;; the top of the loop.
5549 (setq arg-start-pos (cons (point) arg-start-pos)))
5550
5551 (t
5552 ;; Got a character that can't be in an angle bracket
5553 ;; arglist argument. Abort using `throw', since
5554 ;; it's useless to try to find a surrounding arglist
5555 ;; if we're nested.
5556 (throw 'angle-bracket-arglist-escape nil))))))
5557
5558 (if res
5559 (or c-record-found-types t)))))
5560
0386b551
AM
5561(defun c-backward-<>-arglist (all-types &optional limit)
5562 ;; The point is assumed to be directly after a ">". Try to treat it
5563 ;; as the close paren of an angle bracket arglist and move back to
5564 ;; the corresponding "<". If successful, the point is left at
5565 ;; the "<" and t is returned, otherwise the point isn't moved and
5566 ;; nil is returned. ALL-TYPES is passed on to
5567 ;; `c-forward-<>-arglist'.
5568 ;;
5569 ;; If the optional LIMIT is given, it bounds the backward search.
5570 ;; It's then assumed to be at a syntactically relevant position.
5571 ;;
5572 ;; This is a wrapper around `c-forward-<>-arglist'. See that
5573 ;; function for more details.
5574
5575 (let ((start (point)))
5576 (backward-char)
5577 (if (and (not c-parse-and-markup-<>-arglists)
5578 (c-get-char-property (point) 'syntax-table))
5579
5580 (if (and (c-go-up-list-backward)
5581 (eq (char-after) ?<))
5582 t
5583 ;; See corresponding note in `c-forward-<>-arglist'.
5584 (goto-char start)
5585 nil)
5586
51c9af45 5587 (while (progn
0386b551
AM
5588 (c-syntactic-skip-backward "^<;{}" limit t)
5589
51c9af45
AM
5590 (and
5591 (if (eq (char-before) ?<)
5592 t
5593 ;; Stopped at bob or a char that isn't allowed in an
5594 ;; arglist, so we've failed.
5595 (goto-char start)
5596 nil)
0386b551 5597
51c9af45
AM
5598 (if (> (point)
5599 (progn (c-beginning-of-current-token)
5600 (point)))
5601 ;; If we moved then the "<" was part of some
5602 ;; multicharacter token.
5603 t
0386b551 5604
51c9af45
AM
5605 (backward-char)
5606 (let ((beg-pos (point)))
5607 (if (c-forward-<>-arglist all-types)
5608 (cond ((= (point) start)
5609 ;; Matched the arglist. Break the while.
5610 (goto-char beg-pos)
5611 nil)
5612 ((> (point) start)
5613 ;; We started from a non-paren ">" inside an
5614 ;; arglist.
5615 (goto-char start)
5616 nil)
5617 (t
5618 ;; Matched a shorter arglist. Can be a nested
5619 ;; one so continue looking.
5620 (goto-char beg-pos)
5621 t))
5622 t))))))
0386b551
AM
5623
5624 (/= (point) start))))
5625
d9e94c22
MS
5626(defun c-forward-name ()
5627 ;; Move forward over a complete name if at the beginning of one,
5628 ;; stopping at the next following token. If the point is not at
5629 ;; something that are recognized as name then it stays put. A name
5630 ;; could be something as simple as "foo" in C or something as
5631 ;; complex as "X<Y<class A<int>::B, BIT_MAX >> b>, ::operator<> ::
5632 ;; Z<(a>b)> :: operator const X<&foo>::T Q::G<unsigned short
5633 ;; int>::*volatile const" in C++ (this function is actually little
5634 ;; more than a `looking-at' call in all modes except those that,
5635 ;; like C++, have `c-recognize-<>-arglists' set). Return nil if no
5636 ;; name is found, 'template if it's an identifier ending with an
5637 ;; angle bracket arglist, 'operator of it's an operator identifier,
5638 ;; or t if it's some other kind of name.
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.
d9e94c22 5645
0386b551 5646 (let ((pos (point)) (start (point)) res id-start id-end
d9e94c22
MS
5647 ;; Turn off `c-promote-possible-types' here since we might
5648 ;; call `c-forward-<>-arglist' and we don't want it to promote
5649 ;; every suspect thing in the arglist to a type. We're
5650 ;; typically called from `c-forward-type' in this case, and
5651 ;; the caller only wants the top level type that it finds to
5652 ;; be promoted.
5653 c-promote-possible-types)
5654 (while
5655 (and
5656 (looking-at c-identifier-key)
5657
5658 (progn
5659 ;; Check for keyword. We go to the last symbol in
5660 ;; `c-identifier-key' first.
0386b551
AM
5661 (goto-char (setq id-end (match-end 0)))
5662 (c-simple-skip-symbol-backward)
5663 (setq id-start (point))
d9e94c22
MS
5664
5665 (if (looking-at c-keywords-regexp)
5666 (when (and (c-major-mode-is 'c++-mode)
5667 (looking-at
5668 (cc-eval-when-compile
5669 (concat "\\(operator\\|\\(template\\)\\)"
5670 "\\(" (c-lang-const c-nonsymbol-key c++)
5671 "\\|$\\)")))
5672 (if (match-beginning 2)
5673 ;; "template" is only valid inside an
5674 ;; identifier if preceded by "::".
5675 (save-excursion
5676 (c-backward-syntactic-ws)
5677 (and (c-safe (backward-char 2) t)
5678 (looking-at "::")))
5679 t))
5680
5681 ;; Handle a C++ operator or template identifier.
5682 (goto-char id-end)
5683 (c-forward-syntactic-ws)
5684 (cond ((eq (char-before id-end) ?e)
5685 ;; Got "... ::template".
5686 (let ((subres (c-forward-name)))
5687 (when subres
5688 (setq pos (point)
5689 res subres))))
5690
5691 ((looking-at c-identifier-start)
5692 ;; Got a cast operator.
5693 (when (c-forward-type)
5694 (setq pos (point)
5695 res 'operator)
5696 ;; Now we should match a sequence of either
5697 ;; '*', '&' or a name followed by ":: *",
5698 ;; where each can be followed by a sequence
5699 ;; of `c-opt-type-modifier-key'.
5700 (while (cond ((looking-at "[*&]")
5701 (goto-char (match-end 0))
5702 t)
5703 ((looking-at c-identifier-start)
5704 (and (c-forward-name)
5705 (looking-at "::")
5706 (progn
5707 (goto-char (match-end 0))
5708 (c-forward-syntactic-ws)
5709 (eq (char-after) ?*))
5710 (progn
5711 (forward-char)
5712 t))))
5713 (while (progn
5714 (c-forward-syntactic-ws)
5715 (setq pos (point))
5716 (looking-at c-opt-type-modifier-key))
5717 (goto-char (match-end 1))))))
5718
5719 ((looking-at c-overloadable-operators-regexp)
5720 ;; Got some other operator.
0386b551
AM
5721 (setq c-last-identifier-range
5722 (cons (point) (match-end 0)))
d9e94c22
MS
5723 (goto-char (match-end 0))
5724 (c-forward-syntactic-ws)
5725 (setq pos (point)
5726 res 'operator)))
5727
5728 nil)
5729
0386b551
AM
5730 ;; `id-start' is equal to `id-end' if we've jumped over
5731 ;; an identifier that doesn't end with a symbol token.
5732 ;; That can occur e.g. for Java import directives on the
5733 ;; form "foo.bar.*".
5734 (when (and id-start (/= id-start id-end))
d9e94c22
MS
5735 (setq c-last-identifier-range
5736 (cons id-start id-end)))
5737 (goto-char id-end)
5738 (c-forward-syntactic-ws)
5739 (setq pos (point)
5740 res t)))
5741
5742 (progn
5743 (goto-char pos)
5744 (when (or c-opt-identifier-concat-key
5745 c-recognize-<>-arglists)
5746
5747 (cond
5748 ((and c-opt-identifier-concat-key
5749 (looking-at c-opt-identifier-concat-key))
5750 ;; Got a concatenated identifier. This handles the
5751 ;; cases with tricky syntactic whitespace that aren't
5752 ;; covered in `c-identifier-key'.
5753 (goto-char (match-end 0))
5754 (c-forward-syntactic-ws)
5755 t)
5756
5757 ((and c-recognize-<>-arglists
5758 (eq (char-after) ?<))
5759 ;; Maybe an angle bracket arglist.
0386b551
AM
5760
5761 (when (let (c-record-type-identifiers
5762 c-record-found-types)
5763 (c-forward-<>-arglist nil))
5764
5765 (c-add-type start (1+ pos))
d9e94c22 5766 (c-forward-syntactic-ws)
0386b551
AM
5767 (setq pos (point)
5768 c-last-identifier-range nil)
5769
d9e94c22
MS
5770 (if (and c-opt-identifier-concat-key
5771 (looking-at c-opt-identifier-concat-key))
0386b551 5772
d9e94c22
MS
5773 ;; Continue if there's an identifier concatenation
5774 ;; operator after the template argument.
5775 (progn
0386b551
AM
5776 (when (and c-record-type-identifiers id-start)
5777 (c-record-ref-id (cons id-start id-end)))
d9e94c22
MS
5778 (forward-char 2)
5779 (c-forward-syntactic-ws)
5780 t)
0386b551
AM
5781
5782 (when (and c-record-type-identifiers id-start)
5783 (c-record-type-id (cons id-start id-end)))
d9e94c22
MS
5784 (setq res 'template)
5785 nil)))
5786 )))))
5787
5788 (goto-char pos)
5789 res))
5790
5791(defun c-forward-type ()
5792 ;; Move forward over a type spec if at the beginning of one,
5793 ;; stopping at the next following token. Return t if it's a known
2a15eb73
MS
5794 ;; type that can't be a name or other expression, 'known if it's an
5795 ;; otherwise known type (according to `*-font-lock-extra-types'),
5796 ;; 'prefix if it's a known prefix of a type, 'found if it's a type
5797 ;; that matches one in `c-found-types', 'maybe if it's an identfier
5798 ;; that might be a type, or nil if it can't be a type (the point
5799 ;; isn't moved then). The point is assumed to be at the beginning
5800 ;; of a token.
d9e94c22
MS
5801 ;;
5802 ;; Note that this function doesn't skip past the brace definition
5803 ;; that might be considered part of the type, e.g.
5804 ;; "enum {a, b, c} foo".
0386b551
AM
5805 ;;
5806 ;; This function records identifier ranges on
5807 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
5808 ;; `c-record-type-identifiers' is non-nil.
5809 ;;
5810 ;; This function might do hidden buffer changes.
5811
5812 (let ((start (point)) pos res name-res id-start id-end id-range)
d9e94c22
MS
5813
5814 ;; Skip leading type modifiers. If any are found we know it's a
5815 ;; prefix of a type.
5816 (when c-opt-type-modifier-key
5817 (while (looking-at c-opt-type-modifier-key)
5818 (goto-char (match-end 1))
5819 (c-forward-syntactic-ws)
5820 (setq res 'prefix)))
5821
5822 (cond
5823 ((looking-at c-type-prefix-key)
5824 ;; Looking at a keyword that prefixes a type identifier,
5825 ;; e.g. "class".
5826 (goto-char (match-end 1))
5827 (c-forward-syntactic-ws)
5828 (setq pos (point))
0386b551 5829 (if (memq (setq name-res (c-forward-name)) '(t template))
d9e94c22 5830 (progn
0386b551 5831 (when (eq name-res t)
d9e94c22
MS
5832 ;; In many languages the name can be used without the
5833 ;; prefix, so we add it to `c-found-types'.
5834 (c-add-type pos (point))
0386b551
AM
5835 (when (and c-record-type-identifiers
5836 c-last-identifier-range)
d9e94c22
MS
5837 (c-record-type-id c-last-identifier-range)))
5838 (setq res t))
5839 ;; Invalid syntax.
5840 (goto-char start)
5841 (setq res nil)))
5842
5843 ((progn
5844 (setq pos nil)
5845 (if (looking-at c-identifier-start)
5846 (save-excursion
5847 (setq id-start (point)
0386b551
AM
5848 name-res (c-forward-name))
5849 (when name-res
d9e94c22
MS
5850 (setq id-end (point)
5851 id-range c-last-identifier-range))))
5852 (and (cond ((looking-at c-primitive-type-key)
5853 (setq res t))
5854 ((c-with-syntax-table c-identifier-syntax-table
5855 (looking-at c-known-type-key))
5856 (setq res 'known)))
5857 (or (not id-end)
5858 (>= (save-excursion
5859 (save-match-data
5860 (goto-char (match-end 1))
5861 (c-forward-syntactic-ws)
5862 (setq pos (point))))
5863 id-end)
5864 (setq res nil))))
5865 ;; Looking at a primitive or known type identifier. We've
5866 ;; checked for a name first so that we don't go here if the
5867 ;; known type match only is a prefix of another name.
5868
5869 (setq id-end (match-end 1))
5870
5871 (when (and c-record-type-identifiers
5872 (or c-promote-possible-types (eq res t)))
5873 (c-record-type-id (cons (match-beginning 1) (match-end 1))))
5874
5875 (if (and c-opt-type-component-key
5876 (save-match-data
5877 (looking-at c-opt-type-component-key)))
5878 ;; There might be more keywords for the type.
5879 (let (safe-pos)
0386b551 5880 (c-forward-keyword-clause 1)
d9e94c22
MS
5881 (while (progn
5882 (setq safe-pos (point))
5883 (looking-at c-opt-type-component-key))
5884 (when (and c-record-type-identifiers
5885 (looking-at c-primitive-type-key))
5886 (c-record-type-id (cons (match-beginning 1)
5887 (match-end 1))))
0386b551 5888 (c-forward-keyword-clause 1))
d9e94c22
MS
5889 (if (looking-at c-primitive-type-key)
5890 (progn
5891 (when c-record-type-identifiers
5892 (c-record-type-id (cons (match-beginning 1)
5893 (match-end 1))))
0386b551 5894 (c-forward-keyword-clause 1)
d9e94c22
MS
5895 (setq res t))
5896 (goto-char safe-pos)
5897 (setq res 'prefix)))
0386b551 5898 (unless (save-match-data (c-forward-keyword-clause 1))
d9e94c22
MS
5899 (if pos
5900 (goto-char pos)
5901 (goto-char (match-end 1))
5902 (c-forward-syntactic-ws)))))
5903
0386b551
AM
5904 (name-res
5905 (cond ((eq name-res t)
d9e94c22
MS
5906 ;; A normal identifier.
5907 (goto-char id-end)
5908 (if (or res c-promote-possible-types)
5909 (progn
5910 (c-add-type id-start id-end)
0386b551 5911 (when (and c-record-type-identifiers id-range)
d9e94c22
MS
5912 (c-record-type-id id-range))
5913 (unless res
5914 (setq res 'found)))
5915 (setq res (if (c-check-type id-start id-end)
5916 ;; It's an identifier that has been used as
5917 ;; a type somewhere else.
5918 'found
5919 ;; It's an identifier that might be a type.
5920 'maybe))))
0386b551 5921 ((eq name-res 'template)
d9e94c22
MS
5922 ;; A template is a type.
5923 (goto-char id-end)
5924 (setq res t))
5925 (t
5926 ;; Otherwise it's an operator identifier, which is not a type.
5927 (goto-char start)
5928 (setq res nil)))))
5929
5930 (when res
5931 ;; Skip trailing type modifiers. If any are found we know it's
5932 ;; a type.
5933 (when c-opt-type-modifier-key
5934 (while (looking-at c-opt-type-modifier-key)
5935 (goto-char (match-end 1))
5936 (c-forward-syntactic-ws)
5937 (setq res t)))
5938
5939 ;; Step over any type suffix operator. Do not let the existence
5940 ;; of these alter the classification of the found type, since
5941 ;; these operators typically are allowed in normal expressions
5942 ;; too.
5943 (when c-opt-type-suffix-key
5944 (while (looking-at c-opt-type-suffix-key)
5945 (goto-char (match-end 1))
5946 (c-forward-syntactic-ws)))
5947
5948 (when c-opt-type-concat-key
0386b551
AM
5949 ;; Look for a trailing operator that concatenates the type
5950 ;; with a following one, and if so step past that one through
5951 ;; a recursive call. Note that we don't record concatenated
5952 ;; types in `c-found-types' - it's the component types that
5953 ;; are recorded when appropriate.
d9e94c22
MS
5954 (setq pos (point))
5955 (let* ((c-promote-possible-types (or (memq res '(t known))
5956 c-promote-possible-types))
5957 ;; If we can't promote then set `c-record-found-types' so that
5958 ;; we can merge in the types from the second part afterwards if
5959 ;; it turns out to be a known type there.
5960 (c-record-found-types (and c-record-type-identifiers
0386b551
AM
5961 (not c-promote-possible-types)))
5962 subres)
d9e94c22
MS
5963 (if (and (looking-at c-opt-type-concat-key)
5964
5965 (progn
5966 (goto-char (match-end 1))
5967 (c-forward-syntactic-ws)
0386b551 5968 (setq subres (c-forward-type))))
d9e94c22
MS
5969
5970 (progn
5971 ;; If either operand certainly is a type then both are, but we
5972 ;; don't let the existence of the operator itself promote two
5973 ;; uncertain types to a certain one.
5974 (cond ((eq res t))
0386b551
AM
5975 ((eq subres t)
5976 (unless (eq name-res 'template)
5977 (c-add-type id-start id-end))
5978 (when (and c-record-type-identifiers id-range)
d9e94c22
MS
5979 (c-record-type-id id-range))
5980 (setq res t))
2a15eb73 5981 ((eq res 'known))
0386b551 5982 ((eq subres 'known)
2a15eb73 5983 (setq res 'known))
d9e94c22 5984 ((eq res 'found))
0386b551 5985 ((eq subres 'found)
d9e94c22
MS
5986 (setq res 'found))
5987 (t
5988 (setq res 'maybe)))
5989
5990 (when (and (eq res t)
5991 (consp c-record-found-types))
5992 ;; Merge in the ranges of any types found by the second
5993 ;; `c-forward-type'.
5994 (setq c-record-type-identifiers
5995 ;; `nconc' doesn't mind that the tail of
5996 ;; `c-record-found-types' is t.
5997 (nconc c-record-found-types
5998 c-record-type-identifiers))))
5999
6000 (goto-char pos))))
6001
6002 (when (and c-record-found-types (memq res '(known found)) id-range)
6003 (setq c-record-found-types
6004 (cons id-range c-record-found-types))))
6005
6006 ;;(message "c-forward-type %s -> %s: %s" start (point) res)
6007
6008 res))
6009
785eecbb 6010\f
d9e94c22
MS
6011;; Handling of large scale constructs like statements and declarations.
6012
0386b551
AM
6013;; Macro used inside `c-forward-decl-or-cast-1'. It ought to be a
6014;; defsubst or perhaps even a defun, but it contains lots of free
6015;; variables that refer to things inside `c-forward-decl-or-cast-1'.
6016(defmacro c-fdoc-shift-type-backward (&optional short)
6017 ;; `c-forward-decl-or-cast-1' can consume an arbitrary length list
6018 ;; of types when parsing a declaration, which means that it
6019 ;; sometimes consumes the identifier in the declaration as a type.
6020 ;; This is used to "backtrack" and make the last type be treated as
6021 ;; an identifier instead.
6022 `(progn
6023 ,(unless short
6024 ;; These identifiers are bound only in the inner let.
6025 '(setq identifier-type at-type
6026 identifier-start type-start
6027 got-parens nil
6028 got-identifier t
6029 got-suffix t
6030 got-suffix-after-parens id-start
6031 paren-depth 0))
6032
6033 (if (setq at-type (if (eq backup-at-type 'prefix)
6034 t
6035 backup-at-type))
6036 (setq type-start backup-type-start
6037 id-start backup-id-start)
6038 (setq type-start start-pos
6039 id-start start-pos))
6040
6041 ;; When these flags already are set we've found specifiers that
6042 ;; unconditionally signal these attributes - backtracking doesn't
6043 ;; change that. So keep them set in that case.
6044 (or at-type-decl
6045 (setq at-type-decl backup-at-type-decl))
6046 (or maybe-typeless
6047 (setq maybe-typeless backup-maybe-typeless))
6048
6049 ,(unless short
6050 ;; This identifier is bound only in the inner let.
6051 '(setq start id-start))))
6052
6053(defun c-forward-decl-or-cast-1 (preceding-token-end context last-cast-end)
6054 ;; Move forward over a declaration or a cast if at the start of one.
6055 ;; The point is assumed to be at the start of some token. Nil is
6056 ;; returned if no declaration or cast is recognized, and the point
6057 ;; is clobbered in that case.
6058 ;;
6059 ;; If a declaration is parsed:
6060 ;;
6061 ;; The point is left at the first token after the first complete
6062 ;; declarator, if there is one. The return value is a cons where
022d0cf4
AM
6063 ;; the car is the position of the first token in the declarator. (See
6064 ;; below for the cdr.)
0386b551
AM
6065 ;; Some examples:
6066 ;;
6067 ;; void foo (int a, char *b) stuff ...
6068 ;; car ^ ^ point
6069 ;; float (*a)[], b;
6070 ;; car ^ ^ point
6071 ;; unsigned int a = c_style_initializer, b;
6072 ;; car ^ ^ point
6073 ;; unsigned int a (cplusplus_style_initializer), b;
6074 ;; car ^ ^ point (might change)
6075 ;; class Foo : public Bar {}
6076 ;; car ^ ^ point
6077 ;; class PikeClass (int a, string b) stuff ...
6078 ;; car ^ ^ point
6079 ;; enum bool;
6080 ;; car ^ ^ point
6081 ;; enum bool flag;
6082 ;; car ^ ^ point
6083 ;; void cplusplus_function (int x) throw (Bad);
6084 ;; car ^ ^ point
6085 ;; Foo::Foo (int b) : Base (b) {}
6086 ;; car ^ ^ point
6087 ;;
022d0cf4
AM
6088 ;; The cdr of the return value is non-nil iff a `c-typedef-decl-kwds'
6089 ;; specifier (e.g. class, struct, enum, typedef) is found in the
6090 ;; declaration, i.e. the declared identifier(s) are types.
0386b551
AM
6091 ;;
6092 ;; If a cast is parsed:
6093 ;;
6094 ;; The point is left at the first token after the closing paren of
6095 ;; the cast. The return value is `cast'. Note that the start
6096 ;; position must be at the first token inside the cast parenthesis
6097 ;; to recognize it.
6098 ;;
6099 ;; PRECEDING-TOKEN-END is the first position after the preceding
6100 ;; token, i.e. on the other side of the syntactic ws from the point.
6101 ;; Use a value less than or equal to (point-min) if the point is at
6102 ;; the first token in (the visible part of) the buffer.
6103 ;;
6104 ;; CONTEXT is a symbol that describes the context at the point:
022d0cf4 6105 ;; 'decl In a comma-separated declaration context (typically
0386b551
AM
6106 ;; inside a function declaration arglist).
6107 ;; '<> In an angle bracket arglist.
6108 ;; 'arglist Some other type of arglist.
a85fd6da
AM
6109 ;; nil Some other context or unknown context. Includes
6110 ;; within the parens of an if, for, ... construct.
0386b551
AM
6111 ;;
6112 ;; LAST-CAST-END is the first token after the closing paren of a
6113 ;; preceding cast, or nil if none is known. If
6114 ;; `c-forward-decl-or-cast-1' is used in succession, it should be
6115 ;; the position after the closest preceding call where a cast was
6116 ;; matched. In that case it's used to discover chains of casts like
6117 ;; "(a) (b) c".
6118 ;;
6119 ;; This function records identifier ranges on
6120 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
6121 ;; `c-record-type-identifiers' is non-nil.
6122 ;;
6123 ;; This function might do hidden buffer changes.
6124
6125 (let (;; `start-pos' is used below to point to the start of the
6126 ;; first type, i.e. after any leading specifiers. It might
6127 ;; also point at the beginning of the preceding syntactic
6128 ;; whitespace.
6129 (start-pos (point))
6130 ;; Set to the result of `c-forward-type'.
6131 at-type
6132 ;; The position of the first token in what we currently
6133 ;; believe is the type in the declaration or cast, after any
6134 ;; specifiers and their associated clauses.
6135 type-start
6136 ;; The position of the first token in what we currently
6137 ;; believe is the declarator for the first identifier. Set
6138 ;; when the type is found, and moved forward over any
6139 ;; `c-decl-hangon-kwds' and their associated clauses that
6140 ;; occurs after the type.
6141 id-start
6142 ;; These store `at-type', `type-start' and `id-start' of the
6143 ;; identifier before the one in those variables. The previous
6144 ;; identifier might turn out to be the real type in a
6145 ;; declaration if the last one has to be the declarator in it.
6146 ;; If `backup-at-type' is nil then the other variables have
6147 ;; undefined values.
6148 backup-at-type backup-type-start backup-id-start
6149 ;; Set if we've found a specifier that makes the defined
6150 ;; identifier(s) types.
6151 at-type-decl
6152 ;; Set if we've found a specifier that can start a declaration
6153 ;; where there's no type.
6154 maybe-typeless
6155 ;; If a specifier is found that also can be a type prefix,
6156 ;; these flags are set instead of those above. If we need to
6157 ;; back up an identifier, they are copied to the real flag
6158 ;; variables. Thus they only take effect if we fail to
6159 ;; interpret it as a type.
6160 backup-at-type-decl backup-maybe-typeless
6161 ;; Whether we've found a declaration or a cast. We might know
6162 ;; this before we've found the type in it. It's 'ids if we've
6163 ;; found two consecutive identifiers (usually a sure sign, but
6164 ;; we should allow that in labels too), and t if we've found a
6165 ;; specifier keyword (a 100% sure sign).
6166 at-decl-or-cast
6167 ;; Set when we need to back up to parse this as a declaration
6168 ;; but not as a cast.
6169 backup-if-not-cast
6170 ;; For casts, the return position.
6171 cast-end
6172 ;; Save `c-record-type-identifiers' and
6173 ;; `c-record-ref-identifiers' since ranges are recorded
6174 ;; speculatively and should be thrown away if it turns out
6175 ;; that it isn't a declaration or cast.
6176 (save-rec-type-ids c-record-type-identifiers)
6177 (save-rec-ref-ids c-record-ref-identifiers))
6178
6179 ;; Check for a type. Unknown symbols are treated as possible
6180 ;; types, but they could also be specifiers disguised through
6181 ;; macros like __INLINE__, so we recognize both types and known
6182 ;; specifiers after them too.
6183 (while
6184 (let* ((start (point)) kwd-sym kwd-clause-end found-type)
6185
6186 ;; Look for a specifier keyword clause.
6187 (when (looking-at c-prefix-spec-kwds-re)
6188 (setq kwd-sym (c-keyword-sym (match-string 1)))
6189 (save-excursion
6190 (c-forward-keyword-clause 1)
6191 (setq kwd-clause-end (point))))
6192
6193 (when (setq found-type (c-forward-type))
6194 ;; Found a known or possible type or a prefix of a known type.
6195
6196 (when at-type
6197 ;; Got two identifiers with nothing but whitespace
6198 ;; between them. That can only happen in declarations.
6199 (setq at-decl-or-cast 'ids)
6200
6201 (when (eq at-type 'found)
6202 ;; If the previous identifier is a found type we
6203 ;; record it as a real one; it might be some sort of
6204 ;; alias for a prefix like "unsigned".
6205 (save-excursion
6206 (goto-char type-start)
6207 (let ((c-promote-possible-types t))
6208 (c-forward-type)))))
6209
6210 (setq backup-at-type at-type
6211 backup-type-start type-start
6212 backup-id-start id-start
6213 at-type found-type
6214 type-start start
6215 id-start (point)
6216 ;; The previous ambiguous specifier/type turned out
6217 ;; to be a type since we've parsed another one after
6218 ;; it, so clear these backup flags.
6219 backup-at-type-decl nil
6220 backup-maybe-typeless nil))
6221
6222 (if kwd-sym
6223 (progn
6224 ;; Handle known specifier keywords and
6225 ;; `c-decl-hangon-kwds' which can occur after known
6226 ;; types.
6227
6228 (if (c-keyword-member kwd-sym 'c-decl-hangon-kwds)
6229 ;; It's a hang-on keyword that can occur anywhere.
6230 (progn
6231 (setq at-decl-or-cast t)
6232 (if at-type
6233 ;; Move the identifier start position if
6234 ;; we've passed a type.
6235 (setq id-start kwd-clause-end)
6236 ;; Otherwise treat this as a specifier and
6237 ;; move the fallback position.
6238 (setq start-pos kwd-clause-end))
6239 (goto-char kwd-clause-end))
6240
6241 ;; It's an ordinary specifier so we know that
6242 ;; anything before this can't be the type.
6243 (setq backup-at-type nil
6244 start-pos kwd-clause-end)
6245
6246 (if found-type
6247 ;; It's ambiguous whether this keyword is a
6248 ;; specifier or a type prefix, so set the backup
6249 ;; flags. (It's assumed that `c-forward-type'
6250 ;; moved further than `c-forward-keyword-clause'.)
6251 (progn
6252 (when (c-keyword-member kwd-sym 'c-typedef-decl-kwds)
6253 (setq backup-at-type-decl t))
6254 (when (c-keyword-member kwd-sym 'c-typeless-decl-kwds)
6255 (setq backup-maybe-typeless t)))
6256
6257 (when (c-keyword-member kwd-sym 'c-typedef-decl-kwds)
6258 (setq at-type-decl t))
6259 (when (c-keyword-member kwd-sym 'c-typeless-decl-kwds)
6260 (setq maybe-typeless t))
6261
6262 ;; Haven't matched a type so it's an umambiguous
6263 ;; specifier keyword and we know we're in a
6264 ;; declaration.
6265 (setq at-decl-or-cast t)
6266
6267 (goto-char kwd-clause-end))))
6268
6269 ;; If the type isn't known we continue so that we'll jump
6270 ;; over all specifiers and type identifiers. The reason
6271 ;; to do this for a known type prefix is to make things
6272 ;; like "unsigned INT16" work.
6273 (and found-type (not (eq found-type t))))))
6274
6275 (cond
6276 ((eq at-type t)
6277 ;; If a known type was found, we still need to skip over any
6278 ;; hangon keyword clauses after it. Otherwise it has already
6279 ;; been done in the loop above.
6280 (while (looking-at c-decl-hangon-key)
6281 (c-forward-keyword-clause 1))
6282 (setq id-start (point)))
6283
6284 ((eq at-type 'prefix)
6285 ;; A prefix type is itself a primitive type when it's not
6286 ;; followed by another type.
6287 (setq at-type t))
6288
6289 ((not at-type)
6290 ;; Got no type but set things up to continue anyway to handle
6291 ;; the various cases when a declaration doesn't start with a
6292 ;; type.
6293 (setq id-start start-pos))
6294
6295 ((and (eq at-type 'maybe)
6296 (c-major-mode-is 'c++-mode))
6297 ;; If it's C++ then check if the last "type" ends on the form
6298 ;; "foo::foo" or "foo::~foo", i.e. if it's the name of a
6299 ;; (con|de)structor.
6300 (save-excursion
6301 (let (name end-2 end-1)
6302 (goto-char id-start)
6303 (c-backward-syntactic-ws)
6304 (setq end-2 (point))
6305 (when (and
6306 (c-simple-skip-symbol-backward)
6307 (progn
6308 (setq name
6309 (buffer-substring-no-properties (point) end-2))
6310 ;; Cheating in the handling of syntactic ws below.
6311 (< (skip-chars-backward ":~ \t\n\r\v\f") 0))
6312 (progn
6313 (setq end-1 (point))
6314 (c-simple-skip-symbol-backward))
6315 (>= (point) type-start)
6316 (equal (buffer-substring-no-properties (point) end-1)
6317 name))
6318 ;; It is a (con|de)structor name. In that case the
6319 ;; declaration is typeless so zap out any preceding
6320 ;; identifier(s) that we might have taken as types.
6321 (goto-char type-start)
6322 (setq at-type nil
6323 backup-at-type nil
6324 id-start type-start))))))
6325
6326 ;; Check for and step over a type decl expression after the thing
6327 ;; that is or might be a type. This can't be skipped since we
6328 ;; need the correct end position of the declarator for
6329 ;; `max-type-decl-end-*'.
6330 (let ((start (point)) (paren-depth 0) pos
6331 ;; True if there's a non-open-paren match of
6332 ;; `c-type-decl-prefix-key'.
6333 got-prefix
6334 ;; True if the declarator is surrounded by a parenthesis pair.
6335 got-parens
6336 ;; True if there is an identifier in the declarator.
6337 got-identifier
6338 ;; True if there's a non-close-paren match of
6339 ;; `c-type-decl-suffix-key'.
6340 got-suffix
6341 ;; True if there's a prefix match outside the outermost
6342 ;; paren pair that surrounds the declarator.
6343 got-prefix-before-parens
1379f2c5 6344 ;; True if there's a suffix match outside the outermost
0386b551
AM
6345 ;; paren pair that surrounds the declarator. The value is
6346 ;; the position of the first suffix match.
6347 got-suffix-after-parens
6348 ;; True if we've parsed the type decl to a token that is
6349 ;; known to end declarations in this context.
6350 at-decl-end
6351 ;; The earlier values of `at-type' and `type-start' if we've
6352 ;; shifted the type backwards.
6353 identifier-type identifier-start
6354 ;; If `c-parse-and-markup-<>-arglists' is set we need to
6355 ;; turn it off during the name skipping below to avoid
6356 ;; getting `c-type' properties that might be bogus. That
6357 ;; can happen since we don't know if
6358 ;; `c-restricted-<>-arglists' will be correct inside the
6359 ;; arglist paren that gets entered.
6360 c-parse-and-markup-<>-arglists)
6361
6362 (goto-char id-start)
6363
6364 ;; Skip over type decl prefix operators. (Note similar code in
6365 ;; `c-font-lock-declarators'.)
6366 (while (and (looking-at c-type-decl-prefix-key)
6367 (if (and (c-major-mode-is 'c++-mode)
6368 (match-beginning 2))
6369 ;; If the second submatch matches in C++ then
6370 ;; we're looking at an identifier that's a
6371 ;; prefix only if it specifies a member pointer.
6372 (when (setq got-identifier (c-forward-name))
6373 (if (looking-at "\\(::\\)")
6374 ;; We only check for a trailing "::" and
6375 ;; let the "*" that should follow be
6376 ;; matched in the next round.
6377 (progn (setq got-identifier nil) t)
6378 ;; It turned out to be the real identifier,
6379 ;; so stop.
6380 nil))
6381 t))
6382
6383 (if (eq (char-after) ?\()
6384 (progn
6385 (setq paren-depth (1+ paren-depth))
6386 (forward-char))
6387 (unless got-prefix-before-parens
6388 (setq got-prefix-before-parens (= paren-depth 0)))
6389 (setq got-prefix t)
6390 (goto-char (match-end 1)))
6391 (c-forward-syntactic-ws))
6392
6393 (setq got-parens (> paren-depth 0))
6394
6395 ;; Skip over an identifier.
6396 (or got-identifier
6397 (and (looking-at c-identifier-start)
6398 (setq got-identifier (c-forward-name))))
6399
6400 ;; Skip over type decl suffix operators.
6401 (while (if (looking-at c-type-decl-suffix-key)
6402
6403 (if (eq (char-after) ?\))
6404 (when (> paren-depth 0)
6405 (setq paren-depth (1- paren-depth))
6406 (forward-char)
6407 t)
6408 (when (if (save-match-data (looking-at "\\s\("))
6409 (c-safe (c-forward-sexp 1) t)
6410 (goto-char (match-end 1))
6411 t)
6412 (when (and (not got-suffix-after-parens)
6413 (= paren-depth 0))
6414 (setq got-suffix-after-parens (match-beginning 0)))
6415 (setq got-suffix t)))
6416
6417 ;; No suffix matched. We might have matched the
6418 ;; identifier as a type and the open paren of a
6419 ;; function arglist as a type decl prefix. In that
6420 ;; case we should "backtrack": Reinterpret the last
6421 ;; type as the identifier, move out of the arglist and
6422 ;; continue searching for suffix operators.
6423 ;;
6424 ;; Do this even if there's no preceding type, to cope
6425 ;; with old style function declarations in K&R C,
6426 ;; (con|de)structors in C++ and `c-typeless-decl-kwds'
6427 ;; style declarations. That isn't applicable in an
6428 ;; arglist context, though.
6429 (when (and (= paren-depth 1)
6430 (not got-prefix-before-parens)
6431 (not (eq at-type t))
6432 (or backup-at-type
6433 maybe-typeless
6434 backup-maybe-typeless
6435 (when c-recognize-typeless-decls
6436 (not context)))
6437 (setq pos (c-up-list-forward (point)))
6438 (eq (char-before pos) ?\)))
6439 (c-fdoc-shift-type-backward)
6440 (goto-char pos)
6441 t))
6442
6443 (c-forward-syntactic-ws))
6444
6445 (when (and (or maybe-typeless backup-maybe-typeless)
6446 (not got-identifier)
6447 (not got-prefix)
6448 at-type)
6449 ;; Have found no identifier but `c-typeless-decl-kwds' has
6450 ;; matched so we know we're inside a declaration. The
6451 ;; preceding type must be the identifier instead.
6452 (c-fdoc-shift-type-backward))
6453
6454 (setq
6455 at-decl-or-cast
6456 (catch 'at-decl-or-cast
6457
a85fd6da 6458 ;; CASE 1
0386b551
AM
6459 (when (> paren-depth 0)
6460 ;; Encountered something inside parens that isn't matched by
6461 ;; the `c-type-decl-*' regexps, so it's not a type decl
6462 ;; expression. Try to skip out to the same paren depth to
6463 ;; not confuse the cast check below.
6464 (c-safe (goto-char (scan-lists (point) 1 paren-depth)))
6465 ;; If we've found a specifier keyword then it's a
6466 ;; declaration regardless.
6467 (throw 'at-decl-or-cast (eq at-decl-or-cast t)))
6468
6469 (setq at-decl-end
6470 (looking-at (cond ((eq context '<>) "[,>]")
6471 (context "[,\)]")
6472 (t "[,;]"))))
6473
6474 ;; Now we've collected info about various characteristics of
6475 ;; the construct we're looking at. Below follows a decision
6476 ;; tree based on that. It's ordered to check more certain
6477 ;; signs before less certain ones.
6478
6479 (if got-identifier
6480 (progn
6481
a85fd6da 6482 ;; CASE 2
0386b551
AM
6483 (when (and (or at-type maybe-typeless)
6484 (not (or got-prefix got-parens)))
6485 ;; Got another identifier directly after the type, so it's a
6486 ;; declaration.
6487 (throw 'at-decl-or-cast t))
6488
6489 (when (and got-parens
6490 (not got-prefix)
6491 (not got-suffix-after-parens)
6492 (or backup-at-type
6493 maybe-typeless
6494 backup-maybe-typeless))
6495 ;; Got a declaration of the form "foo bar (gnu);" where we've
6496 ;; recognized "bar" as the type and "gnu" as the declarator.
6497 ;; In this case it's however more likely that "bar" is the
6498 ;; declarator and "gnu" a function argument or initializer (if
6499 ;; `c-recognize-paren-inits' is set), since the parens around
6500 ;; "gnu" would be superfluous if it's a declarator. Shift the
6501 ;; type one step backward.
6502 (c-fdoc-shift-type-backward)))
6503
6504 ;; Found no identifier.
6505
6506 (if backup-at-type
6507 (progn
6508
a85fd6da 6509 ;; CASE 3
0386b551
AM
6510 (when (= (point) start)
6511 ;; Got a plain list of identifiers. If a colon follows it's
6512 ;; a valid label. Otherwise the last one probably is the
6513 ;; declared identifier and we should back up to the previous
6514 ;; type, providing it isn't a cast.
6515 (if (eq (char-after) ?:)
6516 ;; If we've found a specifier keyword then it's a
6517 ;; declaration regardless.
6518 (throw 'at-decl-or-cast (eq at-decl-or-cast t))
6519 (setq backup-if-not-cast t)
6520 (throw 'at-decl-or-cast t)))
6521
a85fd6da 6522 ;; CASE 4
0386b551
AM
6523 (when (and got-suffix
6524 (not got-prefix)
6525 (not got-parens))
6526 ;; Got a plain list of identifiers followed by some suffix.
6527 ;; If this isn't a cast then the last identifier probably is
6528 ;; the declared one and we should back up to the previous
6529 ;; type.
6530 (setq backup-if-not-cast t)
6531 (throw 'at-decl-or-cast t)))
6532
a85fd6da 6533 ;; CASE 5
0386b551
AM
6534 (when (eq at-type t)
6535 ;; If the type is known we know that there can't be any
6536 ;; identifier somewhere else, and it's only in declarations in
6537 ;; e.g. function prototypes and in casts that the identifier may
6538 ;; be left out.
6539 (throw 'at-decl-or-cast t))
6540
6541 (when (= (point) start)
6542 ;; Only got a single identifier (parsed as a type so far).
a85fd6da 6543 ;; CASE 6
0386b551
AM
6544 (if (and
6545 ;; Check that the identifier isn't at the start of an
6546 ;; expression.
6547 at-decl-end
6548 (cond
6549 ((eq context 'decl)
6550 ;; Inside an arglist that contains declarations. If K&R
6551 ;; style declarations and parenthesis style initializers
6552 ;; aren't allowed then the single identifier must be a
6553 ;; type, else we require that it's known or found
6554 ;; (primitive types are handled above).
6555 (or (and (not c-recognize-knr-p)
6556 (not c-recognize-paren-inits))
6557 (memq at-type '(known found))))
6558 ((eq context '<>)
6559 ;; Inside a template arglist. Accept known and found
6560 ;; types; other identifiers could just as well be
6561 ;; constants in C++.
6562 (memq at-type '(known found)))))
6563 (throw 'at-decl-or-cast t)
a85fd6da 6564 ;; CASE 7
0386b551
AM
6565 ;; Can't be a valid declaration or cast, but if we've found a
6566 ;; specifier it can't be anything else either, so treat it as
6567 ;; an invalid/unfinished declaration or cast.
6568 (throw 'at-decl-or-cast at-decl-or-cast))))
6569
6570 (if (and got-parens
6571 (not got-prefix)
6572 (not context)
6573 (not (eq at-type t))
6574 (or backup-at-type
6575 maybe-typeless
6576 backup-maybe-typeless
6577 (when c-recognize-typeless-decls
6578 (or (not got-suffix)
6579 (not (looking-at
6580 c-after-suffixed-type-maybe-decl-key))))))
6581 ;; Got an empty paren pair and a preceding type that probably
6582 ;; really is the identifier. Shift the type backwards to make
6583 ;; the last one the identifier. This is analogous to the
6584 ;; "backtracking" done inside the `c-type-decl-suffix-key' loop
6585 ;; above.
6586 ;;
6587 ;; Exception: In addition to the conditions in that
6588 ;; "backtracking" code, do not shift backward if we're not
6589 ;; looking at either `c-after-suffixed-type-decl-key' or "[;,]".
6590 ;; Since there's no preceding type, the shift would mean that
6591 ;; the declaration is typeless. But if the regexp doesn't match
6592 ;; then we will simply fall through in the tests below and not
6593 ;; recognize it at all, so it's better to try it as an abstract
6594 ;; declarator instead.
6595 (c-fdoc-shift-type-backward)
6596
6597 ;; Still no identifier.
a85fd6da 6598 ;; CASE 8
0386b551
AM
6599 (when (and got-prefix (or got-parens got-suffix))
6600 ;; Require `got-prefix' together with either `got-parens' or
6601 ;; `got-suffix' to recognize it as an abstract declarator:
6602 ;; `got-parens' only is probably an empty function call.
6603 ;; `got-suffix' only can build an ordinary expression together
6604 ;; with the preceding identifier which we've taken as a type.
6605 ;; We could actually accept on `got-prefix' only, but that can
6606 ;; easily occur temporarily while writing an expression so we
6607 ;; avoid that case anyway. We could do a better job if we knew
6608 ;; the point when the fontification was invoked.
6609 (throw 'at-decl-or-cast t))
6610
a85fd6da 6611 ;; CASE 9
0386b551
AM
6612 (when (and at-type
6613 (not got-prefix)
6614 (not got-parens)
6615 got-suffix-after-parens
6616 (eq (char-after got-suffix-after-parens) ?\())
6617 ;; Got a type, no declarator but a paren suffix. I.e. it's a
6618 ;; normal function call afterall (or perhaps a C++ style object
6619 ;; instantiation expression).
6620 (throw 'at-decl-or-cast nil))))
6621
a85fd6da 6622 ;; CASE 10
0386b551
AM
6623 (when at-decl-or-cast
6624 ;; By now we've located the type in the declaration that we know
6625 ;; we're in.
6626 (throw 'at-decl-or-cast t))
6627
a85fd6da 6628 ;; CASE 11
0386b551
AM
6629 (when (and got-identifier
6630 (not context)
6631 (looking-at c-after-suffixed-type-decl-key)
6632 (if (and got-parens
6633 (not got-prefix)
6634 (not got-suffix)
6635 (not (eq at-type t)))
6636 ;; Shift the type backward in the case that there's a
6637 ;; single identifier inside parens. That can only
6638 ;; occur in K&R style function declarations so it's
6639 ;; more likely that it really is a function call.
6640 ;; Therefore we only do this after
6641 ;; `c-after-suffixed-type-decl-key' has matched.
6642 (progn (c-fdoc-shift-type-backward) t)
6643 got-suffix-after-parens))
6644 ;; A declaration according to `c-after-suffixed-type-decl-key'.
6645 (throw 'at-decl-or-cast t))
6646
a85fd6da 6647 ;; CASE 12
0386b551
AM
6648 (when (and (or got-prefix (not got-parens))
6649 (memq at-type '(t known)))
6650 ;; It's a declaration if a known type precedes it and it can't be a
6651 ;; function call.
6652 (throw 'at-decl-or-cast t))
6653
6654 ;; If we get here we can't tell if this is a type decl or a normal
6655 ;; expression by looking at it alone. (That's under the assumption
6656 ;; that normal expressions always can look like type decl expressions,
6657 ;; which isn't really true but the cases where it doesn't hold are so
6658 ;; uncommon (e.g. some placements of "const" in C++) it's not worth
6659 ;; the effort to look for them.)
6660
6661 (unless (or at-decl-end (looking-at "=[^=]"))
6662 ;; If this is a declaration it should end here or its initializer(*)
6663 ;; should start here, so check for allowed separation tokens. Note
6664 ;; that this rule doesn't work e.g. with a K&R arglist after a
6665 ;; function header.
6666 ;;
6667 ;; *) Don't check for C++ style initializers using parens
6668 ;; since those already have been matched as suffixes.
6669 ;;
6670 ;; If `at-decl-or-cast' is then we've found some other sign that
6671 ;; it's a declaration or cast, so then it's probably an
6672 ;; invalid/unfinished one.
6673 (throw 'at-decl-or-cast at-decl-or-cast))
6674
6675 ;; Below are tests that only should be applied when we're certain to
6676 ;; not have parsed halfway through an expression.
6677
a85fd6da 6678 ;; CASE 14
0386b551
AM
6679 (when (memq at-type '(t known))
6680 ;; The expression starts with a known type so treat it as a
6681 ;; declaration.
6682 (throw 'at-decl-or-cast t))
6683
a85fd6da 6684 ;; CASE 15
0386b551
AM
6685 (when (and (c-major-mode-is 'c++-mode)
6686 ;; In C++ we check if the identifier is a known type, since
6687 ;; (con|de)structors use the class name as identifier.
6688 ;; We've always shifted over the identifier as a type and
6689 ;; then backed up again in this case.
6690 identifier-type
6691 (or (memq identifier-type '(found known))
6692 (and (eq (char-after identifier-start) ?~)
6693 ;; `at-type' probably won't be 'found for
6694 ;; destructors since the "~" is then part of the
6695 ;; type name being checked against the list of
6696 ;; known types, so do a check without that
6697 ;; operator.
6698 (or (save-excursion
6699 (goto-char (1+ identifier-start))
6700 (c-forward-syntactic-ws)
6701 (c-with-syntax-table
6702 c-identifier-syntax-table
6703 (looking-at c-known-type-key)))
6704 (save-excursion
6705 (goto-char (1+ identifier-start))
6706 ;; We have already parsed the type earlier,
6707 ;; so it'd be possible to cache the end
6708 ;; position instead of redoing it here, but
6709 ;; then we'd need to keep track of another
6710 ;; position everywhere.
6711 (c-check-type (point)
6712 (progn (c-forward-type)
6713 (point))))))))
6714 (throw 'at-decl-or-cast t))
6715
6716 (if got-identifier
6717 (progn
a85fd6da 6718 ;; CASE 16
0386b551
AM
6719 (when (and got-prefix-before-parens
6720 at-type
6721 (or at-decl-end (looking-at "=[^=]"))
6722 (not context)
6723 (not got-suffix))
6724 ;; Got something like "foo * bar;". Since we're not inside an
6725 ;; arglist it would be a meaningless expression because the
6726 ;; result isn't used. We therefore choose to recognize it as
6727 ;; a declaration. Do not allow a suffix since it could then
6728 ;; be a function call.
6729 (throw 'at-decl-or-cast t))
6730
a85fd6da 6731 ;; CASE 17
0386b551
AM
6732 (when (and (or got-suffix-after-parens
6733 (looking-at "=[^=]"))
6734 (eq at-type 'found)
6735 (not (eq context 'arglist)))
6736 ;; Got something like "a (*b) (c);" or "a (b) = c;". It could
6737 ;; be an odd expression or it could be a declaration. Treat
6738 ;; it as a declaration if "a" has been used as a type
6739 ;; somewhere else (if it's a known type we won't get here).
6740 (throw 'at-decl-or-cast t)))
6741
a85fd6da 6742 ;; CASE 18
0386b551
AM
6743 (when (and context
6744 (or got-prefix
6745 (and (eq context 'decl)
6746 (not c-recognize-paren-inits)
6747 (or got-parens got-suffix))))
6748 ;; Got a type followed by an abstract declarator. If `got-prefix'
6749 ;; is set it's something like "a *" without anything after it. If
6750 ;; `got-parens' or `got-suffix' is set it's "a()", "a[]", "a()[]",
6751 ;; or similar, which we accept only if the context rules out
6752 ;; expressions.
6753 (throw 'at-decl-or-cast t)))
6754
6755 ;; If we had a complete symbol table here (which rules out
6756 ;; `c-found-types') we should return t due to the disambiguation rule
6757 ;; (in at least C++) that anything that can be parsed as a declaration
6758 ;; is a declaration. Now we're being more defensive and prefer to
6759 ;; highlight things like "foo (bar);" as a declaration only if we're
6760 ;; inside an arglist that contains declarations.
6761 (eq context 'decl))))
6762
6763 ;; The point is now after the type decl expression.
6764
6765 (cond
6766 ;; Check for a cast.
6767 ((save-excursion
6768 (and
6769 c-cast-parens
6770
6771 ;; Should be the first type/identifier in a cast paren.
6772 (> preceding-token-end (point-min))
6773 (memq (char-before preceding-token-end) c-cast-parens)
6774
6775 ;; The closing paren should follow.
6776 (progn
6777 (c-forward-syntactic-ws)
6778 (looking-at "\\s\)"))
6779
6780 ;; There should be a primary expression after it.
6781 (let (pos)
6782 (forward-char)
6783 (c-forward-syntactic-ws)
6784 (setq cast-end (point))
6785 (and (looking-at c-primary-expr-regexp)
6786 (progn
6787 (setq pos (match-end 0))
6788 (or
6789 ;; Check if the expression begins with a prefix keyword.
6790 (match-beginning 2)
6791 (if (match-beginning 1)
6792 ;; Expression begins with an ambiguous operator. Treat
6793 ;; it as a cast if it's a type decl or if we've
6794 ;; recognized the type somewhere else.
6795 (or at-decl-or-cast
6796 (memq at-type '(t known found)))
6797 ;; Unless it's a keyword, it's the beginning of a primary
6798 ;; expression.
6799 (not (looking-at c-keywords-regexp)))))
6800 ;; If `c-primary-expr-regexp' matched a nonsymbol token, check
6801 ;; that it matched a whole one so that we don't e.g. confuse
6802 ;; the operator '-' with '->'. It's ok if it matches further,
6803 ;; though, since it e.g. can match the float '.5' while the
6804 ;; operator regexp only matches '.'.
6805 (or (not (looking-at c-nonsymbol-token-regexp))
6806 (<= (match-end 0) pos))))
6807
6808 ;; There should either be a cast before it or something that isn't an
6809 ;; identifier or close paren.
6810 (> preceding-token-end (point-min))
6811 (progn
6812 (goto-char (1- preceding-token-end))
6813 (or (eq (point) last-cast-end)
6814 (progn
6815 (c-backward-syntactic-ws)
6816 (if (< (skip-syntax-backward "w_") 0)
6817 ;; It's a symbol. Accept it only if it's one of the
6818 ;; keywords that can precede an expression (without
6819 ;; surrounding parens).
6820 (looking-at c-simple-stmt-key)
6821 (and
6822 ;; Check that it isn't a close paren (block close is ok,
6823 ;; though).
6824 (not (memq (char-before) '(?\) ?\])))
6825 ;; Check that it isn't a nonsymbol identifier.
6826 (not (c-on-identifier)))))))))
6827
6828 ;; Handle the cast.
6829 (when (and c-record-type-identifiers at-type (not (eq at-type t)))
6830 (let ((c-promote-possible-types t))
6831 (goto-char type-start)
6832 (c-forward-type)))
6833
6834 (goto-char cast-end)
6835 'cast)
6836
6837 (at-decl-or-cast
6838 ;; We're at a declaration. Highlight the type and the following
6839 ;; declarators.
6840
6841 (when backup-if-not-cast
6842 (c-fdoc-shift-type-backward t))
6843
6844 (when (and (eq context 'decl) (looking-at ","))
6845 ;; Make sure to propagate the `c-decl-arg-start' property to
6846 ;; the next argument if it's set in this one, to cope with
6847 ;; interactive refontification.
6848 (c-put-c-type-property (point) 'c-decl-arg-start))
6849
6850 (when (and c-record-type-identifiers at-type (not (eq at-type t)))
6851 (let ((c-promote-possible-types t))
6852 (save-excursion
6853 (goto-char type-start)
6854 (c-forward-type))))
6855
6856 (cons id-start at-type-decl))
6857
6858 (t
6859 ;; False alarm. Restore the recorded ranges.
6860 (setq c-record-type-identifiers save-rec-type-ids
6861 c-record-ref-identifiers save-rec-ref-ids)
6862 nil))))
6863
6864(defun c-forward-label (&optional assume-markup preceding-token-end limit)
51c9af45 6865 ;; Assuming that point is at the beginning of a token, check if it starts a
1379f2c5
AM
6866 ;; label and if so move over it and return non-nil (t in default situations,
6867 ;; specific symbols (see below) for interesting situations), otherwise don't
6868 ;; move and return nil. "Label" here means "most things with a colon".
51c9af45
AM
6869 ;;
6870 ;; More precisely, a "label" is regarded as one of:
1379f2c5
AM
6871 ;; (i) a goto target like "foo:" - returns the symbol `goto-target';
6872 ;; (ii) A case label - either the entire construct "case FOO:", or just the
6873 ;; bare "case", should the colon be missing. We return t;
6874 ;; (iii) a keyword which needs a colon, like "default:" or "private:"; We
6875 ;; return t;
51c9af45 6876 ;; (iv) One of QT's "extended" C++ variants of
1379f2c5
AM
6877 ;; "private:"/"protected:"/"public:"/"more:" looking like "public slots:".
6878 ;; Returns the symbol `qt-2kwds-colon'.
6879 ;; (v) QT's construct "signals:". Returns the symbol `qt-1kwd-colon'.
82ba65cf 6880 ;; (vi) One of the keywords matched by `c-opt-extra-label-key' (without any
51c9af45 6881 ;; colon). Currently (2006-03), this applies only to Objective C's
1379f2c5 6882 ;; keywords "@private", "@protected", and "@public". Returns t.
51c9af45
AM
6883 ;;
6884 ;; One of the things which will NOT be recognised as a label is a bit-field
6885 ;; element of a struct, something like "int foo:5".
6886 ;;
6887 ;; The end of the label is taken to be just after the colon, or the end of
6888 ;; the first submatch in `c-opt-extra-label-key'. The point is directly
6889 ;; after the end on return. The terminating char gets marked with
6890 ;; `c-decl-end' to improve recognition of the following declaration or
6891 ;; statement.
0386b551
AM
6892 ;;
6893 ;; If ASSUME-MARKUP is non-nil, it's assumed that the preceding
51c9af45 6894 ;; label, if any, has already been marked up like that.
0386b551
AM
6895 ;;
6896 ;; If PRECEDING-TOKEN-END is given, it should be the first position
6897 ;; after the preceding token, i.e. on the other side of the
6898 ;; syntactic ws from the point. Use a value less than or equal to
6899 ;; (point-min) if the point is at the first token in (the visible
6900 ;; part of) the buffer.
6901 ;;
6902 ;; The optional LIMIT limits the forward scan for the colon.
6903 ;;
6904 ;; This function records the ranges of the label symbols on
6905 ;; `c-record-ref-identifiers' if `c-record-type-identifiers' (!) is
6906 ;; non-nil.
6907 ;;
6908 ;; This function might do hidden buffer changes.
6909
51c9af45 6910 (let ((start (point))
1379f2c5 6911 label-end
51c9af45 6912 qt-symbol-idx
1379f2c5 6913 macro-start ; if we're in one.
f412a567
AM
6914 label-type
6915 kwd)
0386b551 6916 (cond
b414f371 6917 ;; "case" or "default" (Doesn't apply to AWK).
0386b551
AM
6918 ((looking-at c-label-kwds-regexp)
6919 (let ((kwd-end (match-end 1)))
6920 ;; Record only the keyword itself for fontification, since in
6921 ;; case labels the following is a constant expression and not
6922 ;; a label.
6923 (when c-record-type-identifiers
6924 (c-record-ref-id (cons (match-beginning 1) kwd-end)))
6925
6926 ;; Find the label end.
6927 (goto-char kwd-end)
1379f2c5
AM
6928 (setq label-type
6929 (if (and (c-syntactic-re-search-forward
6930 ;; Stop on chars that aren't allowed in expressions,
6931 ;; and on operator chars that would be meaningless
6932 ;; there. FIXME: This doesn't cope with ?: operators.
6933 "[;{=,@]\\|\\(\\=\\|[^:]\\):\\([^:]\\|\\'\\)"
6934 limit t t nil 1)
6935 (match-beginning 2))
6936
6937 (progn ; there's a proper :
6938 (goto-char (match-beginning 2)) ; just after the :
6939 (c-put-c-type-property (1- (point)) 'c-decl-end)
6940 t)
6941
6942 ;; It's an unfinished label. We consider the keyword enough
6943 ;; to recognize it as a label, so that it gets fontified.
6944 ;; Leave the point at the end of it, but don't put any
6945 ;; `c-decl-end' marker.
6946 (goto-char kwd-end)
6947 t))))
0386b551 6948
51c9af45 6949 ;; @private, @protected, @public, in Objective C, or similar.
0386b551
AM
6950 ((and c-opt-extra-label-key
6951 (looking-at c-opt-extra-label-key))
6952 ;; For a `c-opt-extra-label-key' match, we record the whole
6953 ;; thing for fontification. That's to get the leading '@' in
6954 ;; Objective-C protection labels fontified.
6955 (goto-char (match-end 1))
6956 (when c-record-type-identifiers
6957 (c-record-ref-id (cons (match-beginning 1) (point))))
6958 (c-put-c-type-property (1- (point)) 'c-decl-end)
1379f2c5 6959 (setq label-type t))
0386b551 6960
51c9af45
AM
6961 ;; All other cases of labels.
6962 ((and c-recognize-colon-labels ; nil for AWK and IDL, otherwise t.
0386b551
AM
6963
6964 ;; A colon label must have something before the colon.
6965 (not (eq (char-after) ?:))
6966
6967 ;; Check that we're not after a token that can't precede a label.
6968 (or
6969 ;; Trivially succeeds when there's no preceding token.
6970 (if preceding-token-end
6971 (<= preceding-token-end (point-min))
6972 (save-excursion
6973 (c-backward-syntactic-ws)
6974 (setq preceding-token-end (point))
6975 (bobp)))
6976
6977 ;; Check if we're after a label, if we're after a closing
6978 ;; paren that belong to statement, and with
6979 ;; `c-label-prefix-re'. It's done in different order
6980 ;; depending on `assume-markup' since the checks have
6981 ;; different expensiveness.
6982 (if assume-markup
6983 (or
6984 (eq (c-get-char-property (1- preceding-token-end) 'c-type)
6985 'c-decl-end)
6986
6987 (save-excursion
6988 (goto-char (1- preceding-token-end))
6989 (c-beginning-of-current-token)
51c9af45
AM
6990 (or (looking-at c-label-prefix-re)
6991 (looking-at c-block-stmt-1-key)))
0386b551
AM
6992
6993 (and (eq (char-before preceding-token-end) ?\))
6994 (c-after-conditional)))
6995
6996 (or
6997 (save-excursion
6998 (goto-char (1- preceding-token-end))
6999 (c-beginning-of-current-token)
51c9af45
AM
7000 (or (looking-at c-label-prefix-re)
7001 (looking-at c-block-stmt-1-key)))
0386b551
AM
7002
7003 (cond
7004 ((eq (char-before preceding-token-end) ?\))
7005 (c-after-conditional))
7006
7007 ((eq (char-before preceding-token-end) ?:)
7008 ;; Might be after another label, so check it recursively.
51c9af45
AM
7009 (save-restriction
7010 (save-excursion
7011 (goto-char (1- preceding-token-end))
7012 ;; Essentially the same as the
7013 ;; `c-syntactic-re-search-forward' regexp below.
7014 (setq macro-start
7015 (save-excursion (and (c-beginning-of-macro)
7016 (point))))
7017 (if macro-start (narrow-to-region macro-start (point-max)))
7018 (c-syntactic-skip-backward "^-]:?;}=*/%&|,<>!@+" nil t)
7019 ;; Note: the following should work instead of the
7020 ;; narrow-to-region above. Investigate why not,
7021 ;; sometime. ACM, 2006-03-31.
7022 ;; (c-syntactic-skip-backward "^-]:?;}=*/%&|,<>!@+"
7023 ;; macro-start t)
7024 (let ((pte (point))
7025 ;; If the caller turned on recording for us,
7026 ;; it shouldn't apply when we check the
7027 ;; preceding label.
7028 c-record-type-identifiers)
7029 ;; A label can't start at a cpp directive. Check for
7030 ;; this, since c-forward-syntactic-ws would foul up on it.
7031 (unless (and c-opt-cpp-prefix (looking-at c-opt-cpp-prefix))
7032 (c-forward-syntactic-ws)
7033 (c-forward-label nil pte start))))))))))
7034
1379f2c5 7035 ;; Point is still at the beginning of the possible label construct.
b414f371 7036 ;;
51c9af45
AM
7037 ;; Check that the next nonsymbol token is ":", or that we're in one
7038 ;; of QT's "slots" declarations. Allow '(' for the sake of macro
7039 ;; arguments. FIXME: Should build this regexp from the language
7040 ;; constants.
1379f2c5
AM
7041 (cond
7042 ;; public: protected: private:
7043 ((and
7044 (c-major-mode-is 'c++-mode)
7045 (search-forward-regexp
7046 "\\=p\\(r\\(ivate\\|otected\\)\\|ublic\\)\\>[^_]" nil t)
7047 (progn (backward-char)
7048 (c-forward-syntactic-ws limit)
7049 (looking-at ":\\([^:]\\|\\'\\)"))) ; A single colon.
7050 (forward-char)
7051 (setq label-type t))
7052 ;; QT double keyword like "protected slots:" or goto target.
7053 ((progn (goto-char start) nil))
7054 ((when (c-syntactic-re-search-forward
7055 "[ \t\n[:?;{=*/%&|,<>!@+-]" limit t t) ; not at EOB
7056 (backward-char)
7057 (setq label-end (point))
7058 (setq qt-symbol-idx
7059 (and (c-major-mode-is 'c++-mode)
7060 (string-match
7061 "\\(p\\(r\\(ivate\\|otected\\)\\|ublic\\)\\|more\\)\\>"
b414f371 7062 (buffer-substring start (point)))))
1379f2c5
AM
7063 (c-forward-syntactic-ws limit)
7064 (cond
7065 ((looking-at ":\\([^:]\\|\\'\\)") ; A single colon.
7066 (forward-char)
7067 (setq label-type
f412a567
AM
7068 (if (or (string= "signals" ; Special QT macro
7069 (setq kwd (buffer-substring-no-properties start label-end)))
7070 (string= "Q_SIGNALS" kwd))
1379f2c5
AM
7071 'qt-1kwd-colon
7072 'goto-target)))
7073 ((and qt-symbol-idx
f412a567 7074 (search-forward-regexp "\\=\\(slots\\|Q_SLOTS\\)\\>" limit t)
1379f2c5
AM
7075 (progn (c-forward-syntactic-ws limit)
7076 (looking-at ":\\([^:]\\|\\'\\)"))) ; A single colon
7077 (forward-char)
7078 (setq label-type 'qt-2kwds-colon)))))))
0386b551
AM
7079
7080 (save-restriction
7081 (narrow-to-region start (point))
7082
7083 ;; Check that `c-nonlabel-token-key' doesn't match anywhere.
7084 (catch 'check-label
7085 (goto-char start)
7086 (while (progn
7087 (when (looking-at c-nonlabel-token-key)
7088 (goto-char start)
1379f2c5 7089 (setq label-type nil)
0386b551
AM
7090 (throw 'check-label nil))
7091 (and (c-safe (c-forward-sexp)
7092 (c-forward-syntactic-ws)
7093 t)
7094 (not (eobp)))))
7095
7096 ;; Record the identifiers in the label for fontification, unless
7097 ;; it begins with `c-label-kwds' in which case the following
7098 ;; identifiers are part of a (constant) expression that
7099 ;; shouldn't be fontified.
7100 (when (and c-record-type-identifiers
7101 (progn (goto-char start)
7102 (not (looking-at c-label-kwds-regexp))))
7103 (while (c-syntactic-re-search-forward c-symbol-key nil t)
7104 (c-record-ref-id (cons (match-beginning 0)
7105 (match-end 0)))))
7106
7107 (c-put-c-type-property (1- (point-max)) 'c-decl-end)
82ba65cf 7108 (goto-char (point-max)))))
0386b551
AM
7109
7110 (t
7111 ;; Not a label.
1379f2c5
AM
7112 (goto-char start)))
7113 label-type))
0386b551
AM
7114
7115(defun c-forward-objc-directive ()
7116 ;; Assuming the point is at the beginning of a token, try to move
7117 ;; forward to the end of the Objective-C directive that starts
7118 ;; there. Return t if a directive was fully recognized, otherwise
7119 ;; the point is moved as far as one could be successfully parsed and
7120 ;; nil is returned.
7121 ;;
7122 ;; This function records identifier ranges on
7123 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
7124 ;; `c-record-type-identifiers' is non-nil.
7125 ;;
7126 ;; This function might do hidden buffer changes.
7127
7128 (let ((start (point))
7129 start-char
7130 (c-promote-possible-types t)
7131 ;; Turn off recognition of angle bracket arglists while parsing
7132 ;; types here since the protocol reference list might then be
7133 ;; considered part of the preceding name or superclass-name.
7134 c-recognize-<>-arglists)
7135
7136 (if (or
7137 (when (looking-at
7138 (eval-when-compile
7139 (c-make-keywords-re t
7140 (append (c-lang-const c-protection-kwds objc)
7141 '("@end"))
7142 'objc-mode)))
7143 (goto-char (match-end 1))
7144 t)
7145
7146 (and
7147 (looking-at
7148 (eval-when-compile
7149 (c-make-keywords-re t
7150 '("@interface" "@implementation" "@protocol")
7151 'objc-mode)))
7152
7153 ;; Handle the name of the class itself.
7154 (progn
cb694ab7
AM
7155; (c-forward-token-2) ; 2006/1/13 This doesn't move if the token's
7156; at EOB.
7157 (goto-char (match-end 0))
7158 (c-skip-ws-forward)
0386b551
AM
7159 (c-forward-type))
7160
7161 (catch 'break
7162 ;; Look for ": superclass-name" or "( category-name )".
7163 (when (looking-at "[:\(]")
7164 (setq start-char (char-after))
7165 (forward-char)
7166 (c-forward-syntactic-ws)
7167 (unless (c-forward-type) (throw 'break nil))
7168 (when (eq start-char ?\()
7169 (unless (eq (char-after) ?\)) (throw 'break nil))
7170 (forward-char)
7171 (c-forward-syntactic-ws)))
7172
7173 ;; Look for a protocol reference list.
7174 (if (eq (char-after) ?<)
7175 (let ((c-recognize-<>-arglists t)
7176 (c-parse-and-markup-<>-arglists t)
7177 c-restricted-<>-arglists)
7178 (c-forward-<>-arglist t))
7179 t))))
7180
7181 (progn
7182 (c-backward-syntactic-ws)
7183 (c-clear-c-type-property start (1- (point)) 'c-decl-end)
7184 (c-put-c-type-property (1- (point)) 'c-decl-end)
7185 t)
7186
7187 (c-clear-c-type-property start (point) 'c-decl-end)
7188 nil)))
7189
785eecbb
RS
7190(defun c-beginning-of-inheritance-list (&optional lim)
7191 ;; Go to the first non-whitespace after the colon that starts a
7192 ;; multiple inheritance introduction. Optional LIM is the farthest
7193 ;; back we should search.
0386b551
AM
7194 ;;
7195 ;; This function might do hidden buffer changes.
7196 (c-with-syntax-table c++-template-syntax-table
7197 (c-backward-token-2 0 t lim)
7198 (while (and (or (looking-at c-symbol-start)
7199 (looking-at "[<,]\\|::"))
7200 (zerop (c-backward-token-2 1 t lim))))))
785eecbb 7201
785eecbb
RS
7202(defun c-in-method-def-p ()
7203 ;; Return nil if we aren't in a method definition, otherwise the
7204 ;; position of the initial [+-].
0386b551
AM
7205 ;;
7206 ;; This function might do hidden buffer changes.
785eecbb
RS
7207 (save-excursion
7208 (beginning-of-line)
a66cd3ee
MS
7209 (and c-opt-method-key
7210 (looking-at c-opt-method-key)
785eecbb
RS
7211 (point))
7212 ))
7213
a66cd3ee
MS
7214;; Contributed by Kevin Ryde <user42@zip.com.au>.
7215(defun c-in-gcc-asm-p ()
7216 ;; Return non-nil if point is within a gcc \"asm\" block.
7217 ;;
7218 ;; This should be called with point inside an argument list.
7219 ;;
7220 ;; Only one level of enclosing parentheses is considered, so for
7221 ;; instance `nil' is returned when in a function call within an asm
7222 ;; operand.
0386b551
AM
7223 ;;
7224 ;; This function might do hidden buffer changes.
a66cd3ee
MS
7225
7226 (and c-opt-asm-stmt-key
7227 (save-excursion
7228 (beginning-of-line)
7229 (backward-up-list 1)
7230 (c-beginning-of-statement-1 (point-min) nil t)
7231 (looking-at c-opt-asm-stmt-key))))
7232
abb7e5cf 7233(defun c-at-toplevel-p ()
a85fd6da
AM
7234 "Return a determination as to whether point is \"at the top level\".
7235Informally, \"at the top level\" is anywhere where you can write
7236a function.
7237
7238More precisely, being at the top-level means that point is either
7239outside any enclosing block (such as a function definition), or
7240directly inside a class, namespace or other block that contains
7241another declaration level.
abb7e5cf
SM
7242
7243If point is not at the top-level (e.g. it is inside a method
7244definition), then nil is returned. Otherwise, if point is at a
7245top-level not enclosed within a class definition, t is returned.
7246Otherwise, a 2-vector is returned where the zeroth element is the
7247buffer position of the start of the class declaration, and the first
7248element is the buffer position of the enclosing class's opening
0386b551
AM
7249brace.
7250
7251Note that this function might do hidden buffer changes. See the
7252comment at the start of cc-engine.el for more info."
a66cd3ee
MS
7253 (let ((paren-state (c-parse-state)))
7254 (or (not (c-most-enclosing-brace paren-state))
7255 (c-search-uplist-for-classkey paren-state))))
7256
d9e94c22 7257(defun c-just-after-func-arglist-p (&optional lim)
0386b551
AM
7258 ;; Return non-nil if the point is in the region after the argument
7259 ;; list of a function and its opening brace (or semicolon in case it
7260 ;; got no body). If there are K&R style argument declarations in
7261 ;; that region, the point has to be inside the first one for this
7262 ;; function to recognize it.
a66cd3ee 7263 ;;
0386b551
AM
7264 ;; If successful, the point is moved to the first token after the
7265 ;; function header (see `c-forward-decl-or-cast-1' for details) and
7266 ;; the position of the opening paren of the function arglist is
7267 ;; returned.
7268 ;;
7269 ;; The point is clobbered if not successful.
7270 ;;
7271 ;; LIM is used as bound for backward buffer searches.
7272 ;;
7273 ;; This function might do hidden buffer changes.
7274
7275 (let ((beg (point)) end id-start)
7276 (and
7277 (eq (c-beginning-of-statement-1 lim) 'same)
7278
7279 (not (or (c-major-mode-is 'objc-mode)
7280 (c-forward-objc-directive)))
7281
7282 (setq id-start
7283 (car-safe (c-forward-decl-or-cast-1 (c-point 'bosws) nil nil)))
7284 (< id-start beg)
7285
7286 ;; There should not be a '=' or ',' between beg and the
7287 ;; start of the declaration since that means we were in the
7288 ;; "expression part" of the declaration.
7289 (or (> (point) beg)
7290 (not (looking-at "[=,]")))
7291
7292 (save-excursion
7293 ;; Check that there's an arglist paren in the
7294 ;; declaration.
7295 (goto-char id-start)
7296 (cond ((eq (char-after) ?\()
7297 ;; The declarator is a paren expression, so skip past it
7298 ;; so that we don't get stuck on that instead of the
7299 ;; function arglist.
7300 (c-forward-sexp))
51c9af45
AM
7301 ((and c-opt-op-identifier-prefix
7302 (looking-at c-opt-op-identifier-prefix))
0386b551
AM
7303 ;; Don't trip up on "operator ()".
7304 (c-forward-token-2 2 t)))
7305 (and (< (point) beg)
7306 (c-syntactic-re-search-forward "(" beg t t)
7307 (1- (point)))))))
785eecbb 7308
a66cd3ee
MS
7309(defun c-in-knr-argdecl (&optional lim)
7310 ;; Return the position of the first argument declaration if point is
7311 ;; inside a K&R style argument declaration list, nil otherwise.
7312 ;; `c-recognize-knr-p' is not checked. If LIM is non-nil, it's a
7313 ;; position that bounds the backward search for the argument list.
7314 ;;
9cf17ef1
AM
7315 ;; Point must be within a possible K&R region, e.g. just before a top-level
7316 ;; "{". It must be outside of parens and brackets. The test can return
7317 ;; false positives otherwise.
0386b551
AM
7318 ;;
7319 ;; This function might do hidden buffer changes.
d9e94c22 7320
a66cd3ee
MS
7321 (save-excursion
7322 (save-restriction
9cf17ef1
AM
7323 ;; If we're in a macro, our search range is restricted to it. Narrow to
7324 ;; the searchable range.
7325 (let* ((macro-start (c-query-macro-start))
7326 (lim (max (or lim (point-min)) (or macro-start (point-min))))
6b6481ed
AM
7327 before-lparen after-rparen
7328 (pp-count-out 20)) ; Max number of paren/brace constructs before we give up
9cf17ef1
AM
7329 (narrow-to-region lim (c-point 'eol))
7330
7331 ;; Search backwards for the defun's argument list. We give up if we
7332 ;; encounter a "}" (end of a previous defun) or BOB.
7333 ;;
7334 ;; The criterion for a paren structure being the arg list is:
7335 ;; o - there is non-WS stuff after it but before any "{"; AND
7336 ;; o - the token after it isn't a ";" AND
7337 ;; o - it is preceded by either an identifier (the function name) or
7338 ;; a macro expansion like "DEFUN (...)"; AND
7339 ;; o - its content is a non-empty comma-separated list of identifiers
7340 ;; (an empty arg list won't have a knr region).
7341 ;;
7342 ;; The following snippet illustrates these rules:
7343 ;; int foo (bar, baz, yuk)
7344 ;; int bar [] ;
7345 ;; int (*baz) (my_type) ;
7346 ;; int (*) (void) (*yuk) (void) ;
7347 ;; {
7348
7349 (catch 'knr
6b6481ed
AM
7350 (while (> pp-count-out 0) ; go back one paren/bracket pair each time.
7351 (setq pp-count-out (1- pp-count-out))
10489fcb 7352 (c-syntactic-skip-backward "^)]}")
9cf17ef1
AM
7353 (cond ((eq (char-before) ?\))
7354 (setq after-rparen (point)))
10489fcb
AM
7355 ((eq (char-before) ?\])
7356 (setq after-rparen nil))
7357 (t ; either } (hit previous defun) or no more parens/brackets
7358 (throw 'knr nil)))
9cf17ef1
AM
7359
7360 (if after-rparen
7361 ;; We're inside a paren. Could it be our argument list....?
7362 (if
7363 (and
7364 (progn
7365 (goto-char after-rparen)
7366 (unless (c-go-list-backward) (throw 'knr nil)) ;
7367 ;; FIXME!!! What about macros between the parens? 2007/01/20
7368 (setq before-lparen (point)))
d9e94c22 7369
9cf17ef1
AM
7370 ;; It can't be the arg list if next token is ; or {
7371 (progn (goto-char after-rparen)
7372 (c-forward-syntactic-ws)
7373 (not (memq (char-after) '(?\; ?\{))))
d9e94c22 7374
9cf17ef1
AM
7375 ;; Is the thing preceding the list an identifier (the
7376 ;; function name), or a macro expansion?
7377 (progn
7378 (goto-char before-lparen)
7379 (eq (c-backward-token-2) 0)
7380 (or (c-on-identifier)
7381 (and (eq (char-after) ?\))
7382 (c-go-up-list-backward)
7383 (eq (c-backward-token-2) 0)
7384 (c-on-identifier))))
7385
7386 ;; Have we got a non-empty list of comma-separated
7387 ;; identifiers?
7388 (progn
7389 (goto-char before-lparen)
7390 (c-forward-token-2) ; to first token inside parens
7391 (and
7392 (c-on-identifier)
7393 (c-forward-token-2)
7394 (catch 'id-list
7395 (while (eq (char-after) ?\,)
7396 (c-forward-token-2)
7397 (unless (c-on-identifier) (throw 'id-list nil))
7398 (c-forward-token-2))
7399 (eq (char-after) ?\))))))
7400
7401 ;; ...Yes. We've identified the function's argument list.
7402 (throw 'knr
7403 (progn (goto-char after-rparen)
7404 (c-forward-syntactic-ws)
7405 (point)))
7406
7407 ;; ...No. The current parens aren't the function's arg list.
7408 (goto-char before-lparen))
7409
7410 (or (c-go-list-backward) ; backwards over [ .... ]
7411 (throw 'knr nil)))))))))
785eecbb
RS
7412
7413(defun c-skip-conditional ()
7414 ;; skip forward over conditional at point, including any predicate
7415 ;; statements in parentheses. No error checking is performed.
0386b551
AM
7416 ;;
7417 ;; This function might do hidden buffer changes.
0ec8351b
BW
7418 (c-forward-sexp (cond
7419 ;; else if()
a66cd3ee
MS
7420 ((looking-at (concat "\\<else"
7421 "\\([ \t\n]\\|\\\\\n\\)+"
7422 "if\\>\\([^_]\\|$\\)"))
7423 3)
0ec8351b 7424 ;; do, else, try, finally
a66cd3ee
MS
7425 ((looking-at (concat "\\<\\("
7426 "do\\|else\\|try\\|finally"
7427 "\\)\\>\\([^_]\\|$\\)"))
130c507e 7428 1)
ce8c7486 7429 ;; for, if, while, switch, catch, synchronized, foreach
0ec8351b 7430 (t 2))))
785eecbb 7431
a66cd3ee
MS
7432(defun c-after-conditional (&optional lim)
7433 ;; If looking at the token after a conditional then return the
7434 ;; position of its start, otherwise return nil.
0386b551
AM
7435 ;;
7436 ;; This function might do hidden buffer changes.
a66cd3ee 7437 (save-excursion
d9e94c22 7438 (and (zerop (c-backward-token-2 1 t lim))
a66cd3ee
MS
7439 (or (looking-at c-block-stmt-1-key)
7440 (and (eq (char-after) ?\()
d9e94c22 7441 (zerop (c-backward-token-2 1 t lim))
a66cd3ee
MS
7442 (looking-at c-block-stmt-2-key)))
7443 (point))))
7444
0386b551
AM
7445(defun c-after-special-operator-id (&optional lim)
7446 ;; If the point is after an operator identifier that isn't handled
7447 ;; like an ordinary symbol (i.e. like "operator =" in C++) then the
7448 ;; position of the start of that identifier is returned. nil is
7449 ;; returned otherwise. The point may be anywhere in the syntactic
7450 ;; whitespace after the last token of the operator identifier.
7451 ;;
7452 ;; This function might do hidden buffer changes.
7453 (save-excursion
7454 (and c-overloadable-operators-regexp
7455 (zerop (c-backward-token-2 1 nil lim))
7456 (looking-at c-overloadable-operators-regexp)
51c9af45 7457 (or (not c-opt-op-identifier-prefix)
0386b551
AM
7458 (and
7459 (zerop (c-backward-token-2 1 nil lim))
51c9af45 7460 (looking-at c-opt-op-identifier-prefix)))
0386b551
AM
7461 (point))))
7462
a66cd3ee
MS
7463(defsubst c-backward-to-block-anchor (&optional lim)
7464 ;; Assuming point is at a brace that opens a statement block of some
7465 ;; kind, move to the proper anchor point for that block. It might
7466 ;; need to be adjusted further by c-add-stmt-syntax, but the
7467 ;; position at return is suitable as start position for that
7468 ;; function.
0386b551
AM
7469 ;;
7470 ;; This function might do hidden buffer changes.
a66cd3ee
MS
7471 (unless (= (point) (c-point 'boi))
7472 (let ((start (c-after-conditional lim)))
7473 (if start
7474 (goto-char start)))))
7475
037558bf 7476(defsubst c-backward-to-decl-anchor (&optional lim)
a66cd3ee
MS
7477 ;; Assuming point is at a brace that opens the block of a top level
7478 ;; declaration of some kind, move to the proper anchor point for
7479 ;; that block.
0386b551
AM
7480 ;;
7481 ;; This function might do hidden buffer changes.
a66cd3ee 7482 (unless (= (point) (c-point 'boi))
037558bf 7483 (c-beginning-of-statement-1 lim)))
a66cd3ee 7484
ff959bab 7485(defun c-search-decl-header-end ()
a66cd3ee
MS
7486 ;; Search forward for the end of the "header" of the current
7487 ;; declaration. That's the position where the definition body
7488 ;; starts, or the first variable initializer, or the ending
7489 ;; semicolon. I.e. search forward for the closest following
7490 ;; (syntactically relevant) '{', '=' or ';' token. Point is left
7491 ;; _after_ the first found token, or at point-max if none is found.
0386b551
AM
7492 ;;
7493 ;; This function might do hidden buffer changes.
ff959bab
MS
7494
7495 (let ((base (point)))
7496 (if (c-major-mode-is 'c++-mode)
7497
7498 ;; In C++ we need to take special care to handle operator
7499 ;; tokens and those pesky template brackets.
7500 (while (and
7501 (c-syntactic-re-search-forward "[;{<=]" nil 'move t t)
7502 (or
7503 (c-end-of-current-token base)
7504 ;; Handle operator identifiers, i.e. ignore any
7505 ;; operator token preceded by "operator".
7506 (save-excursion
7507 (and (c-safe (c-backward-sexp) t)
51c9af45 7508 (looking-at c-opt-op-identifier-prefix)))
ff959bab
MS
7509 (and (eq (char-before) ?<)
7510 (c-with-syntax-table c++-template-syntax-table
7511 (if (c-safe (goto-char (c-up-list-forward (point))))
7512 t
7513 (goto-char (point-max))
7514 nil)))))
7515 (setq base (point)))
7516
7517 (while (and
7518 (c-syntactic-re-search-forward "[;{=]" nil 'move t t)
7519 (c-end-of-current-token base))
7520 (setq base (point))))))
a66cd3ee
MS
7521
7522(defun c-beginning-of-decl-1 (&optional lim)
7523 ;; Go to the beginning of the current declaration, or the beginning
7524 ;; of the previous one if already at the start of it. Point won't
0386b551 7525 ;; be moved out of any surrounding paren. Return a cons cell of the
a66cd3ee
MS
7526 ;; form (MOVE . KNR-POS). MOVE is like the return value from
7527 ;; `c-beginning-of-statement-1'. If point skipped over some K&R
7528 ;; style argument declarations (and they are to be recognized) then
7529 ;; KNR-POS is set to the start of the first such argument
7530 ;; declaration, otherwise KNR-POS is nil. If LIM is non-nil, it's a
7531 ;; position that bounds the backward search.
7532 ;;
7533 ;; NB: Cases where the declaration continues after the block, as in
7534 ;; "struct foo { ... } bar;", are currently recognized as two
7535 ;; declarations, e.g. "struct foo { ... }" and "bar;" in this case.
0386b551
AM
7536 ;;
7537 ;; This function might do hidden buffer changes.
a66cd3ee
MS
7538 (catch 'return
7539 (let* ((start (point))
d9e94c22 7540 (last-stmt-start (point))
0386b551 7541 (move (c-beginning-of-statement-1 lim nil t)))
a66cd3ee 7542
a66cd3ee
MS
7543 ;; `c-beginning-of-statement-1' stops at a block start, but we
7544 ;; want to continue if the block doesn't begin a top level
2a15eb73
MS
7545 ;; construct, i.e. if it isn't preceded by ';', '}', ':', bob,
7546 ;; or an open paren.
d9e94c22 7547 (let ((beg (point)) tentative-move)
51c9af45
AM
7548 ;; Go back one "statement" each time round the loop until we're just
7549 ;; after a ;, }, or :, or at BOB or the start of a macro or start of
7550 ;; an ObjC method. This will move over a multiple declaration whose
7551 ;; components are comma separated.
d9e94c22
MS
7552 (while (and
7553 ;; Must check with c-opt-method-key in ObjC mode.
7554 (not (and c-opt-method-key
7555 (looking-at c-opt-method-key)))
7556 (/= last-stmt-start (point))
7557 (progn
7558 (c-backward-syntactic-ws lim)
7559 (not (memq (char-before) '(?\; ?} ?: nil))))
2a15eb73
MS
7560 (save-excursion
7561 (backward-char)
7562 (not (looking-at "\\s(")))
d9e94c22
MS
7563 ;; Check that we don't move from the first thing in a
7564 ;; macro to its header.
7565 (not (eq (setq tentative-move
0386b551 7566 (c-beginning-of-statement-1 lim nil t))
d9e94c22
MS
7567 'macro)))
7568 (setq last-stmt-start beg
7569 beg (point)
7570 move tentative-move))
7571 (goto-char beg))
7572
7573 (when c-recognize-knr-p
7574 (let ((fallback-pos (point)) knr-argdecl-start)
7575 ;; Handle K&R argdecls. Back up after the "statement" jumped
7576 ;; over by `c-beginning-of-statement-1', unless it was the
7577 ;; function body, in which case we're sitting on the opening
7578 ;; brace now. Then test if we're in a K&R argdecl region and
7579 ;; that we started at the other side of the first argdecl in
7580 ;; it.
7581 (unless (eq (char-after) ?{)
7582 (goto-char last-stmt-start))
7583 (if (and (setq knr-argdecl-start (c-in-knr-argdecl lim))
7584 (< knr-argdecl-start start)
7585 (progn
7586 (goto-char knr-argdecl-start)
0386b551 7587 (not (eq (c-beginning-of-statement-1 lim nil t) 'macro))))
d9e94c22
MS
7588 (throw 'return
7589 (cons (if (eq (char-after fallback-pos) ?{)
7590 'previous
7591 'same)
7592 knr-argdecl-start))
7593 (goto-char fallback-pos))))
7594
51c9af45
AM
7595 ;; `c-beginning-of-statement-1' counts each brace block as a separate
7596 ;; statement, so the result will be 'previous if we've moved over any.
7597 ;; So change our result back to 'same if necessary.
7598 ;;
7599 ;; If they were brace list initializers we might not have moved over a
7600 ;; declaration boundary though, so change it to 'same if we've moved
7601 ;; past a '=' before '{', but not ';'. (This ought to be integrated
7602 ;; into `c-beginning-of-statement-1', so we avoid this extra pass which
7603 ;; potentially can search over a large amount of text.). Take special
7604 ;; pains not to get mislead by C++'s "operator=", and the like.
d9e94c22
MS
7605 (if (and (eq move 'previous)
7606 (c-with-syntax-table (if (c-major-mode-is 'c++-mode)
7607 c++-template-syntax-table
7608 (syntax-table))
7609 (save-excursion
51c9af45
AM
7610 (and
7611 (progn
7612 (while ; keep going back to "[;={"s until we either find
7613 ; no more, or get to one which isn't an "operator ="
7614 (and (c-syntactic-re-search-forward "[;={]" start t t t)
7615 (eq (char-before) ?=)
7616 c-overloadable-operators-regexp
7617 c-opt-op-identifier-prefix
7618 (save-excursion
7619 (eq (c-backward-token-2) 0)
7620 (looking-at c-overloadable-operators-regexp)
7621 (eq (c-backward-token-2) 0)
7622 (looking-at c-opt-op-identifier-prefix))))
7623 (eq (char-before) ?=))
7624 (c-syntactic-re-search-forward "[;{]" start t t)
7625 (eq (char-before) ?{)
7626 (c-safe (goto-char (c-up-list-forward (point))) t)
7627 (not (c-syntactic-re-search-forward ";" start t t))))))
d9e94c22
MS
7628 (cons 'same nil)
7629 (cons move nil)))))
a66cd3ee
MS
7630
7631(defun c-end-of-decl-1 ()
7632 ;; Assuming point is at the start of a declaration (as detected by
7633 ;; e.g. `c-beginning-of-decl-1'), go to the end of it. Unlike
7634 ;; `c-beginning-of-decl-1', this function handles the case when a
7635 ;; block is followed by identifiers in e.g. struct declarations in C
7636 ;; or C++. If a proper end was found then t is returned, otherwise
7637 ;; point is moved as far as possible within the current sexp and nil
7638 ;; is returned. This function doesn't handle macros; use
7639 ;; `c-end-of-macro' instead in those cases.
0386b551
AM
7640 ;;
7641 ;; This function might do hidden buffer changes.
ce8c7486 7642 (let ((start (point))
a66cd3ee
MS
7643 (decl-syntax-table (if (c-major-mode-is 'c++-mode)
7644 c++-template-syntax-table
7645 (syntax-table))))
7646 (catch 'return
7647 (c-search-decl-header-end)
7648
7649 (when (and c-recognize-knr-p
7650 (eq (char-before) ?\;)
7651 (c-in-knr-argdecl start))
7652 ;; Stopped at the ';' in a K&R argdecl section which is
7653 ;; detected using the same criteria as in
7654 ;; `c-beginning-of-decl-1'. Move to the following block
7655 ;; start.
d9e94c22 7656 (c-syntactic-re-search-forward "{" nil 'move t))
a66cd3ee
MS
7657
7658 (when (eq (char-before) ?{)
7659 ;; Encountered a block in the declaration. Jump over it.
7660 (condition-case nil
7661 (goto-char (c-up-list-forward (point)))
d9e94c22
MS
7662 (error (goto-char (point-max))
7663 (throw 'return nil)))
a66cd3ee
MS
7664 (if (or (not c-opt-block-decls-with-vars-key)
7665 (save-excursion
7666 (c-with-syntax-table decl-syntax-table
7667 (let ((lim (point)))
7668 (goto-char start)
b3cf7e18
MS
7669 (not (and
7670 ;; Check for `c-opt-block-decls-with-vars-key'
7671 ;; before the first paren.
7672 (c-syntactic-re-search-forward
d9e94c22 7673 (concat "[;=\(\[{]\\|\\("
b3cf7e18
MS
7674 c-opt-block-decls-with-vars-key
7675 "\\)")
d9e94c22 7676 lim t t t)
b3cf7e18
MS
7677 (match-beginning 1)
7678 (not (eq (char-before) ?_))
d9e94c22
MS
7679 ;; Check that the first following paren is
7680 ;; the block.
7681 (c-syntactic-re-search-forward "[;=\(\[{]"
7682 lim t t t)
b3cf7e18 7683 (eq (char-before) ?{)))))))
a66cd3ee
MS
7684 ;; The declaration doesn't have any of the
7685 ;; `c-opt-block-decls-with-vars' keywords in the
7686 ;; beginning, so it ends here at the end of the block.
7687 (throw 'return t)))
7688
7689 (c-with-syntax-table decl-syntax-table
7690 (while (progn
7691 (if (eq (char-before) ?\;)
7692 (throw 'return t))
d9e94c22 7693 (c-syntactic-re-search-forward ";" nil 'move t))))
a66cd3ee 7694 nil)))
ce8c7486 7695
0386b551
AM
7696(defun c-looking-at-decl-block (containing-sexp goto-start &optional limit)
7697 ;; Assuming the point is at an open brace, check if it starts a
7698 ;; block that contains another declaration level, i.e. that isn't a
7699 ;; statement block or a brace list, and if so return non-nil.
7700 ;;
7701 ;; If the check is successful, the return value is the start of the
7702 ;; keyword that tells what kind of construct it is, i.e. typically
7703 ;; what `c-decl-block-key' matched. Also, if GOTO-START is set then
7704 ;; the point will be at the start of the construct, before any
7705 ;; leading specifiers, otherwise it's at the returned position.
7706 ;;
7707 ;; The point is clobbered if the check is unsuccessful.
7708 ;;
7709 ;; CONTAINING-SEXP is the position of the open of the surrounding
7710 ;; paren, or nil if none.
7711 ;;
7712 ;; The optional LIMIT limits the backward search for the start of
7713 ;; the construct. It's assumed to be at a syntactically relevant
7714 ;; position.
7715 ;;
7716 ;; If any template arglists are found in the searched region before
7717 ;; the open brace, they get marked with paren syntax.
7718 ;;
7719 ;; This function might do hidden buffer changes.
7720
7721 (let ((open-brace (point)) kwd-start first-specifier-pos)
7722 (c-syntactic-skip-backward c-block-prefix-charset limit t)
7723
7724 (when (and c-recognize-<>-arglists
7725 (eq (char-before) ?>))
7726 ;; Could be at the end of a template arglist.
7727 (let ((c-parse-and-markup-<>-arglists t)
7728 (c-disallow-comma-in-<>-arglists
7729 (and containing-sexp
7730 (not (eq (char-after containing-sexp) ?{)))))
7731 (while (and
7732 (c-backward-<>-arglist nil limit)
7733 (progn
7734 (c-syntactic-skip-backward c-block-prefix-charset limit t)
7735 (eq (char-before) ?>))))))
7736
7737 ;; Note: Can't get bogus hits inside template arglists below since they
7738 ;; have gotten paren syntax above.
7739 (when (and
7740 ;; If `goto-start' is set we begin by searching for the
7741 ;; first possible position of a leading specifier list.
7742 ;; The `c-decl-block-key' search continues from there since
7743 ;; we know it can't match earlier.
7744 (if goto-start
7745 (when (c-syntactic-re-search-forward c-symbol-start
7746 open-brace t t)
7747 (goto-char (setq first-specifier-pos (match-beginning 0)))
7748 t)
7749 t)
7750
7751 (cond
7752 ((c-syntactic-re-search-forward c-decl-block-key open-brace t t t)
7753 (goto-char (setq kwd-start (match-beginning 0)))
7754 (or
7755
7756 ;; Found a keyword that can't be a type?
7757 (match-beginning 1)
7758
7759 ;; Can be a type too, in which case it's the return type of a
7760 ;; function (under the assumption that no declaration level
7761 ;; block construct starts with a type).
7762 (not (c-forward-type))
7763
7764 ;; Jumped over a type, but it could be a declaration keyword
7765 ;; followed by the declared identifier that we've jumped over
7766 ;; instead (e.g. in "class Foo {"). If it indeed is a type
7767 ;; then we should be at the declarator now, so check for a
7768 ;; valid declarator start.
7769 ;;
7770 ;; Note: This doesn't cope with the case when a declared
7771 ;; identifier is followed by e.g. '(' in a language where '('
7772 ;; also might be part of a declarator expression. Currently
7773 ;; there's no such language.
7774 (not (or (looking-at c-symbol-start)
7775 (looking-at c-type-decl-prefix-key)))))
7776
7777 ;; In Pike a list of modifiers may be followed by a brace
7778 ;; to make them apply to many identifiers. Note that the
7779 ;; match data will be empty on return in this case.
7780 ((and (c-major-mode-is 'pike-mode)
7781 (progn
7782 (goto-char open-brace)
7783 (= (c-backward-token-2) 0))
7784 (looking-at c-specifier-key)
7785 ;; Use this variant to avoid yet another special regexp.
7786 (c-keyword-member (c-keyword-sym (match-string 1))
7787 'c-modifier-kwds))
7788 (setq kwd-start (point))
7789 t)))
7790
7791 ;; Got a match.
7792
7793 (if goto-start
7794 ;; Back up over any preceding specifiers and their clauses
7795 ;; by going forward from `first-specifier-pos', which is the
7796 ;; earliest possible position where the specifier list can
7797 ;; start.
7798 (progn
7799 (goto-char first-specifier-pos)
7800
7801 (while (< (point) kwd-start)
7802 (if (looking-at c-symbol-key)
7803 ;; Accept any plain symbol token on the ground that
7804 ;; it's a specifier masked through a macro (just
7805 ;; like `c-forward-decl-or-cast-1' skip forward over
7806 ;; such tokens).
7807 ;;
7808 ;; Could be more restrictive wrt invalid keywords,
7809 ;; but that'd only occur in invalid code so there's
7810 ;; no use spending effort on it.
7811 (let ((end (match-end 0)))
7812 (unless (c-forward-keyword-clause 0)
7813 (goto-char end)
7814 (c-forward-syntactic-ws)))
7815
7816 ;; Can't parse a declaration preamble and is still
7817 ;; before `kwd-start'. That means `first-specifier-pos'
7818 ;; was in some earlier construct. Search again.
7819 (if (c-syntactic-re-search-forward c-symbol-start
7820 kwd-start 'move t)
7821 (goto-char (setq first-specifier-pos (match-beginning 0)))
7822 ;; Got no preamble before the block declaration keyword.
7823 (setq first-specifier-pos kwd-start))))
7824
7825 (goto-char first-specifier-pos))
7826 (goto-char kwd-start))
7827
7828 kwd-start)))
ce8c7486 7829
a66cd3ee 7830(defun c-search-uplist-for-classkey (paren-state)
0386b551
AM
7831 ;; Check if the closest containing paren sexp is a declaration
7832 ;; block, returning a 2 element vector in that case. Aref 0
7833 ;; contains the bufpos at boi of the class key line, and aref 1
7834 ;; contains the bufpos of the open brace. This function is an
7835 ;; obsolete wrapper for `c-looking-at-decl-block'.
7836 ;;
7837 ;; This function might do hidden buffer changes.
7838 (let ((open-paren-pos (c-most-enclosing-brace paren-state)))
7839 (when open-paren-pos
7840 (save-excursion
7841 (goto-char open-paren-pos)
7842 (when (and (eq (char-after) ?{)
7843 (c-looking-at-decl-block
7844 (c-safe-position open-paren-pos paren-state)
7845 nil))
7846 (back-to-indentation)
7847 (vector (point) open-paren-pos))))))
785eecbb 7848
a66cd3ee 7849(defun c-inside-bracelist-p (containing-sexp paren-state)
785eecbb
RS
7850 ;; return the buffer position of the beginning of the brace list
7851 ;; statement if we're inside a brace list, otherwise return nil.
7852 ;; CONTAINING-SEXP is the buffer pos of the innermost containing
0386b551 7853 ;; paren. PAREN-STATE is the remainder of the state of enclosing
130c507e 7854 ;; braces
785eecbb
RS
7855 ;;
7856 ;; N.B.: This algorithm can potentially get confused by cpp macros
0386b551 7857 ;; placed in inconvenient locations. It's a trade-off we make for
785eecbb 7858 ;; speed.
0386b551
AM
7859 ;;
7860 ;; This function might do hidden buffer changes.
785eecbb 7861 (or
d9e94c22 7862 ;; This will pick up brace list declarations.
b2acd789
RS
7863 (c-safe
7864 (save-excursion
7865 (goto-char containing-sexp)
0ec8351b 7866 (c-forward-sexp -1)
b2acd789 7867 (let (bracepos)
d9e94c22 7868 (if (and (or (looking-at c-brace-list-key)
0ec8351b 7869 (progn (c-forward-sexp -1)
d9e94c22 7870 (looking-at c-brace-list-key)))
a66cd3ee 7871 (setq bracepos (c-down-list-forward (point)))
b2acd789
RS
7872 (not (c-crosses-statement-barrier-p (point)
7873 (- bracepos 2))))
7874 (point)))))
785eecbb
RS
7875 ;; this will pick up array/aggregate init lists, even if they are nested.
7876 (save-excursion
0ec8351b
BW
7877 (let ((class-key
7878 ;; Pike can have class definitions anywhere, so we must
7879 ;; check for the class key here.
7880 (and (c-major-mode-is 'pike-mode)
a66cd3ee
MS
7881 c-decl-block-key))
7882 bufpos braceassignp lim next-containing)
785eecbb
RS
7883 (while (and (not bufpos)
7884 containing-sexp)
a66cd3ee
MS
7885 (when paren-state
7886 (if (consp (car paren-state))
7887 (setq lim (cdr (car paren-state))
7888 paren-state (cdr paren-state))
7889 (setq lim (car paren-state)))
7890 (when paren-state
7891 (setq next-containing (car paren-state)
7892 paren-state (cdr paren-state))))
785eecbb 7893 (goto-char containing-sexp)
a66cd3ee
MS
7894 (if (c-looking-at-inexpr-block next-containing next-containing)
7895 ;; We're in an in-expression block of some kind. Do not
7896 ;; check nesting. We deliberately set the limit to the
7897 ;; containing sexp, so that c-looking-at-inexpr-block
7898 ;; doesn't check for an identifier before it.
0ec8351b
BW
7899 (setq containing-sexp nil)
7900 ;; see if the open brace is preceded by = or [...] in
7901 ;; this statement, but watch out for operator=
a66cd3ee 7902 (setq braceassignp 'dontknow)
d9e94c22 7903 (c-backward-token-2 1 t lim)
6393fef2 7904 ;; Checks to do only on the first sexp before the brace.
d9e94c22 7905 (when (and c-opt-inexpr-brace-list-key
6393fef2
RS
7906 (eq (char-after) ?\[))
7907 ;; In Java, an initialization brace list may follow
7908 ;; directly after "new Foo[]", so check for a "new"
7909 ;; earlier.
7910 (while (eq braceassignp 'dontknow)
7911 (setq braceassignp
d9e94c22
MS
7912 (cond ((/= (c-backward-token-2 1 t lim) 0) nil)
7913 ((looking-at c-opt-inexpr-brace-list-key) t)
6393fef2
RS
7914 ((looking-at "\\sw\\|\\s_\\|[.[]")
7915 ;; Carry on looking if this is an
7916 ;; identifier (may contain "." in Java)
7917 ;; or another "[]" sexp.
7918 'dontknow)
7919 (t nil)))))
7920 ;; Checks to do on all sexps before the brace, up to the
7921 ;; beginning of the statement.
7922 (while (eq braceassignp 'dontknow)
0ec8351b
BW
7923 (cond ((eq (char-after) ?\;)
7924 (setq braceassignp nil))
7925 ((and class-key
7926 (looking-at class-key))
7927 (setq braceassignp nil))
7928 ((eq (char-after) ?=)
7929 ;; We've seen a =, but must check earlier tokens so
7930 ;; that it isn't something that should be ignored.
7931 (setq braceassignp 'maybe)
7932 (while (and (eq braceassignp 'maybe)
d9e94c22 7933 (zerop (c-backward-token-2 1 t lim)))
0ec8351b
BW
7934 (setq braceassignp
7935 (cond
7936 ;; Check for operator =
51c9af45
AM
7937 ((and c-opt-op-identifier-prefix
7938 (looking-at c-opt-op-identifier-prefix))
0386b551 7939 nil)
130c507e
GM
7940 ;; Check for `<opchar>= in Pike.
7941 ((and (c-major-mode-is 'pike-mode)
7942 (or (eq (char-after) ?`)
7943 ;; Special case for Pikes
7944 ;; `[]=, since '[' is not in
7945 ;; the punctuation class.
7946 (and (eq (char-after) ?\[)
7947 (eq (char-before) ?`))))
7948 nil)
0ec8351b
BW
7949 ((looking-at "\\s.") 'maybe)
7950 ;; make sure we're not in a C++ template
7951 ;; argument assignment
a66cd3ee
MS
7952 ((and
7953 (c-major-mode-is 'c++-mode)
7954 (save-excursion
7955 (let ((here (point))
7956 (pos< (progn
7957 (skip-chars-backward "^<>")
7958 (point))))
7959 (and (eq (char-before) ?<)
7960 (not (c-crosses-statement-barrier-p
7961 pos< here))
7962 (not (c-in-literal))
7963 ))))
0ec8351b 7964 nil)
6393fef2
RS
7965 (t t))))))
7966 (if (and (eq braceassignp 'dontknow)
d9e94c22 7967 (/= (c-backward-token-2 1 t lim) 0))
6393fef2
RS
7968 (setq braceassignp nil)))
7969 (if (not braceassignp)
0ec8351b
BW
7970 (if (eq (char-after) ?\;)
7971 ;; Brace lists can't contain a semicolon, so we're done.
7972 (setq containing-sexp nil)
a66cd3ee
MS
7973 ;; Go up one level.
7974 (setq containing-sexp next-containing
7975 lim nil
7976 next-containing nil))
0ec8351b
BW
7977 ;; we've hit the beginning of the aggregate list
7978 (c-beginning-of-statement-1
a66cd3ee 7979 (c-most-enclosing-brace paren-state))
0ec8351b 7980 (setq bufpos (point))))
a66cd3ee 7981 )
785eecbb
RS
7982 bufpos))
7983 ))
7984
0ec8351b
BW
7985(defun c-looking-at-special-brace-list (&optional lim)
7986