* faces.el (face-all-attributes): Improved documentation (Bug#6767).
[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 16692.
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
c0209c2c
AM
2248(defun c-state-balance-parens-backwards (here- 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. Except when
2251 ;; there's a macro, HERE- and HERE+ are the same. Like this:
0ec1d2c5 2252 ;;
c0209c2c
AM
2253 ;; ............................................
2254 ;; | |
2255 ;; ( [ ( .........#macro.. ) ( ) ] )
2256 ;; ^ ^ ^ ^
2257 ;; | | | |
2258 ;; return HERE- HERE+ TOP
0ec1d2c5
AM
2259 ;;
2260 ;; If there aren't enough opening paren/brace/brackets, return the position
c0209c2c
AM
2261 ;; of the outermost one found, or HERE- if there are none. If there are no
2262 ;; closeing p/b/bs between HERE+ and TOP, return HERE-. HERE-/+ and TOP
2263 ;; must not be inside literals. Only the accessible portion of the buffer
2264 ;; will be scanned.
2265
2266 ;; PART 1: scan from `here+' up to `top', accumulating ")"s which enclose
2267 ;; `here'. Go round the next loop each time we pass over such a ")". These
2268 ;; probably match "("s before `here-'.
0ec1d2c5
AM
2269 (let (pos pa ren+1 lonely-rens)
2270 (save-excursion
2271 (save-restriction
2272 (narrow-to-region (point-min) top) ; This can move point, sometimes.
c0209c2c 2273 (setq pos here+)
0ec1d2c5
AM
2274 (c-safe
2275 (while
2276 (setq ren+1 (scan-lists pos 1 1)) ; might signal
2277 (setq lonely-rens (cons ren+1 lonely-rens)
2278 pos ren+1)))))
2279
c0209c2c 2280 ;; PART 2: Scan back before `here-' searching for the "("s
0ec1d2c5
AM
2281 ;; matching/mismatching the ")"s found above. We only need to direct the
2282 ;; caller to scan when we've encountered unmatched right parens.
c0209c2c
AM
2283 (setq pos here-)
2284 (when lonely-rens
2285 (c-safe
2286 (while
2287 (and lonely-rens ; actual values aren't used.
2288 (setq pa (scan-lists pos -1 1)))
2289 (setq pos pa)
2290 (setq lonely-rens (cdr lonely-rens)))))
2291 pos))
0ec1d2c5
AM
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
eb2f6eeb 2594 (let* ((in-macro-start ; start of macro containing (point-max) or nil.
0ec1d2c5
AM
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)))
657071fc 2627 (while (and c-state-cache (>= (c-state-cache-top-lparen) upper-lim))
0ec1d2c5
AM
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 2644 (setq pos (if (and good-pos-actual-macro-end
eb2f6eeb
AM
2645 (not (eq good-pos-actual-macro-start
2646 in-macro-start)))
0ec1d2c5
AM
2647 (1+ good-pos-actual-macro-end) ; get outside the macro as
2648 ; marked by a `category' text property.
2649 good-pos))
2650 (goto-char pos)
2651 (while (and c-state-cache
2652 (< (point) (point-max)))
2653 (cond
2654 ((null pps-state) ; first time through
2655 (setq target-depth -1))
2656 ((eq (car pps-state) target-depth) ; found closing ),},]
2657 (setq target-depth (1- (car pps-state))))
2658 ;; Do nothing when we've merely reached pps-point.
2659 )
2660
2661 ;; Scan!
2662 (setq pps-state
2663 (parse-partial-sexp
2664 (point) (if (< (point) pps-point) pps-point (point-max))
2665 target-depth
2666 nil pps-state))
2667
2668 (if (= (point) pps-point)
2669 (setq pps-point-state pps-state))
2670
2671 (when (eq (car pps-state) target-depth)
2672 (setq pos (point)) ; POS is now just after an R-paren/brace.
2673 (cond
2674 ((and (consp (car c-state-cache))
2675 (eq (point) (cdar c-state-cache)))
2676 ;; We've just moved out of the paren pair containing the brace-pair
2677 ;; at (car c-state-cache). `pair-beg' is where the open paren is,
2678 ;; and is potentially where the open brace of a cons in
2679 ;; c-state-cache will be.
2680 (setq pair-beg (car-safe (cdr c-state-cache))
2681 c-state-cache (cdr-safe (cdr c-state-cache)))) ; remove {}pair + containing Lparen.
2682 ((numberp (car c-state-cache))
2683 (setq pair-beg (car c-state-cache)
2684 c-state-cache (cdr c-state-cache))) ; remove this
2685 ; containing Lparen
2686 ((numberp (cadr c-state-cache))
2687 (setq pair-beg (cadr c-state-cache)
2688 c-state-cache (cddr c-state-cache))) ; Remove a paren pair
2689 ; together with enclosed brace pair.
2690 ;; (t nil) ; Ignore an unmated Rparen.
2691 )))
2692
2693 (if (< (point) pps-point)
2694 (setq pps-state (parse-partial-sexp (point) pps-point
2695 nil nil ; TARGETDEPTH, STOPBEFORE
2696 pps-state)))
2697
2698 ;; If the last paren pair we moved out of was actually a brace pair,
2699 ;; insert it into `c-state-cache'.
2700 (when (and pair-beg (eq (char-after pair-beg) ?{))
2701 (if (consp (car-safe c-state-cache))
2702 (setq c-state-cache (cdr c-state-cache)))
2703 (setq c-state-cache (cons (cons pair-beg pos)
2704 c-state-cache)))
2705
2706 (list pos scan-back-pos pps-state)))))
2707
2708(defun c-remove-stale-state-cache-backwards (here cache-pos)
2709 ;; Strip stale elements of `c-state-cache' by moving backwards through the
2710 ;; buffer, and inform the caller of the scenario detected.
2711 ;;
2712 ;; HERE is the position we're setting `c-state-cache' for.
2713 ;; CACHE-POS is just after the latest recorded position in `c-state-cache'
2714 ;; before HERE, or a position at or near point-min which isn't in a
2715 ;; literal.
2716 ;;
2717 ;; This function must only be called only when (> `c-state-cache-good-pos'
2718 ;; HERE). Usually the gap between CACHE-POS and HERE is large. It is thus
2719 ;; optimised to eliminate (or minimise) scanning between these two
2720 ;; positions.
2721 ;;
2722 ;; Return a three element list (GOOD-POS SCAN-BACK-POS FWD-FLAG), where:
2723 ;; o - GOOD-POS is a "good position", where `c-state-cache' is valid, or
2724 ;; could become so after missing elements are inserted into
2725 ;; `c-state-cache'. This is JUST AFTER an opening or closing
2726 ;; brace/paren/bracket which is already in `c-state-cache' or just before
2727 ;; one otherwise. exceptionally (when there's no such b/p/b handy) the BOL
2728 ;; before `here''s line, or the start of the literal containing it.
2729 ;; o - SCAN-BACK-POS, if non-nil, indicates there may be a brace pair
2730 ;; preceding POS which isn't recorded in `c-state-cache'. It is a position
2731 ;; to scan backwards from.
2732 ;; o - FWD-FLAG, if non-nil, indicates there may be parens/braces between
2733 ;; POS and HERE which aren't recorded in `c-state-cache'.
2734 ;;
2735 ;; The comments in this defun use "paren" to mean parenthesis or square
2736 ;; bracket (as contrasted with a brace), and "(" and ")" likewise.
2737 ;;
2738 ;; . {..} (..) (..) ( .. { } ) (...) ( .... . ..)
2739 ;; | | | | | |
2740 ;; CP E here D C good
2741 (let ((pos c-state-cache-good-pos)
2742 pa ren ; positions of "(" and ")"
2743 dropped-cons ; whether the last element dropped from `c-state-cache'
2744 ; was a cons (representing a brace-pair)
2745 good-pos ; see above.
2746 lit ; (START . END) of a literal containing some point.
2747 here-lit-start here-lit-end ; bounds of literal containing `here'
2748 ; or `here' itself.
c0209c2c 2749 here- here+ ; start/end of macro around HERE, or HERE
0ec1d2c5
AM
2750 (here-bol (c-point 'bol here))
2751 (too-far-back (max (- here c-state-cache-too-far) 1)))
2752
2753 ;; Remove completely irrelevant entries from `c-state-cache'.
2754 (while (and c-state-cache
2755 (>= (setq pa (c-state-cache-top-lparen)) here))
2756 (setq dropped-cons (consp (car c-state-cache)))
2757 (setq c-state-cache (cdr c-state-cache))
2758 (setq pos pa))
2759 ;; At this stage, (> pos here);
2760 ;; (< (c-state-cache-top-lparen) here) (or is nil).
2761
c0209c2c
AM
2762 (cond
2763 ((and (consp (car c-state-cache))
2764 (> (cdar c-state-cache) here))
2765 ;; CASE 1: The top of the cache is a brace pair which now encloses
2766 ;; `here'. As good-pos, return the address. of the "{". Since we've no
2767 ;; knowledge of what's inside these braces, we have no alternative but
2768 ;; to direct the caller to scan the buffer from the opening brace.
2769 (setq pos (caar c-state-cache))
2770 (setcar c-state-cache pos)
2771 (list (1+ pos) pos t)) ; return value. We've just converted a brace pair
2772 ; entry into a { entry, so the caller needs to
2773 ; search for a brace pair before the {.
2774
2775 ;; `here' might be inside a literal. Check for this.
2776 ((progn
2777 (setq lit (c-state-literal-at here)
2778 here-lit-start (or (car lit) here)
2779 here-lit-end (or (cdr lit) here))
2780 ;; Has `here' just "newly entered" a macro?
2781 (save-excursion
2782 (goto-char here-lit-start)
2783 (if (and (c-beginning-of-macro)
2784 (or (null c-state-old-cpp-beg)
2785 (not (= (point) c-state-old-cpp-beg))))
2786 (progn
2787 (setq here- (point))
2788 (c-end-of-macro)
2789 (setq here+ (point)))
2790 (setq here- here-lit-start
2791 here+ here-lit-end)))
2792
2793 ;; `here' might be nested inside any depth of parens (or brackets but
2794 ;; not braces). Scan backwards to find the outermost such opening
2795 ;; paren, if there is one. This will be the scan position to return.
2796 (save-restriction
2797 (narrow-to-region cache-pos (point-max))
2798 (setq pos (c-state-balance-parens-backwards here- here+ pos)))
2799 nil)) ; for the cond
2800
2801 ((< pos here-lit-start)
2802 ;; CASE 2: Address of outermost ( or [ which now encloses `here', but
2803 ;; didn't enclose the (previous) `c-state-cache-good-pos'. If there is
2804 ;; a brace pair preceding this, it will already be in `c-state-cache',
2805 ;; unless there was a brace pair after it, i.e. there'll only be one to
2806 ;; scan for if we've just deleted one.
2807 (list pos (and dropped-cons pos) t)) ; Return value.
2808
2809 ;; `here' isn't enclosed in a (previously unrecorded) bracket/paren.
2810 ;; Further forward scanning isn't needed, but we still need to find a
2811 ;; GOOD-POS. Step out of all enclosing "("s on HERE's line.
2812 ((progn
0ec1d2c5
AM
2813 (save-restriction
2814 (narrow-to-region here-bol (point-max))
2815 (setq pos here-lit-start)
2816 (c-safe (while (setq pa (scan-lists pos -1 1))
2817 (setq pos pa)))) ; might signal
c0209c2c
AM
2818 nil)) ; for the cond
2819
2820 ((setq ren (c-safe-scan-lists pos -1 -1 too-far-back))
2821 ;; CASE 3: After a }/)/] before `here''s BOL.
2822 (list (1+ ren) (and dropped-cons pos) nil)) ; Return value
2823
2824 (t
2825 ;; CASE 4; Best of a bad job: BOL before `here-bol', or beginning of
2826 ;; literal containing it.
2827 (setq good-pos (c-state-lit-beg (c-point 'bopl here-bol)))
2828 (list good-pos (and dropped-cons good-pos) nil)))))
0ec1d2c5
AM
2829
2830
2831;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2832;; Externally visible routines.
2833
2834(defun c-state-cache-init ()
2835 (setq c-state-cache nil
2836 c-state-cache-good-pos 1
2837 c-state-nonlit-pos-cache nil
2838 c-state-nonlit-pos-cache-limit 1
2839 c-state-brace-pair-desert nil
2840 c-state-point-min 1
2841 c-state-point-min-lit-type nil
2842 c-state-point-min-lit-start nil
2843 c-state-min-scan-pos 1
2844 c-state-old-cpp-beg nil
2845 c-state-old-cpp-end nil)
2846 (c-state-mark-point-min-literal))
2847
2848(defun c-invalidate-state-cache-1 (here)
2849 ;; Invalidate all info on `c-state-cache' that applies to the buffer at HERE
2850 ;; or higher and set `c-state-cache-good-pos' accordingly. The cache is
2851 ;; left in a consistent state.
2852 ;;
2853 ;; This is much like `c-whack-state-after', but it never changes a paren
2854 ;; pair element into an open paren element. Doing that would mean that the
2855 ;; new open paren wouldn't have the required preceding paren pair element.
785eecbb 2856 ;;
0ec1d2c5
AM
2857 ;; This function is called from c-after-change.
2858
2859 ;; The cache of non-literals:
2860 (if (< here c-state-nonlit-pos-cache-limit)
2861 (setq c-state-nonlit-pos-cache-limit here))
2862
2863 ;; `c-state-cache':
2864 ;; Case 1: if `here' is in a literal containing point-min, everything
2865 ;; becomes (or is already) nil.
2866 (if (or (null c-state-cache-good-pos)
2867 (< here (c-state-get-min-scan-pos)))
2868 (setq c-state-cache nil
2869 c-state-cache-good-pos nil
2870 c-state-min-scan-pos nil)
2871
2872;;; Truncate `c-state-cache' and set `c-state-cache-good-pos' to a value below
2873;;; `here'. To maintain its consistency, we may need to insert a new brace
2874;;; pair.
2875 (let ((here-bol (c-point 'bol here))
2876 too-high-pa ; recorded {/(/[ next above here, or nil.
2877 dropped-cons ; was the last removed element a brace pair?
2878 pa)
2879 ;; The easy bit - knock over-the-top bits off `c-state-cache'.
2880 (while (and c-state-cache
2881 (>= (setq pa (c-state-cache-top-paren)) here))
2882 (setq dropped-cons (consp (car c-state-cache))
2883 too-high-pa (c-state-cache-top-lparen)
2884 c-state-cache (cdr c-state-cache)))
2885
2886 ;; Do we need to add in an earlier brace pair, having lopped one off?
2887 (if (and dropped-cons
2888 (< too-high-pa (+ here c-state-cache-too-far)))
2889 (c-append-lower-brace-pair-to-state-cache too-high-pa here-bol))
2890 (setq c-state-cache-good-pos (or (c-state-cache-after-top-paren)
2891 (c-state-get-min-scan-pos)))))
2892
2893 ;; The brace-pair desert marker:
2894 (when (car c-state-brace-pair-desert)
2895 (if (< here (car c-state-brace-pair-desert))
2896 (setq c-state-brace-pair-desert nil)
2897 (if (< here (cdr c-state-brace-pair-desert))
2898 (setcdr c-state-brace-pair-desert here)))))
2899
2900(defun c-parse-state-1 ()
2901 ;; Find and record all noteworthy parens between some good point earlier in
2902 ;; the file and point. That good point is at least the beginning of the
2903 ;; top-level construct we are in, or the beginning of the preceding
2904 ;; top-level construct if we aren't in one.
2905 ;;
2906 ;; The returned value is a list of the noteworthy parens with the last one
2907 ;; first. If an element in the list is an integer, it's the position of an
2908 ;; open paren (of any type) which has not been closed before the point. If
2909 ;; an element is a cons, it gives the position of a closed BRACE paren
2910 ;; pair[*]; the car is the start brace position and the cdr is the position
2911 ;; following the closing brace. Only the last closed brace paren pair
2912 ;; before each open paren and before the point is recorded, and thus the
2913 ;; state never contains two cons elements in succession. When a close brace
2914 ;; has no matching open brace (e.g., the matching brace is outside the
2915 ;; visible region), it is not represented in the returned value.
2916 ;;
2917 ;; [*] N.B. The close "brace" might be a mismatching close bracket or paren.
2918 ;; This defun explicitly treats mismatching parens/braces/brackets as
2919 ;; matching. It is the open brace which makes it a "brace" pair.
2920 ;;
2921 ;; If POINT is within a macro, open parens and brace pairs within
2922 ;; THIS macro MIGHT be recorded. This depends on whether their
2923 ;; syntactic properties have been suppressed by
2924 ;; `c-neutralize-syntax-in-CPP'. This might need fixing (2008-12-11).
d9e94c22
MS
2925 ;;
2926 ;; Currently no characters which are given paren syntax with the
2927 ;; syntax-table property are recorded, i.e. angle bracket arglist
2928 ;; parens are never present here. Note that this might change.
2929 ;;
0386b551 2930 ;; BUG: This function doesn't cope entirely well with unbalanced
0ec1d2c5
AM
2931 ;; parens in macros. (2008-12-11: this has probably been resolved
2932 ;; by the function `c-neutralize-syntax-in-CPP'.) E.g. in the
2933 ;; following case the brace before the macro isn't balanced with the
2934 ;; one after it:
0386b551
AM
2935 ;;
2936 ;; {
2937 ;; #define X {
2938 ;; }
2939 ;;
0ec1d2c5
AM
2940 ;; Note to maintainers: this function DOES get called with point
2941 ;; within comments and strings, so don't assume it doesn't!
2942 ;;
0386b551 2943 ;; This function might do hidden buffer changes.
0ec1d2c5
AM
2944 (let* ((here (point))
2945 (here-bopl (c-point 'bopl))
2946 strategy ; 'forward, 'backward etc..
2947 ;; Candidate positions to start scanning from:
2948 cache-pos ; highest position below HERE already existing in
2949 ; cache (or 1).
2950 good-pos
2951 start-point
2952 bopl-state
2953 res
2954 scan-backward-pos scan-forward-p) ; used for 'backward.
2955 ;; If POINT-MIN has changed, adjust the cache
2956 (unless (= (point-min) c-state-point-min)
2957 (c-renarrow-state-cache))
2958
2959 ;; Strategy?
2960 (setq res (c-parse-state-get-strategy here c-state-cache-good-pos)
2961 strategy (car res)
2962 cache-pos (cadr res)
2963 start-point (nth 2 res))
2964
2965 (when (eq strategy 'BOD)
2966 (setq c-state-cache nil
2967 c-state-cache-good-pos start-point))
2968
2969 ;; SCAN!
2970 (save-restriction
2971 (cond
2972 ((memq strategy '(forward BOD))
0386b551 2973 (narrow-to-region (point-min) here)
0ec1d2c5
AM
2974 (setq res (c-remove-stale-state-cache start-point here-bopl))
2975 (setq cache-pos (car res)
2976 scan-backward-pos (cadr res)
2977 bopl-state (car (cddr res))) ; will be nil if (< here-bopl
2978 ; start-point)
2979 (if scan-backward-pos
2980 (c-append-lower-brace-pair-to-state-cache scan-backward-pos))
2981 (setq good-pos
2982 (c-append-to-state-cache cache-pos))
2983 (setq c-state-cache-good-pos
2984 (if (and bopl-state
2985 (< good-pos (- here c-state-cache-too-far)))
2986 (c-state-cache-non-literal-place here-bopl bopl-state)
2987 good-pos)))
2988
2989 ((eq strategy 'backward)
2990 (setq res (c-remove-stale-state-cache-backwards here cache-pos)
2991 good-pos (car res)
2992 scan-backward-pos (cadr res)
2993 scan-forward-p (car (cddr res)))
2994 (if scan-backward-pos
2995 (c-append-lower-brace-pair-to-state-cache
2996 scan-backward-pos))
2997 (setq c-state-cache-good-pos
2998 (if scan-forward-p
2999 (progn (narrow-to-region (point-min) here)
3000 (c-append-to-state-cache good-pos))
3001
3002 (c-get-cache-scan-pos good-pos))))
3003
3004 (t ; (eq strategy 'IN-LIT)
3005 (setq c-state-cache nil
3006 c-state-cache-good-pos nil)))))
3007
3008 c-state-cache)
3009
3010(defun c-invalidate-state-cache (here)
3011 ;; This is a wrapper over `c-invalidate-state-cache-1'.
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-invalidate-state-cache-1' without
3016 ;; worrying further about macros and template delimiters.
3017 (c-with-<->-as-parens-suppressed
3e8f7d91
AM
3018 (if (and c-state-old-cpp-beg
3019 (< c-state-old-cpp-beg here))
0ec1d2c5 3020 (c-with-all-but-one-cpps-commented-out
3e8f7d91
AM
3021 c-state-old-cpp-beg
3022 (min c-state-old-cpp-end here)
0ec1d2c5
AM
3023 (c-invalidate-state-cache-1 here))
3024 (c-with-cpps-commented-out
3025 (c-invalidate-state-cache-1 here)))))
3026
3027(defun c-parse-state ()
3028 ;; This is a wrapper over `c-parse-state-1'. See that function for a
3029 ;; description of the functionality and return value.
3030 ;;
3031 ;; It suppresses the syntactic effect of the < and > (template) brackets and
3032 ;; of all parens in preprocessor constructs, except for any such construct
3033 ;; containing point. We can then call `c-parse-state-1' without worrying
3034 ;; further about macros and template delimiters.
3035 (let (here-cpp-beg here-cpp-end)
3036 (save-excursion
3037 (when (c-beginning-of-macro)
3038 (setq here-cpp-beg (point))
3039 (unless
3040 (> (setq here-cpp-end (c-syntactic-end-of-macro))
3041 here-cpp-beg)
3042 (setq here-cpp-beg nil here-cpp-end nil))))
3043 ;; FIXME!!! Put in a `condition-case' here to protect the integrity of the
3044 ;; subsystem.
3045 (prog1
3046 (c-with-<->-as-parens-suppressed
3047 (if (and here-cpp-beg (> here-cpp-end here-cpp-beg))
3048 (c-with-all-but-one-cpps-commented-out
3049 here-cpp-beg here-cpp-end
3050 (c-parse-state-1))
3051 (c-with-cpps-commented-out
3052 (c-parse-state-1))))
3e8f7d91
AM
3053 (setq c-state-old-cpp-beg (and here-cpp-beg (copy-marker here-cpp-beg t))
3054 c-state-old-cpp-end (and here-cpp-end (copy-marker here-cpp-end t)))
3055 )))
0ec1d2c5
AM
3056
3057;; Debug tool to catch cache inconsistencies. This is called from
3058;; 000tests.el.
a66cd3ee
MS
3059(defvar c-debug-parse-state nil)
3060(unless (fboundp 'c-real-parse-state)
3061 (fset 'c-real-parse-state (symbol-function 'c-parse-state)))
3062(cc-bytecomp-defun c-real-parse-state)
3063(defun c-debug-parse-state ()
0ec1d2c5 3064 (let ((here (point)) (res1 (c-real-parse-state)) res2)
0386b551 3065 (let ((c-state-cache nil)
0ec1d2c5
AM
3066 (c-state-cache-good-pos 1)
3067 (c-state-nonlit-pos-cache nil)
3068 (c-state-nonlit-pos-cache-limit 1)
3069 (c-state-brace-pair-desert nil)
3070 (c-state-point-min 1)
3071 (c-state-point-min-lit-type nil)
3072 (c-state-point-min-lit-start nil)
3073 (c-state-min-scan-pos 1)
3074 (c-state-old-cpp-beg nil)
3075 (c-state-old-cpp-end nil))
a66cd3ee
MS
3076 (setq res2 (c-real-parse-state)))
3077 (unless (equal res1 res2)
0386b551
AM
3078 ;; The cache can actually go further back due to the ad-hoc way
3079 ;; the first paren is found, so try to whack off a bit of its
3080 ;; start before complaining.
3081 (save-excursion
3082 (goto-char (or (c-least-enclosing-brace res2) (point)))
3083 (c-beginning-of-defun-1)
3084 (while (not (or (bobp) (eq (char-after) ?{)))
3085 (c-beginning-of-defun-1))
3086 (unless (equal (c-whack-state-before (point) res1) res2)
0ec1d2c5 3087 (message (concat "c-parse-state inconsistency at %s: "
0386b551 3088 "using cache: %s, from scratch: %s")
0ec1d2c5 3089 here res1 res2))))
a66cd3ee 3090 res1))
0ec1d2c5 3091
a66cd3ee
MS
3092(defun c-toggle-parse-state-debug (&optional arg)
3093 (interactive "P")
3094 (setq c-debug-parse-state (c-calculate-state arg c-debug-parse-state))
3095 (fset 'c-parse-state (symbol-function (if c-debug-parse-state
3096 'c-debug-parse-state
3097 'c-real-parse-state)))
3098 (c-keep-region-active))
0386b551
AM
3099(when c-debug-parse-state
3100 (c-toggle-parse-state-debug 1))
a66cd3ee 3101
0ec1d2c5 3102\f
d9e94c22
MS
3103(defun c-whack-state-before (bufpos paren-state)
3104 ;; Whack off any state information from PAREN-STATE which lies
3105 ;; before BUFPOS. Not destructive on PAREN-STATE.
d9e94c22
MS
3106 (let* ((newstate (list nil))
3107 (ptr newstate)
3108 car)
3109 (while paren-state
3110 (setq car (car paren-state)
3111 paren-state (cdr paren-state))
3112 (if (< (if (consp car) (car car) car) bufpos)
3113 (setq paren-state nil)
3114 (setcdr ptr (list car))
3115 (setq ptr (cdr ptr))))
3116 (cdr newstate)))
3117
3118(defun c-whack-state-after (bufpos paren-state)
3119 ;; Whack off any state information from PAREN-STATE which lies at or
3120 ;; after BUFPOS. Not destructive on PAREN-STATE.
d9e94c22
MS
3121 (catch 'done
3122 (while paren-state
3123 (let ((car (car paren-state)))
3124 (if (consp car)
3125 ;; just check the car, because in a balanced brace
3126 ;; expression, it must be impossible for the corresponding
3127 ;; close brace to be before point, but the open brace to
3128 ;; be after.
3129 (if (<= bufpos (car car))
3130 nil ; whack it off
3131 (if (< bufpos (cdr car))
3132 ;; its possible that the open brace is before
3133 ;; bufpos, but the close brace is after. In that
3134 ;; case, convert this to a non-cons element. The
3135 ;; rest of the state is before bufpos, so we're
3136 ;; done.
3137 (throw 'done (cons (car car) (cdr paren-state)))
3138 ;; we know that both the open and close braces are
3139 ;; before bufpos, so we also know that everything else
3140 ;; on state is before bufpos.
3141 (throw 'done paren-state)))
3142 (if (<= bufpos car)
3143 nil ; whack it off
3144 ;; it's before bufpos, so everything else should too.
3145 (throw 'done paren-state)))
3146 (setq paren-state (cdr paren-state)))
3147 nil)))
3148
3149(defun c-most-enclosing-brace (paren-state &optional bufpos)
3150 ;; Return the bufpos of the innermost enclosing open paren before
0386b551 3151 ;; bufpos, or nil if none was found.
d9e94c22
MS
3152 (let (enclosingp)
3153 (or bufpos (setq bufpos 134217727))
3154 (while paren-state
3155 (setq enclosingp (car paren-state)
3156 paren-state (cdr paren-state))
3157 (if (or (consp enclosingp)
3158 (>= enclosingp bufpos))
3159 (setq enclosingp nil)
d9e94c22
MS
3160 (setq paren-state nil)))
3161 enclosingp))
3162
0386b551
AM
3163(defun c-least-enclosing-brace (paren-state)
3164 ;; Return the bufpos of the outermost enclosing open paren, or nil
3165 ;; if none was found.
d9e94c22 3166 (let (pos elem)
d9e94c22
MS
3167 (while paren-state
3168 (setq elem (car paren-state)
3169 paren-state (cdr paren-state))
0386b551
AM
3170 (if (integerp elem)
3171 (setq pos elem)))
d9e94c22
MS
3172 pos))
3173
3174(defun c-safe-position (bufpos paren-state)
0386b551
AM
3175 ;; Return the closest "safe" position recorded on PAREN-STATE that
3176 ;; is higher up than BUFPOS. Return nil if PAREN-STATE doesn't
3177 ;; contain any. Return nil if BUFPOS is nil, which is useful to
3178 ;; find the closest limit before a given limit that might be nil.
d9e94c22 3179 ;;
0386b551
AM
3180 ;; A "safe" position is a position at or after a recorded open
3181 ;; paren, or after a recorded close paren. The returned position is
3182 ;; thus either the first position after a close brace, or the first
3183 ;; position after an enclosing paren, or at the enclosing paren in
3184 ;; case BUFPOS is immediately after it.
d9e94c22
MS
3185 (when bufpos
3186 (let (elem)
3187 (catch 'done
3188 (while paren-state
3189 (setq elem (car paren-state))
3190 (if (consp elem)
3191 (cond ((< (cdr elem) bufpos)
3192 (throw 'done (cdr elem)))
3193 ((< (car elem) bufpos)
3194 ;; See below.
3195 (throw 'done (min (1+ (car elem)) bufpos))))
3196 (if (< elem bufpos)
3197 ;; elem is the position at and not after the opening paren, so
3198 ;; we can go forward one more step unless it's equal to
3199 ;; bufpos. This is useful in some cases avoid an extra paren
3200 ;; level between the safe position and bufpos.
3201 (throw 'done (min (1+ elem) bufpos))))
3202 (setq paren-state (cdr paren-state)))))))
3203
3204(defun c-beginning-of-syntax ()
3205 ;; This is used for `font-lock-beginning-of-syntax-function'. It
3206 ;; goes to the closest previous point that is known to be outside
3207 ;; any string literal or comment. `c-state-cache' is used if it has
3208 ;; a position in the vicinity.
3209 (let* ((paren-state c-state-cache)
3210 elem
3211
3212 (pos (catch 'done
3213 ;; Note: Similar code in `c-safe-position'. The
3214 ;; difference is that we accept a safe position at
3215 ;; the point and don't bother to go forward past open
3216 ;; parens.
3217 (while paren-state
3218 (setq elem (car paren-state))
3219 (if (consp elem)
3220 (cond ((<= (cdr elem) (point))
3221 (throw 'done (cdr elem)))
3222 ((<= (car elem) (point))
3223 (throw 'done (car elem))))
3224 (if (<= elem (point))
3225 (throw 'done elem)))
3226 (setq paren-state (cdr paren-state)))
3227 (point-min))))
3228
3229 (if (> pos (- (point) 4000))
3230 (goto-char pos)
3231 ;; The position is far back. Try `c-beginning-of-defun-1'
3232 ;; (although we can't be entirely sure it will go to a position
3233 ;; outside a comment or string in current emacsen). FIXME:
3234 ;; Consult `syntax-ppss' here.
3235 (c-beginning-of-defun-1)
3236 (if (< (point) pos)
3237 (goto-char pos)))))
3238
3239\f
3240;; Tools for scanning identifiers and other tokens.
3241
3242(defun c-on-identifier ()
3243 "Return non-nil if the point is on or directly after an identifier.
3244Keywords are recognized and not considered identifiers. If an
3245identifier is detected, the returned value is its starting position.
0386b551
AM
3246If an identifier ends at the point and another begins at it \(can only
3247happen in Pike) then the point for the preceding one is returned.
d9e94c22 3248
0386b551
AM
3249Note that this function might do hidden buffer changes. See the
3250comment at the start of cc-engine.el for more info."
3251
3252 ;; FIXME: Shouldn't this function handle "operator" in C++?
d9e94c22
MS
3253
3254 (save-excursion
0386b551
AM
3255 (skip-syntax-backward "w_")
3256
3257 (or
3258
3259 ;; Check for a normal (non-keyword) identifier.
3260 (and (looking-at c-symbol-start)
3261 (not (looking-at c-keywords-regexp))
3262 (point))
3263
3264 (when (c-major-mode-is 'pike-mode)
3265 ;; Handle the `<operator> syntax in Pike.
3266 (let ((pos (point)))
3267 (skip-chars-backward "-!%&*+/<=>^|~[]()")
3268 (and (if (< (skip-chars-backward "`") 0)
3269 t
3270 (goto-char pos)
3271 (eq (char-after) ?\`))
3272 (looking-at c-symbol-key)
3273 (>= (match-end 0) pos)
3274 (point))))
3275
3276 ;; Handle the "operator +" syntax in C++.
3277 (when (and c-overloadable-operators-regexp
3278 (= (c-backward-token-2 0) 0))
3279
3280 (cond ((and (looking-at c-overloadable-operators-regexp)
51c9af45 3281 (or (not c-opt-op-identifier-prefix)
0386b551 3282 (and (= (c-backward-token-2 1) 0)
51c9af45 3283 (looking-at c-opt-op-identifier-prefix))))
0386b551
AM
3284 (point))
3285
3286 ((save-excursion
51c9af45
AM
3287 (and c-opt-op-identifier-prefix
3288 (looking-at c-opt-op-identifier-prefix)
0386b551
AM
3289 (= (c-forward-token-2 1) 0)
3290 (looking-at c-overloadable-operators-regexp)))
3291 (point))))
3292
3293 )))
d9e94c22
MS
3294
3295(defsubst c-simple-skip-symbol-backward ()
3296 ;; If the point is at the end of a symbol then skip backward to the
3297 ;; beginning of it. Don't move otherwise. Return non-nil if point
3298 ;; moved.
0386b551
AM
3299 ;;
3300 ;; This function might do hidden buffer changes.
d9e94c22
MS
3301 (or (< (skip-syntax-backward "w_") 0)
3302 (and (c-major-mode-is 'pike-mode)
3303 ;; Handle the `<operator> syntax in Pike.
3304 (let ((pos (point)))
2a15eb73 3305 (if (and (< (skip-chars-backward "-!%&*+/<=>^|~[]()") 0)
d9e94c22
MS
3306 (< (skip-chars-backward "`") 0)
3307 (looking-at c-symbol-key)
3308 (>= (match-end 0) pos))
3309 t
3310 (goto-char pos)
3311 nil)))))
3312
0386b551 3313(defun c-beginning-of-current-token (&optional back-limit)
d9e94c22
MS
3314 ;; Move to the beginning of the current token. Do not move if not
3315 ;; in the middle of one. BACK-LIMIT may be used to bound the
3316 ;; backward search; if given it's assumed to be at the boundary
a85fd6da 3317 ;; between two tokens. Return non-nil if the point is moved, nil
580fba94 3318 ;; otherwise.
0386b551
AM
3319 ;;
3320 ;; This function might do hidden buffer changes.
d9e94c22 3321 (let ((start (point)))
580fba94
AM
3322 (if (looking-at "\\w\\|\\s_")
3323 (skip-syntax-backward "w_" back-limit)
3324 (when (< (skip-syntax-backward ".()" back-limit) 0)
3325 (while (let ((pos (or (and (looking-at c-nonsymbol-token-regexp)
3326 (match-end 0))
3327 ;; `c-nonsymbol-token-regexp' should always match
3328 ;; since we've skipped backward over punctuator
3329 ;; or paren syntax, but consume one char in case
3330 ;; it doesn't so that we don't leave point before
3331 ;; some earlier incorrect token.
3332 (1+ (point)))))
3333 (if (<= pos start)
3334 (goto-char pos))))))
3335 (< (point) start)))
d9e94c22 3336
ff959bab 3337(defun c-end-of-current-token (&optional back-limit)
d9e94c22
MS
3338 ;; Move to the end of the current token. Do not move if not in the
3339 ;; middle of one. BACK-LIMIT may be used to bound the backward
3340 ;; search; if given it's assumed to be at the boundary between two
ff959bab 3341 ;; tokens. Return non-nil if the point is moved, nil otherwise.
0386b551
AM
3342 ;;
3343 ;; This function might do hidden buffer changes.
d9e94c22
MS
3344 (let ((start (point)))
3345 (cond ((< (skip-syntax-backward "w_" (1- start)) 0)
3346 (skip-syntax-forward "w_"))
3347 ((< (skip-syntax-backward ".()" back-limit) 0)
3348 (while (progn
3349 (if (looking-at c-nonsymbol-token-regexp)
3350 (goto-char (match-end 0))
3351 ;; `c-nonsymbol-token-regexp' should always match since
3352 ;; we've skipped backward over punctuator or paren
3353 ;; syntax, but move forward in case it doesn't so that
3354 ;; we don't leave point earlier than we started with.
3355 (forward-char))
ff959bab
MS
3356 (< (point) start)))))
3357 (> (point) start)))
d9e94c22
MS
3358
3359(defconst c-jump-syntax-balanced
3360 (if (memq 'gen-string-delim c-emacs-features)
3361 "\\w\\|\\s_\\|\\s\(\\|\\s\)\\|\\s\"\\|\\s|"
3362 "\\w\\|\\s_\\|\\s\(\\|\\s\)\\|\\s\""))
3363
3364(defconst c-jump-syntax-unbalanced
3365 (if (memq 'gen-string-delim c-emacs-features)
3366 "\\w\\|\\s_\\|\\s\"\\|\\s|"
3367 "\\w\\|\\s_\\|\\s\""))
3368
3369(defun c-forward-token-2 (&optional count balanced limit)
3370 "Move forward by tokens.
3371A token is defined as all symbols and identifiers which aren't
3372syntactic whitespace \(note that multicharacter tokens like \"==\" are
3373treated properly). Point is always either left at the beginning of a
3374token or not moved at all. COUNT specifies the number of tokens to
3375move; a negative COUNT moves in the opposite direction. A COUNT of 0
3376moves to the next token beginning only if not already at one. If
3377BALANCED is true, move over balanced parens, otherwise move into them.
3378Also, if BALANCED is true, never move out of an enclosing paren.
3379
3380LIMIT sets the limit for the movement and defaults to the point limit.
3381The case when LIMIT is set in the middle of a token, comment or macro
3382is handled correctly, i.e. the point won't be left there.
3383
3384Return the number of tokens left to move \(positive or negative). If
3385BALANCED is true, a move over a balanced paren counts as one. Note
3386that if COUNT is 0 and no appropriate token beginning is found, 1 will
3387be returned. Thus, a return value of 0 guarantees that point is at
3388the requested position and a return value less \(without signs) than
0386b551
AM
3389COUNT guarantees that point is at the beginning of some token.
3390
3391Note that this function might do hidden buffer changes. See the
3392comment at the start of cc-engine.el for more info."
d9e94c22
MS
3393
3394 (or count (setq count 1))
3395 (if (< count 0)
3396 (- (c-backward-token-2 (- count) balanced limit))
3397
3398 (let ((jump-syntax (if balanced
3399 c-jump-syntax-balanced
3400 c-jump-syntax-unbalanced))
3401 (last (point))
3402 (prev (point)))
3403
3404 (if (zerop count)
3405 ;; If count is zero we should jump if in the middle of a token.
3406 (c-end-of-current-token))
3407
3408 (save-restriction
3409 (if limit (narrow-to-region (point-min) limit))
3410 (if (/= (point)
3411 (progn (c-forward-syntactic-ws) (point)))
3412 ;; Skip whitespace. Count this as a move if we did in
3413 ;; fact move.
3414 (setq count (max (1- count) 0)))
3415
3416 (if (eobp)
3417 ;; Moved out of bounds. Make sure the returned count isn't zero.
3418 (progn
3419 (if (zerop count) (setq count 1))
3420 (goto-char last))
3421
3422 ;; Use `condition-case' to avoid having the limit tests
3423 ;; inside the loop.
3424 (condition-case nil
3425 (while (and
3426 (> count 0)
3427 (progn
3428 (setq last (point))
3429 (cond ((looking-at jump-syntax)
3430 (goto-char (scan-sexps (point) 1))
3431 t)
3432 ((looking-at c-nonsymbol-token-regexp)
3433 (goto-char (match-end 0))
3434 t)
3435 ;; `c-nonsymbol-token-regexp' above should always
3436 ;; match if there are correct tokens. Try to
3437 ;; widen to see if the limit was set in the
3438 ;; middle of one, else fall back to treating
3439 ;; the offending thing as a one character token.
3440 ((and limit
3441 (save-restriction
3442 (widen)
3443 (looking-at c-nonsymbol-token-regexp)))
3444 nil)
3445 (t
3446 (forward-char)
3447 t))))
3448 (c-forward-syntactic-ws)
3449 (setq prev last
3450 count (1- count)))
3451 (error (goto-char last)))
3452
3453 (when (eobp)
3454 (goto-char prev)
3455 (setq count (1+ count)))))
3456
3457 count)))
3458
3459(defun c-backward-token-2 (&optional count balanced limit)
3460 "Move backward by tokens.
3461See `c-forward-token-2' for details."
3462
3463 (or count (setq count 1))
3464 (if (< count 0)
3465 (- (c-forward-token-2 (- count) balanced limit))
3466
3467 (or limit (setq limit (point-min)))
3468 (let ((jump-syntax (if balanced
3469 c-jump-syntax-balanced
3470 c-jump-syntax-unbalanced))
3471 (last (point)))
3472
3473 (if (zerop count)
3474 ;; The count is zero so try to skip to the beginning of the
3475 ;; current token.
3476 (if (> (point)
3477 (progn (c-beginning-of-current-token) (point)))
3478 (if (< (point) limit)
3479 ;; The limit is inside the same token, so return 1.
3480 (setq count 1))
3481
3482 ;; We're not in the middle of a token. If there's
3483 ;; whitespace after the point then we must move backward,
3484 ;; so set count to 1 in that case.
3485 (and (looking-at c-syntactic-ws-start)
3486 ;; If we're looking at a '#' that might start a cpp
3487 ;; directive then we have to do a more elaborate check.
3488 (or (/= (char-after) ?#)
3489 (not c-opt-cpp-prefix)
3490 (save-excursion
3491 (and (= (point)
3492 (progn (beginning-of-line)
3493 (looking-at "[ \t]*")
3494 (match-end 0)))
3495 (or (bobp)
3496 (progn (backward-char)
3497 (not (eq (char-before) ?\\)))))))
3498 (setq count 1))))
3499
3500 ;; Use `condition-case' to avoid having to check for buffer
3501 ;; limits in `backward-char', `scan-sexps' and `goto-char' below.
3502 (condition-case nil
3503 (while (and
3504 (> count 0)
3505 (progn
3506 (c-backward-syntactic-ws)
3507 (backward-char)
3508 (if (looking-at jump-syntax)
3509 (goto-char (scan-sexps (1+ (point)) -1))
3510 ;; This can be very inefficient if there's a long
3511 ;; sequence of operator tokens without any separation.
3512 ;; That doesn't happen in practice, anyway.
3513 (c-beginning-of-current-token))
3514 (>= (point) limit)))
3515 (setq last (point)
3516 count (1- count)))
3517 (error (goto-char last)))
3518
3519 (if (< (point) limit)
3520 (goto-char last))
3521
3522 count)))
3523
3524(defun c-forward-token-1 (&optional count balanced limit)
3525 "Like `c-forward-token-2' but doesn't treat multicharacter operator
3526tokens like \"==\" as single tokens, i.e. all sequences of symbol
3527characters are jumped over character by character. This function is
3528for compatibility only; it's only a wrapper over `c-forward-token-2'."
3529 (let ((c-nonsymbol-token-regexp "\\s.\\|\\s\(\\|\\s\)"))
3530 (c-forward-token-2 count balanced limit)))
3531
3532(defun c-backward-token-1 (&optional count balanced limit)
3533 "Like `c-backward-token-2' but doesn't treat multicharacter operator
3534tokens like \"==\" as single tokens, i.e. all sequences of symbol
3535characters are jumped over character by character. This function is
3536for compatibility only; it's only a wrapper over `c-backward-token-2'."
3537 (let ((c-nonsymbol-token-regexp "\\s.\\|\\s\(\\|\\s\)"))
3538 (c-backward-token-2 count balanced limit)))
3539
3540\f
3541;; Tools for doing searches restricted to syntactically relevant text.
3542
3543(defun c-syntactic-re-search-forward (regexp &optional bound noerror
3544 paren-level not-inside-token
3545 lookbehind-submatch)
3546 "Like `re-search-forward', but only report matches that are found
3547in syntactically significant text. I.e. matches in comments, macros
3548or string literals are ignored. The start point is assumed to be
3549outside any comment, macro or string literal, or else the content of
3550that region is taken as syntactically significant text.
3551
3552If PAREN-LEVEL is non-nil, an additional restriction is added to
2a15eb73
MS
3553ignore matches in nested paren sexps. The search will also not go
3554outside the current list sexp, which has the effect that if the point
3555should be moved to BOUND when no match is found \(i.e. NOERROR is
3556neither nil nor t), then it will be at the closing paren if the end of
3557the current list sexp is encountered first.
d9e94c22
MS
3558
3559If NOT-INSIDE-TOKEN is non-nil, matches in the middle of tokens are
3560ignored. Things like multicharacter operators and special symbols
3561\(e.g. \"`()\" in Pike) are handled but currently not floating point
3562constants.
3563
3564If LOOKBEHIND-SUBMATCH is non-nil, it's taken as a number of a
3565subexpression in REGEXP. The end of that submatch is used as the
3566position to check for syntactic significance. If LOOKBEHIND-SUBMATCH
3567isn't used or if that subexpression didn't match then the start
3568position of the whole match is used instead. The \"look behind\"
3569subexpression is never tested before the starting position, so it
3570might be a good idea to include \\=\\= as a match alternative in it.
3571
3572Optimization note: Matches might be missed if the \"look behind\"
2a15eb73 3573subexpression can match the end of nonwhite syntactic whitespace,
d9e94c22 3574i.e. the end of comments or cpp directives. This since the function
2a15eb73
MS
3575skips over such things before resuming the search. It's on the other
3576hand not safe to assume that the \"look behind\" subexpression never
3577matches syntactic whitespace.
3578
3579Bug: Unbalanced parens inside cpp directives are currently not handled
3580correctly \(i.e. they don't get ignored as they should) when
0386b551
AM
3581PAREN-LEVEL is set.
3582
3583Note that this function might do hidden buffer changes. See the
3584comment at the start of cc-engine.el for more info."
d9e94c22
MS
3585
3586 (or bound (setq bound (point-max)))
3587 (if paren-level (setq paren-level -1))
3588
3589 ;;(message "c-syntactic-re-search-forward %s %s %S" (point) bound regexp)
3590
3591 (let ((start (point))
2a15eb73
MS
3592 tmp
3593 ;; Start position for the last search.
3594 search-pos
3595 ;; The `parse-partial-sexp' state between the start position
3596 ;; and the point.
3597 state
3598 ;; The current position after the last state update. The next
3599 ;; `parse-partial-sexp' continues from here.
3600 (state-pos (point))
3601 ;; The position at which to check the state and the state
3602 ;; there. This is separate from `state-pos' since we might
3603 ;; need to back up before doing the next search round.
3604 check-pos check-state
3605 ;; Last position known to end a token.
d9e94c22 3606 (last-token-end-pos (point-min))
2a15eb73
MS
3607 ;; Set when a valid match is found.
3608 found)
d9e94c22
MS
3609
3610 (condition-case err
3611 (while
3612 (and
2a15eb73
MS
3613 (progn
3614 (setq search-pos (point))
3615 (re-search-forward regexp bound noerror))
d9e94c22
MS
3616
3617 (progn
2a15eb73
MS
3618 (setq state (parse-partial-sexp
3619 state-pos (match-beginning 0) paren-level nil state)
3620 state-pos (point))
d9e94c22 3621 (if (setq check-pos (and lookbehind-submatch
2a15eb73
MS
3622 (or (not paren-level)
3623 (>= (car state) 0))
d9e94c22
MS
3624 (match-end lookbehind-submatch)))
3625 (setq check-state (parse-partial-sexp
2a15eb73
MS
3626 state-pos check-pos paren-level nil state))
3627 (setq check-pos state-pos
d9e94c22
MS
3628 check-state state))
3629
2a15eb73
MS
3630 ;; NOTE: If we got a look behind subexpression and get
3631 ;; an insignificant match in something that isn't
d9e94c22
MS
3632 ;; syntactic whitespace (i.e. strings or in nested
3633 ;; parentheses), then we can never skip more than a
2a15eb73
MS
3634 ;; single character from the match start position
3635 ;; (i.e. `state-pos' here) before continuing the
3636 ;; search. That since the look behind subexpression
3637 ;; might match the end of the insignificant region in
3638 ;; the next search.
d9e94c22
MS
3639
3640 (cond
d9e94c22
MS
3641 ((elt check-state 7)
3642 ;; Match inside a line comment. Skip to eol. Use
3643 ;; `re-search-forward' instead of `skip-chars-forward' to get
3644 ;; the right bound behavior.
3645 (re-search-forward "[\n\r]" bound noerror))
3646
3647 ((elt check-state 4)
3648 ;; Match inside a block comment. Skip to the '*/'.
3649 (search-forward "*/" bound noerror))
3650
3651 ((and (not (elt check-state 5))
3652 (eq (char-before check-pos) ?/)
2a15eb73 3653 (not (c-get-char-property (1- check-pos) 'syntax-table))
d9e94c22
MS
3654 (memq (char-after check-pos) '(?/ ?*)))
3655 ;; Match in the middle of the opener of a block or line
3656 ;; comment.
3657 (if (= (char-after check-pos) ?/)
3658 (re-search-forward "[\n\r]" bound noerror)
3659 (search-forward "*/" bound noerror)))
3660
2a15eb73
MS
3661 ;; The last `parse-partial-sexp' above might have
3662 ;; stopped short of the real check position if the end
3663 ;; of the current sexp was encountered in paren-level
3664 ;; mode. The checks above are always false in that
3665 ;; case, and since they can do better skipping in
3666 ;; lookbehind-submatch mode, we do them before
3667 ;; checking the paren level.
3668
3669 ((and paren-level
3670 (/= (setq tmp (car check-state)) 0))
3671 ;; Check the paren level first since we're short of the
3672 ;; syntactic checking position if the end of the
3673 ;; current sexp was encountered by `parse-partial-sexp'.
3674 (if (> tmp 0)
3675
3676 ;; Inside a nested paren sexp.
3677 (if lookbehind-submatch
3678 ;; See the NOTE above.
3679 (progn (goto-char state-pos) t)
3680 ;; Skip out of the paren quickly.
3681 (setq state (parse-partial-sexp state-pos bound 0 nil state)
3682 state-pos (point)))
3683
3684 ;; Have exited the current paren sexp.
3685 (if noerror
3686 (progn
3687 ;; The last `parse-partial-sexp' call above
3688 ;; has left us just after the closing paren
3689 ;; in this case, so we can modify the bound
3690 ;; to leave the point at the right position
3691 ;; upon return.
3692 (setq bound (1- (point)))
3693 nil)
3694 (signal 'search-failed (list regexp)))))
3695
3696 ((setq tmp (elt check-state 3))
3697 ;; Match inside a string.
3698 (if (or lookbehind-submatch
3699 (not (integerp tmp)))
3700 ;; See the NOTE above.
3701 (progn (goto-char state-pos) t)
3702 ;; Skip to the end of the string before continuing.
3703 (let ((ender (make-string 1 tmp)) (continue t))
3704 (while (if (search-forward ender bound noerror)
3705 (progn
3706 (setq state (parse-partial-sexp
3707 state-pos (point) nil nil state)
3708 state-pos (point))
3709 (elt state 3))
3710 (setq continue nil)))
3711 continue)))
d9e94c22
MS
3712
3713 ((save-excursion
3714 (save-match-data
3715 (c-beginning-of-macro start)))
3716 ;; Match inside a macro. Skip to the end of it.
3717 (c-end-of-macro)
3718 (cond ((<= (point) bound) t)
3719 (noerror nil)
2a15eb73 3720 (t (signal 'search-failed (list regexp)))))
d9e94c22 3721
2a15eb73
MS
3722 ((and not-inside-token
3723 (or (< check-pos last-token-end-pos)
3724 (< check-pos
3725 (save-excursion
3726 (goto-char check-pos)
3727 (save-match-data
3728 (c-end-of-current-token last-token-end-pos))
3729 (setq last-token-end-pos (point))))))
3730 ;; Inside a token.
3731 (if lookbehind-submatch
3732 ;; See the NOTE above.
3733 (goto-char state-pos)
3734 (goto-char (min last-token-end-pos bound))))
d9e94c22
MS
3735
3736 (t
3737 ;; A real match.
3738 (setq found t)
2a15eb73
MS
3739 nil)))
3740
3741 ;; Should loop to search again, but take care to avoid
3742 ;; looping on the same spot.
3743 (or (/= search-pos (point))
3744 (if (= (point) bound)
3745 (if noerror
3746 nil
3747 (signal 'search-failed (list regexp)))
3748 (forward-char)
3749 t))))
d9e94c22
MS
3750
3751 (error
3752 (goto-char start)
3753 (signal (car err) (cdr err))))
3754
2a15eb73 3755 ;;(message "c-syntactic-re-search-forward done %s" (or (match-end 0) (point)))
d9e94c22
MS
3756
3757 (if found
3758 (progn
2a15eb73
MS
3759 (goto-char (match-end 0))
3760 (match-end 0))
d9e94c22
MS
3761
3762 ;; Search failed. Set point as appropriate.
2a15eb73
MS
3763 (if (eq noerror t)
3764 (goto-char start)
3765 (goto-char bound))
d9e94c22
MS
3766 nil)))
3767
47641aac
GM
3768(defvar safe-pos-list) ; bound in c-syntactic-skip-backward
3769
d0fcee66
AM
3770(defsubst c-ssb-lit-begin ()
3771 ;; Return the start of the literal point is in, or nil.
3772 ;; We read and write the variables `safe-pos', `safe-pos-list', `state'
3773 ;; bound in the caller.
3774
3775 ;; Use `parse-partial-sexp' from a safe position down to the point to check
3776 ;; if it's outside comments and strings.
3777 (save-excursion
3778 (let ((pos (point)) safe-pos state pps-end-pos)
3779 ;; Pick a safe position as close to the point as possible.
3780 ;;
3781 ;; FIXME: Consult `syntax-ppss' here if our cache doesn't give a good
3782 ;; position.
47641aac 3783
d0fcee66
AM
3784 (while (and safe-pos-list
3785 (> (car safe-pos-list) (point)))
3786 (setq safe-pos-list (cdr safe-pos-list)))
3787 (unless (setq safe-pos (car-safe safe-pos-list))
3788 (setq safe-pos (max (or (c-safe-position
3789 (point) (or c-state-cache
3790 (c-parse-state)))
3791 0)
3792 (point-min))
3793 safe-pos-list (list safe-pos)))
3794
3795 ;; Cache positions along the way to use if we have to back up more. We
3796 ;; cache every closing paren on the same level. If the paren cache is
3797 ;; relevant in this region then we're typically already on the same
3798 ;; level as the target position. Note that we might cache positions
3799 ;; after opening parens in case safe-pos is in a nested list. That's
3800 ;; both uncommon and harmless.
3801 (while (progn
3802 (setq state (parse-partial-sexp
3803 safe-pos pos 0))
3804 (< (point) pos))
3805 (setq safe-pos (point)
3806 safe-pos-list (cons safe-pos safe-pos-list)))
3807
3808 ;; If the state contains the start of the containing sexp we cache that
3809 ;; position too, so that parse-partial-sexp in the next run has a bigger
3810 ;; chance of starting at the same level as the target position and thus
3811 ;; will get more good safe positions into the list.
3812 (if (elt state 1)
3813 (setq safe-pos (1+ (elt state 1))
3814 safe-pos-list (cons safe-pos safe-pos-list)))
3815
3816 (if (or (elt state 3) (elt state 4))
3817 ;; Inside string or comment. Continue search at the
3818 ;; beginning of it.
3819 (elt state 8)))))
3820
0386b551 3821(defun c-syntactic-skip-backward (skip-chars &optional limit paren-level)
d9e94c22
MS
3822 "Like `skip-chars-backward' but only look at syntactically relevant chars,
3823i.e. don't stop at positions inside syntactic whitespace or string
3824literals. Preprocessor directives are also ignored, with the exception
3825of the one that the point starts within, if any. If LIMIT is given,
0386b551
AM
3826it's assumed to be at a syntactically relevant position.
3827
3828If PAREN-LEVEL is non-nil, the function won't stop in nested paren
3829sexps, and the search will also not go outside the current paren sexp.
3830However, if LIMIT or the buffer limit is reached inside a nested paren
3831then the point will be left at the limit.
3832
3833Non-nil is returned if the point moved, nil otherwise.
3834
3835Note that this function might do hidden buffer changes. See the
3836comment at the start of cc-engine.el for more info."
d9e94c22
MS
3837
3838 (let ((start (point))
d0fcee66 3839 state-2
d9e94c22
MS
3840 ;; A list of syntactically relevant positions in descending
3841 ;; order. It's used to avoid scanning repeatedly over
3842 ;; potentially large regions with `parse-partial-sexp' to verify
d0fcee66 3843 ;; each position. Used in `c-ssb-lit-begin'
d9e94c22
MS
3844 safe-pos-list
3845 ;; The result from `c-beginning-of-macro' at the start position or the
3846 ;; start position itself if it isn't within a macro. Evaluated on
3847 ;; demand.
0386b551
AM
3848 start-macro-beg
3849 ;; The earliest position after the current one with the same paren
3850 ;; level. Used only when `paren-level' is set.
d0fcee66 3851 lit-beg
0386b551 3852 (paren-level-pos (point)))
d9e94c22 3853
d0fcee66
AM
3854 (while
3855 (progn
3856 ;; The next loop "tries" to find the end point each time round,
3857 ;; loops when it hasn't succeeded.
3858 (while
3859 (and
3860 (< (skip-chars-backward skip-chars limit) 0)
3861
3862 (let ((pos (point)) state-2 pps-end-pos)
3863
3864 (cond
3865 ;; Don't stop inside a literal
3866 ((setq lit-beg (c-ssb-lit-begin))
3867 (goto-char lit-beg)
3868 t)
3869
3870 ((and paren-level
3871 (save-excursion
3872 (setq state-2 (parse-partial-sexp
3873 pos paren-level-pos -1)
3874 pps-end-pos (point))
3875 (/= (car state-2) 0)))
3876 ;; Not at the right level.
3877
3878 (if (and (< (car state-2) 0)
3879 ;; We stop above if we go out of a paren.
3880 ;; Now check whether it precedes or is
3881 ;; nested in the starting sexp.
3882 (save-excursion
3883 (setq state-2
3884 (parse-partial-sexp
3885 pps-end-pos paren-level-pos
3886 nil nil state-2))
3887 (< (car state-2) 0)))
3888
3889 ;; We've stopped short of the starting position
3890 ;; so the hit was inside a nested list. Go up
3891 ;; until we are at the right level.
3892 (condition-case nil
3893 (progn
3894 (goto-char (scan-lists pos -1
3895 (- (car state-2))))
3896 (setq paren-level-pos (point))
3897 (if (and limit (>= limit paren-level-pos))
0386b551 3898 (progn
d0fcee66
AM
3899 (goto-char limit)
3900 nil)
3901 t))
3902 (error
3903 (goto-char (or limit (point-min)))
3904 nil))
3905
3906 ;; The hit was outside the list at the start
3907 ;; position. Go to the start of the list and exit.
3908 (goto-char (1+ (elt state-2 1)))
3909 nil))
3910
3911 ((c-beginning-of-macro limit)
3912 ;; Inside a macro.
3913 (if (< (point)
3914 (or start-macro-beg
3915 (setq start-macro-beg
3916 (save-excursion
3917 (goto-char start)
3918 (c-beginning-of-macro limit)
3919 (point)))))
3920 t
3921
3922 ;; It's inside the same macro we started in so it's
3923 ;; a relevant match.
3924 (goto-char pos)
3925 nil))))))
3926
3927 (> (point)
3928 (progn
3929 ;; Skip syntactic ws afterwards so that we don't stop at the
3930 ;; end of a comment if `skip-chars' is something like "^/".
3931 (c-backward-syntactic-ws)
3932 (point)))))
d9e94c22 3933
0386b551
AM
3934 ;; We might want to extend this with more useful return values in
3935 ;; the future.
3936 (/= (point) start)))
3937
3938;; The following is an alternative implementation of
3939;; `c-syntactic-skip-backward' that uses backward movement to keep
3940;; track of the syntactic context. It turned out to be generally
3941;; slower than the one above which uses forward checks from earlier
3942;; safe positions.
3943;;
3944;;(defconst c-ssb-stop-re
3945;; ;; The regexp matching chars `c-syntactic-skip-backward' needs to
3946;; ;; stop at to avoid going into comments and literals.
3947;; (concat
3948;; ;; Match comment end syntax and string literal syntax. Also match
3949;; ;; '/' for block comment endings (not covered by comment end
3950;; ;; syntax).
3951;; "\\s>\\|/\\|\\s\""
3952;; (if (memq 'gen-string-delim c-emacs-features)
3953;; "\\|\\s|"
3954;; "")
3955;; (if (memq 'gen-comment-delim c-emacs-features)
3956;; "\\|\\s!"
3957;; "")))
3958;;
3959;;(defconst c-ssb-stop-paren-re
3960;; ;; Like `c-ssb-stop-re' but also stops at paren chars.
3961;; (concat c-ssb-stop-re "\\|\\s(\\|\\s)"))
3962;;
3963;;(defconst c-ssb-sexp-end-re
3964;; ;; Regexp matching the ending syntax of a complex sexp.
3965;; (concat c-string-limit-regexp "\\|\\s)"))
3966;;
3967;;(defun c-syntactic-skip-backward (skip-chars &optional limit paren-level)
3968;; "Like `skip-chars-backward' but only look at syntactically relevant chars,
3969;;i.e. don't stop at positions inside syntactic whitespace or string
3970;;literals. Preprocessor directives are also ignored. However, if the
3971;;point is within a comment, string literal or preprocessor directory to
3972;;begin with, its contents is treated as syntactically relevant chars.
3973;;If LIMIT is given, it limits the backward search and the point will be
3974;;left there if no earlier position is found.
3975;;
3976;;If PAREN-LEVEL is non-nil, the function won't stop in nested paren
3977;;sexps, and the search will also not go outside the current paren sexp.
3978;;However, if LIMIT or the buffer limit is reached inside a nested paren
3979;;then the point will be left at the limit.
3980;;
3981;;Non-nil is returned if the point moved, nil otherwise.
3982;;
3983;;Note that this function might do hidden buffer changes. See the
3984;;comment at the start of cc-engine.el for more info."
3985;;
3986;; (save-restriction
3987;; (when limit
3988;; (narrow-to-region limit (point-max)))
3989;;
3990;; (let ((start (point)))
3991;; (catch 'done
3992;; (while (let ((last-pos (point))
3993;; (stop-pos (progn
3994;; (skip-chars-backward skip-chars)
3995;; (point))))
3996;;
3997;; ;; Skip back over the same region as
3998;; ;; `skip-chars-backward' above, but keep to
3999;; ;; syntactically relevant positions.
4000;; (goto-char last-pos)
4001;; (while (and
4002;; ;; `re-search-backward' with a single char regexp
4003;; ;; should be fast.
4004;; (re-search-backward
4005;; (if paren-level c-ssb-stop-paren-re c-ssb-stop-re)
4006;; stop-pos 'move)
4007;;
4008;; (progn
4009;; (cond
4010;; ((looking-at "\\s(")
4011;; ;; `paren-level' is set and we've found the
4012;; ;; start of the containing paren.
4013;; (forward-char)
4014;; (throw 'done t))
4015;;
4016;; ((looking-at c-ssb-sexp-end-re)
4017;; ;; We're at the end of a string literal or paren
4018;; ;; sexp (if `paren-level' is set).
4019;; (forward-char)
4020;; (condition-case nil
4021;; (c-backward-sexp)
4022;; (error
4023;; (goto-char limit)
4024;; (throw 'done t))))
4025;;
4026;; (t
4027;; (forward-char)
4028;; ;; At the end of some syntactic ws or possibly
4029;; ;; after a plain '/' operator.
4030;; (let ((pos (point)))
4031;; (c-backward-syntactic-ws)
4032;; (if (= pos (point))
4033;; ;; Was a plain '/' operator. Go past it.
4034;; (backward-char)))))
4035;;
4036;; (> (point) stop-pos))))
4037;;
4038;; ;; Now the point is either at `stop-pos' or at some
4039;; ;; position further back if `stop-pos' was at a
4040;; ;; syntactically irrelevant place.
4041;;
4042;; ;; Skip additional syntactic ws so that we don't stop
4043;; ;; at the end of a comment if `skip-chars' is
4044;; ;; something like "^/".
4045;; (c-backward-syntactic-ws)
4046;;
4047;; (< (point) stop-pos))))
4048;;
4049;; ;; We might want to extend this with more useful return values
4050;; ;; in the future.
4051;; (/= (point) start))))
d9e94c22
MS
4052
4053\f
4054;; Tools for handling comments and string literals.
4055
4056(defun c-slow-in-literal (&optional lim detect-cpp)
4057 "Return the type of literal point is in, if any.
4058The return value is `c' if in a C-style comment, `c++' if in a C++
4059style comment, `string' if in a string literal, `pound' if DETECT-CPP
4060is non-nil and in a preprocessor line, or nil if somewhere else.
4061Optional LIM is used as the backward limit of the search. If omitted,
4062or nil, `c-beginning-of-defun' is used.
4063
4064The last point calculated is cached if the cache is enabled, i.e. if
4065`c-in-literal-cache' is bound to a two element vector.
4066
0386b551
AM
4067Note that this function might do hidden buffer changes. See the
4068comment at the start of cc-engine.el for more info."
4069
d9e94c22
MS
4070 (if (and (vectorp c-in-literal-cache)
4071 (= (point) (aref c-in-literal-cache 0)))
4072 (aref c-in-literal-cache 1)
4073 (let ((rtn (save-excursion
4074 (let* ((pos (point))
4075 (lim (or lim (progn
4076 (c-beginning-of-syntax)
4077 (point))))
4078 (state (parse-partial-sexp lim pos)))
4079 (cond
4080 ((elt state 3) 'string)
4081 ((elt state 4) (if (elt state 7) 'c++ 'c))
4082 ((and detect-cpp (c-beginning-of-macro lim)) 'pound)
4083 (t nil))))))
4084 ;; cache this result if the cache is enabled
4085 (if (not c-in-literal-cache)
4086 (setq c-in-literal-cache (vector (point) rtn)))
4087 rtn)))
4088
4089;; XEmacs has a built-in function that should make this much quicker.
4090;; I don't think we even need the cache, which makes our lives more
4091;; complicated anyway. In this case, lim is only used to detect
4092;; cpp directives.
4093;;
4094;; Note that there is a bug in Xemacs's buffer-syntactic-context when used in
4095;; conjunction with syntax-table-properties. The bug is present in, e.g.,
4096;; Xemacs 21.4.4. It manifested itself thus:
4097;;
4098;; Starting with an empty AWK Mode buffer, type
4099;; /regexp/ {<C-j>
4100;; Point gets wrongly left at column 0, rather than being indented to tab-width.
4101;;
4102;; AWK Mode is designed such that when the first / is typed, it gets the
4103;; syntax-table property "string fence". When the second / is typed, BOTH /s
4104;; are given the s-t property "string". However, buffer-syntactic-context
4105;; fails to take account of the change of the s-t property on the opening / to
4106;; "string", and reports that the { is within a string started by the second /.
4107;;
4108;; The workaround for this is for the AWK Mode initialisation to switch the
4109;; defalias for c-in-literal to c-slow-in-literal. This will slow down other
4110;; cc-modes in Xemacs whenever an awk-buffer has been initialised.
b414f371 4111;;
d9e94c22
MS
4112;; (Alan Mackenzie, 2003/4/30).
4113
4114(defun c-fast-in-literal (&optional lim detect-cpp)
0386b551 4115 ;; This function might do hidden buffer changes.
d9e94c22
MS
4116 (let ((context (buffer-syntactic-context)))
4117 (cond
4118 ((eq context 'string) 'string)
4119 ((eq context 'comment) 'c++)
4120 ((eq context 'block-comment) 'c)
4121 ((and detect-cpp (save-excursion (c-beginning-of-macro lim))) 'pound))))
4122
4123(defalias 'c-in-literal
4124 (if (fboundp 'buffer-syntactic-context)
7bfc3fdb 4125 'c-fast-in-literal ; XEmacs
d9e94c22
MS
4126 'c-slow-in-literal)) ; GNU Emacs
4127
4128;; The defalias above isn't enough to shut up the byte compiler.
4129(cc-bytecomp-defun c-in-literal)
4130
4131(defun c-literal-limits (&optional lim near not-in-delimiter)
4132 "Return a cons of the beginning and end positions of the comment or
4133string surrounding point (including both delimiters), or nil if point
4134isn't in one. If LIM is non-nil, it's used as the \"safe\" position
4135to start parsing from. If NEAR is non-nil, then the limits of any
4136literal next to point is returned. \"Next to\" means there's only
4137spaces and tabs between point and the literal. The search for such a
4138literal is done first in forward direction. If NOT-IN-DELIMITER is
4139non-nil, the case when point is inside a starting delimiter won't be
a85fd6da 4140recognized. This only has effect for comments which have starting
d9e94c22
MS
4141delimiters with more than one character.
4142
0386b551
AM
4143Note that this function might do hidden buffer changes. See the
4144comment at the start of cc-engine.el for more info."
d9e94c22
MS
4145
4146 (save-excursion
4147 (let* ((pos (point))
4148 (lim (or lim (progn
4149 (c-beginning-of-syntax)
4150 (point))))
4151 (state (parse-partial-sexp lim pos)))
4152
0386b551
AM
4153 (cond ((elt state 3) ; String.
4154 (goto-char (elt state 8))
d9e94c22
MS
4155 (cons (point) (or (c-safe (c-forward-sexp 1) (point))
4156 (point-max))))
4157
0386b551
AM
4158 ((elt state 4) ; Comment.
4159 (goto-char (elt state 8))
d9e94c22
MS
4160 (cons (point) (progn (c-forward-single-comment) (point))))
4161
4162 ((and (not not-in-delimiter)
4163 (not (elt state 5))
4164 (eq (char-before) ?/)
4165 (looking-at "[/*]"))
4166 ;; We're standing in a comment starter.
4167 (backward-char 1)
4168 (cons (point) (progn (c-forward-single-comment) (point))))
4169
4170 (near
4171 (goto-char pos)
4172
4173 ;; Search forward for a literal.
4174 (skip-chars-forward " \t")
4175
4176 (cond
4177 ((looking-at c-string-limit-regexp) ; String.
4178 (cons (point) (or (c-safe (c-forward-sexp 1) (point))
4179 (point-max))))
4180
4181 ((looking-at c-comment-start-regexp) ; Line or block comment.
4182 (cons (point) (progn (c-forward-single-comment) (point))))
4183
4184 (t
4185 ;; Search backward.
4186 (skip-chars-backward " \t")
4187
4188 (let ((end (point)) beg)
4189 (cond
4190 ((save-excursion
4191 (< (skip-syntax-backward c-string-syntax) 0)) ; String.
4192 (setq beg (c-safe (c-backward-sexp 1) (point))))
4193
4194 ((and (c-safe (forward-char -2) t)
4195 (looking-at "*/"))
4196 ;; Block comment. Due to the nature of line
4197 ;; comments, they will always be covered by the
4198 ;; normal case above.
4199 (goto-char end)
4200 (c-backward-single-comment)
4201 ;; If LIM is bogus, beg will be bogus.
4202 (setq beg (point))))
4203
4204 (if beg (cons beg end))))))
4205 ))))
4206
0386b551
AM
4207;; In case external callers use this; it did have a docstring.
4208(defalias 'c-literal-limits-fast 'c-literal-limits)
d9e94c22 4209
0386b551
AM
4210(defun c-collect-line-comments (range)
4211 "If the argument is a cons of two buffer positions (such as returned by
4212`c-literal-limits'), and that range contains a C++ style line comment,
4213then an extended range is returned that contains all adjacent line
4214comments (i.e. all comments that starts in the same column with no
4215empty lines or non-whitespace characters between them). Otherwise the
4216argument is returned.
d9e94c22 4217
0386b551
AM
4218Note that this function might do hidden buffer changes. See the
4219comment at the start of cc-engine.el for more info."
d9e94c22
MS
4220
4221 (save-excursion
0386b551
AM
4222 (condition-case nil
4223 (if (and (consp range) (progn
4224 (goto-char (car range))
4225 (looking-at c-line-comment-starter)))
b414f371 4226 (let ((col (current-column))
0386b551
AM
4227 (beg (point))
4228 (bopl (c-point 'bopl))
4229 (end (cdr range)))
4230 ;; Got to take care in the backward direction to handle
4231 ;; comments which are preceded by code.
4232 (while (and (c-backward-single-comment)
4233 (>= (point) bopl)
4234 (looking-at c-line-comment-starter)
4235 (= col (current-column)))
4236 (setq beg (point)
4237 bopl (c-point 'bopl)))
4238 (goto-char end)
4239 (while (and (progn (skip-chars-forward " \t")
4240 (looking-at c-line-comment-starter))
4241 (= col (current-column))
4242 (prog1 (zerop (forward-line 1))
4243 (setq end (point)))))
4244 (cons beg end))
4245 range)
4246 (error range))))
d9e94c22
MS
4247
4248(defun c-literal-type (range)
4249 "Convenience function that given the result of `c-literal-limits',
a85fd6da
AM
4250returns nil or the type of literal that the range surrounds, one
4251of the symbols 'c, 'c++ or 'string. It's much faster than using
4252`c-in-literal' and is intended to be used when you need both the
4253type of a literal and its limits.
d9e94c22 4254
0386b551
AM
4255Note that this function might do hidden buffer changes. See the
4256comment at the start of cc-engine.el for more info."
4257
d9e94c22
MS
4258 (if (consp range)
4259 (save-excursion
4260 (goto-char (car range))
4261 (cond ((looking-at c-string-limit-regexp) 'string)
4262 ((or (looking-at "//") ; c++ line comment
4263 (and (looking-at "\\s<") ; comment starter
4264 (looking-at "#"))) ; awk comment.
4265 'c++)
4266 (t 'c))) ; Assuming the range is valid.
4267 range))
4268
4269\f
4270;; `c-find-decl-spots' and accompanying stuff.
4271
4272;; Variables used in `c-find-decl-spots' to cache the search done for
4273;; the first declaration in the last call. When that function starts,
4274;; it needs to back up over syntactic whitespace to look at the last
4275;; token before the region being searched. That can sometimes cause
4276;; moves back and forth over a quite large region of comments and
4277;; macros, which would be repeated for each changed character when
4278;; we're called during fontification, since font-lock refontifies the
4279;; current line for each change. Thus it's worthwhile to cache the
4280;; first match.
4281;;
4282;; `c-find-decl-syntactic-pos' is a syntactically relevant position in
4283;; the syntactic whitespace less or equal to some start position.
4284;; There's no cached value if it's nil.
4285;;
4286;; `c-find-decl-match-pos' is the match position if
4287;; `c-find-decl-prefix-search' matched before the syntactic whitespace
4288;; at `c-find-decl-syntactic-pos', or nil if there's no such match.
4289(defvar c-find-decl-syntactic-pos nil)
4290(make-variable-buffer-local 'c-find-decl-syntactic-pos)
4291(defvar c-find-decl-match-pos nil)
4292(make-variable-buffer-local 'c-find-decl-match-pos)
4293
4294(defsubst c-invalidate-find-decl-cache (change-min-pos)
4295 (and c-find-decl-syntactic-pos
4296 (< change-min-pos c-find-decl-syntactic-pos)
4297 (setq c-find-decl-syntactic-pos nil)))
4298
4299; (defface c-debug-decl-spot-face
4300; '((t (:background "Turquoise")))
4301; "Debug face to mark the spots where `c-find-decl-spots' stopped.")
4302; (defface c-debug-decl-sws-face
4303; '((t (:background "Khaki")))
4304; "Debug face to mark the syntactic whitespace between the declaration
4305; spots and the preceding token end.")
4306
4307(defmacro c-debug-put-decl-spot-faces (match-pos decl-pos)
4308 (when (facep 'c-debug-decl-spot-face)
0386b551 4309 `(c-save-buffer-state ((match-pos ,match-pos) (decl-pos ,decl-pos))
d9e94c22
MS
4310 (c-debug-add-face (max match-pos (point-min)) decl-pos
4311 'c-debug-decl-sws-face)
4312 (c-debug-add-face decl-pos (min (1+ decl-pos) (point-max))
4313 'c-debug-decl-spot-face))))
4314(defmacro c-debug-remove-decl-spot-faces (beg end)
4315 (when (facep 'c-debug-decl-spot-face)
0386b551 4316 `(c-save-buffer-state ()
d9e94c22
MS
4317 (c-debug-remove-face ,beg ,end 'c-debug-decl-spot-face)
4318 (c-debug-remove-face ,beg ,end 'c-debug-decl-sws-face))))
4319
4320(defmacro c-find-decl-prefix-search ()
4321 ;; Macro used inside `c-find-decl-spots'. It ought to be a defun,
4322 ;; but it contains lots of free variables that refer to things
4323 ;; inside `c-find-decl-spots'. The point is left at `cfd-match-pos'
4324 ;; if there is a match, otherwise at `cfd-limit'.
0386b551
AM
4325 ;;
4326 ;; This macro might do hidden buffer changes.
d9e94c22
MS
4327
4328 '(progn
4329 ;; Find the next property match position if we haven't got one already.
4330 (unless cfd-prop-match
4331 (save-excursion
4332 (while (progn
4333 (goto-char (next-single-property-change
4334 (point) 'c-type nil cfd-limit))
4335 (and (< (point) cfd-limit)
4336 (not (eq (c-get-char-property (1- (point)) 'c-type)
4337 'c-decl-end)))))
4338 (setq cfd-prop-match (point))))
4339
0386b551
AM
4340 ;; Find the next `c-decl-prefix-or-start-re' match if we haven't
4341 ;; got one already.
d9e94c22 4342 (unless cfd-re-match
0386b551
AM
4343
4344 (if (> cfd-re-match-end (point))
4345 (goto-char cfd-re-match-end))
4346
4347 (while (if (setq cfd-re-match-end
4348 (re-search-forward c-decl-prefix-or-start-re
4349 cfd-limit 'move))
4350
4351 ;; Match. Check if it's inside a comment or string literal.
4352 (c-got-face-at
4353 (if (setq cfd-re-match (match-end 1))
4354 ;; Matched the end of a token preceding a decl spot.
4355 (progn
4356 (goto-char cfd-re-match)
4357 (1- cfd-re-match))
4358 ;; Matched a token that start a decl spot.
4359 (goto-char (match-beginning 0))
4360 (point))
4361 c-literal-faces)
4362
4363 ;; No match. Finish up and exit the loop.
4364 (setq cfd-re-match cfd-limit)
4365 nil)
4366
4367 ;; Skip out of comments and string literals.
d9e94c22
MS
4368 (while (progn
4369 (goto-char (next-single-property-change
0386b551 4370 (point) 'face nil cfd-limit))
d9e94c22 4371 (and (< (point) cfd-limit)
0386b551
AM
4372 (c-got-face-at (point) c-literal-faces)))))
4373
4374 ;; If we matched at the decl start, we have to back up over the
4375 ;; preceding syntactic ws to set `cfd-match-pos' and to catch
4376 ;; any decl spots in the syntactic ws.
d9e94c22 4377 (unless cfd-re-match
0386b551
AM
4378 (c-backward-syntactic-ws)
4379 (setq cfd-re-match (point))))
d9e94c22
MS
4380
4381 ;; Choose whichever match is closer to the start.
4382 (if (< cfd-re-match cfd-prop-match)
4383 (setq cfd-match-pos cfd-re-match
4384 cfd-re-match nil)
4385 (setq cfd-match-pos cfd-prop-match
4386 cfd-prop-match nil))
4387
4388 (goto-char cfd-match-pos)
4389
4390 (when (< cfd-match-pos cfd-limit)
4391 ;; Skip forward past comments only so we don't skip macros.
4392 (c-forward-comments)
4393 ;; Set the position to continue at. We can avoid going over
4394 ;; the comments skipped above a second time, but it's possible
4395 ;; that the comment skipping has taken us past `cfd-prop-match'
4396 ;; since the property might be used inside comments.
4397 (setq cfd-continue-pos (if cfd-prop-match
4398 (min cfd-prop-match (point))
4399 (point))))))
4400
4401(defun c-find-decl-spots (cfd-limit cfd-decl-re cfd-face-checklist cfd-fun)
0386b551 4402 ;; Call CFD-FUN for each possible spot for a declaration, cast or
a85fd6da
AM
4403 ;; label from the point to CFD-LIMIT.
4404 ;;
4405 ;; CFD-FUN is called with point at the start of the spot. It's
4406 ;; passed two arguments: The first is the end position of the token
4407 ;; preceding the spot, or 0 for the implicit match at bob. The
4408 ;; second is a flag that is t when the match is inside a macro. If
4409 ;; CFD-FUN adds `c-decl-end' properties somewhere below the current
4410 ;; spot, it should return non-nil to ensure that the next search
4411 ;; will find them.
0386b551 4412 ;;
a85fd6da 4413 ;; Such a spot is:
0386b551
AM
4414 ;; o The first token after bob.
4415 ;; o The first token after the end of submatch 1 in
4416 ;; `c-decl-prefix-or-start-re' when that submatch matches.
4417 ;; o The start of each `c-decl-prefix-or-start-re' match when
4418 ;; submatch 1 doesn't match.
5a89f0a7 4419 ;; o The first token after the end of each occurrence of the
0386b551
AM
4420 ;; `c-type' text property with the value `c-decl-end', provided
4421 ;; `c-type-decl-end-used' is set.
4422 ;;
4423 ;; Only a spot that match CFD-DECL-RE and whose face is in the
4424 ;; CFD-FACE-CHECKLIST list causes CFD-FUN to be called. The face
4425 ;; check is disabled if CFD-FACE-CHECKLIST is nil.
d9e94c22
MS
4426 ;;
4427 ;; If the match is inside a macro then the buffer is narrowed to the
4428 ;; end of it, so that CFD-FUN can investigate the following tokens
4429 ;; without matching something that begins inside a macro and ends
4430 ;; outside it. It's to avoid this work that the CFD-DECL-RE and
4431 ;; CFD-FACE-CHECKLIST checks exist.
4432 ;;
0386b551
AM
4433 ;; The spots are visited approximately in order from top to bottom.
4434 ;; It's however the positions where `c-decl-prefix-or-start-re'
4435 ;; matches and where `c-decl-end' properties are found that are in
4436 ;; order. Since the spots often are at the following token, they
4437 ;; might be visited out of order insofar as more spots are reported
4438 ;; later on within the syntactic whitespace between the match
4439 ;; positions and their spots.
4440 ;;
4441 ;; It's assumed that comments and strings are fontified in the
d9e94c22
MS
4442 ;; searched range.
4443 ;;
4444 ;; This is mainly used in fontification, and so has an elaborate
4445 ;; cache to handle repeated calls from the same start position; see
4446 ;; the variables above.
4447 ;;
4448 ;; All variables in this function begin with `cfd-' to avoid name
4449 ;; collision with the (dynamically bound) variables used in CFD-FUN.
0386b551
AM
4450 ;;
4451 ;; This function might do hidden buffer changes.
d9e94c22 4452
0386b551
AM
4453 (let ((cfd-start-pos (point))
4454 (cfd-buffer-end (point-max))
4455 ;; The end of the token preceding the decl spot last found
4456 ;; with `c-decl-prefix-or-start-re'. `cfd-limit' if there's
4457 ;; no match.
d9e94c22 4458 cfd-re-match
0386b551
AM
4459 ;; The end position of the last `c-decl-prefix-or-start-re'
4460 ;; match. If this is greater than `cfd-continue-pos', the
4461 ;; next regexp search is started here instead.
4462 (cfd-re-match-end (point-min))
4463 ;; The end of the last `c-decl-end' found by
4464 ;; `c-find-decl-prefix-search'. `cfd-limit' if there's no
4465 ;; match. If searching for the property isn't needed then we
4466 ;; disable it by setting it to `cfd-limit' directly.
d9e94c22 4467 (cfd-prop-match (unless c-type-decl-end-used cfd-limit))
0386b551
AM
4468 ;; The end of the token preceding the decl spot last found by
4469 ;; `c-find-decl-prefix-search'. 0 for the implicit match at
4470 ;; bob. `cfd-limit' if there's no match. In other words,
4471 ;; this is the minimum of `cfd-re-match' and `cfd-prop-match'.
d9e94c22
MS
4472 (cfd-match-pos cfd-limit)
4473 ;; The position to continue searching at.
4474 cfd-continue-pos
4475 ;; The position of the last "real" token we've stopped at.
4476 ;; This can be greater than `cfd-continue-pos' when we get
4477 ;; hits inside macros or at `c-decl-end' positions inside
4478 ;; comments.
4479 (cfd-token-pos 0)
4480 ;; The end position of the last entered macro.
4481 (cfd-macro-end 0))
4482
4483 ;; Initialize by finding a syntactically relevant start position
0386b551
AM
4484 ;; before the point, and do the first `c-decl-prefix-or-start-re'
4485 ;; search unless we're at bob.
d9e94c22 4486
0386b551 4487 (let (start-in-literal start-in-macro syntactic-pos)
d9e94c22
MS
4488 ;; Must back up a bit since we look for the end of the previous
4489 ;; statement or declaration, which is earlier than the first
4490 ;; returned match.
4491
0386b551
AM
4492 (cond
4493 ;; First we need to move to a syntactically relevant position.
4494 ;; Begin by backing out of comment or string literals.
4495 ((and
4496 (when (c-got-face-at (point) c-literal-faces)
4497 ;; Try to use the faces to back up to the start of the
4498 ;; literal. FIXME: What if the point is on a declaration
4499 ;; inside a comment?
4500 (while (and (not (bobp))
4501 (c-got-face-at (1- (point)) c-literal-faces))
4502 (goto-char (previous-single-property-change
4503 (point) 'face nil (point-min))))
4504
4505 ;; XEmacs doesn't fontify the quotes surrounding string
4506 ;; literals.
4507 (and (featurep 'xemacs)
4508 (eq (get-text-property (point) 'face)
4509 'font-lock-string-face)
4510 (not (bobp))
4511 (progn (backward-char)
4512 (not (looking-at c-string-limit-regexp)))
4513 (forward-char))
4514
4515 ;; Don't trust the literal to contain only literal faces
4516 ;; (the font lock package might not have fontified the
4517 ;; start of it at all, for instance) so check that we have
4518 ;; arrived at something that looks like a start or else
4519 ;; resort to `c-literal-limits'.
4520 (unless (looking-at c-literal-start-regexp)
4521 (let ((range (c-literal-limits)))
4522 (if range (goto-char (car range)))))
4523
4524 (setq start-in-literal (point)))
4525
4526 ;; The start is in a literal. If the limit is in the same
4527 ;; one we don't have to find a syntactic position etc. We
4528 ;; only check that if the limit is at or before bonl to save
4529 ;; time; it covers the by far most common case when font-lock
4530 ;; refontifies the current line only.
4531 (<= cfd-limit (c-point 'bonl cfd-start-pos))
4532 (save-excursion
4533 (goto-char cfd-start-pos)
4534 (while (progn
4535 (goto-char (next-single-property-change
4536 (point) 'face nil cfd-limit))
4537 (and (< (point) cfd-limit)
4538 (c-got-face-at (point) c-literal-faces))))
4539 (= (point) cfd-limit)))
4540
4541 ;; Completely inside a literal. Set up variables to trig the
4542 ;; (< cfd-continue-pos cfd-start-pos) case below and it'll
4543 ;; find a suitable start position.
4544 (setq cfd-continue-pos start-in-literal))
4545
4546 ;; Check if the region might be completely inside a macro, to
4547 ;; optimize that like the completely-inside-literal above.
4548 ((save-excursion
4549 (and (= (forward-line 1) 0)
4550 (bolp) ; forward-line has funny behavior at eob.
4551 (>= (point) cfd-limit)
4552 (progn (backward-char)
4553 (eq (char-before) ?\\))))
4554 ;; (Maybe) completely inside a macro. Only need to trig the
4555 ;; (< cfd-continue-pos cfd-start-pos) case below to make it
4556 ;; set things up.
4557 (setq cfd-continue-pos (1- cfd-start-pos)
4558 start-in-macro t))
d9e94c22 4559
0386b551
AM
4560 (t
4561 ;; Back out of any macro so we don't miss any declaration
4562 ;; that could follow after it.
4563 (when (c-beginning-of-macro)
4564 (setq start-in-macro t))
4565
4566 ;; Now we're at a proper syntactically relevant position so we
4567 ;; can use the cache. But first clear it if it applied
4568 ;; further down.
4569 (c-invalidate-find-decl-cache cfd-start-pos)
4570
4571 (setq syntactic-pos (point))
4572 (unless (eq syntactic-pos c-find-decl-syntactic-pos)
4573 ;; Don't have to do this if the cache is relevant here,
4574 ;; typically if the same line is refontified again. If
4575 ;; we're just some syntactic whitespace further down we can
4576 ;; still use the cache to limit the skipping.
4577 (c-backward-syntactic-ws c-find-decl-syntactic-pos))
4578
4579 ;; If we hit `c-find-decl-syntactic-pos' and
4580 ;; `c-find-decl-match-pos' is set then we install the cached
4581 ;; values. If we hit `c-find-decl-syntactic-pos' and
4582 ;; `c-find-decl-match-pos' is nil then we know there's no decl
4583 ;; prefix in the whitespace before `c-find-decl-syntactic-pos'
4584 ;; and so we can continue the search from this point. If we
4585 ;; didn't hit `c-find-decl-syntactic-pos' then we're now in
4586 ;; the right spot to begin searching anyway.
4587 (if (and (eq (point) c-find-decl-syntactic-pos)
4588 c-find-decl-match-pos)
d9e94c22
MS
4589 (setq cfd-match-pos c-find-decl-match-pos
4590 cfd-continue-pos syntactic-pos)
0386b551
AM
4591
4592 (setq c-find-decl-syntactic-pos syntactic-pos)
4593
4594 (when (if (bobp)
4595 ;; Always consider bob a match to get the first
4596 ;; declaration in the file. Do this separately instead of
4597 ;; letting `c-decl-prefix-or-start-re' match bob, so that
4598 ;; regexp always can consume at least one character to
4599 ;; ensure that we won't get stuck in an infinite loop.
4600 (setq cfd-re-match 0)
4601 (backward-char)
4602 (c-beginning-of-current-token)
4603 (< (point) cfd-limit))
4604 ;; Do an initial search now. In the bob case above it's
4605 ;; only done to search for a `c-decl-end' spot.
4606 (c-find-decl-prefix-search))
4607
4608 (setq c-find-decl-match-pos (and (< cfd-match-pos cfd-start-pos)
4609 cfd-match-pos)))))
4610
4611 ;; Advance `cfd-continue-pos' if it's before the start position.
4612 ;; The closest continue position that might have effect at or
4613 ;; after the start depends on what we started in. This also
4614 ;; finds a suitable start position in the special cases when the
4615 ;; region is completely within a literal or macro.
4616 (when (and cfd-continue-pos (< cfd-continue-pos cfd-start-pos))
4617
4618 (cond
4619 (start-in-macro
4620 ;; If we're in a macro then it's the closest preceding token
4621 ;; in the macro. Check this before `start-in-literal',
4622 ;; since if we're inside a literal in a macro, the preceding
4623 ;; token is earlier than any `c-decl-end' spot inside the
4624 ;; literal (comment).
4625 (goto-char (or start-in-literal cfd-start-pos))
4626 ;; The only syntactic ws in macros are comments.
d9e94c22 4627 (c-backward-comments)
0386b551
AM
4628 (backward-char)
4629 (c-beginning-of-current-token))
4630
4631 (start-in-literal
4632 ;; If we're in a comment it can only be the closest
4633 ;; preceding `c-decl-end' position within that comment, if
4634 ;; any. Go back to the beginning of such a property so that
4635 ;; `c-find-decl-prefix-search' will find the end of it.
4636 ;; (Can't stop at the end and install it directly on
4637 ;; `cfd-prop-match' since that variable might be cleared
4638 ;; after `cfd-fun' below.)
4639 ;;
4640 ;; Note that if the literal is a string then the property
4641 ;; search will simply skip to the beginning of it right
4642 ;; away.
4643 (if (not c-type-decl-end-used)
4644 (goto-char start-in-literal)
4645 (goto-char cfd-start-pos)
4646 (while (progn
4647 (goto-char (previous-single-property-change
4648 (point) 'c-type nil start-in-literal))
4649 (and (> (point) start-in-literal)
4650 (not (eq (c-get-char-property (point) 'c-type)
4651 'c-decl-end))))))
4652
4653 (when (= (point) start-in-literal)
4654 ;; Didn't find any property inside the comment, so we can
4655 ;; skip it entirely. (This won't skip past a string, but
4656 ;; that'll be handled quickly by the next
4657 ;; `c-find-decl-prefix-search' anyway.)
4658 (c-forward-single-comment)
4659 (if (> (point) cfd-limit)
4660 (goto-char cfd-limit))))
d9e94c22 4661
0386b551
AM
4662 (t
4663 ;; If we started in normal code, the only match that might
4664 ;; apply before the start is what we already got in
4665 ;; `cfd-match-pos' so we can continue at the start position.
4666 ;; (Note that we don't get here if the first match is below
4667 ;; it.)
4668 (goto-char cfd-start-pos)))
4669
4670 ;; Delete found matches if they are before our new continue
4671 ;; position, so that `c-find-decl-prefix-search' won't back up
4672 ;; to them later on.
4673 (setq cfd-continue-pos (point))
4674 (when (and cfd-re-match (< cfd-re-match cfd-continue-pos))
4675 (setq cfd-re-match nil))
4676 (when (and cfd-prop-match (< cfd-prop-match cfd-continue-pos))
4677 (setq cfd-prop-match nil)))
4678
4679 (if syntactic-pos
4680 ;; This is the normal case and we got a proper syntactic
4681 ;; position. If there's a match then it's always outside
4682 ;; macros and comments, so advance to the next token and set
4683 ;; `cfd-token-pos'. The loop below will later go back using
4684 ;; `cfd-continue-pos' to fix declarations inside the
4685 ;; syntactic ws.
4686 (when (and cfd-match-pos (< cfd-match-pos syntactic-pos))
4687 (goto-char syntactic-pos)
4688 (c-forward-syntactic-ws)
4689 (and cfd-continue-pos
4690 (< cfd-continue-pos (point))
4691 (setq cfd-token-pos (point))))
4692
4693 ;; Have one of the special cases when the region is completely
4694 ;; within a literal or macro. `cfd-continue-pos' is set to a
4695 ;; good start position for the search, so do it.
4696 (c-find-decl-prefix-search)))
d9e94c22 4697
51c9af45 4698 ;; Now loop. Round what? (ACM, 2006/7/5). We already got the first match.
d9e94c22
MS
4699
4700 (while (progn
4701 (while (and
4702 (< cfd-match-pos cfd-limit)
4703
4704 (or
4705 ;; Kludge to filter out matches on the "<" that
4706 ;; aren't open parens, for the sake of languages
4707 ;; that got `c-recognize-<>-arglists' set.
4708 (and (eq (char-before cfd-match-pos) ?<)
4709 (not (c-get-char-property (1- cfd-match-pos)
4710 'syntax-table)))
4711
4712 ;; If `cfd-continue-pos' is less or equal to
4713 ;; `cfd-token-pos', we've got a hit inside a macro
4714 ;; that's in the syntactic whitespace before the last
4715 ;; "real" declaration we've checked. If they're equal
4716 ;; we've arrived at the declaration a second time, so
4717 ;; there's nothing to do.
4718 (= cfd-continue-pos cfd-token-pos)
4719
4720 (progn
4721 ;; If `cfd-continue-pos' is less than `cfd-token-pos'
4722 ;; we're still searching for declarations embedded in
4723 ;; the syntactic whitespace. In that case we need
4724 ;; only to skip comments and not macros, since they
4725 ;; can't be nested, and that's already been done in
4726 ;; `c-find-decl-prefix-search'.
4727 (when (> cfd-continue-pos cfd-token-pos)
4728 (c-forward-syntactic-ws)
4729 (setq cfd-token-pos (point)))
4730
4731 ;; Continue if the following token fails the
4732 ;; CFD-DECL-RE and CFD-FACE-CHECKLIST checks.
4733 (when (or (>= (point) cfd-limit)
4734 (not (looking-at cfd-decl-re))
4735 (and cfd-face-checklist
4736 (not (c-got-face-at
4737 (point) cfd-face-checklist))))
4738 (goto-char cfd-continue-pos)
4739 t)))
4740
4741 (< (point) cfd-limit))
4742 (c-find-decl-prefix-search))
4743
4744 (< (point) cfd-limit))
4745
0386b551
AM
4746 (when (and
4747 (>= (point) cfd-start-pos)
d9e94c22 4748
0386b551
AM
4749 (progn
4750 ;; Narrow to the end of the macro if we got a hit inside
4751 ;; one, to avoid recognizing things that start inside the
4752 ;; macro and end outside it.
4753 (when (> cfd-match-pos cfd-macro-end)
4754 ;; Not in the same macro as in the previous round.
4755 (save-excursion
4756 (goto-char cfd-match-pos)
4757 (setq cfd-macro-end
4758 (if (save-excursion (and (c-beginning-of-macro)
4759 (< (point) cfd-match-pos)))
4760 (progn (c-end-of-macro)
4761 (point))
4762 0))))
4763
4764 (if (zerop cfd-macro-end)
4765 t
4766 (if (> cfd-macro-end (point))
4767 (progn (narrow-to-region (point-min) cfd-macro-end)
4768 t)
4769 ;; The matched token was the last thing in the macro,
4770 ;; so the whole match is bogus.
4771 (setq cfd-macro-end 0)
4772 nil))))
d9e94c22
MS
4773
4774 (c-debug-put-decl-spot-faces cfd-match-pos (point))
0386b551
AM
4775 (if (funcall cfd-fun cfd-match-pos (/= cfd-macro-end 0))
4776 (setq cfd-prop-match nil))
d9e94c22
MS
4777
4778 (when (/= cfd-macro-end 0)
4779 ;; Restore limits if we did macro narrowment above.
4780 (narrow-to-region (point-min) cfd-buffer-end)))
4781
4782 (goto-char cfd-continue-pos)
4783 (if (= cfd-continue-pos cfd-limit)
4784 (setq cfd-match-pos cfd-limit)
4785 (c-find-decl-prefix-search)))))
4786
4787\f
4788;; A cache for found types.
4789
4790;; Buffer local variable that contains an obarray with the types we've
4791;; found. If a declaration is recognized somewhere we record the
4792;; fully qualified identifier in it to recognize it as a type
4793;; elsewhere in the file too. This is not accurate since we do not
4794;; bother with the scoping rules of the languages, but in practice the
4795;; same name is seldom used as both a type and something else in a
4796;; file, and we only use this as a last resort in ambiguous cases (see
0386b551
AM
4797;; `c-forward-decl-or-cast-1').
4798;;
580fba94
AM
4799;; Not every type need be in this cache. However, things which have
4800;; ceased to be types must be removed from it.
4801;;
0386b551
AM
4802;; Template types in C++ are added here too but with the template
4803;; arglist replaced with "<>" in references or "<" for the one in the
4804;; primary type. E.g. the type "Foo<A,B>::Bar<C>" is stored as
4805;; "Foo<>::Bar<". This avoids storing very long strings (since C++
4806;; template specs can be fairly sized programs in themselves) and
4807;; improves the hit ratio (it's a type regardless of the template
4808;; args; it's just not the same type, but we're only interested in
4809;; recognizing types, not telling distinct types apart). Note that
4810;; template types in references are added here too; from the example
4811;; above there will also be an entry "Foo<".
d9e94c22
MS
4812(defvar c-found-types nil)
4813(make-variable-buffer-local 'c-found-types)
4814
4815(defsubst c-clear-found-types ()
4816 ;; Clears `c-found-types'.
d9e94c22
MS
4817 (setq c-found-types (make-vector 53 0)))
4818
4819(defun c-add-type (from to)
4820 ;; Add the given region as a type in `c-found-types'. If the region
4821 ;; doesn't match an existing type but there is a type which is equal
4822 ;; to the given one except that the last character is missing, then
4823 ;; the shorter type is removed. That's done to avoid adding all
4824 ;; prefixes of a type as it's being entered and font locked. This
4825 ;; doesn't cover cases like when characters are removed from a type
4826 ;; or added in the middle. We'd need the position of point when the
4827 ;; font locking is invoked to solve this well.
0386b551
AM
4828 ;;
4829 ;; This function might do hidden buffer changes.
4830 (let ((type (c-syntactic-content from to c-recognize-<>-arglists)))
4831 (unless (intern-soft type c-found-types)
4832 (unintern (substring type 0 -1) c-found-types)
4833 (intern type c-found-types))))
d9e94c22 4834
580fba94
AM
4835(defun c-unfind-type (name)
4836 ;; Remove the "NAME" from c-found-types, if present.
4837 (unintern name c-found-types))
4838
d9e94c22
MS
4839(defsubst c-check-type (from to)
4840 ;; Return non-nil if the given region contains a type in
4841 ;; `c-found-types'.
0386b551
AM
4842 ;;
4843 ;; This function might do hidden buffer changes.
4844 (intern-soft (c-syntactic-content from to c-recognize-<>-arglists)
4845 c-found-types))
d9e94c22
MS
4846
4847(defun c-list-found-types ()
4848 ;; Return all the types in `c-found-types' as a sorted list of
4849 ;; strings.
4850 (let (type-list)
4851 (mapatoms (lambda (type)
4852 (setq type-list (cons (symbol-name type)
4853 type-list)))
4854 c-found-types)
4855 (sort type-list 'string-lessp)))
a66cd3ee 4856
2f42c75f
DN
4857;; Shut up the byte compiler.
4858(defvar c-maybe-stale-found-type)
4859
580fba94
AM
4860(defun c-trim-found-types (beg end old-len)
4861 ;; An after change function which, in conjunction with the info in
4862 ;; c-maybe-stale-found-type (set in c-before-change), removes a type
4863 ;; from `c-found-types', should this type have become stale. For
4864 ;; example, this happens to "foo" when "foo \n bar();" becomes
4865 ;; "foo(); \n bar();". Such stale types, if not removed, foul up
4866 ;; the fontification.
b414f371 4867 ;;
580fba94
AM
4868 ;; Have we, perhaps, added non-ws characters to the front/back of a found
4869 ;; type?
4870 (when (> end beg)
4871 (save-excursion
4872 (when (< end (point-max))
4873 (goto-char end)
4874 (if (and (c-beginning-of-current-token) ; only moves when we started in the middle
4875 (progn (goto-char end)
4876 (c-end-of-current-token)))
4877 (c-unfind-type (buffer-substring-no-properties
4878 end (point)))))
4879 (when (> beg (point-min))
4880 (goto-char beg)
4881 (if (and (c-end-of-current-token) ; only moves when we started in the middle
4882 (progn (goto-char beg)
4883 (c-beginning-of-current-token)))
4884 (c-unfind-type (buffer-substring-no-properties
4885 (point) beg))))))
b414f371 4886
580fba94
AM
4887 (if c-maybe-stale-found-type ; e.g. (c-decl-id-start "foo" 97 107 " (* ooka) " "o")
4888 (cond
4889 ;; Changing the amount of (already existing) whitespace - don't do anything.
4890 ((and (c-partial-ws-p beg end)
4891 (or (= beg end) ; removal of WS
580fba94
AM
4892 (string-match "^[ \t\n\r\f\v]*$" (nth 5 c-maybe-stale-found-type)))))
4893
4894 ;; The syntactic relationship which defined a "found type" has been
4895 ;; destroyed.
4896 ((eq (car c-maybe-stale-found-type) 'c-decl-id-start)
4897 (c-unfind-type (cadr c-maybe-stale-found-type)))
4898;; ((eq (car c-maybe-stale-found-type) 'c-decl-type-start) FIXME!!!
4899 )))
4900
d9e94c22 4901\f
dd969a56
AM
4902;; Setting and removing syntax properties on < and > in languages (C++
4903;; and Java) where they can be template/generic delimiters as well as
4904;; their normal meaning of "less/greater than".
4905
4906;; Normally, < and > have syntax 'punctuation'. When they are found to
4907;; be delimiters, they are marked as such with the category properties
4908;; c-<-as-paren-syntax, c->-as-paren-syntax respectively.
4909
4910;; STRATEGY:
4911;;
4912;; It is impossible to determine with certainty whether a <..> pair in
4913;; C++ is two comparison operators or is template delimiters, unless
4914;; one duplicates a lot of a C++ compiler. For example, the following
4915;; code fragment:
4916;;
4917;; foo (a < b, c > d) ;
4918;;
4919;; could be a function call with two integer parameters (each a
4920;; relational expression), or it could be a constructor for class foo
4921;; taking one parameter d of templated type "a < b, c >". They are
4922;; somewhat easier to distinguish in Java.
4923;;
4924;; The strategy now (2010-01) adopted is to mark and unmark < and
4925;; > IN MATCHING PAIRS ONLY. [Previously, they were marked
4926;; individually when their context so indicated. This gave rise to
4927;; intractible problems when one of a matching pair was deleted, or
4928;; pulled into a literal.]
4929;;
4930;; At each buffer change, the syntax-table properties are removed in a
4931;; before-change function and reapplied, when needed, in an
4932;; after-change function. It is far more important that the
4933;; properties get removed when they they are spurious than that they
4934;; be present when wanted.
4935;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
4936(defun c-clear-<-pair-props (&optional pos)
4937 ;; POS (default point) is at a < character. If it is marked with
4938 ;; open paren syntax-table text property, remove the property,
4939 ;; together with the close paren property on the matching > (if
4940 ;; any).
4941 (save-excursion
4942 (if pos
4943 (goto-char pos)
4944 (setq pos (point)))
4945 (when (equal (c-get-char-property (point) 'syntax-table)
4946 c-<-as-paren-syntax)
4947 (with-syntax-table c-no-parens-syntax-table ; ignore unbalanced [,{,(,..
4948 (c-go-list-forward))
4949 (when (equal (c-get-char-property (1- (point)) 'syntax-table)
4950 c->-as-paren-syntax) ; should always be true.
8a249abc
AM
4951 (c-clear-char-property (1- (point)) 'category))
4952 (c-clear-char-property pos 'category))))
dd969a56
AM
4953
4954(defun c-clear->-pair-props (&optional pos)
4955 ;; POS (default point) is at a > character. If it is marked with
4956 ;; close paren syntax-table property, remove the property, together
4957 ;; with the open paren property on the matching < (if any).
4958 (save-excursion
4959 (if pos
4960 (goto-char pos)
4961 (setq pos (point)))
4962 (when (equal (c-get-char-property (point) 'syntax-table)
4963 c->-as-paren-syntax)
4964 (with-syntax-table c-no-parens-syntax-table ; ignore unbalanced [,{,(,..
4965 (c-go-up-list-backward))
4966 (when (equal (c-get-char-property (point) 'syntax-table)
4967 c-<-as-paren-syntax) ; should always be true.
8a249abc
AM
4968 (c-clear-char-property (point) 'category))
4969 (c-clear-char-property pos 'category))))
dd969a56
AM
4970
4971(defun c-clear-<>-pair-props (&optional pos)
4972 ;; POS (default point) is at a < or > character. If it has an
4973 ;; open/close paren syntax-table property, remove this property both
4974 ;; from the current character and its partner (which will also be
4975 ;; thusly marked).
4976 (cond
4977 ((eq (char-after) ?\<)
4978 (c-clear-<-pair-props pos))
4979 ((eq (char-after) ?\>)
4980 (c-clear->-pair-props pos))
4981 (t (c-benign-error
4982 "c-clear-<>-pair-props called from wrong position"))))
4983
4984(defun c-clear-<-pair-props-if-match-after (lim &optional pos)
4985 ;; POS (default point) is at a < character. If it is both marked
4986 ;; with open/close paren syntax-table property, and has a matching >
4987 ;; (also marked) which is after LIM, remove the property both from
43a91810
AM
4988 ;; the current > and its partner. Return t when this happens, nil
4989 ;; when it doesn't.
dd969a56
AM
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-list-forward))
4998 (when (and (>= (point) lim)
4999 (equal (c-get-char-property (1- (point)) 'syntax-table)
5000 c->-as-paren-syntax)) ; should always be true.
5001 (c-unmark-<->-as-paren (1- (point)))
43a91810
AM
5002 (c-unmark-<->-as-paren pos))
5003 t)))
dd969a56
AM
5004
5005(defun c-clear->-pair-props-if-match-before (lim &optional pos)
5006 ;; POS (default point) is at a > character. If it is both marked
5007 ;; with open/close paren syntax-table property, and has a matching <
5008 ;; (also marked) which is before LIM, remove the property both from
43a91810
AM
5009 ;; the current < and its partner. Return t when this happens, nil
5010 ;; when it doesn't.
dd969a56
AM
5011 (save-excursion
5012 (if pos
5013 (goto-char pos)
5014 (setq pos (point)))
5015 (when (equal (c-get-char-property (point) 'syntax-table)
5016 c->-as-paren-syntax)
5017 (with-syntax-table c-no-parens-syntax-table ; ignore unbalanced [,{,(,..
5018 (c-go-up-list-backward))
5019 (when (and (<= (point) lim)
5020 (equal (c-get-char-property (point) 'syntax-table)
5021 c-<-as-paren-syntax)) ; should always be true.
5022 (c-unmark-<->-as-paren (point))
43a91810
AM
5023 (c-unmark-<->-as-paren pos))
5024 t)))
dd969a56
AM
5025
5026(defun c-before-change-check-<>-operators (beg end)
5027 ;; Unmark certain pairs of "< .... >" which are currently marked as
5028 ;; template/generic delimiters. (This marking is via syntax-table
5029 ;; text properties).
5030 ;;
5031 ;; These pairs are those which are in the current "statement" (i.e.,
5032 ;; the region between the {, }, or ; before BEG and the one after
5033 ;; END), and which enclose any part of the interval (BEG END).
5034 ;;
5035 ;; Note that in C++ (?and Java), template/generic parens cannot
5036 ;; enclose a brace or semicolon, so we use these as bounds on the
5037 ;; region we must work on.
5038 ;;
5039 ;; This function is called from before-change-functions (via
5040 ;; c-get-state-before-change-functions). Thus the buffer is widened,
5041 ;; and point is undefined, both at entry and exit.
5042 ;;
5043 ;; FIXME!!! This routine ignores the possibility of macros entirely.
5044 ;; 2010-01-29.
5045 (save-excursion
5046 (let ((beg-lit-limits (progn (goto-char beg) (c-literal-limits)))
43a91810
AM
5047 (end-lit-limits (progn (goto-char end) (c-literal-limits)))
5048 new-beg new-end need-new-beg need-new-end)
dd969a56
AM
5049 ;; Locate the barrier before the changed region
5050 (goto-char (if beg-lit-limits (car beg-lit-limits) beg))
5051 (c-syntactic-skip-backward "^;{}" (max (- beg 2048) (point-min)))
43a91810 5052 (setq new-beg (point))
dd969a56
AM
5053
5054 ;; Remove the syntax-table properties from each pertinent <...> pair.
5055 ;; Firsly, the ones with the < before beg and > after beg.
5056 (while (c-search-forward-char-property 'category 'c-<-as-paren-syntax beg)
43a91810
AM
5057 (if (c-clear-<-pair-props-if-match-after beg (1- (point)))
5058 (setq need-new-beg t)))
dd969a56
AM
5059
5060 ;; Locate the barrier after END.
5061 (goto-char (if end-lit-limits (cdr end-lit-limits) end))
5062 (c-syntactic-re-search-forward "[;{}]"
5063 (min (+ end 2048) (point-max)) 'end)
43a91810 5064 (setq new-end (point))
dd969a56
AM
5065
5066 ;; Remove syntax-table properties from the remaining pertinent <...>
5067 ;; pairs, those with a > after end and < before end.
5068 (while (c-search-backward-char-property 'category 'c->-as-paren-syntax end)
43a91810
AM
5069 (if (c-clear->-pair-props-if-match-before end)
5070 (setq need-new-end t)))
5071
5072 ;; Extend the fontification region, if needed.
5073 (when need-new-beg
5074 (goto-char new-beg)
5075 (c-forward-syntactic-ws)
5076 (and (< (point) c-new-BEG) (setq c-new-BEG (point))))
5077
5078 (when need-new-end
5079 (and (> new-end c-new-END) (setq c-new-END new-end))))))
dd969a56
AM
5080
5081
d9e94c22 5082
0386b551
AM
5083(defun c-after-change-check-<>-operators (beg end)
5084 ;; This is called from `after-change-functions' when
5085 ;; c-recognize-<>-arglists' is set. It ensures that no "<" or ">"
5086 ;; chars with paren syntax become part of another operator like "<<"
5087 ;; or ">=".
5088 ;;
5089 ;; This function might do hidden buffer changes.
5090
cb694ab7
AM
5091 (save-excursion
5092 (goto-char beg)
5093 (when (or (looking-at "[<>]")
5094 (< (skip-chars-backward "<>") 0))
5095
0386b551 5096 (goto-char beg)
cb694ab7
AM
5097 (c-beginning-of-current-token)
5098 (when (and (< (point) beg)
5099 (looking-at c-<>-multichar-token-regexp)
5100 (< beg (setq beg (match-end 0))))
5101 (while (progn (skip-chars-forward "^<>" beg)
5102 (< (point) beg))
dd969a56 5103 (c-clear-<>-pair-props)
cb694ab7
AM
5104 (forward-char))))
5105
5106 (when (< beg end)
5107 (goto-char end)
0386b551
AM
5108 (when (or (looking-at "[<>]")
5109 (< (skip-chars-backward "<>") 0))
5110
cb694ab7 5111 (goto-char end)
0386b551 5112 (c-beginning-of-current-token)
cb694ab7 5113 (when (and (< (point) end)
0386b551 5114 (looking-at c-<>-multichar-token-regexp)
cb694ab7
AM
5115 (< end (setq end (match-end 0))))
5116 (while (progn (skip-chars-forward "^<>" end)
5117 (< (point) end))
dd969a56 5118 (c-clear-<>-pair-props)
cb694ab7 5119 (forward-char)))))))
d9e94c22 5120
dd969a56
AM
5121
5122\f
5123;; Handling of small scale constructs like types and names.
5124
d9e94c22
MS
5125;; Dynamically bound variable that instructs `c-forward-type' to also
5126;; treat possible types (i.e. those that it normally returns 'maybe or
5127;; 'found for) as actual types (and always return 'found for them).
5128;; This means that it records them in `c-record-type-identifiers' if
5129;; that is set, and that it adds them to `c-found-types'.
5130(defvar c-promote-possible-types nil)
5131
0386b551
AM
5132;; Dynamically bound variable that instructs `c-forward-<>-arglist' to
5133;; mark up successfully parsed arglists with paren syntax properties on
5134;; the surrounding angle brackets and with `c-<>-arg-sep' in the
5135;; `c-type' property of each argument separating comma.
5136;;
5137;; Setting this variable also makes `c-forward-<>-arglist' recurse into
5138;; all arglists for side effects (i.e. recording types), otherwise it
5139;; exploits any existing paren syntax properties to quickly jump to the
5140;; end of already parsed arglists.
5141;;
5142;; Marking up the arglists is not the default since doing that correctly
5143;; depends on a proper value for `c-restricted-<>-arglists'.
5144(defvar c-parse-and-markup-<>-arglists nil)
5145
d9e94c22 5146;; Dynamically bound variable that instructs `c-forward-<>-arglist' to
037558bf
MS
5147;; not accept arglists that contain binary operators.
5148;;
5149;; This is primarily used to handle C++ template arglists. C++
5150;; disambiguates them by checking whether the preceding name is a
5151;; template or not. We can't do that, so we assume it is a template
5152;; if it can be parsed as one. That usually works well since
5153;; comparison expressions on the forms "a < b > c" or "a < b, c > d"
5154;; in almost all cases would be pointless.
5155;;
5156;; However, in function arglists, e.g. in "foo (a < b, c > d)", we
5157;; should let the comma separate the function arguments instead. And
5158;; in a context where the value of the expression is taken, e.g. in
5159;; "if (a < b || c > d)", it's probably not a template.
5160(defvar c-restricted-<>-arglists nil)
d9e94c22 5161
0386b551
AM
5162;; Dynamically bound variables that instructs
5163;; `c-forward-keyword-clause', `c-forward-<>-arglist',
5164;; `c-forward-name', `c-forward-type', `c-forward-decl-or-cast-1', and
5165;; `c-forward-label' to record the ranges of all the type and
5166;; reference identifiers they encounter. They will build lists on
5167;; these variables where each element is a cons of the buffer
5168;; positions surrounding each identifier. This recording is only
5169;; activated when `c-record-type-identifiers' is non-nil.
d9e94c22
MS
5170;;
5171;; All known types that can't be identifiers are recorded, and also
5172;; other possible types if `c-promote-possible-types' is set.
5173;; Recording is however disabled inside angle bracket arglists that
5174;; are encountered inside names and other angle bracket arglists.
0386b551 5175;; Such occurrences are taken care of by `c-font-lock-<>-arglists'
d9e94c22
MS
5176;; instead.
5177;;
5178;; Only the names in C++ template style references (e.g. "tmpl" in
5179;; "tmpl<a,b>::foo") are recorded as references, other references
5180;; aren't handled here.
0386b551
AM
5181;;
5182;; `c-forward-label' records the label identifier(s) on
5183;; `c-record-ref-identifiers'.
d9e94c22
MS
5184(defvar c-record-type-identifiers nil)
5185(defvar c-record-ref-identifiers nil)
5186
0386b551
AM
5187;; This variable will receive a cons cell of the range of the last
5188;; single identifier symbol stepped over by `c-forward-name' if it's
5189;; successful. This is the range that should be put on one of the
5190;; record lists above by the caller. It's assigned nil if there's no
5191;; such symbol in the name.
d9e94c22
MS
5192(defvar c-last-identifier-range nil)
5193
5194(defmacro c-record-type-id (range)
5195 (if (eq (car-safe range) 'cons)
5196 ;; Always true.
5197 `(setq c-record-type-identifiers
5198 (cons ,range c-record-type-identifiers))
5199 `(let ((range ,range))
5200 (if range
5201 (setq c-record-type-identifiers
5202 (cons range c-record-type-identifiers))))))
5203
5204(defmacro c-record-ref-id (range)
5205 (if (eq (car-safe range) 'cons)
5206 ;; Always true.
5207 `(setq c-record-ref-identifiers
5208 (cons ,range c-record-ref-identifiers))
5209 `(let ((range ,range))
5210 (if range
5211 (setq c-record-ref-identifiers
5212 (cons range c-record-ref-identifiers))))))
5213
5214;; Dynamically bound variable that instructs `c-forward-type' to
5215;; record the ranges of types that only are found. Behaves otherwise
5216;; like `c-record-type-identifiers'.
5217(defvar c-record-found-types nil)
5218
5219(defmacro c-forward-keyword-prefixed-id (type)
5220 ;; Used internally in `c-forward-keyword-clause' to move forward
5221 ;; over a type (if TYPE is 'type) or a name (otherwise) which
5222 ;; possibly is prefixed by keywords and their associated clauses.
5223 ;; Try with a type/name first to not trip up on those that begin
5224 ;; with a keyword. Return t if a known or found type is moved
5225 ;; over. The point is clobbered if nil is returned. If range
5226 ;; recording is enabled, the identifier is recorded on as a type
5227 ;; if TYPE is 'type or as a reference if TYPE is 'ref.
0386b551
AM
5228 ;;
5229 ;; This macro might do hidden buffer changes.
d9e94c22
MS
5230 `(let (res)
5231 (while (if (setq res ,(if (eq type 'type)
5232 `(c-forward-type)
5233 `(c-forward-name)))
5234 nil
5235 (and (looking-at c-keywords-regexp)
0386b551 5236 (c-forward-keyword-clause 1))))
d9e94c22
MS
5237 (when (memq res '(t known found prefix))
5238 ,(when (eq type 'ref)
5239 `(when c-record-type-identifiers
5240 (c-record-ref-id c-last-identifier-range)))
5241 t)))
5242
0386b551 5243(defmacro c-forward-id-comma-list (type update-safe-pos)
d9e94c22
MS
5244 ;; Used internally in `c-forward-keyword-clause' to move forward
5245 ;; over a comma separated list of types or names using
5246 ;; `c-forward-keyword-prefixed-id'.
0386b551
AM
5247 ;;
5248 ;; This macro might do hidden buffer changes.
d9e94c22 5249 `(while (and (progn
0386b551
AM
5250 ,(when update-safe-pos
5251 `(setq safe-pos (point)))
d9e94c22
MS
5252 (eq (char-after) ?,))
5253 (progn
5254 (forward-char)
5255 (c-forward-syntactic-ws)
5256 (c-forward-keyword-prefixed-id ,type)))))
5257
0386b551
AM
5258(defun c-forward-keyword-clause (match)
5259 ;; Submatch MATCH in the current match data is assumed to surround a
5260 ;; token. If it's a keyword, move over it and any immediately
5261 ;; following clauses associated with it, stopping at the start of
5262 ;; the next token. t is returned in that case, otherwise the point
d9e94c22
MS
5263 ;; stays and nil is returned. The kind of clauses that are
5264 ;; recognized are those specified by `c-type-list-kwds',
5265 ;; `c-ref-list-kwds', `c-colon-type-list-kwds',
5266 ;; `c-paren-nontype-kwds', `c-paren-type-kwds', `c-<>-type-kwds',
5267 ;; and `c-<>-arglist-kwds'.
0386b551
AM
5268 ;;
5269 ;; This function records identifier ranges on
5270 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
5271 ;; `c-record-type-identifiers' is non-nil.
5272 ;;
5273 ;; Note that for `c-colon-type-list-kwds', which doesn't necessary
5274 ;; apply directly after the keyword, the type list is moved over
5275 ;; only when there is no unaccounted token before it (i.e. a token
5276 ;; that isn't moved over due to some other keyword list). The
5277 ;; identifier ranges in the list are still recorded if that should
5278 ;; be done, though.
5279 ;;
5280 ;; This function might do hidden buffer changes.
5281
5282 (let ((kwd-sym (c-keyword-sym (match-string match))) safe-pos pos
5283 ;; The call to `c-forward-<>-arglist' below is made after
5284 ;; `c-<>-sexp-kwds' keywords, so we're certain they actually
5285 ;; are angle bracket arglists and `c-restricted-<>-arglists'
5286 ;; should therefore be nil.
5287 (c-parse-and-markup-<>-arglists t)
5288 c-restricted-<>-arglists)
d9e94c22 5289
d9e94c22 5290 (when kwd-sym
0386b551 5291 (goto-char (match-end match))
d9e94c22
MS
5292 (c-forward-syntactic-ws)
5293 (setq safe-pos (point))
a66cd3ee 5294
d9e94c22
MS
5295 (cond
5296 ((and (c-keyword-member kwd-sym 'c-type-list-kwds)
5297 (c-forward-keyword-prefixed-id type))
5298 ;; There's a type directly after a keyword in `c-type-list-kwds'.
0386b551 5299 (c-forward-id-comma-list type t))
d9e94c22
MS
5300
5301 ((and (c-keyword-member kwd-sym 'c-ref-list-kwds)
5302 (c-forward-keyword-prefixed-id ref))
5303 ;; There's a name directly after a keyword in `c-ref-list-kwds'.
0386b551 5304 (c-forward-id-comma-list ref t))
d9e94c22
MS
5305
5306 ((and (c-keyword-member kwd-sym 'c-paren-any-kwds)
5307 (eq (char-after) ?\())
5308 ;; There's an open paren after a keyword in `c-paren-any-kwds'.
5309
5310 (forward-char)
5311 (when (and (setq pos (c-up-list-forward))
5312 (eq (char-before pos) ?\)))
5313 (when (and c-record-type-identifiers
5314 (c-keyword-member kwd-sym 'c-paren-type-kwds))
5315 ;; Use `c-forward-type' on every identifier we can find
5316 ;; inside the paren, to record the types.
5317 (while (c-syntactic-re-search-forward c-symbol-start pos t)
5318 (goto-char (match-beginning 0))
5319 (unless (c-forward-type)
5320 (looking-at c-symbol-key) ; Always matches.
5321 (goto-char (match-end 0)))))
5322
5323 (goto-char pos)
5324 (c-forward-syntactic-ws)
5325 (setq safe-pos (point))))
5326
5327 ((and (c-keyword-member kwd-sym 'c-<>-sexp-kwds)
5328 (eq (char-after) ?<)
0386b551 5329 (c-forward-<>-arglist (c-keyword-member kwd-sym 'c-<>-type-kwds)))
d9e94c22
MS
5330 (c-forward-syntactic-ws)
5331 (setq safe-pos (point)))
5332
5333 ((and (c-keyword-member kwd-sym 'c-nonsymbol-sexp-kwds)
449a2b0d
MS
5334 (not (looking-at c-symbol-start))
5335 (c-safe (c-forward-sexp) t))
d9e94c22
MS
5336 (c-forward-syntactic-ws)
5337 (setq safe-pos (point))))
5338
0386b551
AM
5339 (when (c-keyword-member kwd-sym 'c-colon-type-list-kwds)
5340 (if (eq (char-after) ?:)
5341 ;; If we are at the colon already, we move over the type
5342 ;; list after it.
5343 (progn
5344 (forward-char)
5345 (c-forward-syntactic-ws)
5346 (when (c-forward-keyword-prefixed-id type)
5347 (c-forward-id-comma-list type t)))
5348 ;; Not at the colon, so stop here. But the identifier
5349 ;; ranges in the type list later on should still be
5350 ;; recorded.
5351 (and c-record-type-identifiers
5352 (progn
5353 ;; If a keyword matched both one of the types above and
5354 ;; this one, we match `c-colon-type-list-re' after the
5355 ;; clause matched above.
5356 (goto-char safe-pos)
5357 (looking-at c-colon-type-list-re))
5358 (progn
5359 (goto-char (match-end 0))
5360 (c-forward-syntactic-ws)
5361 (c-forward-keyword-prefixed-id type))
5362 ;; There's a type after the `c-colon-type-list-re' match
5363 ;; after a keyword in `c-colon-type-list-kwds'.
5364 (c-forward-id-comma-list type nil))))
d9e94c22
MS
5365
5366 (goto-char safe-pos)
5367 t)))
5368
0386b551
AM
5369(defun c-forward-<>-arglist (all-types)
5370 ;; The point is assumed to be at a "<". Try to treat it as the open
b4dc7d98 5371 ;; paren of an angle bracket arglist and move forward to the
0386b551
AM
5372 ;; corresponding ">". If successful, the point is left after the
5373 ;; ">" and t is returned, otherwise the point isn't moved and nil is
d9e94c22
MS
5374 ;; returned. If ALL-TYPES is t then all encountered arguments in
5375 ;; the arglist that might be types are treated as found types.
5376 ;;
0386b551
AM
5377 ;; The variable `c-parse-and-markup-<>-arglists' controls how this
5378 ;; function handles text properties on the angle brackets and argument
5379 ;; separating commas.
d9e94c22 5380 ;;
0386b551
AM
5381 ;; `c-restricted-<>-arglists' controls how lenient the template
5382 ;; arglist recognition should be.
5383 ;;
5384 ;; This function records identifier ranges on
5385 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
5386 ;; `c-record-type-identifiers' is non-nil.
5387 ;;
5388 ;; This function might do hidden buffer changes.
d9e94c22
MS
5389
5390 (let ((start (point))
5391 ;; If `c-record-type-identifiers' is set then activate
5392 ;; recording of any found types that constitute an argument in
5393 ;; the arglist.
5394 (c-record-found-types (if c-record-type-identifiers t)))
5395 (if (catch 'angle-bracket-arglist-escape
5396 (setq c-record-found-types
0386b551 5397 (c-forward-<>-arglist-recur all-types)))
d9e94c22
MS
5398 (progn
5399 (when (consp c-record-found-types)
5400 (setq c-record-type-identifiers
5401 ;; `nconc' doesn't mind that the tail of
5402 ;; `c-record-found-types' is t.
5403 (nconc c-record-found-types c-record-type-identifiers)))
5404 t)
5405
5406 (goto-char start)
a66cd3ee 5407 nil)))
785eecbb 5408
0386b551 5409(defun c-forward-<>-arglist-recur (all-types)
d9e94c22 5410 ;; Recursive part of `c-forward-<>-arglist'.
0386b551
AM
5411 ;;
5412 ;; This function might do hidden buffer changes.
d9e94c22
MS
5413
5414 (let ((start (point)) res pos tmp
5415 ;; Cover this so that any recorded found type ranges are
5416 ;; automatically lost if it turns out to not be an angle
5417 ;; bracket arglist. It's propagated through the return value
5418 ;; on successful completion.
5419 (c-record-found-types c-record-found-types)
5420 ;; List that collects the positions after the argument
5421 ;; separating ',' in the arglist.
5422 arg-start-pos)
5423
0386b551
AM
5424 ;; If the '<' has paren open syntax then we've marked it as an angle
5425 ;; bracket arglist before, so skip to the end.
5426 (if (and (not c-parse-and-markup-<>-arglists)
5427 (c-get-char-property (point) 'syntax-table))
5428
5429 (progn
5430 (forward-char)
5431 (if (and (c-go-up-list-forward)
5432 (eq (char-before) ?>))
5433 t
5434
5435 ;; Got unmatched paren angle brackets. We don't clear the paren
5436 ;; syntax properties and retry, on the basis that it's very
5437 ;; unlikely that paren angle brackets become operators by code
5438 ;; manipulation. It's far more likely that it doesn't match due
5439 ;; to narrowing or some temporary change.
5440 (goto-char start)
5441 nil))
d9e94c22
MS
5442
5443 (forward-char)
5444 (unless (looking-at c-<-op-cont-regexp)
5445 (while (and
5446 (progn
5447
5448 (when c-record-type-identifiers
5449 (if all-types
5450
5451 ;; All encountered identifiers are types, so set the
5452 ;; promote flag and parse the type.
5453 (progn
5454 (c-forward-syntactic-ws)
5455 (when (looking-at c-identifier-start)
5456 (let ((c-promote-possible-types t))
5457 (c-forward-type))))
5458
5459 ;; Check if this arglist argument is a sole type. If
5460 ;; it's known then it's recorded in
5461 ;; `c-record-type-identifiers'. If it only is found
5462 ;; then it's recorded in `c-record-found-types' which we
5463 ;; might roll back if it turns out that this isn't an
5464 ;; angle bracket arglist afterall.
5465 (when (memq (char-before) '(?, ?<))
5466 (let ((orig-record-found-types c-record-found-types))
5467 (c-forward-syntactic-ws)
5468 (and (memq (c-forward-type) '(known found))
5469 (not (looking-at "[,>]"))
5470 ;; A found type was recorded but it's not the
5471 ;; only thing in the arglist argument, so reset
5472 ;; `c-record-found-types'.
5473 (setq c-record-found-types
5474 orig-record-found-types))))))
5475
5476 (setq pos (point))
5477 (or (when (eq (char-after) ?>)
5478 ;; Must check for '>' at the very start separately,
5479 ;; since the regexp below has to avoid ">>" without
5480 ;; using \\=.
5481 (forward-char)
5482 t)
5483
037558bf
MS
5484 ;; Note: These regexps exploit the match order in \| so
5485 ;; that "<>" is matched by "<" rather than "[^>:-]>".
d9e94c22 5486 (c-syntactic-re-search-forward
037558bf
MS
5487 (if c-restricted-<>-arglists
5488 ;; Stop on ',', '|', '&', '+' and '-' to catch
5489 ;; common binary operators that could be between
5490 ;; two comparison expressions "a<b" and "c>d".
5491 "[<;{},|&+-]\\|\\([^>:-]>\\)"
5492 ;; Otherwise we still stop on ',' to find the
5493 ;; argument start positions.
5494 "[<;{},]\\|\\([^>:-]>\\)")
5495 nil 'move t t 1)
d9e94c22
MS
5496
5497 ;; If the arglist starter has lost its open paren
5498 ;; syntax but not the closer, we won't find the
5499 ;; closer above since we only search in the
5500 ;; balanced sexp. In that case we stop just short
5501 ;; of it so check if the following char is the closer.
5502 (when (eq (char-after) ?>)
d9e94c22
MS
5503 (forward-char)
5504 t)))
5505
5506 (cond
5507 ((eq (char-before) ?>)
5508 ;; Either an operator starting with '>' or the end of
5509 ;; the angle bracket arglist.
5510
0386b551 5511 (if (looking-at c->-op-cont-regexp)
d9e94c22 5512 (progn
0386b551
AM
5513 (goto-char (match-end 0))
5514 t) ; Continue the loop.
d9e94c22 5515
0386b551
AM
5516 ;; The angle bracket arglist is finished.
5517 (when c-parse-and-markup-<>-arglists
d9e94c22 5518 (while arg-start-pos
0386b551
AM
5519 (c-put-c-type-property (1- (car arg-start-pos))
5520 'c-<>-arg-sep)
d9e94c22
MS
5521 (setq arg-start-pos (cdr arg-start-pos)))
5522 (c-mark-<-as-paren start)
0386b551
AM
5523 (c-mark->-as-paren (1- (point))))
5524 (setq res t)
5525 nil)) ; Exit the loop.
d9e94c22
MS
5526
5527 ((eq (char-before) ?<)
5528 ;; Either an operator starting with '<' or a nested arglist.
5529
5530 (setq pos (point))
5531 (let (id-start id-end subres keyword-match)
5532 (if (if (looking-at c-<-op-cont-regexp)
5533 (setq tmp (match-end 0))
5534 (setq tmp pos)
5535 (backward-char)
5536 (not
5537 (and
5538
5539 (save-excursion
0386b551 5540 ;; There's always an identifier before an angle
d9e94c22
MS
5541 ;; bracket arglist, or a keyword in
5542 ;; `c-<>-type-kwds' or `c-<>-arglist-kwds'.
5543 (c-backward-syntactic-ws)
5544 (setq id-end (point))
5545 (c-simple-skip-symbol-backward)
5546 (when (or (setq keyword-match
5547 (looking-at c-opt-<>-sexp-key))
5548 (not (looking-at c-keywords-regexp)))
5549 (setq id-start (point))))
5550
5551 (setq subres
5552 (let ((c-record-type-identifiers nil)
5553 (c-record-found-types nil))
5554 (c-forward-<>-arglist-recur
5555 (and keyword-match
5556 (c-keyword-member
5557 (c-keyword-sym (match-string 1))
0386b551 5558 'c-<>-type-kwds)))))
d9e94c22
MS
5559 )))
5560
5561 ;; It was not an angle bracket arglist.
0386b551 5562 (goto-char tmp)
d9e94c22
MS
5563
5564 ;; It was an angle bracket arglist.
5565 (setq c-record-found-types subres)
5566
5567 ;; Record the identifier before the template as a type
5568 ;; or reference depending on whether the arglist is last
5569 ;; in a qualified identifier.
5570 (when (and c-record-type-identifiers
5571 (not keyword-match))
5572 (if (and c-opt-identifier-concat-key
5573 (progn
5574 (c-forward-syntactic-ws)
5575 (looking-at c-opt-identifier-concat-key)))
5576 (c-record-ref-id (cons id-start id-end))
5577 (c-record-type-id (cons id-start id-end))))))
5578 t)
5579
5580 ((and (eq (char-before) ?,)
037558bf 5581 (not c-restricted-<>-arglists))
d9e94c22
MS
5582 ;; Just another argument. Record the position. The
5583 ;; type check stuff that made us stop at it is at
5584 ;; the top of the loop.
5585 (setq arg-start-pos (cons (point) arg-start-pos)))
5586
5587 (t
5588 ;; Got a character that can't be in an angle bracket
5589 ;; arglist argument. Abort using `throw', since
5590 ;; it's useless to try to find a surrounding arglist
5591 ;; if we're nested.
5592 (throw 'angle-bracket-arglist-escape nil))))))
5593
5594 (if res
5595 (or c-record-found-types t)))))
5596
0386b551
AM
5597(defun c-backward-<>-arglist (all-types &optional limit)
5598 ;; The point is assumed to be directly after a ">". Try to treat it
5599 ;; as the close paren of an angle bracket arglist and move back to
5600 ;; the corresponding "<". If successful, the point is left at
5601 ;; the "<" and t is returned, otherwise the point isn't moved and
5602 ;; nil is returned. ALL-TYPES is passed on to
5603 ;; `c-forward-<>-arglist'.
5604 ;;
5605 ;; If the optional LIMIT is given, it bounds the backward search.
5606 ;; It's then assumed to be at a syntactically relevant position.
5607 ;;
5608 ;; This is a wrapper around `c-forward-<>-arglist'. See that
5609 ;; function for more details.
5610
5611 (let ((start (point)))
5612 (backward-char)
5613 (if (and (not c-parse-and-markup-<>-arglists)
5614 (c-get-char-property (point) 'syntax-table))
5615
5616 (if (and (c-go-up-list-backward)
5617 (eq (char-after) ?<))
5618 t
5619 ;; See corresponding note in `c-forward-<>-arglist'.
5620 (goto-char start)
5621 nil)
5622
51c9af45 5623 (while (progn
0386b551
AM
5624 (c-syntactic-skip-backward "^<;{}" limit t)
5625
51c9af45
AM
5626 (and
5627 (if (eq (char-before) ?<)
5628 t
5629 ;; Stopped at bob or a char that isn't allowed in an
5630 ;; arglist, so we've failed.
5631 (goto-char start)
5632 nil)
0386b551 5633
51c9af45
AM
5634 (if (> (point)
5635 (progn (c-beginning-of-current-token)
5636 (point)))
5637 ;; If we moved then the "<" was part of some
5638 ;; multicharacter token.
5639 t
0386b551 5640
51c9af45
AM
5641 (backward-char)
5642 (let ((beg-pos (point)))
5643 (if (c-forward-<>-arglist all-types)
5644 (cond ((= (point) start)
5645 ;; Matched the arglist. Break the while.
5646 (goto-char beg-pos)
5647 nil)
5648 ((> (point) start)
5649 ;; We started from a non-paren ">" inside an
5650 ;; arglist.
5651 (goto-char start)
5652 nil)
5653 (t
5654 ;; Matched a shorter arglist. Can be a nested
5655 ;; one so continue looking.
5656 (goto-char beg-pos)
5657 t))
5658 t))))))
0386b551
AM
5659
5660 (/= (point) start))))
5661
d9e94c22
MS
5662(defun c-forward-name ()
5663 ;; Move forward over a complete name if at the beginning of one,
5664 ;; stopping at the next following token. If the point is not at
5665 ;; something that are recognized as name then it stays put. A name
5666 ;; could be something as simple as "foo" in C or something as
5667 ;; complex as "X<Y<class A<int>::B, BIT_MAX >> b>, ::operator<> ::
5668 ;; Z<(a>b)> :: operator const X<&foo>::T Q::G<unsigned short
5669 ;; int>::*volatile const" in C++ (this function is actually little
5670 ;; more than a `looking-at' call in all modes except those that,
5671 ;; like C++, have `c-recognize-<>-arglists' set). Return nil if no
5672 ;; name is found, 'template if it's an identifier ending with an
5673 ;; angle bracket arglist, 'operator of it's an operator identifier,
5674 ;; or t if it's some other kind of name.
0386b551
AM
5675 ;;
5676 ;; This function records identifier ranges on
5677 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
5678 ;; `c-record-type-identifiers' is non-nil.
5679 ;;
5680 ;; This function might do hidden buffer changes.
d9e94c22 5681
0386b551 5682 (let ((pos (point)) (start (point)) res id-start id-end
d9e94c22
MS
5683 ;; Turn off `c-promote-possible-types' here since we might
5684 ;; call `c-forward-<>-arglist' and we don't want it to promote
5685 ;; every suspect thing in the arglist to a type. We're
5686 ;; typically called from `c-forward-type' in this case, and
5687 ;; the caller only wants the top level type that it finds to
5688 ;; be promoted.
5689 c-promote-possible-types)
5690 (while
5691 (and
5692 (looking-at c-identifier-key)
5693
5694 (progn
5695 ;; Check for keyword. We go to the last symbol in
5696 ;; `c-identifier-key' first.
0386b551
AM
5697 (goto-char (setq id-end (match-end 0)))
5698 (c-simple-skip-symbol-backward)
5699 (setq id-start (point))
d9e94c22
MS
5700
5701 (if (looking-at c-keywords-regexp)
5702 (when (and (c-major-mode-is 'c++-mode)
5703 (looking-at
5704 (cc-eval-when-compile
5705 (concat "\\(operator\\|\\(template\\)\\)"
5706 "\\(" (c-lang-const c-nonsymbol-key c++)
5707 "\\|$\\)")))
5708 (if (match-beginning 2)
5709 ;; "template" is only valid inside an
5710 ;; identifier if preceded by "::".
5711 (save-excursion
5712 (c-backward-syntactic-ws)
5713 (and (c-safe (backward-char 2) t)
5714 (looking-at "::")))
5715 t))
5716
5717 ;; Handle a C++ operator or template identifier.
5718 (goto-char id-end)
5719 (c-forward-syntactic-ws)
5720 (cond ((eq (char-before id-end) ?e)
5721 ;; Got "... ::template".
5722 (let ((subres (c-forward-name)))
5723 (when subres
5724 (setq pos (point)
5725 res subres))))
5726
5727 ((looking-at c-identifier-start)
5728 ;; Got a cast operator.
5729 (when (c-forward-type)
5730 (setq pos (point)
5731 res 'operator)
5732 ;; Now we should match a sequence of either
5733 ;; '*', '&' or a name followed by ":: *",
5734 ;; where each can be followed by a sequence
5735 ;; of `c-opt-type-modifier-key'.
5736 (while (cond ((looking-at "[*&]")
5737 (goto-char (match-end 0))
5738 t)
5739 ((looking-at c-identifier-start)
5740 (and (c-forward-name)
5741 (looking-at "::")
5742 (progn
5743 (goto-char (match-end 0))
5744 (c-forward-syntactic-ws)
5745 (eq (char-after) ?*))
5746 (progn
5747 (forward-char)
5748 t))))
5749 (while (progn
5750 (c-forward-syntactic-ws)
5751 (setq pos (point))
5752 (looking-at c-opt-type-modifier-key))
5753 (goto-char (match-end 1))))))
5754
5755 ((looking-at c-overloadable-operators-regexp)
5756 ;; Got some other operator.
0386b551
AM
5757 (setq c-last-identifier-range
5758 (cons (point) (match-end 0)))
d9e94c22
MS
5759 (goto-char (match-end 0))
5760 (c-forward-syntactic-ws)
5761 (setq pos (point)
5762 res 'operator)))
5763
5764 nil)
5765
0386b551
AM
5766 ;; `id-start' is equal to `id-end' if we've jumped over
5767 ;; an identifier that doesn't end with a symbol token.
5768 ;; That can occur e.g. for Java import directives on the
5769 ;; form "foo.bar.*".
5770 (when (and id-start (/= id-start id-end))
d9e94c22
MS
5771 (setq c-last-identifier-range
5772 (cons id-start id-end)))
5773 (goto-char id-end)
5774 (c-forward-syntactic-ws)
5775 (setq pos (point)
5776 res t)))
5777
5778 (progn
5779 (goto-char pos)
5780 (when (or c-opt-identifier-concat-key
5781 c-recognize-<>-arglists)
5782
5783 (cond
5784 ((and c-opt-identifier-concat-key
5785 (looking-at c-opt-identifier-concat-key))
5786 ;; Got a concatenated identifier. This handles the
5787 ;; cases with tricky syntactic whitespace that aren't
5788 ;; covered in `c-identifier-key'.
5789 (goto-char (match-end 0))
5790 (c-forward-syntactic-ws)
5791 t)
5792
5793 ((and c-recognize-<>-arglists
5794 (eq (char-after) ?<))
5795 ;; Maybe an angle bracket arglist.
0386b551
AM
5796
5797 (when (let (c-record-type-identifiers
5798 c-record-found-types)
5799 (c-forward-<>-arglist nil))
5800
5801 (c-add-type start (1+ pos))
d9e94c22 5802 (c-forward-syntactic-ws)
0386b551
AM
5803 (setq pos (point)
5804 c-last-identifier-range nil)
5805
d9e94c22
MS
5806 (if (and c-opt-identifier-concat-key
5807 (looking-at c-opt-identifier-concat-key))
0386b551 5808
d9e94c22
MS
5809 ;; Continue if there's an identifier concatenation
5810 ;; operator after the template argument.
5811 (progn
0386b551
AM
5812 (when (and c-record-type-identifiers id-start)
5813 (c-record-ref-id (cons id-start id-end)))
d9e94c22
MS
5814 (forward-char 2)
5815 (c-forward-syntactic-ws)
5816 t)
0386b551
AM
5817
5818 (when (and c-record-type-identifiers id-start)
5819 (c-record-type-id (cons id-start id-end)))
d9e94c22
MS
5820 (setq res 'template)
5821 nil)))
5822 )))))
5823
5824 (goto-char pos)
5825 res))
5826
5827(defun c-forward-type ()
5828 ;; Move forward over a type spec if at the beginning of one,
5829 ;; stopping at the next following token. Return t if it's a known
2a15eb73
MS
5830 ;; type that can't be a name or other expression, 'known if it's an
5831 ;; otherwise known type (according to `*-font-lock-extra-types'),
5832 ;; 'prefix if it's a known prefix of a type, 'found if it's a type
5833 ;; that matches one in `c-found-types', 'maybe if it's an identfier
5834 ;; that might be a type, or nil if it can't be a type (the point
5835 ;; isn't moved then). The point is assumed to be at the beginning
5836 ;; of a token.
d9e94c22
MS
5837 ;;
5838 ;; Note that this function doesn't skip past the brace definition
5839 ;; that might be considered part of the type, e.g.
5840 ;; "enum {a, b, c} foo".
0386b551
AM
5841 ;;
5842 ;; This function records identifier ranges on
5843 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
5844 ;; `c-record-type-identifiers' is non-nil.
5845 ;;
5846 ;; This function might do hidden buffer changes.
5847
5848 (let ((start (point)) pos res name-res id-start id-end id-range)
d9e94c22
MS
5849
5850 ;; Skip leading type modifiers. If any are found we know it's a
5851 ;; prefix of a type.
5852 (when c-opt-type-modifier-key
5853 (while (looking-at c-opt-type-modifier-key)
5854 (goto-char (match-end 1))
5855 (c-forward-syntactic-ws)
5856 (setq res 'prefix)))
5857
5858 (cond
5859 ((looking-at c-type-prefix-key)
5860 ;; Looking at a keyword that prefixes a type identifier,
5861 ;; e.g. "class".
5862 (goto-char (match-end 1))
5863 (c-forward-syntactic-ws)
5864 (setq pos (point))
0386b551 5865 (if (memq (setq name-res (c-forward-name)) '(t template))
d9e94c22 5866 (progn
0386b551 5867 (when (eq name-res t)
d9e94c22
MS
5868 ;; In many languages the name can be used without the
5869 ;; prefix, so we add it to `c-found-types'.
5870 (c-add-type pos (point))
0386b551
AM
5871 (when (and c-record-type-identifiers
5872 c-last-identifier-range)
d9e94c22
MS
5873 (c-record-type-id c-last-identifier-range)))
5874 (setq res t))
5875 ;; Invalid syntax.
5876 (goto-char start)
5877 (setq res nil)))
5878
5879 ((progn
5880 (setq pos nil)
5881 (if (looking-at c-identifier-start)
5882 (save-excursion
5883 (setq id-start (point)
0386b551
AM
5884 name-res (c-forward-name))
5885 (when name-res
d9e94c22
MS
5886 (setq id-end (point)
5887 id-range c-last-identifier-range))))
5888 (and (cond ((looking-at c-primitive-type-key)
5889 (setq res t))
5890 ((c-with-syntax-table c-identifier-syntax-table
5891 (looking-at c-known-type-key))
5892 (setq res 'known)))
5893 (or (not id-end)
5894 (>= (save-excursion
5895 (save-match-data
5896 (goto-char (match-end 1))
5897 (c-forward-syntactic-ws)
5898 (setq pos (point))))
5899 id-end)
5900 (setq res nil))))
5901 ;; Looking at a primitive or known type identifier. We've
5902 ;; checked for a name first so that we don't go here if the
5903 ;; known type match only is a prefix of another name.
5904
5905 (setq id-end (match-end 1))
5906
5907 (when (and c-record-type-identifiers
5908 (or c-promote-possible-types (eq res t)))
5909 (c-record-type-id (cons (match-beginning 1) (match-end 1))))
5910
5911 (if (and c-opt-type-component-key
5912 (save-match-data
5913 (looking-at c-opt-type-component-key)))
5914 ;; There might be more keywords for the type.
5915 (let (safe-pos)
0386b551 5916 (c-forward-keyword-clause 1)
d9e94c22
MS
5917 (while (progn
5918 (setq safe-pos (point))
5919 (looking-at c-opt-type-component-key))
5920 (when (and c-record-type-identifiers
5921 (looking-at c-primitive-type-key))
5922 (c-record-type-id (cons (match-beginning 1)
5923 (match-end 1))))
0386b551 5924 (c-forward-keyword-clause 1))
d9e94c22
MS
5925 (if (looking-at c-primitive-type-key)
5926 (progn
5927 (when c-record-type-identifiers
5928 (c-record-type-id (cons (match-beginning 1)
5929 (match-end 1))))
0386b551 5930 (c-forward-keyword-clause 1)
d9e94c22
MS
5931 (setq res t))
5932 (goto-char safe-pos)
5933 (setq res 'prefix)))
0386b551 5934 (unless (save-match-data (c-forward-keyword-clause 1))
d9e94c22
MS
5935 (if pos
5936 (goto-char pos)
5937 (goto-char (match-end 1))
5938 (c-forward-syntactic-ws)))))
5939
0386b551
AM
5940 (name-res
5941 (cond ((eq name-res t)
d9e94c22
MS
5942 ;; A normal identifier.
5943 (goto-char id-end)
5944 (if (or res c-promote-possible-types)
5945 (progn
5946 (c-add-type id-start id-end)
0386b551 5947 (when (and c-record-type-identifiers id-range)
d9e94c22
MS
5948 (c-record-type-id id-range))
5949 (unless res
5950 (setq res 'found)))
5951 (setq res (if (c-check-type id-start id-end)
5952 ;; It's an identifier that has been used as
5953 ;; a type somewhere else.
5954 'found
5955 ;; It's an identifier that might be a type.
5956 'maybe))))
0386b551 5957 ((eq name-res 'template)
d9e94c22
MS
5958 ;; A template is a type.
5959 (goto-char id-end)
5960 (setq res t))
5961 (t
5962 ;; Otherwise it's an operator identifier, which is not a type.
5963 (goto-char start)
5964 (setq res nil)))))
5965
5966 (when res
5967 ;; Skip trailing type modifiers. If any are found we know it's
5968 ;; a type.
5969 (when c-opt-type-modifier-key
5970 (while (looking-at c-opt-type-modifier-key)
5971 (goto-char (match-end 1))
5972 (c-forward-syntactic-ws)
5973 (setq res t)))
5974
5975 ;; Step over any type suffix operator. Do not let the existence
5976 ;; of these alter the classification of the found type, since
5977 ;; these operators typically are allowed in normal expressions
5978 ;; too.
5979 (when c-opt-type-suffix-key
5980 (while (looking-at c-opt-type-suffix-key)
5981 (goto-char (match-end 1))
5982 (c-forward-syntactic-ws)))
5983
5984 (when c-opt-type-concat-key
0386b551
AM
5985 ;; Look for a trailing operator that concatenates the type
5986 ;; with a following one, and if so step past that one through
5987 ;; a recursive call. Note that we don't record concatenated
5988 ;; types in `c-found-types' - it's the component types that
5989 ;; are recorded when appropriate.
d9e94c22
MS
5990 (setq pos (point))
5991 (let* ((c-promote-possible-types (or (memq res '(t known))
5992 c-promote-possible-types))
5993 ;; If we can't promote then set `c-record-found-types' so that
5994 ;; we can merge in the types from the second part afterwards if
5995 ;; it turns out to be a known type there.
5996 (c-record-found-types (and c-record-type-identifiers
0386b551
AM
5997 (not c-promote-possible-types)))
5998 subres)
d9e94c22
MS
5999 (if (and (looking-at c-opt-type-concat-key)
6000
6001 (progn
6002 (goto-char (match-end 1))
6003 (c-forward-syntactic-ws)
0386b551 6004 (setq subres (c-forward-type))))
d9e94c22
MS
6005
6006 (progn
6007 ;; If either operand certainly is a type then both are, but we
6008 ;; don't let the existence of the operator itself promote two
6009 ;; uncertain types to a certain one.
6010 (cond ((eq res t))
0386b551
AM
6011 ((eq subres t)
6012 (unless (eq name-res 'template)
6013 (c-add-type id-start id-end))
6014 (when (and c-record-type-identifiers id-range)
d9e94c22
MS
6015 (c-record-type-id id-range))
6016 (setq res t))
2a15eb73 6017 ((eq res 'known))
0386b551 6018 ((eq subres 'known)
2a15eb73 6019 (setq res 'known))
d9e94c22 6020 ((eq res 'found))
0386b551 6021 ((eq subres 'found)
d9e94c22
MS
6022 (setq res 'found))
6023 (t
6024 (setq res 'maybe)))
6025
6026 (when (and (eq res t)
6027 (consp c-record-found-types))
6028 ;; Merge in the ranges of any types found by the second
6029 ;; `c-forward-type'.
6030 (setq c-record-type-identifiers
6031 ;; `nconc' doesn't mind that the tail of
6032 ;; `c-record-found-types' is t.
6033 (nconc c-record-found-types
6034 c-record-type-identifiers))))
6035
6036 (goto-char pos))))
6037
6038 (when (and c-record-found-types (memq res '(known found)) id-range)
6039 (setq c-record-found-types
6040 (cons id-range c-record-found-types))))
6041
6042 ;;(message "c-forward-type %s -> %s: %s" start (point) res)
6043
6044 res))
6045
785eecbb 6046\f
d9e94c22
MS
6047;; Handling of large scale constructs like statements and declarations.
6048
0386b551
AM
6049;; Macro used inside `c-forward-decl-or-cast-1'. It ought to be a
6050;; defsubst or perhaps even a defun, but it contains lots of free
6051;; variables that refer to things inside `c-forward-decl-or-cast-1'.
6052(defmacro c-fdoc-shift-type-backward (&optional short)
6053 ;; `c-forward-decl-or-cast-1' can consume an arbitrary length list
6054 ;; of types when parsing a declaration, which means that it
6055 ;; sometimes consumes the identifier in the declaration as a type.
6056 ;; This is used to "backtrack" and make the last type be treated as
6057 ;; an identifier instead.
6058 `(progn
6059 ,(unless short
6060 ;; These identifiers are bound only in the inner let.
6061 '(setq identifier-type at-type
6062 identifier-start type-start
6063 got-parens nil
6064 got-identifier t
6065 got-suffix t
6066 got-suffix-after-parens id-start
6067 paren-depth 0))
6068
6069 (if (setq at-type (if (eq backup-at-type 'prefix)
6070 t
6071 backup-at-type))
6072 (setq type-start backup-type-start
6073 id-start backup-id-start)
6074 (setq type-start start-pos
6075 id-start start-pos))
6076
6077 ;; When these flags already are set we've found specifiers that
6078 ;; unconditionally signal these attributes - backtracking doesn't
6079 ;; change that. So keep them set in that case.
6080 (or at-type-decl
6081 (setq at-type-decl backup-at-type-decl))
6082 (or maybe-typeless
6083 (setq maybe-typeless backup-maybe-typeless))
6084
6085 ,(unless short
6086 ;; This identifier is bound only in the inner let.
6087 '(setq start id-start))))
6088
6089(defun c-forward-decl-or-cast-1 (preceding-token-end context last-cast-end)
6090 ;; Move forward over a declaration or a cast if at the start of one.
6091 ;; The point is assumed to be at the start of some token. Nil is
6092 ;; returned if no declaration or cast is recognized, and the point
6093 ;; is clobbered in that case.
6094 ;;
6095 ;; If a declaration is parsed:
6096 ;;
6097 ;; The point is left at the first token after the first complete
6098 ;; declarator, if there is one. The return value is a cons where
022d0cf4
AM
6099 ;; the car is the position of the first token in the declarator. (See
6100 ;; below for the cdr.)
0386b551
AM
6101 ;; Some examples:
6102 ;;
6103 ;; void foo (int a, char *b) stuff ...
6104 ;; car ^ ^ point
6105 ;; float (*a)[], b;
6106 ;; car ^ ^ point
6107 ;; unsigned int a = c_style_initializer, b;
6108 ;; car ^ ^ point
6109 ;; unsigned int a (cplusplus_style_initializer), b;
6110 ;; car ^ ^ point (might change)
6111 ;; class Foo : public Bar {}
6112 ;; car ^ ^ point
6113 ;; class PikeClass (int a, string b) stuff ...
6114 ;; car ^ ^ point
6115 ;; enum bool;
6116 ;; car ^ ^ point
6117 ;; enum bool flag;
6118 ;; car ^ ^ point
6119 ;; void cplusplus_function (int x) throw (Bad);
6120 ;; car ^ ^ point
6121 ;; Foo::Foo (int b) : Base (b) {}
6122 ;; car ^ ^ point
6123 ;;
022d0cf4
AM
6124 ;; The cdr of the return value is non-nil iff a `c-typedef-decl-kwds'
6125 ;; specifier (e.g. class, struct, enum, typedef) is found in the
6126 ;; declaration, i.e. the declared identifier(s) are types.
0386b551
AM
6127 ;;
6128 ;; If a cast is parsed:
6129 ;;
6130 ;; The point is left at the first token after the closing paren of
6131 ;; the cast. The return value is `cast'. Note that the start
6132 ;; position must be at the first token inside the cast parenthesis
6133 ;; to recognize it.
6134 ;;
6135 ;; PRECEDING-TOKEN-END is the first position after the preceding
6136 ;; token, i.e. on the other side of the syntactic ws from the point.
6137 ;; Use a value less than or equal to (point-min) if the point is at
6138 ;; the first token in (the visible part of) the buffer.
6139 ;;
6140 ;; CONTEXT is a symbol that describes the context at the point:
022d0cf4 6141 ;; 'decl In a comma-separated declaration context (typically
0386b551
AM
6142 ;; inside a function declaration arglist).
6143 ;; '<> In an angle bracket arglist.
6144 ;; 'arglist Some other type of arglist.
a85fd6da
AM
6145 ;; nil Some other context or unknown context. Includes
6146 ;; within the parens of an if, for, ... construct.
0386b551
AM
6147 ;;
6148 ;; LAST-CAST-END is the first token after the closing paren of a
6149 ;; preceding cast, or nil if none is known. If
6150 ;; `c-forward-decl-or-cast-1' is used in succession, it should be
6151 ;; the position after the closest preceding call where a cast was
6152 ;; matched. In that case it's used to discover chains of casts like
6153 ;; "(a) (b) c".
6154 ;;
6155 ;; This function records identifier ranges on
6156 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
6157 ;; `c-record-type-identifiers' is non-nil.
6158 ;;
6159 ;; This function might do hidden buffer changes.
6160
6161 (let (;; `start-pos' is used below to point to the start of the
6162 ;; first type, i.e. after any leading specifiers. It might
6163 ;; also point at the beginning of the preceding syntactic
6164 ;; whitespace.
6165 (start-pos (point))
6166 ;; Set to the result of `c-forward-type'.
6167 at-type
6168 ;; The position of the first token in what we currently
6169 ;; believe is the type in the declaration or cast, after any
6170 ;; specifiers and their associated clauses.
6171 type-start
6172 ;; The position of the first token in what we currently
6173 ;; believe is the declarator for the first identifier. Set
6174 ;; when the type is found, and moved forward over any
6175 ;; `c-decl-hangon-kwds' and their associated clauses that
6176 ;; occurs after the type.
6177 id-start
6178 ;; These store `at-type', `type-start' and `id-start' of the
6179 ;; identifier before the one in those variables. The previous
6180 ;; identifier might turn out to be the real type in a
6181 ;; declaration if the last one has to be the declarator in it.
6182 ;; If `backup-at-type' is nil then the other variables have
6183 ;; undefined values.
6184 backup-at-type backup-type-start backup-id-start
6185 ;; Set if we've found a specifier that makes the defined
6186 ;; identifier(s) types.
6187 at-type-decl
6188 ;; Set if we've found a specifier that can start a declaration
6189 ;; where there's no type.
6190 maybe-typeless
6191 ;; If a specifier is found that also can be a type prefix,
6192 ;; these flags are set instead of those above. If we need to
6193 ;; back up an identifier, they are copied to the real flag
6194 ;; variables. Thus they only take effect if we fail to
6195 ;; interpret it as a type.
6196 backup-at-type-decl backup-maybe-typeless
6197 ;; Whether we've found a declaration or a cast. We might know
6198 ;; this before we've found the type in it. It's 'ids if we've
6199 ;; found two consecutive identifiers (usually a sure sign, but
6200 ;; we should allow that in labels too), and t if we've found a
6201 ;; specifier keyword (a 100% sure sign).
6202 at-decl-or-cast
6203 ;; Set when we need to back up to parse this as a declaration
6204 ;; but not as a cast.
6205 backup-if-not-cast
6206 ;; For casts, the return position.
6207 cast-end
6208 ;; Save `c-record-type-identifiers' and
6209 ;; `c-record-ref-identifiers' since ranges are recorded
6210 ;; speculatively and should be thrown away if it turns out
6211 ;; that it isn't a declaration or cast.
6212 (save-rec-type-ids c-record-type-identifiers)
6213 (save-rec-ref-ids c-record-ref-identifiers))
6214
6215 ;; Check for a type. Unknown symbols are treated as possible
6216 ;; types, but they could also be specifiers disguised through
6217 ;; macros like __INLINE__, so we recognize both types and known
6218 ;; specifiers after them too.
6219 (while
6220 (let* ((start (point)) kwd-sym kwd-clause-end found-type)
6221
6222 ;; Look for a specifier keyword clause.
6223 (when (looking-at c-prefix-spec-kwds-re)
6224 (setq kwd-sym (c-keyword-sym (match-string 1)))
6225 (save-excursion
6226 (c-forward-keyword-clause 1)
6227 (setq kwd-clause-end (point))))
6228
6229 (when (setq found-type (c-forward-type))
6230 ;; Found a known or possible type or a prefix of a known type.
6231
6232 (when at-type
6233 ;; Got two identifiers with nothing but whitespace
6234 ;; between them. That can only happen in declarations.
6235 (setq at-decl-or-cast 'ids)
6236
6237 (when (eq at-type 'found)
6238 ;; If the previous identifier is a found type we
6239 ;; record it as a real one; it might be some sort of
6240 ;; alias for a prefix like "unsigned".
6241 (save-excursion
6242 (goto-char type-start)
6243 (let ((c-promote-possible-types t))
6244 (c-forward-type)))))
6245
6246 (setq backup-at-type at-type
6247 backup-type-start type-start
6248 backup-id-start id-start
6249 at-type found-type
6250 type-start start
6251 id-start (point)
6252 ;; The previous ambiguous specifier/type turned out
6253 ;; to be a type since we've parsed another one after
6254 ;; it, so clear these backup flags.
6255 backup-at-type-decl nil
6256 backup-maybe-typeless nil))
6257
6258 (if kwd-sym
6259 (progn
6260 ;; Handle known specifier keywords and
6261 ;; `c-decl-hangon-kwds' which can occur after known
6262 ;; types.
6263
6264 (if (c-keyword-member kwd-sym 'c-decl-hangon-kwds)
6265 ;; It's a hang-on keyword that can occur anywhere.
6266 (progn
6267 (setq at-decl-or-cast t)
6268 (if at-type
6269 ;; Move the identifier start position if
6270 ;; we've passed a type.
6271 (setq id-start kwd-clause-end)
6272 ;; Otherwise treat this as a specifier and
6273 ;; move the fallback position.
6274 (setq start-pos kwd-clause-end))
6275 (goto-char kwd-clause-end))
6276
6277 ;; It's an ordinary specifier so we know that
6278 ;; anything before this can't be the type.
6279 (setq backup-at-type nil
6280 start-pos kwd-clause-end)
6281
6282 (if found-type
6283 ;; It's ambiguous whether this keyword is a
6284 ;; specifier or a type prefix, so set the backup
6285 ;; flags. (It's assumed that `c-forward-type'
6286 ;; moved further than `c-forward-keyword-clause'.)
6287 (progn
6288 (when (c-keyword-member kwd-sym 'c-typedef-decl-kwds)
6289 (setq backup-at-type-decl t))
6290 (when (c-keyword-member kwd-sym 'c-typeless-decl-kwds)
6291 (setq backup-maybe-typeless t)))
6292
6293 (when (c-keyword-member kwd-sym 'c-typedef-decl-kwds)
6294 (setq at-type-decl t))
6295 (when (c-keyword-member kwd-sym 'c-typeless-decl-kwds)
6296 (setq maybe-typeless t))
6297
6298 ;; Haven't matched a type so it's an umambiguous
6299 ;; specifier keyword and we know we're in a
6300 ;; declaration.
6301 (setq at-decl-or-cast t)
6302
6303 (goto-char kwd-clause-end))))
6304
6305 ;; If the type isn't known we continue so that we'll jump
6306 ;; over all specifiers and type identifiers. The reason
6307 ;; to do this for a known type prefix is to make things
6308 ;; like "unsigned INT16" work.
6309 (and found-type (not (eq found-type t))))))
6310
6311 (cond
6312 ((eq at-type t)
6313 ;; If a known type was found, we still need to skip over any
6314 ;; hangon keyword clauses after it. Otherwise it has already
6315 ;; been done in the loop above.
6316 (while (looking-at c-decl-hangon-key)
6317 (c-forward-keyword-clause 1))
6318 (setq id-start (point)))
6319
6320 ((eq at-type 'prefix)
6321 ;; A prefix type is itself a primitive type when it's not
6322 ;; followed by another type.
6323 (setq at-type t))
6324
6325 ((not at-type)
6326 ;; Got no type but set things up to continue anyway to handle
6327 ;; the various cases when a declaration doesn't start with a
6328 ;; type.
6329 (setq id-start start-pos))
6330
6331 ((and (eq at-type 'maybe)
6332 (c-major-mode-is 'c++-mode))
6333 ;; If it's C++ then check if the last "type" ends on the form
6334 ;; "foo::foo" or "foo::~foo", i.e. if it's the name of a
6335 ;; (con|de)structor.
6336 (save-excursion
6337 (let (name end-2 end-1)
6338 (goto-char id-start)
6339 (c-backward-syntactic-ws)
6340 (setq end-2 (point))
6341 (when (and
6342 (c-simple-skip-symbol-backward)
6343 (progn
6344 (setq name
6345 (buffer-substring-no-properties (point) end-2))
6346 ;; Cheating in the handling of syntactic ws below.
6347 (< (skip-chars-backward ":~ \t\n\r\v\f") 0))
6348 (progn
6349 (setq end-1 (point))
6350 (c-simple-skip-symbol-backward))
6351 (>= (point) type-start)
6352 (equal (buffer-substring-no-properties (point) end-1)
6353 name))
6354 ;; It is a (con|de)structor name. In that case the
6355 ;; declaration is typeless so zap out any preceding
6356 ;; identifier(s) that we might have taken as types.
6357 (goto-char type-start)
6358 (setq at-type nil
6359 backup-at-type nil
6360 id-start type-start))))))
6361
6362 ;; Check for and step over a type decl expression after the thing
6363 ;; that is or might be a type. This can't be skipped since we
6364 ;; need the correct end position of the declarator for
6365 ;; `max-type-decl-end-*'.
6366 (let ((start (point)) (paren-depth 0) pos
6367 ;; True if there's a non-open-paren match of
6368 ;; `c-type-decl-prefix-key'.
6369 got-prefix
6370 ;; True if the declarator is surrounded by a parenthesis pair.
6371 got-parens
6372 ;; True if there is an identifier in the declarator.
6373 got-identifier
6374 ;; True if there's a non-close-paren match of
6375 ;; `c-type-decl-suffix-key'.
6376 got-suffix
6377 ;; True if there's a prefix match outside the outermost
6378 ;; paren pair that surrounds the declarator.
6379 got-prefix-before-parens
1379f2c5 6380 ;; True if there's a suffix match outside the outermost
0386b551
AM
6381 ;; paren pair that surrounds the declarator. The value is
6382 ;; the position of the first suffix match.
6383 got-suffix-after-parens
6384 ;; True if we've parsed the type decl to a token that is
6385 ;; known to end declarations in this context.
6386 at-decl-end
6387 ;; The earlier values of `at-type' and `type-start' if we've
6388 ;; shifted the type backwards.
6389 identifier-type identifier-start
6390 ;; If `c-parse-and-markup-<>-arglists' is set we need to
6391 ;; turn it off during the name skipping below to avoid
6392 ;; getting `c-type' properties that might be bogus. That
6393 ;; can happen since we don't know if
6394 ;; `c-restricted-<>-arglists' will be correct inside the
6395 ;; arglist paren that gets entered.
6396 c-parse-and-markup-<>-arglists)
6397
6398 (goto-char id-start)
6399
6400 ;; Skip over type decl prefix operators. (Note similar code in
6401 ;; `c-font-lock-declarators'.)
6402 (while (and (looking-at c-type-decl-prefix-key)
6403 (if (and (c-major-mode-is 'c++-mode)
6404 (match-beginning 2))
6405 ;; If the second submatch matches in C++ then
6406 ;; we're looking at an identifier that's a
6407 ;; prefix only if it specifies a member pointer.
6408 (when (setq got-identifier (c-forward-name))
6409 (if (looking-at "\\(::\\)")
6410 ;; We only check for a trailing "::" and
6411 ;; let the "*" that should follow be
6412 ;; matched in the next round.
6413 (progn (setq got-identifier nil) t)
6414 ;; It turned out to be the real identifier,
6415 ;; so stop.
6416 nil))
6417 t))
6418
6419 (if (eq (char-after) ?\()
6420 (progn
6421 (setq paren-depth (1+ paren-depth))
6422 (forward-char))
6423 (unless got-prefix-before-parens
6424 (setq got-prefix-before-parens (= paren-depth 0)))
6425 (setq got-prefix t)
6426 (goto-char (match-end 1)))
6427 (c-forward-syntactic-ws))
6428
6429 (setq got-parens (> paren-depth 0))
6430
6431 ;; Skip over an identifier.
6432 (or got-identifier
6433 (and (looking-at c-identifier-start)
6434 (setq got-identifier (c-forward-name))))
6435
6436 ;; Skip over type decl suffix operators.
6437 (while (if (looking-at c-type-decl-suffix-key)
6438
6439 (if (eq (char-after) ?\))
6440 (when (> paren-depth 0)
6441 (setq paren-depth (1- paren-depth))
6442 (forward-char)
6443 t)
6444 (when (if (save-match-data (looking-at "\\s\("))
6445 (c-safe (c-forward-sexp 1) t)
6446 (goto-char (match-end 1))
6447 t)
6448 (when (and (not got-suffix-after-parens)
6449 (= paren-depth 0))
6450 (setq got-suffix-after-parens (match-beginning 0)))
6451 (setq got-suffix t)))
6452
6453 ;; No suffix matched. We might have matched the
6454 ;; identifier as a type and the open paren of a
6455 ;; function arglist as a type decl prefix. In that
6456 ;; case we should "backtrack": Reinterpret the last
6457 ;; type as the identifier, move out of the arglist and
6458 ;; continue searching for suffix operators.
6459 ;;
6460 ;; Do this even if there's no preceding type, to cope
6461 ;; with old style function declarations in K&R C,
6462 ;; (con|de)structors in C++ and `c-typeless-decl-kwds'
6463 ;; style declarations. That isn't applicable in an
6464 ;; arglist context, though.
6465 (when (and (= paren-depth 1)
6466 (not got-prefix-before-parens)
6467 (not (eq at-type t))
6468 (or backup-at-type
6469 maybe-typeless
6470 backup-maybe-typeless
6471 (when c-recognize-typeless-decls
6472 (not context)))
6473 (setq pos (c-up-list-forward (point)))
6474 (eq (char-before pos) ?\)))
6475 (c-fdoc-shift-type-backward)
6476 (goto-char pos)
6477 t))
6478
6479 (c-forward-syntactic-ws))
6480
6481 (when (and (or maybe-typeless backup-maybe-typeless)
6482 (not got-identifier)
6483 (not got-prefix)
6484 at-type)
6485 ;; Have found no identifier but `c-typeless-decl-kwds' has
6486 ;; matched so we know we're inside a declaration. The
6487 ;; preceding type must be the identifier instead.
6488 (c-fdoc-shift-type-backward))
6489
6490 (setq
6491 at-decl-or-cast
6492 (catch 'at-decl-or-cast
6493
a85fd6da 6494 ;; CASE 1
0386b551
AM
6495 (when (> paren-depth 0)
6496 ;; Encountered something inside parens that isn't matched by
6497 ;; the `c-type-decl-*' regexps, so it's not a type decl
6498 ;; expression. Try to skip out to the same paren depth to
6499 ;; not confuse the cast check below.
6500 (c-safe (goto-char (scan-lists (point) 1 paren-depth)))
6501 ;; If we've found a specifier keyword then it's a
6502 ;; declaration regardless.
6503 (throw 'at-decl-or-cast (eq at-decl-or-cast t)))
6504
6505 (setq at-decl-end
6506 (looking-at (cond ((eq context '<>) "[,>]")
6507 (context "[,\)]")
6508 (t "[,;]"))))
6509
6510 ;; Now we've collected info about various characteristics of
6511 ;; the construct we're looking at. Below follows a decision
6512 ;; tree based on that. It's ordered to check more certain
6513 ;; signs before less certain ones.
6514
6515 (if got-identifier
6516 (progn
6517
a85fd6da 6518 ;; CASE 2
0386b551
AM
6519 (when (and (or at-type maybe-typeless)
6520 (not (or got-prefix got-parens)))
6521 ;; Got another identifier directly after the type, so it's a
6522 ;; declaration.
6523 (throw 'at-decl-or-cast t))
6524
6525 (when (and got-parens
6526 (not got-prefix)
6527 (not got-suffix-after-parens)
6528 (or backup-at-type
6529 maybe-typeless
6530 backup-maybe-typeless))
6531 ;; Got a declaration of the form "foo bar (gnu);" where we've
6532 ;; recognized "bar" as the type and "gnu" as the declarator.
6533 ;; In this case it's however more likely that "bar" is the
6534 ;; declarator and "gnu" a function argument or initializer (if
6535 ;; `c-recognize-paren-inits' is set), since the parens around
6536 ;; "gnu" would be superfluous if it's a declarator. Shift the
6537 ;; type one step backward.
6538 (c-fdoc-shift-type-backward)))
6539
6540 ;; Found no identifier.
6541
6542 (if backup-at-type
6543 (progn
6544
a85fd6da 6545 ;; CASE 3
0386b551
AM
6546 (when (= (point) start)
6547 ;; Got a plain list of identifiers. If a colon follows it's
6548 ;; a valid label. Otherwise the last one probably is the
6549 ;; declared identifier and we should back up to the previous
6550 ;; type, providing it isn't a cast.
6551 (if (eq (char-after) ?:)
6552 ;; If we've found a specifier keyword then it's a
6553 ;; declaration regardless.
6554 (throw 'at-decl-or-cast (eq at-decl-or-cast t))
6555 (setq backup-if-not-cast t)
6556 (throw 'at-decl-or-cast t)))
6557
a85fd6da 6558 ;; CASE 4
0386b551
AM
6559 (when (and got-suffix
6560 (not got-prefix)
6561 (not got-parens))
6562 ;; Got a plain list of identifiers followed by some suffix.
6563 ;; If this isn't a cast then the last identifier probably is
6564 ;; the declared one and we should back up to the previous
6565 ;; type.
6566 (setq backup-if-not-cast t)
6567 (throw 'at-decl-or-cast t)))
6568
a85fd6da 6569 ;; CASE 5
0386b551
AM
6570 (when (eq at-type t)
6571 ;; If the type is known we know that there can't be any
6572 ;; identifier somewhere else, and it's only in declarations in
6573 ;; e.g. function prototypes and in casts that the identifier may
6574 ;; be left out.
6575 (throw 'at-decl-or-cast t))
6576
6577 (when (= (point) start)
6578 ;; Only got a single identifier (parsed as a type so far).
a85fd6da 6579 ;; CASE 6
0386b551
AM
6580 (if (and
6581 ;; Check that the identifier isn't at the start of an
6582 ;; expression.
6583 at-decl-end
6584 (cond
6585 ((eq context 'decl)
6586 ;; Inside an arglist that contains declarations. If K&R
6587 ;; style declarations and parenthesis style initializers
6588 ;; aren't allowed then the single identifier must be a
6589 ;; type, else we require that it's known or found
6590 ;; (primitive types are handled above).
6591 (or (and (not c-recognize-knr-p)
6592 (not c-recognize-paren-inits))
6593 (memq at-type '(known found))))
6594 ((eq context '<>)
6595 ;; Inside a template arglist. Accept known and found
6596 ;; types; other identifiers could just as well be
6597 ;; constants in C++.
6598 (memq at-type '(known found)))))
6599 (throw 'at-decl-or-cast t)
a85fd6da 6600 ;; CASE 7
0386b551
AM
6601 ;; Can't be a valid declaration or cast, but if we've found a
6602 ;; specifier it can't be anything else either, so treat it as
6603 ;; an invalid/unfinished declaration or cast.
6604 (throw 'at-decl-or-cast at-decl-or-cast))))
6605
6606 (if (and got-parens
6607 (not got-prefix)
6608 (not context)
6609 (not (eq at-type t))
6610 (or backup-at-type
6611 maybe-typeless
6612 backup-maybe-typeless
6613 (when c-recognize-typeless-decls
6614 (or (not got-suffix)
6615 (not (looking-at
6616 c-after-suffixed-type-maybe-decl-key))))))
6617 ;; Got an empty paren pair and a preceding type that probably
6618 ;; really is the identifier. Shift the type backwards to make
6619 ;; the last one the identifier. This is analogous to the
6620 ;; "backtracking" done inside the `c-type-decl-suffix-key' loop
6621 ;; above.
6622 ;;
6623 ;; Exception: In addition to the conditions in that
6624 ;; "backtracking" code, do not shift backward if we're not
6625 ;; looking at either `c-after-suffixed-type-decl-key' or "[;,]".
6626 ;; Since there's no preceding type, the shift would mean that
6627 ;; the declaration is typeless. But if the regexp doesn't match
6628 ;; then we will simply fall through in the tests below and not
6629 ;; recognize it at all, so it's better to try it as an abstract
6630 ;; declarator instead.
6631 (c-fdoc-shift-type-backward)
6632
6633 ;; Still no identifier.
a85fd6da 6634 ;; CASE 8
0386b551
AM
6635 (when (and got-prefix (or got-parens got-suffix))
6636 ;; Require `got-prefix' together with either `got-parens' or
6637 ;; `got-suffix' to recognize it as an abstract declarator:
6638 ;; `got-parens' only is probably an empty function call.
6639 ;; `got-suffix' only can build an ordinary expression together
6640 ;; with the preceding identifier which we've taken as a type.
6641 ;; We could actually accept on `got-prefix' only, but that can
6642 ;; easily occur temporarily while writing an expression so we
6643 ;; avoid that case anyway. We could do a better job if we knew
6644 ;; the point when the fontification was invoked.
6645 (throw 'at-decl-or-cast t))
6646
a85fd6da 6647 ;; CASE 9
0386b551
AM
6648 (when (and at-type
6649 (not got-prefix)
6650 (not got-parens)
6651 got-suffix-after-parens
6652 (eq (char-after got-suffix-after-parens) ?\())
6653 ;; Got a type, no declarator but a paren suffix. I.e. it's a
6654 ;; normal function call afterall (or perhaps a C++ style object
6655 ;; instantiation expression).
6656 (throw 'at-decl-or-cast nil))))
6657
a85fd6da 6658 ;; CASE 10
0386b551
AM
6659 (when at-decl-or-cast
6660 ;; By now we've located the type in the declaration that we know
6661 ;; we're in.
6662 (throw 'at-decl-or-cast t))
6663
a85fd6da 6664 ;; CASE 11
0386b551
AM
6665 (when (and got-identifier
6666 (not context)
6667 (looking-at c-after-suffixed-type-decl-key)
6668 (if (and got-parens
6669 (not got-prefix)
6670 (not got-suffix)
6671 (not (eq at-type t)))
6672 ;; Shift the type backward in the case that there's a
6673 ;; single identifier inside parens. That can only
6674 ;; occur in K&R style function declarations so it's
6675 ;; more likely that it really is a function call.
6676 ;; Therefore we only do this after
6677 ;; `c-after-suffixed-type-decl-key' has matched.
6678 (progn (c-fdoc-shift-type-backward) t)
6679 got-suffix-after-parens))
6680 ;; A declaration according to `c-after-suffixed-type-decl-key'.
6681 (throw 'at-decl-or-cast t))
6682
a85fd6da 6683 ;; CASE 12
0386b551
AM
6684 (when (and (or got-prefix (not got-parens))
6685 (memq at-type '(t known)))
6686 ;; It's a declaration if a known type precedes it and it can't be a
6687 ;; function call.
6688 (throw 'at-decl-or-cast t))
6689
6690 ;; If we get here we can't tell if this is a type decl or a normal
6691 ;; expression by looking at it alone. (That's under the assumption
6692 ;; that normal expressions always can look like type decl expressions,
6693 ;; which isn't really true but the cases where it doesn't hold are so
6694 ;; uncommon (e.g. some placements of "const" in C++) it's not worth
6695 ;; the effort to look for them.)
6696
6697 (unless (or at-decl-end (looking-at "=[^=]"))
6698 ;; If this is a declaration it should end here or its initializer(*)
6699 ;; should start here, so check for allowed separation tokens. Note
6700 ;; that this rule doesn't work e.g. with a K&R arglist after a
6701 ;; function header.
6702 ;;
6703 ;; *) Don't check for C++ style initializers using parens
6704 ;; since those already have been matched as suffixes.
6705 ;;
6706 ;; If `at-decl-or-cast' is then we've found some other sign that
6707 ;; it's a declaration or cast, so then it's probably an
6708 ;; invalid/unfinished one.
6709 (throw 'at-decl-or-cast at-decl-or-cast))
6710
6711 ;; Below are tests that only should be applied when we're certain to
6712 ;; not have parsed halfway through an expression.
6713
a85fd6da 6714 ;; CASE 14
0386b551
AM
6715 (when (memq at-type '(t known))
6716 ;; The expression starts with a known type so treat it as a
6717 ;; declaration.
6718 (throw 'at-decl-or-cast t))
6719
a85fd6da 6720 ;; CASE 15
0386b551
AM
6721 (when (and (c-major-mode-is 'c++-mode)
6722 ;; In C++ we check if the identifier is a known type, since
6723 ;; (con|de)structors use the class name as identifier.
6724 ;; We've always shifted over the identifier as a type and
6725 ;; then backed up again in this case.
6726 identifier-type
6727 (or (memq identifier-type '(found known))
6728 (and (eq (char-after identifier-start) ?~)
6729 ;; `at-type' probably won't be 'found for
6730 ;; destructors since the "~" is then part of the
6731 ;; type name being checked against the list of
6732 ;; known types, so do a check without that
6733 ;; operator.
6734 (or (save-excursion
6735 (goto-char (1+ identifier-start))
6736 (c-forward-syntactic-ws)
6737 (c-with-syntax-table
6738 c-identifier-syntax-table
6739 (looking-at c-known-type-key)))
6740 (save-excursion
6741 (goto-char (1+ identifier-start))
6742 ;; We have already parsed the type earlier,
6743 ;; so it'd be possible to cache the end
6744 ;; position instead of redoing it here, but
6745 ;; then we'd need to keep track of another
6746 ;; position everywhere.
6747 (c-check-type (point)
6748 (progn (c-forward-type)
6749 (point))))))))
6750 (throw 'at-decl-or-cast t))
6751
6752 (if got-identifier
6753 (progn
a85fd6da 6754 ;; CASE 16
0386b551
AM
6755 (when (and got-prefix-before-parens
6756 at-type
6757 (or at-decl-end (looking-at "=[^=]"))
6758 (not context)
6759 (not got-suffix))
6760 ;; Got something like "foo * bar;". Since we're not inside an
6761 ;; arglist it would be a meaningless expression because the
6762 ;; result isn't used. We therefore choose to recognize it as
6763 ;; a declaration. Do not allow a suffix since it could then
6764 ;; be a function call.
6765 (throw 'at-decl-or-cast t))
6766
a85fd6da 6767 ;; CASE 17
0386b551
AM
6768 (when (and (or got-suffix-after-parens
6769 (looking-at "=[^=]"))
6770 (eq at-type 'found)
6771 (not (eq context 'arglist)))
6772 ;; Got something like "a (*b) (c);" or "a (b) = c;". It could
6773 ;; be an odd expression or it could be a declaration. Treat
6774 ;; it as a declaration if "a" has been used as a type
6775 ;; somewhere else (if it's a known type we won't get here).
6776 (throw 'at-decl-or-cast t)))
6777
a85fd6da 6778 ;; CASE 18
0386b551
AM
6779 (when (and context
6780 (or got-prefix
6781 (and (eq context 'decl)
6782 (not c-recognize-paren-inits)
6783 (or got-parens got-suffix))))
6784 ;; Got a type followed by an abstract declarator. If `got-prefix'
6785 ;; is set it's something like "a *" without anything after it. If
6786 ;; `got-parens' or `got-suffix' is set it's "a()", "a[]", "a()[]",
6787 ;; or similar, which we accept only if the context rules out
6788 ;; expressions.
6789 (throw 'at-decl-or-cast t)))
6790
6791 ;; If we had a complete symbol table here (which rules out
6792 ;; `c-found-types') we should return t due to the disambiguation rule
6793 ;; (in at least C++) that anything that can be parsed as a declaration
6794 ;; is a declaration. Now we're being more defensive and prefer to
6795 ;; highlight things like "foo (bar);" as a declaration only if we're
6796 ;; inside an arglist that contains declarations.
6797 (eq context 'decl))))
6798
6799 ;; The point is now after the type decl expression.
6800
6801 (cond
6802 ;; Check for a cast.
6803 ((save-excursion
6804 (and
6805 c-cast-parens
6806
6807 ;; Should be the first type/identifier in a cast paren.
6808 (> preceding-token-end (point-min))
6809 (memq (char-before preceding-token-end) c-cast-parens)
6810
6811 ;; The closing paren should follow.
6812 (progn
6813 (c-forward-syntactic-ws)
6814 (looking-at "\\s\)"))
6815
6816 ;; There should be a primary expression after it.
6817 (let (pos)
6818 (forward-char)
6819 (c-forward-syntactic-ws)
6820 (setq cast-end (point))
6821 (and (looking-at c-primary-expr-regexp)
6822 (progn
6823 (setq pos (match-end 0))
6824 (or
6825 ;; Check if the expression begins with a prefix keyword.
6826 (match-beginning 2)
6827 (if (match-beginning 1)
6828 ;; Expression begins with an ambiguous operator. Treat
6829 ;; it as a cast if it's a type decl or if we've
6830 ;; recognized the type somewhere else.
6831 (or at-decl-or-cast
6832 (memq at-type '(t known found)))
6833 ;; Unless it's a keyword, it's the beginning of a primary
6834 ;; expression.
6835 (not (looking-at c-keywords-regexp)))))
6836 ;; If `c-primary-expr-regexp' matched a nonsymbol token, check
6837 ;; that it matched a whole one so that we don't e.g. confuse
6838 ;; the operator '-' with '->'. It's ok if it matches further,
6839 ;; though, since it e.g. can match the float '.5' while the
6840 ;; operator regexp only matches '.'.
6841 (or (not (looking-at c-nonsymbol-token-regexp))
6842 (<= (match-end 0) pos))))
6843
6844 ;; There should either be a cast before it or something that isn't an
6845 ;; identifier or close paren.
6846 (> preceding-token-end (point-min))
6847 (progn
6848 (goto-char (1- preceding-token-end))
6849 (or (eq (point) last-cast-end)
6850 (progn
6851 (c-backward-syntactic-ws)
6852 (if (< (skip-syntax-backward "w_") 0)
6853 ;; It's a symbol. Accept it only if it's one of the
6854 ;; keywords that can precede an expression (without
6855 ;; surrounding parens).
6856 (looking-at c-simple-stmt-key)
6857 (and
6858 ;; Check that it isn't a close paren (block close is ok,
6859 ;; though).
6860 (not (memq (char-before) '(?\) ?\])))
6861 ;; Check that it isn't a nonsymbol identifier.
6862 (not (c-on-identifier)))))))))
6863
6864 ;; Handle the cast.
6865 (when (and c-record-type-identifiers at-type (not (eq at-type t)))
6866 (let ((c-promote-possible-types t))
6867 (goto-char type-start)
6868 (c-forward-type)))
6869
6870 (goto-char cast-end)
6871 'cast)
6872
6873 (at-decl-or-cast
6874 ;; We're at a declaration. Highlight the type and the following
6875 ;; declarators.
6876
6877 (when backup-if-not-cast
6878 (c-fdoc-shift-type-backward t))
6879
6880 (when (and (eq context 'decl) (looking-at ","))
6881 ;; Make sure to propagate the `c-decl-arg-start' property to
6882 ;; the next argument if it's set in this one, to cope with
6883 ;; interactive refontification.
6884 (c-put-c-type-property (point) 'c-decl-arg-start))
6885
6886 (when (and c-record-type-identifiers at-type (not (eq at-type t)))
6887 (let ((c-promote-possible-types t))
6888 (save-excursion
6889 (goto-char type-start)
6890 (c-forward-type))))
6891
6892 (cons id-start at-type-decl))
6893
6894 (t
6895 ;; False alarm. Restore the recorded ranges.
6896 (setq c-record-type-identifiers save-rec-type-ids
6897 c-record-ref-identifiers save-rec-ref-ids)
6898 nil))))
6899
6900(defun c-forward-label (&optional assume-markup preceding-token-end limit)
51c9af45 6901 ;; Assuming that point is at the beginning of a token, check if it starts a
1379f2c5
AM
6902 ;; label and if so move over it and return non-nil (t in default situations,
6903 ;; specific symbols (see below) for interesting situations), otherwise don't
6904 ;; move and return nil. "Label" here means "most things with a colon".
51c9af45
AM
6905 ;;
6906 ;; More precisely, a "label" is regarded as one of:
1379f2c5
AM
6907 ;; (i) a goto target like "foo:" - returns the symbol `goto-target';
6908 ;; (ii) A case label - either the entire construct "case FOO:", or just the
6909 ;; bare "case", should the colon be missing. We return t;
6910 ;; (iii) a keyword which needs a colon, like "default:" or "private:"; We
6911 ;; return t;
51c9af45 6912 ;; (iv) One of QT's "extended" C++ variants of
1379f2c5
AM
6913 ;; "private:"/"protected:"/"public:"/"more:" looking like "public slots:".
6914 ;; Returns the symbol `qt-2kwds-colon'.
6915 ;; (v) QT's construct "signals:". Returns the symbol `qt-1kwd-colon'.
82ba65cf 6916 ;; (vi) One of the keywords matched by `c-opt-extra-label-key' (without any
51c9af45 6917 ;; colon). Currently (2006-03), this applies only to Objective C's
1379f2c5 6918 ;; keywords "@private", "@protected", and "@public". Returns t.
51c9af45
AM
6919 ;;
6920 ;; One of the things which will NOT be recognised as a label is a bit-field
6921 ;; element of a struct, something like "int foo:5".
6922 ;;
6923 ;; The end of the label is taken to be just after the colon, or the end of
6924 ;; the first submatch in `c-opt-extra-label-key'. The point is directly
6925 ;; after the end on return. The terminating char gets marked with
6926 ;; `c-decl-end' to improve recognition of the following declaration or
6927 ;; statement.
0386b551
AM
6928 ;;
6929 ;; If ASSUME-MARKUP is non-nil, it's assumed that the preceding
51c9af45 6930 ;; label, if any, has already been marked up like that.
0386b551
AM
6931 ;;
6932 ;; If PRECEDING-TOKEN-END is given, it should be the first position
6933 ;; after the preceding token, i.e. on the other side of the
6934 ;; syntactic ws from the point. Use a value less than or equal to
6935 ;; (point-min) if the point is at the first token in (the visible
6936 ;; part of) the buffer.
6937 ;;
6938 ;; The optional LIMIT limits the forward scan for the colon.
6939 ;;
6940 ;; This function records the ranges of the label symbols on
6941 ;; `c-record-ref-identifiers' if `c-record-type-identifiers' (!) is
6942 ;; non-nil.
6943 ;;
6944 ;; This function might do hidden buffer changes.
6945
51c9af45 6946 (let ((start (point))
1379f2c5 6947 label-end
51c9af45 6948 qt-symbol-idx
1379f2c5 6949 macro-start ; if we're in one.
f412a567
AM
6950 label-type
6951 kwd)
0386b551 6952 (cond
b414f371 6953 ;; "case" or "default" (Doesn't apply to AWK).
0386b551
AM
6954 ((looking-at c-label-kwds-regexp)
6955 (let ((kwd-end (match-end 1)))
6956 ;; Record only the keyword itself for fontification, since in
6957 ;; case labels the following is a constant expression and not
6958 ;; a label.
6959 (when c-record-type-identifiers
6960 (c-record-ref-id (cons (match-beginning 1) kwd-end)))
6961
6962 ;; Find the label end.
6963 (goto-char kwd-end)
1379f2c5
AM
6964 (setq label-type
6965 (if (and (c-syntactic-re-search-forward
6966 ;; Stop on chars that aren't allowed in expressions,
6967 ;; and on operator chars that would be meaningless
6968 ;; there. FIXME: This doesn't cope with ?: operators.
6969 "[;{=,@]\\|\\(\\=\\|[^:]\\):\\([^:]\\|\\'\\)"
6970 limit t t nil 1)
6971 (match-beginning 2))
6972
6973 (progn ; there's a proper :
6974 (goto-char (match-beginning 2)) ; just after the :
6975 (c-put-c-type-property (1- (point)) 'c-decl-end)
6976 t)
6977
6978 ;; It's an unfinished label. We consider the keyword enough
6979 ;; to recognize it as a label, so that it gets fontified.
6980 ;; Leave the point at the end of it, but don't put any
6981 ;; `c-decl-end' marker.
6982 (goto-char kwd-end)
6983 t))))
0386b551 6984
51c9af45 6985 ;; @private, @protected, @public, in Objective C, or similar.
0386b551
AM
6986 ((and c-opt-extra-label-key
6987 (looking-at c-opt-extra-label-key))
6988 ;; For a `c-opt-extra-label-key' match, we record the whole
6989 ;; thing for fontification. That's to get the leading '@' in
6990 ;; Objective-C protection labels fontified.
6991 (goto-char (match-end 1))
6992 (when c-record-type-identifiers
6993 (c-record-ref-id (cons (match-beginning 1) (point))))
6994 (c-put-c-type-property (1- (point)) 'c-decl-end)
1379f2c5 6995 (setq label-type t))
0386b551 6996
51c9af45
AM
6997 ;; All other cases of labels.
6998 ((and c-recognize-colon-labels ; nil for AWK and IDL, otherwise t.
0386b551
AM
6999
7000 ;; A colon label must have something before the colon.
7001 (not (eq (char-after) ?:))
7002
7003 ;; Check that we're not after a token that can't precede a label.
7004 (or
7005 ;; Trivially succeeds when there's no preceding token.
7006 (if preceding-token-end
7007 (<= preceding-token-end (point-min))
7008 (save-excursion
7009 (c-backward-syntactic-ws)
7010 (setq preceding-token-end (point))
7011 (bobp)))
7012
7013 ;; Check if we're after a label, if we're after a closing
7014 ;; paren that belong to statement, and with
7015 ;; `c-label-prefix-re'. It's done in different order
7016 ;; depending on `assume-markup' since the checks have
7017 ;; different expensiveness.
7018 (if assume-markup
7019 (or
7020 (eq (c-get-char-property (1- preceding-token-end) 'c-type)
7021 'c-decl-end)
7022
7023 (save-excursion
7024 (goto-char (1- preceding-token-end))
7025 (c-beginning-of-current-token)
51c9af45
AM
7026 (or (looking-at c-label-prefix-re)
7027 (looking-at c-block-stmt-1-key)))
0386b551
AM
7028
7029 (and (eq (char-before preceding-token-end) ?\))
7030 (c-after-conditional)))
7031
7032 (or
7033 (save-excursion
7034 (goto-char (1- preceding-token-end))
7035 (c-beginning-of-current-token)
51c9af45
AM
7036 (or (looking-at c-label-prefix-re)
7037 (looking-at c-block-stmt-1-key)))
0386b551
AM
7038
7039 (cond
7040 ((eq (char-before preceding-token-end) ?\))
7041 (c-after-conditional))
7042
7043 ((eq (char-before preceding-token-end) ?:)
7044 ;; Might be after another label, so check it recursively.
51c9af45
AM
7045 (save-restriction
7046 (save-excursion
7047 (goto-char (1- preceding-token-end))
7048 ;; Essentially the same as the
7049 ;; `c-syntactic-re-search-forward' regexp below.
7050 (setq macro-start
7051 (save-excursion (and (c-beginning-of-macro)
7052 (point))))
7053 (if macro-start (narrow-to-region macro-start (point-max)))
7054 (c-syntactic-skip-backward "^-]:?;}=*/%&|,<>!@+" nil t)
7055 ;; Note: the following should work instead of the
7056 ;; narrow-to-region above. Investigate why not,
7057 ;; sometime. ACM, 2006-03-31.
7058 ;; (c-syntactic-skip-backward "^-]:?;}=*/%&|,<>!@+"
7059 ;; macro-start t)
7060 (let ((pte (point))
7061 ;; If the caller turned on recording for us,
7062 ;; it shouldn't apply when we check the
7063 ;; preceding label.
7064 c-record-type-identifiers)
7065 ;; A label can't start at a cpp directive. Check for
7066 ;; this, since c-forward-syntactic-ws would foul up on it.
7067 (unless (and c-opt-cpp-prefix (looking-at c-opt-cpp-prefix))
7068 (c-forward-syntactic-ws)
7069 (c-forward-label nil pte start))))))))))
7070
1379f2c5 7071 ;; Point is still at the beginning of the possible label construct.
b414f371 7072 ;;
51c9af45
AM
7073 ;; Check that the next nonsymbol token is ":", or that we're in one
7074 ;; of QT's "slots" declarations. Allow '(' for the sake of macro
7075 ;; arguments. FIXME: Should build this regexp from the language
7076 ;; constants.
1379f2c5
AM
7077 (cond
7078 ;; public: protected: private:
7079 ((and
7080 (c-major-mode-is 'c++-mode)
7081 (search-forward-regexp
7082 "\\=p\\(r\\(ivate\\|otected\\)\\|ublic\\)\\>[^_]" nil t)
7083 (progn (backward-char)
7084 (c-forward-syntactic-ws limit)
7085 (looking-at ":\\([^:]\\|\\'\\)"))) ; A single colon.
7086 (forward-char)
7087 (setq label-type t))
7088 ;; QT double keyword like "protected slots:" or goto target.
7089 ((progn (goto-char start) nil))
7090 ((when (c-syntactic-re-search-forward
7091 "[ \t\n[:?;{=*/%&|,<>!@+-]" limit t t) ; not at EOB
7092 (backward-char)
7093 (setq label-end (point))
7094 (setq qt-symbol-idx
7095 (and (c-major-mode-is 'c++-mode)
7096 (string-match
7097 "\\(p\\(r\\(ivate\\|otected\\)\\|ublic\\)\\|more\\)\\>"
b414f371 7098 (buffer-substring start (point)))))
1379f2c5
AM
7099 (c-forward-syntactic-ws limit)
7100 (cond
7101 ((looking-at ":\\([^:]\\|\\'\\)") ; A single colon.
7102 (forward-char)
7103 (setq label-type
f412a567
AM
7104 (if (or (string= "signals" ; Special QT macro
7105 (setq kwd (buffer-substring-no-properties start label-end)))
7106 (string= "Q_SIGNALS" kwd))
1379f2c5
AM
7107 'qt-1kwd-colon
7108 'goto-target)))
7109 ((and qt-symbol-idx
f412a567 7110 (search-forward-regexp "\\=\\(slots\\|Q_SLOTS\\)\\>" limit t)
1379f2c5
AM
7111 (progn (c-forward-syntactic-ws limit)
7112 (looking-at ":\\([^:]\\|\\'\\)"))) ; A single colon
7113 (forward-char)
7114 (setq label-type 'qt-2kwds-colon)))))))
0386b551
AM
7115
7116 (save-restriction
7117 (narrow-to-region start (point))
7118
7119 ;; Check that `c-nonlabel-token-key' doesn't match anywhere.
7120 (catch 'check-label
7121 (goto-char start)
7122 (while (progn
7123 (when (looking-at c-nonlabel-token-key)
7124 (goto-char start)
1379f2c5 7125 (setq label-type nil)
0386b551
AM
7126 (throw 'check-label nil))
7127 (and (c-safe (c-forward-sexp)
7128 (c-forward-syntactic-ws)
7129 t)
7130 (not (eobp)))))
7131
7132 ;; Record the identifiers in the label for fontification, unless
7133 ;; it begins with `c-label-kwds' in which case the following
7134 ;; identifiers are part of a (constant) expression that
7135 ;; shouldn't be fontified.
7136 (when (and c-record-type-identifiers
7137 (progn (goto-char start)
7138 (not (looking-at c-label-kwds-regexp))))
7139 (while (c-syntactic-re-search-forward c-symbol-key nil t)
7140 (c-record-ref-id (cons (match-beginning 0)
7141 (match-end 0)))))
7142
7143 (c-put-c-type-property (1- (point-max)) 'c-decl-end)
82ba65cf 7144 (goto-char (point-max)))))
0386b551
AM
7145
7146 (t
7147 ;; Not a label.
1379f2c5
AM
7148 (goto-char start)))
7149 label-type))
0386b551
AM
7150
7151(defun c-forward-objc-directive ()
7152 ;; Assuming the point is at the beginning of a token, try to move
7153 ;; forward to the end of the Objective-C directive that starts
7154 ;; there. Return t if a directive was fully recognized, otherwise
7155 ;; the point is moved as far as one could be successfully parsed and
7156 ;; nil is returned.
7157 ;;
7158 ;; This function records identifier ranges on
7159 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
7160 ;; `c-record-type-identifiers' is non-nil.
7161 ;;
7162 ;; This function might do hidden buffer changes.
7163
7164 (let ((start (point))
7165 start-char
7166 (c-promote-possible-types t)
7167 ;; Turn off recognition of angle bracket arglists while parsing
7168 ;; types here since the protocol reference list might then be
7169 ;; considered part of the preceding name or superclass-name.
7170 c-recognize-<>-arglists)
7171
7172 (if (or
7173 (when (looking-at
7174 (eval-when-compile
7175 (c-make-keywords-re t
7176 (append (c-lang-const c-protection-kwds objc)
7177 '("@end"))
7178 'objc-mode)))
7179 (goto-char (match-end 1))
7180 t)
7181
7182 (and
7183 (looking-at
7184 (eval-when-compile
7185 (c-make-keywords-re t
7186 '("@interface" "@implementation" "@protocol")
7187 'objc-mode)))
7188
7189 ;; Handle the name of the class itself.
7190 (progn
cb694ab7
AM
7191; (c-forward-token-2) ; 2006/1/13 This doesn't move if the token's
7192; at EOB.
7193 (goto-char (match-end 0))
7194 (c-skip-ws-forward)
0386b551
AM
7195 (c-forward-type))
7196
7197 (catch 'break
7198 ;; Look for ": superclass-name" or "( category-name )".
7199 (when (looking-at "[:\(]")
7200 (setq start-char (char-after))
7201 (forward-char)
7202 (c-forward-syntactic-ws)
7203 (unless (c-forward-type) (throw 'break nil))
7204 (when (eq start-char ?\()
7205 (unless (eq (char-after) ?\)) (throw 'break nil))
7206 (forward-char)
7207 (c-forward-syntactic-ws)))
7208
7209 ;; Look for a protocol reference list.
7210 (if (eq (char-after) ?<)
7211 (let ((c-recognize-<>-arglists t)
7212 (c-parse-and-markup-<>-arglists t)
7213 c-restricted-<>-arglists)
7214 (c-forward-<>-arglist t))
7215 t))))
7216
7217 (progn
7218 (c-backward-syntactic-ws)
7219 (c-clear-c-type-property start (1- (point)) 'c-decl-end)
7220 (c-put-c-type-property (1- (point)) 'c-decl-end)
7221 t)
7222
7223 (c-clear-c-type-property start (point) 'c-decl-end)
7224 nil)))
7225
785eecbb
RS
7226(defun c-beginning-of-inheritance-list (&optional lim)
7227 ;; Go to the first non-whitespace after the colon that starts a
7228 ;; multiple inheritance introduction. Optional LIM is the farthest
7229 ;; back we should search.
0386b551
AM
7230 ;;
7231 ;; This function might do hidden buffer changes.
7232 (c-with-syntax-table c++-template-syntax-table
7233 (c-backward-token-2 0 t lim)
7234 (while (and (or (looking-at c-symbol-start)
7235 (looking-at "[<,]\\|::"))
7236 (zerop (c-backward-token-2 1 t lim))))))
785eecbb 7237
785eecbb
RS
7238(defun c-in-method-def-p ()
7239 ;; Return nil if we aren't in a method definition, otherwise the
7240 ;; position of the initial [+-].
0386b551
AM
7241 ;;
7242 ;; This function might do hidden buffer changes.
785eecbb
RS
7243 (save-excursion
7244 (beginning-of-line)
a66cd3ee
MS
7245 (and c-opt-method-key
7246 (looking-at c-opt-method-key)
785eecbb
RS
7247 (point))
7248 ))
7249
a66cd3ee
MS
7250;; Contributed by Kevin Ryde <user42@zip.com.au>.
7251(defun c-in-gcc-asm-p ()
7252 ;; Return non-nil if point is within a gcc \"asm\" block.
7253 ;;
7254 ;; This should be called with point inside an argument list.
7255 ;;
7256 ;; Only one level of enclosing parentheses is considered, so for
7257 ;; instance `nil' is returned when in a function call within an asm
7258 ;; operand.
0386b551
AM
7259 ;;
7260 ;; This function might do hidden buffer changes.
a66cd3ee
MS
7261
7262 (and c-opt-asm-stmt-key
7263 (save-excursion
7264 (beginning-of-line)
7265 (backward-up-list 1)
7266 (c-beginning-of-statement-1 (point-min) nil t)
7267 (looking-at c-opt-asm-stmt-key))))
7268
abb7e5cf 7269(defun c-at-toplevel-p ()
a85fd6da
AM
7270 "Return a determination as to whether point is \"at the top level\".
7271Informally, \"at the top level\" is anywhere where you can write
7272a function.
7273
7274More precisely, being at the top-level means that point is either
7275outside any enclosing block (such as a function definition), or
7276directly inside a class, namespace or other block that contains
7277another declaration level.
abb7e5cf
SM
7278
7279If point is not at the top-level (e.g. it is inside a method
7280definition), then nil is returned. Otherwise, if point is at a
7281top-level not enclosed within a class definition, t is returned.
7282Otherwise, a 2-vector is returned where the zeroth element is the
7283buffer position of the start of the class declaration, and the first
7284element is the buffer position of the enclosing class's opening
0386b551
AM
7285brace.
7286
7287Note that this function might do hidden buffer changes. See the
7288comment at the start of cc-engine.el for more info."
a66cd3ee
MS
7289 (let ((paren-state (c-parse-state)))
7290 (or (not (c-most-enclosing-brace paren-state))
7291 (c-search-uplist-for-classkey paren-state))))
7292
d9e94c22 7293(defun c-just-after-func-arglist-p (&optional lim)
0386b551
AM
7294 ;; Return non-nil if the point is in the region after the argument
7295 ;; list of a function and its opening brace (or semicolon in case it
7296 ;; got no body). If there are K&R style argument declarations in
7297 ;; that region, the point has to be inside the first one for this
7298 ;; function to recognize it.
a66cd3ee 7299 ;;
0386b551
AM
7300 ;; If successful, the point is moved to the first token after the
7301 ;; function header (see `c-forward-decl-or-cast-1' for details) and
7302 ;; the position of the opening paren of the function arglist is
7303 ;; returned.
7304 ;;
7305 ;; The point is clobbered if not successful.
7306 ;;
7307 ;; LIM is used as bound for backward buffer searches.
7308 ;;
7309 ;; This function might do hidden buffer changes.
7310
7311 (let ((beg (point)) end id-start)
7312 (and
7313 (eq (c-beginning-of-statement-1 lim) 'same)
7314
7315 (not (or (c-major-mode-is 'objc-mode)
7316 (c-forward-objc-directive)))
7317
7318 (setq id-start
7319 (car-safe (c-forward-decl-or-cast-1 (c-point 'bosws) nil nil)))
7320 (< id-start beg)
7321
7322 ;; There should not be a '=' or ',' between beg and the
7323 ;; start of the declaration since that means we were in the
7324 ;; "expression part" of the declaration.
7325 (or (> (point) beg)
7326 (not (looking-at "[=,]")))
7327
7328 (save-excursion
7329 ;; Check that there's an arglist paren in the
7330 ;; declaration.
7331 (goto-char id-start)
7332 (cond ((eq (char-after) ?\()
7333 ;; The declarator is a paren expression, so skip past it
7334 ;; so that we don't get stuck on that instead of the
7335 ;; function arglist.
7336 (c-forward-sexp))
51c9af45
AM
7337 ((and c-opt-op-identifier-prefix
7338 (looking-at c-opt-op-identifier-prefix))
0386b551
AM
7339 ;; Don't trip up on "operator ()".
7340 (c-forward-token-2 2 t)))
7341 (and (< (point) beg)
7342 (c-syntactic-re-search-forward "(" beg t t)
7343 (1- (point)))))))
785eecbb 7344
a66cd3ee
MS
7345(defun c-in-knr-argdecl (&optional lim)
7346 ;; Return the position of the first argument declaration if point is
7347 ;; inside a K&R style argument declaration list, nil otherwise.
7348 ;; `c-recognize-knr-p' is not checked. If LIM is non-nil, it's a
7349 ;; position that bounds the backward search for the argument list.
7350 ;;
9cf17ef1
AM
7351 ;; Point must be within a possible K&R region, e.g. just before a top-level
7352 ;; "{". It must be outside of parens and brackets. The test can return
7353 ;; false positives otherwise.
0386b551
AM
7354 ;;
7355 ;; This function might do hidden buffer changes.
d9e94c22 7356
a66cd3ee
MS
7357 (save-excursion
7358 (save-restriction
9cf17ef1
AM
7359 ;; If we're in a macro, our search range is restricted to it. Narrow to
7360 ;; the searchable range.
7361 (let* ((macro-start (c-query-macro-start))
7362 (lim (max (or lim (point-min)) (or macro-start (point-min))))
6b6481ed
AM
7363 before-lparen after-rparen
7364 (pp-count-out 20)) ; Max number of paren/brace constructs before we give up
9cf17ef1
AM
7365 (narrow-to-region lim (c-point 'eol))
7366
7367 ;; Search backwards for the defun's argument list. We give up if we
7368 ;; encounter a "}" (end of a previous defun) or BOB.
7369 ;;
7370 ;; The criterion for a paren structure being the arg list is:
7371 ;; o - there is non-WS stuff after it but before any "{"; AND
7372 ;; o - the token after it isn't a ";" AND
7373 ;; o - it is preceded by either an identifier (the function name) or
7374 ;; a macro expansion like "DEFUN (...)"; AND
7375 ;; o - its content is a non-empty comma-separated list of identifiers
7376 ;; (an empty arg list won't have a knr region).
7377 ;;
7378 ;; The following snippet illustrates these rules:
7379 ;; int foo (bar, baz, yuk)
7380 ;; int bar [] ;
7381 ;; int (*baz) (my_type) ;
7382 ;; int (*) (void) (*yuk) (void) ;
7383 ;; {
7384
7385 (catch 'knr
6b6481ed
AM
7386 (while (> pp-count-out 0) ; go back one paren/bracket pair each time.
7387 (setq pp-count-out (1- pp-count-out))
10489fcb 7388 (c-syntactic-skip-backward "^)]}")
9cf17ef1
AM
7389 (cond ((eq (char-before) ?\))
7390 (setq after-rparen (point)))
10489fcb
AM
7391 ((eq (char-before) ?\])
7392 (setq after-rparen nil))
7393 (t ; either } (hit previous defun) or no more parens/brackets
7394 (throw 'knr nil)))
9cf17ef1
AM
7395
7396 (if after-rparen
7397 ;; We're inside a paren. Could it be our argument list....?
7398 (if
7399 (and
7400 (progn
7401 (goto-char after-rparen)
7402 (unless (c-go-list-backward) (throw 'knr nil)) ;
7403 ;; FIXME!!! What about macros between the parens? 2007/01/20
7404 (setq before-lparen (point)))
d9e94c22 7405
9cf17ef1
AM
7406 ;; It can't be the arg list if next token is ; or {
7407 (progn (goto-char after-rparen)
7408 (c-forward-syntactic-ws)
7409 (not (memq (char-after) '(?\; ?\{))))
d9e94c22 7410
9cf17ef1
AM
7411 ;; Is the thing preceding the list an identifier (the
7412 ;; function name), or a macro expansion?
7413 (progn
7414 (goto-char before-lparen)
7415 (eq (c-backward-token-2) 0)
7416 (or (c-on-identifier)
7417 (and (eq (char-after) ?\))
7418 (c-go-up-list-backward)
7419 (eq (c-backward-token-2) 0)
7420 (c-on-identifier))))
7421
7422 ;; Have we got a non-empty list of comma-separated
7423 ;; identifiers?
7424 (progn
7425 (goto-char before-lparen)
7426 (c-forward-token-2) ; to first token inside parens
7427 (and
7428 (c-on-identifier)
7429 (c-forward-token-2)
7430 (catch 'id-list
7431 (while (eq (char-after) ?\,)
7432 (c-forward-token-2)
7433 (unless (c-on-identifier) (throw 'id-list nil))
7434 (c-forward-token-2))
7435 (eq (char-after) ?\))))))
7436
7437 ;; ...Yes. We've identified the function's argument list.
7438 (throw 'knr
7439 (progn (goto-char after-rparen)
7440 (c-forward-syntactic-ws)
7441 (point)))
7442
7443 ;; ...No. The current parens aren't the function's arg list.
7444 (goto-char before-lparen))
7445
7446 (or (c-go-list-backward) ; backwards over [ .... ]
7447 (throw 'knr nil)))))))))
785eecbb
RS
7448
7449(defun c-skip-conditional ()
7450 ;; skip forward over conditional at point, including any predicate
7451 ;; statements in parentheses. No error checking is performed.
0386b551
AM
7452 ;;
7453 ;; This function might do hidden buffer changes.
0ec8351b
BW
7454 (c-forward-sexp (cond
7455 ;; else if()
a66cd3ee
MS
7456 ((looking-at (concat "\\<else"
7457 "\\([ \t\n]\\|\\\\\n\\)+"
7458 "if\\>\\([^_]\\|$\\)"))
7459 3)
0ec8351b 7460 ;; do, else, try, finally
a66cd3ee
MS
7461 ((looking-at (concat "\\<\\("
7462 "do\\|else\\|try\\|finally"
7463 "\\)\\>\\([^_]\\|$\\)"))
130c507e 7464 1)
ce8c7486 7465 ;; for, if, while, switch, catch, synchronized, foreach
0ec8351b 7466 (t 2))))
785eecbb 7467
a66cd3ee
MS
7468(defun c-after-conditional (&optional lim)
7469 ;; If looking at the token after a conditional then return the
7470 ;; position of its start, otherwise return nil.
0386b551
AM
7471 ;;
7472 ;; This function might do hidden buffer changes.
a66cd3ee 7473 (save-excursion
d9e94c22 7474 (and (zerop (c-backward-token-2 1 t lim))
a66cd3ee
MS
7475 (or (looking-at c-block-stmt-1-key)
7476 (and (eq (char-after) ?\()
d9e94c22 7477 (zerop (c-backward-token-2 1 t lim))
a66cd3ee
MS
7478 (looking-at c-block-stmt-2-key)))
7479 (point))))
7480
0386b551
AM
7481(defun c-after-special-operator-id (&optional lim)
7482 ;; If the point is after an operator identifier that isn't handled
7483 ;; like an ordinary symbol (i.e. like "operator =" in C++) then the
7484 ;; position of the start of that identifier is returned. nil is
7485 ;; returned otherwise. The point may be anywhere in the syntactic
7486 ;; whitespace after the last token of the operator identifier.
7487 ;;
7488 ;; This function might do hidden buffer changes.
7489 (save-excursion
7490 (and c-overloadable-operators-regexp
7491 (zerop (c-backward-token-2 1 nil lim))
7492 (looking-at c-overloadable-operators-regexp)
51c9af45 7493 (or (not c-opt-op-identifier-prefix)
0386b551
AM
7494 (and
7495 (zerop (c-backward-token-2 1 nil lim))
51c9af45 7496 (looking-at c-opt-op-identifier-prefix)))
0386b551
AM
7497 (point))))
7498
a66cd3ee
MS
7499(defsubst c-backward-to-block-anchor (&optional lim)
7500 ;; Assuming point is at a brace that opens a statement block of some
7501 ;; kind, move to the proper anchor point for that block. It might
7502 ;; need to be adjusted further by c-add-stmt-syntax, but the
7503 ;; position at return is suitable as start position for that
7504 ;; function.
0386b551
AM
7505 ;;
7506 ;; This function might do hidden buffer changes.
a66cd3ee
MS
7507 (unless (= (point) (c-point 'boi))
7508 (let ((start (c-after-conditional lim)))
7509 (if start
7510 (goto-char start)))))
7511
037558bf 7512(defsubst c-backward-to-decl-anchor (&optional lim)
a66cd3ee
MS
7513 ;; Assuming point is at a brace that opens the block of a top level
7514 ;; declaration of some kind, move to the proper anchor point for
7515 ;; that block.
0386b551
AM
7516 ;;
7517 ;; This function might do hidden buffer changes.
a66cd3ee 7518 (unless (= (point) (c-point 'boi))
037558bf 7519 (c-beginning-of-statement-1 lim)))
a66cd3ee 7520
ff959bab 7521(defun c-search-decl-header-end ()
a66cd3ee
MS
7522 ;; Search forward for the end of the "header" of the current
7523 ;; declaration. That's the position where the definition body
7524 ;; starts, or the first variable initializer, or the ending
7525 ;; semicolon. I.e. search forward for the closest following
7526 ;; (syntactically relevant) '{', '=' or ';' token. Point is left
7527 ;; _after_ the first found token, or at point-max if none is found.
0386b551
AM
7528 ;;
7529 ;; This function might do hidden buffer changes.
ff959bab
MS
7530
7531 (let ((base (point)))
7532 (if (c-major-mode-is 'c++-mode)
7533
7534 ;; In C++ we need to take special care to handle operator
7535 ;; tokens and those pesky template brackets.
7536 (while (and
7537 (c-syntactic-re-search-forward "[;{<=]" nil 'move t t)
7538 (or
7539 (c-end-of-current-token base)
7540 ;; Handle operator identifiers, i.e. ignore any
7541 ;; operator token preceded by "operator".
7542 (save-excursion
7543 (and (c-safe (c-backward-sexp) t)
51c9af45 7544 (looking-at c-opt-op-identifier-prefix)))
ff959bab
MS
7545 (and (eq (char-before) ?<)
7546 (c-with-syntax-table c++-template-syntax-table
7547 (if (c-safe (goto-char (c-up-list-forward (point))))
7548 t
7549 (goto-char (point-max))
7550 nil)))))
7551 (setq base (point)))
7552
7553 (while (and
7554 (c-syntactic-re-search-forward "[;{=]" nil 'move t t)
7555 (c-end-of-current-token base))
7556 (setq base (point))))))
a66cd3ee
MS
7557
7558(defun c-beginning-of-decl-1 (&optional lim)
7559 ;; Go to the beginning of the current declaration, or the beginning
7560 ;; of the previous one if already at the start of it. Point won't
0386b551 7561 ;; be moved out of any surrounding paren. Return a cons cell of the
a66cd3ee
MS
7562 ;; form (MOVE . KNR-POS). MOVE is like the return value from
7563 ;; `c-beginning-of-statement-1'. If point skipped over some K&R
7564 ;; style argument declarations (and they are to be recognized) then
7565 ;; KNR-POS is set to the start of the first such argument
7566 ;; declaration, otherwise KNR-POS is nil. If LIM is non-nil, it's a
7567 ;; position that bounds the backward search.
7568 ;;
7569 ;; NB: Cases where the declaration continues after the block, as in
7570 ;; "struct foo { ... } bar;", are currently recognized as two
7571 ;; declarations, e.g. "struct foo { ... }" and "bar;" in this case.
0386b551
AM
7572 ;;
7573 ;; This function might do hidden buffer changes.
a66cd3ee
MS
7574 (catch 'return
7575 (let* ((start (point))
d9e94c22 7576 (last-stmt-start (point))
0386b551 7577 (move (c-beginning-of-statement-1 lim nil t)))
a66cd3ee 7578
a66cd3ee
MS
7579 ;; `c-beginning-of-statement-1' stops at a block start, but we
7580 ;; want to continue if the block doesn't begin a top level
2a15eb73
MS
7581 ;; construct, i.e. if it isn't preceded by ';', '}', ':', bob,
7582 ;; or an open paren.
d9e94c22 7583 (let ((beg (point)) tentative-move)
51c9af45
AM
7584 ;; Go back one "statement" each time round the loop until we're just
7585 ;; after a ;, }, or :, or at BOB or the start of a macro or start of
7586 ;; an ObjC method. This will move over a multiple declaration whose
7587 ;; components are comma separated.
d9e94c22
MS
7588 (while (and
7589 ;; Must check with c-opt-method-key in ObjC mode.
7590 (not (and c-opt-method-key
7591 (looking-at c-opt-method-key)))
7592 (/= last-stmt-start (point))
7593 (progn
7594 (c-backward-syntactic-ws lim)
7595 (not (memq (char-before) '(?\; ?} ?: nil))))
2a15eb73
MS
7596 (save-excursion
7597 (backward-char)
7598 (not (looking-at "\\s(")))
d9e94c22
MS
7599 ;; Check that we don't move from the first thing in a
7600 ;; macro to its header.
7601 (not (eq (setq tentative-move
0386b551 7602 (c-beginning-of-statement-1 lim nil t))
d9e94c22
MS
7603 'macro)))
7604 (setq last-stmt-start beg
7605 beg (point)
7606 move tentative-move))
7607 (goto-char beg))
7608
7609 (when c-recognize-knr-p
7610 (let ((fallback-pos (point)) knr-argdecl-start)
7611 ;; Handle K&R argdecls. Back up after the "statement" jumped
7612 ;; over by `c-beginning-of-statement-1', unless it was the
7613 ;; function body, in which case we're sitting on the opening
7614 ;; brace now. Then test if we're in a K&R argdecl region and
7615 ;; that we started at the other side of the first argdecl in
7616 ;; it.
7617 (unless (eq (char-after) ?{)
7618 (goto-char last-stmt-start))
7619 (if (and (setq knr-argdecl-start (c-in-knr-argdecl lim))
7620 (< knr-argdecl-start start)
7621 (progn
7622 (goto-char knr-argdecl-start)
0386b551 7623 (not (eq (c-beginning-of-statement-1 lim nil t) 'macro))))
d9e94c22
MS
7624 (throw 'return
7625 (cons (if (eq (char-after fallback-pos) ?{)
7626 'previous
7627 'same)
7628 knr-argdecl-start))
7629 (goto-char fallback-pos))))
7630
51c9af45
AM
7631 ;; `c-beginning-of-statement-1' counts each brace block as a separate
7632 ;; statement, so the result will be 'previous if we've moved over any.
7633 ;; So change our result back to 'same if necessary.
7634 ;;
7635 ;; If they were brace list initializers we might not have moved over a
7636 ;; declaration boundary though, so change it to 'same if we've moved
7637 ;; past a '=' before '{', but not ';'. (This ought to be integrated
7638 ;; into `c-beginning-of-statement-1', so we avoid this extra pass which
7639 ;; potentially can search over a large amount of text.). Take special
7640 ;; pains not to get mislead by C++'s "operator=", and the like.
d9e94c22
MS
7641 (if (and (eq move 'previous)
7642 (c-with-syntax-table (if (c-major-mode-is 'c++-mode)
7643 c++-template-syntax-table
7644 (syntax-table))
7645 (save-excursion
51c9af45
AM
7646 (and
7647 (progn
7648 (while ; keep going back to "[;={"s until we either find
7649 ; no more, or get to one which isn't an "operator ="
7650 (and (c-syntactic-re-search-forward "[;={]" start t t t)
7651 (eq (char-before) ?=)
7652 c-overloadable-operators-regexp
7653 c-opt-op-identifier-prefix
7654 (save-excursion
7655 (eq (c-backward-token-2) 0)
7656 (looking-at c-overloadable-operators-regexp)
7657 (eq (c-backward-token-2) 0)
7658 (looking-at c-opt-op-identifier-prefix))))
7659 (eq (char-before) ?=))
7660 (c-syntactic-re-search-forward "[;{]" start t t)
7661 (eq (char-before) ?{)
7662 (c-safe (goto-char (c-up-list-forward (point))) t)
7663 (not (c-syntactic-re-search-forward ";" start t t))))))
d9e94c22
MS
7664 (cons 'same nil)
7665 (cons move nil)))))
a66cd3ee
MS
7666
7667(defun c-end-of-decl-1 ()
7668 ;; Assuming point is at the start of a declaration (as detected by
7669 ;; e.g. `c-beginning-of-decl-1'), go to the end of it. Unlike
7670 ;; `c-beginning-of-decl-1', this function handles the case when a
7671 ;; block is followed by identifiers in e.g. struct declarations in C
7672 ;; or C++. If a proper end was found then t is returned, otherwise
7673 ;; point is moved as far as possible within the current sexp and nil
7674 ;; is returned. This function doesn't handle macros; use
7675 ;; `c-end-of-macro' instead in those cases.
0386b551
AM
7676 ;;
7677 ;; This function might do hidden buffer changes.
ce8c7486 7678 (let ((start (point))
a66cd3ee
MS
7679 (decl-syntax-table (if (c-major-mode-is 'c++-mode)
7680 c++-template-syntax-table
7681 (syntax-table))))
7682 (catch 'return
7683 (c-search-decl-header-end)
7684
7685 (when (and c-recognize-knr-p
7686 (eq (char-before) ?\;)
7687 (c-in-knr-argdecl start))
7688 ;; Stopped at the ';' in a K&R argdecl section which is
7689 ;; detected using the same criteria as in
7690 ;; `c-beginning-of-decl-1'. Move to the following block
7691 ;; start.
d9e94c22 7692 (c-syntactic-re-search-forward "{" nil 'move t))
a66cd3ee
MS
7693
7694 (when (eq (char-before) ?{)
7695 ;; Encountered a block in the declaration. Jump over it.
7696 (condition-case nil
7697 (goto-char (c-up-list-forward (point)))
d9e94c22
MS
7698 (error (goto-char (point-max))
7699 (throw 'return nil)))
a66cd3ee
MS
7700 (if (or (not c-opt-block-decls-with-vars-key)
7701 (save-excursion
7702 (c-with-syntax-table decl-syntax-table
7703 (let ((lim (point)))
7704 (goto-char start)
b3cf7e18
MS
7705 (not (and
7706 ;; Check for `c-opt-block-decls-with-vars-key'
7707 ;; before the first paren.
7708 (c-syntactic-re-search-forward
d9e94c22 7709 (concat "[;=\(\[{]\\|\\("
b3cf7e18
MS
7710 c-opt-block-decls-with-vars-key
7711 "\\)")
d9e94c22 7712 lim t t t)
b3cf7e18
MS
7713 (match-beginning 1)
7714 (not (eq (char-before) ?_))
d9e94c22
MS
7715 ;; Check that the first following paren is
7716 ;; the block.
7717 (c-syntactic-re-search-forward "[;=\(\[{]"
7718 lim t t t)
b3cf7e18 7719 (eq (char-before) ?{)))))))
a66cd3ee
MS
7720 ;; The declaration doesn't have any of the
7721 ;; `c-opt-block-decls-with-vars' keywords in the
7722 ;; beginning, so it ends here at the end of the block.
7723 (throw 'return t)))
7724
7725 (c-with-syntax-table decl-syntax-table
7726 (while (progn
7727 (if (eq (char-before) ?\;)
7728 (throw 'return t))
d9e94c22 7729 (c-syntactic-re-search-forward ";" nil 'move t))))
a66cd3ee 7730 nil)))
ce8c7486 7731
0386b551
AM
7732(defun c-looking-at-decl-block (containing-sexp goto-start &optional limit)
7733 ;; Assuming the point is at an open brace, check if it starts a
7734 ;; block that contains another declaration level, i.e. that isn't a
7735 ;; statement block or a brace list, and if so return non-nil.
7736 ;;
7737 ;; If the check is successful, the return value is the start of the
7738 ;; keyword that tells what kind of construct it is, i.e. typically
7739 ;; what `c-decl-block-key' matched. Also, if GOTO-START is set then
7740 ;; the point will be at the start of the construct, before any
7741 ;; leading specifiers, otherwise it's at the returned position.
7742 ;;
7743 ;; The point is clobbered if the check is unsuccessful.
7744 ;;
7745 ;; CONTAINING-SEXP is the position of the open of the surrounding
7746 ;; paren, or nil if none.
7747 ;;
7748 ;; The optional LIMIT limits the backward search for the start of
7749 ;; the construct. It's assumed to be at a syntactically relevant
7750 ;; position.
7751 ;;
7752 ;; If any template arglists are found in the searched region before
7753 ;; the open brace, they get marked with paren syntax.
7754 ;;
7755 ;; This function might do hidden buffer changes.
7756
7757 (let ((open-brace (point)) kwd-start first-specifier-pos)
7758 (c-syntactic-skip-backward c-block-prefix-charset limit t)
7759
7760 (when (and c-recognize-<>-arglists
7761 (eq (char-before) ?>))
7762 ;; Could be at the end of a template arglist.
7763 (let ((c-parse-and-markup-<>-arglists t)
7764 (c-disallow-comma-in-<>-arglists
7765 (and containing-sexp
7766 (not (eq (char-after containing-sexp) ?{)))))
7767 (while (and
7768 (c-backward-<>-arglist nil limit)
7769 (progn
7770 (c-syntactic-skip-backward c-block-prefix-charset limit t)
7771 (eq (char-before) ?>))))))
7772
7773 ;; Note: Can't get bogus hits inside template arglists below since they
7774 ;; have gotten paren syntax above.
7775 (when (and
7776 ;; If `goto-start' is set we begin by searching for the
7777 ;; first possible position of a leading specifier list.
7778 ;; The `c-decl-block-key' search continues from there since
7779 ;; we know it can't match earlier.
7780 (if goto-start
7781 (when (c-syntactic-re-search-forward c-symbol-start
7782 open-brace t t)
7783 (goto-char (setq first-specifier-pos (match-beginning 0)))
7784 t)
7785 t)
7786
7787 (cond
7788 ((c-syntactic-re-search-forward c-decl-block-key open-brace t t t)
7789 (goto-char (setq kwd-start (match-beginning 0)))
7790 (or
7791
7792 ;; Found a keyword that can't be a type?
7793 (match-beginning 1)
7794
7795 ;; Can be a type too, in which case it's the return type of a
7796 ;; function (under the assumption that no declaration level
7797 ;; block construct starts with a type).
7798 (not (c-forward-type))
7799
7800 ;; Jumped over a type, but it could be a declaration keyword
7801 ;; followed by the declared identifier that we've jumped over
7802 ;; instead (e.g. in "class Foo {"). If it indeed is a type
7803 ;; then we should be at the declarator now, so check for a
7804 ;; valid declarator start.
7805 ;;
7806 ;; Note: This doesn't cope with the case when a declared
7807 ;; identifier is followed by e.g. '(' in a language where '('
7808 ;; also might be part of a declarator expression. Currently
7809 ;; there's no such language.
7810 (not (or (looking-at c-symbol-start)
7811 (looking-at c-type-decl-prefix-key)))))
7812
7813 ;; In Pike a list of modifiers may be followed by a brace
7814 ;; to make them apply to many identifiers. Note that the
7815 ;; match data will be empty on return in this case.
7816 ((and (c-major-mode-is 'pike-mode)
7817 (progn
7818 (goto-char open-brace)
7819 (= (c-backward-token-2) 0))
7820 (looking-at c-specifier-key)
7821 ;; Use this variant to avoid yet another special regexp.
7822 (c-keyword-member (c-keyword-sym (match-string 1))
7823 'c-modifier-kwds))
7824 (setq kwd-start (point))
7825 t)))
7826
7827 ;; Got a match.
7828
7829 (if goto-start
7830 ;; Back up over any preceding specifiers and their clauses
7831 ;; by going forward from `first-specifier-pos', which is the
7832 ;; earliest possible position where the specifier list can
7833 ;; start.
7834 (progn
7835 (goto-char first-specifier-pos)
7836
7837 (while (< (point) kwd-start)
7838 (if (looking-at c-symbol-key)
7839 ;; Accept any plain symbol token on the ground that
7840 ;; it's a specifier masked through a macro (just
7841 ;; like `c-forward-decl-or-cast-1' skip forward over
7842 ;; such tokens).
7843 ;;
7844 ;; Could be more restrictive wrt invalid keywords,
7845 ;; but that'd only occur in invalid code so there's
7846 ;; no use spending effort on it.
7847 (let ((end (match-end 0)))
7848 (unless (c-forward-keyword-clause 0)
7849 (goto-char end)
7850 (c-forward-syntactic-ws)))
7851
7852 ;; Can't parse a declaration preamble and is still
7853 ;; before `kwd-start'. That means `first-specifier-pos'
7854 ;; was in some earlier construct. Search again.
7855 (if (c-syntactic-re-search-forward c-symbol-start
7856 kwd-start 'move t)
7857 (goto-char (setq first-specifier-pos (match-beginning 0)))
7858 ;; Got no preamble before the block declaration keyword.
7859 (setq first-specifier-pos kwd-start))))
7860
7861 (goto-char first-specifier-pos))
7862 (goto-char kwd-start))
7863
7864 kwd-start)))
ce8c7486 7865
a66cd3ee 7866(defun c-search-uplist-for-classkey (paren-state)
0386b551
AM
7867 ;; Check if the closest containing paren sexp is a declaration
7868 ;; block, returning a 2 element vector in that case. Aref 0
7869 ;; contains the bufpos at boi of the class key line, and aref 1
7870 ;; contains the bufpos of the open brace. This function is an
7871 ;; obsolete wrapper for `c-looking-at-decl-block'.
7872 ;;
7873 ;; This function might do hidden buffer changes.
7874 (let ((open-paren-pos (c-most-enclosing-brace paren-state)))
7875 (when open-paren-pos
7876 (save-excursion
7877 (goto-char open-paren-pos)
7878 (when (and (eq (char-after) ?{)
7879 (c-looking-at-decl-block
7880 (c-safe-position open-paren-pos paren-state)
7881 nil))
7882 (back-to-indentation)
7883 (vector (point) open-paren-pos))))))
785eecbb 7884
a66cd3ee 7885(defun c-inside-bracelist-p (containing-sexp paren-state)
785eecbb
RS
7886 ;; return the buffer position of the beginning of the brace list
7887 ;; statement if we're inside a brace list, otherwise return nil.
7888 ;; CONTAINING-SEXP is the buffer pos of the innermost containing
0386b551 7889 ;; paren. PAREN-STATE is the remainder of the state of enclosing
130c507e 7890 ;; braces
785eecbb
RS
7891 ;;
7892 ;; N.B.: This algorithm can potentially get confused by cpp macros
0386b551 7893 ;; placed in inconvenient locations. It's a trade-off we make for
785eecbb 7894 ;; speed.
0386b551
AM
7895 ;;
7896 ;; This function might do hidden buffer changes.
785eecbb 7897 (or
d9e94c22 7898 ;; This will pick up brace list declarations.
b2acd789
RS
7899 (c-safe
7900 (save-excursion
7901 (goto-char containing-sexp)
0ec8351b 7902 (c-forward-sexp -1)
b2acd789 7903 (let (bracepos)
d9e94c22 7904 (if (and (or (looking-at c-brace-list-key)
0ec8351b 7905 (progn (c-forward-sexp -1)
d9e94c22 7906 (looking-at c-brace-list-key)))
a66cd3ee 7907 (setq bracepos (c-down-list-forward (point)))
b2acd789
RS
7908 (not (c-crosses-statement-barrier-p (point)
7909 (- bracepos 2))))
7910 (point)))))
785eecbb
RS
7911 ;; this will pick up array/aggregate init lists, even if they are nested.
7912 (save-excursion
0ec8351b
BW
7913 (let ((class-key
7914 ;; Pike can have class definitions anywhere, so we must
7915 ;; check for the class key here.
7916 (and (c-major-mode-is 'pike-mode)
a66cd3ee
MS
7917 c-decl-block-key))
7918 bufpos braceassignp lim next-containing)
785eecbb
RS
7919 (while (and (not bufpos)
7920 containing-sexp)
a66cd3ee
MS
7921 (when paren-state
7922 (if (consp (car paren-state))
7923 (setq lim (cdr (car paren-state))
7924 paren-state (cdr paren-state))
7925 (setq lim (car paren-state)))
7926 (when paren-state
7927 (setq next-containing (car paren-state)
7928 paren-state (cdr paren-state))))
785eecbb 7929 (goto-char containing-sexp)
a66cd3ee
MS
7930 (if (c-looking-at-inexpr-block next-containing next-containing)
7931 ;; We're in an in-expression block of some kind. Do not
7932 ;; check nesting. We deliberately set the limit to the
7933 ;; containing sexp, so that c-looking-at-inexpr-block
7934 ;; doesn't check for an identifier before it.
0ec8351b
BW
7935 (setq containing-sexp nil)
7936 ;; see if the open brace is preceded by = or [...] in
7937 ;; this statement, but watch out for operator=
a66cd3ee 7938 (setq braceassignp 'dontknow)
d9e94c22 7939 (c-backward-token-2 1 t lim)
6393fef2 7940 ;; Checks to do only on the first sexp before the brace.
d9e94c22 7941 (when (and c-opt-inexpr-brace-list-key
6393fef2
RS
7942 (eq (char-after) ?\[))
7943 ;; In Java, an initialization brace list may follow
7944 ;; directly after "new Foo[]", so check for a "new"
7945 ;; earlier.
7946 (while (eq braceassignp 'dontknow)
7947 (setq braceassignp
d9e94c22
MS
7948 (cond ((/= (c-backward-token-2 1 t lim) 0) nil)
7949 ((looking-at c-opt-inexpr-brace-list-key) t)
6393fef2
RS
7950 ((looking-at "\\sw\\|\\s_\\|[.[]")
7951 ;; Carry on looking if this is an
7952 ;; identifier (may contain "." in Java)
7953 ;; or another "[]" sexp.
7954 'dontknow)
7955 (t nil)))))
7956 ;; Checks to do on all sexps before the brace, up to the
7957 ;; beginning of the statement.
7958 (while (eq braceassignp 'dontknow)
0ec8351b
BW
7959 (cond ((eq (char-after) ?\;)
7960 (setq braceassignp nil))
7961 ((and class-key
7962 (looking-at class-key))
7963 (setq braceassignp nil))
7964 ((eq (char-after) ?=)
7965 ;; We've seen a =, but must check earlier tokens so
7966 ;; that it isn't something that should be ignored.
7967 (setq braceassignp 'maybe)
7968 (while (and (eq braceassignp 'maybe)
d9e94c22 7969 (zerop (c-backward-token-2 1 t lim)))
0ec8351b
BW
7970 (setq braceassignp
7971 (cond
7972 ;; Check for operator =
51c9af45
AM
7973 ((and c-opt-op-identifier-prefix
7974 (looking-at c-opt-op-identifier-prefix))
0386b551 7975 nil)
130c507e
GM
7976 ;; Check for `<opchar>= in Pike.
7977 ((and (c-major-mode-is 'pike-mode)
7978 (or (eq (char-after) ?`)
7979 ;; Special case for Pikes
7980 ;; `[]=, since '[' is not in
7981 ;; the punctuation class.
7982 (and (eq (char-after) ?\[)
7983 (eq (char-before) ?`))))
7984 nil)
0ec8351b
BW
7985 ((looking-at "\\s.") 'maybe)
7986 ;; make sure we're not in a C++ template
7987 ;; argument assignment
a66cd3ee
MS
7988 ((and
7989 (c-major-mode-is 'c++-mode)
7990 (save-excursion
7991 (let ((here (point))
7992 (pos< (progn
7993 (skip-chars-backward "^<>")
7994 (point))))
7995 (and (eq (char-before) ?<)
7996 (not (c-crosses-statement-barrier-p
7997 pos< here))
7998 (not (c-in-literal))
7999 ))))
0ec8351b 8000 nil)
6393fef2
RS
8001 (t t))))))
8002 (if (and (eq braceassignp 'dontknow)
d9e94c22 8003 (/= (c-backward-token-2 1 t lim) 0))
6393fef2
RS
8004 (setq braceassignp nil)))
8005 (if (not braceassignp)
0ec8351b
BW
8006 (if (eq (char-after) ?\;)
8007 ;; Brace lists can't contain a semicolon, so we're done.
8008 (setq containing-sexp nil)
a66cd3ee
MS
8009 ;; Go up one level.
8010 (setq containing-sexp next-containing
8011 lim nil
8012 next-containing nil))
0ec8351b
BW
8013 ;; we've hit the beginning of the aggregate list
8014 (c-beginning-of-statement-1
a66cd3ee 8015 (c-most-enclosing-brace paren-state))
0ec8351b 8016 (setq bufpos (point))))
a66cd3ee 8017 )
785eecbb
RS
8018 bufpos))
8019 ))
8020
0ec8351b
BW
8021(defun c-looking-at-special-brace-list (&optional lim)
8022