Add 2012 to FSF copyright years for Emacs files
[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 17313.
CommitLineData
785eecbb
RS
1;;; cc-engine.el --- core syntax guessing engine for CC mode
2
acaf905b 3;; Copyright (C) 1985, 1987, 1992-2012 Free Software Foundation, Inc.
785eecbb 4
e309f66c
AM
5;; Authors: 2001- Alan Mackenzie
6;; 1998- Martin Stjernholm
d9e94c22 7;; 1992-1999 Barry A. Warsaw
5858f68c
GM
8;; 1987 Dave Detlefs
9;; 1987 Stewart Clamen
785eecbb 10;; 1985 Richard M. Stallman
0ec8351b 11;; Maintainer: bug-cc-mode@gnu.org
785eecbb 12;; Created: 22-Apr-1997 (split from cc-mode.el)
bd78fa1d
CY
13;; Keywords: c languages
14;; Package: cc-mode
785eecbb
RS
15
16;; This file is part of GNU Emacs.
17
b1fc2b50 18;; GNU Emacs is free software: you can redistribute it and/or modify
785eecbb 19;; it under the terms of the GNU General Public License as published by
b1fc2b50
GM
20;; the Free Software Foundation, either version 3 of the License, or
21;; (at your option) any later version.
785eecbb
RS
22
23;; GNU Emacs is distributed in the hope that it will be useful,
24;; but WITHOUT ANY WARRANTY; without even the implied warranty of
25;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
26;; GNU General Public License for more details.
27
28;; You should have received a copy of the GNU General Public License
b1fc2b50 29;; along with GNU Emacs. If not, see <http://www.gnu.org/licenses/>.
785eecbb 30
3afbc435
PJ
31;;; Commentary:
32
a66cd3ee
MS
33;; The functions which have docstring documentation can be considered
34;; part of an API which other packages can use in CC Mode buffers.
35;; Otoh, undocumented functions and functions with the documentation
36;; in comments are considered purely internal and can change semantics
37;; or even disappear in the future.
38;;
39;; (This policy applies to CC Mode as a whole, not just this file. It
40;; probably also applies to many other Emacs packages, but here it's
41;; clearly spelled out.)
42
d9e94c22
MS
43;; Hidden buffer changes
44;;
45;; Various functions in CC Mode use text properties for caching and
46;; syntactic markup purposes, and those of them that might modify such
0386b551
AM
47;; properties but still don't modify the buffer in a visible way are
48;; said to do "hidden buffer changes". They should be used within
49;; `c-save-buffer-state' or a similar function that saves and restores
50;; buffer modifiedness, disables buffer change hooks, etc.
d9e94c22 51;;
0386b551
AM
52;; Interactive functions are assumed to not do hidden buffer changes,
53;; except in the specific parts of them that do real changes.
d9e94c22 54;;
0386b551
AM
55;; Lineup functions are assumed to do hidden buffer changes. They
56;; must not do real changes, though.
d9e94c22 57;;
0386b551
AM
58;; All other functions that do hidden buffer changes have that noted
59;; in their doc string or comment.
60;;
61;; The intention with this system is to avoid wrapping every leaf
62;; function that do hidden buffer changes inside
63;; `c-save-buffer-state'. It should be used as near the top of the
64;; interactive functions as possible.
65;;
66;; Functions called during font locking are allowed to do hidden
67;; buffer changes since the font-lock package run them in a context
68;; similar to `c-save-buffer-state' (in fact, that function is heavily
69;; inspired by `save-buffer-state' in the font-lock package).
d9e94c22
MS
70
71;; Use of text properties
72;;
73;; CC Mode uses several text properties internally to mark up various
74;; positions, e.g. to improve speed and to eliminate glitches in
75;; interactive refontification.
76;;
7bfc3fdb
MS
77;; Note: This doc is for internal use only. Other packages should not
78;; assume that these text properties are used as described here.
79;;
0ec1d2c5
AM
80;; 'category
81;; Used for "indirection". With its help, some other property can
82;; be cheaply and easily switched on or off everywhere it occurs.
83;;
d9e94c22 84;; 'syntax-table
c8018ede
AM
85;; Used to modify the syntax of some characters. It is used to
86;; mark the "<" and ">" of angle bracket parens with paren syntax, and
87;; to "hide" obtrusive characters in preprocessor lines.
d9e94c22
MS
88;;
89;; This property is used on single characters and is therefore
90;; always treated as front and rear nonsticky (or start and end open
91;; in XEmacs vocabulary). It's therefore installed on
92;; `text-property-default-nonsticky' if that variable exists (Emacs
93;; >= 21).
94;;
95;; 'c-is-sws and 'c-in-sws
96;; Used by `c-forward-syntactic-ws' and `c-backward-syntactic-ws' to
97;; speed them up. See the comment blurb before `c-put-is-sws'
98;; below for further details.
99;;
100;; 'c-type
101;; This property is used on single characters to mark positions with
0386b551
AM
102;; special syntactic relevance of various sorts. Its primary use is
103;; to avoid glitches when multiline constructs are refontified
d9e94c22
MS
104;; interactively (on font lock decoration level 3). It's cleared in
105;; a region before it's fontified and is then put on relevant chars
106;; in that region as they are encountered during the fontification.
107;; The value specifies the kind of position:
108;;
109;; 'c-decl-arg-start
110;; Put on the last char of the token preceding each declaration
111;; inside a declaration style arglist (typically in a function
112;; prototype).
113;;
114;; 'c-decl-end
115;; Put on the last char of the token preceding a declaration.
116;; This is used in cases where declaration boundaries can't be
117;; recognized simply by looking for a token like ";" or "}".
118;; `c-type-decl-end-used' must be set if this is used (see also
119;; `c-find-decl-spots').
120;;
121;; 'c-<>-arg-sep
122;; Put on the commas that separate arguments in angle bracket
123;; arglists like C++ template arglists.
124;;
125;; 'c-decl-id-start and 'c-decl-type-start
126;; Put on the last char of the token preceding each declarator
127;; in the declarator list of a declaration. They are also used
128;; between the identifiers cases like enum declarations.
129;; 'c-decl-type-start is used when the declarators are types,
130;; 'c-decl-id-start otherwise.
131;;
132;; 'c-awk-NL-prop
133;; Used in AWK mode to mark the various kinds of newlines. See
134;; cc-awk.el.
135
3afbc435
PJ
136;;; Code:
137
0ec8351b 138(eval-when-compile
51f606de 139 (let ((load-path
130c507e
GM
140 (if (and (boundp 'byte-compile-dest-file)
141 (stringp byte-compile-dest-file))
142 (cons (file-name-directory byte-compile-dest-file) load-path)
51f606de 143 load-path)))
d9e94c22 144 (load "cc-bytecomp" nil t)))
130c507e
GM
145
146(cc-require 'cc-defs)
d9e94c22 147(cc-require-when-compile 'cc-langs)
130c507e 148(cc-require 'cc-vars)
d9e94c22 149
130c507e
GM
150;; Silence the compiler.
151(cc-bytecomp-defun buffer-syntactic-context) ; XEmacs
0ec8351b 152
51f606de 153\f
d9e94c22
MS
154;; Make declarations for all the `c-lang-defvar' variables in cc-langs.
155
156(defmacro c-declare-lang-variables ()
157 `(progn
485fe977
RS
158 ,@(apply 'nconc
159 (mapcar (lambda (init)
160 `(,(if (elt init 2)
161 `(defvar ,(car init) nil ,(elt init 2))
162 `(defvar ,(car init) nil))
163 (make-variable-buffer-local ',(car init))))
164 (cdr c-lang-variable-inits)))))
d9e94c22
MS
165(c-declare-lang-variables)
166
167\f
168;;; Internal state variables.
169
170;; Internal state of hungry delete key feature
171(defvar c-hungry-delete-key nil)
172(make-variable-buffer-local 'c-hungry-delete-key)
173
0386b551
AM
174;; The electric flag (toggled by `c-toggle-electric-state').
175;; If t, electric actions (like automatic reindentation, and (if
176;; c-auto-newline is also set) auto newlining) will happen when an electric
177;; key like `{' is pressed (or an electric keyword like `else').
178(defvar c-electric-flag t)
179(make-variable-buffer-local 'c-electric-flag)
180
d9e94c22
MS
181;; Internal state of auto newline feature.
182(defvar c-auto-newline nil)
183(make-variable-buffer-local 'c-auto-newline)
184
0386b551 185;; Included in the mode line to indicate the active submodes.
cb694ab7
AM
186;; (defvar c-submode-indicators nil)
187;; (make-variable-buffer-local 'c-submode-indicators)
d9e94c22 188
a66cd3ee
MS
189(defun c-calculate-state (arg prevstate)
190 ;; Calculate the new state of PREVSTATE, t or nil, based on arg. If
191 ;; arg is nil or zero, toggle the state. If arg is negative, turn
192 ;; the state off, and if arg is positive, turn the state on
193 (if (or (not arg)
194 (zerop (setq arg (prefix-numeric-value arg))))
195 (not prevstate)
196 (> arg 0)))
197
d9e94c22 198\f
037558bf
MS
199;; Basic handling of preprocessor directives.
200
201;; This is a dynamically bound cache used together with
202;; `c-query-macro-start' and `c-query-and-set-macro-start'. It only
203;; works as long as point doesn't cross a macro boundary.
204(defvar c-macro-start 'unknown)
205
206(defsubst c-query-and-set-macro-start ()
037558bf
MS
207 (if (symbolp c-macro-start)
208 (setq c-macro-start (save-excursion
0386b551
AM
209 (c-save-buffer-state ()
210 (and (c-beginning-of-macro)
211 (point)))))
037558bf
MS
212 c-macro-start))
213
214(defsubst c-query-macro-start ()
037558bf
MS
215 (if (symbolp c-macro-start)
216 (save-excursion
0386b551
AM
217 (c-save-buffer-state ()
218 (and (c-beginning-of-macro)
219 (point))))
037558bf
MS
220 c-macro-start))
221
222(defun c-beginning-of-macro (&optional lim)
223 "Go to the beginning of a preprocessor directive.
224Leave point at the beginning of the directive and return t if in one,
225otherwise return nil and leave point unchanged.
226
0386b551
AM
227Note that this function might do hidden buffer changes. See the
228comment at the start of cc-engine.el for more info."
037558bf
MS
229 (when c-opt-cpp-prefix
230 (let ((here (point)))
231 (save-restriction
232 (if lim (narrow-to-region lim (point-max)))
233 (beginning-of-line)
234 (while (eq (char-before (1- (point))) ?\\)
235 (forward-line -1))
236 (back-to-indentation)
237 (if (and (<= (point) here)
238 (looking-at c-opt-cpp-start))
239 t
240 (goto-char here)
241 nil)))))
242
243(defun c-end-of-macro ()
244 "Go to the end of a preprocessor directive.
0386b551
AM
245More accurately, move the point to the end of the closest following
246line that doesn't end with a line continuation backslash - no check is
247done that the point is inside a cpp directive to begin with.
037558bf 248
0386b551
AM
249Note that this function might do hidden buffer changes. See the
250comment at the start of cc-engine.el for more info."
037558bf
MS
251 (while (progn
252 (end-of-line)
253 (when (and (eq (char-before) ?\\)
254 (not (eobp)))
255 (forward-char)
256 t))))
257
0ec1d2c5
AM
258(defun c-syntactic-end-of-macro ()
259 ;; Go to the end of a CPP directive, or a "safe" pos just before.
260 ;;
261 ;; This is normally the end of the next non-escaped line. A "safe"
262 ;; position is one not within a string or comment. (The EOL on a line
263 ;; comment is NOT "safe").
264 ;;
265 ;; This function must only be called from the beginning of a CPP construct.
266 ;;
267 ;; Note that this function might do hidden buffer changes. See the comment
268 ;; at the start of cc-engine.el for more info.
269 (let* ((here (point))
270 (there (progn (c-end-of-macro) (point)))
271 (s (parse-partial-sexp here there)))
272 (while (and (or (nth 3 s) ; in a string
273 (nth 4 s)) ; in a comment (maybe at end of line comment)
274 (> there here)) ; No infinite loops, please.
275 (setq there (1- (nth 8 s)))
276 (setq s (parse-partial-sexp here there)))
277 (point)))
278
51c9af45
AM
279(defun c-forward-over-cpp-define-id ()
280 ;; Assuming point is at the "#" that introduces a preprocessor
281 ;; directive, it's moved forward to the end of the identifier which is
282 ;; "#define"d (or whatever c-opt-cpp-macro-define specifies). Non-nil
283 ;; is returned in this case, in all other cases nil is returned and
284 ;; point isn't moved.
285 ;;
286 ;; This function might do hidden buffer changes.
287 (when (and c-opt-cpp-macro-define-id
288 (looking-at c-opt-cpp-macro-define-id))
289 (goto-char (match-end 0))))
290
037558bf
MS
291(defun c-forward-to-cpp-define-body ()
292 ;; Assuming point is at the "#" that introduces a preprocessor
293 ;; directive, it's moved forward to the start of the definition body
0386b551
AM
294 ;; if it's a "#define" (or whatever c-opt-cpp-macro-define
295 ;; specifies). Non-nil is returned in this case, in all other cases
296 ;; nil is returned and point isn't moved.
297 ;;
298 ;; This function might do hidden buffer changes.
299 (when (and c-opt-cpp-macro-define-start
300 (looking-at c-opt-cpp-macro-define-start)
037558bf
MS
301 (not (= (match-end 0) (c-point 'eol))))
302 (goto-char (match-end 0))))
303
304\f
d9e94c22
MS
305;;; Basic utility functions.
306
0386b551 307(defun c-syntactic-content (from to paren-level)
d9e94c22
MS
308 ;; Return the given region as a string where all syntactic
309 ;; whitespace is removed or, where necessary, replaced with a single
0386b551
AM
310 ;; space. If PAREN-LEVEL is given then all parens in the region are
311 ;; collapsed to "()", "[]" etc.
312 ;;
313 ;; This function might do hidden buffer changes.
314
d9e94c22 315 (save-excursion
0386b551
AM
316 (save-restriction
317 (narrow-to-region from to)
318 (goto-char from)
319 (let* ((parts (list nil)) (tail parts) pos in-paren)
320
321 (while (re-search-forward c-syntactic-ws-start to t)
322 (goto-char (setq pos (match-beginning 0)))
323 (c-forward-syntactic-ws)
324 (if (= (point) pos)
325 (forward-char)
326
327 (when paren-level
328 (save-excursion
329 (setq in-paren (= (car (parse-partial-sexp from pos 1)) 1)
330 pos (point))))
331
332 (if (and (> pos from)
333 (< (point) to)
334 (looking-at "\\w\\|\\s_")
335 (save-excursion
336 (goto-char (1- pos))
337 (looking-at "\\w\\|\\s_")))
338 (progn
339 (setcdr tail (list (buffer-substring-no-properties from pos)
340 " "))
341 (setq tail (cddr tail)))
342 (setcdr tail (list (buffer-substring-no-properties from pos)))
343 (setq tail (cdr tail)))
344
345 (when in-paren
346 (when (= (car (parse-partial-sexp pos to -1)) -1)
347 (setcdr tail (list (buffer-substring-no-properties
348 (1- (point)) (point))))
349 (setq tail (cdr tail))))
350
351 (setq from (point))))
352
353 (setcdr tail (list (buffer-substring-no-properties from to)))
354 (apply 'concat (cdr parts))))))
355
356(defun c-shift-line-indentation (shift-amt)
357 ;; Shift the indentation of the current line with the specified
358 ;; amount (positive inwards). The buffer is modified only if
359 ;; SHIFT-AMT isn't equal to zero.
360 (let ((pos (- (point-max) (point)))
361 (c-macro-start c-macro-start)
362 tmp-char-inserted)
363 (if (zerop shift-amt)
364 nil
365 ;; If we're on an empty line inside a macro, we take the point
366 ;; to be at the current indentation and shift it to the
367 ;; appropriate column. This way we don't treat the extra
368 ;; whitespace out to the line continuation as indentation.
369 (when (and (c-query-and-set-macro-start)
370 (looking-at "[ \t]*\\\\$")
371 (save-excursion
372 (skip-chars-backward " \t")
373 (bolp)))
374 (insert ?x)
375 (backward-char)
376 (setq tmp-char-inserted t))
377 (unwind-protect
378 (let ((col (current-indentation)))
379 (delete-region (c-point 'bol) (c-point 'boi))
380 (beginning-of-line)
381 (indent-to (+ col shift-amt)))
382 (when tmp-char-inserted
383 (delete-char 1))))
384 ;; If initial point was within line's indentation and we're not on
385 ;; a line with a line continuation in a macro, position after the
386 ;; indentation. Else stay at same point in text.
387 (if (and (< (point) (c-point 'boi))
388 (not tmp-char-inserted))
389 (back-to-indentation)
390 (if (> (- (point-max) pos) (point))
391 (goto-char (- (point-max) pos))))))
d9e94c22
MS
392
393(defsubst c-keyword-sym (keyword)
394 ;; Return non-nil if the string KEYWORD is a known keyword. More
395 ;; precisely, the value is the symbol for the keyword in
396 ;; `c-keywords-obarray'.
397 (intern-soft keyword c-keywords-obarray))
398
399(defsubst c-keyword-member (keyword-sym lang-constant)
400 ;; Return non-nil if the symbol KEYWORD-SYM, as returned by
401 ;; `c-keyword-sym', is a member of LANG-CONSTANT, which is the name
402 ;; of a language constant that ends with "-kwds". If KEYWORD-SYM is
403 ;; nil then the result is nil.
404 (get keyword-sym lang-constant))
405
406;; String syntax chars, suitable for skip-syntax-(forward|backward).
407(defconst c-string-syntax (if (memq 'gen-string-delim c-emacs-features)
408 "\"|"
409 "\""))
410
0386b551 411;; Regexp matching string limit syntax.
d9e94c22
MS
412(defconst c-string-limit-regexp (if (memq 'gen-string-delim c-emacs-features)
413 "\\s\"\\|\\s|"
414 "\\s\""))
415
0386b551
AM
416;; Regexp matching WS followed by string limit syntax.
417(defconst c-ws*-string-limit-regexp
418 (concat "[ \t]*\\(" c-string-limit-regexp "\\)"))
419
d9e94c22
MS
420;; Holds formatted error strings for the few cases where parse errors
421;; are reported.
a66cd3ee 422(defvar c-parsing-error nil)
d9e94c22
MS
423(make-variable-buffer-local 'c-parsing-error)
424
425(defun c-echo-parsing-error (&optional quiet)
d9e94c22
MS
426 (when (and c-report-syntactic-errors c-parsing-error (not quiet))
427 (c-benign-error "%s" c-parsing-error))
428 c-parsing-error)
429
430;; Faces given to comments and string literals. This is used in some
431;; situations to speed up recognition; it isn't mandatory that font
432;; locking is in use. This variable is extended with the face in
433;; `c-doc-face-name' when fontification is activated in cc-fonts.el.
1e330469 434(defvar c-literal-faces
0386b551
AM
435 (append '(font-lock-comment-face font-lock-string-face)
436 (when (facep 'font-lock-comment-delimiter-face)
437 ;; New in Emacs 22.
438 '(font-lock-comment-delimiter-face))))
439
440(defsubst c-put-c-type-property (pos value)
441 ;; Put a c-type property with the given value at POS.
442 (c-put-char-property pos 'c-type value))
443
444(defun c-clear-c-type-property (from to value)
5a89f0a7 445 ;; Remove all occurrences of the c-type property that has the given
0386b551
AM
446 ;; value in the region between FROM and TO. VALUE is assumed to not
447 ;; be nil.
448 ;;
449 ;; Note: This assumes that c-type is put on single chars only; it's
450 ;; very inefficient if matching properties cover large regions.
451 (save-excursion
452 (goto-char from)
453 (while (progn
454 (when (eq (get-text-property (point) 'c-type) value)
455 (c-clear-char-property (point) 'c-type))
456 (goto-char (next-single-property-change (point) 'c-type nil to))
457 (< (point) to)))))
037558bf 458
d9e94c22
MS
459\f
460;; Some debug tools to visualize various special positions. This
461;; debug code isn't as portable as the rest of CC Mode.
462
463(cc-bytecomp-defun overlays-in)
464(cc-bytecomp-defun overlay-get)
465(cc-bytecomp-defun overlay-start)
466(cc-bytecomp-defun overlay-end)
467(cc-bytecomp-defun delete-overlay)
468(cc-bytecomp-defun overlay-put)
469(cc-bytecomp-defun make-overlay)
470
471(defun c-debug-add-face (beg end face)
472 (c-save-buffer-state ((overlays (overlays-in beg end)) overlay)
473 (while overlays
474 (setq overlay (car overlays)
475 overlays (cdr overlays))
476 (when (eq (overlay-get overlay 'face) face)
477 (setq beg (min beg (overlay-start overlay))
478 end (max end (overlay-end overlay)))
479 (delete-overlay overlay)))
480 (overlay-put (make-overlay beg end) 'face face)))
481
482(defun c-debug-remove-face (beg end face)
483 (c-save-buffer-state ((overlays (overlays-in beg end)) overlay
484 (ol-beg beg) (ol-end end))
485 (while overlays
486 (setq overlay (car overlays)
487 overlays (cdr overlays))
488 (when (eq (overlay-get overlay 'face) face)
489 (setq ol-beg (min ol-beg (overlay-start overlay))
490 ol-end (max ol-end (overlay-end overlay)))
491 (delete-overlay overlay)))
492 (when (< ol-beg beg)
493 (overlay-put (make-overlay ol-beg beg) 'face face))
494 (when (> ol-end end)
495 (overlay-put (make-overlay end ol-end) 'face face))))
496
497\f
498;; `c-beginning-of-statement-1' and accompanying stuff.
130c507e 499
64001211
RS
500;; KLUDGE ALERT: c-maybe-labelp is used to pass information between
501;; c-crosses-statement-barrier-p and c-beginning-of-statement-1. A
502;; better way should be implemented, but this will at least shut up
503;; the byte compiler.
0386b551 504(defvar c-maybe-labelp)
64001211 505
d9e94c22
MS
506;; New awk-compatible version of c-beginning-of-statement-1, ACM 2002/6/22
507
a66cd3ee
MS
508;; Macros used internally in c-beginning-of-statement-1 for the
509;; automaton actions.
510(defmacro c-bos-push-state ()
511 '(setq stack (cons (cons state saved-pos)
512 stack)))
513(defmacro c-bos-pop-state (&optional do-if-done)
514 `(if (setq state (car (car stack))
515 saved-pos (cdr (car stack))
516 stack (cdr stack))
517 t
518 ,do-if-done
519 (throw 'loop nil)))
520(defmacro c-bos-pop-state-and-retry ()
521 '(throw 'loop (setq state (car (car stack))
522 saved-pos (cdr (car stack))
523 ;; Throw nil if stack is empty, else throw non-nil.
524 stack (cdr stack))))
525(defmacro c-bos-save-pos ()
526 '(setq saved-pos (vector pos tok ptok pptok)))
527(defmacro c-bos-restore-pos ()
528 '(unless (eq (elt saved-pos 0) start)
529 (setq pos (elt saved-pos 0)
530 tok (elt saved-pos 1)
531 ptok (elt saved-pos 2)
532 pptok (elt saved-pos 3))
533 (goto-char pos)
534 (setq sym nil)))
535(defmacro c-bos-save-error-info (missing got)
536 `(setq saved-pos (vector pos ,missing ,got)))
537(defmacro c-bos-report-error ()
538 '(unless noerror
539 (setq c-parsing-error
540 (format "No matching `%s' found for `%s' on line %d"
541 (elt saved-pos 1)
542 (elt saved-pos 2)
543 (1+ (count-lines (point-min)
544 (c-point 'bol (elt saved-pos 0))))))))
545
546(defun c-beginning-of-statement-1 (&optional lim ignore-labels
547 noerror comma-delim)
548 "Move to the start of the current statement or declaration, or to
549the previous one if already at the beginning of one. Only
550statements/declarations on the same level are considered, i.e. don't
551move into or out of sexps (not even normal expression parentheses).
552
5a89f0a7 553If point is already at the earliest statement within braces or parens,
a85fd6da
AM
554this function doesn't move back into any whitespace preceding it; it
555returns 'same in this case.
556
d9e94c22
MS
557Stop at statement continuation tokens like \"else\", \"catch\",
558\"finally\" and the \"while\" in \"do ... while\" if the start point
559is within the continuation. If starting at such a token, move to the
560corresponding statement start. If at the beginning of a statement,
561move to the closest containing statement if there is any. This might
562also stop at a continuation clause.
a66cd3ee 563
0386b551
AM
564Labels are treated as part of the following statements if
565IGNORE-LABELS is non-nil. (FIXME: Doesn't work if we stop at a known
cb694ab7
AM
566statement start keyword.) Otherwise, each label is treated as a
567separate statement.
a66cd3ee 568
cb694ab7
AM
569Macros are ignored \(i.e. skipped over) unless point is within one, in
570which case the content of the macro is treated as normal code. Aside
571from any normal statement starts found in it, stop at the first token
572of the content in the macro, i.e. the expression of an \"#if\" or the
573start of the definition in a \"#define\". Also stop at start of
574macros before leaving them.
a66cd3ee 575
a85fd6da 576Return:
d28e7f28 577'label if stopped at a label or \"case...:\" or \"default:\";
a85fd6da
AM
578'same if stopped at the beginning of the current statement;
579'up if stepped to a containing statement;
580'previous if stepped to a preceding statement;
581'beginning if stepped from a statement continuation clause to
582 its start clause; or
583'macro if stepped to a macro start.
584Note that 'same and not 'label is returned if stopped at the same
585label without crossing the colon character.
a66cd3ee
MS
586
587LIM may be given to limit the search. If the search hits the limit,
588point will be left at the closest following token, or at the start
589position if that is less ('same is returned in this case).
590
591NOERROR turns off error logging to `c-parsing-error'.
592
cb694ab7
AM
593Normally only ';' and virtual semicolons are considered to delimit
594statements, but if COMMA-DELIM is non-nil then ',' is treated
595as a delimiter too.
0386b551
AM
596
597Note that this function might do hidden buffer changes. See the
598comment at the start of cc-engine.el for more info."
a66cd3ee 599
d9e94c22
MS
600 ;; The bulk of this function is a pushdown automaton that looks at statement
601 ;; boundaries and the tokens (such as "while") in c-opt-block-stmt-key. Its
602 ;; purpose is to keep track of nested statements, ensuring that such
5a89f0a7 603 ;; statements are skipped over in their entirety (somewhat akin to what C-M-p
d9e94c22 604 ;; does with nested braces/brackets/parentheses).
a66cd3ee
MS
605 ;;
606 ;; Note: The position of a boundary is the following token.
607 ;;
d9e94c22
MS
608 ;; Beginning with the current token (the one following point), move back one
609 ;; sexp at a time (where a sexp is, more or less, either a token or the
610 ;; entire contents of a brace/bracket/paren pair). Each time a statement
611 ;; boundary is crossed or a "while"-like token is found, update the state of
612 ;; the PDA. Stop at the beginning of a statement when the stack (holding
613 ;; nested statement info) is empty and the position has been moved.
614 ;;
615 ;; The following variables constitute the PDA:
616 ;;
617 ;; sym: This is either the "while"-like token (e.g. 'for) we've just
618 ;; scanned back over, 'boundary if we've just gone back over a
619 ;; statement boundary, or nil otherwise.
620 ;; state: takes one of the values (nil else else-boundary while
621 ;; while-boundary catch catch-boundary).
622 ;; nil means "no "while"-like token yet scanned".
623 ;; 'else, for example, means "just gone back over an else".
624 ;; 'else-boundary means "just gone back over a statement boundary
625 ;; immediately after having gone back over an else".
626 ;; saved-pos: A vector of either saved positions (tok ptok pptok, etc.) or
627 ;; of error reporting information.
628 ;; stack: The stack onto which the PDA pushes its state. Each entry
629 ;; consists of a saved value of state and saved-pos. An entry is
630 ;; pushed when we move back over a "continuation" token (e.g. else)
631 ;; and popped when we encounter the corresponding opening token
632 ;; (e.g. if).
633 ;;
634 ;;
b414f371 635 ;; The following diagram briefly outlines the PDA.
a66cd3ee
MS
636 ;;
637 ;; Common state:
d9e94c22
MS
638 ;; "else": Push state, goto state `else'.
639 ;; "while": Push state, goto state `while'.
640 ;; "catch" or "finally": Push state, goto state `catch'.
641 ;; boundary: Pop state.
a66cd3ee
MS
642 ;; other: Do nothing special.
643 ;;
d9e94c22
MS
644 ;; State `else':
645 ;; boundary: Goto state `else-boundary'.
646 ;; other: Error, pop state, retry token.
647 ;;
648 ;; State `else-boundary':
649 ;; "if": Pop state.
650 ;; boundary: Error, pop state.
651 ;; other: See common state.
652 ;;
653 ;; State `while':
654 ;; boundary: Save position, goto state `while-boundary'.
655 ;; other: Pop state, retry token.
656 ;;
657 ;; State `while-boundary':
658 ;; "do": Pop state.
659 ;; boundary: Restore position if it's not at start, pop state. [*see below]
660 ;; other: See common state.
661 ;;
662 ;; State `catch':
663 ;; boundary: Goto state `catch-boundary'.
664 ;; other: Error, pop state, retry token.
665 ;;
666 ;; State `catch-boundary':
667 ;; "try": Pop state.
668 ;; "catch": Goto state `catch'.
669 ;; boundary: Error, pop state.
670 ;; other: See common state.
671 ;;
672 ;; [*] In the `while-boundary' state, we had pushed a 'while state, and were
673 ;; searching for a "do" which would have opened a do-while. If we didn't
674 ;; find it, we discard the analysis done since the "while", go back to this
675 ;; token in the buffer and restart the scanning there, this time WITHOUT
676 ;; pushing the 'while state onto the stack.
677 ;;
a66cd3ee
MS
678 ;; In addition to the above there is some special handling of labels
679 ;; and macros.
680
681 (let ((case-fold-search nil)
682 (start (point))
683 macro-start
684 (delims (if comma-delim '(?\; ?,) '(?\;)))
685 (c-stmt-delim-chars (if comma-delim
686 c-stmt-delim-chars-with-comma
687 c-stmt-delim-chars))
d28e7f28 688 c-in-literal-cache c-maybe-labelp after-case:-pos saved
0386b551
AM
689 ;; Current position.
690 pos
691 ;; Position of last stmt boundary character (e.g. ;).
692 boundary-pos
693 ;; The position of the last sexp or bound that follows the
694 ;; first found colon, i.e. the start of the nonlabel part of
695 ;; the statement. It's `start' if a colon is found just after
696 ;; the start.
697 after-labels-pos
698 ;; Like `after-labels-pos', but the first such position inside
699 ;; a label, i.e. the start of the last label before the start
700 ;; of the nonlabel part of the statement.
701 last-label-pos
702 ;; The last position where a label is possible provided the
703 ;; statement started there. It's nil as long as no invalid
704 ;; label content has been found (according to
020716e1 705 ;; `c-nonlabel-token-key'). It's `start' if no valid label
0386b551
AM
706 ;; content was found in the label. Note that we might still
707 ;; regard it a label if it starts with `c-label-kwds'.
708 label-good-pos
b248a85d
AM
709 ;; Putative positions of the components of a bitfield declaration,
710 ;; e.g. "int foo : NUM_FOO_BITS ;"
711 bitfield-type-pos bitfield-id-pos bitfield-size-pos
0386b551
AM
712 ;; Symbol just scanned back over (e.g. 'while or 'boundary).
713 ;; See above.
714 sym
715 ;; Current state in the automaton. See above.
716 state
717 ;; Current saved positions. See above.
718 saved-pos
719 ;; Stack of conses (state . saved-pos).
720 stack
721 ;; Regexp which matches "for", "if", etc.
722 (cond-key (or c-opt-block-stmt-key
a66cd3ee 723 "\\<\\>")) ; Matches nothing.
0386b551
AM
724 ;; Return value.
725 (ret 'same)
726 ;; Positions of the last three sexps or bounds we've stopped at.
727 tok ptok pptok)
a66cd3ee
MS
728
729 (save-restriction
730 (if lim (narrow-to-region lim (point-max)))
731
732 (if (save-excursion
733 (and (c-beginning-of-macro)
734 (/= (point) start)))
735 (setq macro-start (point)))
736
d9e94c22 737 ;; Try to skip back over unary operator characters, to register
a66cd3ee
MS
738 ;; that we've moved.
739 (while (progn
740 (setq pos (point))
0386b551
AM
741 (c-backward-syntactic-ws)
742 ;; Protect post-++/-- operators just before a virtual semicolon.
743 (and (not (c-at-vsemi-p))
744 (/= (skip-chars-backward "-+!*&~@`#") 0))))
d9e94c22
MS
745
746 ;; Skip back over any semicolon here. If it was a bare semicolon, we're
0386b551 747 ;; done. Later on we ignore the boundaries for statements that don't
d9e94c22
MS
748 ;; contain any sexp. The only thing that is affected is that the error
749 ;; checking is a little less strict, and we really don't bother.
a66cd3ee
MS
750 (if (and (memq (char-before) delims)
751 (progn (forward-char -1)
752 (setq saved (point))
0386b551 753 (c-backward-syntactic-ws)
a66cd3ee
MS
754 (or (memq (char-before) delims)
755 (memq (char-before) '(?: nil))
d9e94c22 756 (eq (char-syntax (char-before)) ?\()
0386b551 757 (c-at-vsemi-p))))
a66cd3ee
MS
758 (setq ret 'previous
759 pos saved)
760
761 ;; Begin at start and not pos to detect macros if we stand
762 ;; directly after the #.
763 (goto-char start)
764 (if (looking-at "\\<\\|\\W")
765 ;; Record this as the first token if not starting inside it.
766 (setq tok start))
767
b248a85d
AM
768
769 ;; The following while loop goes back one sexp (balanced parens,
770 ;; etc. with contents, or symbol or suchlike) each iteration. This
771 ;; movement is accomplished with a call to c-backward-sexp approx 170
772 ;; lines below.
773 ;;
774 ;; The loop is exited only by throwing nil to the (catch 'loop ...):
775 ;; 1. On reaching the start of a macro;
776 ;; 2. On having passed a stmt boundary with the PDA stack empty;
777 ;; 3. On reaching the start of an Objective C method def;
778 ;; 4. From macro `c-bos-pop-state'; when the stack is empty;
779 ;; 5. From macro `c-bos-pop-state-and-retry' when the stack is empty.
a66cd3ee
MS
780 (while
781 (catch 'loop ;; Throw nil to break, non-nil to continue.
782 (cond
b248a85d 783 ;; Are we in a macro, just after the opening #?
a66cd3ee 784 ((save-excursion
0386b551 785 (and macro-start ; Always NIL for AWK.
a66cd3ee
MS
786 (progn (skip-chars-backward " \t")
787 (eq (char-before) ?#))
788 (progn (setq saved (1- (point)))
789 (beginning-of-line)
790 (not (eq (char-before (1- (point))) ?\\)))
d9e94c22 791 (looking-at c-opt-cpp-start)
a66cd3ee
MS
792 (progn (skip-chars-forward " \t")
793 (eq (point) saved))))
794 (goto-char saved)
795 (if (and (c-forward-to-cpp-define-body)
796 (progn (c-forward-syntactic-ws start)
797 (< (point) start)))
798 ;; Stop at the first token in the content of the macro.
799 (setq pos (point)
800 ignore-labels t) ; Avoid the label check on exit.
801 (setq pos saved
802 ret 'macro
803 ignore-labels t))
b248a85d 804 (throw 'loop nil)) ; 1. Start of macro.
a66cd3ee 805
d9e94c22
MS
806 ;; Do a round through the automaton if we've just passed a
807 ;; statement boundary or passed a "while"-like token.
a66cd3ee
MS
808 ((or sym
809 (and (looking-at cond-key)
810 (setq sym (intern (match-string 1)))))
811
812 (when (and (< pos start) (null stack))
b248a85d 813 (throw 'loop nil)) ; 2. Statement boundary.
a66cd3ee 814
d9e94c22
MS
815 ;; The PDA state handling.
816 ;;
037558bf 817 ;; Refer to the description of the PDA in the opening
d9e94c22
MS
818 ;; comments. In the following OR form, the first leaf
819 ;; attempts to handles one of the specific actions detailed
820 ;; (e.g., finding token "if" whilst in state `else-boundary').
821 ;; We drop through to the second leaf (which handles common
822 ;; state) if no specific handler is found in the first cond.
823 ;; If a parsing error is detected (e.g. an "else" with no
824 ;; preceding "if"), we throw to the enclosing catch.
825 ;;
826 ;; Note that the (eq state 'else) means
827 ;; "we've just passed an else", NOT "we're looking for an
828 ;; else".
a66cd3ee
MS
829 (or (cond
830 ((eq state 'else)
831 (if (eq sym 'boundary)
832 (setq state 'else-boundary)
833 (c-bos-report-error)
834 (c-bos-pop-state-and-retry)))
835
836 ((eq state 'else-boundary)
837 (cond ((eq sym 'if)
838 (c-bos-pop-state (setq ret 'beginning)))
839 ((eq sym 'boundary)
840 (c-bos-report-error)
841 (c-bos-pop-state))))
842
843 ((eq state 'while)
844 (if (and (eq sym 'boundary)
845 ;; Since this can cause backtracking we do a
846 ;; little more careful analysis to avoid it:
847 ;; If there's a label in front of the while
848 ;; it can't be part of a do-while.
849 (not after-labels-pos))
850 (progn (c-bos-save-pos)
851 (setq state 'while-boundary))
d9e94c22 852 (c-bos-pop-state-and-retry))) ; Can't be a do-while
a66cd3ee
MS
853
854 ((eq state 'while-boundary)
855 (cond ((eq sym 'do)
856 (c-bos-pop-state (setq ret 'beginning)))
d9e94c22
MS
857 ((eq sym 'boundary) ; isn't a do-while
858 (c-bos-restore-pos) ; the position of the while
859 (c-bos-pop-state)))) ; no longer searching for do.
a66cd3ee
MS
860
861 ((eq state 'catch)
862 (if (eq sym 'boundary)
863 (setq state 'catch-boundary)
864 (c-bos-report-error)
865 (c-bos-pop-state-and-retry)))
866
867 ((eq state 'catch-boundary)
868 (cond
869 ((eq sym 'try)
870 (c-bos-pop-state (setq ret 'beginning)))
871 ((eq sym 'catch)
872 (setq state 'catch))
873 ((eq sym 'boundary)
874 (c-bos-report-error)
875 (c-bos-pop-state)))))
876
d9e94c22
MS
877 ;; This is state common. We get here when the previous
878 ;; cond statement found no particular state handler.
a66cd3ee 879 (cond ((eq sym 'boundary)
d9e94c22
MS
880 ;; If we have a boundary at the start
881 ;; position we push a frame to go to the
882 ;; previous statement.
883 (if (>= pos start)
884 (c-bos-push-state)
885 (c-bos-pop-state)))
a66cd3ee
MS
886 ((eq sym 'else)
887 (c-bos-push-state)
888 (c-bos-save-error-info 'if 'else)
889 (setq state 'else))
890 ((eq sym 'while)
0386b551
AM
891 ;; Is this a real while, or a do-while?
892 ;; The next `when' triggers unless we are SURE that
40ba43b4 893 ;; the `while' is not the tail end of a `do-while'.
a66cd3ee 894 (when (or (not pptok)
d9e94c22 895 (memq (char-after pptok) delims)
0386b551
AM
896 ;; The following kludge is to prevent
897 ;; infinite recursion when called from
898 ;; c-awk-after-if-for-while-condition-p,
899 ;; or the like.
900 (and (eq (point) start)
901 (c-vsemi-status-unknown-p))
902 (c-at-vsemi-p pptok))
a66cd3ee
MS
903 ;; Since this can cause backtracking we do a
904 ;; little more careful analysis to avoid it: If
0386b551
AM
905 ;; the while isn't followed by a (possibly
906 ;; virtual) semicolon it can't be a do-while.
a66cd3ee
MS
907 (c-bos-push-state)
908 (setq state 'while)))
909 ((memq sym '(catch finally))
910 (c-bos-push-state)
911 (c-bos-save-error-info 'try sym)
912 (setq state 'catch))))
913
914 (when c-maybe-labelp
915 ;; We're either past a statement boundary or at the
916 ;; start of a statement, so throw away any label data
917 ;; for the previous one.
918 (setq after-labels-pos nil
919 last-label-pos nil
920 c-maybe-labelp nil))))
921
d9e94c22
MS
922 ;; Step to the previous sexp, but not if we crossed a
923 ;; boundary, since that doesn't consume an sexp.
a66cd3ee
MS
924 (if (eq sym 'boundary)
925 (setq ret 'previous)
d9e94c22
MS
926
927 ;; HERE IS THE SINGLE PLACE INSIDE THE PDA LOOP WHERE WE MOVE
0386b551
AM
928 ;; BACKWARDS THROUGH THE SOURCE.
929
0386b551 930 (c-backward-syntactic-ws)
0386b551 931 (let ((before-sws-pos (point))
0386b551
AM
932 ;; The end position of the area to search for statement
933 ;; barriers in this round.
b248a85d 934 (maybe-after-boundary-pos pos))
0386b551 935
b248a85d
AM
936 ;; Go back over exactly one logical sexp, taking proper
937 ;; account of macros and escaped EOLs.
0386b551
AM
938 (while
939 (progn
940 (unless (c-safe (c-backward-sexp) t)
941 ;; Give up if we hit an unbalanced block. Since the
942 ;; stack won't be empty the code below will report a
943 ;; suitable error.
a66cd3ee 944 (throw 'loop nil))
b248a85d
AM
945 (cond
946 ;; Have we moved into a macro?
947 ((and (not macro-start)
948 (c-beginning-of-macro))
949 ;; Have we crossed a statement boundary? If not,
950 ;; keep going back until we find one or a "real" sexp.
951 (and
952 (save-excursion
953 (c-end-of-macro)
954 (not (c-crosses-statement-barrier-p
955 (point) maybe-after-boundary-pos)))
956 (setq maybe-after-boundary-pos (point))))
957 ;; Have we just gone back over an escaped NL? This
958 ;; doesn't count as a sexp.
959 ((looking-at "\\\\$")))))
960
961 ;; Have we crossed a statement boundary?
962 (setq boundary-pos
963 (cond
964 ;; Are we at a macro beginning?
965 ((and (not macro-start)
966 c-opt-cpp-prefix
967 (looking-at c-opt-cpp-prefix))
968 (save-excursion
969 (c-end-of-macro)
970 (c-crosses-statement-barrier-p
971 (point) maybe-after-boundary-pos)))
972 ;; Just gone back over a brace block?
973 ((and
974 (eq (char-after) ?{)
975 (not (c-looking-at-inexpr-block lim nil t)))
976 (save-excursion
977 (c-forward-sexp) (point)))
978 ;; Just gone back over some paren block?
979 ((looking-at "\\s\(")
980 (save-excursion
981 (goto-char (1+ (c-down-list-backward
982 before-sws-pos)))
983 (c-crosses-statement-barrier-p
984 (point) maybe-after-boundary-pos)))
985 ;; Just gone back over an ordinary symbol of some sort?
986 (t (c-crosses-statement-barrier-p
987 (point) maybe-after-boundary-pos))))
91af3942 988
b248a85d
AM
989 (when boundary-pos
990 (setq pptok ptok
991 ptok tok
992 tok boundary-pos
993 sym 'boundary)
994 ;; Like a C "continue". Analyze the next sexp.
995 (throw 'loop t))))
a66cd3ee
MS
996
997 ;; ObjC method def?
998 (when (and c-opt-method-key
999 (setq saved (c-in-method-def-p)))
1000 (setq pos saved
1001 ignore-labels t) ; Avoid the label check on exit.
b248a85d
AM
1002 (throw 'loop nil)) ; 3. ObjC method def.
1003
1004 ;; Might we have a bitfield declaration, "<type> <id> : <size>"?
1005 (if c-has-bitfields
1006 (cond
1007 ;; The : <size> and <id> fields?
1008 ((and (numberp c-maybe-labelp)
1009 (not bitfield-size-pos)
1010 (save-excursion
1011 (goto-char (or tok start))
1012 (not (looking-at c-keywords-regexp)))
1013 (not (looking-at c-keywords-regexp))
1014 (not (c-punctuation-in (point) c-maybe-labelp)))
1015 (setq bitfield-size-pos (or tok start)
1016 bitfield-id-pos (point)))
1017 ;; The <type> field?
1018 ((and bitfield-id-pos
1019 (not bitfield-type-pos))
1020 (if (and (looking-at c-symbol-key) ; Can only be an integer type. :-)
1021 (not (looking-at c-not-primitive-type-keywords-regexp))
1022 (not (c-punctuation-in (point) tok)))
1023 (setq bitfield-type-pos (point))
1024 (setq bitfield-size-pos nil
1025 bitfield-id-pos nil)))))
a66cd3ee 1026
0386b551
AM
1027 ;; Handle labels.
1028 (unless (eq ignore-labels t)
1029 (when (numberp c-maybe-labelp)
cb694ab7
AM
1030 ;; `c-crosses-statement-barrier-p' has found a colon, so we
1031 ;; might be in a label now. Have we got a real label
1032 ;; (including a case label) or something like C++'s "public:"?
d28e7f28
AM
1033 ;; A case label might use an expression rather than a token.
1034 (setq after-case:-pos (or tok start))
020716e1
AM
1035 (if (or (looking-at c-nonlabel-token-key) ; e.g. "while" or "'a'"
1036 ;; Catch C++'s inheritance construct "class foo : bar".
1037 (save-excursion
1038 (and
1039 (c-safe (c-backward-sexp) t)
1040 (looking-at c-nonlabel-token-2-key))))
d28e7f28
AM
1041 (setq c-maybe-labelp nil)
1042 (if after-labels-pos ; Have we already encountered a label?
1043 (if (not last-label-pos)
1044 (setq last-label-pos (or tok start)))
1045 (setq after-labels-pos (or tok start)))
1046 (setq c-maybe-labelp t
1047 label-good-pos nil))) ; bogus "label"
cb694ab7
AM
1048
1049 (when (and (not label-good-pos) ; i.e. no invalid "label"'s yet
1050 ; been found.
1051 (looking-at c-nonlabel-token-key)) ; e.g. "while :"
0386b551
AM
1052 ;; We're in a potential label and it's the first
1053 ;; time we've found something that isn't allowed in
1054 ;; one.
1055 (setq label-good-pos (or tok start))))
1056
1057 ;; We've moved back by a sexp, so update the token positions.
a66cd3ee
MS
1058 (setq sym nil
1059 pptok ptok
1060 ptok tok
1061 tok (point)
b248a85d
AM
1062 pos tok) ; always non-nil
1063 ) ; end of (catch loop ....)
1064 ) ; end of sexp-at-a-time (while ....)
91af3942 1065
a66cd3ee
MS
1066 ;; If the stack isn't empty there might be errors to report.
1067 (while stack
1068 (if (and (vectorp saved-pos) (eq (length saved-pos) 3))
1069 (c-bos-report-error))
1070 (setq saved-pos (cdr (car stack))
1071 stack (cdr stack)))
1072
1073 (when (and (eq ret 'same)
1074 (not (memq sym '(boundary ignore nil))))
1075 ;; Need to investigate closer whether we've crossed
1076 ;; between a substatement and its containing statement.
1077 (if (setq saved (if (looking-at c-block-stmt-1-key)
1078 ptok
1079 pptok))
1080 (cond ((> start saved) (setq pos saved))
1081 ((= start saved) (setq ret 'up)))))
1082
0386b551
AM
1083 (when (and (not ignore-labels)
1084 (eq c-maybe-labelp t)
d9e94c22 1085 (not (eq ret 'beginning))
0386b551 1086 after-labels-pos
b248a85d 1087 (not bitfield-type-pos) ; Bitfields take precedence over labels.
0386b551
AM
1088 (or (not label-good-pos)
1089 (<= label-good-pos pos)
1090 (progn
1091 (goto-char (if (and last-label-pos
1092 (< last-label-pos start))
1093 last-label-pos
1094 pos))
1095 (looking-at c-label-kwds-regexp))))
a66cd3ee
MS
1096 ;; We're in a label. Maybe we should step to the statement
1097 ;; after it.
1098 (if (< after-labels-pos start)
1099 (setq pos after-labels-pos)
1100 (setq ret 'label)
0386b551
AM
1101 (if (and last-label-pos (< last-label-pos start))
1102 ;; Might have jumped over several labels. Go to the last one.
a66cd3ee
MS
1103 (setq pos last-label-pos)))))
1104
d28e7f28 1105 ;; Have we got "case <expression>:"?
a66cd3ee 1106 (goto-char pos)
d28e7f28
AM
1107 (when (and after-case:-pos
1108 (not (eq ret 'beginning))
1109 (looking-at c-case-kwds-regexp))
1110 (if (< after-case:-pos start)
0000ee90
AM
1111 (setq pos after-case:-pos))
1112 (if (eq ret 'same)
1113 (setq ret 'label)))
d28e7f28
AM
1114
1115 ;; Skip over the unary operators that can start the statement.
a66cd3ee 1116 (while (progn
0386b551
AM
1117 (c-backward-syntactic-ws)
1118 ;; protect AWK post-inc/decrement operators, etc.
1119 (and (not (c-at-vsemi-p (point)))
1120 (/= (skip-chars-backward "-+!*&~@`#") 0)))
a66cd3ee
MS
1121 (setq pos (point)))
1122 (goto-char pos)
1123 ret)))
785eecbb 1124
b248a85d
AM
1125(defun c-punctuation-in (from to)
1126 "Return non-nil if there is a non-comment non-macro punctuation character
1127between FROM and TO. FROM must not be in a string or comment. The returned
1128value is the position of the first such character."
1129 (save-excursion
1130 (goto-char from)
1131 (let ((pos (point)))
1132 (while (progn (skip-chars-forward c-symbol-chars to)
1133 (c-forward-syntactic-ws to)
1134 (> (point) pos))
1135 (setq pos (point))))
1136 (and (< (point) to) (point))))
1137
785eecbb 1138(defun c-crosses-statement-barrier-p (from to)
a66cd3ee
MS
1139 "Return non-nil if buffer positions FROM to TO cross one or more
1140statement or declaration boundaries. The returned value is actually
d9e94c22
MS
1141the position of the earliest boundary char. FROM must not be within
1142a string or comment.
a66cd3ee
MS
1143
1144The variable `c-maybe-labelp' is set to the position of the first `:' that
1145might start a label (i.e. not part of `::' and not preceded by `?'). If a
0386b551
AM
1146single `?' is found, then `c-maybe-labelp' is cleared.
1147
1148For AWK, a statement which is terminated by an EOL (not a \; or a }) is
1149regarded as having a \"virtual semicolon\" immediately after the last token on
b414f371 1150the line. If this virtual semicolon is _at_ from, the function recognizes it.
0386b551
AM
1151
1152Note that this function might do hidden buffer changes. See the
1153comment at the start of cc-engine.el for more info."
536610a4
AM
1154 (let* ((skip-chars
1155 ;; If the current language has CPP macros, insert # into skip-chars.
1156 (if c-opt-cpp-symbol
1157 (concat (substring c-stmt-delim-chars 0 1) ; "^"
1158 c-opt-cpp-symbol ; usually "#"
1159 (substring c-stmt-delim-chars 1)) ; e.g. ";{}?:"
1160 c-stmt-delim-chars))
1161 (non-skip-list
1162 (append (substring skip-chars 1) nil)) ; e.g. (?# ?\; ?{ ?} ?? ?:)
1163 lit-range vsemi-pos)
1164 (save-restriction
1165 (widen)
1166 (save-excursion
1167 (catch 'done
1168 (goto-char from)
1169 (while (progn (skip-chars-forward
1170 skip-chars
1171 (min to (c-point 'bonl)))
1172 (< (point) to))
1173 (cond
1174 ;; Virtual semicolon?
1175 ((and (bolp)
1176 (save-excursion
1177 (progn
1178 (if (setq lit-range (c-literal-limits from)) ; Have we landed in a string/comment?
1179 (goto-char (car lit-range)))
1180 (c-backward-syntactic-ws) ; ? put a limit here, maybe?
1181 (setq vsemi-pos (point))
1182 (c-at-vsemi-p))))
1183 (throw 'done vsemi-pos))
1184 ;; In a string/comment?
1185 ((setq lit-range (c-literal-limits))
1186 (goto-char (cdr lit-range)))
1187 ((eq (char-after) ?:)
1188 (forward-char)
1189 (if (and (eq (char-after) ?:)
1190 (< (point) to))
1191 ;; Ignore scope operators.
1192 (forward-char)
1193 (setq c-maybe-labelp (1- (point)))))
1194 ((eq (char-after) ??)
1195 ;; A question mark. Can't be a label, so stop
1196 ;; looking for more : and ?.
1197 (setq c-maybe-labelp nil
1198 skip-chars (substring c-stmt-delim-chars 0 -2)))
1199 ;; At a CPP construct?
1200 ((and c-opt-cpp-symbol (looking-at c-opt-cpp-symbol)
1201 (save-excursion
1202 (forward-line 0)
1203 (looking-at c-opt-cpp-prefix)))
1204 (c-end-of-macro))
1205 ((memq (char-after) non-skip-list)
1206 (throw 'done (point)))))
1207 ;; In trailing space after an as yet undetected virtual semicolon?
1208 (c-backward-syntactic-ws from)
1209 (if (and (< (point) to)
1210 (c-at-vsemi-p))
1211 (point)
1212 nil))))))
0386b551
AM
1213
1214(defun c-at-statement-start-p ()
1215 "Return non-nil if the point is at the first token in a statement
1216or somewhere in the syntactic whitespace before it.
1217
1218A \"statement\" here is not restricted to those inside code blocks.
1219Any kind of declaration-like construct that occur outside function
1220bodies is also considered a \"statement\".
1221
1222Note that this function might do hidden buffer changes. See the
1223comment at the start of cc-engine.el for more info."
1224
1225 (save-excursion
1226 (let ((end (point))
1227 c-maybe-labelp)
1228 (c-syntactic-skip-backward (substring c-stmt-delim-chars 1) nil t)
1229 (or (bobp)
1230 (eq (char-before) ?})
1231 (and (eq (char-before) ?{)
1232 (not (and c-special-brace-lists
1233 (progn (backward-char)
1234 (c-looking-at-special-brace-list)))))
1235 (c-crosses-statement-barrier-p (point) end)))))
1236
1237(defun c-at-expression-start-p ()
1238 "Return non-nil if the point is at the first token in an expression or
1239statement, or somewhere in the syntactic whitespace before it.
1240
1241An \"expression\" here is a bit different from the normal language
1242grammar sense: It's any sequence of expression tokens except commas,
1243unless they are enclosed inside parentheses of some kind. Also, an
1244expression never continues past an enclosing parenthesis, but it might
1245contain parenthesis pairs of any sort except braces.
1246
1247Since expressions never cross statement boundaries, this function also
1248recognizes statement beginnings, just like `c-at-statement-start-p'.
1249
1250Note that this function might do hidden buffer changes. See the
1251comment at the start of cc-engine.el for more info."
1252
1253 (save-excursion
1254 (let ((end (point))
1255 (c-stmt-delim-chars c-stmt-delim-chars-with-comma)
1256 c-maybe-labelp)
1257 (c-syntactic-skip-backward (substring c-stmt-delim-chars 1) nil t)
1258 (or (bobp)
1259 (memq (char-before) '(?{ ?}))
1260 (save-excursion (backward-char)
1261 (looking-at "\\s("))
1262 (c-crosses-statement-barrier-p (point) end)))))
785eecbb
RS
1263
1264\f
d9e94c22
MS
1265;; A set of functions that covers various idiosyncrasies in
1266;; implementations of `forward-comment'.
1267
1268;; Note: Some emacsen considers incorrectly that any line comment
1269;; ending with a backslash continues to the next line. I can't think
1270;; of any way to work around that in a reliable way without changing
1271;; the buffer, though. Suggestions welcome. ;) (No, temporarily
1272;; changing the syntax for backslash doesn't work since we must treat
1273;; escapes in string literals correctly.)
1274
1275(defun c-forward-single-comment ()
1276 "Move forward past whitespace and the closest following comment, if any.
1277Return t if a comment was found, nil otherwise. In either case, the
1278point is moved past the following whitespace. Line continuations,
1279i.e. a backslashes followed by line breaks, are treated as whitespace.
1280The line breaks that end line comments are considered to be the
1281comment enders, so the point will be put on the beginning of the next
1282line if it moved past a line comment.
1283
1284This function does not do any hidden buffer changes."
1285
1286 (let ((start (point)))
1287 (when (looking-at "\\([ \t\n\r\f\v]\\|\\\\[\n\r]\\)+")
1288 (goto-char (match-end 0)))
1289
1290 (when (forward-comment 1)
1291 (if (eobp)
1292 ;; Some emacsen (e.g. XEmacs 21) return t when moving
1293 ;; forwards at eob.
1294 nil
1295
1296 ;; Emacs includes the ending newline in a b-style (c++)
1297 ;; comment, but XEmacs doesn't. We depend on the Emacs
1298 ;; behavior (which also is symmetric).
1299 (if (and (eolp) (elt (parse-partial-sexp start (point)) 7))
1300 (condition-case nil (forward-char 1)))
1301
1302 t))))
1303
1304(defsubst c-forward-comments ()
1305 "Move forward past all following whitespace and comments.
1306Line continuations, i.e. a backslashes followed by line breaks, are
1307treated as whitespace.
1308
0386b551
AM
1309Note that this function might do hidden buffer changes. See the
1310comment at the start of cc-engine.el for more info."
d9e94c22
MS
1311
1312 (while (or
1313 ;; If forward-comment in at least XEmacs 21 is given a large
1314 ;; positive value, it'll loop all the way through if it hits
1315 ;; eob.
1316 (and (forward-comment 5)
1317 ;; Some emacsen (e.g. XEmacs 21) return t when moving
1318 ;; forwards at eob.
1319 (not (eobp)))
1320
1321 (when (looking-at "\\\\[\n\r]")
1322 (forward-char 2)
1323 t))))
1324
1325(defun c-backward-single-comment ()
1326 "Move backward past whitespace and the closest preceding comment, if any.
1327Return t if a comment was found, nil otherwise. In either case, the
1328point is moved past the preceding whitespace. Line continuations,
1329i.e. a backslashes followed by line breaks, are treated as whitespace.
1330The line breaks that end line comments are considered to be the
1331comment enders, so the point cannot be at the end of the same line to
1332move over a line comment.
1333
1334This function does not do any hidden buffer changes."
1335
1336 (let ((start (point)))
1337 ;; When we got newline terminated comments, forward-comment in all
1338 ;; supported emacsen so far will stop at eol of each line not
1339 ;; ending with a comment when moving backwards. This corrects for
1340 ;; that, and at the same time handles line continuations.
1341 (while (progn
1342 (skip-chars-backward " \t\n\r\f\v")
1343 (and (looking-at "[\n\r]")
0386b551 1344 (eq (char-before) ?\\)))
d9e94c22
MS
1345 (backward-char))
1346
1347 (if (bobp)
1348 ;; Some emacsen (e.g. Emacs 19.34) return t when moving
1349 ;; backwards at bob.
1350 nil
1351
1352 ;; Leave point after the closest following newline if we've
1353 ;; backed up over any above, since forward-comment won't move
1354 ;; backward over a line comment if point is at the end of the
1355 ;; same line.
1356 (re-search-forward "\\=\\s *[\n\r]" start t)
1357
bf2c1571 1358 (if (if (let (open-paren-in-column-0-is-defun-start) (forward-comment -1))
d9e94c22
MS
1359 (if (eolp)
1360 ;; If forward-comment above succeeded and we're at eol
1361 ;; then the newline we moved over above didn't end a
1362 ;; line comment, so we give it another go.
bf2c1571
AM
1363 (let (open-paren-in-column-0-is-defun-start)
1364 (forward-comment -1))
d9e94c22
MS
1365 t))
1366
1367 ;; Emacs <= 20 and XEmacs move back over the closer of a
1368 ;; block comment that lacks an opener.
1369 (if (looking-at "\\*/")
1370 (progn (forward-char 2) nil)
1371 t)))))
1372
1373(defsubst c-backward-comments ()
1374 "Move backward past all preceding whitespace and comments.
1375Line continuations, i.e. a backslashes followed by line breaks, are
1376treated as whitespace. The line breaks that end line comments are
1377considered to be the comment enders, so the point cannot be at the end
0386b551
AM
1378of the same line to move over a line comment. Unlike
1379c-backward-syntactic-ws, this function doesn't move back over
1380preprocessor directives.
d9e94c22 1381
0386b551
AM
1382Note that this function might do hidden buffer changes. See the
1383comment at the start of cc-engine.el for more info."
d9e94c22
MS
1384
1385 (let ((start (point)))
1386 (while (and
0386b551 1387 ;; `forward-comment' in some emacsen (e.g. XEmacs 21.4)
d9e94c22
MS
1388 ;; return t when moving backwards at bob.
1389 (not (bobp))
1390
bf2c1571
AM
1391 (if (let (open-paren-in-column-0-is-defun-start)
1392 (forward-comment -1))
d9e94c22
MS
1393 (if (looking-at "\\*/")
1394 ;; Emacs <= 20 and XEmacs move back over the
1395 ;; closer of a block comment that lacks an opener.
1396 (progn (forward-char 2) nil)
1397 t)
1398
1399 ;; XEmacs treats line continuations as whitespace but
1400 ;; only in the backward direction, which seems a bit
1401 ;; odd. Anyway, this is necessary for Emacs.
1402 (when (and (looking-at "[\n\r]")
1403 (eq (char-before) ?\\)
1404 (< (point) start))
1405 (backward-char)
1406 t))))))
1407
1408\f
d9e94c22 1409;; Tools for skipping over syntactic whitespace.
a66cd3ee 1410
d9e94c22
MS
1411;; The following functions use text properties to cache searches over
1412;; large regions of syntactic whitespace. It works as follows:
1413;;
1414;; o If a syntactic whitespace region contains anything but simple
1415;; whitespace (i.e. space, tab and line breaks), the text property
1416;; `c-in-sws' is put over it. At places where we have stopped
1417;; within that region there's also a `c-is-sws' text property.
1418;; That since there typically are nested whitespace inside that
1419;; must be handled separately, e.g. whitespace inside a comment or
1420;; cpp directive. Thus, from one point with `c-is-sws' it's safe
1421;; to jump to another point with that property within the same
1422;; `c-in-sws' region. It can be likened to a ladder where
1423;; `c-in-sws' marks the bars and `c-is-sws' the rungs.
1424;;
1425;; o The `c-is-sws' property is put on the simple whitespace chars at
1426;; a "rung position" and also maybe on the first following char.
1427;; As many characters as can be conveniently found in this range
1428;; are marked, but no assumption can be made that the whole range
1429;; is marked (it could be clobbered by later changes, for
1430;; instance).
1431;;
1432;; Note that some part of the beginning of a sequence of simple
1433;; whitespace might be part of the end of a preceding line comment
1434;; or cpp directive and must not be considered part of the "rung".
1435;; Such whitespace is some amount of horizontal whitespace followed
1436;; by a newline. In the case of cpp directives it could also be
1437;; two newlines with horizontal whitespace between them.
1438;;
1439;; The reason to include the first following char is to cope with
1440;; "rung positions" that doesn't have any ordinary whitespace. If
1441;; `c-is-sws' is put on a token character it does not have
1442;; `c-in-sws' set simultaneously. That's the only case when that
1443;; can occur, and the reason for not extending the `c-in-sws'
1444;; region to cover it is that the `c-in-sws' region could then be
1445;; accidentally merged with a following one if the token is only
1446;; one character long.
1447;;
1448;; o On buffer changes the `c-in-sws' and `c-is-sws' properties are
1449;; removed in the changed region. If the change was inside
1450;; syntactic whitespace that means that the "ladder" is broken, but
1451;; a later call to `c-forward-sws' or `c-backward-sws' will use the
1452;; parts on either side and use an ordinary search only to "repair"
1453;; the gap.
1454;;
1455;; Special care needs to be taken if a region is removed: If there
1456;; are `c-in-sws' on both sides of it which do not connect inside
1457;; the region then they can't be joined. If e.g. a marked macro is
1458;; broken, syntactic whitespace inside the new text might be
1459;; marked. If those marks would become connected with the old
1460;; `c-in-sws' range around the macro then we could get a ladder
1461;; with one end outside the macro and the other at some whitespace
1462;; within it.
1463;;
1464;; The main motivation for this system is to increase the speed in
1465;; skipping over the large whitespace regions that can occur at the
1466;; top level in e.g. header files that contain a lot of comments and
1467;; cpp directives. For small comments inside code it's probably
1468;; slower than using `forward-comment' straightforwardly, but speed is
1469;; not a significant factor there anyway.
1470
1471; (defface c-debug-is-sws-face
1472; '((t (:background "GreenYellow")))
1473; "Debug face to mark the `c-is-sws' property.")
1474; (defface c-debug-in-sws-face
1475; '((t (:underline t)))
1476; "Debug face to mark the `c-in-sws' property.")
1477
1478; (defun c-debug-put-sws-faces ()
1479; ;; Put the sws debug faces on all the `c-is-sws' and `c-in-sws'
1480; ;; properties in the buffer.
1481; (interactive)
1482; (save-excursion
0386b551 1483; (c-save-buffer-state (in-face)
d9e94c22
MS
1484; (goto-char (point-min))
1485; (setq in-face (if (get-text-property (point) 'c-is-sws)
1486; (point)))
1487; (while (progn
1488; (goto-char (next-single-property-change
1489; (point) 'c-is-sws nil (point-max)))
1490; (if in-face
1491; (progn
1492; (c-debug-add-face in-face (point) 'c-debug-is-sws-face)
1493; (setq in-face nil))
1494; (setq in-face (point)))
1495; (not (eobp))))
1496; (goto-char (point-min))
1497; (setq in-face (if (get-text-property (point) 'c-in-sws)
1498; (point)))
1499; (while (progn
1500; (goto-char (next-single-property-change
1501; (point) 'c-in-sws nil (point-max)))
1502; (if in-face
1503; (progn
1504; (c-debug-add-face in-face (point) 'c-debug-in-sws-face)
1505; (setq in-face nil))
1506; (setq in-face (point)))
1507; (not (eobp)))))))
1508
1509(defmacro c-debug-sws-msg (&rest args)
1510 ;;`(message ,@args)
1511 )
1512
1513(defmacro c-put-is-sws (beg end)
0386b551 1514 ;; This macro does a hidden buffer change.
d9e94c22
MS
1515 `(let ((beg ,beg) (end ,end))
1516 (put-text-property beg end 'c-is-sws t)
1517 ,@(when (facep 'c-debug-is-sws-face)
1518 `((c-debug-add-face beg end 'c-debug-is-sws-face)))))
1519
1520(defmacro c-put-in-sws (beg end)
0386b551 1521 ;; This macro does a hidden buffer change.
d9e94c22
MS
1522 `(let ((beg ,beg) (end ,end))
1523 (put-text-property beg end 'c-in-sws t)
1524 ,@(when (facep 'c-debug-is-sws-face)
1525 `((c-debug-add-face beg end 'c-debug-in-sws-face)))))
1526
1527(defmacro c-remove-is-sws (beg end)
0386b551 1528 ;; This macro does a hidden buffer change.
d9e94c22
MS
1529 `(let ((beg ,beg) (end ,end))
1530 (remove-text-properties beg end '(c-is-sws nil))
1531 ,@(when (facep 'c-debug-is-sws-face)
1532 `((c-debug-remove-face beg end 'c-debug-is-sws-face)))))
1533
1534(defmacro c-remove-in-sws (beg end)
0386b551 1535 ;; This macro does a hidden buffer change.
d9e94c22
MS
1536 `(let ((beg ,beg) (end ,end))
1537 (remove-text-properties beg end '(c-in-sws nil))
1538 ,@(when (facep 'c-debug-is-sws-face)
1539 `((c-debug-remove-face beg end 'c-debug-in-sws-face)))))
1540
1541(defmacro c-remove-is-and-in-sws (beg end)
0386b551 1542 ;; This macro does a hidden buffer change.
d9e94c22
MS
1543 `(let ((beg ,beg) (end ,end))
1544 (remove-text-properties beg end '(c-is-sws nil c-in-sws nil))
1545 ,@(when (facep 'c-debug-is-sws-face)
1546 `((c-debug-remove-face beg end 'c-debug-is-sws-face)
1547 (c-debug-remove-face beg end 'c-debug-in-sws-face)))))
1548
1549(defsubst c-invalidate-sws-region-after (beg end)
1550 ;; Called from `after-change-functions'. Note that if
1551 ;; `c-forward-sws' or `c-backward-sws' are used outside
1552 ;; `c-save-buffer-state' or similar then this will remove the cache
1553 ;; properties right after they're added.
0386b551
AM
1554 ;;
1555 ;; This function does hidden buffer changes.
d9e94c22
MS
1556
1557 (save-excursion
1558 ;; Adjust the end to remove the properties in any following simple
1559 ;; ws up to and including the next line break, if there is any
1560 ;; after the changed region. This is necessary e.g. when a rung
1561 ;; marked empty line is converted to a line comment by inserting
1562 ;; "//" before the line break. In that case the line break would
1563 ;; keep the rung mark which could make a later `c-backward-sws'
1564 ;; move into the line comment instead of over it.
1565 (goto-char end)
1566 (skip-chars-forward " \t\f\v")
1567 (when (and (eolp) (not (eobp)))
1568 (setq end (1+ (point)))))
1569
1570 (when (and (= beg end)
1571 (get-text-property beg 'c-in-sws)
2a15eb73 1572 (> beg (point-min))
d9e94c22
MS
1573 (get-text-property (1- beg) 'c-in-sws))
1574 ;; Ensure that an `c-in-sws' range gets broken. Note that it isn't
1575 ;; safe to keep a range that was continuous before the change. E.g:
1576 ;;
1577 ;; #define foo
1578 ;; \
1579 ;; bar
1580 ;;
1581 ;; There can be a "ladder" between "#" and "b". Now, if the newline
1582 ;; after "foo" is removed then "bar" will become part of the cpp
1583 ;; directive instead of a syntactically relevant token. In that
1584 ;; case there's no longer syntactic ws from "#" to "b".
1585 (setq beg (1- beg)))
1586
1587 (c-debug-sws-msg "c-invalidate-sws-region-after [%s..%s]" beg end)
1588 (c-remove-is-and-in-sws beg end))
1589
1590(defun c-forward-sws ()
1591 ;; Used by `c-forward-syntactic-ws' to implement the unbounded search.
0386b551
AM
1592 ;;
1593 ;; This function might do hidden buffer changes.
d9e94c22
MS
1594
1595 (let (;; `rung-pos' is set to a position as early as possible in the
1596 ;; unmarked part of the simple ws region.
1597 (rung-pos (point)) next-rung-pos rung-end-pos last-put-in-sws-pos
1598 rung-is-marked next-rung-is-marked simple-ws-end
1599 ;; `safe-start' is set when it's safe to cache the start position.
1600 ;; It's not set if we've initially skipped over comments and line
1601 ;; continuations since we might have gone out through the end of a
1602 ;; macro then. This provision makes `c-forward-sws' not populate the
1603 ;; cache in the majority of cases, but otoh is `c-backward-sws' by far
1604 ;; more common.
1605 safe-start)
1606
1607 ;; Skip simple ws and do a quick check on the following character to see
1608 ;; if it's anything that can't start syntactic ws, so we can bail out
1609 ;; early in the majority of cases when there just are a few ws chars.
1610 (skip-chars-forward " \t\n\r\f\v")
1611 (when (looking-at c-syntactic-ws-start)
1612
1613 (setq rung-end-pos (min (1+ (point)) (point-max)))
1614 (if (setq rung-is-marked (text-property-any rung-pos rung-end-pos
1615 'c-is-sws t))
1616 ;; Find the last rung position to avoid setting properties in all
1617 ;; the cases when the marked rung is complete.
1618 ;; (`next-single-property-change' is certain to move at least one
1619 ;; step forward.)
1620 (setq rung-pos (1- (next-single-property-change
1621 rung-is-marked 'c-is-sws nil rung-end-pos)))
1622 ;; Got no marked rung here. Since the simple ws might have started
1623 ;; inside a line comment or cpp directive we must set `rung-pos' as
1624 ;; high as possible.
1625 (setq rung-pos (point)))
1626
1627 (while
1628 (progn
1629 (while
1630 (when (and rung-is-marked
1631 (get-text-property (point) 'c-in-sws))
1632
1633 ;; The following search is the main reason that `c-in-sws'
1634 ;; and `c-is-sws' aren't combined to one property.
1635 (goto-char (next-single-property-change
1636 (point) 'c-in-sws nil (point-max)))
1637 (unless (get-text-property (point) 'c-is-sws)
1638 ;; If the `c-in-sws' region extended past the last
1639 ;; `c-is-sws' char we have to go back a bit.
1640 (or (get-text-property (1- (point)) 'c-is-sws)
1641 (goto-char (previous-single-property-change
1642 (point) 'c-is-sws)))
1643 (backward-char))
1644
1645 (c-debug-sws-msg
1646 "c-forward-sws cached move %s -> %s (max %s)"
1647 rung-pos (point) (point-max))
1648
1649 (setq rung-pos (point))
1650 (and (> (skip-chars-forward " \t\n\r\f\v") 0)
1651 (not (eobp))))
1652
1653 ;; We'll loop here if there is simple ws after the last rung.
1654 ;; That means that there's been some change in it and it's
1655 ;; possible that we've stepped into another ladder, so extend
1656 ;; the previous one to join with it if there is one, and try to
1657 ;; use the cache again.
1658 (c-debug-sws-msg
1659 "c-forward-sws extending rung with [%s..%s] (max %s)"
1660 (1+ rung-pos) (1+ (point)) (point-max))
1661 (unless (get-text-property (point) 'c-is-sws)
1662 ;; Remove any `c-in-sws' property from the last char of
1663 ;; the rung before we mark it with `c-is-sws', so that we
1664 ;; won't connect with the remains of a broken "ladder".
1665 (c-remove-in-sws (point) (1+ (point))))
1666 (c-put-is-sws (1+ rung-pos)
1667 (1+ (point)))
1668 (c-put-in-sws rung-pos
1669 (setq rung-pos (point)
1670 last-put-in-sws-pos rung-pos)))
1671
1672 (setq simple-ws-end (point))
1673 (c-forward-comments)
1674
1675 (cond
1676 ((/= (point) simple-ws-end)
1677 ;; Skipped over comments. Don't cache at eob in case the buffer
1678 ;; is narrowed.
1679 (not (eobp)))
1680
1681 ((save-excursion
1682 (and c-opt-cpp-prefix
1683 (looking-at c-opt-cpp-start)
1684 (progn (skip-chars-backward " \t")
1685 (bolp))
1686 (or (bobp)
1687 (progn (backward-char)
1688 (not (eq (char-before) ?\\))))))
1689 ;; Skip a preprocessor directive.
1690 (end-of-line)
1691 (while (and (eq (char-before) ?\\)
1692 (= (forward-line 1) 0))
1693 (end-of-line))
1694 (forward-line 1)
1695 (setq safe-start t)
1696 ;; Don't cache at eob in case the buffer is narrowed.
1697 (not (eobp)))))
1698
1699 ;; We've searched over a piece of non-white syntactic ws. See if this
1700 ;; can be cached.
1701 (setq next-rung-pos (point))
1702 (skip-chars-forward " \t\n\r\f\v")
1703 (setq rung-end-pos (min (1+ (point)) (point-max)))
1704
1705 (if (or
1706 ;; Cache if we haven't skipped comments only, and if we started
1707 ;; either from a marked rung or from a completely uncached
1708 ;; position.
1709 (and safe-start
1710 (or rung-is-marked
1711 (not (get-text-property simple-ws-end 'c-in-sws))))
1712
1713 ;; See if there's a marked rung in the encountered simple ws. If
1714 ;; so then we can cache, unless `safe-start' is nil. Even then
1715 ;; we need to do this to check if the cache can be used for the
1716 ;; next step.
1717 (and (setq next-rung-is-marked
1718 (text-property-any next-rung-pos rung-end-pos
1719 'c-is-sws t))
1720 safe-start))
b2acd789 1721
0ec8351b 1722 (progn
d9e94c22
MS
1723 (c-debug-sws-msg
1724 "c-forward-sws caching [%s..%s] - [%s..%s] (max %s)"
1725 rung-pos (1+ simple-ws-end) next-rung-pos rung-end-pos
1726 (point-max))
1727
1728 ;; Remove the properties for any nested ws that might be cached.
1729 ;; Only necessary for `c-is-sws' since `c-in-sws' will be set
1730 ;; anyway.
1731 (c-remove-is-sws (1+ simple-ws-end) next-rung-pos)
1732 (unless (and rung-is-marked (= rung-pos simple-ws-end))
1733 (c-put-is-sws rung-pos
1734 (1+ simple-ws-end))
1735 (setq rung-is-marked t))
1736 (c-put-in-sws rung-pos
1737 (setq rung-pos (point)
1738 last-put-in-sws-pos rung-pos))
1739 (unless (get-text-property (1- rung-end-pos) 'c-is-sws)
1740 ;; Remove any `c-in-sws' property from the last char of
1741 ;; the rung before we mark it with `c-is-sws', so that we
1742 ;; won't connect with the remains of a broken "ladder".
1743 (c-remove-in-sws (1- rung-end-pos) rung-end-pos))
1744 (c-put-is-sws next-rung-pos
1745 rung-end-pos))
1746
1747 (c-debug-sws-msg
1748 "c-forward-sws not caching [%s..%s] - [%s..%s] (max %s)"
1749 rung-pos (1+ simple-ws-end) next-rung-pos rung-end-pos
1750 (point-max))
1751
1752 ;; Set `rung-pos' for the next rung. It's the same thing here as
1753 ;; initially, except that the rung position is set as early as
1754 ;; possible since we can't be in the ending ws of a line comment or
1755 ;; cpp directive now.
1756 (if (setq rung-is-marked next-rung-is-marked)
1757 (setq rung-pos (1- (next-single-property-change
1758 rung-is-marked 'c-is-sws nil rung-end-pos)))
1759 (setq rung-pos next-rung-pos))
1760 (setq safe-start t)))
1761
1762 ;; Make sure that the newly marked `c-in-sws' region doesn't connect to
1763 ;; another one after the point (which might occur when editing inside a
1764 ;; comment or macro).
1765 (when (eq last-put-in-sws-pos (point))
1766 (cond ((< last-put-in-sws-pos (point-max))
1767 (c-debug-sws-msg
1768 "c-forward-sws clearing at %s for cache separation"
1769 last-put-in-sws-pos)
1770 (c-remove-in-sws last-put-in-sws-pos
1771 (1+ last-put-in-sws-pos)))
1772 (t
1773 ;; If at eob we have to clear the last character before the end
1774 ;; instead since the buffer might be narrowed and there might
1775 ;; be a `c-in-sws' after (point-max). In this case it's
1776 ;; necessary to clear both properties.
1777 (c-debug-sws-msg
1778 "c-forward-sws clearing thoroughly at %s for cache separation"
1779 (1- last-put-in-sws-pos))
1780 (c-remove-is-and-in-sws (1- last-put-in-sws-pos)
1781 last-put-in-sws-pos))))
1782 )))
b2acd789 1783
d9e94c22
MS
1784(defun c-backward-sws ()
1785 ;; Used by `c-backward-syntactic-ws' to implement the unbounded search.
0386b551
AM
1786 ;;
1787 ;; This function might do hidden buffer changes.
d9e94c22
MS
1788
1789 (let (;; `rung-pos' is set to a position as late as possible in the unmarked
1790 ;; part of the simple ws region.
1791 (rung-pos (point)) next-rung-pos last-put-in-sws-pos
1792 rung-is-marked simple-ws-beg cmt-skip-pos)
1793
1794 ;; Skip simple horizontal ws and do a quick check on the preceding
91af3942 1795 ;; character to see if it's anything that can't end syntactic ws, so we can
d9e94c22
MS
1796 ;; bail out early in the majority of cases when there just are a few ws
1797 ;; chars. Newlines are complicated in the backward direction, so we can't
1798 ;; skip over them.
1799 (skip-chars-backward " \t\f")
1800 (when (and (not (bobp))
1801 (save-excursion
1802 (backward-char)
1803 (looking-at c-syntactic-ws-end)))
1804
1805 ;; Try to find a rung position in the simple ws preceding point, so that
1806 ;; we can get a cache hit even if the last bit of the simple ws has
1807 ;; changed recently.
1808 (setq simple-ws-beg (point))
1809 (skip-chars-backward " \t\n\r\f\v")
1810 (if (setq rung-is-marked (text-property-any
1811 (point) (min (1+ rung-pos) (point-max))
1812 'c-is-sws t))
1813 ;; `rung-pos' will be the earliest marked position, which means that
1814 ;; there might be later unmarked parts in the simple ws region.
1815 ;; It's not worth the effort to fix that; the last part of the
1816 ;; simple ws is also typically edited often, so it could be wasted.
1817 (goto-char (setq rung-pos rung-is-marked))
1818 (goto-char simple-ws-beg))
1819
1820 (while
1821 (progn
1822 (while
1823 (when (and rung-is-marked
1824 (not (bobp))
1825 (get-text-property (1- (point)) 'c-in-sws))
1826
1827 ;; The following search is the main reason that `c-in-sws'
1828 ;; and `c-is-sws' aren't combined to one property.
1829 (goto-char (previous-single-property-change
1830 (point) 'c-in-sws nil (point-min)))
1831 (unless (get-text-property (point) 'c-is-sws)
1832 ;; If the `c-in-sws' region extended past the first
1833 ;; `c-is-sws' char we have to go forward a bit.
1834 (goto-char (next-single-property-change
1835 (point) 'c-is-sws)))
1836
1837 (c-debug-sws-msg
1838 "c-backward-sws cached move %s <- %s (min %s)"
1839 (point) rung-pos (point-min))
1840
1841 (setq rung-pos (point))
1842 (if (and (< (min (skip-chars-backward " \t\f\v")
1843 (progn
1844 (setq simple-ws-beg (point))
1845 (skip-chars-backward " \t\n\r\f\v")))
1846 0)
1847 (setq rung-is-marked
1848 (text-property-any (point) rung-pos
1849 'c-is-sws t)))
1850 t
1851 (goto-char simple-ws-beg)
1852 nil))
1853
1854 ;; We'll loop here if there is simple ws before the first rung.
1855 ;; That means that there's been some change in it and it's
1856 ;; possible that we've stepped into another ladder, so extend
1857 ;; the previous one to join with it if there is one, and try to
1858 ;; use the cache again.
1859 (c-debug-sws-msg
1860 "c-backward-sws extending rung with [%s..%s] (min %s)"
1861 rung-is-marked rung-pos (point-min))
1862 (unless (get-text-property (1- rung-pos) 'c-is-sws)
1863 ;; Remove any `c-in-sws' property from the last char of
1864 ;; the rung before we mark it with `c-is-sws', so that we
1865 ;; won't connect with the remains of a broken "ladder".
1866 (c-remove-in-sws (1- rung-pos) rung-pos))
1867 (c-put-is-sws rung-is-marked
1868 rung-pos)
1869 (c-put-in-sws rung-is-marked
1870 (1- rung-pos))
1871 (setq rung-pos rung-is-marked
1872 last-put-in-sws-pos rung-pos))
1873
1874 (c-backward-comments)
1875 (setq cmt-skip-pos (point))
a66cd3ee 1876
d9e94c22
MS
1877 (cond
1878 ((and c-opt-cpp-prefix
1879 (/= cmt-skip-pos simple-ws-beg)
1880 (c-beginning-of-macro))
1881 ;; Inside a cpp directive. See if it should be skipped over.
1882 (let ((cpp-beg (point)))
1883
1884 ;; Move back over all line continuations in the region skipped
1885 ;; over by `c-backward-comments'. If we go past it then we
1886 ;; started inside the cpp directive.
1887 (goto-char simple-ws-beg)
1888 (beginning-of-line)
1889 (while (and (> (point) cmt-skip-pos)
1890 (progn (backward-char)
1891 (eq (char-before) ?\\)))
1892 (beginning-of-line))
1893
1894 (if (< (point) cmt-skip-pos)
1895 ;; Don't move past the cpp directive if we began inside
1896 ;; it. Note that the position at the end of the last line
1897 ;; of the macro is also considered to be within it.
1898 (progn (goto-char cmt-skip-pos)
1899 nil)
1900
1901 ;; It's worthwhile to spend a little bit of effort on finding
1902 ;; the end of the macro, to get a good `simple-ws-beg'
1903 ;; position for the cache. Note that `c-backward-comments'
1904 ;; could have stepped over some comments before going into
1905 ;; the macro, and then `simple-ws-beg' must be kept on the
1906 ;; same side of those comments.
1907 (goto-char simple-ws-beg)
1908 (skip-chars-backward " \t\n\r\f\v")
1909 (if (eq (char-before) ?\\)
1910 (forward-char))
1911 (forward-line 1)
1912 (if (< (point) simple-ws-beg)
1913 ;; Might happen if comments after the macro were skipped
1914 ;; over.
1915 (setq simple-ws-beg (point)))
1916
1917 (goto-char cpp-beg)
1918 t)))
1919
1920 ((/= (save-excursion
1921 (skip-chars-forward " \t\n\r\f\v" simple-ws-beg)
1922 (setq next-rung-pos (point)))
1923 simple-ws-beg)
1924 ;; Skipped over comments. Must put point at the end of
1925 ;; the simple ws at point since we might be after a line
1926 ;; comment or cpp directive that's been partially
1927 ;; narrowed out, and we can't risk marking the simple ws
1928 ;; at the end of it.
1929 (goto-char next-rung-pos)
1930 t)))
1931
1932 ;; We've searched over a piece of non-white syntactic ws. See if this
1933 ;; can be cached.
1934 (setq next-rung-pos (point))
1935 (skip-chars-backward " \t\f\v")
1936
1937 (if (or
1938 ;; Cache if we started either from a marked rung or from a
1939 ;; completely uncached position.
1940 rung-is-marked
1941 (not (get-text-property (1- simple-ws-beg) 'c-in-sws))
1942
1943 ;; Cache if there's a marked rung in the encountered simple ws.
1944 (save-excursion
1945 (skip-chars-backward " \t\n\r\f\v")
1946 (text-property-any (point) (min (1+ next-rung-pos) (point-max))
1947 'c-is-sws t)))
a66cd3ee 1948
d9e94c22
MS
1949 (progn
1950 (c-debug-sws-msg
1951 "c-backward-sws caching [%s..%s] - [%s..%s] (min %s)"
1952 (point) (1+ next-rung-pos)
1953 simple-ws-beg (min (1+ rung-pos) (point-max))
1954 (point-min))
1955
1956 ;; Remove the properties for any nested ws that might be cached.
1957 ;; Only necessary for `c-is-sws' since `c-in-sws' will be set
1958 ;; anyway.
1959 (c-remove-is-sws (1+ next-rung-pos) simple-ws-beg)
1960 (unless (and rung-is-marked (= simple-ws-beg rung-pos))
1961 (let ((rung-end-pos (min (1+ rung-pos) (point-max))))
1962 (unless (get-text-property (1- rung-end-pos) 'c-is-sws)
1963 ;; Remove any `c-in-sws' property from the last char of
1964 ;; the rung before we mark it with `c-is-sws', so that we
1965 ;; won't connect with the remains of a broken "ladder".
1966 (c-remove-in-sws (1- rung-end-pos) rung-end-pos))
1967 (c-put-is-sws simple-ws-beg
1968 rung-end-pos)
1969 (setq rung-is-marked t)))
1970 (c-put-in-sws (setq simple-ws-beg (point)
1971 last-put-in-sws-pos simple-ws-beg)
1972 rung-pos)
1973 (c-put-is-sws (setq rung-pos simple-ws-beg)
1974 (1+ next-rung-pos)))
1975
1976 (c-debug-sws-msg
1977 "c-backward-sws not caching [%s..%s] - [%s..%s] (min %s)"
1978 (point) (1+ next-rung-pos)
1979 simple-ws-beg (min (1+ rung-pos) (point-max))
1980 (point-min))
1981 (setq rung-pos next-rung-pos
1982 simple-ws-beg (point))
1983 ))
1984
1985 ;; Make sure that the newly marked `c-in-sws' region doesn't connect to
1986 ;; another one before the point (which might occur when editing inside a
1987 ;; comment or macro).
1988 (when (eq last-put-in-sws-pos (point))
1989 (cond ((< (point-min) last-put-in-sws-pos)
1990 (c-debug-sws-msg
1991 "c-backward-sws clearing at %s for cache separation"
1992 (1- last-put-in-sws-pos))
1993 (c-remove-in-sws (1- last-put-in-sws-pos)
1994 last-put-in-sws-pos))
1995 ((> (point-min) 1)
1996 ;; If at bob and the buffer is narrowed, we have to clear the
1997 ;; character we're standing on instead since there might be a
1998 ;; `c-in-sws' before (point-min). In this case it's necessary
1999 ;; to clear both properties.
2000 (c-debug-sws-msg
2001 "c-backward-sws clearing thoroughly at %s for cache separation"
2002 last-put-in-sws-pos)
2003 (c-remove-is-and-in-sws last-put-in-sws-pos
2004 (1+ last-put-in-sws-pos)))))
2005 )))
785eecbb 2006
d9e94c22 2007\f
580fba94
AM
2008;; Other whitespace tools
2009(defun c-partial-ws-p (beg end)
2010 ;; Is the region (beg end) WS, and is there WS (or BOB/EOB) next to the
2011 ;; region? This is a "heuristic" function. .....
b414f371 2012 ;;
c421028a
AM
2013 ;; The motivation for the second bit is to check whether removing this
2014 ;; region would coalesce two symbols.
580fba94
AM
2015 ;;
2016 ;; FIXME!!! This function doesn't check virtual semicolons in any way. Be
2017 ;; careful about using this function for, e.g. AWK. (2007/3/7)
2018 (save-excursion
2019 (let ((end+1 (min (1+ end) (point-max))))
2020 (or (progn (goto-char (max (point-min) (1- beg)))
2021 (c-skip-ws-forward end)
2022 (eq (point) end))
2023 (progn (goto-char beg)
2024 (c-skip-ws-forward end+1)
2025 (eq (point) end+1))))))
2026\f
0386b551 2027;; A system for finding noteworthy parens before the point.
e1c458ae 2028
0ec1d2c5
AM
2029(defconst c-state-cache-too-far 5000)
2030;; A maximum comfortable scanning distance, e.g. between
2031;; `c-state-cache-good-pos' and "HERE" (where we call c-parse-state). When
0d26e0b6 2032;; this distance is exceeded, we take "emergency measures", e.g. by clearing
0ec1d2c5
AM
2033;; the cache and starting again from point-min or a beginning of defun. This
2034;; value can be tuned for efficiency or set to a lower value for testing.
2035
d9e94c22
MS
2036(defvar c-state-cache nil)
2037(make-variable-buffer-local 'c-state-cache)
2038;; The state cache used by `c-parse-state' to cut down the amount of
0ec1d2c5
AM
2039;; searching. It's the result from some earlier `c-parse-state' call. See
2040;; `c-parse-state''s doc string for details of its structure.
0386b551 2041;;
d9e94c22
MS
2042;; The use of the cached info is more effective if the next
2043;; `c-parse-state' call is on a line close by the one the cached state
2044;; was made at; the cache can actually slow down a little if the
2045;; cached state was made very far back in the buffer. The cache is
2046;; most effective if `c-parse-state' is used on each line while moving
2047;; forward.
e1c458ae 2048
0386b551
AM
2049(defvar c-state-cache-good-pos 1)
2050(make-variable-buffer-local 'c-state-cache-good-pos)
0ec1d2c5
AM
2051;; This is a position where `c-state-cache' is known to be correct, or
2052;; nil (see below). It's a position inside one of the recorded unclosed
2053;; parens or the top level, but not further nested inside any literal or
2054;; subparen that is closed before the last recorded position.
0386b551
AM
2055;;
2056;; The exact position is chosen to try to be close to yet earlier than
2057;; the position where `c-state-cache' will be called next. Right now
2058;; the heuristic is to set it to the position after the last found
2059;; closing paren (of any type) before the line on which
2060;; `c-parse-state' was called. That is chosen primarily to work well
2061;; with refontification of the current line.
0ec1d2c5
AM
2062;;
2063;; 2009-07-28: When `c-state-point-min' and the last position where
2064;; `c-parse-state' or for which `c-invalidate-state-cache' was called, are
2065;; both in the same literal, there is no such "good position", and
2066;; c-state-cache-good-pos is then nil. This is the ONLY circumstance in which
2067;; it can be nil. In this case, `c-state-point-min-literal' will be non-nil.
2068;;
2069;; 2009-06-12: In a brace desert, c-state-cache-good-pos may also be in
2070;; the middle of the desert, as long as it is not within a brace pair
2071;; recorded in `c-state-cache' or a paren/bracket pair.
2072
2073
2074;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2075;; We maintain a simple cache of positions which aren't in a literal, so as to
2076;; speed up testing for non-literality.
2077(defconst c-state-nonlit-pos-interval 10000)
2078;; The approximate interval between entries in `c-state-nonlit-pos-cache'.
2079
2080(defvar c-state-nonlit-pos-cache nil)
2081(make-variable-buffer-local 'c-state-nonlit-pos-cache)
e84efb70
AM
2082;; A list of buffer positions which are known not to be in a literal or a cpp
2083;; construct. This is ordered with higher positions at the front of the list.
2084;; Only those which are less than `c-state-nonlit-pos-cache-limit' are valid.
0ec1d2c5
AM
2085
2086(defvar c-state-nonlit-pos-cache-limit 1)
2087(make-variable-buffer-local 'c-state-nonlit-pos-cache-limit)
2088;; An upper limit on valid entries in `c-state-nonlit-pos-cache'. This is
2089;; reduced by buffer changes, and increased by invocations of
2090;; `c-state-literal-at'.
2091
2092(defsubst c-state-pp-to-literal (from to)
13d49cbb
AM
2093 ;; Do a parse-partial-sexp from FROM to TO, returning either
2094 ;; (STATE TYPE (BEG . END)) if TO is in a literal; or
2095 ;; (STATE) otherwise,
2096 ;; where STATE is the parsing state at TO, TYPE is the type of the literal
2097 ;; (one of 'c, 'c++, 'string) and (BEG . END) is the boundaries of the literal.
2098 ;;
2099 ;; Only elements 3 (in a string), 4 (in a comment), 5 (following a quote),
2100 ;; 7 (comment type) and 8 (start of comment/string) (and possibly 9) of
2101 ;; STATE are valid.
0ec1d2c5 2102 (save-excursion
13d49cbb
AM
2103 (let ((s (parse-partial-sexp from to))
2104 ty)
0ec1d2c5 2105 (when (or (nth 3 s) (nth 4 s)) ; in a string or comment
13d49cbb
AM
2106 (setq ty (cond
2107 ((nth 3 s) 'string)
2108 ((eq (nth 7 s) t) 'c++)
2109 (t 'c)))
0ec1d2c5
AM
2110 (parse-partial-sexp (point) (point-max)
2111 nil ; TARGETDEPTH
2112 nil ; STOPBEFORE
2113 s ; OLDSTATE
13d49cbb
AM
2114 'syntax-table)) ; stop at end of literal
2115 (if ty
2116 `(,s ,ty (,(nth 8 s) . ,(point)))
2117 `(,s)))))
2118
2119(defun c-state-safe-place (here)
2120 ;; Return a buffer position before HERE which is "safe", i.e. outside any
2121 ;; string, comment, or macro.
e84efb70
AM
2122 ;;
2123 ;; NOTE: This function manipulates `c-state-nonlit-pos-cache'. This cache
2124 ;; MAY NOT contain any positions within macros, since macros are frequently
2125 ;; turned into comments by use of the `c-cpp-delimiter' category properties.
2126 ;; We cannot rely on this mechanism whilst determining a cache pos since
2127 ;; this function is also called from outwith `c-parse-state'.
0ec1d2c5
AM
2128 (save-restriction
2129 (widen)
2130 (save-excursion
2131 (let ((c c-state-nonlit-pos-cache)
2132 pos npos lit)
2133 ;; Trim the cache to take account of buffer changes.
2134 (while (and c (> (car c) c-state-nonlit-pos-cache-limit))
2135 (setq c (cdr c)))
2136 (setq c-state-nonlit-pos-cache c)
2137
2138 (while (and c (> (car c) here))
2139 (setq c (cdr c)))
2140 (setq pos (or (car c) (point-min)))
2141
2142 (while (<= (setq npos (+ pos c-state-nonlit-pos-interval))
2143 here)
13d49cbb 2144 (setq lit (car (cddr (c-state-pp-to-literal pos npos))))
0ec1d2c5 2145 (setq pos (or (cdr lit) npos)) ; end of literal containing npos.
e84efb70
AM
2146 (goto-char pos)
2147 (when (and (c-beginning-of-macro) (/= (point) pos))
2148 (c-syntactic-end-of-macro)
2149 (or (eobp) (forward-char))
2150 (setq pos (point)))
0ec1d2c5
AM
2151 (setq c-state-nonlit-pos-cache (cons pos c-state-nonlit-pos-cache)))
2152
2153 (if (> pos c-state-nonlit-pos-cache-limit)
2154 (setq c-state-nonlit-pos-cache-limit pos))
13d49cbb 2155 pos))))
9858f6c3 2156
13d49cbb
AM
2157(defun c-state-literal-at (here)
2158 ;; If position HERE is inside a literal, return (START . END), the
2159 ;; boundaries of the literal (which may be outside the accessible bit of the
2160 ;; buffer). Otherwise, return nil.
2161 ;;
2162 ;; This function is almost the same as `c-literal-limits'. Previously, it
9858f6c3 2163 ;; differed in that it was a lower level function, and that it rigorously
13d49cbb
AM
2164 ;; followed the syntax from BOB. `c-literal-limits' is now (2011-12)
2165 ;; virtually identical to this function.
2166 (save-restriction
2167 (widen)
2168 (save-excursion
2169 (let ((pos (c-state-safe-place here)))
2170 (car (cddr (c-state-pp-to-literal pos here)))))))
0ec1d2c5
AM
2171
2172(defsubst c-state-lit-beg (pos)
2173 ;; Return the start of the literal containing POS, or POS itself.
2174 (or (car (c-state-literal-at pos))
2175 pos))
2176
2177(defsubst c-state-cache-non-literal-place (pos state)
9c4aeabf 2178 ;; Return a position outside of a string/comment/macro at or before POS.
0ec1d2c5 2179 ;; STATE is the parse-partial-sexp state at POS.
9c4aeabf
AM
2180 (let ((res (if (or (nth 3 state) ; in a string?
2181 (nth 4 state)) ; in a comment?
2182 (nth 8 state)
2183 pos)))
2184 (save-excursion
2185 (goto-char res)
2186 (if (c-beginning-of-macro)
2187 (point)
2188 res))))
0ec1d2c5
AM
2189
2190;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2191;; Stuff to do with point-min, and coping with any literal there.
2192(defvar c-state-point-min 1)
2193(make-variable-buffer-local 'c-state-point-min)
2194;; This is (point-min) when `c-state-cache' was last calculated. A change of
2195;; narrowing is likely to affect the parens that are visible before the point.
2196
2197(defvar c-state-point-min-lit-type nil)
2198(make-variable-buffer-local 'c-state-point-min-lit-type)
2199(defvar c-state-point-min-lit-start nil)
2200(make-variable-buffer-local 'c-state-point-min-lit-start)
2201;; These two variables define the literal, if any, containing point-min.
2202;; Their values are, respectively, 'string, c, or c++, and the start of the
2203;; literal. If there's no literal there, they're both nil.
2204
2205(defvar c-state-min-scan-pos 1)
2206(make-variable-buffer-local 'c-state-min-scan-pos)
2207;; This is the earliest buffer-pos from which scanning can be done. It is
2208;; either the end of the literal containing point-min, or point-min itself.
2209;; It becomes nil if the buffer is changed earlier than this point.
2210(defun c-state-get-min-scan-pos ()
2211 ;; Return the lowest valid scanning pos. This will be the end of the
2212 ;; literal enclosing point-min, or point-min itself.
2213 (or c-state-min-scan-pos
2214 (save-restriction
2215 (save-excursion
2216 (widen)
2217 (goto-char c-state-point-min-lit-start)
2218 (if (eq c-state-point-min-lit-type 'string)
2219 (forward-sexp)
2220 (forward-comment 1))
2221 (setq c-state-min-scan-pos (point))))))
2222
2223(defun c-state-mark-point-min-literal ()
2224 ;; Determine the properties of any literal containing POINT-MIN, setting the
2225 ;; variables `c-state-point-min-lit-type', `c-state-point-min-lit-start',
2226 ;; and `c-state-min-scan-pos' accordingly. The return value is meaningless.
2227 (let ((p-min (point-min))
2228 lit)
2229 (save-restriction
2230 (widen)
2231 (setq lit (c-state-literal-at p-min))
2232 (if lit
2233 (setq c-state-point-min-lit-type
2234 (save-excursion
2235 (goto-char (car lit))
2236 (cond
2237 ((looking-at c-block-comment-start-regexp) 'c)
2238 ((looking-at c-line-comment-starter) 'c++)
2239 (t 'string)))
2240 c-state-point-min-lit-start (car lit)
2241 c-state-min-scan-pos (cdr lit))
2242 (setq c-state-point-min-lit-type nil
2243 c-state-point-min-lit-start nil
2244 c-state-min-scan-pos p-min)))))
2245
2246
2247;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2248;; A variable which signals a brace dessert - helpful for reducing the number
2249;; of fruitless backward scans.
2250(defvar c-state-brace-pair-desert nil)
2251(make-variable-buffer-local 'c-state-brace-pair-desert)
e84efb70 2252;; Used only in `c-append-lower-brace-pair-to-state-cache'. It is set when
0ec1d2c5
AM
2253;; that defun has searched backwards for a brace pair and not found one. Its
2254;; value is either nil or a cons (PA . FROM), where PA is the position of the
2255;; enclosing opening paren/brace/bracket which bounds the backwards search (or
2256;; nil when at top level) and FROM is where the backward search started. It
2257;; is reset to nil in `c-invalidate-state-cache'.
2258
2259
2260;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2261;; Lowish level functions/macros which work directly on `c-state-cache', or a
2262;; list of like structure.
2263(defmacro c-state-cache-top-lparen (&optional cache)
2264 ;; Return the address of the top left brace/bracket/paren recorded in CACHE
2265 ;; (default `c-state-cache') (or nil).
2266 (let ((cash (or cache 'c-state-cache)))
2267 `(if (consp (car ,cash))
2268 (caar ,cash)
2269 (car ,cash))))
2270
2271(defmacro c-state-cache-top-paren (&optional cache)
2272 ;; Return the address of the latest brace/bracket/paren (whether left or
2273 ;; right) recorded in CACHE (default `c-state-cache') or nil.
2274 (let ((cash (or cache 'c-state-cache)))
2275 `(if (consp (car ,cash))
2276 (cdar ,cash)
2277 (car ,cash))))
2278
2279(defmacro c-state-cache-after-top-paren (&optional cache)
2280 ;; Return the position just after the latest brace/bracket/paren (whether
2281 ;; left or right) recorded in CACHE (default `c-state-cache') or nil.
2282 (let ((cash (or cache 'c-state-cache)))
2283 `(if (consp (car ,cash))
2284 (cdar ,cash)
2285 (and (car ,cash)
2286 (1+ (car ,cash))))))
2287
2288(defun c-get-cache-scan-pos (here)
2289 ;; From the state-cache, determine the buffer position from which we might
2290 ;; scan forward to HERE to update this cache. This position will be just
2291 ;; after a paren/brace/bracket recorded in the cache, if possible, otherwise
2292 ;; return the earliest position in the accessible region which isn't within
2293 ;; a literal. If the visible portion of the buffer is entirely within a
2294 ;; literal, return NIL.
2295 (let ((c c-state-cache) elt)
2296 ;(while (>= (or (c-state-cache-top-lparen c) 1) here)
2297 (while (and c
2298 (>= (c-state-cache-top-lparen c) here))
2299 (setq c (cdr c)))
2300
2301 (setq elt (car c))
2302 (cond
2303 ((consp elt)
2304 (if (> (cdr elt) here)
2305 (1+ (car elt))
2306 (cdr elt)))
2307 (elt (1+ elt))
2308 ((<= (c-state-get-min-scan-pos) here)
2309 (c-state-get-min-scan-pos))
2310 (t nil))))
2311
2312;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2313;; Variables which keep track of preprocessor constructs.
2314(defvar c-state-old-cpp-beg nil)
2315(make-variable-buffer-local 'c-state-old-cpp-beg)
2316(defvar c-state-old-cpp-end nil)
2317(make-variable-buffer-local 'c-state-old-cpp-end)
2318;; These are the limits of the macro containing point at the previous call of
2319;; `c-parse-state', or nil.
2320
2321;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
c7015153 2322;; Defuns which analyze the buffer, yet don't change `c-state-cache'.
0ec1d2c5
AM
2323(defun c-get-fallback-scan-pos (here)
2324 ;; Return a start position for building `c-state-cache' from
2325 ;; scratch. This will be at the top level, 2 defuns back.
0386b551
AM
2326 (save-excursion
2327 ;; Go back 2 bods, but ignore any bogus positions returned by
2328 ;; beginning-of-defun (i.e. open paren in column zero).
2329 (goto-char here)
2330 (let ((cnt 2))
2331 (while (not (or (bobp) (zerop cnt)))
0ec1d2c5 2332 (c-beginning-of-defun-1) ; Pure elisp BOD.
0386b551
AM
2333 (if (eq (char-after) ?\{)
2334 (setq cnt (1- cnt)))))
2335 (point)))
2336
c0209c2c
AM
2337(defun c-state-balance-parens-backwards (here- here+ top)
2338 ;; Return the position of the opening paren/brace/bracket before HERE- which
2339 ;; matches the outermost close p/b/b between HERE+ and TOP. Except when
2340 ;; there's a macro, HERE- and HERE+ are the same. Like this:
0ec1d2c5 2341 ;;
c0209c2c
AM
2342 ;; ............................................
2343 ;; | |
2344 ;; ( [ ( .........#macro.. ) ( ) ] )
2345 ;; ^ ^ ^ ^
2346 ;; | | | |
2347 ;; return HERE- HERE+ TOP
0ec1d2c5
AM
2348 ;;
2349 ;; If there aren't enough opening paren/brace/brackets, return the position
c0209c2c 2350 ;; of the outermost one found, or HERE- if there are none. If there are no
da6062e6 2351 ;; closing p/b/bs between HERE+ and TOP, return HERE-. HERE-/+ and TOP
c0209c2c
AM
2352 ;; must not be inside literals. Only the accessible portion of the buffer
2353 ;; will be scanned.
2354
2355 ;; PART 1: scan from `here+' up to `top', accumulating ")"s which enclose
2356 ;; `here'. Go round the next loop each time we pass over such a ")". These
2357 ;; probably match "("s before `here-'.
0ec1d2c5
AM
2358 (let (pos pa ren+1 lonely-rens)
2359 (save-excursion
2360 (save-restriction
2361 (narrow-to-region (point-min) top) ; This can move point, sometimes.
c0209c2c 2362 (setq pos here+)
0ec1d2c5
AM
2363 (c-safe
2364 (while
2365 (setq ren+1 (scan-lists pos 1 1)) ; might signal
2366 (setq lonely-rens (cons ren+1 lonely-rens)
2367 pos ren+1)))))
2368
c0209c2c 2369 ;; PART 2: Scan back before `here-' searching for the "("s
0ec1d2c5
AM
2370 ;; matching/mismatching the ")"s found above. We only need to direct the
2371 ;; caller to scan when we've encountered unmatched right parens.
c0209c2c
AM
2372 (setq pos here-)
2373 (when lonely-rens
2374 (c-safe
2375 (while
2376 (and lonely-rens ; actual values aren't used.
2377 (setq pa (scan-lists pos -1 1)))
2378 (setq pos pa)
2379 (setq lonely-rens (cdr lonely-rens)))))
2380 pos))
0ec1d2c5
AM
2381
2382(defun c-parse-state-get-strategy (here good-pos)
2383 ;; Determine the scanning strategy for adjusting `c-parse-state', attempting
53964682 2384 ;; to minimize the amount of scanning. HERE is the pertinent position in
0ec1d2c5
AM
2385 ;; the buffer, GOOD-POS is a position where `c-state-cache' (possibly with
2386 ;; its head trimmed) is known to be good, or nil if there is no such
2387 ;; position.
2388 ;;
2389 ;; The return value is a list, one of the following:
2390 ;;
2391 ;; o - ('forward CACHE-POS START-POINT) - scan forward from START-POINT,
2392 ;; which is not less than CACHE-POS.
2393 ;; o - ('backward CACHE-POS nil) - scan backwards (from HERE).
2394 ;; o - ('BOD nil START-POINT) - scan forwards from START-POINT, which is at the
2395 ;; top level.
2396 ;; o - ('IN-LIT nil nil) - point is inside the literal containing point-min.
2397 ;; , where CACHE-POS is the highest position recorded in `c-state-cache' at
2398 ;; or below HERE.
2399 (let ((cache-pos (c-get-cache-scan-pos here)) ; highest position below HERE in cache (or 1)
2400 BOD-pos ; position of 2nd BOD before HERE.
2401 strategy ; 'forward, 'backward, 'BOD, or 'IN-LIT.
2402 start-point
2403 how-far) ; putative scanning distance.
2404 (setq good-pos (or good-pos (c-state-get-min-scan-pos)))
2405 (cond
2406 ((< here (c-state-get-min-scan-pos))
2407 (setq strategy 'IN-LIT
2408 start-point nil
2409 cache-pos nil
2410 how-far 0))
2411 ((<= good-pos here)
2412 (setq strategy 'forward
2413 start-point (max good-pos cache-pos)
2414 how-far (- here start-point)))
2415 ((< (- good-pos here) (- here cache-pos)) ; FIXME!!! ; apply some sort of weighting.
2416 (setq strategy 'backward
2417 how-far (- good-pos here)))
2418 (t
2419 (setq strategy 'forward
2420 how-far (- here cache-pos)
2421 start-point cache-pos)))
2422
2423 ;; Might we be better off starting from the top level, two defuns back,
2424 ;; instead?
2425 (when (> how-far c-state-cache-too-far)
2426 (setq BOD-pos (c-get-fallback-scan-pos here)) ; somewhat EXPENSIVE!!!
2427 (if (< (- here BOD-pos) how-far)
2428 (setq strategy 'BOD
2429 start-point BOD-pos)))
2430
2431 (list
2432 strategy
2433 (and (memq strategy '(forward backward)) cache-pos)
2434 (and (memq strategy '(forward BOD)) start-point))))
2435
2436
2437;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2438;; Routines which change `c-state-cache' and associated values.
2439(defun c-renarrow-state-cache ()
2440 ;; The region (more precisely, point-min) has changed since we
2441 ;; calculated `c-state-cache'. Amend `c-state-cache' accordingly.
2442 (if (< (point-min) c-state-point-min)
2443 ;; If point-min has MOVED BACKWARDS then we drop the state completely.
2444 ;; It would be possible to do a better job here and recalculate the top
2445 ;; only.
2446 (progn
2447 (c-state-mark-point-min-literal)
2448 (setq c-state-cache nil
2449 c-state-cache-good-pos c-state-min-scan-pos
2450 c-state-brace-pair-desert nil))
2451
2452 ;; point-min has MOVED FORWARD.
2453
2454 ;; Is the new point-min inside a (different) literal?
2455 (unless (and c-state-point-min-lit-start ; at prev. point-min
2456 (< (point-min) (c-state-get-min-scan-pos)))
2457 (c-state-mark-point-min-literal))
2458
2459 ;; Cut off a bit of the tail from `c-state-cache'.
2460 (let ((ptr (cons nil c-state-cache))
2461 pa)
2462 (while (and (setq pa (c-state-cache-top-lparen (cdr ptr)))
2463 (>= pa (point-min)))
2464 (setq ptr (cdr ptr)))
2465
2466 (when (consp ptr)
2467 (if (eq (cdr ptr) c-state-cache)
2468 (setq c-state-cache nil
2469 c-state-cache-good-pos c-state-min-scan-pos)
2470 (setcdr ptr nil)
2471 (setq c-state-cache-good-pos (1+ (c-state-cache-top-lparen))))
2472 )))
2473
2474 (setq c-state-point-min (point-min)))
2475
2476(defun c-append-lower-brace-pair-to-state-cache (from &optional upper-lim)
2477 ;; If there is a brace pair preceding FROM in the buffer (not necessarily
2478 ;; immediately preceding), push a cons onto `c-state-cache' to represent it.
2479 ;; FROM must not be inside a literal. If UPPER-LIM is non-nil, we append
2480 ;; the highest brace pair whose "}" is below UPPER-LIM.
2481 ;;
2482 ;; Return non-nil when this has been done.
2483 ;;
2484 ;; This routine should be fast. Since it can get called a LOT, we maintain
2485 ;; `c-state-brace-pair-desert', a small cache of "failures", such that we
2486 ;; reduce the time wasted in repeated fruitless searches in brace deserts.
2487 (save-excursion
2488 (save-restriction
2489 (let ((bra from) ce ; Positions of "{" and "}".
2490 new-cons
2491 (cache-pos (c-state-cache-top-lparen)) ; might be nil.
2492 (macro-start-or-from
2493 (progn (goto-char from)
2494 (c-beginning-of-macro)
2495 (point))))
2496 (or upper-lim (setq upper-lim from))
2497
2498 ;; If we're essentially repeating a fruitless search, just give up.
2499 (unless (and c-state-brace-pair-desert
2500 (eq cache-pos (car c-state-brace-pair-desert))
2501 (<= from (cdr c-state-brace-pair-desert)))
2502 ;; Only search what we absolutely need to:
2503 (if (and c-state-brace-pair-desert
cd5495ff 2504 (eq cache-pos (car c-state-brace-pair-desert)))
0ec1d2c5
AM
2505 (narrow-to-region (cdr c-state-brace-pair-desert) (point-max)))
2506
2507 ;; In the next pair of nested loops, the inner one moves back past a
2508 ;; pair of (mis-)matching parens or brackets; the outer one moves
2509 ;; back over a sequence of unmatched close brace/paren/bracket each
2510 ;; time round.
2511 (while
2512 (progn
2513 (c-safe
2514 (while
2515 (and (setq ce (scan-lists bra -1 -1)) ; back past )/]/}; might signal
2516 (setq bra (scan-lists ce -1 1)) ; back past (/[/{; might signal
2517 (or (> ce upper-lim)
2518 (not (eq (char-after bra) ?\{))
2519 (and (goto-char bra)
2520 (c-beginning-of-macro)
2521 (< (point) macro-start-or-from))))))
2522 (and ce (< ce bra)))
2523 (setq bra ce)) ; If we just backed over an unbalanced closing
2524 ; brace, ignore it.
2525
2526 (if (and ce (< bra ce) (eq (char-after bra) ?\{))
2527 ;; We've found the desired brace-pair.
2528 (progn
2529 (setq new-cons (cons bra (1+ ce)))
2530 (cond
2531 ((consp (car c-state-cache))
2532 (setcar c-state-cache new-cons))
2533 ((and (numberp (car c-state-cache)) ; probably never happens
2534 (< ce (car c-state-cache)))
2535 (setcdr c-state-cache
2536 (cons new-cons (cdr c-state-cache))))
2537 (t (setq c-state-cache (cons new-cons c-state-cache)))))
2538
2539 ;; We haven't found a brace pair. Record this.
2540 (setq c-state-brace-pair-desert (cons cache-pos from))))))))
2541
2542(defsubst c-state-push-any-brace-pair (bra+1 macro-start-or-here)
2543 ;; If BRA+1 is nil, do nothing. Otherwise, BRA+1 is the buffer position
2544 ;; following a {, and that brace has a (mis-)matching } (or ]), and we
2545 ;; "push" "a" brace pair onto `c-state-cache'.
2546 ;;
2547 ;; Here "push" means overwrite the top element if it's itself a brace-pair,
2548 ;; otherwise push it normally.
2549 ;;
2550 ;; The brace pair we push is normally the one surrounding BRA+1, but if the
2551 ;; latter is inside a macro, not being a macro containing
2552 ;; MACRO-START-OR-HERE, we scan backwards through the buffer for a non-macro
2553 ;; base pair. This latter case is assumed to be rare.
2554 ;;
2555 ;; Note: POINT is not preserved in this routine.
2556 (if bra+1
2557 (if (or (> bra+1 macro-start-or-here)
2558 (progn (goto-char bra+1)
2559 (not (c-beginning-of-macro))))
2560 (setq c-state-cache
2561 (cons (cons (1- bra+1)
2562 (scan-lists bra+1 1 1))
2563 (if (consp (car c-state-cache))
2564 (cdr c-state-cache)
2565 c-state-cache)))
2566 ;; N.B. This defsubst codes one method for the simple, normal case,
2567 ;; and a more sophisticated, slower way for the general case. Don't
e1dbe924 2568 ;; eliminate this defsubst - it's a speed optimization.
0ec1d2c5
AM
2569 (c-append-lower-brace-pair-to-state-cache (1- bra+1)))))
2570
2571(defun c-append-to-state-cache (from)
2572 ;; Scan the buffer from FROM to (point-max), adding elements into
2573 ;; `c-state-cache' for braces etc. Return a candidate for
2574 ;; `c-state-cache-good-pos'.
2575 ;;
2576 ;; FROM must be after the latest brace/paren/bracket in `c-state-cache', if
2577 ;; any. Typically, it is immediately after it. It must not be inside a
2578 ;; literal.
2579 (let ((here-bol (c-point 'bol (point-max)))
2580 (macro-start-or-here
2581 (save-excursion (goto-char (point-max))
2582 (if (c-beginning-of-macro)
2583 (point)
2584 (point-max))))
2585 pa+1 ; pos just after an opening PAren (or brace).
2586 (ren+1 from) ; usually a pos just after an closing paREN etc.
2587 ; Is actually the pos. to scan for a (/{/[ from,
2588 ; which sometimes is after a silly )/}/].
2589 paren+1 ; Pos after some opening or closing paren.
2590 paren+1s ; A list of `paren+1's; used to determine a
2591 ; good-pos.
2592 bra+1 ce+1 ; just after L/R bra-ces.
2593 bra+1s ; list of OLD values of bra+1.
2594 mstart) ; start of a macro.
2595
2596 (save-excursion
22bcf204 2597 ;; Each time round the following loop, we enter a successively deeper
0ec1d2c5
AM
2598 ;; level of brace/paren nesting. (Except sometimes we "continue at
2599 ;; the existing level".) `pa+1' is a pos inside an opening
2600 ;; brace/paren/bracket, usually just after it.
2601 (while
2602 (progn
2603 ;; Each time round the next loop moves forward over an opening then
2604 ;; a closing brace/bracket/paren. This loop is white hot, so it
2605 ;; plays ugly tricks to go fast. DON'T PUT ANYTHING INTO THIS
2606 ;; LOOP WHICH ISN'T ABSOLUTELY NECESSARY!!! It terminates when a
2607 ;; call of `scan-lists' signals an error, which happens when there
2608 ;; are no more b/b/p's to scan.
2609 (c-safe
2610 (while t
2611 (setq pa+1 (scan-lists ren+1 1 -1) ; Into (/{/[; might signal
2612 paren+1s (cons pa+1 paren+1s))
2613 (setq ren+1 (scan-lists pa+1 1 1)) ; Out of )/}/]; might signal
2614 (if (and (eq (char-before pa+1) ?{)) ; Check for a macro later.
2615 (setq bra+1 pa+1))
2616 (setcar paren+1s ren+1)))
2617
2618 (if (and pa+1 (> pa+1 ren+1))
2619 ;; We've just entered a deeper nesting level.
2620 (progn
2621 ;; Insert the brace pair (if present) and the single open
2622 ;; paren/brace/bracket into `c-state-cache' It cannot be
2623 ;; inside a macro, except one around point, because of what
2624 ;; `c-neutralize-syntax-in-CPP' has done.
2625 (c-state-push-any-brace-pair bra+1 macro-start-or-here)
2626 ;; Insert the opening brace/bracket/paren position.
2627 (setq c-state-cache (cons (1- pa+1) c-state-cache))
2628 ;; Clear admin stuff for the next more nested part of the scan.
2629 (setq ren+1 pa+1 pa+1 nil bra+1 nil bra+1s nil)
2630 t) ; Carry on the loop
2631
2632 ;; All open p/b/b's at this nesting level, if any, have probably
2633 ;; been closed by matching/mismatching ones. We're probably
2634 ;; finished - we just need to check for having found an
2635 ;; unmatched )/}/], which we ignore. Such a )/}/] can't be in a
2636 ;; macro, due the action of `c-neutralize-syntax-in-CPP'.
2637 (c-safe (setq ren+1 (scan-lists ren+1 1 1)))))) ; acts as loop control.
2638
2639 ;; Record the final, innermost, brace-pair if there is one.
2640 (c-state-push-any-brace-pair bra+1 macro-start-or-here)
2641
2642 ;; Determine a good pos
2643 (while (and (setq paren+1 (car paren+1s))
2644 (> (if (> paren+1 macro-start-or-here)
2645 paren+1
2646 (goto-char paren+1)
2647 (setq mstart (and (c-beginning-of-macro)
2648 (point)))
2649 (or mstart paren+1))
2650 here-bol))
2651 (setq paren+1s (cdr paren+1s)))
2652 (cond
2653 ((and paren+1 mstart)
2654 (min paren+1 mstart))
2655 (paren+1)
2656 (t from)))))
2657
2658(defun c-remove-stale-state-cache (good-pos pps-point)
2659 ;; Remove stale entries from the `c-cache-state', i.e. those which will
2660 ;; not be in it when it is amended for position (point-max).
2661 ;; Additionally, the "outermost" open-brace entry before (point-max)
2662 ;; will be converted to a cons if the matching close-brace is scanned.
2663 ;;
2664 ;; GOOD-POS is a "maximal" "safe position" - there must be no open
2665 ;; parens/braces/brackets between GOOD-POS and (point-max).
2666 ;;
2667 ;; As a second thing, calculate the result of parse-partial-sexp at
2668 ;; PPS-POINT, w.r.t. GOOD-POS. The motivation here is that
2669 ;; `c-state-cache-good-pos' may become PPS-POINT, but the caller may need to
2670 ;; adjust it to get outside a string/comment. (Sorry about this! The code
2671 ;; needs to be FAST).
2672 ;;
2673 ;; Return a list (GOOD-POS SCAN-BACK-POS PPS-STATE), where
2674 ;; o - GOOD-POS is a position where the new value `c-state-cache' is known
2675 ;; to be good (we aim for this to be as high as possible);
2676 ;; o - SCAN-BACK-POS, if not nil, indicates there may be a brace pair
2677 ;; preceding POS which needs to be recorded in `c-state-cache'. It is a
2678 ;; position to scan backwards from.
2679 ;; o - PPS-STATE is the parse-partial-sexp state at PPS-POINT.
2680 (save-restriction
2681 (narrow-to-region 1 (point-max))
2682 (save-excursion
eb2f6eeb 2683 (let* ((in-macro-start ; start of macro containing (point-max) or nil.
0ec1d2c5
AM
2684 (save-excursion
2685 (goto-char (point-max))
2686 (and (c-beginning-of-macro)
2687 (point))))
2688 (good-pos-actual-macro-start ; Start of macro containing good-pos
2689 ; or nil
2690 (and (< good-pos (point-max))
2691 (save-excursion
2692 (goto-char good-pos)
2693 (and (c-beginning-of-macro)
2694 (point)))))
2695 (good-pos-actual-macro-end ; End of this macro, (maybe
2696 ; (point-max)), or nil.
2697 (and good-pos-actual-macro-start
2698 (save-excursion
2699 (goto-char good-pos-actual-macro-start)
2700 (c-end-of-macro)
2701 (point))))
2702 pps-state ; Will be 9 or 10 elements long.
2703 pos
2704 upper-lim ; ,beyond which `c-state-cache' entries are removed
2705 scan-back-pos
2706 pair-beg pps-point-state target-depth)
2707
2708 ;; Remove entries beyond (point-max). Also remove any entries inside
2709 ;; a macro, unless (point-max) is in the same macro.
2710 (setq upper-lim
2711 (if (or (null c-state-old-cpp-beg)
2712 (and (> (point-max) c-state-old-cpp-beg)
2713 (< (point-max) c-state-old-cpp-end)))
2714 (point-max)
2715 (min (point-max) c-state-old-cpp-beg)))
657071fc 2716 (while (and c-state-cache (>= (c-state-cache-top-lparen) upper-lim))
0ec1d2c5
AM
2717 (setq c-state-cache (cdr c-state-cache)))
2718 ;; If `upper-lim' is inside the last recorded brace pair, remove its
2719 ;; RBrace and indicate we'll need to search backwards for a previous
2720 ;; brace pair.
2721 (when (and c-state-cache
2722 (consp (car c-state-cache))
2723 (> (cdar c-state-cache) upper-lim))
2724 (setcar c-state-cache (caar c-state-cache))
2725 (setq scan-back-pos (car c-state-cache)))
2726
2727 ;; The next loop jumps forward out of a nested level of parens each
2728 ;; time round; the corresponding elements in `c-state-cache' are
2729 ;; removed. `pos' is just after the brace-pair or the open paren at
2730 ;; (car c-state-cache). There can be no open parens/braces/brackets
2731 ;; between `good-pos'/`good-pos-actual-macro-start' and (point-max),
2732 ;; due to the interface spec to this function.
dd21b621 2733 (setq pos (if (and good-pos-actual-macro-end
eb2f6eeb
AM
2734 (not (eq good-pos-actual-macro-start
2735 in-macro-start)))
0ec1d2c5
AM
2736 (1+ good-pos-actual-macro-end) ; get outside the macro as
2737 ; marked by a `category' text property.
2738 good-pos))
2739 (goto-char pos)
2740 (while (and c-state-cache
2741 (< (point) (point-max)))
2742 (cond
2743 ((null pps-state) ; first time through
2744 (setq target-depth -1))
2745 ((eq (car pps-state) target-depth) ; found closing ),},]
2746 (setq target-depth (1- (car pps-state))))
2747 ;; Do nothing when we've merely reached pps-point.
2748 )
2749
2750 ;; Scan!
2751 (setq pps-state
2752 (parse-partial-sexp
2753 (point) (if (< (point) pps-point) pps-point (point-max))
2754 target-depth
2755 nil pps-state))
2756
2757 (if (= (point) pps-point)
2758 (setq pps-point-state pps-state))
2759
2760 (when (eq (car pps-state) target-depth)
2761 (setq pos (point)) ; POS is now just after an R-paren/brace.
2762 (cond
2763 ((and (consp (car c-state-cache))
2764 (eq (point) (cdar c-state-cache)))
2765 ;; We've just moved out of the paren pair containing the brace-pair
2766 ;; at (car c-state-cache). `pair-beg' is where the open paren is,
2767 ;; and is potentially where the open brace of a cons in
2768 ;; c-state-cache will be.
2769 (setq pair-beg (car-safe (cdr c-state-cache))
2770 c-state-cache (cdr-safe (cdr c-state-cache)))) ; remove {}pair + containing Lparen.
2771 ((numberp (car c-state-cache))
2772 (setq pair-beg (car c-state-cache)
2773 c-state-cache (cdr c-state-cache))) ; remove this
2774 ; containing Lparen
2775 ((numberp (cadr c-state-cache))
2776 (setq pair-beg (cadr c-state-cache)
2777 c-state-cache (cddr c-state-cache))) ; Remove a paren pair
2778 ; together with enclosed brace pair.
2779 ;; (t nil) ; Ignore an unmated Rparen.
2780 )))
2781
2782 (if (< (point) pps-point)
2783 (setq pps-state (parse-partial-sexp (point) pps-point
2784 nil nil ; TARGETDEPTH, STOPBEFORE
2785 pps-state)))
2786
2787 ;; If the last paren pair we moved out of was actually a brace pair,
2788 ;; insert it into `c-state-cache'.
2789 (when (and pair-beg (eq (char-after pair-beg) ?{))
2790 (if (consp (car-safe c-state-cache))
2791 (setq c-state-cache (cdr c-state-cache)))
2792 (setq c-state-cache (cons (cons pair-beg pos)
2793 c-state-cache)))
2794
2795 (list pos scan-back-pos pps-state)))))
2796
2797(defun c-remove-stale-state-cache-backwards (here cache-pos)
2798 ;; Strip stale elements of `c-state-cache' by moving backwards through the
2799 ;; buffer, and inform the caller of the scenario detected.
2800 ;;
2801 ;; HERE is the position we're setting `c-state-cache' for.
2802 ;; CACHE-POS is just after the latest recorded position in `c-state-cache'
2803 ;; before HERE, or a position at or near point-min which isn't in a
2804 ;; literal.
2805 ;;
2806 ;; This function must only be called only when (> `c-state-cache-good-pos'
2807 ;; HERE). Usually the gap between CACHE-POS and HERE is large. It is thus
e1dbe924 2808 ;; optimized to eliminate (or minimize) scanning between these two
0ec1d2c5
AM
2809 ;; positions.
2810 ;;
2811 ;; Return a three element list (GOOD-POS SCAN-BACK-POS FWD-FLAG), where:
2812 ;; o - GOOD-POS is a "good position", where `c-state-cache' is valid, or
2813 ;; could become so after missing elements are inserted into
2814 ;; `c-state-cache'. This is JUST AFTER an opening or closing
2815 ;; brace/paren/bracket which is already in `c-state-cache' or just before
2816 ;; one otherwise. exceptionally (when there's no such b/p/b handy) the BOL
2817 ;; before `here''s line, or the start of the literal containing it.
2818 ;; o - SCAN-BACK-POS, if non-nil, indicates there may be a brace pair
2819 ;; preceding POS which isn't recorded in `c-state-cache'. It is a position
2820 ;; to scan backwards from.
2821 ;; o - FWD-FLAG, if non-nil, indicates there may be parens/braces between
2822 ;; POS and HERE which aren't recorded in `c-state-cache'.
2823 ;;
2824 ;; The comments in this defun use "paren" to mean parenthesis or square
2825 ;; bracket (as contrasted with a brace), and "(" and ")" likewise.
2826 ;;
2827 ;; . {..} (..) (..) ( .. { } ) (...) ( .... . ..)
2828 ;; | | | | | |
2829 ;; CP E here D C good
2830 (let ((pos c-state-cache-good-pos)
2831 pa ren ; positions of "(" and ")"
2832 dropped-cons ; whether the last element dropped from `c-state-cache'
2833 ; was a cons (representing a brace-pair)
2834 good-pos ; see above.
2835 lit ; (START . END) of a literal containing some point.
2836 here-lit-start here-lit-end ; bounds of literal containing `here'
2837 ; or `here' itself.
c0209c2c 2838 here- here+ ; start/end of macro around HERE, or HERE
0ec1d2c5 2839 (here-bol (c-point 'bol here))
c400c4d7 2840 (too-far-back (max (- here c-state-cache-too-far) (point-min))))
0ec1d2c5
AM
2841
2842 ;; Remove completely irrelevant entries from `c-state-cache'.
2843 (while (and c-state-cache
2844 (>= (setq pa (c-state-cache-top-lparen)) here))
2845 (setq dropped-cons (consp (car c-state-cache)))
2846 (setq c-state-cache (cdr c-state-cache))
2847 (setq pos pa))
2848 ;; At this stage, (> pos here);
2849 ;; (< (c-state-cache-top-lparen) here) (or is nil).
2850
c0209c2c
AM
2851 (cond
2852 ((and (consp (car c-state-cache))
2853 (> (cdar c-state-cache) here))
2854 ;; CASE 1: The top of the cache is a brace pair which now encloses
2855 ;; `here'. As good-pos, return the address. of the "{". Since we've no
2856 ;; knowledge of what's inside these braces, we have no alternative but
2857 ;; to direct the caller to scan the buffer from the opening brace.
2858 (setq pos (caar c-state-cache))
2859 (setcar c-state-cache pos)
2860 (list (1+ pos) pos t)) ; return value. We've just converted a brace pair
2861 ; entry into a { entry, so the caller needs to
2862 ; search for a brace pair before the {.
2863
2864 ;; `here' might be inside a literal. Check for this.
2865 ((progn
2866 (setq lit (c-state-literal-at here)
2867 here-lit-start (or (car lit) here)
2868 here-lit-end (or (cdr lit) here))
2869 ;; Has `here' just "newly entered" a macro?
2870 (save-excursion
2871 (goto-char here-lit-start)
2872 (if (and (c-beginning-of-macro)
2873 (or (null c-state-old-cpp-beg)
2874 (not (= (point) c-state-old-cpp-beg))))
2875 (progn
2876 (setq here- (point))
2877 (c-end-of-macro)
2878 (setq here+ (point)))
2879 (setq here- here-lit-start
2880 here+ here-lit-end)))
2881
2882 ;; `here' might be nested inside any depth of parens (or brackets but
2883 ;; not braces). Scan backwards to find the outermost such opening
2884 ;; paren, if there is one. This will be the scan position to return.
2885 (save-restriction
2886 (narrow-to-region cache-pos (point-max))
2887 (setq pos (c-state-balance-parens-backwards here- here+ pos)))
2888 nil)) ; for the cond
2889
2890 ((< pos here-lit-start)
2891 ;; CASE 2: Address of outermost ( or [ which now encloses `here', but
2892 ;; didn't enclose the (previous) `c-state-cache-good-pos'. If there is
2893 ;; a brace pair preceding this, it will already be in `c-state-cache',
2894 ;; unless there was a brace pair after it, i.e. there'll only be one to
2895 ;; scan for if we've just deleted one.
2896 (list pos (and dropped-cons pos) t)) ; Return value.
2897
2898 ;; `here' isn't enclosed in a (previously unrecorded) bracket/paren.
2899 ;; Further forward scanning isn't needed, but we still need to find a
2900 ;; GOOD-POS. Step out of all enclosing "("s on HERE's line.
2901 ((progn
0ec1d2c5
AM
2902 (save-restriction
2903 (narrow-to-region here-bol (point-max))
2904 (setq pos here-lit-start)
2905 (c-safe (while (setq pa (scan-lists pos -1 1))
2906 (setq pos pa)))) ; might signal
c0209c2c
AM
2907 nil)) ; for the cond
2908
2909 ((setq ren (c-safe-scan-lists pos -1 -1 too-far-back))
2910 ;; CASE 3: After a }/)/] before `here''s BOL.
2911 (list (1+ ren) (and dropped-cons pos) nil)) ; Return value
2912
2913 (t
2914 ;; CASE 4; Best of a bad job: BOL before `here-bol', or beginning of
2915 ;; literal containing it.
2916 (setq good-pos (c-state-lit-beg (c-point 'bopl here-bol)))
2917 (list good-pos (and dropped-cons good-pos) nil)))))
0ec1d2c5
AM
2918
2919
2920;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2921;; Externally visible routines.
2922
2923(defun c-state-cache-init ()
2924 (setq c-state-cache nil
2925 c-state-cache-good-pos 1
2926 c-state-nonlit-pos-cache nil
2927 c-state-nonlit-pos-cache-limit 1
2928 c-state-brace-pair-desert nil
2929 c-state-point-min 1
2930 c-state-point-min-lit-type nil
2931 c-state-point-min-lit-start nil
2932 c-state-min-scan-pos 1
2933 c-state-old-cpp-beg nil
2934 c-state-old-cpp-end nil)
2935 (c-state-mark-point-min-literal))
2936
e84efb70
AM
2937;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2938;; Debugging routines to dump `c-state-cache' in a "replayable" form.
2939;; (defmacro c-sc-de (elt) ; "c-state-cache-dump-element"
2940;; `(format ,(concat "(setq " (symbol-name elt) " %s) ") ,elt))
2941;; (defmacro c-sc-qde (elt) ; "c-state-cache-quote-dump-element"
2942;; `(format ,(concat "(setq " (symbol-name elt) " '%s) ") ,elt))
2943;; (defun c-state-dump ()
2944;; ;; For debugging.
2945;; ;(message
2946;; (concat
2947;; (c-sc-qde c-state-cache)
2948;; (c-sc-de c-state-cache-good-pos)
2949;; (c-sc-qde c-state-nonlit-pos-cache)
2950;; (c-sc-de c-state-nonlit-pos-cache-limit)
2951;; (c-sc-qde c-state-brace-pair-desert)
2952;; (c-sc-de c-state-point-min)
2953;; (c-sc-de c-state-point-min-lit-type)
2954;; (c-sc-de c-state-point-min-lit-start)
2955;; (c-sc-de c-state-min-scan-pos)
2956;; (c-sc-de c-state-old-cpp-beg)
2957;; (c-sc-de c-state-old-cpp-end)))
2958;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2959
0ec1d2c5
AM
2960(defun c-invalidate-state-cache-1 (here)
2961 ;; Invalidate all info on `c-state-cache' that applies to the buffer at HERE
2962 ;; or higher and set `c-state-cache-good-pos' accordingly. The cache is
2963 ;; left in a consistent state.
2964 ;;
2965 ;; This is much like `c-whack-state-after', but it never changes a paren
2966 ;; pair element into an open paren element. Doing that would mean that the
2967 ;; new open paren wouldn't have the required preceding paren pair element.
785eecbb 2968 ;;
0ec1d2c5
AM
2969 ;; This function is called from c-after-change.
2970
2971 ;; The cache of non-literals:
2972 (if (< here c-state-nonlit-pos-cache-limit)
2973 (setq c-state-nonlit-pos-cache-limit here))
2974
2975 ;; `c-state-cache':
2976 ;; Case 1: if `here' is in a literal containing point-min, everything
2977 ;; becomes (or is already) nil.
2978 (if (or (null c-state-cache-good-pos)
2979 (< here (c-state-get-min-scan-pos)))
2980 (setq c-state-cache nil
2981 c-state-cache-good-pos nil
2982 c-state-min-scan-pos nil)
2983
c400c4d7
SM
2984 ;; Truncate `c-state-cache' and set `c-state-cache-good-pos' to a value
2985 ;; below `here'. To maintain its consistency, we may need to insert a new
2986 ;; brace pair.
0ec1d2c5
AM
2987 (let ((here-bol (c-point 'bol here))
2988 too-high-pa ; recorded {/(/[ next above here, or nil.
2989 dropped-cons ; was the last removed element a brace pair?
2990 pa)
2991 ;; The easy bit - knock over-the-top bits off `c-state-cache'.
2992 (while (and c-state-cache
2993 (>= (setq pa (c-state-cache-top-paren)) here))
2994 (setq dropped-cons (consp (car c-state-cache))
2995 too-high-pa (c-state-cache-top-lparen)
2996 c-state-cache (cdr c-state-cache)))
2997
2998 ;; Do we need to add in an earlier brace pair, having lopped one off?
2999 (if (and dropped-cons
3000 (< too-high-pa (+ here c-state-cache-too-far)))
3001 (c-append-lower-brace-pair-to-state-cache too-high-pa here-bol))
3002 (setq c-state-cache-good-pos (or (c-state-cache-after-top-paren)
3003 (c-state-get-min-scan-pos)))))
3004
3005 ;; The brace-pair desert marker:
3006 (when (car c-state-brace-pair-desert)
3007 (if (< here (car c-state-brace-pair-desert))
3008 (setq c-state-brace-pair-desert nil)
3009 (if (< here (cdr c-state-brace-pair-desert))
3010 (setcdr c-state-brace-pair-desert here)))))
3011
3012(defun c-parse-state-1 ()
3013 ;; Find and record all noteworthy parens between some good point earlier in
3014 ;; the file and point. That good point is at least the beginning of the
3015 ;; top-level construct we are in, or the beginning of the preceding
3016 ;; top-level construct if we aren't in one.
3017 ;;
3018 ;; The returned value is a list of the noteworthy parens with the last one
3019 ;; first. If an element in the list is an integer, it's the position of an
3020 ;; open paren (of any type) which has not been closed before the point. If
3021 ;; an element is a cons, it gives the position of a closed BRACE paren
3022 ;; pair[*]; the car is the start brace position and the cdr is the position
3023 ;; following the closing brace. Only the last closed brace paren pair
3024 ;; before each open paren and before the point is recorded, and thus the
3025 ;; state never contains two cons elements in succession. When a close brace
3026 ;; has no matching open brace (e.g., the matching brace is outside the
3027 ;; visible region), it is not represented in the returned value.
3028 ;;
3029 ;; [*] N.B. The close "brace" might be a mismatching close bracket or paren.
3030 ;; This defun explicitly treats mismatching parens/braces/brackets as
3031 ;; matching. It is the open brace which makes it a "brace" pair.
3032 ;;
3033 ;; If POINT is within a macro, open parens and brace pairs within
3034 ;; THIS macro MIGHT be recorded. This depends on whether their
3035 ;; syntactic properties have been suppressed by
3036 ;; `c-neutralize-syntax-in-CPP'. This might need fixing (2008-12-11).
d9e94c22
MS
3037 ;;
3038 ;; Currently no characters which are given paren syntax with the
3039 ;; syntax-table property are recorded, i.e. angle bracket arglist
3040 ;; parens are never present here. Note that this might change.
3041 ;;
0386b551 3042 ;; BUG: This function doesn't cope entirely well with unbalanced
0ec1d2c5
AM
3043 ;; parens in macros. (2008-12-11: this has probably been resolved
3044 ;; by the function `c-neutralize-syntax-in-CPP'.) E.g. in the
3045 ;; following case the brace before the macro isn't balanced with the
3046 ;; one after it:
0386b551
AM
3047 ;;
3048 ;; {
3049 ;; #define X {
3050 ;; }
3051 ;;
0ec1d2c5
AM
3052 ;; Note to maintainers: this function DOES get called with point
3053 ;; within comments and strings, so don't assume it doesn't!
3054 ;;
0386b551 3055 ;; This function might do hidden buffer changes.
0ec1d2c5
AM
3056 (let* ((here (point))
3057 (here-bopl (c-point 'bopl))
3058 strategy ; 'forward, 'backward etc..
3059 ;; Candidate positions to start scanning from:
3060 cache-pos ; highest position below HERE already existing in
3061 ; cache (or 1).
3062 good-pos
3063 start-point
3064 bopl-state
3065 res
3066 scan-backward-pos scan-forward-p) ; used for 'backward.
3067 ;; If POINT-MIN has changed, adjust the cache
3068 (unless (= (point-min) c-state-point-min)
3069 (c-renarrow-state-cache))
3070
3071 ;; Strategy?
3072 (setq res (c-parse-state-get-strategy here c-state-cache-good-pos)
3073 strategy (car res)
3074 cache-pos (cadr res)
3075 start-point (nth 2 res))
3076
3077 (when (eq strategy 'BOD)
3078 (setq c-state-cache nil
3079 c-state-cache-good-pos start-point))
3080
3081 ;; SCAN!
3082 (save-restriction
3083 (cond
3084 ((memq strategy '(forward BOD))
0386b551 3085 (narrow-to-region (point-min) here)
0ec1d2c5
AM
3086 (setq res (c-remove-stale-state-cache start-point here-bopl))
3087 (setq cache-pos (car res)
3088 scan-backward-pos (cadr res)
3089 bopl-state (car (cddr res))) ; will be nil if (< here-bopl
3090 ; start-point)
3091 (if scan-backward-pos
3092 (c-append-lower-brace-pair-to-state-cache scan-backward-pos))
3093 (setq good-pos
3094 (c-append-to-state-cache cache-pos))
3095 (setq c-state-cache-good-pos
3096 (if (and bopl-state
3097 (< good-pos (- here c-state-cache-too-far)))
3098 (c-state-cache-non-literal-place here-bopl bopl-state)
3099 good-pos)))
3100
3101 ((eq strategy 'backward)
3102 (setq res (c-remove-stale-state-cache-backwards here cache-pos)
3103 good-pos (car res)
3104 scan-backward-pos (cadr res)
3105 scan-forward-p (car (cddr res)))
3106 (if scan-backward-pos
3107 (c-append-lower-brace-pair-to-state-cache
3108 scan-backward-pos))
3109 (setq c-state-cache-good-pos
3110 (if scan-forward-p
3111 (progn (narrow-to-region (point-min) here)
3112 (c-append-to-state-cache good-pos))
3113
3114 (c-get-cache-scan-pos good-pos))))
3115
3116 (t ; (eq strategy 'IN-LIT)
3117 (setq c-state-cache nil
3118 c-state-cache-good-pos nil)))))
3119
3120 c-state-cache)
3121
3122(defun c-invalidate-state-cache (here)
3123 ;; This is a wrapper over `c-invalidate-state-cache-1'.
3124 ;;
3125 ;; It suppresses the syntactic effect of the < and > (template) brackets and
3126 ;; of all parens in preprocessor constructs, except for any such construct
3127 ;; containing point. We can then call `c-invalidate-state-cache-1' without
3128 ;; worrying further about macros and template delimiters.
3129 (c-with-<->-as-parens-suppressed
3e8f7d91
AM
3130 (if (and c-state-old-cpp-beg
3131 (< c-state-old-cpp-beg here))
0ec1d2c5 3132 (c-with-all-but-one-cpps-commented-out
3e8f7d91
AM
3133 c-state-old-cpp-beg
3134 (min c-state-old-cpp-end here)
0ec1d2c5
AM
3135 (c-invalidate-state-cache-1 here))
3136 (c-with-cpps-commented-out
3137 (c-invalidate-state-cache-1 here)))))
3138
3139(defun c-parse-state ()
3140 ;; This is a wrapper over `c-parse-state-1'. See that function for a
3141 ;; description of the functionality and return value.
3142 ;;
3143 ;; It suppresses the syntactic effect of the < and > (template) brackets and
3144 ;; of all parens in preprocessor constructs, except for any such construct
3145 ;; containing point. We can then call `c-parse-state-1' without worrying
3146 ;; further about macros and template delimiters.
3147 (let (here-cpp-beg here-cpp-end)
3148 (save-excursion
3149 (when (c-beginning-of-macro)
3150 (setq here-cpp-beg (point))
3151 (unless
3152 (> (setq here-cpp-end (c-syntactic-end-of-macro))
3153 here-cpp-beg)
3154 (setq here-cpp-beg nil here-cpp-end nil))))
3155 ;; FIXME!!! Put in a `condition-case' here to protect the integrity of the
3156 ;; subsystem.
3157 (prog1
3158 (c-with-<->-as-parens-suppressed
3159 (if (and here-cpp-beg (> here-cpp-end here-cpp-beg))
3160 (c-with-all-but-one-cpps-commented-out
3161 here-cpp-beg here-cpp-end
3162 (c-parse-state-1))
3163 (c-with-cpps-commented-out
3164 (c-parse-state-1))))
3e8f7d91
AM
3165 (setq c-state-old-cpp-beg (and here-cpp-beg (copy-marker here-cpp-beg t))
3166 c-state-old-cpp-end (and here-cpp-end (copy-marker here-cpp-end t)))
3167 )))
0ec1d2c5
AM
3168
3169;; Debug tool to catch cache inconsistencies. This is called from
3170;; 000tests.el.
a66cd3ee
MS
3171(defvar c-debug-parse-state nil)
3172(unless (fboundp 'c-real-parse-state)
3173 (fset 'c-real-parse-state (symbol-function 'c-parse-state)))
3174(cc-bytecomp-defun c-real-parse-state)
cd5495ff
AM
3175
3176(defvar c-parse-state-state nil)
3177(defun c-record-parse-state-state ()
3178 (setq c-parse-state-state
3179 (mapcar
3180 (lambda (arg)
3181 (cons arg (symbol-value arg)))
3182 '(c-state-cache
3183 c-state-cache-good-pos
3184 c-state-nonlit-pos-cache
3185 c-state-nonlit-pos-cache-limit
3186 c-state-brace-pair-desert
3187 c-state-point-min
3188 c-state-point-min-lit-type
3189 c-state-point-min-lit-start
3190 c-state-min-scan-pos
3191 c-state-old-cpp-beg
3192 c-state-old-cpp-end))))
3193(defun c-replay-parse-state-state ()
3194 (message
3195 (concat "(setq "
3196 (mapconcat
3197 (lambda (arg)
3198 (format "%s %s%s" (car arg) (if (atom (cdr arg)) "" "'") (cdr arg)))
3199 c-parse-state-state " ")
3200 ")")))
3201
a66cd3ee 3202(defun c-debug-parse-state ()
0ec1d2c5 3203 (let ((here (point)) (res1 (c-real-parse-state)) res2)
0386b551 3204 (let ((c-state-cache nil)
0ec1d2c5
AM
3205 (c-state-cache-good-pos 1)
3206 (c-state-nonlit-pos-cache nil)
3207 (c-state-nonlit-pos-cache-limit 1)
3208 (c-state-brace-pair-desert nil)
3209 (c-state-point-min 1)
3210 (c-state-point-min-lit-type nil)
3211 (c-state-point-min-lit-start nil)
3212 (c-state-min-scan-pos 1)
3213 (c-state-old-cpp-beg nil)
3214 (c-state-old-cpp-end nil))
a66cd3ee
MS
3215 (setq res2 (c-real-parse-state)))
3216 (unless (equal res1 res2)
0386b551
AM
3217 ;; The cache can actually go further back due to the ad-hoc way
3218 ;; the first paren is found, so try to whack off a bit of its
3219 ;; start before complaining.
cd5495ff
AM
3220 ;; (save-excursion
3221 ;; (goto-char (or (c-least-enclosing-brace res2) (point)))
3222 ;; (c-beginning-of-defun-1)
3223 ;; (while (not (or (bobp) (eq (char-after) ?{)))
3224 ;; (c-beginning-of-defun-1))
3225 ;; (unless (equal (c-whack-state-before (point) res1) res2)
3226 ;; (message (concat "c-parse-state inconsistency at %s: "
3227 ;; "using cache: %s, from scratch: %s")
3228 ;; here res1 res2)))
3229 (message (concat "c-parse-state inconsistency at %s: "
3230 "using cache: %s, from scratch: %s")
3231 here res1 res2)
3232 (message "Old state:")
3233 (c-replay-parse-state-state))
3234 (c-record-parse-state-state)
a66cd3ee 3235 res1))
0ec1d2c5 3236
a66cd3ee
MS
3237(defun c-toggle-parse-state-debug (&optional arg)
3238 (interactive "P")
3239 (setq c-debug-parse-state (c-calculate-state arg c-debug-parse-state))
3240 (fset 'c-parse-state (symbol-function (if c-debug-parse-state
3241 'c-debug-parse-state
3242 'c-real-parse-state)))
3243 (c-keep-region-active))
0386b551
AM
3244(when c-debug-parse-state
3245 (c-toggle-parse-state-debug 1))
a66cd3ee 3246
0ec1d2c5 3247\f
d9e94c22
MS
3248(defun c-whack-state-before (bufpos paren-state)
3249 ;; Whack off any state information from PAREN-STATE which lies
3250 ;; before BUFPOS. Not destructive on PAREN-STATE.
d9e94c22
MS
3251 (let* ((newstate (list nil))
3252 (ptr newstate)
3253 car)
3254 (while paren-state
3255 (setq car (car paren-state)
3256 paren-state (cdr paren-state))
3257 (if (< (if (consp car) (car car) car) bufpos)
3258 (setq paren-state nil)
3259 (setcdr ptr (list car))
3260 (setq ptr (cdr ptr))))
3261 (cdr newstate)))
3262
3263(defun c-whack-state-after (bufpos paren-state)
3264 ;; Whack off any state information from PAREN-STATE which lies at or
3265 ;; after BUFPOS. Not destructive on PAREN-STATE.
d9e94c22
MS
3266 (catch 'done
3267 (while paren-state
3268 (let ((car (car paren-state)))
3269 (if (consp car)
3270 ;; just check the car, because in a balanced brace
3271 ;; expression, it must be impossible for the corresponding
3272 ;; close brace to be before point, but the open brace to
3273 ;; be after.
3274 (if (<= bufpos (car car))
3275 nil ; whack it off
3276 (if (< bufpos (cdr car))
3277 ;; its possible that the open brace is before
3278 ;; bufpos, but the close brace is after. In that
3279 ;; case, convert this to a non-cons element. The
3280 ;; rest of the state is before bufpos, so we're
3281 ;; done.
3282 (throw 'done (cons (car car) (cdr paren-state)))
3283 ;; we know that both the open and close braces are
3284 ;; before bufpos, so we also know that everything else
3285 ;; on state is before bufpos.
3286 (throw 'done paren-state)))
3287 (if (<= bufpos car)
3288 nil ; whack it off
3289 ;; it's before bufpos, so everything else should too.
3290 (throw 'done paren-state)))
3291 (setq paren-state (cdr paren-state)))
3292 nil)))
3293
3294(defun c-most-enclosing-brace (paren-state &optional bufpos)
3295 ;; Return the bufpos of the innermost enclosing open paren before
0386b551 3296 ;; bufpos, or nil if none was found.
d9e94c22
MS
3297 (let (enclosingp)
3298 (or bufpos (setq bufpos 134217727))
3299 (while paren-state
3300 (setq enclosingp (car paren-state)
3301 paren-state (cdr paren-state))
3302 (if (or (consp enclosingp)
3303 (>= enclosingp bufpos))
3304 (setq enclosingp nil)
d9e94c22
MS
3305 (setq paren-state nil)))
3306 enclosingp))
3307
0386b551
AM
3308(defun c-least-enclosing-brace (paren-state)
3309 ;; Return the bufpos of the outermost enclosing open paren, or nil
3310 ;; if none was found.
d9e94c22 3311 (let (pos elem)
d9e94c22
MS
3312 (while paren-state
3313 (setq elem (car paren-state)
3314 paren-state (cdr paren-state))
0386b551
AM
3315 (if (integerp elem)
3316 (setq pos elem)))
d9e94c22
MS
3317 pos))
3318
3319(defun c-safe-position (bufpos paren-state)
0386b551
AM
3320 ;; Return the closest "safe" position recorded on PAREN-STATE that
3321 ;; is higher up than BUFPOS. Return nil if PAREN-STATE doesn't
3322 ;; contain any. Return nil if BUFPOS is nil, which is useful to
3323 ;; find the closest limit before a given limit that might be nil.
d9e94c22 3324 ;;
0386b551
AM
3325 ;; A "safe" position is a position at or after a recorded open
3326 ;; paren, or after a recorded close paren. The returned position is
3327 ;; thus either the first position after a close brace, or the first
3328 ;; position after an enclosing paren, or at the enclosing paren in
3329 ;; case BUFPOS is immediately after it.
d9e94c22
MS
3330 (when bufpos
3331 (let (elem)
3332 (catch 'done
3333 (while paren-state
3334 (setq elem (car paren-state))
3335 (if (consp elem)
3336 (cond ((< (cdr elem) bufpos)
3337 (throw 'done (cdr elem)))
3338 ((< (car elem) bufpos)
3339 ;; See below.
3340 (throw 'done (min (1+ (car elem)) bufpos))))
3341 (if (< elem bufpos)
3342 ;; elem is the position at and not after the opening paren, so
3343 ;; we can go forward one more step unless it's equal to
3344 ;; bufpos. This is useful in some cases avoid an extra paren
3345 ;; level between the safe position and bufpos.
3346 (throw 'done (min (1+ elem) bufpos))))
3347 (setq paren-state (cdr paren-state)))))))
3348
3349(defun c-beginning-of-syntax ()
3350 ;; This is used for `font-lock-beginning-of-syntax-function'. It
3351 ;; goes to the closest previous point that is known to be outside
3352 ;; any string literal or comment. `c-state-cache' is used if it has
3353 ;; a position in the vicinity.
3354 (let* ((paren-state c-state-cache)
3355 elem
3356
3357 (pos (catch 'done
3358 ;; Note: Similar code in `c-safe-position'. The
3359 ;; difference is that we accept a safe position at
3360 ;; the point and don't bother to go forward past open
3361 ;; parens.
3362 (while paren-state
3363 (setq elem (car paren-state))
3364 (if (consp elem)
3365 (cond ((<= (cdr elem) (point))
3366 (throw 'done (cdr elem)))
3367 ((<= (car elem) (point))
3368 (throw 'done (car elem))))
3369 (if (<= elem (point))
3370 (throw 'done elem)))
3371 (setq paren-state (cdr paren-state)))
3372 (point-min))))
3373
3374 (if (> pos (- (point) 4000))
3375 (goto-char pos)
3376 ;; The position is far back. Try `c-beginning-of-defun-1'
3377 ;; (although we can't be entirely sure it will go to a position
3378 ;; outside a comment or string in current emacsen). FIXME:
3379 ;; Consult `syntax-ppss' here.
3380 (c-beginning-of-defun-1)
3381 (if (< (point) pos)
3382 (goto-char pos)))))
3383
3384\f
3385;; Tools for scanning identifiers and other tokens.
3386
3387(defun c-on-identifier ()
3388 "Return non-nil if the point is on or directly after an identifier.
3389Keywords are recognized and not considered identifiers. If an
3390identifier is detected, the returned value is its starting position.
0386b551
AM
3391If an identifier ends at the point and another begins at it \(can only
3392happen in Pike) then the point for the preceding one is returned.
d9e94c22 3393
0386b551
AM
3394Note that this function might do hidden buffer changes. See the
3395comment at the start of cc-engine.el for more info."
3396
3397 ;; FIXME: Shouldn't this function handle "operator" in C++?
d9e94c22
MS
3398
3399 (save-excursion
0386b551
AM
3400 (skip-syntax-backward "w_")
3401
3402 (or
3403
3404 ;; Check for a normal (non-keyword) identifier.
3405 (and (looking-at c-symbol-start)
3406 (not (looking-at c-keywords-regexp))
3407 (point))
3408
3409 (when (c-major-mode-is 'pike-mode)
3410 ;; Handle the `<operator> syntax in Pike.
3411 (let ((pos (point)))
3412 (skip-chars-backward "-!%&*+/<=>^|~[]()")
3413 (and (if (< (skip-chars-backward "`") 0)
3414 t
3415 (goto-char pos)
3416 (eq (char-after) ?\`))
3417 (looking-at c-symbol-key)
3418 (>= (match-end 0) pos)
3419 (point))))
3420
3421 ;; Handle the "operator +" syntax in C++.
3422 (when (and c-overloadable-operators-regexp
3423 (= (c-backward-token-2 0) 0))
3424
3425 (cond ((and (looking-at c-overloadable-operators-regexp)
51c9af45 3426 (or (not c-opt-op-identifier-prefix)
0386b551 3427 (and (= (c-backward-token-2 1) 0)
51c9af45 3428 (looking-at c-opt-op-identifier-prefix))))
0386b551
AM
3429 (point))
3430
3431 ((save-excursion
51c9af45
AM
3432 (and c-opt-op-identifier-prefix
3433 (looking-at c-opt-op-identifier-prefix)
0386b551
AM
3434 (= (c-forward-token-2 1) 0)
3435 (looking-at c-overloadable-operators-regexp)))
3436 (point))))
3437
3438 )))
d9e94c22
MS
3439
3440(defsubst c-simple-skip-symbol-backward ()
3441 ;; If the point is at the end of a symbol then skip backward to the
3442 ;; beginning of it. Don't move otherwise. Return non-nil if point
3443 ;; moved.
0386b551
AM
3444 ;;
3445 ;; This function might do hidden buffer changes.
d9e94c22
MS
3446 (or (< (skip-syntax-backward "w_") 0)
3447 (and (c-major-mode-is 'pike-mode)
3448 ;; Handle the `<operator> syntax in Pike.
3449 (let ((pos (point)))
2a15eb73 3450 (if (and (< (skip-chars-backward "-!%&*+/<=>^|~[]()") 0)
d9e94c22
MS
3451 (< (skip-chars-backward "`") 0)
3452 (looking-at c-symbol-key)
3453 (>= (match-end 0) pos))
3454 t
3455 (goto-char pos)
3456 nil)))))
3457
0386b551 3458(defun c-beginning-of-current-token (&optional back-limit)
d9e94c22
MS
3459 ;; Move to the beginning of the current token. Do not move if not
3460 ;; in the middle of one. BACK-LIMIT may be used to bound the
3461 ;; backward search; if given it's assumed to be at the boundary
a85fd6da 3462 ;; between two tokens. Return non-nil if the point is moved, nil
580fba94 3463 ;; otherwise.
0386b551
AM
3464 ;;
3465 ;; This function might do hidden buffer changes.
d9e94c22 3466 (let ((start (point)))
580fba94
AM
3467 (if (looking-at "\\w\\|\\s_")
3468 (skip-syntax-backward "w_" back-limit)
3469 (when (< (skip-syntax-backward ".()" back-limit) 0)
3470 (while (let ((pos (or (and (looking-at c-nonsymbol-token-regexp)
3471 (match-end 0))
3472 ;; `c-nonsymbol-token-regexp' should always match
3473 ;; since we've skipped backward over punctuator
3474 ;; or paren syntax, but consume one char in case
3475 ;; it doesn't so that we don't leave point before
3476 ;; some earlier incorrect token.
3477 (1+ (point)))))
3478 (if (<= pos start)
3479 (goto-char pos))))))
3480 (< (point) start)))
d9e94c22 3481
ff959bab 3482(defun c-end-of-current-token (&optional back-limit)
d9e94c22
MS
3483 ;; Move to the end of the current token. Do not move if not in the
3484 ;; middle of one. BACK-LIMIT may be used to bound the backward
3485 ;; search; if given it's assumed to be at the boundary between two
ff959bab 3486 ;; tokens. Return non-nil if the point is moved, nil otherwise.
0386b551
AM
3487 ;;
3488 ;; This function might do hidden buffer changes.
d9e94c22
MS
3489 (let ((start (point)))
3490 (cond ((< (skip-syntax-backward "w_" (1- start)) 0)
3491 (skip-syntax-forward "w_"))
3492 ((< (skip-syntax-backward ".()" back-limit) 0)
3493 (while (progn
3494 (if (looking-at c-nonsymbol-token-regexp)
3495 (goto-char (match-end 0))
3496 ;; `c-nonsymbol-token-regexp' should always match since
3497 ;; we've skipped backward over punctuator or paren
3498 ;; syntax, but move forward in case it doesn't so that
3499 ;; we don't leave point earlier than we started with.
3500 (forward-char))
ff959bab
MS
3501 (< (point) start)))))
3502 (> (point) start)))
d9e94c22
MS
3503
3504(defconst c-jump-syntax-balanced
3505 (if (memq 'gen-string-delim c-emacs-features)
3506 "\\w\\|\\s_\\|\\s\(\\|\\s\)\\|\\s\"\\|\\s|"
3507 "\\w\\|\\s_\\|\\s\(\\|\\s\)\\|\\s\""))
3508
3509(defconst c-jump-syntax-unbalanced
3510 (if (memq 'gen-string-delim c-emacs-features)
3511 "\\w\\|\\s_\\|\\s\"\\|\\s|"
3512 "\\w\\|\\s_\\|\\s\""))
3513
3514(defun c-forward-token-2 (&optional count balanced limit)
3515 "Move forward by tokens.
3516A token is defined as all symbols and identifiers which aren't
3517syntactic whitespace \(note that multicharacter tokens like \"==\" are
3518treated properly). Point is always either left at the beginning of a
3519token or not moved at all. COUNT specifies the number of tokens to
3520move; a negative COUNT moves in the opposite direction. A COUNT of 0
3521moves to the next token beginning only if not already at one. If
3522BALANCED is true, move over balanced parens, otherwise move into them.
3523Also, if BALANCED is true, never move out of an enclosing paren.
3524
3525LIMIT sets the limit for the movement and defaults to the point limit.
3526The case when LIMIT is set in the middle of a token, comment or macro
3527is handled correctly, i.e. the point won't be left there.
3528
3529Return the number of tokens left to move \(positive or negative). If
3530BALANCED is true, a move over a balanced paren counts as one. Note
3531that if COUNT is 0 and no appropriate token beginning is found, 1 will
3532be returned. Thus, a return value of 0 guarantees that point is at
3533the requested position and a return value less \(without signs) than
0386b551
AM
3534COUNT guarantees that point is at the beginning of some token.
3535
3536Note that this function might do hidden buffer changes. See the
3537comment at the start of cc-engine.el for more info."
d9e94c22
MS
3538
3539 (or count (setq count 1))
3540 (if (< count 0)
3541 (- (c-backward-token-2 (- count) balanced limit))
3542
3543 (let ((jump-syntax (if balanced
3544 c-jump-syntax-balanced
3545 c-jump-syntax-unbalanced))
3546 (last (point))
3547 (prev (point)))
3548
3549 (if (zerop count)
3550 ;; If count is zero we should jump if in the middle of a token.
3551 (c-end-of-current-token))
3552
3553 (save-restriction
3554 (if limit (narrow-to-region (point-min) limit))
3555 (if (/= (point)
3556 (progn (c-forward-syntactic-ws) (point)))
3557 ;; Skip whitespace. Count this as a move if we did in
3558 ;; fact move.
3559 (setq count (max (1- count) 0)))
3560
3561 (if (eobp)
3562 ;; Moved out of bounds. Make sure the returned count isn't zero.
3563 (progn
3564 (if (zerop count) (setq count 1))
3565 (goto-char last))
3566
3567 ;; Use `condition-case' to avoid having the limit tests
3568 ;; inside the loop.
3569 (condition-case nil
3570 (while (and
3571 (> count 0)
3572 (progn
3573 (setq last (point))
3574 (cond ((looking-at jump-syntax)
3575 (goto-char (scan-sexps (point) 1))
3576 t)
3577 ((looking-at c-nonsymbol-token-regexp)
3578 (goto-char (match-end 0))
3579 t)
3580 ;; `c-nonsymbol-token-regexp' above should always
3581 ;; match if there are correct tokens. Try to
3582 ;; widen to see if the limit was set in the
3583 ;; middle of one, else fall back to treating
3584 ;; the offending thing as a one character token.
3585 ((and limit
3586 (save-restriction
3587 (widen)
3588 (looking-at c-nonsymbol-token-regexp)))
3589 nil)
3590 (t
3591 (forward-char)
3592 t))))
3593 (c-forward-syntactic-ws)
3594 (setq prev last
3595 count (1- count)))
3596 (error (goto-char last)))
3597
3598 (when (eobp)
3599 (goto-char prev)
3600 (setq count (1+ count)))))
3601
3602 count)))
3603
3604(defun c-backward-token-2 (&optional count balanced limit)
3605 "Move backward by tokens.
3606See `c-forward-token-2' for details."
3607
3608 (or count (setq count 1))
3609 (if (< count 0)
3610 (- (c-forward-token-2 (- count) balanced limit))
3611
3612 (or limit (setq limit (point-min)))
3613 (let ((jump-syntax (if balanced
3614 c-jump-syntax-balanced
3615 c-jump-syntax-unbalanced))
3616 (last (point)))
3617
3618 (if (zerop count)
3619 ;; The count is zero so try to skip to the beginning of the
3620 ;; current token.
3621 (if (> (point)
3622 (progn (c-beginning-of-current-token) (point)))
3623 (if (< (point) limit)
3624 ;; The limit is inside the same token, so return 1.
3625 (setq count 1))
3626
3627 ;; We're not in the middle of a token. If there's
3628 ;; whitespace after the point then we must move backward,
3629 ;; so set count to 1 in that case.
3630 (and (looking-at c-syntactic-ws-start)
3631 ;; If we're looking at a '#' that might start a cpp
3632 ;; directive then we have to do a more elaborate check.
3633 (or (/= (char-after) ?#)
3634 (not c-opt-cpp-prefix)
3635 (save-excursion
3636 (and (= (point)
3637 (progn (beginning-of-line)
3638 (looking-at "[ \t]*")
3639 (match-end 0)))
3640 (or (bobp)
3641 (progn (backward-char)
3642 (not (eq (char-before) ?\\)))))))
3643 (setq count 1))))
3644
3645 ;; Use `condition-case' to avoid having to check for buffer
3646 ;; limits in `backward-char', `scan-sexps' and `goto-char' below.
3647 (condition-case nil
3648 (while (and
3649 (> count 0)
3650 (progn
3651 (c-backward-syntactic-ws)
3652 (backward-char)
3653 (if (looking-at jump-syntax)
3654 (goto-char (scan-sexps (1+ (point)) -1))
3655 ;; This can be very inefficient if there's a long
3656 ;; sequence of operator tokens without any separation.
3657 ;; That doesn't happen in practice, anyway.
3658 (c-beginning-of-current-token))
3659 (>= (point) limit)))
3660 (setq last (point)
3661 count (1- count)))
3662 (error (goto-char last)))
3663
3664 (if (< (point) limit)
3665 (goto-char last))
3666
3667 count)))
3668
3669(defun c-forward-token-1 (&optional count balanced limit)
3670 "Like `c-forward-token-2' but doesn't treat multicharacter operator
3671tokens like \"==\" as single tokens, i.e. all sequences of symbol
3672characters are jumped over character by character. This function is
3673for compatibility only; it's only a wrapper over `c-forward-token-2'."
3674 (let ((c-nonsymbol-token-regexp "\\s.\\|\\s\(\\|\\s\)"))
3675 (c-forward-token-2 count balanced limit)))
3676
3677(defun c-backward-token-1 (&optional count balanced limit)
3678 "Like `c-backward-token-2' but doesn't treat multicharacter operator
3679tokens like \"==\" as single tokens, i.e. all sequences of symbol
3680characters are jumped over character by character. This function is
3681for compatibility only; it's only a wrapper over `c-backward-token-2'."
3682 (let ((c-nonsymbol-token-regexp "\\s.\\|\\s\(\\|\\s\)"))
3683 (c-backward-token-2 count balanced limit)))
3684
3685\f
3686;; Tools for doing searches restricted to syntactically relevant text.
3687
3688(defun c-syntactic-re-search-forward (regexp &optional bound noerror
3689 paren-level not-inside-token
3690 lookbehind-submatch)
3691 "Like `re-search-forward', but only report matches that are found
3692in syntactically significant text. I.e. matches in comments, macros
3693or string literals are ignored. The start point is assumed to be
3694outside any comment, macro or string literal, or else the content of
3695that region is taken as syntactically significant text.
3696
3697If PAREN-LEVEL is non-nil, an additional restriction is added to
2a15eb73
MS
3698ignore matches in nested paren sexps. The search will also not go
3699outside the current list sexp, which has the effect that if the point
3700should be moved to BOUND when no match is found \(i.e. NOERROR is
3701neither nil nor t), then it will be at the closing paren if the end of
3702the current list sexp is encountered first.
d9e94c22
MS
3703
3704If NOT-INSIDE-TOKEN is non-nil, matches in the middle of tokens are
3705ignored. Things like multicharacter operators and special symbols
3706\(e.g. \"`()\" in Pike) are handled but currently not floating point
3707constants.
3708
3709If LOOKBEHIND-SUBMATCH is non-nil, it's taken as a number of a
3710subexpression in REGEXP. The end of that submatch is used as the
3711position to check for syntactic significance. If LOOKBEHIND-SUBMATCH
3712isn't used or if that subexpression didn't match then the start
3713position of the whole match is used instead. The \"look behind\"
3714subexpression is never tested before the starting position, so it
3715might be a good idea to include \\=\\= as a match alternative in it.
3716
3717Optimization note: Matches might be missed if the \"look behind\"
2a15eb73 3718subexpression can match the end of nonwhite syntactic whitespace,
d9e94c22 3719i.e. the end of comments or cpp directives. This since the function
2a15eb73
MS
3720skips over such things before resuming the search. It's on the other
3721hand not safe to assume that the \"look behind\" subexpression never
3722matches syntactic whitespace.
3723
3724Bug: Unbalanced parens inside cpp directives are currently not handled
3725correctly \(i.e. they don't get ignored as they should) when
0386b551
AM
3726PAREN-LEVEL is set.
3727
3728Note that this function might do hidden buffer changes. See the
3729comment at the start of cc-engine.el for more info."
d9e94c22
MS
3730
3731 (or bound (setq bound (point-max)))
3732 (if paren-level (setq paren-level -1))
3733
3734 ;;(message "c-syntactic-re-search-forward %s %s %S" (point) bound regexp)
3735
3736 (let ((start (point))
2a15eb73
MS
3737 tmp
3738 ;; Start position for the last search.
3739 search-pos
3740 ;; The `parse-partial-sexp' state between the start position
3741 ;; and the point.
3742 state
3743 ;; The current position after the last state update. The next
3744 ;; `parse-partial-sexp' continues from here.
3745 (state-pos (point))
3746 ;; The position at which to check the state and the state
3747 ;; there. This is separate from `state-pos' since we might
3748 ;; need to back up before doing the next search round.
3749 check-pos check-state
3750 ;; Last position known to end a token.
d9e94c22 3751 (last-token-end-pos (point-min))
2a15eb73
MS
3752 ;; Set when a valid match is found.
3753 found)
d9e94c22
MS
3754
3755 (condition-case err
3756 (while
3757 (and
2a15eb73
MS
3758 (progn
3759 (setq search-pos (point))
3760 (re-search-forward regexp bound noerror))
d9e94c22
MS
3761
3762 (progn
2a15eb73
MS
3763 (setq state (parse-partial-sexp
3764 state-pos (match-beginning 0) paren-level nil state)
3765 state-pos (point))
d9e94c22 3766 (if (setq check-pos (and lookbehind-submatch
2a15eb73
MS
3767 (or (not paren-level)
3768 (>= (car state) 0))
d9e94c22
MS
3769 (match-end lookbehind-submatch)))
3770 (setq check-state (parse-partial-sexp
2a15eb73
MS
3771 state-pos check-pos paren-level nil state))
3772 (setq check-pos state-pos
d9e94c22
MS
3773 check-state state))
3774
2a15eb73
MS
3775 ;; NOTE: If we got a look behind subexpression and get
3776 ;; an insignificant match in something that isn't
d9e94c22
MS
3777 ;; syntactic whitespace (i.e. strings or in nested
3778 ;; parentheses), then we can never skip more than a
2a15eb73
MS
3779 ;; single character from the match start position
3780 ;; (i.e. `state-pos' here) before continuing the
3781 ;; search. That since the look behind subexpression
3782 ;; might match the end of the insignificant region in
3783 ;; the next search.
d9e94c22
MS
3784
3785 (cond
d9e94c22
MS
3786 ((elt check-state 7)
3787 ;; Match inside a line comment. Skip to eol. Use
3788 ;; `re-search-forward' instead of `skip-chars-forward' to get
3789 ;; the right bound behavior.
3790 (re-search-forward "[\n\r]" bound noerror))
3791
3792 ((elt check-state 4)
3793 ;; Match inside a block comment. Skip to the '*/'.
3794 (search-forward "*/" bound noerror))
3795
3796 ((and (not (elt check-state 5))
3797 (eq (char-before check-pos) ?/)
2a15eb73 3798 (not (c-get-char-property (1- check-pos) 'syntax-table))
d9e94c22
MS
3799 (memq (char-after check-pos) '(?/ ?*)))
3800 ;; Match in the middle of the opener of a block or line
3801 ;; comment.
3802 (if (= (char-after check-pos) ?/)
3803 (re-search-forward "[\n\r]" bound noerror)
3804 (search-forward "*/" bound noerror)))
3805
2a15eb73
MS
3806 ;; The last `parse-partial-sexp' above might have
3807 ;; stopped short of the real check position if the end
3808 ;; of the current sexp was encountered in paren-level
3809 ;; mode. The checks above are always false in that
3810 ;; case, and since they can do better skipping in
3811 ;; lookbehind-submatch mode, we do them before
3812 ;; checking the paren level.
3813
3814 ((and paren-level
3815 (/= (setq tmp (car check-state)) 0))
3816 ;; Check the paren level first since we're short of the
3817 ;; syntactic checking position if the end of the
3818 ;; current sexp was encountered by `parse-partial-sexp'.
3819 (if (> tmp 0)
3820
3821 ;; Inside a nested paren sexp.
3822 (if lookbehind-submatch
3823 ;; See the NOTE above.
3824 (progn (goto-char state-pos) t)
3825 ;; Skip out of the paren quickly.
3826 (setq state (parse-partial-sexp state-pos bound 0 nil state)
3827 state-pos (point)))
3828
3829 ;; Have exited the current paren sexp.
3830 (if noerror
3831 (progn
3832 ;; The last `parse-partial-sexp' call above
3833 ;; has left us just after the closing paren
3834 ;; in this case, so we can modify the bound
3835 ;; to leave the point at the right position
3836 ;; upon return.
3837 (setq bound (1- (point)))
3838 nil)
3839 (signal 'search-failed (list regexp)))))
3840
3841 ((setq tmp (elt check-state 3))
3842 ;; Match inside a string.
3843 (if (or lookbehind-submatch
3844 (not (integerp tmp)))
3845 ;; See the NOTE above.
3846 (progn (goto-char state-pos) t)
3847 ;; Skip to the end of the string before continuing.
3848 (let ((ender (make-string 1 tmp)) (continue t))
3849 (while (if (search-forward ender bound noerror)
3850 (progn
3851 (setq state (parse-partial-sexp
3852 state-pos (point) nil nil state)
3853 state-pos (point))
3854 (elt state 3))
3855 (setq continue nil)))
3856 continue)))
d9e94c22
MS
3857
3858 ((save-excursion
3859 (save-match-data
3860 (c-beginning-of-macro start)))
3861 ;; Match inside a macro. Skip to the end of it.
3862 (c-end-of-macro)
3863 (cond ((<= (point) bound) t)
3864 (noerror nil)
2a15eb73 3865 (t (signal 'search-failed (list regexp)))))
d9e94c22 3866
2a15eb73
MS
3867 ((and not-inside-token
3868 (or (< check-pos last-token-end-pos)
3869 (< check-pos
3870 (save-excursion
3871 (goto-char check-pos)
3872 (save-match-data
3873 (c-end-of-current-token last-token-end-pos))
3874 (setq last-token-end-pos (point))))))
3875 ;; Inside a token.
3876 (if lookbehind-submatch
3877 ;; See the NOTE above.
3878 (goto-char state-pos)
3879 (goto-char (min last-token-end-pos bound))))
d9e94c22
MS
3880
3881 (t
3882 ;; A real match.
3883 (setq found t)
2a15eb73
MS
3884 nil)))
3885
3886 ;; Should loop to search again, but take care to avoid
3887 ;; looping on the same spot.
3888 (or (/= search-pos (point))
3889 (if (= (point) bound)
3890 (if noerror
3891 nil
3892 (signal 'search-failed (list regexp)))
3893 (forward-char)
3894 t))))
d9e94c22
MS
3895
3896 (error
3897 (goto-char start)
3898 (signal (car err) (cdr err))))
3899
2a15eb73 3900 ;;(message "c-syntactic-re-search-forward done %s" (or (match-end 0) (point)))
d9e94c22
MS
3901
3902 (if found
3903 (progn
2a15eb73
MS
3904 (goto-char (match-end 0))
3905 (match-end 0))
d9e94c22
MS
3906
3907 ;; Search failed. Set point as appropriate.
2a15eb73
MS
3908 (if (eq noerror t)
3909 (goto-char start)
3910 (goto-char bound))
d9e94c22
MS
3911 nil)))
3912
47641aac
GM
3913(defvar safe-pos-list) ; bound in c-syntactic-skip-backward
3914
d0fcee66
AM
3915(defsubst c-ssb-lit-begin ()
3916 ;; Return the start of the literal point is in, or nil.
3917 ;; We read and write the variables `safe-pos', `safe-pos-list', `state'
3918 ;; bound in the caller.
3919
3920 ;; Use `parse-partial-sexp' from a safe position down to the point to check
3921 ;; if it's outside comments and strings.
3922 (save-excursion
3923 (let ((pos (point)) safe-pos state pps-end-pos)
3924 ;; Pick a safe position as close to the point as possible.
3925 ;;
3926 ;; FIXME: Consult `syntax-ppss' here if our cache doesn't give a good
3927 ;; position.
47641aac 3928
d0fcee66
AM
3929 (while (and safe-pos-list
3930 (> (car safe-pos-list) (point)))
3931 (setq safe-pos-list (cdr safe-pos-list)))
3932 (unless (setq safe-pos (car-safe safe-pos-list))
3933 (setq safe-pos (max (or (c-safe-position
3934 (point) (or c-state-cache
3935 (c-parse-state)))
3936 0)
3937 (point-min))
3938 safe-pos-list (list safe-pos)))
3939
3940 ;; Cache positions along the way to use if we have to back up more. We
3941 ;; cache every closing paren on the same level. If the paren cache is
3942 ;; relevant in this region then we're typically already on the same
3943 ;; level as the target position. Note that we might cache positions
3944 ;; after opening parens in case safe-pos is in a nested list. That's
3945 ;; both uncommon and harmless.
3946 (while (progn
3947 (setq state (parse-partial-sexp
3948 safe-pos pos 0))
3949 (< (point) pos))
3950 (setq safe-pos (point)
3951 safe-pos-list (cons safe-pos safe-pos-list)))
3952
3953 ;; If the state contains the start of the containing sexp we cache that
3954 ;; position too, so that parse-partial-sexp in the next run has a bigger
3955 ;; chance of starting at the same level as the target position and thus
3956 ;; will get more good safe positions into the list.
3957 (if (elt state 1)
3958 (setq safe-pos (1+ (elt state 1))
3959 safe-pos-list (cons safe-pos safe-pos-list)))
3960
3961 (if (or (elt state 3) (elt state 4))
3962 ;; Inside string or comment. Continue search at the
3963 ;; beginning of it.
3964 (elt state 8)))))
3965
0386b551 3966(defun c-syntactic-skip-backward (skip-chars &optional limit paren-level)
d9e94c22
MS
3967 "Like `skip-chars-backward' but only look at syntactically relevant chars,
3968i.e. don't stop at positions inside syntactic whitespace or string
3969literals. Preprocessor directives are also ignored, with the exception
3970of the one that the point starts within, if any. If LIMIT is given,
0386b551
AM
3971it's assumed to be at a syntactically relevant position.
3972
3973If PAREN-LEVEL is non-nil, the function won't stop in nested paren
3974sexps, and the search will also not go outside the current paren sexp.
3975However, if LIMIT or the buffer limit is reached inside a nested paren
3976then the point will be left at the limit.
3977
3978Non-nil is returned if the point moved, nil otherwise.
3979
3980Note that this function might do hidden buffer changes. See the
3981comment at the start of cc-engine.el for more info."
d9e94c22
MS
3982
3983 (let ((start (point))
d0fcee66 3984 state-2
d9e94c22
MS
3985 ;; A list of syntactically relevant positions in descending
3986 ;; order. It's used to avoid scanning repeatedly over
3987 ;; potentially large regions with `parse-partial-sexp' to verify
d0fcee66 3988 ;; each position. Used in `c-ssb-lit-begin'
d9e94c22
MS
3989 safe-pos-list
3990 ;; The result from `c-beginning-of-macro' at the start position or the
3991 ;; start position itself if it isn't within a macro. Evaluated on
3992 ;; demand.
0386b551
AM
3993 start-macro-beg
3994 ;; The earliest position after the current one with the same paren
3995 ;; level. Used only when `paren-level' is set.
d0fcee66 3996 lit-beg
0386b551 3997 (paren-level-pos (point)))
d9e94c22 3998
d0fcee66
AM
3999 (while
4000 (progn
4001 ;; The next loop "tries" to find the end point each time round,
4002 ;; loops when it hasn't succeeded.
4003 (while
4004 (and
4005 (< (skip-chars-backward skip-chars limit) 0)
4006
4007 (let ((pos (point)) state-2 pps-end-pos)
4008
4009 (cond
4010 ;; Don't stop inside a literal
4011 ((setq lit-beg (c-ssb-lit-begin))
4012 (goto-char lit-beg)
4013 t)
4014
4015 ((and paren-level
4016 (save-excursion
4017 (setq state-2 (parse-partial-sexp
4018 pos paren-level-pos -1)
4019 pps-end-pos (point))
4020 (/= (car state-2) 0)))
4021 ;; Not at the right level.
4022
4023 (if (and (< (car state-2) 0)
4024 ;; We stop above if we go out of a paren.
4025 ;; Now check whether it precedes or is
4026 ;; nested in the starting sexp.
4027 (save-excursion
4028 (setq state-2
4029 (parse-partial-sexp
4030 pps-end-pos paren-level-pos
4031 nil nil state-2))
4032 (< (car state-2) 0)))
4033
4034 ;; We've stopped short of the starting position
4035 ;; so the hit was inside a nested list. Go up
4036 ;; until we are at the right level.
4037 (condition-case nil
4038 (progn
4039 (goto-char (scan-lists pos -1
4040 (- (car state-2))))
4041 (setq paren-level-pos (point))
4042 (if (and limit (>= limit paren-level-pos))
0386b551 4043 (progn
d0fcee66
AM
4044 (goto-char limit)
4045 nil)
4046 t))
4047 (error
4048 (goto-char (or limit (point-min)))
4049 nil))
4050
4051 ;; The hit was outside the list at the start
4052 ;; position. Go to the start of the list and exit.
4053 (goto-char (1+ (elt state-2 1)))
4054 nil))
4055
4056 ((c-beginning-of-macro limit)
4057 ;; Inside a macro.
4058 (if (< (point)
4059 (or start-macro-beg
4060 (setq start-macro-beg
4061 (save-excursion
4062 (goto-char start)
4063 (c-beginning-of-macro limit)
4064 (point)))))
4065 t
4066
4067 ;; It's inside the same macro we started in so it's
4068 ;; a relevant match.
4069 (goto-char pos)
4070 nil))))))
91af3942 4071
d0fcee66
AM
4072 (> (point)
4073 (progn
4074 ;; Skip syntactic ws afterwards so that we don't stop at the
4075 ;; end of a comment if `skip-chars' is something like "^/".
4076 (c-backward-syntactic-ws)
4077 (point)))))
d9e94c22 4078
0386b551
AM
4079 ;; We might want to extend this with more useful return values in
4080 ;; the future.
4081 (/= (point) start)))
4082
4083;; The following is an alternative implementation of
4084;; `c-syntactic-skip-backward' that uses backward movement to keep
4085;; track of the syntactic context. It turned out to be generally
4086;; slower than the one above which uses forward checks from earlier
4087;; safe positions.
4088;;
4089;;(defconst c-ssb-stop-re
4090;; ;; The regexp matching chars `c-syntactic-skip-backward' needs to
4091;; ;; stop at to avoid going into comments and literals.
4092;; (concat
4093;; ;; Match comment end syntax and string literal syntax. Also match
4094;; ;; '/' for block comment endings (not covered by comment end
4095;; ;; syntax).
4096;; "\\s>\\|/\\|\\s\""
4097;; (if (memq 'gen-string-delim c-emacs-features)
4098;; "\\|\\s|"
4099;; "")
4100;; (if (memq 'gen-comment-delim c-emacs-features)
4101;; "\\|\\s!"
4102;; "")))
4103;;
4104;;(defconst c-ssb-stop-paren-re
4105;; ;; Like `c-ssb-stop-re' but also stops at paren chars.
4106;; (concat c-ssb-stop-re "\\|\\s(\\|\\s)"))
4107;;
4108;;(defconst c-ssb-sexp-end-re
4109;; ;; Regexp matching the ending syntax of a complex sexp.
4110;; (concat c-string-limit-regexp "\\|\\s)"))
4111;;
4112;;(defun c-syntactic-skip-backward (skip-chars &optional limit paren-level)
4113;; "Like `skip-chars-backward' but only look at syntactically relevant chars,
4114;;i.e. don't stop at positions inside syntactic whitespace or string
4115;;literals. Preprocessor directives are also ignored. However, if the
4116;;point is within a comment, string literal or preprocessor directory to
4117;;begin with, its contents is treated as syntactically relevant chars.
4118;;If LIMIT is given, it limits the backward search and the point will be
4119;;left there if no earlier position is found.
4120;;
4121;;If PAREN-LEVEL is non-nil, the function won't stop in nested paren
4122;;sexps, and the search will also not go outside the current paren sexp.
4123;;However, if LIMIT or the buffer limit is reached inside a nested paren
4124;;then the point will be left at the limit.
4125;;
4126;;Non-nil is returned if the point moved, nil otherwise.
4127;;
4128;;Note that this function might do hidden buffer changes. See the
4129;;comment at the start of cc-engine.el for more info."
4130;;
4131;; (save-restriction
4132;; (when limit
4133;; (narrow-to-region limit (point-max)))
4134;;
4135;; (let ((start (point)))
4136;; (catch 'done
4137;; (while (let ((last-pos (point))
4138;; (stop-pos (progn
4139;; (skip-chars-backward skip-chars)
4140;; (point))))
4141;;
4142;; ;; Skip back over the same region as
4143;; ;; `skip-chars-backward' above, but keep to
4144;; ;; syntactically relevant positions.
4145;; (goto-char last-pos)
4146;; (while (and
4147;; ;; `re-search-backward' with a single char regexp
4148;; ;; should be fast.
4149;; (re-search-backward
4150;; (if paren-level c-ssb-stop-paren-re c-ssb-stop-re)
4151;; stop-pos 'move)
4152;;
4153;; (progn
4154;; (cond
4155;; ((looking-at "\\s(")
4156;; ;; `paren-level' is set and we've found the
4157;; ;; start of the containing paren.
4158;; (forward-char)
4159;; (throw 'done t))
4160;;
4161;; ((looking-at c-ssb-sexp-end-re)
4162;; ;; We're at the end of a string literal or paren
4163;; ;; sexp (if `paren-level' is set).
4164;; (forward-char)
4165;; (condition-case nil
4166;; (c-backward-sexp)
4167;; (error
4168;; (goto-char limit)
4169;; (throw 'done t))))
4170;;
4171;; (t
4172;; (forward-char)
4173;; ;; At the end of some syntactic ws or possibly
4174;; ;; after a plain '/' operator.
4175;; (let ((pos (point)))
4176;; (c-backward-syntactic-ws)
4177;; (if (= pos (point))
4178;; ;; Was a plain '/' operator. Go past it.
4179;; (backward-char)))))
4180;;
4181;; (> (point) stop-pos))))
4182;;
4183;; ;; Now the point is either at `stop-pos' or at some
4184;; ;; position further back if `stop-pos' was at a
4185;; ;; syntactically irrelevant place.
4186;;
4187;; ;; Skip additional syntactic ws so that we don't stop
4188;; ;; at the end of a comment if `skip-chars' is
4189;; ;; something like "^/".
4190;; (c-backward-syntactic-ws)
4191;;
4192;; (< (point) stop-pos))))
4193;;
4194;; ;; We might want to extend this with more useful return values
4195;; ;; in the future.
4196;; (/= (point) start))))
d9e94c22
MS
4197
4198\f
4199;; Tools for handling comments and string literals.
4200
13d49cbb 4201(defun c-in-literal (&optional lim detect-cpp)
d9e94c22
MS
4202 "Return the type of literal point is in, if any.
4203The return value is `c' if in a C-style comment, `c++' if in a C++
4204style comment, `string' if in a string literal, `pound' if DETECT-CPP
4205is non-nil and in a preprocessor line, or nil if somewhere else.
4206Optional LIM is used as the backward limit of the search. If omitted,
4207or nil, `c-beginning-of-defun' is used.
4208
4209The last point calculated is cached if the cache is enabled, i.e. if
4210`c-in-literal-cache' is bound to a two element vector.
4211
0386b551
AM
4212Note that this function might do hidden buffer changes. See the
4213comment at the start of cc-engine.el for more info."
13d49cbb
AM
4214 (let* ((safe-place (c-state-safe-place (point)))
4215 (lit (c-state-pp-to-literal safe-place (point))))
4216 (or (cadr lit)
4217 (and detect-cpp
4218 (save-excursion (c-beginning-of-macro))
4219 'pound))))
d9e94c22
MS
4220
4221(defun c-literal-limits (&optional lim near not-in-delimiter)
4222 "Return a cons of the beginning and end positions of the comment or
4223string surrounding point (including both delimiters), or nil if point
4224isn't in one. If LIM is non-nil, it's used as the \"safe\" position
4225to start parsing from. If NEAR is non-nil, then the limits of any
4226literal next to point is returned. \"Next to\" means there's only
4227spaces and tabs between point and the literal. The search for such a
4228literal is done first in forward direction. If NOT-IN-DELIMITER is
4229non-nil, the case when point is inside a starting delimiter won't be
a85fd6da 4230recognized. This only has effect for comments which have starting
d9e94c22
MS
4231delimiters with more than one character.
4232
0386b551
AM
4233Note that this function might do hidden buffer changes. See the
4234comment at the start of cc-engine.el for more info."
d9e94c22
MS
4235
4236 (save-excursion
4237 (let* ((pos (point))
13d49cbb
AM
4238 (lim (or lim (c-state-safe-place pos)))
4239 (pp-to-lit (c-state-pp-to-literal lim pos))
4240 (state (car pp-to-lit))
4241 (lit-type (cadr pp-to-lit))
4242 (lit-limits (car (cddr pp-to-lit))))
d9e94c22 4243
13d49cbb
AM
4244 (cond
4245 (lit-limits)
4246 ((and (not not-in-delimiter)
4247 (not (elt state 5))
4248 (eq (char-before) ?/)
4249 (looking-at "[/*]")) ; FIXME!!! use c-line/block-comment-starter. 2008-09-28.
4250 ;; We're standing in a comment starter.
4251 (backward-char 1)
4252 (cons (point) (progn (c-forward-single-comment) (point))))
4253
4254 (near
4255 (goto-char pos)
4256 ;; Search forward for a literal.
4257 (skip-chars-forward " \t")
4258 (cond
4259 ((looking-at c-string-limit-regexp) ; String.
4260 (cons (point) (or (c-safe (c-forward-sexp 1) (point))
4261 (point-max))))
d9e94c22 4262
13d49cbb
AM
4263 ((looking-at c-comment-start-regexp) ; Line or block comment.
4264 (cons (point) (progn (c-forward-single-comment) (point))))
4265
4266 (t
4267 ;; Search backward.
4268 (skip-chars-backward " \t")
4269
4270 (let ((end (point)) beg)
4271 (cond
4272 ((save-excursion
4273 (< (skip-syntax-backward c-string-syntax) 0)) ; String.
4274 (setq beg (c-safe (c-backward-sexp 1) (point))))
4275
4276 ((and (c-safe (forward-char -2) t)
4277 (looking-at "*/"))
4278 ;; Block comment. Due to the nature of line
4279 ;; comments, they will always be covered by the
4280 ;; normal case above.
4281 (goto-char end)
4282 (c-backward-single-comment)
4283 ;; If LIM is bogus, beg will be bogus.
4284 (setq beg (point))))
4285
4286 (if beg (cons beg end))))))
4287 ))))
d9e94c22 4288
0386b551
AM
4289;; In case external callers use this; it did have a docstring.
4290(defalias 'c-literal-limits-fast 'c-literal-limits)
d9e94c22 4291
0386b551
AM
4292(defun c-collect-line-comments (range)
4293 "If the argument is a cons of two buffer positions (such as returned by
4294`c-literal-limits'), and that range contains a C++ style line comment,
4295then an extended range is returned that contains all adjacent line
4296comments (i.e. all comments that starts in the same column with no
4297empty lines or non-whitespace characters between them). Otherwise the
4298argument is returned.
d9e94c22 4299
0386b551
AM
4300Note that this function might do hidden buffer changes. See the
4301comment at the start of cc-engine.el for more info."
d9e94c22
MS
4302
4303 (save-excursion
0386b551
AM
4304 (condition-case nil
4305 (if (and (consp range) (progn
4306 (goto-char (car range))
4307 (looking-at c-line-comment-starter)))
b414f371 4308 (let ((col (current-column))
0386b551
AM
4309 (beg (point))
4310 (bopl (c-point 'bopl))
4311 (end (cdr range)))
4312 ;; Got to take care in the backward direction to handle
4313 ;; comments which are preceded by code.
4314 (while (and (c-backward-single-comment)
4315 (>= (point) bopl)
4316 (looking-at c-line-comment-starter)
4317 (= col (current-column)))
4318 (setq beg (point)
4319 bopl (c-point 'bopl)))
4320 (goto-char end)
4321 (while (and (progn (skip-chars-forward " \t")
4322 (looking-at c-line-comment-starter))
4323 (= col (current-column))
4324 (prog1 (zerop (forward-line 1))
4325 (setq end (point)))))
4326 (cons beg end))
4327 range)
4328 (error range))))
d9e94c22
MS
4329
4330(defun c-literal-type (range)
4331 "Convenience function that given the result of `c-literal-limits',
a85fd6da
AM
4332returns nil or the type of literal that the range surrounds, one
4333of the symbols 'c, 'c++ or 'string. It's much faster than using
4334`c-in-literal' and is intended to be used when you need both the
4335type of a literal and its limits.
d9e94c22 4336
0386b551
AM
4337Note that this function might do hidden buffer changes. See the
4338comment at the start of cc-engine.el for more info."
4339
d9e94c22
MS
4340 (if (consp range)
4341 (save-excursion
4342 (goto-char (car range))
4343 (cond ((looking-at c-string-limit-regexp) 'string)
4344 ((or (looking-at "//") ; c++ line comment
4345 (and (looking-at "\\s<") ; comment starter
4346 (looking-at "#"))) ; awk comment.
4347 'c++)
4348 (t 'c))) ; Assuming the range is valid.
4349 range))
4350
4351\f
4352;; `c-find-decl-spots' and accompanying stuff.
4353
4354;; Variables used in `c-find-decl-spots' to cache the search done for
4355;; the first declaration in the last call. When that function starts,
4356;; it needs to back up over syntactic whitespace to look at the last
4357;; token before the region being searched. That can sometimes cause
4358;; moves back and forth over a quite large region of comments and
4359;; macros, which would be repeated for each changed character when
4360;; we're called during fontification, since font-lock refontifies the
4361;; current line for each change. Thus it's worthwhile to cache the
4362;; first match.
4363;;
4364;; `c-find-decl-syntactic-pos' is a syntactically relevant position in
4365;; the syntactic whitespace less or equal to some start position.
4366;; There's no cached value if it's nil.
4367;;
4368;; `c-find-decl-match-pos' is the match position if
4369;; `c-find-decl-prefix-search' matched before the syntactic whitespace
4370;; at `c-find-decl-syntactic-pos', or nil if there's no such match.
4371(defvar c-find-decl-syntactic-pos nil)
4372(make-variable-buffer-local 'c-find-decl-syntactic-pos)
4373(defvar c-find-decl-match-pos nil)
4374(make-variable-buffer-local 'c-find-decl-match-pos)
4375
4376(defsubst c-invalidate-find-decl-cache (change-min-pos)
4377 (and c-find-decl-syntactic-pos
4378 (< change-min-pos c-find-decl-syntactic-pos)
4379 (setq c-find-decl-syntactic-pos nil)))
4380
4381; (defface c-debug-decl-spot-face
4382; '((t (:background "Turquoise")))
4383; "Debug face to mark the spots where `c-find-decl-spots' stopped.")
4384; (defface c-debug-decl-sws-face
4385; '((t (:background "Khaki")))
4386; "Debug face to mark the syntactic whitespace between the declaration
4387; spots and the preceding token end.")
4388
4389(defmacro c-debug-put-decl-spot-faces (match-pos decl-pos)
4390 (when (facep 'c-debug-decl-spot-face)
0386b551 4391 `(c-save-buffer-state ((match-pos ,match-pos) (decl-pos ,decl-pos))
d9e94c22
MS
4392 (c-debug-add-face (max match-pos (point-min)) decl-pos
4393 'c-debug-decl-sws-face)
4394 (c-debug-add-face decl-pos (min (1+ decl-pos) (point-max))
4395 'c-debug-decl-spot-face))))
4396(defmacro c-debug-remove-decl-spot-faces (beg end)
4397 (when (facep 'c-debug-decl-spot-face)
0386b551 4398 `(c-save-buffer-state ()
d9e94c22
MS
4399 (c-debug-remove-face ,beg ,end 'c-debug-decl-spot-face)
4400 (c-debug-remove-face ,beg ,end 'c-debug-decl-sws-face))))
4401
4402(defmacro c-find-decl-prefix-search ()
4403 ;; Macro used inside `c-find-decl-spots'. It ought to be a defun,
4404 ;; but it contains lots of free variables that refer to things
4405 ;; inside `c-find-decl-spots'. The point is left at `cfd-match-pos'
4406 ;; if there is a match, otherwise at `cfd-limit'.
0386b551
AM
4407 ;;
4408 ;; This macro might do hidden buffer changes.
d9e94c22
MS
4409
4410 '(progn
4411 ;; Find the next property match position if we haven't got one already.
4412 (unless cfd-prop-match
4413 (save-excursion
4414 (while (progn
4415 (goto-char (next-single-property-change
4416 (point) 'c-type nil cfd-limit))
4417 (and (< (point) cfd-limit)
4418 (not (eq (c-get-char-property (1- (point)) 'c-type)
4419 'c-decl-end)))))
4420 (setq cfd-prop-match (point))))
4421
0386b551
AM
4422 ;; Find the next `c-decl-prefix-or-start-re' match if we haven't
4423 ;; got one already.
d9e94c22 4424 (unless cfd-re-match
0386b551
AM
4425
4426 (if (> cfd-re-match-end (point))
4427 (goto-char cfd-re-match-end))
4428
4429 (while (if (setq cfd-re-match-end
4430 (re-search-forward c-decl-prefix-or-start-re
4431 cfd-limit 'move))
4432
4433 ;; Match. Check if it's inside a comment or string literal.
4434 (c-got-face-at
4435 (if (setq cfd-re-match (match-end 1))
4436 ;; Matched the end of a token preceding a decl spot.
4437 (progn
4438 (goto-char cfd-re-match)
4439 (1- cfd-re-match))
4440 ;; Matched a token that start a decl spot.
4441 (goto-char (match-beginning 0))
4442 (point))
4443 c-literal-faces)
4444
4445 ;; No match. Finish up and exit the loop.
4446 (setq cfd-re-match cfd-limit)
4447 nil)
4448
4449 ;; Skip out of comments and string literals.
d9e94c22
MS
4450 (while (progn
4451 (goto-char (next-single-property-change
0386b551 4452 (point) 'face nil cfd-limit))
d9e94c22 4453 (and (< (point) cfd-limit)
0386b551
AM
4454 (c-got-face-at (point) c-literal-faces)))))
4455
4456 ;; If we matched at the decl start, we have to back up over the
4457 ;; preceding syntactic ws to set `cfd-match-pos' and to catch
4458 ;; any decl spots in the syntactic ws.
d9e94c22 4459 (unless cfd-re-match
0386b551
AM
4460 (c-backward-syntactic-ws)
4461 (setq cfd-re-match (point))))
d9e94c22
MS
4462
4463 ;; Choose whichever match is closer to the start.
4464 (if (< cfd-re-match cfd-prop-match)
4465 (setq cfd-match-pos cfd-re-match
4466 cfd-re-match nil)
4467 (setq cfd-match-pos cfd-prop-match
4468 cfd-prop-match nil))
4469
4470 (goto-char cfd-match-pos)
4471
4472 (when (< cfd-match-pos cfd-limit)
4473 ;; Skip forward past comments only so we don't skip macros.
4474 (c-forward-comments)
4475 ;; Set the position to continue at. We can avoid going over
4476 ;; the comments skipped above a second time, but it's possible
4477 ;; that the comment skipping has taken us past `cfd-prop-match'
4478 ;; since the property might be used inside comments.
4479 (setq cfd-continue-pos (if cfd-prop-match
4480 (min cfd-prop-match (point))
4481 (point))))))
4482
4483(defun c-find-decl-spots (cfd-limit cfd-decl-re cfd-face-checklist cfd-fun)
0386b551 4484 ;; Call CFD-FUN for each possible spot for a declaration, cast or
a85fd6da
AM
4485 ;; label from the point to CFD-LIMIT.
4486 ;;
4487 ;; CFD-FUN is called with point at the start of the spot. It's
4488 ;; passed two arguments: The first is the end position of the token
4489 ;; preceding the spot, or 0 for the implicit match at bob. The
4490 ;; second is a flag that is t when the match is inside a macro. If
4491 ;; CFD-FUN adds `c-decl-end' properties somewhere below the current
4492 ;; spot, it should return non-nil to ensure that the next search
4493 ;; will find them.
0386b551 4494 ;;
a85fd6da 4495 ;; Such a spot is:
0386b551
AM
4496 ;; o The first token after bob.
4497 ;; o The first token after the end of submatch 1 in
4498 ;; `c-decl-prefix-or-start-re' when that submatch matches.
4499 ;; o The start of each `c-decl-prefix-or-start-re' match when
4500 ;; submatch 1 doesn't match.
5a89f0a7 4501 ;; o The first token after the end of each occurrence of the
0386b551
AM
4502 ;; `c-type' text property with the value `c-decl-end', provided
4503 ;; `c-type-decl-end-used' is set.
4504 ;;
4505 ;; Only a spot that match CFD-DECL-RE and whose face is in the
4506 ;; CFD-FACE-CHECKLIST list causes CFD-FUN to be called. The face
4507 ;; check is disabled if CFD-FACE-CHECKLIST is nil.
d9e94c22
MS
4508 ;;
4509 ;; If the match is inside a macro then the buffer is narrowed to the
4510 ;; end of it, so that CFD-FUN can investigate the following tokens
4511 ;; without matching something that begins inside a macro and ends
4512 ;; outside it. It's to avoid this work that the CFD-DECL-RE and
4513 ;; CFD-FACE-CHECKLIST checks exist.
4514 ;;
0386b551
AM
4515 ;; The spots are visited approximately in order from top to bottom.
4516 ;; It's however the positions where `c-decl-prefix-or-start-re'
4517 ;; matches and where `c-decl-end' properties are found that are in
4518 ;; order. Since the spots often are at the following token, they
4519 ;; might be visited out of order insofar as more spots are reported
4520 ;; later on within the syntactic whitespace between the match
4521 ;; positions and their spots.
4522 ;;
4523 ;; It's assumed that comments and strings are fontified in the
d9e94c22
MS
4524 ;; searched range.
4525 ;;
4526 ;; This is mainly used in fontification, and so has an elaborate
4527 ;; cache to handle repeated calls from the same start position; see
4528 ;; the variables above.
4529 ;;
4530 ;; All variables in this function begin with `cfd-' to avoid name
4531 ;; collision with the (dynamically bound) variables used in CFD-FUN.
0386b551
AM
4532 ;;
4533 ;; This function might do hidden buffer changes.
d9e94c22 4534
0386b551
AM
4535 (let ((cfd-start-pos (point))
4536 (cfd-buffer-end (point-max))
4537 ;; The end of the token preceding the decl spot last found
4538 ;; with `c-decl-prefix-or-start-re'. `cfd-limit' if there's
4539 ;; no match.
d9e94c22 4540 cfd-re-match
0386b551
AM
4541 ;; The end position of the last `c-decl-prefix-or-start-re'
4542 ;; match. If this is greater than `cfd-continue-pos', the
4543 ;; next regexp search is started here instead.
4544 (cfd-re-match-end (point-min))
4545 ;; The end of the last `c-decl-end' found by
4546 ;; `c-find-decl-prefix-search'. `cfd-limit' if there's no
4547 ;; match. If searching for the property isn't needed then we
4548 ;; disable it by setting it to `cfd-limit' directly.
d9e94c22 4549 (cfd-prop-match (unless c-type-decl-end-used cfd-limit))
0386b551
AM
4550 ;; The end of the token preceding the decl spot last found by
4551 ;; `c-find-decl-prefix-search'. 0 for the implicit match at
4552 ;; bob. `cfd-limit' if there's no match. In other words,
4553 ;; this is the minimum of `cfd-re-match' and `cfd-prop-match'.
d9e94c22
MS
4554 (cfd-match-pos cfd-limit)
4555 ;; The position to continue searching at.
4556 cfd-continue-pos
4557 ;; The position of the last "real" token we've stopped at.
4558 ;; This can be greater than `cfd-continue-pos' when we get
4559 ;; hits inside macros or at `c-decl-end' positions inside
4560 ;; comments.
4561 (cfd-token-pos 0)
4562 ;; The end position of the last entered macro.
4563 (cfd-macro-end 0))
4564
4565 ;; Initialize by finding a syntactically relevant start position
0386b551
AM
4566 ;; before the point, and do the first `c-decl-prefix-or-start-re'
4567 ;; search unless we're at bob.
d9e94c22 4568
0386b551 4569 (let (start-in-literal start-in-macro syntactic-pos)
d9e94c22
MS
4570 ;; Must back up a bit since we look for the end of the previous
4571 ;; statement or declaration, which is earlier than the first
4572 ;; returned match.
4573
0386b551
AM
4574 (cond
4575 ;; First we need to move to a syntactically relevant position.
4576 ;; Begin by backing out of comment or string literals.
4577 ((and
4578 (when (c-got-face-at (point) c-literal-faces)
4579 ;; Try to use the faces to back up to the start of the
4580 ;; literal. FIXME: What if the point is on a declaration
4581 ;; inside a comment?
4582 (while (and (not (bobp))
4583 (c-got-face-at (1- (point)) c-literal-faces))
4584 (goto-char (previous-single-property-change
4585 (point) 'face nil (point-min))))
4586
4587 ;; XEmacs doesn't fontify the quotes surrounding string
4588 ;; literals.
4589 (and (featurep 'xemacs)
4590 (eq (get-text-property (point) 'face)
4591 'font-lock-string-face)
4592 (not (bobp))
4593 (progn (backward-char)
4594 (not (looking-at c-string-limit-regexp)))
4595 (forward-char))
4596
4597 ;; Don't trust the literal to contain only literal faces
4598 ;; (the font lock package might not have fontified the
4599 ;; start of it at all, for instance) so check that we have
4600 ;; arrived at something that looks like a start or else
4601 ;; resort to `c-literal-limits'.
4602 (unless (looking-at c-literal-start-regexp)
4603 (let ((range (c-literal-limits)))
4604 (if range (goto-char (car range)))))
4605
4606 (setq start-in-literal (point)))
4607
4608 ;; The start is in a literal. If the limit is in the same
4609 ;; one we don't have to find a syntactic position etc. We
4610 ;; only check that if the limit is at or before bonl to save
4611 ;; time; it covers the by far most common case when font-lock
4612 ;; refontifies the current line only.
4613 (<= cfd-limit (c-point 'bonl cfd-start-pos))
4614 (save-excursion
4615 (goto-char cfd-start-pos)
4616 (while (progn
4617 (goto-char (next-single-property-change
4618 (point) 'face nil cfd-limit))
4619 (and (< (point) cfd-limit)
4620 (c-got-face-at (point) c-literal-faces))))
4621 (= (point) cfd-limit)))
4622
4623 ;; Completely inside a literal. Set up variables to trig the
4624 ;; (< cfd-continue-pos cfd-start-pos) case below and it'll
4625 ;; find a suitable start position.
4626 (setq cfd-continue-pos start-in-literal))
4627
4628 ;; Check if the region might be completely inside a macro, to
4629 ;; optimize that like the completely-inside-literal above.
4630 ((save-excursion
4631 (and (= (forward-line 1) 0)
4632 (bolp) ; forward-line has funny behavior at eob.
4633 (>= (point) cfd-limit)
4634 (progn (backward-char)
4635 (eq (char-before) ?\\))))
4636 ;; (Maybe) completely inside a macro. Only need to trig the
4637 ;; (< cfd-continue-pos cfd-start-pos) case below to make it
4638 ;; set things up.
4639 (setq cfd-continue-pos (1- cfd-start-pos)
4640 start-in-macro t))
d9e94c22 4641
0386b551
AM
4642 (t
4643 ;; Back out of any macro so we don't miss any declaration
4644 ;; that could follow after it.
4645 (when (c-beginning-of-macro)
4646 (setq start-in-macro t))
4647
4648 ;; Now we're at a proper syntactically relevant position so we
4649 ;; can use the cache. But first clear it if it applied
4650 ;; further down.
4651 (c-invalidate-find-decl-cache cfd-start-pos)
4652
4653 (setq syntactic-pos (point))
4654 (unless (eq syntactic-pos c-find-decl-syntactic-pos)
4655 ;; Don't have to do this if the cache is relevant here,
4656 ;; typically if the same line is refontified again. If
4657 ;; we're just some syntactic whitespace further down we can
4658 ;; still use the cache to limit the skipping.
4659 (c-backward-syntactic-ws c-find-decl-syntactic-pos))
4660
4661 ;; If we hit `c-find-decl-syntactic-pos' and
4662 ;; `c-find-decl-match-pos' is set then we install the cached
4663 ;; values. If we hit `c-find-decl-syntactic-pos' and
4664 ;; `c-find-decl-match-pos' is nil then we know there's no decl
4665 ;; prefix in the whitespace before `c-find-decl-syntactic-pos'
4666 ;; and so we can continue the search from this point. If we
4667 ;; didn't hit `c-find-decl-syntactic-pos' then we're now in
4668 ;; the right spot to begin searching anyway.
4669 (if (and (eq (point) c-find-decl-syntactic-pos)
4670 c-find-decl-match-pos)
d9e94c22
MS
4671 (setq cfd-match-pos c-find-decl-match-pos
4672 cfd-continue-pos syntactic-pos)
0386b551
AM
4673
4674 (setq c-find-decl-syntactic-pos syntactic-pos)
4675
4676 (when (if (bobp)
4677 ;; Always consider bob a match to get the first
4678 ;; declaration in the file. Do this separately instead of
4679 ;; letting `c-decl-prefix-or-start-re' match bob, so that
4680 ;; regexp always can consume at least one character to
4681 ;; ensure that we won't get stuck in an infinite loop.
4682 (setq cfd-re-match 0)
4683 (backward-char)
4684 (c-beginning-of-current-token)
4685 (< (point) cfd-limit))
4686 ;; Do an initial search now. In the bob case above it's
4687 ;; only done to search for a `c-decl-end' spot.
4688 (c-find-decl-prefix-search))
4689
4690 (setq c-find-decl-match-pos (and (< cfd-match-pos cfd-start-pos)
4691 cfd-match-pos)))))
4692
4693 ;; Advance `cfd-continue-pos' if it's before the start position.
4694 ;; The closest continue position that might have effect at or
4695 ;; after the start depends on what we started in. This also
4696 ;; finds a suitable start position in the special cases when the
4697 ;; region is completely within a literal or macro.
4698 (when (and cfd-continue-pos (< cfd-continue-pos cfd-start-pos))
4699
4700 (cond
4701 (start-in-macro
4702 ;; If we're in a macro then it's the closest preceding token
4703 ;; in the macro. Check this before `start-in-literal',
4704 ;; since if we're inside a literal in a macro, the preceding
4705 ;; token is earlier than any `c-decl-end' spot inside the
4706 ;; literal (comment).
4707 (goto-char (or start-in-literal cfd-start-pos))
4708 ;; The only syntactic ws in macros are comments.
d9e94c22 4709 (c-backward-comments)
0386b551
AM
4710 (backward-char)
4711 (c-beginning-of-current-token))
4712
4713 (start-in-literal
4714 ;; If we're in a comment it can only be the closest
4715 ;; preceding `c-decl-end' position within that comment, if
4716 ;; any. Go back to the beginning of such a property so that
4717 ;; `c-find-decl-prefix-search' will find the end of it.
4718 ;; (Can't stop at the end and install it directly on
4719 ;; `cfd-prop-match' since that variable might be cleared
4720 ;; after `cfd-fun' below.)
4721 ;;
4722 ;; Note that if the literal is a string then the property
4723 ;; search will simply skip to the beginning of it right
4724 ;; away.
4725 (if (not c-type-decl-end-used)
4726 (goto-char start-in-literal)
4727 (goto-char cfd-start-pos)
4728 (while (progn
4729 (goto-char (previous-single-property-change
4730 (point) 'c-type nil start-in-literal))
4731 (and (> (point) start-in-literal)
4732 (not (eq (c-get-char-property (point) 'c-type)
4733 'c-decl-end))))))
4734
4735 (when (= (point) start-in-literal)
4736 ;; Didn't find any property inside the comment, so we can
4737 ;; skip it entirely. (This won't skip past a string, but
4738 ;; that'll be handled quickly by the next
4739 ;; `c-find-decl-prefix-search' anyway.)
4740 (c-forward-single-comment)
4741 (if (> (point) cfd-limit)
4742 (goto-char cfd-limit))))
d9e94c22 4743
0386b551
AM
4744 (t
4745 ;; If we started in normal code, the only match that might
4746 ;; apply before the start is what we already got in
4747 ;; `cfd-match-pos' so we can continue at the start position.
4748 ;; (Note that we don't get here if the first match is below
4749 ;; it.)
4750 (goto-char cfd-start-pos)))
4751
4752 ;; Delete found matches if they are before our new continue
4753 ;; position, so that `c-find-decl-prefix-search' won't back up
4754 ;; to them later on.
4755 (setq cfd-continue-pos (point))
4756 (when (and cfd-re-match (< cfd-re-match cfd-continue-pos))
4757 (setq cfd-re-match nil))
4758 (when (and cfd-prop-match (< cfd-prop-match cfd-continue-pos))
4759 (setq cfd-prop-match nil)))
4760
4761 (if syntactic-pos
4762 ;; This is the normal case and we got a proper syntactic
4763 ;; position. If there's a match then it's always outside
4764 ;; macros and comments, so advance to the next token and set
4765 ;; `cfd-token-pos'. The loop below will later go back using
4766 ;; `cfd-continue-pos' to fix declarations inside the
4767 ;; syntactic ws.
4768 (when (and cfd-match-pos (< cfd-match-pos syntactic-pos))
4769 (goto-char syntactic-pos)
4770 (c-forward-syntactic-ws)
4771 (and cfd-continue-pos
4772 (< cfd-continue-pos (point))
4773 (setq cfd-token-pos (point))))
4774
4775 ;; Have one of the special cases when the region is completely
4776 ;; within a literal or macro. `cfd-continue-pos' is set to a
4777 ;; good start position for the search, so do it.
4778 (c-find-decl-prefix-search)))
d9e94c22 4779
51c9af45 4780 ;; Now loop. Round what? (ACM, 2006/7/5). We already got the first match.
d9e94c22
MS
4781
4782 (while (progn
4783 (while (and
4784 (< cfd-match-pos cfd-limit)
4785
4786 (or
4787 ;; Kludge to filter out matches on the "<" that
4788 ;; aren't open parens, for the sake of languages
4789 ;; that got `c-recognize-<>-arglists' set.
4790 (and (eq (char-before cfd-match-pos) ?<)
4791 (not (c-get-char-property (1- cfd-match-pos)
4792 'syntax-table)))
4793
4794 ;; If `cfd-continue-pos' is less or equal to
4795 ;; `cfd-token-pos', we've got a hit inside a macro
4796 ;; that's in the syntactic whitespace before the last
4797 ;; "real" declaration we've checked. If they're equal
4798 ;; we've arrived at the declaration a second time, so
4799 ;; there's nothing to do.
4800 (= cfd-continue-pos cfd-token-pos)
4801
4802 (progn
4803 ;; If `cfd-continue-pos' is less than `cfd-token-pos'
4804 ;; we're still searching for declarations embedded in
4805 ;; the syntactic whitespace. In that case we need
4806 ;; only to skip comments and not macros, since they
4807 ;; can't be nested, and that's already been done in
4808 ;; `c-find-decl-prefix-search'.
4809 (when (> cfd-continue-pos cfd-token-pos)
4810 (c-forward-syntactic-ws)
4811 (setq cfd-token-pos (point)))
4812
4813 ;; Continue if the following token fails the
4814 ;; CFD-DECL-RE and CFD-FACE-CHECKLIST checks.
4815 (when (or (>= (point) cfd-limit)
4816 (not (looking-at cfd-decl-re))
4817 (and cfd-face-checklist
4818 (not (c-got-face-at
4819 (point) cfd-face-checklist))))
4820 (goto-char cfd-continue-pos)
4821 t)))
4822
4823 (< (point) cfd-limit))
4824 (c-find-decl-prefix-search))
4825
4826 (< (point) cfd-limit))
4827
0386b551
AM
4828 (when (and
4829 (>= (point) cfd-start-pos)
d9e94c22 4830
0386b551
AM
4831 (progn
4832 ;; Narrow to the end of the macro if we got a hit inside
4833 ;; one, to avoid recognizing things that start inside the
4834 ;; macro and end outside it.
4835 (when (> cfd-match-pos cfd-macro-end)
4836 ;; Not in the same macro as in the previous round.
4837 (save-excursion
4838 (goto-char cfd-match-pos)
4839 (setq cfd-macro-end
4840 (if (save-excursion (and (c-beginning-of-macro)
4841 (< (point) cfd-match-pos)))
4842 (progn (c-end-of-macro)
4843 (point))
4844 0))))
4845
4846 (if (zerop cfd-macro-end)
4847 t
4848 (if (> cfd-macro-end (point))
4849 (progn (narrow-to-region (point-min) cfd-macro-end)
4850 t)
4851 ;; The matched token was the last thing in the macro,
4852 ;; so the whole match is bogus.
4853 (setq cfd-macro-end 0)
4854 nil))))
d9e94c22
MS
4855
4856 (c-debug-put-decl-spot-faces cfd-match-pos (point))
0386b551
AM
4857 (if (funcall cfd-fun cfd-match-pos (/= cfd-macro-end 0))
4858 (setq cfd-prop-match nil))
d9e94c22
MS
4859
4860 (when (/= cfd-macro-end 0)
e1dbe924 4861 ;; Restore limits if we did macro narrowing above.
d9e94c22
MS
4862 (narrow-to-region (point-min) cfd-buffer-end)))
4863
4864 (goto-char cfd-continue-pos)
4865 (if (= cfd-continue-pos cfd-limit)
4866 (setq cfd-match-pos cfd-limit)
4867 (c-find-decl-prefix-search)))))
4868
4869\f
4870;; A cache for found types.
4871
4872;; Buffer local variable that contains an obarray with the types we've
4873;; found. If a declaration is recognized somewhere we record the
4874;; fully qualified identifier in it to recognize it as a type
4875;; elsewhere in the file too. This is not accurate since we do not
4876;; bother with the scoping rules of the languages, but in practice the
4877;; same name is seldom used as both a type and something else in a
4878;; file, and we only use this as a last resort in ambiguous cases (see
0386b551
AM
4879;; `c-forward-decl-or-cast-1').
4880;;
580fba94
AM
4881;; Not every type need be in this cache. However, things which have
4882;; ceased to be types must be removed from it.
4883;;
0386b551
AM
4884;; Template types in C++ are added here too but with the template
4885;; arglist replaced with "<>" in references or "<" for the one in the
4886;; primary type. E.g. the type "Foo<A,B>::Bar<C>" is stored as
4887;; "Foo<>::Bar<". This avoids storing very long strings (since C++
4888;; template specs can be fairly sized programs in themselves) and
4889;; improves the hit ratio (it's a type regardless of the template
4890;; args; it's just not the same type, but we're only interested in
4891;; recognizing types, not telling distinct types apart). Note that
4892;; template types in references are added here too; from the example
4893;; above there will also be an entry "Foo<".
d9e94c22
MS
4894(defvar c-found-types nil)
4895(make-variable-buffer-local 'c-found-types)
4896
4897(defsubst c-clear-found-types ()
4898 ;; Clears `c-found-types'.
d9e94c22
MS
4899 (setq c-found-types (make-vector 53 0)))
4900
4901(defun c-add-type (from to)
4902 ;; Add the given region as a type in `c-found-types'. If the region
4903 ;; doesn't match an existing type but there is a type which is equal
4904 ;; to the given one except that the last character is missing, then
4905 ;; the shorter type is removed. That's done to avoid adding all
4906 ;; prefixes of a type as it's being entered and font locked. This
4907 ;; doesn't cover cases like when characters are removed from a type
4908 ;; or added in the middle. We'd need the position of point when the
4909 ;; font locking is invoked to solve this well.
0386b551
AM
4910 ;;
4911 ;; This function might do hidden buffer changes.
4912 (let ((type (c-syntactic-content from to c-recognize-<>-arglists)))
4913 (unless (intern-soft type c-found-types)
4914 (unintern (substring type 0 -1) c-found-types)
4915 (intern type c-found-types))))
d9e94c22 4916
580fba94
AM
4917(defun c-unfind-type (name)
4918 ;; Remove the "NAME" from c-found-types, if present.
4919 (unintern name c-found-types))
4920
d9e94c22
MS
4921(defsubst c-check-type (from to)
4922 ;; Return non-nil if the given region contains a type in
4923 ;; `c-found-types'.
0386b551
AM
4924 ;;
4925 ;; This function might do hidden buffer changes.
4926 (intern-soft (c-syntactic-content from to c-recognize-<>-arglists)
4927 c-found-types))
d9e94c22
MS
4928
4929(defun c-list-found-types ()
4930 ;; Return all the types in `c-found-types' as a sorted list of
4931 ;; strings.
4932 (let (type-list)
4933 (mapatoms (lambda (type)
4934 (setq type-list (cons (symbol-name type)
4935 type-list)))
4936 c-found-types)
4937 (sort type-list 'string-lessp)))
a66cd3ee 4938
2f42c75f
DN
4939;; Shut up the byte compiler.
4940(defvar c-maybe-stale-found-type)
4941
580fba94
AM
4942(defun c-trim-found-types (beg end old-len)
4943 ;; An after change function which, in conjunction with the info in
4944 ;; c-maybe-stale-found-type (set in c-before-change), removes a type
4945 ;; from `c-found-types', should this type have become stale. For
4946 ;; example, this happens to "foo" when "foo \n bar();" becomes
4947 ;; "foo(); \n bar();". Such stale types, if not removed, foul up
4948 ;; the fontification.
b414f371 4949 ;;
580fba94
AM
4950 ;; Have we, perhaps, added non-ws characters to the front/back of a found
4951 ;; type?
4952 (when (> end beg)
4953 (save-excursion
4954 (when (< end (point-max))
4955 (goto-char end)
4956 (if (and (c-beginning-of-current-token) ; only moves when we started in the middle
4957 (progn (goto-char end)
4958 (c-end-of-current-token)))
4959 (c-unfind-type (buffer-substring-no-properties
4960 end (point)))))
4961 (when (> beg (point-min))
4962 (goto-char beg)
4963 (if (and (c-end-of-current-token) ; only moves when we started in the middle
4964 (progn (goto-char beg)
4965 (c-beginning-of-current-token)))
4966 (c-unfind-type (buffer-substring-no-properties
4967 (point) beg))))))
b414f371 4968
580fba94
AM
4969 (if c-maybe-stale-found-type ; e.g. (c-decl-id-start "foo" 97 107 " (* ooka) " "o")
4970 (cond
4971 ;; Changing the amount of (already existing) whitespace - don't do anything.
4972 ((and (c-partial-ws-p beg end)
4973 (or (= beg end) ; removal of WS
580fba94
AM
4974 (string-match "^[ \t\n\r\f\v]*$" (nth 5 c-maybe-stale-found-type)))))
4975
4976 ;; The syntactic relationship which defined a "found type" has been
4977 ;; destroyed.
4978 ((eq (car c-maybe-stale-found-type) 'c-decl-id-start)
4979 (c-unfind-type (cadr c-maybe-stale-found-type)))
4980;; ((eq (car c-maybe-stale-found-type) 'c-decl-type-start) FIXME!!!
4981 )))
4982
d9e94c22 4983\f
dd969a56
AM
4984;; Setting and removing syntax properties on < and > in languages (C++
4985;; and Java) where they can be template/generic delimiters as well as
4986;; their normal meaning of "less/greater than".
4987
4988;; Normally, < and > have syntax 'punctuation'. When they are found to
4989;; be delimiters, they are marked as such with the category properties
4990;; c-<-as-paren-syntax, c->-as-paren-syntax respectively.
4991
4992;; STRATEGY:
4993;;
4994;; It is impossible to determine with certainty whether a <..> pair in
4995;; C++ is two comparison operators or is template delimiters, unless
4996;; one duplicates a lot of a C++ compiler. For example, the following
4997;; code fragment:
4998;;
4999;; foo (a < b, c > d) ;
5000;;
5001;; could be a function call with two integer parameters (each a
5002;; relational expression), or it could be a constructor for class foo
5003;; taking one parameter d of templated type "a < b, c >". They are
5004;; somewhat easier to distinguish in Java.
5005;;
5006;; The strategy now (2010-01) adopted is to mark and unmark < and
5007;; > IN MATCHING PAIRS ONLY. [Previously, they were marked
5008;; individually when their context so indicated. This gave rise to
53964682 5009;; intractable problems when one of a matching pair was deleted, or
dd969a56
AM
5010;; pulled into a literal.]
5011;;
5012;; At each buffer change, the syntax-table properties are removed in a
5013;; before-change function and reapplied, when needed, in an
5014;; after-change function. It is far more important that the
5015;; properties get removed when they they are spurious than that they
5016;; be present when wanted.
5017;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
5018(defun c-clear-<-pair-props (&optional pos)
5019 ;; POS (default point) is at a < character. If it is marked with
5020 ;; open paren syntax-table text property, remove the property,
5021 ;; together with the close paren property on the matching > (if
5022 ;; any).
5023 (save-excursion
5024 (if pos
5025 (goto-char pos)
5026 (setq pos (point)))
5027 (when (equal (c-get-char-property (point) 'syntax-table)
5028 c-<-as-paren-syntax)
5029 (with-syntax-table c-no-parens-syntax-table ; ignore unbalanced [,{,(,..
5030 (c-go-list-forward))
5031 (when (equal (c-get-char-property (1- (point)) 'syntax-table)
5032 c->-as-paren-syntax) ; should always be true.
8a249abc
AM
5033 (c-clear-char-property (1- (point)) 'category))
5034 (c-clear-char-property pos 'category))))
dd969a56
AM
5035
5036(defun c-clear->-pair-props (&optional pos)
5037 ;; POS (default point) is at a > character. If it is marked with
5038 ;; close paren syntax-table property, remove the property, together
5039 ;; with the open paren property on the matching < (if any).
5040 (save-excursion
5041 (if pos
5042 (goto-char pos)
5043 (setq pos (point)))
5044 (when (equal (c-get-char-property (point) 'syntax-table)
5045 c->-as-paren-syntax)
5046 (with-syntax-table c-no-parens-syntax-table ; ignore unbalanced [,{,(,..
5047 (c-go-up-list-backward))
5048 (when (equal (c-get-char-property (point) 'syntax-table)
5049 c-<-as-paren-syntax) ; should always be true.
8a249abc
AM
5050 (c-clear-char-property (point) 'category))
5051 (c-clear-char-property pos 'category))))
dd969a56
AM
5052
5053(defun c-clear-<>-pair-props (&optional pos)
5054 ;; POS (default point) is at a < or > character. If it has an
5055 ;; open/close paren syntax-table property, remove this property both
5056 ;; from the current character and its partner (which will also be
5057 ;; thusly marked).
5058 (cond
5059 ((eq (char-after) ?\<)
5060 (c-clear-<-pair-props pos))
5061 ((eq (char-after) ?\>)
5062 (c-clear->-pair-props pos))
5063 (t (c-benign-error
5064 "c-clear-<>-pair-props called from wrong position"))))
5065
5066(defun c-clear-<-pair-props-if-match-after (lim &optional pos)
5067 ;; POS (default point) is at a < character. If it is both marked
5068 ;; with open/close paren syntax-table property, and has a matching >
5069 ;; (also marked) which is after LIM, remove the property both from
43a91810
AM
5070 ;; the current > and its partner. Return t when this happens, nil
5071 ;; when it doesn't.
dd969a56
AM
5072 (save-excursion
5073 (if pos
5074 (goto-char pos)
5075 (setq pos (point)))
5076 (when (equal (c-get-char-property (point) 'syntax-table)
5077 c-<-as-paren-syntax)
5078 (with-syntax-table c-no-parens-syntax-table ; ignore unbalanced [,{,(,..
5079 (c-go-list-forward))
5080 (when (and (>= (point) lim)
5081 (equal (c-get-char-property (1- (point)) 'syntax-table)
5082 c->-as-paren-syntax)) ; should always be true.
5083 (c-unmark-<->-as-paren (1- (point)))
43a91810
AM
5084 (c-unmark-<->-as-paren pos))
5085 t)))
dd969a56
AM
5086
5087(defun c-clear->-pair-props-if-match-before (lim &optional pos)
5088 ;; POS (default point) is at a > character. If it is both marked
5089 ;; with open/close paren syntax-table property, and has a matching <
5090 ;; (also marked) which is before LIM, remove the property both from
43a91810
AM
5091 ;; the current < and its partner. Return t when this happens, nil
5092 ;; when it doesn't.
dd969a56
AM
5093 (save-excursion
5094 (if pos
5095 (goto-char pos)
5096 (setq pos (point)))
5097 (when (equal (c-get-char-property (point) 'syntax-table)
5098 c->-as-paren-syntax)
5099 (with-syntax-table c-no-parens-syntax-table ; ignore unbalanced [,{,(,..
5100 (c-go-up-list-backward))
5101 (when (and (<= (point) lim)
5102 (equal (c-get-char-property (point) 'syntax-table)
5103 c-<-as-paren-syntax)) ; should always be true.
5104 (c-unmark-<->-as-paren (point))
43a91810
AM
5105 (c-unmark-<->-as-paren pos))
5106 t)))
dd969a56 5107
f3b554af
GM
5108;; Set by c-common-init in cc-mode.el.
5109(defvar c-new-BEG)
5110(defvar c-new-END)
5111
dd969a56
AM
5112(defun c-before-change-check-<>-operators (beg end)
5113 ;; Unmark certain pairs of "< .... >" which are currently marked as
5114 ;; template/generic delimiters. (This marking is via syntax-table
5115 ;; text properties).
5116 ;;
5117 ;; These pairs are those which are in the current "statement" (i.e.,
5118 ;; the region between the {, }, or ; before BEG and the one after
5119 ;; END), and which enclose any part of the interval (BEG END).
5120 ;;
5121 ;; Note that in C++ (?and Java), template/generic parens cannot
5122 ;; enclose a brace or semicolon, so we use these as bounds on the
5123 ;; region we must work on.
5124 ;;
5125 ;; This function is called from before-change-functions (via
5126 ;; c-get-state-before-change-functions). Thus the buffer is widened,
5127 ;; and point is undefined, both at entry and exit.
5128 ;;
5129 ;; FIXME!!! This routine ignores the possibility of macros entirely.
5130 ;; 2010-01-29.
5131 (save-excursion
5132 (let ((beg-lit-limits (progn (goto-char beg) (c-literal-limits)))
43a91810
AM
5133 (end-lit-limits (progn (goto-char end) (c-literal-limits)))
5134 new-beg new-end need-new-beg need-new-end)
dd969a56
AM
5135 ;; Locate the barrier before the changed region
5136 (goto-char (if beg-lit-limits (car beg-lit-limits) beg))
5137 (c-syntactic-skip-backward "^;{}" (max (- beg 2048) (point-min)))
43a91810 5138 (setq new-beg (point))
dd969a56
AM
5139
5140 ;; Remove the syntax-table properties from each pertinent <...> pair.
5141 ;; Firsly, the ones with the < before beg and > after beg.
5142 (while (c-search-forward-char-property 'category 'c-<-as-paren-syntax beg)
43a91810
AM
5143 (if (c-clear-<-pair-props-if-match-after beg (1- (point)))
5144 (setq need-new-beg t)))
dd969a56
AM
5145
5146 ;; Locate the barrier after END.
5147 (goto-char (if end-lit-limits (cdr end-lit-limits) end))
5148 (c-syntactic-re-search-forward "[;{}]"
5149 (min (+ end 2048) (point-max)) 'end)
43a91810 5150 (setq new-end (point))
dd969a56
AM
5151
5152 ;; Remove syntax-table properties from the remaining pertinent <...>
5153 ;; pairs, those with a > after end and < before end.
5154 (while (c-search-backward-char-property 'category 'c->-as-paren-syntax end)
43a91810
AM
5155 (if (c-clear->-pair-props-if-match-before end)
5156 (setq need-new-end t)))
5157
5158 ;; Extend the fontification region, if needed.
5159 (when need-new-beg
5160 (goto-char new-beg)
5161 (c-forward-syntactic-ws)
5162 (and (< (point) c-new-BEG) (setq c-new-BEG (point))))
5163
5164 (when need-new-end
5165 (and (> new-end c-new-END) (setq c-new-END new-end))))))
dd969a56
AM
5166
5167
d9e94c22 5168
0386b551
AM
5169(defun c-after-change-check-<>-operators (beg end)
5170 ;; This is called from `after-change-functions' when
5171 ;; c-recognize-<>-arglists' is set. It ensures that no "<" or ">"
5172 ;; chars with paren syntax become part of another operator like "<<"
5173 ;; or ">=".
5174 ;;
5175 ;; This function might do hidden buffer changes.
5176
cb694ab7
AM
5177 (save-excursion
5178 (goto-char beg)
5179 (when (or (looking-at "[<>]")
5180 (< (skip-chars-backward "<>") 0))
5181
0386b551 5182 (goto-char beg)
cb694ab7
AM
5183 (c-beginning-of-current-token)
5184 (when (and (< (point) beg)
5185 (looking-at c-<>-multichar-token-regexp)
5186 (< beg (setq beg (match-end 0))))
5187 (while (progn (skip-chars-forward "^<>" beg)
5188 (< (point) beg))
dd969a56 5189 (c-clear-<>-pair-props)
cb694ab7
AM
5190 (forward-char))))
5191
5192 (when (< beg end)
5193 (goto-char end)
0386b551
AM
5194 (when (or (looking-at "[<>]")
5195 (< (skip-chars-backward "<>") 0))
5196
cb694ab7 5197 (goto-char end)
0386b551 5198 (c-beginning-of-current-token)
cb694ab7 5199 (when (and (< (point) end)
0386b551 5200 (looking-at c-<>-multichar-token-regexp)
cb694ab7
AM
5201 (< end (setq end (match-end 0))))
5202 (while (progn (skip-chars-forward "^<>" end)
5203 (< (point) end))
dd969a56 5204 (c-clear-<>-pair-props)
cb694ab7 5205 (forward-char)))))))
d9e94c22 5206
dd969a56
AM
5207
5208\f
5209;; Handling of small scale constructs like types and names.
5210
d9e94c22
MS
5211;; Dynamically bound variable that instructs `c-forward-type' to also
5212;; treat possible types (i.e. those that it normally returns 'maybe or
5213;; 'found for) as actual types (and always return 'found for them).
5214;; This means that it records them in `c-record-type-identifiers' if
5215;; that is set, and that it adds them to `c-found-types'.
5216(defvar c-promote-possible-types nil)
5217
0386b551
AM
5218;; Dynamically bound variable that instructs `c-forward-<>-arglist' to
5219;; mark up successfully parsed arglists with paren syntax properties on
5220;; the surrounding angle brackets and with `c-<>-arg-sep' in the
5221;; `c-type' property of each argument separating comma.
5222;;
5223;; Setting this variable also makes `c-forward-<>-arglist' recurse into
5224;; all arglists for side effects (i.e. recording types), otherwise it
5225;; exploits any existing paren syntax properties to quickly jump to the
5226;; end of already parsed arglists.
5227;;
5228;; Marking up the arglists is not the default since doing that correctly
5229;; depends on a proper value for `c-restricted-<>-arglists'.
5230(defvar c-parse-and-markup-<>-arglists nil)
5231
d9e94c22 5232;; Dynamically bound variable that instructs `c-forward-<>-arglist' to
037558bf
MS
5233;; not accept arglists that contain binary operators.
5234;;
5235;; This is primarily used to handle C++ template arglists. C++
5236;; disambiguates them by checking whether the preceding name is a
5237;; template or not. We can't do that, so we assume it is a template
5238;; if it can be parsed as one. That usually works well since
5239;; comparison expressions on the forms "a < b > c" or "a < b, c > d"
5240;; in almost all cases would be pointless.
5241;;
5242;; However, in function arglists, e.g. in "foo (a < b, c > d)", we
5243;; should let the comma separate the function arguments instead. And
5244;; in a context where the value of the expression is taken, e.g. in
5245;; "if (a < b || c > d)", it's probably not a template.
5246(defvar c-restricted-<>-arglists nil)
d9e94c22 5247
0386b551
AM
5248;; Dynamically bound variables that instructs
5249;; `c-forward-keyword-clause', `c-forward-<>-arglist',
5250;; `c-forward-name', `c-forward-type', `c-forward-decl-or-cast-1', and
5251;; `c-forward-label' to record the ranges of all the type and
5252;; reference identifiers they encounter. They will build lists on
5253;; these variables where each element is a cons of the buffer
5254;; positions surrounding each identifier. This recording is only
5255;; activated when `c-record-type-identifiers' is non-nil.
d9e94c22
MS
5256;;
5257;; All known types that can't be identifiers are recorded, and also
5258;; other possible types if `c-promote-possible-types' is set.
5259;; Recording is however disabled inside angle bracket arglists that
5260;; are encountered inside names and other angle bracket arglists.
0386b551 5261;; Such occurrences are taken care of by `c-font-lock-<>-arglists'
d9e94c22
MS
5262;; instead.
5263;;
5264;; Only the names in C++ template style references (e.g. "tmpl" in
5265;; "tmpl<a,b>::foo") are recorded as references, other references
5266;; aren't handled here.
0386b551
AM
5267;;
5268;; `c-forward-label' records the label identifier(s) on
5269;; `c-record-ref-identifiers'.
d9e94c22
MS
5270(defvar c-record-type-identifiers nil)
5271(defvar c-record-ref-identifiers nil)
5272
0386b551
AM
5273;; This variable will receive a cons cell of the range of the last
5274;; single identifier symbol stepped over by `c-forward-name' if it's
5275;; successful. This is the range that should be put on one of the
5276;; record lists above by the caller. It's assigned nil if there's no
5277;; such symbol in the name.
d9e94c22
MS
5278(defvar c-last-identifier-range nil)
5279
5280(defmacro c-record-type-id (range)
5281 (if (eq (car-safe range) 'cons)
5282 ;; Always true.
5283 `(setq c-record-type-identifiers
5284 (cons ,range c-record-type-identifiers))
5285 `(let ((range ,range))
5286 (if range
5287 (setq c-record-type-identifiers
5288 (cons range c-record-type-identifiers))))))
5289
5290(defmacro c-record-ref-id (range)
5291 (if (eq (car-safe range) 'cons)
5292 ;; Always true.
5293 `(setq c-record-ref-identifiers
5294 (cons ,range c-record-ref-identifiers))
5295 `(let ((range ,range))
5296 (if range
5297 (setq c-record-ref-identifiers
5298 (cons range c-record-ref-identifiers))))))
5299
5300;; Dynamically bound variable that instructs `c-forward-type' to
5301;; record the ranges of types that only are found. Behaves otherwise
5302;; like `c-record-type-identifiers'.
5303(defvar c-record-found-types nil)
5304
5305(defmacro c-forward-keyword-prefixed-id (type)
5306 ;; Used internally in `c-forward-keyword-clause' to move forward
5307 ;; over a type (if TYPE is 'type) or a name (otherwise) which
5308 ;; possibly is prefixed by keywords and their associated clauses.
5309 ;; Try with a type/name first to not trip up on those that begin
5310 ;; with a keyword. Return t if a known or found type is moved
5311 ;; over. The point is clobbered if nil is returned. If range
5312 ;; recording is enabled, the identifier is recorded on as a type
5313 ;; if TYPE is 'type or as a reference if TYPE is 'ref.
0386b551
AM
5314 ;;
5315 ;; This macro might do hidden buffer changes.
d9e94c22
MS
5316 `(let (res)
5317 (while (if (setq res ,(if (eq type 'type)
5318 `(c-forward-type)
5319 `(c-forward-name)))
5320 nil
5321 (and (looking-at c-keywords-regexp)
0386b551 5322 (c-forward-keyword-clause 1))))
d9e94c22
MS
5323 (when (memq res '(t known found prefix))
5324 ,(when (eq type 'ref)
5325 `(when c-record-type-identifiers
5326 (c-record-ref-id c-last-identifier-range)))
5327 t)))
5328
0386b551 5329(defmacro c-forward-id-comma-list (type update-safe-pos)
d9e94c22
MS
5330 ;; Used internally in `c-forward-keyword-clause' to move forward
5331 ;; over a comma separated list of types or names using
5332 ;; `c-forward-keyword-prefixed-id'.
0386b551
AM
5333 ;;
5334 ;; This macro might do hidden buffer changes.
d9e94c22 5335 `(while (and (progn
0386b551
AM
5336 ,(when update-safe-pos
5337 `(setq safe-pos (point)))
d9e94c22
MS
5338 (eq (char-after) ?,))
5339 (progn
5340 (forward-char)
5341 (c-forward-syntactic-ws)
5342 (c-forward-keyword-prefixed-id ,type)))))
5343
0386b551
AM
5344(defun c-forward-keyword-clause (match)
5345 ;; Submatch MATCH in the current match data is assumed to surround a
5346 ;; token. If it's a keyword, move over it and any immediately
5347 ;; following clauses associated with it, stopping at the start of
5348 ;; the next token. t is returned in that case, otherwise the point
d9e94c22
MS
5349 ;; stays and nil is returned. The kind of clauses that are
5350 ;; recognized are those specified by `c-type-list-kwds',
5351 ;; `c-ref-list-kwds', `c-colon-type-list-kwds',
5352 ;; `c-paren-nontype-kwds', `c-paren-type-kwds', `c-<>-type-kwds',
5353 ;; and `c-<>-arglist-kwds'.
0386b551
AM
5354 ;;
5355 ;; This function records identifier ranges on
5356 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
5357 ;; `c-record-type-identifiers' is non-nil.
5358 ;;
5359 ;; Note that for `c-colon-type-list-kwds', which doesn't necessary
5360 ;; apply directly after the keyword, the type list is moved over
5361 ;; only when there is no unaccounted token before it (i.e. a token
5362 ;; that isn't moved over due to some other keyword list). The
5363 ;; identifier ranges in the list are still recorded if that should
5364 ;; be done, though.
5365 ;;
5366 ;; This function might do hidden buffer changes.
5367
5368 (let ((kwd-sym (c-keyword-sym (match-string match))) safe-pos pos
5369 ;; The call to `c-forward-<>-arglist' below is made after
5370 ;; `c-<>-sexp-kwds' keywords, so we're certain they actually
5371 ;; are angle bracket arglists and `c-restricted-<>-arglists'
5372 ;; should therefore be nil.
5373 (c-parse-and-markup-<>-arglists t)
5374 c-restricted-<>-arglists)
d9e94c22 5375
d9e94c22 5376 (when kwd-sym
0386b551 5377 (goto-char (match-end match))
d9e94c22
MS
5378 (c-forward-syntactic-ws)
5379 (setq safe-pos (point))
a66cd3ee 5380
d9e94c22
MS
5381 (cond
5382 ((and (c-keyword-member kwd-sym 'c-type-list-kwds)
5383 (c-forward-keyword-prefixed-id type))
5384 ;; There's a type directly after a keyword in `c-type-list-kwds'.
0386b551 5385 (c-forward-id-comma-list type t))
d9e94c22
MS
5386
5387 ((and (c-keyword-member kwd-sym 'c-ref-list-kwds)
5388 (c-forward-keyword-prefixed-id ref))
5389 ;; There's a name directly after a keyword in `c-ref-list-kwds'.
0386b551 5390 (c-forward-id-comma-list ref t))
d9e94c22
MS
5391
5392 ((and (c-keyword-member kwd-sym 'c-paren-any-kwds)
5393 (eq (char-after) ?\())
5394 ;; There's an open paren after a keyword in `c-paren-any-kwds'.
5395
5396 (forward-char)
5397 (when (and (setq pos (c-up-list-forward))
5398 (eq (char-before pos) ?\)))
5399 (when (and c-record-type-identifiers
5400 (c-keyword-member kwd-sym 'c-paren-type-kwds))
5401 ;; Use `c-forward-type' on every identifier we can find
5402 ;; inside the paren, to record the types.
5403 (while (c-syntactic-re-search-forward c-symbol-start pos t)
5404 (goto-char (match-beginning 0))
5405 (unless (c-forward-type)
5406 (looking-at c-symbol-key) ; Always matches.
5407 (goto-char (match-end 0)))))
5408
5409 (goto-char pos)
5410 (c-forward-syntactic-ws)
5411 (setq safe-pos (point))))
5412
5413 ((and (c-keyword-member kwd-sym 'c-<>-sexp-kwds)
5414 (eq (char-after) ?<)
0386b551 5415 (c-forward-<>-arglist (c-keyword-member kwd-sym 'c-<>-type-kwds)))
d9e94c22
MS
5416 (c-forward-syntactic-ws)
5417 (setq safe-pos (point)))
5418
5419 ((and (c-keyword-member kwd-sym 'c-nonsymbol-sexp-kwds)
449a2b0d
MS
5420 (not (looking-at c-symbol-start))
5421 (c-safe (c-forward-sexp) t))
d9e94c22
MS
5422 (c-forward-syntactic-ws)
5423 (setq safe-pos (point))))
5424
0386b551
AM
5425 (when (c-keyword-member kwd-sym 'c-colon-type-list-kwds)
5426 (if (eq (char-after) ?:)
5427 ;; If we are at the colon already, we move over the type
5428 ;; list after it.
5429 (progn
5430 (forward-char)
5431 (c-forward-syntactic-ws)
5432 (when (c-forward-keyword-prefixed-id type)
5433 (c-forward-id-comma-list type t)))
5434 ;; Not at the colon, so stop here. But the identifier
5435 ;; ranges in the type list later on should still be
5436 ;; recorded.
5437 (and c-record-type-identifiers
5438 (progn
5439 ;; If a keyword matched both one of the types above and
5440 ;; this one, we match `c-colon-type-list-re' after the
5441 ;; clause matched above.
5442 (goto-char safe-pos)
5443 (looking-at c-colon-type-list-re))
5444 (progn
5445 (goto-char (match-end 0))
5446 (c-forward-syntactic-ws)
5447 (c-forward-keyword-prefixed-id type))
5448 ;; There's a type after the `c-colon-type-list-re' match
5449 ;; after a keyword in `c-colon-type-list-kwds'.
5450 (c-forward-id-comma-list type nil))))
d9e94c22
MS
5451
5452 (goto-char safe-pos)
5453 t)))
5454
f3b554af
GM
5455;; cc-mode requires cc-fonts.
5456(declare-function c-fontify-recorded-types-and-refs "cc-fonts" ())
5457
0386b551
AM
5458(defun c-forward-<>-arglist (all-types)
5459 ;; The point is assumed to be at a "<". Try to treat it as the open
b4dc7d98 5460 ;; paren of an angle bracket arglist and move forward to the
0386b551
AM
5461 ;; corresponding ">". If successful, the point is left after the
5462 ;; ">" and t is returned, otherwise the point isn't moved and nil is
d9e94c22
MS
5463 ;; returned. If ALL-TYPES is t then all encountered arguments in
5464 ;; the arglist that might be types are treated as found types.
5465 ;;
0386b551
AM
5466 ;; The variable `c-parse-and-markup-<>-arglists' controls how this
5467 ;; function handles text properties on the angle brackets and argument
5468 ;; separating commas.
d9e94c22 5469 ;;
0386b551
AM
5470 ;; `c-restricted-<>-arglists' controls how lenient the template
5471 ;; arglist recognition should be.
5472 ;;
5473 ;; This function records identifier ranges on
5474 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
5475 ;; `c-record-type-identifiers' is non-nil.
5476 ;;
5477 ;; This function might do hidden buffer changes.
d9e94c22
MS
5478
5479 (let ((start (point))
5480 ;; If `c-record-type-identifiers' is set then activate
5481 ;; recording of any found types that constitute an argument in
5482 ;; the arglist.
abfc152b 5483 (c-record-found-types (if c-record-type-identifiers t)))
d9e94c22
MS
5484 (if (catch 'angle-bracket-arglist-escape
5485 (setq c-record-found-types
0386b551 5486 (c-forward-<>-arglist-recur all-types)))
d9e94c22
MS
5487 (progn
5488 (when (consp c-record-found-types)
5489 (setq c-record-type-identifiers
5490 ;; `nconc' doesn't mind that the tail of
5491 ;; `c-record-found-types' is t.
5492 (nconc c-record-found-types c-record-type-identifiers)))
452ea855 5493 (if (c-major-mode-is 'java-mode) (c-fontify-recorded-types-and-refs))
d9e94c22
MS
5494 t)
5495
5496 (goto-char start)
a66cd3ee 5497 nil)))
785eecbb 5498
0386b551 5499(defun c-forward-<>-arglist-recur (all-types)
d9e94c22 5500 ;; Recursive part of `c-forward-<>-arglist'.
0386b551
AM
5501 ;;
5502 ;; This function might do hidden buffer changes.
d9e94c22
MS
5503
5504 (let ((start (point)) res pos tmp
5505 ;; Cover this so that any recorded found type ranges are
5506 ;; automatically lost if it turns out to not be an angle
5507 ;; bracket arglist. It's propagated through the return value
5508 ;; on successful completion.
5509 (c-record-found-types c-record-found-types)
5510 ;; List that collects the positions after the argument
5511 ;; separating ',' in the arglist.
5512 arg-start-pos)
0386b551
AM
5513 ;; If the '<' has paren open syntax then we've marked it as an angle
5514 ;; bracket arglist before, so skip to the end.
5515 (if (and (not c-parse-and-markup-<>-arglists)
5516 (c-get-char-property (point) 'syntax-table))
5517
5518 (progn
5519 (forward-char)
5520 (if (and (c-go-up-list-forward)
5521 (eq (char-before) ?>))
5522 t
0386b551
AM
5523 ;; Got unmatched paren angle brackets. We don't clear the paren
5524 ;; syntax properties and retry, on the basis that it's very
5525 ;; unlikely that paren angle brackets become operators by code
5526 ;; manipulation. It's far more likely that it doesn't match due
5527 ;; to narrowing or some temporary change.
5528 (goto-char start)
5529 nil))
d9e94c22 5530
abfc152b 5531 (forward-char) ; Forward over the opening '<'.
452ea855 5532
d9e94c22 5533 (unless (looking-at c-<-op-cont-regexp)
abfc152b
AM
5534 ;; go forward one non-alphanumeric character (group) per iteration of
5535 ;; this loop.
22c3ce97 5536 (while (and
d9e94c22 5537 (progn
22c3ce97
AM
5538 (c-forward-syntactic-ws)
5539 (let ((orig-record-found-types c-record-found-types))
5540 (when (or (and c-record-type-identifiers all-types)
5541 (c-major-mode-is 'java-mode))
5542 ;; All encountered identifiers are types, so set the
5543 ;; promote flag and parse the type.
5544 (progn
5545 (c-forward-syntactic-ws)
5546 (if (looking-at "\\?")
5547 (forward-char)
5548 (when (looking-at c-identifier-start)
5549 (let ((c-promote-possible-types t)
5550 (c-record-found-types t))
5551 (c-forward-type))))
5552
5553 (c-forward-syntactic-ws)
5554
5555 (when (or (looking-at "extends")
5556 (looking-at "super"))
5557 (forward-word)
5558 (c-forward-syntactic-ws)
5559 (let ((c-promote-possible-types t)
5560 (c-record-found-types t))
5561 (c-forward-type)
5562 (c-forward-syntactic-ws))))))
5563
abfc152b 5564 (setq pos (point)) ; e.g. first token inside the '<'
22c3ce97 5565
a4ee83cc
AM
5566 ;; Note: These regexps exploit the match order in \| so
5567 ;; that "<>" is matched by "<" rather than "[^>:-]>".
5568 (c-syntactic-re-search-forward
5569 ;; Stop on ',', '|', '&', '+' and '-' to catch
5570 ;; common binary operators that could be between
5571 ;; two comparison expressions "a<b" and "c>d".
5572 "[<;{},|+&-]\\|[>)]"
5573 nil t t))
22c3ce97
AM
5574
5575 (cond
5576 ((eq (char-before) ?>)
d9e94c22
MS
5577 ;; Either an operator starting with '>' or the end of
5578 ;; the angle bracket arglist.
5579
0386b551 5580 (if (looking-at c->-op-cont-regexp)
d9e94c22 5581 (progn
0386b551
AM
5582 (goto-char (match-end 0))
5583 t) ; Continue the loop.
d9e94c22 5584
0386b551
AM
5585 ;; The angle bracket arglist is finished.
5586 (when c-parse-and-markup-<>-arglists
d9e94c22 5587 (while arg-start-pos
0386b551
AM
5588 (c-put-c-type-property (1- (car arg-start-pos))
5589 'c-<>-arg-sep)
d9e94c22
MS
5590 (setq arg-start-pos (cdr arg-start-pos)))
5591 (c-mark-<-as-paren start)
0386b551
AM
5592 (c-mark->-as-paren (1- (point))))
5593 (setq res t)
5594 nil)) ; Exit the loop.
d9e94c22
MS
5595
5596 ((eq (char-before) ?<)
5597 ;; Either an operator starting with '<' or a nested arglist.
d9e94c22
MS
5598 (setq pos (point))
5599 (let (id-start id-end subres keyword-match)
abfc152b
AM
5600 (cond
5601 ;; The '<' begins a multi-char operator.
5602 ((looking-at c-<-op-cont-regexp)
5603 (setq tmp (match-end 0))
5604 (goto-char (match-end 0)))
5605 ;; We're at a nested <.....>
5606 ((progn
5607 (setq tmp pos)
5608 (backward-char) ; to the '<'
5609 (and
5610 (save-excursion
5611 ;; There's always an identifier before an angle
5612 ;; bracket arglist, or a keyword in `c-<>-type-kwds'
5613 ;; or `c-<>-arglist-kwds'.
5614 (c-backward-syntactic-ws)
5615 (setq id-end (point))
5616 (c-simple-skip-symbol-backward)
5617 (when (or (setq keyword-match
5618 (looking-at c-opt-<>-sexp-key))
5619 (not (looking-at c-keywords-regexp)))
5620 (setq id-start (point))))
5621 (setq subres
5622 (let ((c-promote-possible-types t)
5623 (c-record-found-types t))
5624 (c-forward-<>-arglist-recur
5625 (and keyword-match
5626 (c-keyword-member
5627 (c-keyword-sym (match-string 1))
5628 'c-<>-type-kwds)))))))
d9e94c22
MS
5629
5630 ;; It was an angle bracket arglist.
5631 (setq c-record-found-types subres)
5632
5633 ;; Record the identifier before the template as a type
5634 ;; or reference depending on whether the arglist is last
5635 ;; in a qualified identifier.
5636 (when (and c-record-type-identifiers
5637 (not keyword-match))
5638 (if (and c-opt-identifier-concat-key
5639 (progn
5640 (c-forward-syntactic-ws)
5641 (looking-at c-opt-identifier-concat-key)))
5642 (c-record-ref-id (cons id-start id-end))
abfc152b
AM
5643 (c-record-type-id (cons id-start id-end)))))
5644
5645 ;; At a "less than" operator.
5646 (t
5647 (forward-char)
5648 )))
5649 t) ; carry on looping.
22c3ce97
AM
5650
5651 ((and (not c-restricted-<>-arglists)
5652 (or (and (eq (char-before) ?&)
5653 (not (eq (char-after) ?&)))
5654 (eq (char-before) ?,)))
5655 ;; Just another argument. Record the position. The
5656 ;; type check stuff that made us stop at it is at
5657 ;; the top of the loop.
d9e94c22
MS
5658 (setq arg-start-pos (cons (point) arg-start-pos)))
5659
5660 (t
5661 ;; Got a character that can't be in an angle bracket
5662 ;; arglist argument. Abort using `throw', since
5663 ;; it's useless to try to find a surrounding arglist
5664 ;; if we're nested.
5665 (throw 'angle-bracket-arglist-escape nil))))))
d9e94c22
MS
5666 (if res
5667 (or c-record-found-types t)))))
5668
0386b551
AM
5669(defun c-backward-<>-arglist (all-types &optional limit)
5670 ;; The point is assumed to be directly after a ">". Try to treat it
5671 ;; as the close paren of an angle bracket arglist and move back to
5672 ;; the corresponding "<". If successful, the point is left at
5673 ;; the "<" and t is returned, otherwise the point isn't moved and
5674 ;; nil is returned. ALL-TYPES is passed on to
5675 ;; `c-forward-<>-arglist'.
5676 ;;
5677 ;; If the optional LIMIT is given, it bounds the backward search.
5678 ;; It's then assumed to be at a syntactically relevant position.
5679 ;;
5680 ;; This is a wrapper around `c-forward-<>-arglist'. See that
5681 ;; function for more details.
5682
5683 (let ((start (point)))
5684 (backward-char)
5685 (if (and (not c-parse-and-markup-<>-arglists)
5686 (c-get-char-property (point) 'syntax-table))
5687
5688 (if (and (c-go-up-list-backward)
5689 (eq (char-after) ?<))
5690 t
5691 ;; See corresponding note in `c-forward-<>-arglist'.
5692 (goto-char start)
5693 nil)
5694
51c9af45 5695 (while (progn
0386b551
AM
5696 (c-syntactic-skip-backward "^<;{}" limit t)
5697
51c9af45
AM
5698 (and
5699 (if (eq (char-before) ?<)
5700 t
5701 ;; Stopped at bob or a char that isn't allowed in an
5702 ;; arglist, so we've failed.
5703 (goto-char start)
5704 nil)
0386b551 5705
51c9af45
AM
5706 (if (> (point)
5707 (progn (c-beginning-of-current-token)
5708 (point)))
5709 ;; If we moved then the "<" was part of some
5710 ;; multicharacter token.
5711 t
0386b551 5712
51c9af45
AM
5713 (backward-char)
5714 (let ((beg-pos (point)))
5715 (if (c-forward-<>-arglist all-types)
5716 (cond ((= (point) start)
5717 ;; Matched the arglist. Break the while.
5718 (goto-char beg-pos)
5719 nil)
5720 ((> (point) start)
5721 ;; We started from a non-paren ">" inside an
5722 ;; arglist.
5723 (goto-char start)
5724 nil)
5725 (t
5726 ;; Matched a shorter arglist. Can be a nested
5727 ;; one so continue looking.
5728 (goto-char beg-pos)
5729 t))
5730 t))))))
0386b551
AM
5731
5732 (/= (point) start))))
5733
d9e94c22
MS
5734(defun c-forward-name ()
5735 ;; Move forward over a complete name if at the beginning of one,
e15f8aaa
AM
5736 ;; stopping at the next following token. A keyword, as such,
5737 ;; doesn't count as a name. If the point is not at something that
5738 ;; is recognized as a name then it stays put.
5739 ;;
5740 ;; A name could be something as simple as "foo" in C or something as
d9e94c22
MS
5741 ;; complex as "X<Y<class A<int>::B, BIT_MAX >> b>, ::operator<> ::
5742 ;; Z<(a>b)> :: operator const X<&foo>::T Q::G<unsigned short
5743 ;; int>::*volatile const" in C++ (this function is actually little
5744 ;; more than a `looking-at' call in all modes except those that,
e15f8aaa
AM
5745 ;; like C++, have `c-recognize-<>-arglists' set).
5746 ;;
5747 ;; Return
5748 ;; o - nil if no name is found;
5749 ;; o - 'template if it's an identifier ending with an angle bracket
5750 ;; arglist;
5751 ;; o - 'operator of it's an operator identifier;
5752 ;; o - t if it's some other kind of name.
0386b551
AM
5753 ;;
5754 ;; This function records identifier ranges on
5755 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
5756 ;; `c-record-type-identifiers' is non-nil.
5757 ;;
5758 ;; This function might do hidden buffer changes.
d9e94c22 5759
0386b551 5760 (let ((pos (point)) (start (point)) res id-start id-end
d9e94c22
MS
5761 ;; Turn off `c-promote-possible-types' here since we might
5762 ;; call `c-forward-<>-arglist' and we don't want it to promote
5763 ;; every suspect thing in the arglist to a type. We're
5764 ;; typically called from `c-forward-type' in this case, and
5765 ;; the caller only wants the top level type that it finds to
5766 ;; be promoted.
5767 c-promote-possible-types)
5768 (while
5769 (and
5770 (looking-at c-identifier-key)
5771
5772 (progn
5773 ;; Check for keyword. We go to the last symbol in
5774 ;; `c-identifier-key' first.
0386b551
AM
5775 (goto-char (setq id-end (match-end 0)))
5776 (c-simple-skip-symbol-backward)
5777 (setq id-start (point))
d9e94c22
MS
5778
5779 (if (looking-at c-keywords-regexp)
5780 (when (and (c-major-mode-is 'c++-mode)
5781 (looking-at
5782 (cc-eval-when-compile
5783 (concat "\\(operator\\|\\(template\\)\\)"
5784 "\\(" (c-lang-const c-nonsymbol-key c++)
5785 "\\|$\\)")))
5786 (if (match-beginning 2)
5787 ;; "template" is only valid inside an
5788 ;; identifier if preceded by "::".
5789 (save-excursion
5790 (c-backward-syntactic-ws)
5791 (and (c-safe (backward-char 2) t)
5792 (looking-at "::")))
5793 t))
5794
5795 ;; Handle a C++ operator or template identifier.
5796 (goto-char id-end)
5797 (c-forward-syntactic-ws)
5798 (cond ((eq (char-before id-end) ?e)
5799 ;; Got "... ::template".
5800 (let ((subres (c-forward-name)))
5801 (when subres
5802 (setq pos (point)
5803 res subres))))
5804
5805 ((looking-at c-identifier-start)
5806 ;; Got a cast operator.
5807 (when (c-forward-type)
5808 (setq pos (point)
5809 res 'operator)
5810 ;; Now we should match a sequence of either
5811 ;; '*', '&' or a name followed by ":: *",
5812 ;; where each can be followed by a sequence
5813 ;; of `c-opt-type-modifier-key'.
5814 (while (cond ((looking-at "[*&]")
5815 (goto-char (match-end 0))
5816 t)
5817 ((looking-at c-identifier-start)
5818 (and (c-forward-name)
5819 (looking-at "::")
5820 (progn
5821 (goto-char (match-end 0))
5822 (c-forward-syntactic-ws)
5823 (eq (char-after) ?*))
5824 (progn
5825 (forward-char)
5826 t))))
5827 (while (progn
5828 (c-forward-syntactic-ws)
5829 (setq pos (point))
5830 (looking-at c-opt-type-modifier-key))
5831 (goto-char (match-end 1))))))
5832
5833 ((looking-at c-overloadable-operators-regexp)
5834 ;; Got some other operator.
0386b551
AM
5835 (setq c-last-identifier-range
5836 (cons (point) (match-end 0)))
d9e94c22
MS
5837 (goto-char (match-end 0))
5838 (c-forward-syntactic-ws)
5839 (setq pos (point)
5840 res 'operator)))
5841
5842 nil)
5843
0386b551
AM
5844 ;; `id-start' is equal to `id-end' if we've jumped over
5845 ;; an identifier that doesn't end with a symbol token.
5846 ;; That can occur e.g. for Java import directives on the
5847 ;; form "foo.bar.*".
5848 (when (and id-start (/= id-start id-end))
d9e94c22
MS
5849 (setq c-last-identifier-range
5850 (cons id-start id-end)))
5851 (goto-char id-end)
5852 (c-forward-syntactic-ws)
5853 (setq pos (point)
5854 res t)))
5855
5856 (progn
5857 (goto-char pos)
5858 (when (or c-opt-identifier-concat-key
5859 c-recognize-<>-arglists)
5860
5861 (cond
5862 ((and c-opt-identifier-concat-key
5863 (looking-at c-opt-identifier-concat-key))
5864 ;; Got a concatenated identifier. This handles the
5865 ;; cases with tricky syntactic whitespace that aren't
5866 ;; covered in `c-identifier-key'.
5867 (goto-char (match-end 0))
5868 (c-forward-syntactic-ws)
5869 t)
5870
5871 ((and c-recognize-<>-arglists
5872 (eq (char-after) ?<))
5873 ;; Maybe an angle bracket arglist.
452ea855
AM
5874 (when (let ((c-record-type-identifiers t)
5875 (c-record-found-types t))
0386b551
AM
5876 (c-forward-<>-arglist nil))
5877
5878 (c-add-type start (1+ pos))
d9e94c22 5879 (c-forward-syntactic-ws)
0386b551
AM
5880 (setq pos (point)
5881 c-last-identifier-range nil)
5882
d9e94c22
MS
5883 (if (and c-opt-identifier-concat-key
5884 (looking-at c-opt-identifier-concat-key))
0386b551 5885
d9e94c22
MS
5886 ;; Continue if there's an identifier concatenation
5887 ;; operator after the template argument.
5888 (progn
0386b551
AM
5889 (when (and c-record-type-identifiers id-start)
5890 (c-record-ref-id (cons id-start id-end)))
d9e94c22
MS
5891 (forward-char 2)
5892 (c-forward-syntactic-ws)
5893 t)
0386b551
AM
5894
5895 (when (and c-record-type-identifiers id-start)
5896 (c-record-type-id (cons id-start id-end)))
d9e94c22
MS
5897 (setq res 'template)
5898 nil)))
5899 )))))
5900
5901 (goto-char pos)
5902 res))
5903
e15f8aaa 5904(defun c-forward-type (&optional brace-block-too)
d9e94c22 5905 ;; Move forward over a type spec if at the beginning of one,
e15f8aaa
AM
5906 ;; stopping at the next following token. The keyword "typedef"
5907 ;; isn't part of a type spec here.
5908 ;;
5909 ;; BRACE-BLOCK-TOO, when non-nil, means move over the brace block in
5910 ;; constructs like "struct foo {...} bar ;" or "struct {...} bar;".
5911 ;; The current (2009-03-10) intention is to convert all uses of
5912 ;; `c-forward-type' to call with this parameter set, then to
5913 ;; eliminate it.
5914 ;;
5915 ;; Return
5916 ;; o - t if it's a known type that can't be a name or other
5917 ;; expression;
5918 ;; o - 'known if it's an otherwise known type (according to
5919 ;; `*-font-lock-extra-types');
5920 ;; o - 'prefix if it's a known prefix of a type;
5921 ;; o - 'found if it's a type that matches one in `c-found-types';
53964682 5922 ;; o - 'maybe if it's an identifier that might be a type; or
e15f8aaa
AM
5923 ;; o - nil if it can't be a type (the point isn't moved then).
5924 ;;
5925 ;; The point is assumed to be at the beginning of a token.
d9e94c22
MS
5926 ;;
5927 ;; Note that this function doesn't skip past the brace definition
5928 ;; that might be considered part of the type, e.g.
5929 ;; "enum {a, b, c} foo".
0386b551
AM
5930 ;;
5931 ;; This function records identifier ranges on
5932 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
5933 ;; `c-record-type-identifiers' is non-nil.
5934 ;;
5935 ;; This function might do hidden buffer changes.
0e5cf2b8
AM
5936 (when (and c-recognize-<>-arglists
5937 (looking-at "<"))
452ea855
AM
5938 (c-forward-<>-arglist t)
5939 (c-forward-syntactic-ws))
0386b551
AM
5940
5941 (let ((start (point)) pos res name-res id-start id-end id-range)
d9e94c22
MS
5942
5943 ;; Skip leading type modifiers. If any are found we know it's a
5944 ;; prefix of a type.
e15f8aaa 5945 (when c-opt-type-modifier-key ; e.g. "const" "volatile", but NOT "typedef"
d9e94c22
MS
5946 (while (looking-at c-opt-type-modifier-key)
5947 (goto-char (match-end 1))
5948 (c-forward-syntactic-ws)
5949 (setq res 'prefix)))
5950
5951 (cond
e15f8aaa
AM
5952 ((looking-at c-type-prefix-key) ; e.g. "struct", "class", but NOT
5953 ; "typedef".
d9e94c22
MS
5954 (goto-char (match-end 1))
5955 (c-forward-syntactic-ws)
5956 (setq pos (point))
e15f8aaa
AM
5957
5958 (setq name-res (c-forward-name))
5959 (setq res (not (null name-res)))
5960 (when (eq name-res t)
5961 ;; In many languages the name can be used without the
5962 ;; prefix, so we add it to `c-found-types'.
5963 (c-add-type pos (point))
5964 (when (and c-record-type-identifiers
5965 c-last-identifier-range)
5966 (c-record-type-id c-last-identifier-range)))
5967 (when (and brace-block-too
5968 (memq res '(t nil))
5969 (eq (char-after) ?\{)
5970 (save-excursion
5971 (c-safe
5972 (progn (c-forward-sexp)
5973 (c-forward-syntactic-ws)
5974 (setq pos (point))))))
5975 (goto-char pos)
5976 (setq res t))
5977 (unless res (goto-char start))) ; invalid syntax
d9e94c22
MS
5978
5979 ((progn
5980 (setq pos nil)
5981 (if (looking-at c-identifier-start)
5982 (save-excursion
5983 (setq id-start (point)
0386b551
AM
5984 name-res (c-forward-name))
5985 (when name-res
d9e94c22
MS
5986 (setq id-end (point)
5987 id-range c-last-identifier-range))))
5988 (and (cond ((looking-at c-primitive-type-key)
5989 (setq res t))
5990 ((c-with-syntax-table c-identifier-syntax-table
5991 (looking-at c-known-type-key))
5992 (setq res 'known)))
5993 (or (not id-end)
5994 (>= (save-excursion
5995 (save-match-data
5996 (goto-char (match-end 1))
5997 (c-forward-syntactic-ws)
5998 (setq pos (point))))
5999 id-end)
6000 (setq res nil))))
6001 ;; Looking at a primitive or known type identifier. We've
6002 ;; checked for a name first so that we don't go here if the
6003 ;; known type match only is a prefix of another name.
6004
6005 (setq id-end (match-end 1))
6006
6007 (when (and c-record-type-identifiers
6008 (or c-promote-possible-types (eq res t)))
6009 (c-record-type-id (cons (match-beginning 1) (match-end 1))))
6010
6011 (if (and c-opt-type-component-key
6012 (save-match-data
6013 (looking-at c-opt-type-component-key)))
6014 ;; There might be more keywords for the type.
6015 (let (safe-pos)
0386b551 6016 (c-forward-keyword-clause 1)
d9e94c22
MS
6017 (while (progn
6018 (setq safe-pos (point))
6019 (looking-at c-opt-type-component-key))
6020 (when (and c-record-type-identifiers
6021 (looking-at c-primitive-type-key))
6022 (c-record-type-id (cons (match-beginning 1)
6023 (match-end 1))))
0386b551 6024 (c-forward-keyword-clause 1))
d9e94c22
MS
6025 (if (looking-at c-primitive-type-key)
6026 (progn
6027 (when c-record-type-identifiers
6028 (c-record-type-id (cons (match-beginning 1)
6029 (match-end 1))))
0386b551 6030 (c-forward-keyword-clause 1)
d9e94c22
MS
6031 (setq res t))
6032 (goto-char safe-pos)
6033 (setq res 'prefix)))
0386b551 6034 (unless (save-match-data (c-forward-keyword-clause 1))
d9e94c22
MS
6035 (if pos
6036 (goto-char pos)
6037 (goto-char (match-end 1))
6038 (c-forward-syntactic-ws)))))
6039
0386b551
AM
6040 (name-res
6041 (cond ((eq name-res t)
d9e94c22
MS
6042 ;; A normal identifier.
6043 (goto-char id-end)
6044 (if (or res c-promote-possible-types)
6045 (progn
6046 (c-add-type id-start id-end)
0386b551 6047 (when (and c-record-type-identifiers id-range)
d9e94c22
MS
6048 (c-record-type-id id-range))
6049 (unless res
6050 (setq res 'found)))
6051 (setq res (if (c-check-type id-start id-end)
6052 ;; It's an identifier that has been used as
6053 ;; a type somewhere else.
6054 'found
6055 ;; It's an identifier that might be a type.
6056 'maybe))))
0386b551 6057 ((eq name-res 'template)
d9e94c22
MS
6058 ;; A template is a type.
6059 (goto-char id-end)
6060 (setq res t))
6061 (t
6062 ;; Otherwise it's an operator identifier, which is not a type.
6063 (goto-char start)
6064 (setq res nil)))))
6065
6066 (when res
e15f8aaa 6067 ;; Skip trailing type modifiers. If any are found we know it's
d9e94c22
MS
6068 ;; a type.
6069 (when c-opt-type-modifier-key
e15f8aaa 6070 (while (looking-at c-opt-type-modifier-key) ; e.g. "const", "volatile"
d9e94c22
MS
6071 (goto-char (match-end 1))
6072 (c-forward-syntactic-ws)
6073 (setq res t)))
d9e94c22
MS
6074 ;; Step over any type suffix operator. Do not let the existence
6075 ;; of these alter the classification of the found type, since
6076 ;; these operators typically are allowed in normal expressions
6077 ;; too.
6078 (when c-opt-type-suffix-key
6079 (while (looking-at c-opt-type-suffix-key)
6080 (goto-char (match-end 1))
6081 (c-forward-syntactic-ws)))
6082
e15f8aaa 6083 (when c-opt-type-concat-key ; Only/mainly for pike.
0386b551
AM
6084 ;; Look for a trailing operator that concatenates the type
6085 ;; with a following one, and if so step past that one through
6086 ;; a recursive call. Note that we don't record concatenated
6087 ;; types in `c-found-types' - it's the component types that
6088 ;; are recorded when appropriate.
d9e94c22
MS
6089 (setq pos (point))
6090 (let* ((c-promote-possible-types (or (memq res '(t known))
6091 c-promote-possible-types))
6092 ;; If we can't promote then set `c-record-found-types' so that
6093 ;; we can merge in the types from the second part afterwards if
6094 ;; it turns out to be a known type there.
6095 (c-record-found-types (and c-record-type-identifiers
0386b551
AM
6096 (not c-promote-possible-types)))
6097 subres)
d9e94c22
MS
6098 (if (and (looking-at c-opt-type-concat-key)
6099
6100 (progn
6101 (goto-char (match-end 1))
6102 (c-forward-syntactic-ws)
0386b551 6103 (setq subres (c-forward-type))))
d9e94c22
MS
6104
6105 (progn
6106 ;; If either operand certainly is a type then both are, but we
6107 ;; don't let the existence of the operator itself promote two
6108 ;; uncertain types to a certain one.
6109 (cond ((eq res t))
0386b551
AM
6110 ((eq subres t)
6111 (unless (eq name-res 'template)
6112 (c-add-type id-start id-end))
6113 (when (and c-record-type-identifiers id-range)
d9e94c22
MS
6114 (c-record-type-id id-range))
6115 (setq res t))
2a15eb73 6116 ((eq res 'known))
0386b551 6117 ((eq subres 'known)
2a15eb73 6118 (setq res 'known))
d9e94c22 6119 ((eq res 'found))
0386b551 6120 ((eq subres 'found)
d9e94c22
MS
6121 (setq res 'found))
6122 (t
6123 (setq res 'maybe)))
6124
6125 (when (and (eq res t)
6126 (consp c-record-found-types))
6127 ;; Merge in the ranges of any types found by the second
6128 ;; `c-forward-type'.
6129 (setq c-record-type-identifiers
6130 ;; `nconc' doesn't mind that the tail of
6131 ;; `c-record-found-types' is t.
6132 (nconc c-record-found-types
6133 c-record-type-identifiers))))
6134
6135 (goto-char pos))))
6136
6137 (when (and c-record-found-types (memq res '(known found)) id-range)
6138 (setq c-record-found-types
6139 (cons id-range c-record-found-types))))
6140
6141 ;;(message "c-forward-type %s -> %s: %s" start (point) res)
6142
6143 res))
6144
452ea855
AM
6145(defun c-forward-annotation ()
6146 ;; Used for Java code only at the moment. Assumes point is on the
6147 ;; @, moves forward an annotation. returns nil if there is no
6148 ;; annotation at point.
6149 (and (looking-at "@")
6150 (progn (forward-char) t)
6151 (c-forward-type)
6152 (progn (c-forward-syntactic-ws) t)
6153 (if (looking-at "(")
6154 (c-go-list-forward)
6155 t)))
6156
785eecbb 6157\f
d9e94c22
MS
6158;; Handling of large scale constructs like statements and declarations.
6159
0386b551
AM
6160;; Macro used inside `c-forward-decl-or-cast-1'. It ought to be a
6161;; defsubst or perhaps even a defun, but it contains lots of free
6162;; variables that refer to things inside `c-forward-decl-or-cast-1'.
6163(defmacro c-fdoc-shift-type-backward (&optional short)
6164 ;; `c-forward-decl-or-cast-1' can consume an arbitrary length list
6165 ;; of types when parsing a declaration, which means that it
6166 ;; sometimes consumes the identifier in the declaration as a type.
6167 ;; This is used to "backtrack" and make the last type be treated as
6168 ;; an identifier instead.
6169 `(progn
6170 ,(unless short
6171 ;; These identifiers are bound only in the inner let.
6172 '(setq identifier-type at-type
6173 identifier-start type-start
6174 got-parens nil
6175 got-identifier t
6176 got-suffix t
6177 got-suffix-after-parens id-start
6178 paren-depth 0))
6179
6180 (if (setq at-type (if (eq backup-at-type 'prefix)
6181 t
6182 backup-at-type))
6183 (setq type-start backup-type-start
6184 id-start backup-id-start)
6185 (setq type-start start-pos
6186 id-start start-pos))
6187
6188 ;; When these flags already are set we've found specifiers that
6189 ;; unconditionally signal these attributes - backtracking doesn't
6190 ;; change that. So keep them set in that case.
6191 (or at-type-decl
6192 (setq at-type-decl backup-at-type-decl))
6193 (or maybe-typeless
6194 (setq maybe-typeless backup-maybe-typeless))
6195
6196 ,(unless short
6197 ;; This identifier is bound only in the inner let.
6198 '(setq start id-start))))
6199
6200(defun c-forward-decl-or-cast-1 (preceding-token-end context last-cast-end)
6201 ;; Move forward over a declaration or a cast if at the start of one.
6202 ;; The point is assumed to be at the start of some token. Nil is
6203 ;; returned if no declaration or cast is recognized, and the point
6204 ;; is clobbered in that case.
6205 ;;
6206 ;; If a declaration is parsed:
6207 ;;
6208 ;; The point is left at the first token after the first complete
6209 ;; declarator, if there is one. The return value is a cons where
022d0cf4
AM
6210 ;; the car is the position of the first token in the declarator. (See
6211 ;; below for the cdr.)
0386b551
AM
6212 ;; Some examples:
6213 ;;
6214 ;; void foo (int a, char *b) stuff ...
6215 ;; car ^ ^ point
6216 ;; float (*a)[], b;
6217 ;; car ^ ^ point
6218 ;; unsigned int a = c_style_initializer, b;
6219 ;; car ^ ^ point
6220 ;; unsigned int a (cplusplus_style_initializer), b;
6221 ;; car ^ ^ point (might change)
6222 ;; class Foo : public Bar {}
6223 ;; car ^ ^ point
6224 ;; class PikeClass (int a, string b) stuff ...
6225 ;; car ^ ^ point
6226 ;; enum bool;
6227 ;; car ^ ^ point
6228 ;; enum bool flag;
6229 ;; car ^ ^ point
6230 ;; void cplusplus_function (int x) throw (Bad);
6231 ;; car ^ ^ point
6232 ;; Foo::Foo (int b) : Base (b) {}
6233 ;; car ^ ^ point
91af3942 6234 ;;
e15f8aaa
AM
6235 ;; The cdr of the return value is non-nil when a
6236 ;; `c-typedef-decl-kwds' specifier is found in the declaration.
6237 ;; Specifically it is a dotted pair (A . B) where B is t when a
6238 ;; `c-typedef-kwds' ("typedef") is present, and A is t when some
6239 ;; other `c-typedef-decl-kwds' (e.g. class, struct, enum)
6240 ;; specifier is present. I.e., (some of) the declared
6241 ;; identifier(s) are types.
91af3942 6242 ;;
0386b551
AM
6243 ;; If a cast is parsed:
6244 ;;
6245 ;; The point is left at the first token after the closing paren of
6246 ;; the cast. The return value is `cast'. Note that the start
6247 ;; position must be at the first token inside the cast parenthesis
6248 ;; to recognize it.
6249 ;;
6250 ;; PRECEDING-TOKEN-END is the first position after the preceding
6251 ;; token, i.e. on the other side of the syntactic ws from the point.
6252 ;; Use a value less than or equal to (point-min) if the point is at
6253 ;; the first token in (the visible part of) the buffer.
6254 ;;
6255 ;; CONTEXT is a symbol that describes the context at the point:
022d0cf4 6256 ;; 'decl In a comma-separated declaration context (typically
0386b551
AM
6257 ;; inside a function declaration arglist).
6258 ;; '<> In an angle bracket arglist.
6259 ;; 'arglist Some other type of arglist.
a85fd6da
AM
6260 ;; nil Some other context or unknown context. Includes
6261 ;; within the parens of an if, for, ... construct.
0386b551
AM
6262 ;;
6263 ;; LAST-CAST-END is the first token after the closing paren of a
6264 ;; preceding cast, or nil if none is known. If
6265 ;; `c-forward-decl-or-cast-1' is used in succession, it should be
6266 ;; the position after the closest preceding call where a cast was
6267 ;; matched. In that case it's used to discover chains of casts like
6268 ;; "(a) (b) c".
6269 ;;
6270 ;; This function records identifier ranges on
6271 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
6272 ;; `c-record-type-identifiers' is non-nil.
6273 ;;
6274 ;; This function might do hidden buffer changes.
6275
6276 (let (;; `start-pos' is used below to point to the start of the
6277 ;; first type, i.e. after any leading specifiers. It might
6278 ;; also point at the beginning of the preceding syntactic
6279 ;; whitespace.
6280 (start-pos (point))
6281 ;; Set to the result of `c-forward-type'.
6282 at-type
6283 ;; The position of the first token in what we currently
6284 ;; believe is the type in the declaration or cast, after any
6285 ;; specifiers and their associated clauses.
6286 type-start
6287 ;; The position of the first token in what we currently
6288 ;; believe is the declarator for the first identifier. Set
6289 ;; when the type is found, and moved forward over any
6290 ;; `c-decl-hangon-kwds' and their associated clauses that
6291 ;; occurs after the type.
6292 id-start
6293 ;; These store `at-type', `type-start' and `id-start' of the
6294 ;; identifier before the one in those variables. The previous
6295 ;; identifier might turn out to be the real type in a
6296 ;; declaration if the last one has to be the declarator in it.
6297 ;; If `backup-at-type' is nil then the other variables have
6298 ;; undefined values.
6299 backup-at-type backup-type-start backup-id-start
e15f8aaa
AM
6300 ;; Set if we've found a specifier (apart from "typedef") that makes
6301 ;; the defined identifier(s) types.
0386b551 6302 at-type-decl
e15f8aaa
AM
6303 ;; Set if we've a "typedef" keyword.
6304 at-typedef
0386b551
AM
6305 ;; Set if we've found a specifier that can start a declaration
6306 ;; where there's no type.
6307 maybe-typeless
6308 ;; If a specifier is found that also can be a type prefix,
6309 ;; these flags are set instead of those above. If we need to
6310 ;; back up an identifier, they are copied to the real flag
6311 ;; variables. Thus they only take effect if we fail to
6312 ;; interpret it as a type.
6313 backup-at-type-decl backup-maybe-typeless
6314 ;; Whether we've found a declaration or a cast. We might know
6315 ;; this before we've found the type in it. It's 'ids if we've
6316 ;; found two consecutive identifiers (usually a sure sign, but
6317 ;; we should allow that in labels too), and t if we've found a
6318 ;; specifier keyword (a 100% sure sign).
6319 at-decl-or-cast
6320 ;; Set when we need to back up to parse this as a declaration
6321 ;; but not as a cast.
6322 backup-if-not-cast
6323 ;; For casts, the return position.
6324 cast-end
6325 ;; Save `c-record-type-identifiers' and
6326 ;; `c-record-ref-identifiers' since ranges are recorded
6327 ;; speculatively and should be thrown away if it turns out
6328 ;; that it isn't a declaration or cast.
6329 (save-rec-type-ids c-record-type-identifiers)
6330 (save-rec-ref-ids c-record-ref-identifiers))
6331
452ea855
AM
6332 (while (c-forward-annotation)
6333 (c-forward-syntactic-ws))
6334
0386b551
AM
6335 ;; Check for a type. Unknown symbols are treated as possible
6336 ;; types, but they could also be specifiers disguised through
6337 ;; macros like __INLINE__, so we recognize both types and known
6338 ;; specifiers after them too.
6339 (while
6340 (let* ((start (point)) kwd-sym kwd-clause-end found-type)
6341
6342 ;; Look for a specifier keyword clause.
e3715033
AM
6343 (when (or (looking-at c-prefix-spec-kwds-re)
6344 (and (c-major-mode-is 'java-mode)
6345 (looking-at "@[A-Za-z0-9]+")))
e15f8aaa
AM
6346 (if (looking-at c-typedef-key)
6347 (setq at-typedef t))
0386b551
AM
6348 (setq kwd-sym (c-keyword-sym (match-string 1)))
6349 (save-excursion
6350 (c-forward-keyword-clause 1)
6351 (setq kwd-clause-end (point))))
6352
e15f8aaa 6353 (when (setq found-type (c-forward-type t)) ; brace-block-too
0386b551
AM
6354 ;; Found a known or possible type or a prefix of a known type.
6355
6356 (when at-type
6357 ;; Got two identifiers with nothing but whitespace
6358 ;; between them. That can only happen in declarations.
6359 (setq at-decl-or-cast 'ids)
6360
6361 (when (eq at-type 'found)
6362 ;; If the previous identifier is a found type we
6363 ;; record it as a real one; it might be some sort of
6364 ;; alias for a prefix like "unsigned".
6365 (save-excursion
6366 (goto-char type-start)
6367 (let ((c-promote-possible-types t))
6368 (c-forward-type)))))
6369
6370 (setq backup-at-type at-type
6371 backup-type-start type-start
6372 backup-id-start id-start
6373 at-type found-type
6374 type-start start
6375 id-start (point)
6376 ;; The previous ambiguous specifier/type turned out
6377 ;; to be a type since we've parsed another one after
6378 ;; it, so clear these backup flags.
6379 backup-at-type-decl nil
6380 backup-maybe-typeless nil))
6381
6382 (if kwd-sym
6383 (progn
6384 ;; Handle known specifier keywords and
6385 ;; `c-decl-hangon-kwds' which can occur after known
6386 ;; types.
6387
6388 (if (c-keyword-member kwd-sym 'c-decl-hangon-kwds)
6389 ;; It's a hang-on keyword that can occur anywhere.
6390 (progn
6391 (setq at-decl-or-cast t)
6392 (if at-type
6393 ;; Move the identifier start position if
6394 ;; we've passed a type.
6395 (setq id-start kwd-clause-end)
6396 ;; Otherwise treat this as a specifier and
6397 ;; move the fallback position.
6398 (setq start-pos kwd-clause-end))
6399 (goto-char kwd-clause-end))
6400
6401 ;; It's an ordinary specifier so we know that
6402 ;; anything before this can't be the type.
6403 (setq backup-at-type nil
6404 start-pos kwd-clause-end)
6405
6406 (if found-type
6407 ;; It's ambiguous whether this keyword is a
6408 ;; specifier or a type prefix, so set the backup
6409 ;; flags. (It's assumed that `c-forward-type'
6410 ;; moved further than `c-forward-keyword-clause'.)
6411 (progn
6412 (when (c-keyword-member kwd-sym 'c-typedef-decl-kwds)
6413 (setq backup-at-type-decl t))
6414 (when (c-keyword-member kwd-sym 'c-typeless-decl-kwds)
6415 (setq backup-maybe-typeless t)))
6416
6417 (when (c-keyword-member kwd-sym 'c-typedef-decl-kwds)
e15f8aaa
AM
6418 ;; This test only happens after we've scanned a type.
6419 ;; So, with valid syntax, kwd-sym can't be 'typedef.
0386b551
AM
6420 (setq at-type-decl t))
6421 (when (c-keyword-member kwd-sym 'c-typeless-decl-kwds)
6422 (setq maybe-typeless t))
6423
ee7683eb 6424 ;; Haven't matched a type so it's an unambiguous
0386b551
AM
6425 ;; specifier keyword and we know we're in a
6426 ;; declaration.
6427 (setq at-decl-or-cast t)
6428
6429 (goto-char kwd-clause-end))))
6430
6431 ;; If the type isn't known we continue so that we'll jump
6432 ;; over all specifiers and type identifiers. The reason
6433 ;; to do this for a known type prefix is to make things
6434 ;; like "unsigned INT16" work.
6435 (and found-type (not (eq found-type t))))))
6436
6437 (cond
6438 ((eq at-type t)
6439 ;; If a known type was found, we still need to skip over any
6440 ;; hangon keyword clauses after it. Otherwise it has already
6441 ;; been done in the loop above.
6442 (while (looking-at c-decl-hangon-key)
6443 (c-forward-keyword-clause 1))
6444 (setq id-start (point)))
6445
6446 ((eq at-type 'prefix)
6447 ;; A prefix type is itself a primitive type when it's not
6448 ;; followed by another type.
6449 (setq at-type t))
6450
6451 ((not at-type)
6452 ;; Got no type but set things up to continue anyway to handle
6453 ;; the various cases when a declaration doesn't start with a
6454 ;; type.
6455 (setq id-start start-pos))
6456
6457 ((and (eq at-type 'maybe)
6458 (c-major-mode-is 'c++-mode))
6459 ;; If it's C++ then check if the last "type" ends on the form
6460 ;; "foo::foo" or "foo::~foo", i.e. if it's the name of a
6461 ;; (con|de)structor.
6462 (save-excursion
6463 (let (name end-2 end-1)
6464 (goto-char id-start)
6465 (c-backward-syntactic-ws)
6466 (setq end-2 (point))
6467 (when (and
6468 (c-simple-skip-symbol-backward)
6469 (progn
6470 (setq name
6471 (buffer-substring-no-properties (point) end-2))
6472 ;; Cheating in the handling of syntactic ws below.
6473 (< (skip-chars-backward ":~ \t\n\r\v\f") 0))
6474 (progn
6475 (setq end-1 (point))
6476 (c-simple-skip-symbol-backward))
6477 (>= (point) type-start)
6478 (equal (buffer-substring-no-properties (point) end-1)
6479 name))
6480 ;; It is a (con|de)structor name. In that case the
6481 ;; declaration is typeless so zap out any preceding
6482 ;; identifier(s) that we might have taken as types.
6483 (goto-char type-start)
6484 (setq at-type nil
6485 backup-at-type nil
6486 id-start type-start))))))
6487
6488 ;; Check for and step over a type decl expression after the thing
6489 ;; that is or might be a type. This can't be skipped since we
6490 ;; need the correct end position of the declarator for
6491 ;; `max-type-decl-end-*'.
6492 (let ((start (point)) (paren-depth 0) pos
6493 ;; True if there's a non-open-paren match of
6494 ;; `c-type-decl-prefix-key'.
6495 got-prefix
6496 ;; True if the declarator is surrounded by a parenthesis pair.
6497 got-parens
6498 ;; True if there is an identifier in the declarator.
6499 got-identifier
6500 ;; True if there's a non-close-paren match of
6501 ;; `c-type-decl-suffix-key'.
6502 got-suffix
6503 ;; True if there's a prefix match outside the outermost
6504 ;; paren pair that surrounds the declarator.
6505 got-prefix-before-parens
1379f2c5 6506 ;; True if there's a suffix match outside the outermost
0386b551
AM
6507 ;; paren pair that surrounds the declarator. The value is
6508 ;; the position of the first suffix match.
6509 got-suffix-after-parens
6510 ;; True if we've parsed the type decl to a token that is
6511 ;; known to end declarations in this context.
6512 at-decl-end
6513 ;; The earlier values of `at-type' and `type-start' if we've
6514 ;; shifted the type backwards.
6515 identifier-type identifier-start
6516 ;; If `c-parse-and-markup-<>-arglists' is set we need to
6517 ;; turn it off during the name skipping below to avoid
6518 ;; getting `c-type' properties that might be bogus. That
6519 ;; can happen since we don't know if
6520 ;; `c-restricted-<>-arglists' will be correct inside the
6521 ;; arglist paren that gets entered.
6522 c-parse-and-markup-<>-arglists)
6523
6524 (goto-char id-start)
6525
6526 ;; Skip over type decl prefix operators. (Note similar code in
6527 ;; `c-font-lock-declarators'.)
6528 (while (and (looking-at c-type-decl-prefix-key)
6529 (if (and (c-major-mode-is 'c++-mode)
05842630 6530 (match-beginning 3))
0386b551
AM
6531 ;; If the second submatch matches in C++ then
6532 ;; we're looking at an identifier that's a
6533 ;; prefix only if it specifies a member pointer.
6534 (when (setq got-identifier (c-forward-name))
6535 (if (looking-at "\\(::\\)")
6536 ;; We only check for a trailing "::" and
6537 ;; let the "*" that should follow be
6538 ;; matched in the next round.
6539 (progn (setq got-identifier nil) t)
6540 ;; It turned out to be the real identifier,
6541 ;; so stop.
6542 nil))
6543 t))
6544
6545 (if (eq (char-after) ?\()
6546 (progn
6547 (setq paren-depth (1+ paren-depth))
6548 (forward-char))
6549 (unless got-prefix-before-parens
6550 (setq got-prefix-before-parens (= paren-depth 0)))
6551 (setq got-prefix t)
6552 (goto-char (match-end 1)))
6553 (c-forward-syntactic-ws))
6554
6555 (setq got-parens (> paren-depth 0))
6556
6557 ;; Skip over an identifier.
6558 (or got-identifier
6559 (and (looking-at c-identifier-start)
6560 (setq got-identifier (c-forward-name))))
6561
6562 ;; Skip over type decl suffix operators.
6563 (while (if (looking-at c-type-decl-suffix-key)
6564
6565 (if (eq (char-after) ?\))
6566 (when (> paren-depth 0)
6567 (setq paren-depth (1- paren-depth))
6568 (forward-char)
6569 t)
6570 (when (if (save-match-data (looking-at "\\s\("))
6571 (c-safe (c-forward-sexp 1) t)
6572 (goto-char (match-end 1))
6573 t)
6574 (when (and (not got-suffix-after-parens)
6575 (= paren-depth 0))
6576 (setq got-suffix-after-parens (match-beginning 0)))
6577 (setq got-suffix t)))
6578
6579 ;; No suffix matched. We might have matched the
6580 ;; identifier as a type and the open paren of a
6581 ;; function arglist as a type decl prefix. In that
6582 ;; case we should "backtrack": Reinterpret the last
6583 ;; type as the identifier, move out of the arglist and
6584 ;; continue searching for suffix operators.
6585 ;;
6586 ;; Do this even if there's no preceding type, to cope
6587 ;; with old style function declarations in K&R C,
6588 ;; (con|de)structors in C++ and `c-typeless-decl-kwds'
6589 ;; style declarations. That isn't applicable in an
6590 ;; arglist context, though.
6591 (when (and (= paren-depth 1)
6592 (not got-prefix-before-parens)
6593 (not (eq at-type t))
6594 (or backup-at-type
6595 maybe-typeless
6596 backup-maybe-typeless
6597 (when c-recognize-typeless-decls
6598 (not context)))
6599 (setq pos (c-up-list-forward (point)))
6600 (eq (char-before pos) ?\)))
6601 (c-fdoc-shift-type-backward)
6602 (goto-char pos)
6603 t))
6604
6605 (c-forward-syntactic-ws))
6606
6607 (when (and (or maybe-typeless backup-maybe-typeless)
6608 (not got-identifier)
6609 (not got-prefix)
6610 at-type)
6611 ;; Have found no identifier but `c-typeless-decl-kwds' has
6612 ;; matched so we know we're inside a declaration. The
6613 ;; preceding type must be the identifier instead.
6614 (c-fdoc-shift-type-backward))
6615
6616 (setq
6617 at-decl-or-cast
6618 (catch 'at-decl-or-cast
6619
a85fd6da 6620 ;; CASE 1
0386b551
AM
6621 (when (> paren-depth 0)
6622 ;; Encountered something inside parens that isn't matched by
6623 ;; the `c-type-decl-*' regexps, so it's not a type decl
6624 ;; expression. Try to skip out to the same paren depth to
6625 ;; not confuse the cast check below.
6626 (c-safe (goto-char (scan-lists (point) 1 paren-depth)))
6627 ;; If we've found a specifier keyword then it's a
6628 ;; declaration regardless.
6629 (throw 'at-decl-or-cast (eq at-decl-or-cast t)))
6630
6631 (setq at-decl-end
6632 (looking-at (cond ((eq context '<>) "[,>]")
6633 (context "[,\)]")
6634 (t "[,;]"))))
6635
6636 ;; Now we've collected info about various characteristics of
6637 ;; the construct we're looking at. Below follows a decision
6638 ;; tree based on that. It's ordered to check more certain
6639 ;; signs before less certain ones.
6640
6641 (if got-identifier
6642 (progn
6643
a85fd6da 6644 ;; CASE 2
0386b551
AM
6645 (when (and (or at-type maybe-typeless)
6646 (not (or got-prefix got-parens)))
6647 ;; Got another identifier directly after the type, so it's a
6648 ;; declaration.
6649 (throw 'at-decl-or-cast t))
6650
6651 (when (and got-parens
6652 (not got-prefix)
6653 (not got-suffix-after-parens)
6654 (or backup-at-type
6655 maybe-typeless
6656 backup-maybe-typeless))
6657 ;; Got a declaration of the form "foo bar (gnu);" where we've
6658 ;; recognized "bar" as the type and "gnu" as the declarator.
6659 ;; In this case it's however more likely that "bar" is the
6660 ;; declarator and "gnu" a function argument or initializer (if
6661 ;; `c-recognize-paren-inits' is set), since the parens around
6662 ;; "gnu" would be superfluous if it's a declarator. Shift the
6663 ;; type one step backward.
6664 (c-fdoc-shift-type-backward)))
6665
6666 ;; Found no identifier.
6667
6668 (if backup-at-type
6669 (progn
6670
b248a85d
AM
6671
6672 ;; CASE 3
6673 (when (= (point) start)
6674 ;; Got a plain list of identifiers. If a colon follows it's
6675 ;; a valid label, or maybe a bitfield. Otherwise the last
6676 ;; one probably is the declared identifier and we should
6677 ;; back up to the previous type, providing it isn't a cast.
452ea855
AM
6678 (if (and (eq (char-after) ?:)
6679 (not (c-major-mode-is 'java-mode)))
b248a85d
AM
6680 (cond
6681 ;; If we've found a specifier keyword then it's a
6682 ;; declaration regardless.
6683 ((eq at-decl-or-cast t)
6684 (throw 'at-decl-or-cast t))
6685 ((and c-has-bitfields
6686 (eq at-decl-or-cast 'ids)) ; bitfield.
6687 (setq backup-if-not-cast t)
6688 (throw 'at-decl-or-cast t)))
6689
6690 (setq backup-if-not-cast t)
6691 (throw 'at-decl-or-cast t)))
0386b551 6692
a85fd6da 6693 ;; CASE 4
0386b551
AM
6694 (when (and got-suffix
6695 (not got-prefix)
6696 (not got-parens))
6697 ;; Got a plain list of identifiers followed by some suffix.
6698 ;; If this isn't a cast then the last identifier probably is
6699 ;; the declared one and we should back up to the previous
6700 ;; type.
6701 (setq backup-if-not-cast t)
6702 (throw 'at-decl-or-cast t)))
6703
a85fd6da 6704 ;; CASE 5
0386b551
AM
6705 (when (eq at-type t)
6706 ;; If the type is known we know that there can't be any
6707 ;; identifier somewhere else, and it's only in declarations in
6708 ;; e.g. function prototypes and in casts that the identifier may
6709 ;; be left out.
6710 (throw 'at-decl-or-cast t))
6711
6712 (when (= (point) start)
6713 ;; Only got a single identifier (parsed as a type so far).
a85fd6da 6714 ;; CASE 6
0386b551
AM
6715 (if (and
6716 ;; Check that the identifier isn't at the start of an
6717 ;; expression.
6718 at-decl-end
6719 (cond
6720 ((eq context 'decl)
6721 ;; Inside an arglist that contains declarations. If K&R
6722 ;; style declarations and parenthesis style initializers
6723 ;; aren't allowed then the single identifier must be a
6724 ;; type, else we require that it's known or found
6725 ;; (primitive types are handled above).
6726 (or (and (not c-recognize-knr-p)
6727 (not c-recognize-paren-inits))
6728 (memq at-type '(known found))))
6729 ((eq context '<>)
6730 ;; Inside a template arglist. Accept known and found
6731 ;; types; other identifiers could just as well be
6732 ;; constants in C++.
6733 (memq at-type '(known found)))))
6734 (throw 'at-decl-or-cast t)
a85fd6da 6735 ;; CASE 7
0386b551
AM
6736 ;; Can't be a valid declaration or cast, but if we've found a
6737 ;; specifier it can't be anything else either, so treat it as
6738 ;; an invalid/unfinished declaration or cast.
6739 (throw 'at-decl-or-cast at-decl-or-cast))))
6740
6741 (if (and got-parens
6742 (not got-prefix)
6743 (not context)
6744 (not (eq at-type t))
6745 (or backup-at-type
6746 maybe-typeless
6747 backup-maybe-typeless
6748 (when c-recognize-typeless-decls
6749 (or (not got-suffix)
6750 (not (looking-at
6751 c-after-suffixed-type-maybe-decl-key))))))
6752 ;; Got an empty paren pair and a preceding type that probably
6753 ;; really is the identifier. Shift the type backwards to make
6754 ;; the last one the identifier. This is analogous to the
6755 ;; "backtracking" done inside the `c-type-decl-suffix-key' loop
6756 ;; above.
6757 ;;
6758 ;; Exception: In addition to the conditions in that
6759 ;; "backtracking" code, do not shift backward if we're not
6760 ;; looking at either `c-after-suffixed-type-decl-key' or "[;,]".
6761 ;; Since there's no preceding type, the shift would mean that
6762 ;; the declaration is typeless. But if the regexp doesn't match
6763 ;; then we will simply fall through in the tests below and not
6764 ;; recognize it at all, so it's better to try it as an abstract
6765 ;; declarator instead.
6766 (c-fdoc-shift-type-backward)
6767
6768 ;; Still no identifier.
a85fd6da 6769 ;; CASE 8
0386b551
AM
6770 (when (and got-prefix (or got-parens got-suffix))
6771 ;; Require `got-prefix' together with either `got-parens' or
6772 ;; `got-suffix' to recognize it as an abstract declarator:
6773 ;; `got-parens' only is probably an empty function call.
6774 ;; `got-suffix' only can build an ordinary expression together
6775 ;; with the preceding identifier which we've taken as a type.
6776 ;; We could actually accept on `got-prefix' only, but that can
6777 ;; easily occur temporarily while writing an expression so we
6778 ;; avoid that case anyway. We could do a better job if we knew
6779 ;; the point when the fontification was invoked.
6780 (throw 'at-decl-or-cast t))
6781
a85fd6da 6782 ;; CASE 9
0386b551
AM
6783 (when (and at-type
6784 (not got-prefix)
6785 (not got-parens)
6786 got-suffix-after-parens
6787 (eq (char-after got-suffix-after-parens) ?\())
6788 ;; Got a type, no declarator but a paren suffix. I.e. it's a
fa463103 6789 ;; normal function call after all (or perhaps a C++ style object
0386b551
AM
6790 ;; instantiation expression).
6791 (throw 'at-decl-or-cast nil))))
6792
a85fd6da 6793 ;; CASE 10
0386b551
AM
6794 (when at-decl-or-cast
6795 ;; By now we've located the type in the declaration that we know
6796 ;; we're in.
6797 (throw 'at-decl-or-cast t))
6798
a85fd6da 6799 ;; CASE 11
0386b551
AM
6800 (when (and got-identifier
6801 (not context)
6802 (looking-at c-after-suffixed-type-decl-key)
6803 (if (and got-parens
6804 (not got-prefix)
6805 (not got-suffix)
6806 (not (eq at-type t)))
6807 ;; Shift the type backward in the case that there's a
6808 ;; single identifier inside parens. That can only
6809 ;; occur in K&R style function declarations so it's
6810 ;; more likely that it really is a function call.
6811 ;; Therefore we only do this after
6812 ;; `c-after-suffixed-type-decl-key' has matched.
6813 (progn (c-fdoc-shift-type-backward) t)
6814 got-suffix-after-parens))
6815 ;; A declaration according to `c-after-suffixed-type-decl-key'.
6816 (throw 'at-decl-or-cast t))
6817
a85fd6da 6818 ;; CASE 12
0386b551
AM
6819 (when (and (or got-prefix (not got-parens))
6820 (memq at-type '(t known)))
6821 ;; It's a declaration if a known type precedes it and it can't be a
6822 ;; function call.
6823 (throw 'at-decl-or-cast t))
6824
6825 ;; If we get here we can't tell if this is a type decl or a normal
6826 ;; expression by looking at it alone. (That's under the assumption
6827 ;; that normal expressions always can look like type decl expressions,
6828 ;; which isn't really true but the cases where it doesn't hold are so
6829 ;; uncommon (e.g. some placements of "const" in C++) it's not worth
6830 ;; the effort to look for them.)
6831
6832 (unless (or at-decl-end (looking-at "=[^=]"))
6833 ;; If this is a declaration it should end here or its initializer(*)
6834 ;; should start here, so check for allowed separation tokens. Note
6835 ;; that this rule doesn't work e.g. with a K&R arglist after a
6836 ;; function header.
6837 ;;
6838 ;; *) Don't check for C++ style initializers using parens
6839 ;; since those already have been matched as suffixes.
6840 ;;
6841 ;; If `at-decl-or-cast' is then we've found some other sign that
6842 ;; it's a declaration or cast, so then it's probably an
6843 ;; invalid/unfinished one.
6844 (throw 'at-decl-or-cast at-decl-or-cast))
6845
6846 ;; Below are tests that only should be applied when we're certain to
6847 ;; not have parsed halfway through an expression.
6848
a85fd6da 6849 ;; CASE 14
0386b551
AM
6850 (when (memq at-type '(t known))
6851 ;; The expression starts with a known type so treat it as a
6852 ;; declaration.
6853 (throw 'at-decl-or-cast t))
6854
a85fd6da 6855 ;; CASE 15
0386b551
AM
6856 (when (and (c-major-mode-is 'c++-mode)
6857 ;; In C++ we check if the identifier is a known type, since
6858 ;; (con|de)structors use the class name as identifier.
6859 ;; We've always shifted over the identifier as a type and
6860 ;; then backed up again in this case.
6861 identifier-type
6862 (or (memq identifier-type '(found known))
6863 (and (eq (char-after identifier-start) ?~)
6864 ;; `at-type' probably won't be 'found for
6865 ;; destructors since the "~" is then part of the
6866 ;; type name being checked against the list of
6867 ;; known types, so do a check without that
6868 ;; operator.
6869 (or (save-excursion
6870 (goto-char (1+ identifier-start))
6871 (c-forward-syntactic-ws)
6872 (c-with-syntax-table
6873 c-identifier-syntax-table
6874 (looking-at c-known-type-key)))
6875 (save-excursion
6876 (goto-char (1+ identifier-start))
6877 ;; We have already parsed the type earlier,
6878 ;; so it'd be possible to cache the end
6879 ;; position instead of redoing it here, but
6880 ;; then we'd need to keep track of another
6881 ;; position everywhere.
6882 (c-check-type (point)
6883 (progn (c-forward-type)
6884 (point))))))))
6885 (throw 'at-decl-or-cast t))
6886
6887 (if got-identifier
6888 (progn
a85fd6da 6889 ;; CASE 16
0386b551
AM
6890 (when (and got-prefix-before-parens
6891 at-type
6892 (or at-decl-end (looking-at "=[^=]"))
6893 (not context)
6894 (not got-suffix))
6895 ;; Got something like "foo * bar;". Since we're not inside an
6896 ;; arglist it would be a meaningless expression because the
6897 ;; result isn't used. We therefore choose to recognize it as
6898 ;; a declaration. Do not allow a suffix since it could then
6899 ;; be a function call.
6900 (throw 'at-decl-or-cast t))
6901
a85fd6da 6902 ;; CASE 17
0386b551
AM
6903 (when (and (or got-suffix-after-parens
6904 (looking-at "=[^=]"))
6905 (eq at-type 'found)
6906 (not (eq context 'arglist)))
6907 ;; Got something like "a (*b) (c);" or "a (b) = c;". It could
6908 ;; be an odd expression or it could be a declaration. Treat
6909 ;; it as a declaration if "a" has been used as a type
6910 ;; somewhere else (if it's a known type we won't get here).
6911 (throw 'at-decl-or-cast t)))
6912
a85fd6da 6913 ;; CASE 18
0386b551
AM
6914 (when (and context
6915 (or got-prefix
6916 (and (eq context 'decl)
6917 (not c-recognize-paren-inits)
6918 (or got-parens got-suffix))))
6919 ;; Got a type followed by an abstract declarator. If `got-prefix'
6920 ;; is set it's something like "a *" without anything after it. If
6921 ;; `got-parens' or `got-suffix' is set it's "a()", "a[]", "a()[]",
6922 ;; or similar, which we accept only if the context rules out
6923 ;; expressions.
6924 (throw 'at-decl-or-cast t)))
6925
6926 ;; If we had a complete symbol table here (which rules out
6927 ;; `c-found-types') we should return t due to the disambiguation rule
6928 ;; (in at least C++) that anything that can be parsed as a declaration
6929 ;; is a declaration. Now we're being more defensive and prefer to
6930 ;; highlight things like "foo (bar);" as a declaration only if we're
6931 ;; inside an arglist that contains declarations.
6932 (eq context 'decl))))
6933
6934 ;; The point is now after the type decl expression.
6935
6936 (cond
6937 ;; Check for a cast.
6938 ((save-excursion
6939 (and
6940 c-cast-parens
6941
6942 ;; Should be the first type/identifier in a cast paren.
6943 (> preceding-token-end (point-min))
6944 (memq (char-before preceding-token-end) c-cast-parens)
6945
6946 ;; The closing paren should follow.
6947 (progn
6948 (c-forward-syntactic-ws)
6949 (looking-at "\\s\)"))
6950
6951 ;; There should be a primary expression after it.
6952 (let (pos)
6953 (forward-char)
6954 (c-forward-syntactic-ws)
6955 (setq cast-end (point))
6956 (and (looking-at c-primary-expr-regexp)
6957 (progn
6958 (setq pos (match-end 0))
6959 (or
6960 ;; Check if the expression begins with a prefix keyword.
6961 (match-beginning 2)
6962 (if (match-beginning 1)
6963 ;; Expression begins with an ambiguous operator. Treat
6964 ;; it as a cast if it's a type decl or if we've
6965 ;; recognized the type somewhere else.
6966 (or at-decl-or-cast
6967 (memq at-type '(t known found)))
6968 ;; Unless it's a keyword, it's the beginning of a primary
6969 ;; expression.
6970 (not (looking-at c-keywords-regexp)))))
6971 ;; If `c-primary-expr-regexp' matched a nonsymbol token, check
6972 ;; that it matched a whole one so that we don't e.g. confuse
6973 ;; the operator '-' with '->'. It's ok if it matches further,
6974 ;; though, since it e.g. can match the float '.5' while the
6975 ;; operator regexp only matches '.'.
6976 (or (not (looking-at c-nonsymbol-token-regexp))
6977 (<= (match-end 0) pos))))
6978
6979 ;; There should either be a cast before it or something that isn't an
6980 ;; identifier or close paren.
6981 (> preceding-token-end (point-min))
6982 (progn
6983 (goto-char (1- preceding-token-end))
6984 (or (eq (point) last-cast-end)
6985 (progn
6986 (c-backward-syntactic-ws)
6987 (if (< (skip-syntax-backward "w_") 0)
6988 ;; It's a symbol. Accept it only if it's one of the
6989 ;; keywords that can precede an expression (without
6990 ;; surrounding parens).
6991 (looking-at c-simple-stmt-key)
6992 (and
6993 ;; Check that it isn't a close paren (block close is ok,
6994 ;; though).
6995 (not (memq (char-before) '(?\) ?\])))
6996 ;; Check that it isn't a nonsymbol identifier.
6997 (not (c-on-identifier)))))))))
6998
6999 ;; Handle the cast.
7000 (when (and c-record-type-identifiers at-type (not (eq at-type t)))
7001 (let ((c-promote-possible-types t))
7002 (goto-char type-start)
7003 (c-forward-type)))
7004
7005 (goto-char cast-end)
7006 'cast)
7007
7008 (at-decl-or-cast
7009 ;; We're at a declaration. Highlight the type and the following
7010 ;; declarators.
7011
7012 (when backup-if-not-cast
7013 (c-fdoc-shift-type-backward t))
7014
7015 (when (and (eq context 'decl) (looking-at ","))
7016 ;; Make sure to propagate the `c-decl-arg-start' property to
7017 ;; the next argument if it's set in this one, to cope with
7018 ;; interactive refontification.
7019 (c-put-c-type-property (point) 'c-decl-arg-start))
7020
7021 (when (and c-record-type-identifiers at-type (not (eq at-type t)))
7022 (let ((c-promote-possible-types t))
7023 (save-excursion
7024 (goto-char type-start)
7025 (c-forward-type))))
7026
e15f8aaa
AM
7027 (cons id-start
7028 (and (or at-type-decl at-typedef)
7029 (cons at-type-decl at-typedef))))
0386b551
AM
7030
7031 (t
7032 ;; False alarm. Restore the recorded ranges.
7033 (setq c-record-type-identifiers save-rec-type-ids
7034 c-record-ref-identifiers save-rec-ref-ids)
7035 nil))))
7036
7037(defun c-forward-label (&optional assume-markup preceding-token-end limit)
51c9af45 7038 ;; Assuming that point is at the beginning of a token, check if it starts a
1379f2c5
AM
7039 ;; label and if so move over it and return non-nil (t in default situations,
7040 ;; specific symbols (see below) for interesting situations), otherwise don't
7041 ;; move and return nil. "Label" here means "most things with a colon".
51c9af45
AM
7042 ;;
7043 ;; More precisely, a "label" is regarded as one of:
1379f2c5
AM
7044 ;; (i) a goto target like "foo:" - returns the symbol `goto-target';
7045 ;; (ii) A case label - either the entire construct "case FOO:", or just the
7046 ;; bare "case", should the colon be missing. We return t;
7047 ;; (iii) a keyword which needs a colon, like "default:" or "private:"; We
7048 ;; return t;
51c9af45 7049 ;; (iv) One of QT's "extended" C++ variants of
1379f2c5
AM
7050 ;; "private:"/"protected:"/"public:"/"more:" looking like "public slots:".
7051 ;; Returns the symbol `qt-2kwds-colon'.
7052 ;; (v) QT's construct "signals:". Returns the symbol `qt-1kwd-colon'.
82ba65cf 7053 ;; (vi) One of the keywords matched by `c-opt-extra-label-key' (without any
51c9af45 7054 ;; colon). Currently (2006-03), this applies only to Objective C's
1379f2c5 7055 ;; keywords "@private", "@protected", and "@public". Returns t.
51c9af45 7056 ;;
e1dbe924 7057 ;; One of the things which will NOT be recognized as a label is a bit-field
51c9af45
AM
7058 ;; element of a struct, something like "int foo:5".
7059 ;;
7060 ;; The end of the label is taken to be just after the colon, or the end of
7061 ;; the first submatch in `c-opt-extra-label-key'. The point is directly
7062 ;; after the end on return. The terminating char gets marked with
7063 ;; `c-decl-end' to improve recognition of the following declaration or
7064 ;; statement.
0386b551
AM
7065 ;;
7066 ;; If ASSUME-MARKUP is non-nil, it's assumed that the preceding
51c9af45 7067 ;; label, if any, has already been marked up like that.
0386b551
AM
7068 ;;
7069 ;; If PRECEDING-TOKEN-END is given, it should be the first position
7070 ;; after the preceding token, i.e. on the other side of the
7071 ;; syntactic ws from the point. Use a value less than or equal to
7072 ;; (point-min) if the point is at the first token in (the visible
7073 ;; part of) the buffer.
7074 ;;
7075 ;; The optional LIMIT limits the forward scan for the colon.
7076 ;;
7077 ;; This function records the ranges of the label symbols on
7078 ;; `c-record-ref-identifiers' if `c-record-type-identifiers' (!) is
7079 ;; non-nil.
7080 ;;
7081 ;; This function might do hidden buffer changes.
7082
51c9af45 7083 (let ((start (point))
1379f2c5 7084 label-end
51c9af45 7085 qt-symbol-idx
1379f2c5 7086 macro-start ; if we're in one.
f412a567
AM
7087 label-type
7088 kwd)
0386b551 7089 (cond
b414f371 7090 ;; "case" or "default" (Doesn't apply to AWK).
0386b551
AM
7091 ((looking-at c-label-kwds-regexp)
7092 (let ((kwd-end (match-end 1)))
7093 ;; Record only the keyword itself for fontification, since in
7094 ;; case labels the following is a constant expression and not
7095 ;; a label.
7096 (when c-record-type-identifiers
7097 (c-record-ref-id (cons (match-beginning 1) kwd-end)))
7098
7099 ;; Find the label end.
7100 (goto-char kwd-end)
1379f2c5
AM
7101 (setq label-type
7102 (if (and (c-syntactic-re-search-forward
7103 ;; Stop on chars that aren't allowed in expressions,
7104 ;; and on operator chars that would be meaningless
7105 ;; there. FIXME: This doesn't cope with ?: operators.
7106 "[;{=,@]\\|\\(\\=\\|[^:]\\):\\([^:]\\|\\'\\)"
7107 limit t t nil 1)
7108 (match-beginning 2))
7109
7110 (progn ; there's a proper :
7111 (goto-char (match-beginning 2)) ; just after the :
7112 (c-put-c-type-property (1- (point)) 'c-decl-end)
7113 t)
7114
7115 ;; It's an unfinished label. We consider the keyword enough
7116 ;; to recognize it as a label, so that it gets fontified.
7117 ;; Leave the point at the end of it, but don't put any
7118 ;; `c-decl-end' marker.
7119 (goto-char kwd-end)
7120 t))))
0386b551 7121
51c9af45 7122 ;; @private, @protected, @public, in Objective C, or similar.
0386b551
AM
7123 ((and c-opt-extra-label-key
7124 (looking-at c-opt-extra-label-key))
7125 ;; For a `c-opt-extra-label-key' match, we record the whole
7126 ;; thing for fontification. That's to get the leading '@' in
7127 ;; Objective-C protection labels fontified.
7128 (goto-char (match-end 1))
7129 (when c-record-type-identifiers
7130 (c-record-ref-id (cons (match-beginning 1) (point))))
7131 (c-put-c-type-property (1- (point)) 'c-decl-end)
1379f2c5 7132 (setq label-type t))
0386b551 7133
51c9af45
AM
7134 ;; All other cases of labels.
7135 ((and c-recognize-colon-labels ; nil for AWK and IDL, otherwise t.
0386b551
AM
7136
7137 ;; A colon label must have something before the colon.
7138 (not (eq (char-after) ?:))
7139
7140 ;; Check that we're not after a token that can't precede a label.
7141 (or
7142 ;; Trivially succeeds when there's no preceding token.
536610a4 7143 ;; Succeeds when we're at a virtual semicolon.
0386b551
AM
7144 (if preceding-token-end
7145 (<= preceding-token-end (point-min))
7146 (save-excursion
7147 (c-backward-syntactic-ws)
7148 (setq preceding-token-end (point))
536610a4
AM
7149 (or (bobp)
7150 (c-at-vsemi-p))))
0386b551
AM
7151
7152 ;; Check if we're after a label, if we're after a closing
7153 ;; paren that belong to statement, and with
7154 ;; `c-label-prefix-re'. It's done in different order
7155 ;; depending on `assume-markup' since the checks have
7156 ;; different expensiveness.
7157 (if assume-markup
7158 (or
7159 (eq (c-get-char-property (1- preceding-token-end) 'c-type)
7160 'c-decl-end)
7161
7162 (save-excursion
7163 (goto-char (1- preceding-token-end))
7164 (c-beginning-of-current-token)
51c9af45
AM
7165 (or (looking-at c-label-prefix-re)
7166 (looking-at c-block-stmt-1-key)))
0386b551
AM
7167
7168 (and (eq (char-before preceding-token-end) ?\))
7169 (c-after-conditional)))
7170
7171 (or
7172 (save-excursion
7173 (goto-char (1- preceding-token-end))
7174 (c-beginning-of-current-token)
51c9af45
AM
7175 (or (looking-at c-label-prefix-re)
7176 (looking-at c-block-stmt-1-key)))
0386b551
AM
7177
7178 (cond
7179 ((eq (char-before preceding-token-end) ?\))
7180 (c-after-conditional))
7181
7182 ((eq (char-before preceding-token-end) ?:)
7183 ;; Might be after another label, so check it recursively.
51c9af45
AM
7184 (save-restriction
7185 (save-excursion
7186 (goto-char (1- preceding-token-end))
7187 ;; Essentially the same as the
7188 ;; `c-syntactic-re-search-forward' regexp below.
7189 (setq macro-start
7190 (save-excursion (and (c-beginning-of-macro)
7191 (point))))
7192 (if macro-start (narrow-to-region macro-start (point-max)))
7193 (c-syntactic-skip-backward "^-]:?;}=*/%&|,<>!@+" nil t)
7194 ;; Note: the following should work instead of the
7195 ;; narrow-to-region above. Investigate why not,
7196 ;; sometime. ACM, 2006-03-31.
7197 ;; (c-syntactic-skip-backward "^-]:?;}=*/%&|,<>!@+"
7198 ;; macro-start t)
7199 (let ((pte (point))
7200 ;; If the caller turned on recording for us,
7201 ;; it shouldn't apply when we check the
7202 ;; preceding label.
7203 c-record-type-identifiers)
7204 ;; A label can't start at a cpp directive. Check for
7205 ;; this, since c-forward-syntactic-ws would foul up on it.
7206 (unless (and c-opt-cpp-prefix (looking-at c-opt-cpp-prefix))
7207 (c-forward-syntactic-ws)
7208 (c-forward-label nil pte start))))))))))
7209
1379f2c5 7210 ;; Point is still at the beginning of the possible label construct.
b414f371 7211 ;;
51c9af45
AM
7212 ;; Check that the next nonsymbol token is ":", or that we're in one
7213 ;; of QT's "slots" declarations. Allow '(' for the sake of macro
7214 ;; arguments. FIXME: Should build this regexp from the language
7215 ;; constants.
1379f2c5
AM
7216 (cond
7217 ;; public: protected: private:
7218 ((and
7219 (c-major-mode-is 'c++-mode)
7220 (search-forward-regexp
7221 "\\=p\\(r\\(ivate\\|otected\\)\\|ublic\\)\\>[^_]" nil t)
7222 (progn (backward-char)
7223 (c-forward-syntactic-ws limit)
7224 (looking-at ":\\([^:]\\|\\'\\)"))) ; A single colon.
7225 (forward-char)
7226 (setq label-type t))
7227 ;; QT double keyword like "protected slots:" or goto target.
7228 ((progn (goto-char start) nil))
7229 ((when (c-syntactic-re-search-forward
7230 "[ \t\n[:?;{=*/%&|,<>!@+-]" limit t t) ; not at EOB
7231 (backward-char)
7232 (setq label-end (point))
7233 (setq qt-symbol-idx
7234 (and (c-major-mode-is 'c++-mode)
7235 (string-match
7236 "\\(p\\(r\\(ivate\\|otected\\)\\|ublic\\)\\|more\\)\\>"
b414f371 7237 (buffer-substring start (point)))))
1379f2c5
AM
7238 (c-forward-syntactic-ws limit)
7239 (cond
7240 ((looking-at ":\\([^:]\\|\\'\\)") ; A single colon.
7241 (forward-char)
7242 (setq label-type
f412a567
AM
7243 (if (or (string= "signals" ; Special QT macro
7244 (setq kwd (buffer-substring-no-properties start label-end)))
7245 (string= "Q_SIGNALS" kwd))
1379f2c5
AM
7246 'qt-1kwd-colon
7247 'goto-target)))
7248 ((and qt-symbol-idx
f412a567 7249 (search-forward-regexp "\\=\\(slots\\|Q_SLOTS\\)\\>" limit t)
1379f2c5
AM
7250 (progn (c-forward-syntactic-ws limit)
7251 (looking-at ":\\([^:]\\|\\'\\)"))) ; A single colon
7252 (forward-char)
7253 (setq label-type 'qt-2kwds-colon)))))))
0386b551
AM
7254
7255 (save-restriction
7256 (narrow-to-region start (point))
7257
7258 ;; Check that `c-nonlabel-token-key' doesn't match anywhere.
7259 (catch 'check-label
7260 (goto-char start)
7261 (while (progn
7262 (when (looking-at c-nonlabel-token-key)
7263 (goto-char start)
1379f2c5 7264 (setq label-type nil)
0386b551
AM
7265 (throw 'check-label nil))
7266 (and (c-safe (c-forward-sexp)
7267 (c-forward-syntactic-ws)
7268 t)
7269 (not (eobp)))))
7270
7271 ;; Record the identifiers in the label for fontification, unless
7272 ;; it begins with `c-label-kwds' in which case the following
7273 ;; identifiers are part of a (constant) expression that
7274 ;; shouldn't be fontified.
7275 (when (and c-record-type-identifiers
7276 (progn (goto-char start)
7277 (not (looking-at c-label-kwds-regexp))))
7278 (while (c-syntactic-re-search-forward c-symbol-key nil t)
7279 (c-record-ref-id (cons (match-beginning 0)
7280 (match-end 0)))))
7281
7282 (c-put-c-type-property (1- (point-max)) 'c-decl-end)
82ba65cf 7283 (goto-char (point-max)))))
0386b551
AM
7284
7285 (t
7286 ;; Not a label.
1379f2c5
AM
7287 (goto-char start)))
7288 label-type))
0386b551
AM
7289
7290(defun c-forward-objc-directive ()
7291 ;; Assuming the point is at the beginning of a token, try to move
7292 ;; forward to the end of the Objective-C directive that starts
7293 ;; there. Return t if a directive was fully recognized, otherwise
7294 ;; the point is moved as far as one could be successfully parsed and
7295 ;; nil is returned.
7296 ;;
7297 ;; This function records identifier ranges on
7298 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
7299 ;; `c-record-type-identifiers' is non-nil.
7300 ;;
7301 ;; This function might do hidden buffer changes.
7302
7303 (let ((start (point))
7304 start-char
7305 (c-promote-possible-types t)
7306 ;; Turn off recognition of angle bracket arglists while parsing
7307 ;; types here since the protocol reference list might then be
7308 ;; considered part of the preceding name or superclass-name.
7309 c-recognize-<>-arglists)
7310
7311 (if (or
7312 (when (looking-at
7313 (eval-when-compile
7314 (c-make-keywords-re t
7315 (append (c-lang-const c-protection-kwds objc)
7316 '("@end"))
7317 'objc-mode)))
7318 (goto-char (match-end 1))
7319 t)
7320
7321 (and
7322 (looking-at
7323 (eval-when-compile
7324 (c-make-keywords-re t
7325 '("@interface" "@implementation" "@protocol")
7326 'objc-mode)))
7327
7328 ;; Handle the name of the class itself.
7329 (progn
cb694ab7
AM
7330; (c-forward-token-2) ; 2006/1/13 This doesn't move if the token's
7331; at EOB.
7332 (goto-char (match-end 0))
7333 (c-skip-ws-forward)
0386b551
AM
7334 (c-forward-type))
7335
7336 (catch 'break
7337 ;; Look for ": superclass-name" or "( category-name )".
7338 (when (looking-at "[:\(]")
7339 (setq start-char (char-after))
7340 (forward-char)
7341 (c-forward-syntactic-ws)
7342 (unless (c-forward-type) (throw 'break nil))
7343 (when (eq start-char ?\()
7344 (unless (eq (char-after) ?\)) (throw 'break nil))
7345 (forward-char)
7346 (c-forward-syntactic-ws)))
7347
7348 ;; Look for a protocol reference list.
7349 (if (eq (char-after) ?<)
7350 (let ((c-recognize-<>-arglists t)
7351 (c-parse-and-markup-<>-arglists t)
7352 c-restricted-<>-arglists)
7353 (c-forward-<>-arglist t))
7354 t))))
7355
7356 (progn
7357 (c-backward-syntactic-ws)
7358 (c-clear-c-type-property start (1- (point)) 'c-decl-end)
7359 (c-put-c-type-property (1- (point)) 'c-decl-end)
7360 t)
7361
7362 (c-clear-c-type-property start (point) 'c-decl-end)
7363 nil)))
7364
785eecbb
RS
7365(defun c-beginning-of-inheritance-list (&optional lim)
7366 ;; Go to the first non-whitespace after the colon that starts a
7367 ;; multiple inheritance introduction. Optional LIM is the farthest
7368 ;; back we should search.
0386b551
AM
7369 ;;
7370 ;; This function might do hidden buffer changes.
7371 (c-with-syntax-table c++-template-syntax-table
7372 (c-backward-token-2 0 t lim)
7373 (while (and (or (looking-at c-symbol-start)
7374 (looking-at "[<,]\\|::"))
7375 (zerop (c-backward-token-2 1 t lim))))))
785eecbb 7376
785eecbb
RS
7377(defun c-in-method-def-p ()
7378 ;; Return nil if we aren't in a method definition, otherwise the
7379 ;; position of the initial [+-].
0386b551
AM
7380 ;;
7381 ;; This function might do hidden buffer changes.
785eecbb
RS
7382 (save-excursion
7383 (beginning-of-line)
a66cd3ee
MS
7384 (and c-opt-method-key
7385 (looking-at c-opt-method-key)
785eecbb
RS
7386 (point))
7387 ))
7388
a66cd3ee
MS
7389;; Contributed by Kevin Ryde <user42@zip.com.au>.
7390(defun c-in-gcc-asm-p ()
7391 ;; Return non-nil if point is within a gcc \"asm\" block.
7392 ;;
7393 ;; This should be called with point inside an argument list.
7394 ;;
7395 ;; Only one level of enclosing parentheses is considered, so for
7396 ;; instance `nil' is returned when in a function call within an asm
7397 ;; operand.
0386b551
AM
7398 ;;
7399 ;; This function might do hidden buffer changes.
a66cd3ee
MS
7400
7401 (and c-opt-asm-stmt-key
7402 (save-excursion
7403 (beginning-of-line)
7404 (backward-up-list 1)
7405 (c-beginning-of-statement-1 (point-min) nil t)
7406 (looking-at c-opt-asm-stmt-key))))
7407
abb7e5cf 7408(defun c-at-toplevel-p ()
a85fd6da
AM
7409 "Return a determination as to whether point is \"at the top level\".
7410Informally, \"at the top level\" is anywhere where you can write
7411a function.
7412
7413More precisely, being at the top-level means that point is either
7414outside any enclosing block (such as a function definition), or
7415directly inside a class, namespace or other block that contains
7416another declaration level.
abb7e5cf
SM
7417
7418If point is not at the top-level (e.g. it is inside a method
7419definition), then nil is returned. Otherwise, if point is at a
7420top-level not enclosed within a class definition, t is returned.
7421Otherwise, a 2-vector is returned where the zeroth element is the
7422buffer position of the start of the class declaration, and the first
7423element is the buffer position of the enclosing class's opening
0386b551
AM
7424brace.
7425
7426Note that this function might do hidden buffer changes. See the
7427comment at the start of cc-engine.el for more info."
a66cd3ee
MS
7428 (let ((paren-state (c-parse-state)))
7429 (or (not (c-most-enclosing-brace paren-state))
7430 (c-search-uplist-for-classkey paren-state))))
7431
d9e94c22 7432(defun c-just-after-func-arglist-p (&optional lim)
0386b551
AM
7433 ;; Return non-nil if the point is in the region after the argument
7434 ;; list of a function and its opening brace (or semicolon in case it
7435 ;; got no body). If there are K&R style argument declarations in
7436 ;; that region, the point has to be inside the first one for this
7437 ;; function to recognize it.
a66cd3ee 7438 ;;
0386b551
AM
7439 ;; If successful, the point is moved to the first token after the
7440 ;; function header (see `c-forward-decl-or-cast-1' for details) and
7441 ;; the position of the opening paren of the function arglist is
7442 ;; returned.
7443 ;;
7444 ;; The point is clobbered if not successful.
7445 ;;
7446 ;; LIM is used as bound for backward buffer searches.
7447 ;;
7448 ;; This function might do hidden buffer changes.
7449
7450 (let ((beg (point)) end id-start)
7451 (and
7452 (eq (c-beginning-of-statement-1 lim) 'same)
7453
7454 (not (or (c-major-mode-is 'objc-mode)
7455 (c-forward-objc-directive)))
7456
7457 (setq id-start
7458 (car-safe (c-forward-decl-or-cast-1 (c-point 'bosws) nil nil)))
7459 (< id-start beg)
7460
7461 ;; There should not be a '=' or ',' between beg and the
7462 ;; start of the declaration since that means we were in the
7463 ;; "expression part" of the declaration.
7464 (or (> (point) beg)
7465 (not (looking-at "[=,]")))
7466
7467 (save-excursion
7468 ;; Check that there's an arglist paren in the
7469 ;; declaration.
7470 (goto-char id-start)
7471 (cond ((eq (char-after) ?\()
7472 ;; The declarator is a paren expression, so skip past it
7473 ;; so that we don't get stuck on that instead of the
7474 ;; function arglist.
7475 (c-forward-sexp))
51c9af45
AM
7476 ((and c-opt-op-identifier-prefix
7477 (looking-at c-opt-op-identifier-prefix))
0386b551
AM
7478 ;; Don't trip up on "operator ()".
7479 (c-forward-token-2 2 t)))
7480 (and (< (point) beg)
7481 (c-syntactic-re-search-forward "(" beg t t)
7482 (1- (point)))))))
785eecbb 7483
a66cd3ee
MS
7484(defun c-in-knr-argdecl (&optional lim)
7485 ;; Return the position of the first argument declaration if point is
7486 ;; inside a K&R style argument declaration list, nil otherwise.
7487 ;; `c-recognize-knr-p' is not checked. If LIM is non-nil, it's a
7488 ;; position that bounds the backward search for the argument list.
7489 ;;
9cf17ef1
AM
7490 ;; Point must be within a possible K&R region, e.g. just before a top-level
7491 ;; "{". It must be outside of parens and brackets. The test can return
7492 ;; false positives otherwise.
0386b551
AM
7493 ;;
7494 ;; This function might do hidden buffer changes.
d9e94c22 7495
a66cd3ee
MS
7496 (save-excursion
7497 (save-restriction
9cf17ef1
AM
7498 ;; If we're in a macro, our search range is restricted to it. Narrow to
7499 ;; the searchable range.
7500 (let* ((macro-start (c-query-macro-start))
7501 (lim (max (or lim (point-min)) (or macro-start (point-min))))
6b6481ed
AM
7502 before-lparen after-rparen
7503 (pp-count-out 20)) ; Max number of paren/brace constructs before we give up
9cf17ef1
AM
7504 (narrow-to-region lim (c-point 'eol))
7505
7506 ;; Search backwards for the defun's argument list. We give up if we
7507 ;; encounter a "}" (end of a previous defun) or BOB.
7508 ;;
7509 ;; The criterion for a paren structure being the arg list is:
7510 ;; o - there is non-WS stuff after it but before any "{"; AND
7511 ;; o - the token after it isn't a ";" AND
7512 ;; o - it is preceded by either an identifier (the function name) or
7513 ;; a macro expansion like "DEFUN (...)"; AND
7514 ;; o - its content is a non-empty comma-separated list of identifiers
7515 ;; (an empty arg list won't have a knr region).
7516 ;;
7517 ;; The following snippet illustrates these rules:
7518 ;; int foo (bar, baz, yuk)
7519 ;; int bar [] ;
7520 ;; int (*baz) (my_type) ;
7521 ;; int (*) (void) (*yuk) (void) ;
7522 ;; {
7523
7524 (catch 'knr
6b6481ed
AM
7525 (while (> pp-count-out 0) ; go back one paren/bracket pair each time.
7526 (setq pp-count-out (1- pp-count-out))
10489fcb 7527 (c-syntactic-skip-backward "^)]}")
9cf17ef1
AM
7528 (cond ((eq (char-before) ?\))
7529 (setq after-rparen (point)))
10489fcb
AM
7530 ((eq (char-before) ?\])
7531 (setq after-rparen nil))
7532 (t ; either } (hit previous defun) or no more parens/brackets
7533 (throw 'knr nil)))
9cf17ef1
AM
7534
7535 (if after-rparen
7536 ;; We're inside a paren. Could it be our argument list....?
7537 (if
7538 (and
7539 (progn
7540 (goto-char after-rparen)
7541 (unless (c-go-list-backward) (throw 'knr nil)) ;
7542 ;; FIXME!!! What about macros between the parens? 2007/01/20
7543 (setq before-lparen (point)))
d9e94c22 7544
9cf17ef1
AM
7545 ;; It can't be the arg list if next token is ; or {
7546 (progn (goto-char after-rparen)
7547 (c-forward-syntactic-ws)
7548 (not (memq (char-after) '(?\; ?\{))))
d9e94c22 7549
9cf17ef1
AM
7550 ;; Is the thing preceding the list an identifier (the
7551 ;; function name), or a macro expansion?
7552 (progn
7553 (goto-char before-lparen)
7554 (eq (c-backward-token-2) 0)
7555 (or (c-on-identifier)
7556 (and (eq (char-after) ?\))
7557 (c-go-up-list-backward)
7558 (eq (c-backward-token-2) 0)
7559 (c-on-identifier))))
7560
7561 ;; Have we got a non-empty list of comma-separated
7562 ;; identifiers?
7563 (progn
7564 (goto-char before-lparen)
7565 (c-forward-token-2) ; to first token inside parens
7566 (and
7567 (c-on-identifier)
7568 (c-forward-token-2)
7569 (catch 'id-list
7570 (while (eq (char-after) ?\,)
7571 (c-forward-token-2)
7572 (unless (c-on-identifier) (throw 'id-list nil))
7573 (c-forward-token-2))
7574 (eq (char-after) ?\))))))
7575
7576 ;; ...Yes. We've identified the function's argument list.
7577 (throw 'knr
7578 (progn (goto-char after-rparen)
7579 (c-forward-syntactic-ws)
7580 (point)))
7581
7582 ;; ...No. The current parens aren't the function's arg list.
7583 (goto-char before-lparen))
7584
7585 (or (c-go-list-backward) ; backwards over [ .... ]
7586 (throw 'knr nil)))))))))
785eecbb
RS
7587
7588(defun c-skip-conditional ()
7589 ;; skip forward over conditional at point, including any predicate
7590 ;; statements in parentheses. No error checking is performed.
0386b551
AM
7591 ;;
7592 ;; This function might do hidden buffer changes.
0ec8351b
BW
7593 (c-forward-sexp (cond
7594 ;; else if()
a66cd3ee
MS
7595 ((looking-at (concat "\\<else"
7596 "\\([ \t\n]\\|\\\\\n\\)+"
7597 "if\\>\\([^_]\\|$\\)"))
7598 3)
0ec8351b 7599 ;; do, else, try, finally
a66cd3ee
MS
7600 ((looking-at (concat "\\<\\("
7601 "do\\|else\\|try\\|finally"
7602 "\\)\\>\\([^_]\\|$\\)"))
130c507e 7603 1)
ce8c7486 7604 ;; for, if, while, switch, catch, synchronized, foreach
0ec8351b 7605 (t 2))))
785eecbb 7606
a66cd3ee
MS
7607(defun c-after-conditional (&optional lim)
7608 ;; If looking at the token after a conditional then return the
7609 ;; position of its start, otherwise return nil.
0386b551
AM
7610 ;;
7611 ;; This function might do hidden buffer changes.
a66cd3ee 7612 (save-excursion
d9e94c22 7613 (and (zerop (c-backward-token-2 1 t lim))
a66cd3ee
MS
7614 (or (looking-at c-block-stmt-1-key)
7615 (and (eq (char-after) ?\()
d9e94c22 7616 (zerop (c-backward-token-2 1 t lim))
a66cd3ee
MS
7617 (looking-at c-block-stmt-2-key)))
7618 (point))))
7619
0386b551
AM
7620(defun c-after-special-operator-id (&optional lim)
7621 ;; If the point is after an operator identifier that isn't handled
7622 ;; like an ordinary symbol (i.e. like "operator =" in C++) then the
7623 ;; position of the start of that identifier is returned. nil is
7624 ;; returned otherwise. The point may be anywhere in the syntactic
7625 ;; whitespace after the last token of the operator identifier.
7626 ;;
7627 ;; This function might do hidden buffer changes.
7628 (save-excursion
7629 (and c-overloadable-operators-regexp
7630 (zerop (c-backward-token-2 1 nil lim))
7631 (looking-at c-overloadable-operators-regexp)
51c9af45 7632 (or (not c-opt-op-identifier-prefix)
0386b551
AM
7633 (and
7634 (zerop (c-backward-token-2 1 nil lim))
51c9af45 7635 (looking-at c-opt-op-identifier-prefix)))
0386b551
AM
7636 (point))))
7637
a66cd3ee
MS
7638(defsubst c-backward-to-block-anchor (&optional lim)
7639 ;; Assuming point is at a brace that opens a statement block of some
7640 ;; kind, move to the proper anchor point for that block. It might
7641 ;; need to be adjusted further by c-add-stmt-syntax, but the
7642 ;; position at return is suitable as start position for that
7643 ;; function.
0386b551
AM
7644 ;;
7645 ;; This function might do hidden buffer changes.
a66cd3ee
MS
7646 (unless (= (point) (c-point 'boi))
7647 (let ((start (c-after-conditional lim)))
7648 (if start
7649 (goto-char start)))))
7650
037558bf 7651(defsubst c-backward-to-decl-anchor (&optional lim)
a66cd3ee
MS
7652 ;; Assuming point is at a brace that opens the block of a top level
7653 ;; declaration of some kind, move to the proper anchor point for
7654 ;; that block.
0386b551
AM
7655 ;;
7656 ;; This function might do hidden buffer changes.
a66cd3ee 7657 (unless (= (point) (c-point 'boi))
037558bf 7658 (c-beginning-of-statement-1 lim)))
a66cd3ee 7659
ff959bab 7660(defun c-search-decl-header-end ()
a66cd3ee
MS
7661 ;; Search forward for the end of the "header" of the current
7662 ;; declaration. That's the position where the definition body
7663 ;; starts, or the first variable initializer, or the ending
7664 ;; semicolon. I.e. search forward for the closest following
7665 ;; (syntactically relevant) '{', '=' or ';' token. Point is left
7666 ;; _after_ the first found token, or at point-max if none is found.
0386b551
AM
7667 ;;
7668 ;; This function might do hidden buffer changes.
ff959bab
MS
7669
7670 (let ((base (point)))
7671 (if (c-major-mode-is 'c++-mode)
7672
7673 ;; In C++ we need to take special care to handle operator
7674 ;; tokens and those pesky template brackets.
7675 (while (and
7676 (c-syntactic-re-search-forward "[;{<=]" nil 'move t t)
7677 (or
7678 (c-end-of-current-token base)
7679 ;; Handle operator identifiers, i.e. ignore any
7680 ;; operator token preceded by "operator".
7681 (save-excursion
7682 (and (c-safe (c-backward-sexp) t)
51c9af45 7683 (looking-at c-opt-op-identifier-prefix)))
ff959bab
MS
7684 (and (eq (char-before) ?<)
7685 (c-with-syntax-table c++-template-syntax-table
7686 (if (c-safe (goto-char (c-up-list-forward (point))))
7687 t
7688 (goto-char (point-max))
7689 nil)))))
7690 (setq base (point)))
7691
7692 (while (and
7693 (c-syntactic-re-search-forward "[;{=]" nil 'move t t)
7694 (c-end-of-current-token base))
7695 (setq base (point))))))
a66cd3ee
MS
7696
7697(defun c-beginning-of-decl-1 (&optional lim)
7698 ;; Go to the beginning of the current declaration, or the beginning
7699 ;; of the previous one if already at the start of it. Point won't
0386b551 7700 ;; be moved out of any surrounding paren. Return a cons cell of the
a66cd3ee
MS
7701 ;; form (MOVE . KNR-POS). MOVE is like the return value from
7702 ;; `c-beginning-of-statement-1'. If point skipped over some K&R
7703 ;; style argument declarations (and they are to be recognized) then
7704 ;; KNR-POS is set to the start of the first such argument
7705 ;; declaration, otherwise KNR-POS is nil. If LIM is non-nil, it's a
7706 ;; position that bounds the backward search.
7707 ;;
7708 ;; NB: Cases where the declaration continues after the block, as in
7709 ;; "struct foo { ... } bar;", are currently recognized as two
7710 ;; declarations, e.g. "struct foo { ... }" and "bar;" in this case.
0386b551
AM
7711 ;;
7712 ;; This function might do hidden buffer changes.
a66cd3ee
MS
7713 (catch 'return
7714 (let* ((start (point))
d9e94c22 7715 (last-stmt-start (point))
0386b551 7716 (move (c-beginning-of-statement-1 lim nil t)))
a66cd3ee 7717
a66cd3ee
MS
7718 ;; `c-beginning-of-statement-1' stops at a block start, but we
7719 ;; want to continue if the block doesn't begin a top level
2a15eb73
MS
7720 ;; construct, i.e. if it isn't preceded by ';', '}', ':', bob,
7721 ;; or an open paren.
d9e94c22 7722 (let ((beg (point)) tentative-move)
51c9af45
AM
7723 ;; Go back one "statement" each time round the loop until we're just
7724 ;; after a ;, }, or :, or at BOB or the start of a macro or start of
7725 ;; an ObjC method. This will move over a multiple declaration whose
7726 ;; components are comma separated.
d9e94c22
MS
7727 (while (and
7728 ;; Must check with c-opt-method-key in ObjC mode.
7729 (not (and c-opt-method-key
7730 (looking-at c-opt-method-key)))
7731 (/= last-stmt-start (point))
7732 (progn
7733 (c-backward-syntactic-ws lim)
7734 (not (memq (char-before) '(?\; ?} ?: nil))))
2a15eb73
MS
7735 (save-excursion
7736 (backward-char)
7737 (not (looking-at "\\s(")))
d9e94c22
MS
7738 ;; Check that we don't move from the first thing in a
7739 ;; macro to its header.
7740 (not (eq (setq tentative-move
0386b551 7741 (c-beginning-of-statement-1 lim nil t))
d9e94c22
MS
7742 'macro)))
7743 (setq last-stmt-start beg
7744 beg (point)
7745 move tentative-move))
7746 (goto-char beg))
7747
7748 (when c-recognize-knr-p
7749 (let ((fallback-pos (point)) knr-argdecl-start)
7750 ;; Handle K&R argdecls. Back up after the "statement" jumped
7751 ;; over by `c-beginning-of-statement-1', unless it was the
7752 ;; function body, in which case we're sitting on the opening
7753 ;; brace now. Then test if we're in a K&R argdecl region and
7754 ;; that we started at the other side of the first argdecl in
7755 ;; it.
7756 (unless (eq (char-after) ?{)
7757 (goto-char last-stmt-start))
7758 (if (and (setq knr-argdecl-start (c-in-knr-argdecl lim))
7759 (< knr-argdecl-start start)
7760 (progn
7761 (goto-char knr-argdecl-start)
0386b551 7762 (not (eq (c-beginning-of-statement-1 lim nil t) 'macro))))
d9e94c22
MS
7763 (throw 'return
7764 (cons (if (eq (char-after fallback-pos) ?{)
7765 'previous
7766 'same)
7767 knr-argdecl-start))
7768 (goto-char fallback-pos))))
7769
51c9af45
AM
7770 ;; `c-beginning-of-statement-1' counts each brace block as a separate
7771 ;; statement, so the result will be 'previous if we've moved over any.
7772 ;; So change our result back to 'same if necessary.
7773 ;;
7774 ;; If they were brace list initializers we might not have moved over a
7775 ;; declaration boundary though, so change it to 'same if we've moved
7776 ;; past a '=' before '{', but not ';'. (This ought to be integrated
7777 ;; into `c-beginning-of-statement-1', so we avoid this extra pass which
7778 ;; potentially can search over a large amount of text.). Take special
7779 ;; pains not to get mislead by C++'s "operator=", and the like.
d9e94c22
MS
7780 (if (and (eq move 'previous)
7781 (c-with-syntax-table (if (c-major-mode-is 'c++-mode)
7782 c++-template-syntax-table
7783 (syntax-table))
7784 (save-excursion
51c9af45
AM
7785 (and
7786 (progn
7787 (while ; keep going back to "[;={"s until we either find
7788 ; no more, or get to one which isn't an "operator ="
7789 (and (c-syntactic-re-search-forward "[;={]" start t t t)
7790 (eq (char-before) ?=)
7791 c-overloadable-operators-regexp
7792 c-opt-op-identifier-prefix
7793 (save-excursion
7794 (eq (c-backward-token-2) 0)
7795 (looking-at c-overloadable-operators-regexp)
7796 (eq (c-backward-token-2) 0)
7797 (looking-at c-opt-op-identifier-prefix))))
7798 (eq (char-before) ?=))
7799 (c-syntactic-re-search-forward "[;{]" start t t)
7800 (eq (char-before) ?{)
7801 (c-safe (goto-char (c-up-list-forward (point))) t)
7802 (not (c-syntactic-re-search-forward ";" start t t))))))
d9e94c22
MS
7803 (cons 'same nil)
7804 (cons move nil)))))
a66cd3ee
MS
7805
7806(defun c-end-of-decl-1 ()
7807 ;; Assuming point is at the start of a declaration (as detected by
7808 ;; e.g. `c-beginning-of-decl-1'), go to the end of it. Unlike
7809 ;; `c-beginning-of-decl-1', this function handles the case when a
7810 ;; block is followed by identifiers in e.g. struct declarations in C
7811 ;; or C++. If a proper end was found then t is returned, otherwise
7812 ;; point is moved as far as possible within the current sexp and nil
7813 ;; is returned. This function doesn't handle macros; use
7814 ;; `c-end-of-macro' instead in those cases.
0386b551
AM
7815 ;;
7816 ;; This function might do hidden buffer changes.
ce8c7486 7817 (let ((start (point))
a66cd3ee
MS
7818 (decl-syntax-table (if (c-major-mode-is 'c++-mode)
7819 c++-template-syntax-table
7820 (syntax-table))))
7821 (catch 'return
7822 (c-search-decl-header-end)
7823
7824 (when (and c-recognize-knr-p
7825 (eq (char-before) ?\;)
7826 (c-in-knr-argdecl start))
7827 ;; Stopped at the ';' in a K&R argdecl section which is
7828 ;; detected using the same criteria as in
7829 ;; `c-beginning-of-decl-1'. Move to the following block
7830 ;; start.
d9e94c22 7831 (c-syntactic-re-search-forward "{" nil 'move t))
a66cd3ee
MS
7832
7833 (when (eq (char-before) ?{)
7834 ;; Encountered a block in the declaration. Jump over it.
7835 (condition-case nil
7836 (goto-char (c-up-list-forward (point)))
d9e94c22
MS
7837 (error (goto-char (point-max))
7838 (throw 'return nil)))
a66cd3ee
MS
7839 (if (or (not c-opt-block-decls-with-vars-key)
7840 (save-excursion
7841 (c-with-syntax-table decl-syntax-table
7842 (let ((lim (point)))
7843 (goto-char start)
b3cf7e18
MS
7844 (not (and
7845 ;; Check for `c-opt-block-decls-with-vars-key'
7846 ;; before the first paren.
7847 (c-syntactic-re-search-forward
d9e94c22 7848 (concat "[;=\(\[{]\\|\\("
b3cf7e18
MS
7849 c-opt-block-decls-with-vars-key
7850 "\\)")
d9e94c22 7851 lim t t t)
b3cf7e18
MS
7852 (match-beginning 1)
7853 (not (eq (char-before) ?_))
d9e94c22
MS
7854 ;; Check that the first following paren is
7855 ;; the block.
7856 (c-syntactic-re-search-forward "[;=\(\[{]"
7857 lim t t t)
b3cf7e18 7858 (eq (char-before) ?{)))))))
a66cd3ee
MS
7859 ;; The declaration doesn't have any of the
7860 ;; `c-opt-block-decls-with-vars' keywords in the
7861 ;; beginning, so it ends here at the end of the block.
7862 (throw 'return t)))
7863
7864 (c-with-syntax-table decl-syntax-table
7865 (while (progn
7866 (if (eq (char-before) ?\;)
7867 (throw 'return t))
d9e94c22 7868 (c-syntactic-re-search-forward ";" nil 'move t))))
a66cd3ee 7869 nil)))
ce8c7486 7870
0386b551
AM
7871(defun c-looking-at-decl-block (containing-sexp goto-start &optional limit)
7872 ;; Assuming the point is at an open brace, check if it starts a
7873 ;; block that contains another declaration level, i.e. that isn't a
7874 ;; statement block or a brace list, and if so return non-nil.
7875 ;;
7876 ;; If the check is successful, the return value is the start of the
7877 ;; keyword that tells what kind of construct it is, i.e. typically
7878 ;; what `c-decl-block-key' matched. Also, if GOTO-START is set then
7879 ;; the point will be at the start of the construct, before any
7880 ;; leading specifiers, otherwise it's at the returned position.
7881 ;;
7882 ;; The point is clobbered if the check is unsuccessful.
7883 ;;
7884 ;; CONTAINING-SEXP is the position of the open of the surrounding
7885 ;; paren, or nil if none.
7886 ;;
7887 ;; The optional LIMIT limits the backward search for the start of
7888 ;; the construct. It's assumed to be at a syntactically relevant
7889 ;; position.
7890 ;;
7891 ;; If any template arglists are found in the searched region before
7892 ;; the open brace, they get marked with paren syntax.
7893 ;;
7894 ;; This function might do hidden buffer changes.
7895
7896 (let ((open-brace (point)) kwd-start first-specifier-pos)
7897 (c-syntactic-skip-backward c-block-prefix-charset limit t)
7898
7899 (when (and c-recognize-<>-arglists
7900 (eq (char-before) ?>))
7901 ;; Could be at the end of a template arglist.
7902 (let ((c-parse-and-markup-<>-arglists t)
7903 (c-disallow-comma-in-<>-arglists
7904 (and containing-sexp
7905 (not (eq (char-after containing-sexp) ?{)))))
7906 (while (and
7907 (c-backward-<>-arglist nil limit)
7908 (progn
7909 (c-syntactic-skip-backward c-block-prefix-charset limit t)
7910 (eq (char-before) ?>))))))
7911
7912 ;; Note: Can't get bogus hits inside template arglists below since they
7913 ;; have gotten paren syntax above.
7914 (when (and
7915 ;; If `goto-start' is set we begin by searching for the
7916 ;; first possible position of a leading specifier list.
7917 ;; The `c-decl-block-key' search continues from there since
7918 ;; we know it can't match earlier.
7919 (if goto-start
7920 (when (c-syntactic-re-search-forward c-symbol-start
7921 open-brace t t)
7922 (goto-char (setq first-specifier-pos (match-beginning 0)))
7923 t)
7924 t)
7925
7926 (cond
7927 ((c-syntactic-re-search-forward c-decl-block-key open-brace t t t)
7928 (goto-char (setq kwd-start (match-beginning 0)))
7929 (or
7930
7931 ;; Found a keyword that can't be a type?
7932 (match-beginning 1)
7933
7934 ;; Can be a type too, in which case it's the return type of a
7935 ;; function (under the assumption that no declaration level
7936 ;; block construct starts with a type).
7937 (not (c-forward-type))
7938
7939 ;; Jumped over a type, but it could be a declaration keyword
7940 ;; followed by the declared identifier that we've jumped over
7941 ;; instead (e.g. in "class Foo {"). If it indeed is a type
7942 ;; then we should be at the declarator now, so check for a
7943 ;; valid declarator start.
7944 ;;
7945 ;; Note: This doesn't cope with the case when a declared
7946 ;; identifier is followed by e.g. '(' in a language where '('
7947 ;; also might be part of a declarator expression. Currently
7948 ;; there's no such language.
7949 (not (or (looking-at c-symbol-start)
7950 (looking-at c-type-decl-prefix-key)))))
7951
7952 ;; In Pike a list of modifiers may be followed by a brace
7953 ;; to make them apply to many identifiers. Note that the
7954 ;; match data will be empty on return in this case.
7955 ((and (c-major-mode-is 'pike-mode)
7956 (progn
7957 (goto-char open-brace)
7958 (= (c-backward-token-2) 0))
7959 (looking-at c-specifier-key)
7960 ;; Use this variant to avoid yet another special regexp.
7961 (c-keyword-member (c-keyword-sym (match-string 1))
7962 'c-modifier-kwds))
7963 (setq kwd-start (point))
7964 t)))
7965
7966 ;; Got a match.
7967
7968 (if goto-start
7969 ;; Back up over any preceding specifiers and their clauses
7970 ;; by going forward from `first-specifier-pos', which is the
7971 ;; earliest possible position where the specifier list can
7972 ;; start.
7973 (progn
7974 (goto-char first-specifier-pos)
7975
7976 (while (< (point) kwd-start)
7977 (if (looking-at c-symbol-key)
7978 ;; Accept any plain symbol token on the ground that
7979 ;; it's a specifier masked through a macro (just
7980 ;; like `c-forward-decl-or-cast-1' skip forward over
7981 ;; such tokens).
7982 ;;
7983 ;; Could be more restrictive wrt invalid keywords,
7984 ;; but that'd only occur in invalid code so there's
7985 ;; no use spending effort on it.
7986 (let ((end (match-end 0)))
7987 (unless (c-forward-keyword-clause 0)
7988 (goto-char end)
7989 (c-forward-syntactic-ws)))
7990
7991 ;; Can't parse a declaration preamble and is still
7992 ;; before `kwd-start'. That means `first-specifier-pos'
7993 ;; was in some earlier construct. Search again.
7994 (if (c-syntactic-re-search-forward c-symbol-start
7995 kwd-start 'move t)
7996 (goto-char (setq first-specifier-pos (match-beginning 0)))
7997 ;; Got no preamble before the block declaration keyword.
7998 (setq first-specifier-pos kwd-start))))
7999
8000 (goto-char first-specifier-pos))
8001 (goto-char kwd-start))
8002
8003 kwd-start)))
ce8c7486 8004
a66cd3ee 8005(defun c-search-uplist-for-classkey (paren-state)
0386b551
AM
8006 ;; Check if the closest containing paren sexp is a declaration
8007 ;; block, returning a 2 element vector in that case. Aref 0
8008 ;; contains the bufpos at boi of the class key line, and aref 1
8009 ;; contains the bufpos of the open brace. This function is an
8010 ;; obsolete wrapper for `c-looking-at-decl-block'.
8011 ;;
8012 ;; This function might do hidden buffer changes.
8013 (let ((open-paren-pos (c-most-enclosing-brace paren-state)))
8014 (when open-paren-pos
8015 (save-excursion
8016 (goto-char open-paren-pos)
8017 (when (and (eq (char-after) ?{)
8018 (c-looking-at-decl-block
8019 (c-safe-position open-paren-pos paren-state)
8020 nil))
8021 (back-to-indentation)
8022 (vector (point) open-paren-pos))))))
785eecbb 8023
020716e1
AM
8024(defmacro c-pull-open-brace (ps)
8025 ;; Pull the next open brace from PS (which has the form of paren-state),
8026 ;; skipping over any brace pairs. Returns NIL when PS is exhausted.
8027 `(progn
8028 (while (consp (car ,ps))
8029 (setq ,ps (cdr ,ps)))
8030 (prog1 (car ,ps)
8031 (setq ,ps (cdr ,ps)))))
8032
8033(defun c-most-enclosing-decl-block (paren-state)
8034 ;; Return the buffer position of the most enclosing decl-block brace (in the
8035 ;; sense of c-looking-at-decl-block) in the PAREN-STATE structure, or nil if
8036 ;; none was found.
8037 (let* ((open-brace (c-pull-open-brace paren-state))
8038 (next-open-brace (c-pull-open-brace paren-state)))
8039 (while (and open-brace
8040 (save-excursion
8041 (goto-char open-brace)
8042 (not (c-looking-at-decl-block next-open-brace nil))))
8043 (setq open-brace next-open-brace
8044 next-open-brace (c-pull-open-brace paren-state)))
8045 open-brace))
8046
a66cd3ee 8047(defun c-inside-bracelist-p (containing-sexp paren-state)
785eecbb
RS
8048 ;; return the buffer position of the beginning of the brace list
8049 ;; statement if we're inside a brace list, otherwise return nil.
8050 ;; CONTAINING-SEXP is the buffer pos of the innermost containing
0386b551 8051 ;; paren. PAREN-STATE is the remainder of the state of enclosing
130c507e 8052 ;; braces
785eecbb
RS
8053 ;;
8054 ;; N.B.: This algorithm can potentially get confused by cpp macros
0386b551 8055 ;; placed in inconvenient locations. It's a trade-off we make for
785eecbb 8056 ;; speed.
0386b551
AM
8057 ;;
8058 ;; This function might do hidden buffer changes.
785eecbb 8059 (or
d9e94c22 8060 ;; This will pick up brace list declarations.
b2acd789
RS
8061 (c-safe
8062 (save-excursion
8063 (goto-char containing-sexp)
0ec8351b 8064 (c-forward-sexp -1)
b2acd789 8065 (let (bracepos)
d9e94c22 8066 (if (and (or (looking-at c-brace-list-key)
0ec8351b 8067 (progn (c-forward-sexp -1)
d9e94c22 8068 (looking-at c-brace-list-key)))
a66cd3ee 8069 (setq bracepos (c-down-list-forward (point)))
b2acd789
RS
8070 (not (c-crosses-statement-barrier-p (point)
8071 (- bracepos 2))))
8072 (point)))))
785eecbb
RS
8073 ;; this will pick up array/aggregate init lists, even if they are nested.
8074 (save-excursion
0ec8351b
BW
8075 (let ((class-key
8076 ;; Pike can have class definitions anywhere, so we must
8077 ;; check for the class key here.
8078 (and (c-major-mode-is 'pike-mode)
a66cd3ee
MS
8079 c-decl-block-key))
8080 bufpos braceassignp lim next-containing)
785eecbb
RS
8081 (while (and (not bufpos)
8082 containing-sexp)
a66cd3ee
MS
8083 (when paren-state
8084 (if (consp (car paren-state))
8085 (setq lim (cdr (car paren-state))
8086 paren-state (cdr paren-state))
8087 (setq lim (car paren-state)))
8088 (when paren-state
8089 (setq next-containing (car paren-state)
8090 paren-state (cdr paren-state))))
785eecbb 8091 (goto-char containing-sexp)
a66cd3ee
MS
8092 (if (c-looking-at-inexpr-block next-containing next-containing)
8093 ;; We're in an in-expression block of some kind. Do not
8094 ;; check nesting. We deliberately set the limit to the
8095 ;; containing sexp, so that c-looking-at-inexpr-block
8096 ;; doesn't check for an identifier before it.
0ec8351b
BW
8097 (setq containing-sexp nil)
8098 ;; see if the open brace is preceded by = or [...] in
8099 ;; this statement, but watch out for operator=
a66cd3ee 8100 (setq braceassignp 'dontknow)
d9e94c22 8101 (c-backward-token-2 1 t lim)
6393fef2 8102 ;; Checks to do only on the first sexp before the brace.
d9e94c22 8103 (when (and c-opt-inexpr-brace-list-key
6393fef2
RS
8104 (eq (char-after) ?\[))
8105 ;; In Java, an initialization brace list may follow
8106 ;; directly after "new Foo[]", so check for a "new"
8107 ;; earlier.
8108 (while (eq braceassignp 'dontknow)
8109 (setq braceassignp
d9e94c22
MS
8110 (cond ((/= (c-backward-token-2 1 t lim) 0) nil)
8111 ((looking-at c-opt-inexpr-brace-list-key) t)
6393fef2
RS
8112 ((looking-at "\\sw\\|\\s_\\|[.[]")
8113 ;; Carry on looking if this is an
8114 ;; identifier (may contain "." in Java)
8115 ;; or another "[]" sexp.
8116 'dontknow)
8117 (t nil)))))
8118 ;; Checks to do on all sexps before the brace, up to the
8119 ;; beginning of the statement.
8120 (while (eq braceassignp 'dontknow)
0ec8351b
BW
8121 (cond ((eq (char-after) ?\;)
8122 (setq braceassignp nil))
8123 ((and class-key
8124 (looking-at class-key))
8125 (setq braceassignp nil))
8126 ((eq (char-after) ?=)
8127 ;; We've seen a =, but must check earlier tokens so
8128 ;; that it isn't something that should be ignored.
8129 (setq braceassignp 'maybe)
8130 (while (and (eq braceassignp 'maybe)
d9e94c22 8131 (zerop (c-backward-token-2 1 t lim)))
0ec8351b
BW
8132 (setq braceassignp
8133 (cond
8134 ;; Check for operator =
51c9af45
AM
8135 ((and c-opt-op-identifier-prefix
8136 (looking-at c-opt-op-identifier-prefix))
0386b551 8137 nil)
130c507e
GM
8138 ;; Check for `<opchar>= in Pike.
8139 ((and (c-major-mode-is 'pike-mode)
8140 (or (eq (char-after) ?`)
8141 ;; Special case for Pikes
8142 ;; `[]=, since '[' is not in
8143 ;; the punctuation class.
8144 (and (eq (char-after) ?\[)
8145 (eq (char-before) ?`))))
8146 nil)
0ec8351b
BW
8147 ((looking-at "\\s.") 'maybe)
8148 ;; make sure we're not in a C++ template
8149 ;; argument assignment
a66cd3ee
MS
8150 ((and
8151 (c-major-mode-is 'c++-mode)
8152 (save-excursion
8153 (let ((here (point))
8154 (pos< (progn
8155 (skip-chars-backward "^<>")
8156 (point))))
8157 (and (eq (char-before) ?<)
8158 (not (c-crosses-statement-barrier-p
8159 pos< here))
8160 (not (c-in-literal))
8161 ))))
0ec8351b 8162 nil)
6393fef2
RS
8163 (t t))))))
8164 (if (and (eq braceassignp 'dontknow)
d9e94c22 8165 (/= (c-backward-token-2 1 t lim) 0))
6393fef2
RS
8166 (setq braceassignp nil)))
8167 (if (not braceassignp)
0ec8351b
BW
8168 (if (eq (char-after) ?\;)
8169 ;; Brace lists can't contain a semicolon, so we're done.
8170 (setq containing-sexp nil)
a66cd3ee
MS
8171 ;; Go up one level.
8172 (setq containing-sexp next-containing
8173 lim nil
8174 next-containing nil))
0ec8351b
BW
8175 ;; we've hit the beginning of the aggregate list
8176 (c-beginning-of-statement-1
a66cd3ee 8177 (c-most-enclosing-brace paren-state))
0ec8351b 8178 (setq bufpos (point))))
a66cd3ee 8179 )
785eecbb
RS
8180 bufpos))
8181 ))
8182
0ec8351b
BW
8183(defun c-looking-at-special-brace-list (&optional lim)
8184