Small rmail fixes.
[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 17328.
CommitLineData
785eecbb
RS
1;;; cc-engine.el --- core syntax guessing engine for CC mode
2
73b0cd50 3;; Copyright (C) 1985, 1987, 1992-2011 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;; Dynamically bound cache for `c-in-literal'.
130c507e 199(defvar c-in-literal-cache t)
d9e94c22 200
d9e94c22 201\f
037558bf
MS
202;; Basic handling of preprocessor directives.
203
204;; This is a dynamically bound cache used together with
205;; `c-query-macro-start' and `c-query-and-set-macro-start'. It only
206;; works as long as point doesn't cross a macro boundary.
207(defvar c-macro-start 'unknown)
208
209(defsubst c-query-and-set-macro-start ()
037558bf
MS
210 (if (symbolp c-macro-start)
211 (setq c-macro-start (save-excursion
0386b551
AM
212 (c-save-buffer-state ()
213 (and (c-beginning-of-macro)
214 (point)))))
037558bf
MS
215 c-macro-start))
216
217(defsubst c-query-macro-start ()
037558bf
MS
218 (if (symbolp c-macro-start)
219 (save-excursion
0386b551
AM
220 (c-save-buffer-state ()
221 (and (c-beginning-of-macro)
222 (point))))
037558bf
MS
223 c-macro-start))
224
225(defun c-beginning-of-macro (&optional lim)
226 "Go to the beginning of a preprocessor directive.
227Leave point at the beginning of the directive and return t if in one,
228otherwise return nil and leave point unchanged.
229
0386b551
AM
230Note that this function might do hidden buffer changes. See the
231comment at the start of cc-engine.el for more info."
037558bf
MS
232 (when c-opt-cpp-prefix
233 (let ((here (point)))
234 (save-restriction
235 (if lim (narrow-to-region lim (point-max)))
236 (beginning-of-line)
237 (while (eq (char-before (1- (point))) ?\\)
238 (forward-line -1))
239 (back-to-indentation)
240 (if (and (<= (point) here)
241 (looking-at c-opt-cpp-start))
242 t
243 (goto-char here)
244 nil)))))
245
246(defun c-end-of-macro ()
247 "Go to the end of a preprocessor directive.
0386b551
AM
248More accurately, move the point to the end of the closest following
249line that doesn't end with a line continuation backslash - no check is
250done that the point is inside a cpp directive to begin with.
037558bf 251
0386b551
AM
252Note that this function might do hidden buffer changes. See the
253comment at the start of cc-engine.el for more info."
037558bf
MS
254 (while (progn
255 (end-of-line)
256 (when (and (eq (char-before) ?\\)
257 (not (eobp)))
258 (forward-char)
259 t))))
260
0ec1d2c5
AM
261(defun c-syntactic-end-of-macro ()
262 ;; Go to the end of a CPP directive, or a "safe" pos just before.
263 ;;
264 ;; This is normally the end of the next non-escaped line. A "safe"
265 ;; position is one not within a string or comment. (The EOL on a line
266 ;; comment is NOT "safe").
267 ;;
268 ;; This function must only be called from the beginning of a CPP construct.
269 ;;
270 ;; Note that this function might do hidden buffer changes. See the comment
271 ;; at the start of cc-engine.el for more info.
272 (let* ((here (point))
273 (there (progn (c-end-of-macro) (point)))
274 (s (parse-partial-sexp here there)))
275 (while (and (or (nth 3 s) ; in a string
276 (nth 4 s)) ; in a comment (maybe at end of line comment)
277 (> there here)) ; No infinite loops, please.
278 (setq there (1- (nth 8 s)))
279 (setq s (parse-partial-sexp here there)))
280 (point)))
281
51c9af45
AM
282(defun c-forward-over-cpp-define-id ()
283 ;; Assuming point is at the "#" that introduces a preprocessor
284 ;; directive, it's moved forward to the end of the identifier which is
285 ;; "#define"d (or whatever c-opt-cpp-macro-define specifies). Non-nil
286 ;; is returned in this case, in all other cases nil is returned and
287 ;; point isn't moved.
288 ;;
289 ;; This function might do hidden buffer changes.
290 (when (and c-opt-cpp-macro-define-id
291 (looking-at c-opt-cpp-macro-define-id))
292 (goto-char (match-end 0))))
293
037558bf
MS
294(defun c-forward-to-cpp-define-body ()
295 ;; Assuming point is at the "#" that introduces a preprocessor
296 ;; directive, it's moved forward to the start of the definition body
0386b551
AM
297 ;; if it's a "#define" (or whatever c-opt-cpp-macro-define
298 ;; specifies). Non-nil is returned in this case, in all other cases
299 ;; nil is returned and point isn't moved.
300 ;;
301 ;; This function might do hidden buffer changes.
302 (when (and c-opt-cpp-macro-define-start
303 (looking-at c-opt-cpp-macro-define-start)
037558bf
MS
304 (not (= (match-end 0) (c-point 'eol))))
305 (goto-char (match-end 0))))
306
307\f
d9e94c22
MS
308;;; Basic utility functions.
309
0386b551 310(defun c-syntactic-content (from to paren-level)
d9e94c22
MS
311 ;; Return the given region as a string where all syntactic
312 ;; whitespace is removed or, where necessary, replaced with a single
0386b551
AM
313 ;; space. If PAREN-LEVEL is given then all parens in the region are
314 ;; collapsed to "()", "[]" etc.
315 ;;
316 ;; This function might do hidden buffer changes.
317
d9e94c22 318 (save-excursion
0386b551
AM
319 (save-restriction
320 (narrow-to-region from to)
321 (goto-char from)
322 (let* ((parts (list nil)) (tail parts) pos in-paren)
323
324 (while (re-search-forward c-syntactic-ws-start to t)
325 (goto-char (setq pos (match-beginning 0)))
326 (c-forward-syntactic-ws)
327 (if (= (point) pos)
328 (forward-char)
329
330 (when paren-level
331 (save-excursion
332 (setq in-paren (= (car (parse-partial-sexp from pos 1)) 1)
333 pos (point))))
334
335 (if (and (> pos from)
336 (< (point) to)
337 (looking-at "\\w\\|\\s_")
338 (save-excursion
339 (goto-char (1- pos))
340 (looking-at "\\w\\|\\s_")))
341 (progn
342 (setcdr tail (list (buffer-substring-no-properties from pos)
343 " "))
344 (setq tail (cddr tail)))
345 (setcdr tail (list (buffer-substring-no-properties from pos)))
346 (setq tail (cdr tail)))
347
348 (when in-paren
349 (when (= (car (parse-partial-sexp pos to -1)) -1)
350 (setcdr tail (list (buffer-substring-no-properties
351 (1- (point)) (point))))
352 (setq tail (cdr tail))))
353
354 (setq from (point))))
355
356 (setcdr tail (list (buffer-substring-no-properties from to)))
357 (apply 'concat (cdr parts))))))
358
359(defun c-shift-line-indentation (shift-amt)
360 ;; Shift the indentation of the current line with the specified
361 ;; amount (positive inwards). The buffer is modified only if
362 ;; SHIFT-AMT isn't equal to zero.
363 (let ((pos (- (point-max) (point)))
364 (c-macro-start c-macro-start)
365 tmp-char-inserted)
366 (if (zerop shift-amt)
367 nil
368 ;; If we're on an empty line inside a macro, we take the point
369 ;; to be at the current indentation and shift it to the
370 ;; appropriate column. This way we don't treat the extra
371 ;; whitespace out to the line continuation as indentation.
372 (when (and (c-query-and-set-macro-start)
373 (looking-at "[ \t]*\\\\$")
374 (save-excursion
375 (skip-chars-backward " \t")
376 (bolp)))
377 (insert ?x)
378 (backward-char)
379 (setq tmp-char-inserted t))
380 (unwind-protect
381 (let ((col (current-indentation)))
382 (delete-region (c-point 'bol) (c-point 'boi))
383 (beginning-of-line)
384 (indent-to (+ col shift-amt)))
385 (when tmp-char-inserted
386 (delete-char 1))))
387 ;; If initial point was within line's indentation and we're not on
388 ;; a line with a line continuation in a macro, position after the
389 ;; indentation. Else stay at same point in text.
390 (if (and (< (point) (c-point 'boi))
391 (not tmp-char-inserted))
392 (back-to-indentation)
393 (if (> (- (point-max) pos) (point))
394 (goto-char (- (point-max) pos))))))
d9e94c22
MS
395
396(defsubst c-keyword-sym (keyword)
397 ;; Return non-nil if the string KEYWORD is a known keyword. More
398 ;; precisely, the value is the symbol for the keyword in
399 ;; `c-keywords-obarray'.
400 (intern-soft keyword c-keywords-obarray))
401
402(defsubst c-keyword-member (keyword-sym lang-constant)
403 ;; Return non-nil if the symbol KEYWORD-SYM, as returned by
404 ;; `c-keyword-sym', is a member of LANG-CONSTANT, which is the name
405 ;; of a language constant that ends with "-kwds". If KEYWORD-SYM is
406 ;; nil then the result is nil.
407 (get keyword-sym lang-constant))
408
409;; String syntax chars, suitable for skip-syntax-(forward|backward).
410(defconst c-string-syntax (if (memq 'gen-string-delim c-emacs-features)
411 "\"|"
412 "\""))
413
0386b551 414;; Regexp matching string limit syntax.
d9e94c22
MS
415(defconst c-string-limit-regexp (if (memq 'gen-string-delim c-emacs-features)
416 "\\s\"\\|\\s|"
417 "\\s\""))
418
0386b551
AM
419;; Regexp matching WS followed by string limit syntax.
420(defconst c-ws*-string-limit-regexp
421 (concat "[ \t]*\\(" c-string-limit-regexp "\\)"))
422
d9e94c22
MS
423;; Holds formatted error strings for the few cases where parse errors
424;; are reported.
a66cd3ee 425(defvar c-parsing-error nil)
d9e94c22
MS
426(make-variable-buffer-local 'c-parsing-error)
427
428(defun c-echo-parsing-error (&optional quiet)
d9e94c22
MS
429 (when (and c-report-syntactic-errors c-parsing-error (not quiet))
430 (c-benign-error "%s" c-parsing-error))
431 c-parsing-error)
432
433;; Faces given to comments and string literals. This is used in some
434;; situations to speed up recognition; it isn't mandatory that font
435;; locking is in use. This variable is extended with the face in
436;; `c-doc-face-name' when fontification is activated in cc-fonts.el.
1e330469 437(defvar c-literal-faces
0386b551
AM
438 (append '(font-lock-comment-face font-lock-string-face)
439 (when (facep 'font-lock-comment-delimiter-face)
440 ;; New in Emacs 22.
441 '(font-lock-comment-delimiter-face))))
442
443(defsubst c-put-c-type-property (pos value)
444 ;; Put a c-type property with the given value at POS.
445 (c-put-char-property pos 'c-type value))
446
447(defun c-clear-c-type-property (from to value)
5a89f0a7 448 ;; Remove all occurrences of the c-type property that has the given
0386b551
AM
449 ;; value in the region between FROM and TO. VALUE is assumed to not
450 ;; be nil.
451 ;;
452 ;; Note: This assumes that c-type is put on single chars only; it's
453 ;; very inefficient if matching properties cover large regions.
454 (save-excursion
455 (goto-char from)
456 (while (progn
457 (when (eq (get-text-property (point) 'c-type) value)
458 (c-clear-char-property (point) 'c-type))
459 (goto-char (next-single-property-change (point) 'c-type nil to))
460 (< (point) to)))))
037558bf 461
d9e94c22
MS
462\f
463;; Some debug tools to visualize various special positions. This
464;; debug code isn't as portable as the rest of CC Mode.
465
466(cc-bytecomp-defun overlays-in)
467(cc-bytecomp-defun overlay-get)
468(cc-bytecomp-defun overlay-start)
469(cc-bytecomp-defun overlay-end)
470(cc-bytecomp-defun delete-overlay)
471(cc-bytecomp-defun overlay-put)
472(cc-bytecomp-defun make-overlay)
473
474(defun c-debug-add-face (beg end face)
475 (c-save-buffer-state ((overlays (overlays-in beg end)) overlay)
476 (while overlays
477 (setq overlay (car overlays)
478 overlays (cdr overlays))
479 (when (eq (overlay-get overlay 'face) face)
480 (setq beg (min beg (overlay-start overlay))
481 end (max end (overlay-end overlay)))
482 (delete-overlay overlay)))
483 (overlay-put (make-overlay beg end) 'face face)))
484
485(defun c-debug-remove-face (beg end face)
486 (c-save-buffer-state ((overlays (overlays-in beg end)) overlay
487 (ol-beg beg) (ol-end end))
488 (while overlays
489 (setq overlay (car overlays)
490 overlays (cdr overlays))
491 (when (eq (overlay-get overlay 'face) face)
492 (setq ol-beg (min ol-beg (overlay-start overlay))
493 ol-end (max ol-end (overlay-end overlay)))
494 (delete-overlay overlay)))
495 (when (< ol-beg beg)
496 (overlay-put (make-overlay ol-beg beg) 'face face))
497 (when (> ol-end end)
498 (overlay-put (make-overlay end ol-end) 'face face))))
499
500\f
501;; `c-beginning-of-statement-1' and accompanying stuff.
130c507e 502
64001211
RS
503;; KLUDGE ALERT: c-maybe-labelp is used to pass information between
504;; c-crosses-statement-barrier-p and c-beginning-of-statement-1. A
505;; better way should be implemented, but this will at least shut up
506;; the byte compiler.
0386b551 507(defvar c-maybe-labelp)
64001211 508
d9e94c22
MS
509;; New awk-compatible version of c-beginning-of-statement-1, ACM 2002/6/22
510
a66cd3ee
MS
511;; Macros used internally in c-beginning-of-statement-1 for the
512;; automaton actions.
513(defmacro c-bos-push-state ()
514 '(setq stack (cons (cons state saved-pos)
515 stack)))
516(defmacro c-bos-pop-state (&optional do-if-done)
517 `(if (setq state (car (car stack))
518 saved-pos (cdr (car stack))
519 stack (cdr stack))
520 t
521 ,do-if-done
522 (throw 'loop nil)))
523(defmacro c-bos-pop-state-and-retry ()
524 '(throw 'loop (setq state (car (car stack))
525 saved-pos (cdr (car stack))
526 ;; Throw nil if stack is empty, else throw non-nil.
527 stack (cdr stack))))
528(defmacro c-bos-save-pos ()
529 '(setq saved-pos (vector pos tok ptok pptok)))
530(defmacro c-bos-restore-pos ()
531 '(unless (eq (elt saved-pos 0) start)
532 (setq pos (elt saved-pos 0)
533 tok (elt saved-pos 1)
534 ptok (elt saved-pos 2)
535 pptok (elt saved-pos 3))
536 (goto-char pos)
537 (setq sym nil)))
538(defmacro c-bos-save-error-info (missing got)
539 `(setq saved-pos (vector pos ,missing ,got)))
540(defmacro c-bos-report-error ()
541 '(unless noerror
542 (setq c-parsing-error
543 (format "No matching `%s' found for `%s' on line %d"
544 (elt saved-pos 1)
545 (elt saved-pos 2)
546 (1+ (count-lines (point-min)
547 (c-point 'bol (elt saved-pos 0))))))))
548
549(defun c-beginning-of-statement-1 (&optional lim ignore-labels
550 noerror comma-delim)
551 "Move to the start of the current statement or declaration, or to
552the previous one if already at the beginning of one. Only
553statements/declarations on the same level are considered, i.e. don't
554move into or out of sexps (not even normal expression parentheses).
555
5a89f0a7 556If point is already at the earliest statement within braces or parens,
a85fd6da
AM
557this function doesn't move back into any whitespace preceding it; it
558returns 'same in this case.
559
d9e94c22
MS
560Stop at statement continuation tokens like \"else\", \"catch\",
561\"finally\" and the \"while\" in \"do ... while\" if the start point
562is within the continuation. If starting at such a token, move to the
563corresponding statement start. If at the beginning of a statement,
564move to the closest containing statement if there is any. This might
565also stop at a continuation clause.
a66cd3ee 566
0386b551
AM
567Labels are treated as part of the following statements if
568IGNORE-LABELS is non-nil. (FIXME: Doesn't work if we stop at a known
cb694ab7
AM
569statement start keyword.) Otherwise, each label is treated as a
570separate statement.
a66cd3ee 571
cb694ab7
AM
572Macros are ignored \(i.e. skipped over) unless point is within one, in
573which case the content of the macro is treated as normal code. Aside
574from any normal statement starts found in it, stop at the first token
575of the content in the macro, i.e. the expression of an \"#if\" or the
576start of the definition in a \"#define\". Also stop at start of
577macros before leaving them.
a66cd3ee 578
a85fd6da 579Return:
d28e7f28 580'label if stopped at a label or \"case...:\" or \"default:\";
a85fd6da
AM
581'same if stopped at the beginning of the current statement;
582'up if stepped to a containing statement;
583'previous if stepped to a preceding statement;
584'beginning if stepped from a statement continuation clause to
585 its start clause; or
586'macro if stepped to a macro start.
587Note that 'same and not 'label is returned if stopped at the same
588label without crossing the colon character.
a66cd3ee
MS
589
590LIM may be given to limit the search. If the search hits the limit,
591point will be left at the closest following token, or at the start
592position if that is less ('same is returned in this case).
593
594NOERROR turns off error logging to `c-parsing-error'.
595
cb694ab7
AM
596Normally only ';' and virtual semicolons are considered to delimit
597statements, but if COMMA-DELIM is non-nil then ',' is treated
598as a delimiter too.
0386b551
AM
599
600Note that this function might do hidden buffer changes. See the
601comment at the start of cc-engine.el for more info."
a66cd3ee 602
d9e94c22
MS
603 ;; The bulk of this function is a pushdown automaton that looks at statement
604 ;; boundaries and the tokens (such as "while") in c-opt-block-stmt-key. Its
605 ;; purpose is to keep track of nested statements, ensuring that such
5a89f0a7 606 ;; statements are skipped over in their entirety (somewhat akin to what C-M-p
d9e94c22 607 ;; does with nested braces/brackets/parentheses).
a66cd3ee
MS
608 ;;
609 ;; Note: The position of a boundary is the following token.
610 ;;
d9e94c22
MS
611 ;; Beginning with the current token (the one following point), move back one
612 ;; sexp at a time (where a sexp is, more or less, either a token or the
613 ;; entire contents of a brace/bracket/paren pair). Each time a statement
614 ;; boundary is crossed or a "while"-like token is found, update the state of
615 ;; the PDA. Stop at the beginning of a statement when the stack (holding
616 ;; nested statement info) is empty and the position has been moved.
617 ;;
618 ;; The following variables constitute the PDA:
619 ;;
620 ;; sym: This is either the "while"-like token (e.g. 'for) we've just
621 ;; scanned back over, 'boundary if we've just gone back over a
622 ;; statement boundary, or nil otherwise.
623 ;; state: takes one of the values (nil else else-boundary while
624 ;; while-boundary catch catch-boundary).
625 ;; nil means "no "while"-like token yet scanned".
626 ;; 'else, for example, means "just gone back over an else".
627 ;; 'else-boundary means "just gone back over a statement boundary
628 ;; immediately after having gone back over an else".
629 ;; saved-pos: A vector of either saved positions (tok ptok pptok, etc.) or
630 ;; of error reporting information.
631 ;; stack: The stack onto which the PDA pushes its state. Each entry
632 ;; consists of a saved value of state and saved-pos. An entry is
633 ;; pushed when we move back over a "continuation" token (e.g. else)
634 ;; and popped when we encounter the corresponding opening token
635 ;; (e.g. if).
636 ;;
637 ;;
b414f371 638 ;; The following diagram briefly outlines the PDA.
a66cd3ee
MS
639 ;;
640 ;; Common state:
d9e94c22
MS
641 ;; "else": Push state, goto state `else'.
642 ;; "while": Push state, goto state `while'.
643 ;; "catch" or "finally": Push state, goto state `catch'.
644 ;; boundary: Pop state.
a66cd3ee
MS
645 ;; other: Do nothing special.
646 ;;
d9e94c22
MS
647 ;; State `else':
648 ;; boundary: Goto state `else-boundary'.
649 ;; other: Error, pop state, retry token.
650 ;;
651 ;; State `else-boundary':
652 ;; "if": Pop state.
653 ;; boundary: Error, pop state.
654 ;; other: See common state.
655 ;;
656 ;; State `while':
657 ;; boundary: Save position, goto state `while-boundary'.
658 ;; other: Pop state, retry token.
659 ;;
660 ;; State `while-boundary':
661 ;; "do": Pop state.
662 ;; boundary: Restore position if it's not at start, pop state. [*see below]
663 ;; other: See common state.
664 ;;
665 ;; State `catch':
666 ;; boundary: Goto state `catch-boundary'.
667 ;; other: Error, pop state, retry token.
668 ;;
669 ;; State `catch-boundary':
670 ;; "try": Pop state.
671 ;; "catch": Goto state `catch'.
672 ;; boundary: Error, pop state.
673 ;; other: See common state.
674 ;;
675 ;; [*] In the `while-boundary' state, we had pushed a 'while state, and were
676 ;; searching for a "do" which would have opened a do-while. If we didn't
677 ;; find it, we discard the analysis done since the "while", go back to this
678 ;; token in the buffer and restart the scanning there, this time WITHOUT
679 ;; pushing the 'while state onto the stack.
680 ;;
a66cd3ee
MS
681 ;; In addition to the above there is some special handling of labels
682 ;; and macros.
683
684 (let ((case-fold-search nil)
685 (start (point))
686 macro-start
687 (delims (if comma-delim '(?\; ?,) '(?\;)))
688 (c-stmt-delim-chars (if comma-delim
689 c-stmt-delim-chars-with-comma
690 c-stmt-delim-chars))
d28e7f28 691 c-in-literal-cache c-maybe-labelp after-case:-pos saved
0386b551
AM
692 ;; Current position.
693 pos
694 ;; Position of last stmt boundary character (e.g. ;).
695 boundary-pos
696 ;; The position of the last sexp or bound that follows the
697 ;; first found colon, i.e. the start of the nonlabel part of
698 ;; the statement. It's `start' if a colon is found just after
699 ;; the start.
700 after-labels-pos
701 ;; Like `after-labels-pos', but the first such position inside
702 ;; a label, i.e. the start of the last label before the start
703 ;; of the nonlabel part of the statement.
704 last-label-pos
705 ;; The last position where a label is possible provided the
706 ;; statement started there. It's nil as long as no invalid
707 ;; label content has been found (according to
020716e1 708 ;; `c-nonlabel-token-key'). It's `start' if no valid label
0386b551
AM
709 ;; content was found in the label. Note that we might still
710 ;; regard it a label if it starts with `c-label-kwds'.
711 label-good-pos
b248a85d
AM
712 ;; Putative positions of the components of a bitfield declaration,
713 ;; e.g. "int foo : NUM_FOO_BITS ;"
714 bitfield-type-pos bitfield-id-pos bitfield-size-pos
0386b551
AM
715 ;; Symbol just scanned back over (e.g. 'while or 'boundary).
716 ;; See above.
717 sym
718 ;; Current state in the automaton. See above.
719 state
720 ;; Current saved positions. See above.
721 saved-pos
722 ;; Stack of conses (state . saved-pos).
723 stack
724 ;; Regexp which matches "for", "if", etc.
725 (cond-key (or c-opt-block-stmt-key
a66cd3ee 726 "\\<\\>")) ; Matches nothing.
0386b551
AM
727 ;; Return value.
728 (ret 'same)
729 ;; Positions of the last three sexps or bounds we've stopped at.
730 tok ptok pptok)
a66cd3ee
MS
731
732 (save-restriction
733 (if lim (narrow-to-region lim (point-max)))
734
735 (if (save-excursion
736 (and (c-beginning-of-macro)
737 (/= (point) start)))
738 (setq macro-start (point)))
739
d9e94c22 740 ;; Try to skip back over unary operator characters, to register
a66cd3ee
MS
741 ;; that we've moved.
742 (while (progn
743 (setq pos (point))
0386b551
AM
744 (c-backward-syntactic-ws)
745 ;; Protect post-++/-- operators just before a virtual semicolon.
746 (and (not (c-at-vsemi-p))
747 (/= (skip-chars-backward "-+!*&~@`#") 0))))
d9e94c22
MS
748
749 ;; Skip back over any semicolon here. If it was a bare semicolon, we're
0386b551 750 ;; done. Later on we ignore the boundaries for statements that don't
d9e94c22
MS
751 ;; contain any sexp. The only thing that is affected is that the error
752 ;; checking is a little less strict, and we really don't bother.
a66cd3ee
MS
753 (if (and (memq (char-before) delims)
754 (progn (forward-char -1)
755 (setq saved (point))
0386b551 756 (c-backward-syntactic-ws)
a66cd3ee
MS
757 (or (memq (char-before) delims)
758 (memq (char-before) '(?: nil))
d9e94c22 759 (eq (char-syntax (char-before)) ?\()
0386b551 760 (c-at-vsemi-p))))
a66cd3ee
MS
761 (setq ret 'previous
762 pos saved)
763
764 ;; Begin at start and not pos to detect macros if we stand
765 ;; directly after the #.
766 (goto-char start)
767 (if (looking-at "\\<\\|\\W")
768 ;; Record this as the first token if not starting inside it.
769 (setq tok start))
770
b248a85d
AM
771
772 ;; The following while loop goes back one sexp (balanced parens,
773 ;; etc. with contents, or symbol or suchlike) each iteration. This
774 ;; movement is accomplished with a call to c-backward-sexp approx 170
775 ;; lines below.
776 ;;
777 ;; The loop is exited only by throwing nil to the (catch 'loop ...):
778 ;; 1. On reaching the start of a macro;
779 ;; 2. On having passed a stmt boundary with the PDA stack empty;
780 ;; 3. On reaching the start of an Objective C method def;
781 ;; 4. From macro `c-bos-pop-state'; when the stack is empty;
782 ;; 5. From macro `c-bos-pop-state-and-retry' when the stack is empty.
a66cd3ee
MS
783 (while
784 (catch 'loop ;; Throw nil to break, non-nil to continue.
785 (cond
b248a85d 786 ;; Are we in a macro, just after the opening #?
a66cd3ee 787 ((save-excursion
0386b551 788 (and macro-start ; Always NIL for AWK.
a66cd3ee
MS
789 (progn (skip-chars-backward " \t")
790 (eq (char-before) ?#))
791 (progn (setq saved (1- (point)))
792 (beginning-of-line)
793 (not (eq (char-before (1- (point))) ?\\)))
d9e94c22 794 (looking-at c-opt-cpp-start)
a66cd3ee
MS
795 (progn (skip-chars-forward " \t")
796 (eq (point) saved))))
797 (goto-char saved)
798 (if (and (c-forward-to-cpp-define-body)
799 (progn (c-forward-syntactic-ws start)
800 (< (point) start)))
801 ;; Stop at the first token in the content of the macro.
802 (setq pos (point)
803 ignore-labels t) ; Avoid the label check on exit.
804 (setq pos saved
805 ret 'macro
806 ignore-labels t))
b248a85d 807 (throw 'loop nil)) ; 1. Start of macro.
a66cd3ee 808
d9e94c22
MS
809 ;; Do a round through the automaton if we've just passed a
810 ;; statement boundary or passed a "while"-like token.
a66cd3ee
MS
811 ((or sym
812 (and (looking-at cond-key)
813 (setq sym (intern (match-string 1)))))
814
815 (when (and (< pos start) (null stack))
b248a85d 816 (throw 'loop nil)) ; 2. Statement boundary.
a66cd3ee 817
d9e94c22
MS
818 ;; The PDA state handling.
819 ;;
037558bf 820 ;; Refer to the description of the PDA in the opening
d9e94c22
MS
821 ;; comments. In the following OR form, the first leaf
822 ;; attempts to handles one of the specific actions detailed
823 ;; (e.g., finding token "if" whilst in state `else-boundary').
824 ;; We drop through to the second leaf (which handles common
825 ;; state) if no specific handler is found in the first cond.
826 ;; If a parsing error is detected (e.g. an "else" with no
827 ;; preceding "if"), we throw to the enclosing catch.
828 ;;
829 ;; Note that the (eq state 'else) means
830 ;; "we've just passed an else", NOT "we're looking for an
831 ;; else".
a66cd3ee
MS
832 (or (cond
833 ((eq state 'else)
834 (if (eq sym 'boundary)
835 (setq state 'else-boundary)
836 (c-bos-report-error)
837 (c-bos-pop-state-and-retry)))
838
839 ((eq state 'else-boundary)
840 (cond ((eq sym 'if)
841 (c-bos-pop-state (setq ret 'beginning)))
842 ((eq sym 'boundary)
843 (c-bos-report-error)
844 (c-bos-pop-state))))
845
846 ((eq state 'while)
847 (if (and (eq sym 'boundary)
848 ;; Since this can cause backtracking we do a
849 ;; little more careful analysis to avoid it:
850 ;; If there's a label in front of the while
851 ;; it can't be part of a do-while.
852 (not after-labels-pos))
853 (progn (c-bos-save-pos)
854 (setq state 'while-boundary))
d9e94c22 855 (c-bos-pop-state-and-retry))) ; Can't be a do-while
a66cd3ee
MS
856
857 ((eq state 'while-boundary)
858 (cond ((eq sym 'do)
859 (c-bos-pop-state (setq ret 'beginning)))
d9e94c22
MS
860 ((eq sym 'boundary) ; isn't a do-while
861 (c-bos-restore-pos) ; the position of the while
862 (c-bos-pop-state)))) ; no longer searching for do.
a66cd3ee
MS
863
864 ((eq state 'catch)
865 (if (eq sym 'boundary)
866 (setq state 'catch-boundary)
867 (c-bos-report-error)
868 (c-bos-pop-state-and-retry)))
869
870 ((eq state 'catch-boundary)
871 (cond
872 ((eq sym 'try)
873 (c-bos-pop-state (setq ret 'beginning)))
874 ((eq sym 'catch)
875 (setq state 'catch))
876 ((eq sym 'boundary)
877 (c-bos-report-error)
878 (c-bos-pop-state)))))
879
d9e94c22
MS
880 ;; This is state common. We get here when the previous
881 ;; cond statement found no particular state handler.
a66cd3ee 882 (cond ((eq sym 'boundary)
d9e94c22
MS
883 ;; If we have a boundary at the start
884 ;; position we push a frame to go to the
885 ;; previous statement.
886 (if (>= pos start)
887 (c-bos-push-state)
888 (c-bos-pop-state)))
a66cd3ee
MS
889 ((eq sym 'else)
890 (c-bos-push-state)
891 (c-bos-save-error-info 'if 'else)
892 (setq state 'else))
893 ((eq sym 'while)
0386b551
AM
894 ;; Is this a real while, or a do-while?
895 ;; The next `when' triggers unless we are SURE that
896 ;; the `while' is not the tailend of a `do-while'.
a66cd3ee 897 (when (or (not pptok)
d9e94c22 898 (memq (char-after pptok) delims)
0386b551
AM
899 ;; The following kludge is to prevent
900 ;; infinite recursion when called from
901 ;; c-awk-after-if-for-while-condition-p,
902 ;; or the like.
903 (and (eq (point) start)
904 (c-vsemi-status-unknown-p))
905 (c-at-vsemi-p pptok))
a66cd3ee
MS
906 ;; Since this can cause backtracking we do a
907 ;; little more careful analysis to avoid it: If
0386b551
AM
908 ;; the while isn't followed by a (possibly
909 ;; virtual) semicolon it can't be a do-while.
a66cd3ee
MS
910 (c-bos-push-state)
911 (setq state 'while)))
912 ((memq sym '(catch finally))
913 (c-bos-push-state)
914 (c-bos-save-error-info 'try sym)
915 (setq state 'catch))))
916
917 (when c-maybe-labelp
918 ;; We're either past a statement boundary or at the
919 ;; start of a statement, so throw away any label data
920 ;; for the previous one.
921 (setq after-labels-pos nil
922 last-label-pos nil
923 c-maybe-labelp nil))))
924
d9e94c22
MS
925 ;; Step to the previous sexp, but not if we crossed a
926 ;; boundary, since that doesn't consume an sexp.
a66cd3ee
MS
927 (if (eq sym 'boundary)
928 (setq ret 'previous)
d9e94c22
MS
929
930 ;; HERE IS THE SINGLE PLACE INSIDE THE PDA LOOP WHERE WE MOVE
0386b551
AM
931 ;; BACKWARDS THROUGH THE SOURCE.
932
0386b551 933 (c-backward-syntactic-ws)
0386b551 934 (let ((before-sws-pos (point))
0386b551
AM
935 ;; The end position of the area to search for statement
936 ;; barriers in this round.
b248a85d 937 (maybe-after-boundary-pos pos))
0386b551 938
b248a85d
AM
939 ;; Go back over exactly one logical sexp, taking proper
940 ;; account of macros and escaped EOLs.
0386b551
AM
941 (while
942 (progn
943 (unless (c-safe (c-backward-sexp) t)
944 ;; Give up if we hit an unbalanced block. Since the
945 ;; stack won't be empty the code below will report a
946 ;; suitable error.
a66cd3ee 947 (throw 'loop nil))
b248a85d
AM
948 (cond
949 ;; Have we moved into a macro?
950 ((and (not macro-start)
951 (c-beginning-of-macro))
952 ;; Have we crossed a statement boundary? If not,
953 ;; keep going back until we find one or a "real" sexp.
954 (and
955 (save-excursion
956 (c-end-of-macro)
957 (not (c-crosses-statement-barrier-p
958 (point) maybe-after-boundary-pos)))
959 (setq maybe-after-boundary-pos (point))))
960 ;; Have we just gone back over an escaped NL? This
961 ;; doesn't count as a sexp.
962 ((looking-at "\\\\$")))))
963
964 ;; Have we crossed a statement boundary?
965 (setq boundary-pos
966 (cond
967 ;; Are we at a macro beginning?
968 ((and (not macro-start)
969 c-opt-cpp-prefix
970 (looking-at c-opt-cpp-prefix))
971 (save-excursion
972 (c-end-of-macro)
973 (c-crosses-statement-barrier-p
974 (point) maybe-after-boundary-pos)))
975 ;; Just gone back over a brace block?
976 ((and
977 (eq (char-after) ?{)
978 (not (c-looking-at-inexpr-block lim nil t)))
979 (save-excursion
980 (c-forward-sexp) (point)))
981 ;; Just gone back over some paren block?
982 ((looking-at "\\s\(")
983 (save-excursion
984 (goto-char (1+ (c-down-list-backward
985 before-sws-pos)))
986 (c-crosses-statement-barrier-p
987 (point) maybe-after-boundary-pos)))
988 ;; Just gone back over an ordinary symbol of some sort?
989 (t (c-crosses-statement-barrier-p
990 (point) maybe-after-boundary-pos))))
91af3942 991
b248a85d
AM
992 (when boundary-pos
993 (setq pptok ptok
994 ptok tok
995 tok boundary-pos
996 sym 'boundary)
997 ;; Like a C "continue". Analyze the next sexp.
998 (throw 'loop t))))
a66cd3ee
MS
999
1000 ;; ObjC method def?
1001 (when (and c-opt-method-key
1002 (setq saved (c-in-method-def-p)))
1003 (setq pos saved
1004 ignore-labels t) ; Avoid the label check on exit.
b248a85d
AM
1005 (throw 'loop nil)) ; 3. ObjC method def.
1006
1007 ;; Might we have a bitfield declaration, "<type> <id> : <size>"?
1008 (if c-has-bitfields
1009 (cond
1010 ;; The : <size> and <id> fields?
1011 ((and (numberp c-maybe-labelp)
1012 (not bitfield-size-pos)
1013 (save-excursion
1014 (goto-char (or tok start))
1015 (not (looking-at c-keywords-regexp)))
1016 (not (looking-at c-keywords-regexp))
1017 (not (c-punctuation-in (point) c-maybe-labelp)))
1018 (setq bitfield-size-pos (or tok start)
1019 bitfield-id-pos (point)))
1020 ;; The <type> field?
1021 ((and bitfield-id-pos
1022 (not bitfield-type-pos))
1023 (if (and (looking-at c-symbol-key) ; Can only be an integer type. :-)
1024 (not (looking-at c-not-primitive-type-keywords-regexp))
1025 (not (c-punctuation-in (point) tok)))
1026 (setq bitfield-type-pos (point))
1027 (setq bitfield-size-pos nil
1028 bitfield-id-pos nil)))))
a66cd3ee 1029
0386b551
AM
1030 ;; Handle labels.
1031 (unless (eq ignore-labels t)
1032 (when (numberp c-maybe-labelp)
cb694ab7
AM
1033 ;; `c-crosses-statement-barrier-p' has found a colon, so we
1034 ;; might be in a label now. Have we got a real label
1035 ;; (including a case label) or something like C++'s "public:"?
d28e7f28
AM
1036 ;; A case label might use an expression rather than a token.
1037 (setq after-case:-pos (or tok start))
020716e1
AM
1038 (if (or (looking-at c-nonlabel-token-key) ; e.g. "while" or "'a'"
1039 ;; Catch C++'s inheritance construct "class foo : bar".
1040 (save-excursion
1041 (and
1042 (c-safe (c-backward-sexp) t)
1043 (looking-at c-nonlabel-token-2-key))))
d28e7f28
AM
1044 (setq c-maybe-labelp nil)
1045 (if after-labels-pos ; Have we already encountered a label?
1046 (if (not last-label-pos)
1047 (setq last-label-pos (or tok start)))
1048 (setq after-labels-pos (or tok start)))
1049 (setq c-maybe-labelp t
1050 label-good-pos nil))) ; bogus "label"
cb694ab7
AM
1051
1052 (when (and (not label-good-pos) ; i.e. no invalid "label"'s yet
1053 ; been found.
1054 (looking-at c-nonlabel-token-key)) ; e.g. "while :"
0386b551
AM
1055 ;; We're in a potential label and it's the first
1056 ;; time we've found something that isn't allowed in
1057 ;; one.
1058 (setq label-good-pos (or tok start))))
1059
1060 ;; We've moved back by a sexp, so update the token positions.
a66cd3ee
MS
1061 (setq sym nil
1062 pptok ptok
1063 ptok tok
1064 tok (point)
b248a85d
AM
1065 pos tok) ; always non-nil
1066 ) ; end of (catch loop ....)
1067 ) ; end of sexp-at-a-time (while ....)
91af3942 1068
a66cd3ee
MS
1069 ;; If the stack isn't empty there might be errors to report.
1070 (while stack
1071 (if (and (vectorp saved-pos) (eq (length saved-pos) 3))
1072 (c-bos-report-error))
1073 (setq saved-pos (cdr (car stack))
1074 stack (cdr stack)))
1075
1076 (when (and (eq ret 'same)
1077 (not (memq sym '(boundary ignore nil))))
1078 ;; Need to investigate closer whether we've crossed
1079 ;; between a substatement and its containing statement.
1080 (if (setq saved (if (looking-at c-block-stmt-1-key)
1081 ptok
1082 pptok))
1083 (cond ((> start saved) (setq pos saved))
1084 ((= start saved) (setq ret 'up)))))
1085
0386b551
AM
1086 (when (and (not ignore-labels)
1087 (eq c-maybe-labelp t)
d9e94c22 1088 (not (eq ret 'beginning))
0386b551 1089 after-labels-pos
b248a85d 1090 (not bitfield-type-pos) ; Bitfields take precedence over labels.
0386b551
AM
1091 (or (not label-good-pos)
1092 (<= label-good-pos pos)
1093 (progn
1094 (goto-char (if (and last-label-pos
1095 (< last-label-pos start))
1096 last-label-pos
1097 pos))
1098 (looking-at c-label-kwds-regexp))))
a66cd3ee
MS
1099 ;; We're in a label. Maybe we should step to the statement
1100 ;; after it.
1101 (if (< after-labels-pos start)
1102 (setq pos after-labels-pos)
1103 (setq ret 'label)
0386b551
AM
1104 (if (and last-label-pos (< last-label-pos start))
1105 ;; Might have jumped over several labels. Go to the last one.
a66cd3ee
MS
1106 (setq pos last-label-pos)))))
1107
d28e7f28 1108 ;; Have we got "case <expression>:"?
a66cd3ee 1109 (goto-char pos)
d28e7f28
AM
1110 (when (and after-case:-pos
1111 (not (eq ret 'beginning))
1112 (looking-at c-case-kwds-regexp))
1113 (if (< after-case:-pos start)
0000ee90
AM
1114 (setq pos after-case:-pos))
1115 (if (eq ret 'same)
1116 (setq ret 'label)))
d28e7f28
AM
1117
1118 ;; Skip over the unary operators that can start the statement.
a66cd3ee 1119 (while (progn
0386b551
AM
1120 (c-backward-syntactic-ws)
1121 ;; protect AWK post-inc/decrement operators, etc.
1122 (and (not (c-at-vsemi-p (point)))
1123 (/= (skip-chars-backward "-+!*&~@`#") 0)))
a66cd3ee
MS
1124 (setq pos (point)))
1125 (goto-char pos)
1126 ret)))
785eecbb 1127
b248a85d
AM
1128(defun c-punctuation-in (from to)
1129 "Return non-nil if there is a non-comment non-macro punctuation character
1130between FROM and TO. FROM must not be in a string or comment. The returned
1131value is the position of the first such character."
1132 (save-excursion
1133 (goto-char from)
1134 (let ((pos (point)))
1135 (while (progn (skip-chars-forward c-symbol-chars to)
1136 (c-forward-syntactic-ws to)
1137 (> (point) pos))
1138 (setq pos (point))))
1139 (and (< (point) to) (point))))
1140
785eecbb 1141(defun c-crosses-statement-barrier-p (from to)
a66cd3ee
MS
1142 "Return non-nil if buffer positions FROM to TO cross one or more
1143statement or declaration boundaries. The returned value is actually
d9e94c22
MS
1144the position of the earliest boundary char. FROM must not be within
1145a string or comment.
a66cd3ee
MS
1146
1147The variable `c-maybe-labelp' is set to the position of the first `:' that
1148might start a label (i.e. not part of `::' and not preceded by `?'). If a
0386b551
AM
1149single `?' is found, then `c-maybe-labelp' is cleared.
1150
1151For AWK, a statement which is terminated by an EOL (not a \; or a }) is
1152regarded as having a \"virtual semicolon\" immediately after the last token on
b414f371 1153the line. If this virtual semicolon is _at_ from, the function recognizes it.
0386b551
AM
1154
1155Note that this function might do hidden buffer changes. See the
1156comment at the start of cc-engine.el for more info."
536610a4
AM
1157 (let* ((skip-chars
1158 ;; If the current language has CPP macros, insert # into skip-chars.
1159 (if c-opt-cpp-symbol
1160 (concat (substring c-stmt-delim-chars 0 1) ; "^"
1161 c-opt-cpp-symbol ; usually "#"
1162 (substring c-stmt-delim-chars 1)) ; e.g. ";{}?:"
1163 c-stmt-delim-chars))
1164 (non-skip-list
1165 (append (substring skip-chars 1) nil)) ; e.g. (?# ?\; ?{ ?} ?? ?:)
1166 lit-range vsemi-pos)
1167 (save-restriction
1168 (widen)
1169 (save-excursion
1170 (catch 'done
1171 (goto-char from)
1172 (while (progn (skip-chars-forward
1173 skip-chars
1174 (min to (c-point 'bonl)))
1175 (< (point) to))
1176 (cond
1177 ;; Virtual semicolon?
1178 ((and (bolp)
1179 (save-excursion
1180 (progn
1181 (if (setq lit-range (c-literal-limits from)) ; Have we landed in a string/comment?
1182 (goto-char (car lit-range)))
1183 (c-backward-syntactic-ws) ; ? put a limit here, maybe?
1184 (setq vsemi-pos (point))
1185 (c-at-vsemi-p))))
1186 (throw 'done vsemi-pos))
1187 ;; In a string/comment?
1188 ((setq lit-range (c-literal-limits))
1189 (goto-char (cdr lit-range)))
1190 ((eq (char-after) ?:)
1191 (forward-char)
1192 (if (and (eq (char-after) ?:)
1193 (< (point) to))
1194 ;; Ignore scope operators.
1195 (forward-char)
1196 (setq c-maybe-labelp (1- (point)))))
1197 ((eq (char-after) ??)
1198 ;; A question mark. Can't be a label, so stop
1199 ;; looking for more : and ?.
1200 (setq c-maybe-labelp nil
1201 skip-chars (substring c-stmt-delim-chars 0 -2)))
1202 ;; At a CPP construct?
1203 ((and c-opt-cpp-symbol (looking-at c-opt-cpp-symbol)
1204 (save-excursion
1205 (forward-line 0)
1206 (looking-at c-opt-cpp-prefix)))
1207 (c-end-of-macro))
1208 ((memq (char-after) non-skip-list)
1209 (throw 'done (point)))))
1210 ;; In trailing space after an as yet undetected virtual semicolon?
1211 (c-backward-syntactic-ws from)
1212 (if (and (< (point) to)
1213 (c-at-vsemi-p))
1214 (point)
1215 nil))))))
0386b551
AM
1216
1217(defun c-at-statement-start-p ()
1218 "Return non-nil if the point is at the first token in a statement
1219or somewhere in the syntactic whitespace before it.
1220
1221A \"statement\" here is not restricted to those inside code blocks.
1222Any kind of declaration-like construct that occur outside function
1223bodies is also considered a \"statement\".
1224
1225Note that this function might do hidden buffer changes. See the
1226comment at the start of cc-engine.el for more info."
1227
1228 (save-excursion
1229 (let ((end (point))
1230 c-maybe-labelp)
1231 (c-syntactic-skip-backward (substring c-stmt-delim-chars 1) nil t)
1232 (or (bobp)
1233 (eq (char-before) ?})
1234 (and (eq (char-before) ?{)
1235 (not (and c-special-brace-lists
1236 (progn (backward-char)
1237 (c-looking-at-special-brace-list)))))
1238 (c-crosses-statement-barrier-p (point) end)))))
1239
1240(defun c-at-expression-start-p ()
1241 "Return non-nil if the point is at the first token in an expression or
1242statement, or somewhere in the syntactic whitespace before it.
1243
1244An \"expression\" here is a bit different from the normal language
1245grammar sense: It's any sequence of expression tokens except commas,
1246unless they are enclosed inside parentheses of some kind. Also, an
1247expression never continues past an enclosing parenthesis, but it might
1248contain parenthesis pairs of any sort except braces.
1249
1250Since expressions never cross statement boundaries, this function also
1251recognizes statement beginnings, just like `c-at-statement-start-p'.
1252
1253Note that this function might do hidden buffer changes. See the
1254comment at the start of cc-engine.el for more info."
1255
1256 (save-excursion
1257 (let ((end (point))
1258 (c-stmt-delim-chars c-stmt-delim-chars-with-comma)
1259 c-maybe-labelp)
1260 (c-syntactic-skip-backward (substring c-stmt-delim-chars 1) nil t)
1261 (or (bobp)
1262 (memq (char-before) '(?{ ?}))
1263 (save-excursion (backward-char)
1264 (looking-at "\\s("))
1265 (c-crosses-statement-barrier-p (point) end)))))
785eecbb
RS
1266
1267\f
d9e94c22
MS
1268;; A set of functions that covers various idiosyncrasies in
1269;; implementations of `forward-comment'.
1270
1271;; Note: Some emacsen considers incorrectly that any line comment
1272;; ending with a backslash continues to the next line. I can't think
1273;; of any way to work around that in a reliable way without changing
1274;; the buffer, though. Suggestions welcome. ;) (No, temporarily
1275;; changing the syntax for backslash doesn't work since we must treat
1276;; escapes in string literals correctly.)
1277
1278(defun c-forward-single-comment ()
1279 "Move forward past whitespace and the closest following comment, if any.
1280Return t if a comment was found, nil otherwise. In either case, the
1281point is moved past the following whitespace. Line continuations,
1282i.e. a backslashes followed by line breaks, are treated as whitespace.
1283The line breaks that end line comments are considered to be the
1284comment enders, so the point will be put on the beginning of the next
1285line if it moved past a line comment.
1286
1287This function does not do any hidden buffer changes."
1288
1289 (let ((start (point)))
1290 (when (looking-at "\\([ \t\n\r\f\v]\\|\\\\[\n\r]\\)+")
1291 (goto-char (match-end 0)))
1292
1293 (when (forward-comment 1)
1294 (if (eobp)
1295 ;; Some emacsen (e.g. XEmacs 21) return t when moving
1296 ;; forwards at eob.
1297 nil
1298
1299 ;; Emacs includes the ending newline in a b-style (c++)
1300 ;; comment, but XEmacs doesn't. We depend on the Emacs
1301 ;; behavior (which also is symmetric).
1302 (if (and (eolp) (elt (parse-partial-sexp start (point)) 7))
1303 (condition-case nil (forward-char 1)))
1304
1305 t))))
1306
1307(defsubst c-forward-comments ()
1308 "Move forward past all following whitespace and comments.
1309Line continuations, i.e. a backslashes followed by line breaks, are
1310treated as whitespace.
1311
0386b551
AM
1312Note that this function might do hidden buffer changes. See the
1313comment at the start of cc-engine.el for more info."
d9e94c22
MS
1314
1315 (while (or
1316 ;; If forward-comment in at least XEmacs 21 is given a large
1317 ;; positive value, it'll loop all the way through if it hits
1318 ;; eob.
1319 (and (forward-comment 5)
1320 ;; Some emacsen (e.g. XEmacs 21) return t when moving
1321 ;; forwards at eob.
1322 (not (eobp)))
1323
1324 (when (looking-at "\\\\[\n\r]")
1325 (forward-char 2)
1326 t))))
1327
1328(defun c-backward-single-comment ()
1329 "Move backward past whitespace and the closest preceding comment, if any.
1330Return t if a comment was found, nil otherwise. In either case, the
1331point is moved past the preceding whitespace. Line continuations,
1332i.e. a backslashes followed by line breaks, are treated as whitespace.
1333The line breaks that end line comments are considered to be the
1334comment enders, so the point cannot be at the end of the same line to
1335move over a line comment.
1336
1337This function does not do any hidden buffer changes."
1338
1339 (let ((start (point)))
1340 ;; When we got newline terminated comments, forward-comment in all
1341 ;; supported emacsen so far will stop at eol of each line not
1342 ;; ending with a comment when moving backwards. This corrects for
1343 ;; that, and at the same time handles line continuations.
1344 (while (progn
1345 (skip-chars-backward " \t\n\r\f\v")
1346 (and (looking-at "[\n\r]")
0386b551 1347 (eq (char-before) ?\\)))
d9e94c22
MS
1348 (backward-char))
1349
1350 (if (bobp)
1351 ;; Some emacsen (e.g. Emacs 19.34) return t when moving
1352 ;; backwards at bob.
1353 nil
1354
1355 ;; Leave point after the closest following newline if we've
1356 ;; backed up over any above, since forward-comment won't move
1357 ;; backward over a line comment if point is at the end of the
1358 ;; same line.
1359 (re-search-forward "\\=\\s *[\n\r]" start t)
1360
bf2c1571 1361 (if (if (let (open-paren-in-column-0-is-defun-start) (forward-comment -1))
d9e94c22
MS
1362 (if (eolp)
1363 ;; If forward-comment above succeeded and we're at eol
1364 ;; then the newline we moved over above didn't end a
1365 ;; line comment, so we give it another go.
bf2c1571
AM
1366 (let (open-paren-in-column-0-is-defun-start)
1367 (forward-comment -1))
d9e94c22
MS
1368 t))
1369
1370 ;; Emacs <= 20 and XEmacs move back over the closer of a
1371 ;; block comment that lacks an opener.
1372 (if (looking-at "\\*/")
1373 (progn (forward-char 2) nil)
1374 t)))))
1375
1376(defsubst c-backward-comments ()
1377 "Move backward past all preceding whitespace and comments.
1378Line continuations, i.e. a backslashes followed by line breaks, are
1379treated as whitespace. The line breaks that end line comments are
1380considered to be the comment enders, so the point cannot be at the end
0386b551
AM
1381of the same line to move over a line comment. Unlike
1382c-backward-syntactic-ws, this function doesn't move back over
1383preprocessor directives.
d9e94c22 1384
0386b551
AM
1385Note that this function might do hidden buffer changes. See the
1386comment at the start of cc-engine.el for more info."
d9e94c22
MS
1387
1388 (let ((start (point)))
1389 (while (and
0386b551 1390 ;; `forward-comment' in some emacsen (e.g. XEmacs 21.4)
d9e94c22
MS
1391 ;; return t when moving backwards at bob.
1392 (not (bobp))
1393
bf2c1571
AM
1394 (if (let (open-paren-in-column-0-is-defun-start)
1395 (forward-comment -1))
d9e94c22
MS
1396 (if (looking-at "\\*/")
1397 ;; Emacs <= 20 and XEmacs move back over the
1398 ;; closer of a block comment that lacks an opener.
1399 (progn (forward-char 2) nil)
1400 t)
1401
1402 ;; XEmacs treats line continuations as whitespace but
1403 ;; only in the backward direction, which seems a bit
1404 ;; odd. Anyway, this is necessary for Emacs.
1405 (when (and (looking-at "[\n\r]")
1406 (eq (char-before) ?\\)
1407 (< (point) start))
1408 (backward-char)
1409 t))))))
1410
1411\f
d9e94c22 1412;; Tools for skipping over syntactic whitespace.
a66cd3ee 1413
d9e94c22
MS
1414;; The following functions use text properties to cache searches over
1415;; large regions of syntactic whitespace. It works as follows:
1416;;
1417;; o If a syntactic whitespace region contains anything but simple
1418;; whitespace (i.e. space, tab and line breaks), the text property
1419;; `c-in-sws' is put over it. At places where we have stopped
1420;; within that region there's also a `c-is-sws' text property.
1421;; That since there typically are nested whitespace inside that
1422;; must be handled separately, e.g. whitespace inside a comment or
1423;; cpp directive. Thus, from one point with `c-is-sws' it's safe
1424;; to jump to another point with that property within the same
1425;; `c-in-sws' region. It can be likened to a ladder where
1426;; `c-in-sws' marks the bars and `c-is-sws' the rungs.
1427;;
1428;; o The `c-is-sws' property is put on the simple whitespace chars at
1429;; a "rung position" and also maybe on the first following char.
1430;; As many characters as can be conveniently found in this range
1431;; are marked, but no assumption can be made that the whole range
1432;; is marked (it could be clobbered by later changes, for
1433;; instance).
1434;;
1435;; Note that some part of the beginning of a sequence of simple
1436;; whitespace might be part of the end of a preceding line comment
1437;; or cpp directive and must not be considered part of the "rung".
1438;; Such whitespace is some amount of horizontal whitespace followed
1439;; by a newline. In the case of cpp directives it could also be
1440;; two newlines with horizontal whitespace between them.
1441;;
1442;; The reason to include the first following char is to cope with
1443;; "rung positions" that doesn't have any ordinary whitespace. If
1444;; `c-is-sws' is put on a token character it does not have
1445;; `c-in-sws' set simultaneously. That's the only case when that
1446;; can occur, and the reason for not extending the `c-in-sws'
1447;; region to cover it is that the `c-in-sws' region could then be
1448;; accidentally merged with a following one if the token is only
1449;; one character long.
1450;;
1451;; o On buffer changes the `c-in-sws' and `c-is-sws' properties are
1452;; removed in the changed region. If the change was inside
1453;; syntactic whitespace that means that the "ladder" is broken, but
1454;; a later call to `c-forward-sws' or `c-backward-sws' will use the
1455;; parts on either side and use an ordinary search only to "repair"
1456;; the gap.
1457;;
1458;; Special care needs to be taken if a region is removed: If there
1459;; are `c-in-sws' on both sides of it which do not connect inside
1460;; the region then they can't be joined. If e.g. a marked macro is
1461;; broken, syntactic whitespace inside the new text might be
1462;; marked. If those marks would become connected with the old
1463;; `c-in-sws' range around the macro then we could get a ladder
1464;; with one end outside the macro and the other at some whitespace
1465;; within it.
1466;;
1467;; The main motivation for this system is to increase the speed in
1468;; skipping over the large whitespace regions that can occur at the
1469;; top level in e.g. header files that contain a lot of comments and
1470;; cpp directives. For small comments inside code it's probably
1471;; slower than using `forward-comment' straightforwardly, but speed is
1472;; not a significant factor there anyway.
1473
1474; (defface c-debug-is-sws-face
1475; '((t (:background "GreenYellow")))
1476; "Debug face to mark the `c-is-sws' property.")
1477; (defface c-debug-in-sws-face
1478; '((t (:underline t)))
1479; "Debug face to mark the `c-in-sws' property.")
1480
1481; (defun c-debug-put-sws-faces ()
1482; ;; Put the sws debug faces on all the `c-is-sws' and `c-in-sws'
1483; ;; properties in the buffer.
1484; (interactive)
1485; (save-excursion
0386b551 1486; (c-save-buffer-state (in-face)
d9e94c22
MS
1487; (goto-char (point-min))
1488; (setq in-face (if (get-text-property (point) 'c-is-sws)
1489; (point)))
1490; (while (progn
1491; (goto-char (next-single-property-change
1492; (point) 'c-is-sws nil (point-max)))
1493; (if in-face
1494; (progn
1495; (c-debug-add-face in-face (point) 'c-debug-is-sws-face)
1496; (setq in-face nil))
1497; (setq in-face (point)))
1498; (not (eobp))))
1499; (goto-char (point-min))
1500; (setq in-face (if (get-text-property (point) 'c-in-sws)
1501; (point)))
1502; (while (progn
1503; (goto-char (next-single-property-change
1504; (point) 'c-in-sws nil (point-max)))
1505; (if in-face
1506; (progn
1507; (c-debug-add-face in-face (point) 'c-debug-in-sws-face)
1508; (setq in-face nil))
1509; (setq in-face (point)))
1510; (not (eobp)))))))
1511
1512(defmacro c-debug-sws-msg (&rest args)
1513 ;;`(message ,@args)
1514 )
1515
1516(defmacro c-put-is-sws (beg end)
0386b551 1517 ;; This macro does a hidden buffer change.
d9e94c22
MS
1518 `(let ((beg ,beg) (end ,end))
1519 (put-text-property beg end 'c-is-sws t)
1520 ,@(when (facep 'c-debug-is-sws-face)
1521 `((c-debug-add-face beg end 'c-debug-is-sws-face)))))
1522
1523(defmacro c-put-in-sws (beg end)
0386b551 1524 ;; This macro does a hidden buffer change.
d9e94c22
MS
1525 `(let ((beg ,beg) (end ,end))
1526 (put-text-property beg end 'c-in-sws t)
1527 ,@(when (facep 'c-debug-is-sws-face)
1528 `((c-debug-add-face beg end 'c-debug-in-sws-face)))))
1529
1530(defmacro c-remove-is-sws (beg end)
0386b551 1531 ;; This macro does a hidden buffer change.
d9e94c22
MS
1532 `(let ((beg ,beg) (end ,end))
1533 (remove-text-properties beg end '(c-is-sws nil))
1534 ,@(when (facep 'c-debug-is-sws-face)
1535 `((c-debug-remove-face beg end 'c-debug-is-sws-face)))))
1536
1537(defmacro c-remove-in-sws (beg end)
0386b551 1538 ;; This macro does a hidden buffer change.
d9e94c22
MS
1539 `(let ((beg ,beg) (end ,end))
1540 (remove-text-properties beg end '(c-in-sws nil))
1541 ,@(when (facep 'c-debug-is-sws-face)
1542 `((c-debug-remove-face beg end 'c-debug-in-sws-face)))))
1543
1544(defmacro c-remove-is-and-in-sws (beg end)
0386b551 1545 ;; This macro does a hidden buffer change.
d9e94c22
MS
1546 `(let ((beg ,beg) (end ,end))
1547 (remove-text-properties beg end '(c-is-sws nil c-in-sws nil))
1548 ,@(when (facep 'c-debug-is-sws-face)
1549 `((c-debug-remove-face beg end 'c-debug-is-sws-face)
1550 (c-debug-remove-face beg end 'c-debug-in-sws-face)))))
1551
1552(defsubst c-invalidate-sws-region-after (beg end)
1553 ;; Called from `after-change-functions'. Note that if
1554 ;; `c-forward-sws' or `c-backward-sws' are used outside
1555 ;; `c-save-buffer-state' or similar then this will remove the cache
1556 ;; properties right after they're added.
0386b551
AM
1557 ;;
1558 ;; This function does hidden buffer changes.
d9e94c22
MS
1559
1560 (save-excursion
1561 ;; Adjust the end to remove the properties in any following simple
1562 ;; ws up to and including the next line break, if there is any
1563 ;; after the changed region. This is necessary e.g. when a rung
1564 ;; marked empty line is converted to a line comment by inserting
1565 ;; "//" before the line break. In that case the line break would
1566 ;; keep the rung mark which could make a later `c-backward-sws'
1567 ;; move into the line comment instead of over it.
1568 (goto-char end)
1569 (skip-chars-forward " \t\f\v")
1570 (when (and (eolp) (not (eobp)))
1571 (setq end (1+ (point)))))
1572
1573 (when (and (= beg end)
1574 (get-text-property beg 'c-in-sws)
2a15eb73 1575 (> beg (point-min))
d9e94c22
MS
1576 (get-text-property (1- beg) 'c-in-sws))
1577 ;; Ensure that an `c-in-sws' range gets broken. Note that it isn't
1578 ;; safe to keep a range that was continuous before the change. E.g:
1579 ;;
1580 ;; #define foo
1581 ;; \
1582 ;; bar
1583 ;;
1584 ;; There can be a "ladder" between "#" and "b". Now, if the newline
1585 ;; after "foo" is removed then "bar" will become part of the cpp
1586 ;; directive instead of a syntactically relevant token. In that
1587 ;; case there's no longer syntactic ws from "#" to "b".
1588 (setq beg (1- beg)))
1589
1590 (c-debug-sws-msg "c-invalidate-sws-region-after [%s..%s]" beg end)
1591 (c-remove-is-and-in-sws beg end))
1592
1593(defun c-forward-sws ()
1594 ;; Used by `c-forward-syntactic-ws' to implement the unbounded search.
0386b551
AM
1595 ;;
1596 ;; This function might do hidden buffer changes.
d9e94c22
MS
1597
1598 (let (;; `rung-pos' is set to a position as early as possible in the
1599 ;; unmarked part of the simple ws region.
1600 (rung-pos (point)) next-rung-pos rung-end-pos last-put-in-sws-pos
1601 rung-is-marked next-rung-is-marked simple-ws-end
1602 ;; `safe-start' is set when it's safe to cache the start position.
1603 ;; It's not set if we've initially skipped over comments and line
1604 ;; continuations since we might have gone out through the end of a
1605 ;; macro then. This provision makes `c-forward-sws' not populate the
1606 ;; cache in the majority of cases, but otoh is `c-backward-sws' by far
1607 ;; more common.
1608 safe-start)
1609
1610 ;; Skip simple ws and do a quick check on the following character to see
1611 ;; if it's anything that can't start syntactic ws, so we can bail out
1612 ;; early in the majority of cases when there just are a few ws chars.
1613 (skip-chars-forward " \t\n\r\f\v")
1614 (when (looking-at c-syntactic-ws-start)
1615
1616 (setq rung-end-pos (min (1+ (point)) (point-max)))
1617 (if (setq rung-is-marked (text-property-any rung-pos rung-end-pos
1618 'c-is-sws t))
1619 ;; Find the last rung position to avoid setting properties in all
1620 ;; the cases when the marked rung is complete.
1621 ;; (`next-single-property-change' is certain to move at least one
1622 ;; step forward.)
1623 (setq rung-pos (1- (next-single-property-change
1624 rung-is-marked 'c-is-sws nil rung-end-pos)))
1625 ;; Got no marked rung here. Since the simple ws might have started
1626 ;; inside a line comment or cpp directive we must set `rung-pos' as
1627 ;; high as possible.
1628 (setq rung-pos (point)))
1629
1630 (while
1631 (progn
1632 (while
1633 (when (and rung-is-marked
1634 (get-text-property (point) 'c-in-sws))
1635
1636 ;; The following search is the main reason that `c-in-sws'
1637 ;; and `c-is-sws' aren't combined to one property.
1638 (goto-char (next-single-property-change
1639 (point) 'c-in-sws nil (point-max)))
1640 (unless (get-text-property (point) 'c-is-sws)
1641 ;; If the `c-in-sws' region extended past the last
1642 ;; `c-is-sws' char we have to go back a bit.
1643 (or (get-text-property (1- (point)) 'c-is-sws)
1644 (goto-char (previous-single-property-change
1645 (point) 'c-is-sws)))
1646 (backward-char))
1647
1648 (c-debug-sws-msg
1649 "c-forward-sws cached move %s -> %s (max %s)"
1650 rung-pos (point) (point-max))
1651
1652 (setq rung-pos (point))
1653 (and (> (skip-chars-forward " \t\n\r\f\v") 0)
1654 (not (eobp))))
1655
1656 ;; We'll loop here if there is simple ws after the last rung.
1657 ;; That means that there's been some change in it and it's
1658 ;; possible that we've stepped into another ladder, so extend
1659 ;; the previous one to join with it if there is one, and try to
1660 ;; use the cache again.
1661 (c-debug-sws-msg
1662 "c-forward-sws extending rung with [%s..%s] (max %s)"
1663 (1+ rung-pos) (1+ (point)) (point-max))
1664 (unless (get-text-property (point) 'c-is-sws)
1665 ;; Remove any `c-in-sws' property from the last char of
1666 ;; the rung before we mark it with `c-is-sws', so that we
1667 ;; won't connect with the remains of a broken "ladder".
1668 (c-remove-in-sws (point) (1+ (point))))
1669 (c-put-is-sws (1+ rung-pos)
1670 (1+ (point)))
1671 (c-put-in-sws rung-pos
1672 (setq rung-pos (point)
1673 last-put-in-sws-pos rung-pos)))
1674
1675 (setq simple-ws-end (point))
1676 (c-forward-comments)
1677
1678 (cond
1679 ((/= (point) simple-ws-end)
1680 ;; Skipped over comments. Don't cache at eob in case the buffer
1681 ;; is narrowed.
1682 (not (eobp)))
1683
1684 ((save-excursion
1685 (and c-opt-cpp-prefix
1686 (looking-at c-opt-cpp-start)
1687 (progn (skip-chars-backward " \t")
1688 (bolp))
1689 (or (bobp)
1690 (progn (backward-char)
1691 (not (eq (char-before) ?\\))))))
1692 ;; Skip a preprocessor directive.
1693 (end-of-line)
1694 (while (and (eq (char-before) ?\\)
1695 (= (forward-line 1) 0))
1696 (end-of-line))
1697 (forward-line 1)
1698 (setq safe-start t)
1699 ;; Don't cache at eob in case the buffer is narrowed.
1700 (not (eobp)))))
1701
1702 ;; We've searched over a piece of non-white syntactic ws. See if this
1703 ;; can be cached.
1704 (setq next-rung-pos (point))
1705 (skip-chars-forward " \t\n\r\f\v")
1706 (setq rung-end-pos (min (1+ (point)) (point-max)))
1707
1708 (if (or
1709 ;; Cache if we haven't skipped comments only, and if we started
1710 ;; either from a marked rung or from a completely uncached
1711 ;; position.
1712 (and safe-start
1713 (or rung-is-marked
1714 (not (get-text-property simple-ws-end 'c-in-sws))))
1715
1716 ;; See if there's a marked rung in the encountered simple ws. If
1717 ;; so then we can cache, unless `safe-start' is nil. Even then
1718 ;; we need to do this to check if the cache can be used for the
1719 ;; next step.
1720 (and (setq next-rung-is-marked
1721 (text-property-any next-rung-pos rung-end-pos
1722 'c-is-sws t))
1723 safe-start))
b2acd789 1724
0ec8351b 1725 (progn
d9e94c22
MS
1726 (c-debug-sws-msg
1727 "c-forward-sws caching [%s..%s] - [%s..%s] (max %s)"
1728 rung-pos (1+ simple-ws-end) next-rung-pos rung-end-pos
1729 (point-max))
1730
1731 ;; Remove the properties for any nested ws that might be cached.
1732 ;; Only necessary for `c-is-sws' since `c-in-sws' will be set
1733 ;; anyway.
1734 (c-remove-is-sws (1+ simple-ws-end) next-rung-pos)
1735 (unless (and rung-is-marked (= rung-pos simple-ws-end))
1736 (c-put-is-sws rung-pos
1737 (1+ simple-ws-end))
1738 (setq rung-is-marked t))
1739 (c-put-in-sws rung-pos
1740 (setq rung-pos (point)
1741 last-put-in-sws-pos rung-pos))
1742 (unless (get-text-property (1- rung-end-pos) 'c-is-sws)
1743 ;; Remove any `c-in-sws' property from the last char of
1744 ;; the rung before we mark it with `c-is-sws', so that we
1745 ;; won't connect with the remains of a broken "ladder".
1746 (c-remove-in-sws (1- rung-end-pos) rung-end-pos))
1747 (c-put-is-sws next-rung-pos
1748 rung-end-pos))
1749
1750 (c-debug-sws-msg
1751 "c-forward-sws not caching [%s..%s] - [%s..%s] (max %s)"
1752 rung-pos (1+ simple-ws-end) next-rung-pos rung-end-pos
1753 (point-max))
1754
1755 ;; Set `rung-pos' for the next rung. It's the same thing here as
1756 ;; initially, except that the rung position is set as early as
1757 ;; possible since we can't be in the ending ws of a line comment or
1758 ;; cpp directive now.
1759 (if (setq rung-is-marked next-rung-is-marked)
1760 (setq rung-pos (1- (next-single-property-change
1761 rung-is-marked 'c-is-sws nil rung-end-pos)))
1762 (setq rung-pos next-rung-pos))
1763 (setq safe-start t)))
1764
1765 ;; Make sure that the newly marked `c-in-sws' region doesn't connect to
1766 ;; another one after the point (which might occur when editing inside a
1767 ;; comment or macro).
1768 (when (eq last-put-in-sws-pos (point))
1769 (cond ((< last-put-in-sws-pos (point-max))
1770 (c-debug-sws-msg
1771 "c-forward-sws clearing at %s for cache separation"
1772 last-put-in-sws-pos)
1773 (c-remove-in-sws last-put-in-sws-pos
1774 (1+ last-put-in-sws-pos)))
1775 (t
1776 ;; If at eob we have to clear the last character before the end
1777 ;; instead since the buffer might be narrowed and there might
1778 ;; be a `c-in-sws' after (point-max). In this case it's
1779 ;; necessary to clear both properties.
1780 (c-debug-sws-msg
1781 "c-forward-sws clearing thoroughly at %s for cache separation"
1782 (1- last-put-in-sws-pos))
1783 (c-remove-is-and-in-sws (1- last-put-in-sws-pos)
1784 last-put-in-sws-pos))))
1785 )))
b2acd789 1786
d9e94c22
MS
1787(defun c-backward-sws ()
1788 ;; Used by `c-backward-syntactic-ws' to implement the unbounded search.
0386b551
AM
1789 ;;
1790 ;; This function might do hidden buffer changes.
d9e94c22
MS
1791
1792 (let (;; `rung-pos' is set to a position as late as possible in the unmarked
1793 ;; part of the simple ws region.
1794 (rung-pos (point)) next-rung-pos last-put-in-sws-pos
1795 rung-is-marked simple-ws-beg cmt-skip-pos)
1796
1797 ;; Skip simple horizontal ws and do a quick check on the preceding
91af3942 1798 ;; character to see if it's anything that can't end syntactic ws, so we can
d9e94c22
MS
1799 ;; bail out early in the majority of cases when there just are a few ws
1800 ;; chars. Newlines are complicated in the backward direction, so we can't
1801 ;; skip over them.
1802 (skip-chars-backward " \t\f")
1803 (when (and (not (bobp))
1804 (save-excursion
1805 (backward-char)
1806 (looking-at c-syntactic-ws-end)))
1807
1808 ;; Try to find a rung position in the simple ws preceding point, so that
1809 ;; we can get a cache hit even if the last bit of the simple ws has
1810 ;; changed recently.
1811 (setq simple-ws-beg (point))
1812 (skip-chars-backward " \t\n\r\f\v")
1813 (if (setq rung-is-marked (text-property-any
1814 (point) (min (1+ rung-pos) (point-max))
1815 'c-is-sws t))
1816 ;; `rung-pos' will be the earliest marked position, which means that
1817 ;; there might be later unmarked parts in the simple ws region.
1818 ;; It's not worth the effort to fix that; the last part of the
1819 ;; simple ws is also typically edited often, so it could be wasted.
1820 (goto-char (setq rung-pos rung-is-marked))
1821 (goto-char simple-ws-beg))
1822
1823 (while
1824 (progn
1825 (while
1826 (when (and rung-is-marked
1827 (not (bobp))
1828 (get-text-property (1- (point)) 'c-in-sws))
1829
1830 ;; The following search is the main reason that `c-in-sws'
1831 ;; and `c-is-sws' aren't combined to one property.
1832 (goto-char (previous-single-property-change
1833 (point) 'c-in-sws nil (point-min)))
1834 (unless (get-text-property (point) 'c-is-sws)
1835 ;; If the `c-in-sws' region extended past the first
1836 ;; `c-is-sws' char we have to go forward a bit.
1837 (goto-char (next-single-property-change
1838 (point) 'c-is-sws)))
1839
1840 (c-debug-sws-msg
1841 "c-backward-sws cached move %s <- %s (min %s)"
1842 (point) rung-pos (point-min))
1843
1844 (setq rung-pos (point))
1845 (if (and (< (min (skip-chars-backward " \t\f\v")
1846 (progn
1847 (setq simple-ws-beg (point))
1848 (skip-chars-backward " \t\n\r\f\v")))
1849 0)
1850 (setq rung-is-marked
1851 (text-property-any (point) rung-pos
1852 'c-is-sws t)))
1853 t
1854 (goto-char simple-ws-beg)
1855 nil))
1856
1857 ;; We'll loop here if there is simple ws before the first rung.
1858 ;; That means that there's been some change in it and it's
1859 ;; possible that we've stepped into another ladder, so extend
1860 ;; the previous one to join with it if there is one, and try to
1861 ;; use the cache again.
1862 (c-debug-sws-msg
1863 "c-backward-sws extending rung with [%s..%s] (min %s)"
1864 rung-is-marked rung-pos (point-min))
1865 (unless (get-text-property (1- rung-pos) 'c-is-sws)
1866 ;; Remove any `c-in-sws' property from the last char of
1867 ;; the rung before we mark it with `c-is-sws', so that we
1868 ;; won't connect with the remains of a broken "ladder".
1869 (c-remove-in-sws (1- rung-pos) rung-pos))
1870 (c-put-is-sws rung-is-marked
1871 rung-pos)
1872 (c-put-in-sws rung-is-marked
1873 (1- rung-pos))
1874 (setq rung-pos rung-is-marked
1875 last-put-in-sws-pos rung-pos))
1876
1877 (c-backward-comments)
1878 (setq cmt-skip-pos (point))
a66cd3ee 1879
d9e94c22
MS
1880 (cond
1881 ((and c-opt-cpp-prefix
1882 (/= cmt-skip-pos simple-ws-beg)
1883 (c-beginning-of-macro))
1884 ;; Inside a cpp directive. See if it should be skipped over.
1885 (let ((cpp-beg (point)))
1886
1887 ;; Move back over all line continuations in the region skipped
1888 ;; over by `c-backward-comments'. If we go past it then we
1889 ;; started inside the cpp directive.
1890 (goto-char simple-ws-beg)
1891 (beginning-of-line)
1892 (while (and (> (point) cmt-skip-pos)
1893 (progn (backward-char)
1894 (eq (char-before) ?\\)))
1895 (beginning-of-line))
1896
1897 (if (< (point) cmt-skip-pos)
1898 ;; Don't move past the cpp directive if we began inside
1899 ;; it. Note that the position at the end of the last line
1900 ;; of the macro is also considered to be within it.
1901 (progn (goto-char cmt-skip-pos)
1902 nil)
1903
1904 ;; It's worthwhile to spend a little bit of effort on finding
1905 ;; the end of the macro, to get a good `simple-ws-beg'
1906 ;; position for the cache. Note that `c-backward-comments'
1907 ;; could have stepped over some comments before going into
1908 ;; the macro, and then `simple-ws-beg' must be kept on the
1909 ;; same side of those comments.
1910 (goto-char simple-ws-beg)
1911 (skip-chars-backward " \t\n\r\f\v")
1912 (if (eq (char-before) ?\\)
1913 (forward-char))
1914 (forward-line 1)
1915 (if (< (point) simple-ws-beg)
1916 ;; Might happen if comments after the macro were skipped
1917 ;; over.
1918 (setq simple-ws-beg (point)))
1919
1920 (goto-char cpp-beg)
1921 t)))
1922
1923 ((/= (save-excursion
1924 (skip-chars-forward " \t\n\r\f\v" simple-ws-beg)
1925 (setq next-rung-pos (point)))
1926 simple-ws-beg)
1927 ;; Skipped over comments. Must put point at the end of
1928 ;; the simple ws at point since we might be after a line
1929 ;; comment or cpp directive that's been partially
1930 ;; narrowed out, and we can't risk marking the simple ws
1931 ;; at the end of it.
1932 (goto-char next-rung-pos)
1933 t)))
1934
1935 ;; We've searched over a piece of non-white syntactic ws. See if this
1936 ;; can be cached.
1937 (setq next-rung-pos (point))
1938 (skip-chars-backward " \t\f\v")
1939
1940 (if (or
1941 ;; Cache if we started either from a marked rung or from a
1942 ;; completely uncached position.
1943 rung-is-marked
1944 (not (get-text-property (1- simple-ws-beg) 'c-in-sws))
1945
1946 ;; Cache if there's a marked rung in the encountered simple ws.
1947 (save-excursion
1948 (skip-chars-backward " \t\n\r\f\v")
1949 (text-property-any (point) (min (1+ next-rung-pos) (point-max))
1950 'c-is-sws t)))
a66cd3ee 1951
d9e94c22
MS
1952 (progn
1953 (c-debug-sws-msg
1954 "c-backward-sws caching [%s..%s] - [%s..%s] (min %s)"
1955 (point) (1+ next-rung-pos)
1956 simple-ws-beg (min (1+ rung-pos) (point-max))
1957 (point-min))
1958
1959 ;; Remove the properties for any nested ws that might be cached.
1960 ;; Only necessary for `c-is-sws' since `c-in-sws' will be set
1961 ;; anyway.
1962 (c-remove-is-sws (1+ next-rung-pos) simple-ws-beg)
1963 (unless (and rung-is-marked (= simple-ws-beg rung-pos))
1964 (let ((rung-end-pos (min (1+ rung-pos) (point-max))))
1965 (unless (get-text-property (1- rung-end-pos) 'c-is-sws)
1966 ;; Remove any `c-in-sws' property from the last char of
1967 ;; the rung before we mark it with `c-is-sws', so that we
1968 ;; won't connect with the remains of a broken "ladder".
1969 (c-remove-in-sws (1- rung-end-pos) rung-end-pos))
1970 (c-put-is-sws simple-ws-beg
1971 rung-end-pos)
1972 (setq rung-is-marked t)))
1973 (c-put-in-sws (setq simple-ws-beg (point)
1974 last-put-in-sws-pos simple-ws-beg)
1975 rung-pos)
1976 (c-put-is-sws (setq rung-pos simple-ws-beg)
1977 (1+ next-rung-pos)))
1978
1979 (c-debug-sws-msg
1980 "c-backward-sws not caching [%s..%s] - [%s..%s] (min %s)"
1981 (point) (1+ next-rung-pos)
1982 simple-ws-beg (min (1+ rung-pos) (point-max))
1983 (point-min))
1984 (setq rung-pos next-rung-pos
1985 simple-ws-beg (point))
1986 ))
1987
1988 ;; Make sure that the newly marked `c-in-sws' region doesn't connect to
1989 ;; another one before the point (which might occur when editing inside a
1990 ;; comment or macro).
1991 (when (eq last-put-in-sws-pos (point))
1992 (cond ((< (point-min) last-put-in-sws-pos)
1993 (c-debug-sws-msg
1994 "c-backward-sws clearing at %s for cache separation"
1995 (1- last-put-in-sws-pos))
1996 (c-remove-in-sws (1- last-put-in-sws-pos)
1997 last-put-in-sws-pos))
1998 ((> (point-min) 1)
1999 ;; If at bob and the buffer is narrowed, we have to clear the
2000 ;; character we're standing on instead since there might be a
2001 ;; `c-in-sws' before (point-min). In this case it's necessary
2002 ;; to clear both properties.
2003 (c-debug-sws-msg
2004 "c-backward-sws clearing thoroughly at %s for cache separation"
2005 last-put-in-sws-pos)
2006 (c-remove-is-and-in-sws last-put-in-sws-pos
2007 (1+ last-put-in-sws-pos)))))
2008 )))
785eecbb 2009
d9e94c22 2010\f
580fba94
AM
2011;; Other whitespace tools
2012(defun c-partial-ws-p (beg end)
2013 ;; Is the region (beg end) WS, and is there WS (or BOB/EOB) next to the
2014 ;; region? This is a "heuristic" function. .....
b414f371 2015 ;;
c421028a
AM
2016 ;; The motivation for the second bit is to check whether removing this
2017 ;; region would coalesce two symbols.
580fba94
AM
2018 ;;
2019 ;; FIXME!!! This function doesn't check virtual semicolons in any way. Be
2020 ;; careful about using this function for, e.g. AWK. (2007/3/7)
2021 (save-excursion
2022 (let ((end+1 (min (1+ end) (point-max))))
2023 (or (progn (goto-char (max (point-min) (1- beg)))
2024 (c-skip-ws-forward end)
2025 (eq (point) end))
2026 (progn (goto-char beg)
2027 (c-skip-ws-forward end+1)
2028 (eq (point) end+1))))))
2029\f
0386b551 2030;; A system for finding noteworthy parens before the point.
e1c458ae 2031
0ec1d2c5
AM
2032(defconst c-state-cache-too-far 5000)
2033;; A maximum comfortable scanning distance, e.g. between
2034;; `c-state-cache-good-pos' and "HERE" (where we call c-parse-state). When
0d26e0b6 2035;; this distance is exceeded, we take "emergency measures", e.g. by clearing
0ec1d2c5
AM
2036;; the cache and starting again from point-min or a beginning of defun. This
2037;; value can be tuned for efficiency or set to a lower value for testing.
2038
d9e94c22
MS
2039(defvar c-state-cache nil)
2040(make-variable-buffer-local 'c-state-cache)
2041;; The state cache used by `c-parse-state' to cut down the amount of
0ec1d2c5
AM
2042;; searching. It's the result from some earlier `c-parse-state' call. See
2043;; `c-parse-state''s doc string for details of its structure.
0386b551 2044;;
d9e94c22
MS
2045;; The use of the cached info is more effective if the next
2046;; `c-parse-state' call is on a line close by the one the cached state
2047;; was made at; the cache can actually slow down a little if the
2048;; cached state was made very far back in the buffer. The cache is
2049;; most effective if `c-parse-state' is used on each line while moving
2050;; forward.
e1c458ae 2051
0386b551
AM
2052(defvar c-state-cache-good-pos 1)
2053(make-variable-buffer-local 'c-state-cache-good-pos)
0ec1d2c5
AM
2054;; This is a position where `c-state-cache' is known to be correct, or
2055;; nil (see below). It's a position inside one of the recorded unclosed
2056;; parens or the top level, but not further nested inside any literal or
2057;; subparen that is closed before the last recorded position.
0386b551
AM
2058;;
2059;; The exact position is chosen to try to be close to yet earlier than
2060;; the position where `c-state-cache' will be called next. Right now
2061;; the heuristic is to set it to the position after the last found
2062;; closing paren (of any type) before the line on which
2063;; `c-parse-state' was called. That is chosen primarily to work well
2064;; with refontification of the current line.
0ec1d2c5
AM
2065;;
2066;; 2009-07-28: When `c-state-point-min' and the last position where
2067;; `c-parse-state' or for which `c-invalidate-state-cache' was called, are
2068;; both in the same literal, there is no such "good position", and
2069;; c-state-cache-good-pos is then nil. This is the ONLY circumstance in which
2070;; it can be nil. In this case, `c-state-point-min-literal' will be non-nil.
2071;;
2072;; 2009-06-12: In a brace desert, c-state-cache-good-pos may also be in
2073;; the middle of the desert, as long as it is not within a brace pair
2074;; recorded in `c-state-cache' or a paren/bracket pair.
2075
2076
2077;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2078;; We maintain a simple cache of positions which aren't in a literal, so as to
2079;; speed up testing for non-literality.
2080(defconst c-state-nonlit-pos-interval 10000)
2081;; The approximate interval between entries in `c-state-nonlit-pos-cache'.
2082
2083(defvar c-state-nonlit-pos-cache nil)
2084(make-variable-buffer-local 'c-state-nonlit-pos-cache)
e84efb70
AM
2085;; A list of buffer positions which are known not to be in a literal or a cpp
2086;; construct. This is ordered with higher positions at the front of the list.
2087;; Only those which are less than `c-state-nonlit-pos-cache-limit' are valid.
0ec1d2c5
AM
2088
2089(defvar c-state-nonlit-pos-cache-limit 1)
2090(make-variable-buffer-local 'c-state-nonlit-pos-cache-limit)
2091;; An upper limit on valid entries in `c-state-nonlit-pos-cache'. This is
2092;; reduced by buffer changes, and increased by invocations of
2093;; `c-state-literal-at'.
2094
2095(defsubst c-state-pp-to-literal (from to)
2096 ;; Do a parse-partial-sexp from FROM to TO, returning the bounds of any
2097 ;; literal at TO as a cons, otherwise NIL.
2098 ;; FROM must not be in a literal, and the buffer should already be wide
2099 ;; enough.
2100 (save-excursion
2101 (let ((s (parse-partial-sexp from to)))
2102 (when (or (nth 3 s) (nth 4 s)) ; in a string or comment
2103 (parse-partial-sexp (point) (point-max)
2104 nil ; TARGETDEPTH
2105 nil ; STOPBEFORE
2106 s ; OLDSTATE
2107 'syntax-table) ; stop at end of literal
2108 (cons (nth 8 s) (point))))))
2109
2110(defun c-state-literal-at (here)
2111 ;; If position HERE is inside a literal, return (START . END), the
2112 ;; boundaries of the literal (which may be outside the accessible bit of the
2113 ;; buffer). Otherwise, return nil.
2114 ;;
2115 ;; This function is almost the same as `c-literal-limits'. It differs in
2116 ;; that it is a lower level function, and that it rigourously follows the
2117 ;; syntax from BOB, whereas `c-literal-limits' uses a "local" safe position.
e84efb70
AM
2118 ;;
2119 ;; NOTE: This function manipulates `c-state-nonlit-pos-cache'. This cache
2120 ;; MAY NOT contain any positions within macros, since macros are frequently
2121 ;; turned into comments by use of the `c-cpp-delimiter' category properties.
2122 ;; We cannot rely on this mechanism whilst determining a cache pos since
2123 ;; this function is also called from outwith `c-parse-state'.
0ec1d2c5
AM
2124 (save-restriction
2125 (widen)
2126 (save-excursion
2127 (let ((c c-state-nonlit-pos-cache)
2128 pos npos lit)
2129 ;; Trim the cache to take account of buffer changes.
2130 (while (and c (> (car c) c-state-nonlit-pos-cache-limit))
2131 (setq c (cdr c)))
2132 (setq c-state-nonlit-pos-cache c)
2133
2134 (while (and c (> (car c) here))
2135 (setq c (cdr c)))
2136 (setq pos (or (car c) (point-min)))
2137
2138 (while (<= (setq npos (+ pos c-state-nonlit-pos-interval))
2139 here)
2140 (setq lit (c-state-pp-to-literal pos npos))
2141 (setq pos (or (cdr lit) npos)) ; end of literal containing npos.
e84efb70
AM
2142 (goto-char pos)
2143 (when (and (c-beginning-of-macro) (/= (point) pos))
2144 (c-syntactic-end-of-macro)
2145 (or (eobp) (forward-char))
2146 (setq pos (point)))
0ec1d2c5
AM
2147 (setq c-state-nonlit-pos-cache (cons pos c-state-nonlit-pos-cache)))
2148
2149 (if (> pos c-state-nonlit-pos-cache-limit)
2150 (setq c-state-nonlit-pos-cache-limit pos))
2151 (if (< pos here)
2152 (setq lit (c-state-pp-to-literal pos here)))
2153 lit))))
2154
2155(defsubst c-state-lit-beg (pos)
2156 ;; Return the start of the literal containing POS, or POS itself.
2157 (or (car (c-state-literal-at pos))
2158 pos))
2159
2160(defsubst c-state-cache-non-literal-place (pos state)
9c4aeabf 2161 ;; Return a position outside of a string/comment/macro at or before POS.
0ec1d2c5 2162 ;; STATE is the parse-partial-sexp state at POS.
9c4aeabf
AM
2163 (let ((res (if (or (nth 3 state) ; in a string?
2164 (nth 4 state)) ; in a comment?
2165 (nth 8 state)
2166 pos)))
2167 (save-excursion
2168 (goto-char res)
2169 (if (c-beginning-of-macro)
2170 (point)
2171 res))))
0ec1d2c5
AM
2172
2173;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2174;; Stuff to do with point-min, and coping with any literal there.
2175(defvar c-state-point-min 1)
2176(make-variable-buffer-local 'c-state-point-min)
2177;; This is (point-min) when `c-state-cache' was last calculated. A change of
2178;; narrowing is likely to affect the parens that are visible before the point.
2179
2180(defvar c-state-point-min-lit-type nil)
2181(make-variable-buffer-local 'c-state-point-min-lit-type)
2182(defvar c-state-point-min-lit-start nil)
2183(make-variable-buffer-local 'c-state-point-min-lit-start)
2184;; These two variables define the literal, if any, containing point-min.
2185;; Their values are, respectively, 'string, c, or c++, and the start of the
2186;; literal. If there's no literal there, they're both nil.
2187
2188(defvar c-state-min-scan-pos 1)
2189(make-variable-buffer-local 'c-state-min-scan-pos)
2190;; This is the earliest buffer-pos from which scanning can be done. It is
2191;; either the end of the literal containing point-min, or point-min itself.
2192;; It becomes nil if the buffer is changed earlier than this point.
2193(defun c-state-get-min-scan-pos ()
2194 ;; Return the lowest valid scanning pos. This will be the end of the
2195 ;; literal enclosing point-min, or point-min itself.
2196 (or c-state-min-scan-pos
2197 (save-restriction
2198 (save-excursion
2199 (widen)
2200 (goto-char c-state-point-min-lit-start)
2201 (if (eq c-state-point-min-lit-type 'string)
2202 (forward-sexp)
2203 (forward-comment 1))
2204 (setq c-state-min-scan-pos (point))))))
2205
2206(defun c-state-mark-point-min-literal ()
2207 ;; Determine the properties of any literal containing POINT-MIN, setting the
2208 ;; variables `c-state-point-min-lit-type', `c-state-point-min-lit-start',
2209 ;; and `c-state-min-scan-pos' accordingly. The return value is meaningless.
2210 (let ((p-min (point-min))
2211 lit)
2212 (save-restriction
2213 (widen)
2214 (setq lit (c-state-literal-at p-min))
2215 (if lit
2216 (setq c-state-point-min-lit-type
2217 (save-excursion
2218 (goto-char (car lit))
2219 (cond
2220 ((looking-at c-block-comment-start-regexp) 'c)
2221 ((looking-at c-line-comment-starter) 'c++)
2222 (t 'string)))
2223 c-state-point-min-lit-start (car lit)
2224 c-state-min-scan-pos (cdr lit))
2225 (setq c-state-point-min-lit-type nil
2226 c-state-point-min-lit-start nil
2227 c-state-min-scan-pos p-min)))))
2228
2229
2230;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2231;; A variable which signals a brace dessert - helpful for reducing the number
2232;; of fruitless backward scans.
2233(defvar c-state-brace-pair-desert nil)
2234(make-variable-buffer-local 'c-state-brace-pair-desert)
e84efb70 2235;; Used only in `c-append-lower-brace-pair-to-state-cache'. It is set when
0ec1d2c5
AM
2236;; that defun has searched backwards for a brace pair and not found one. Its
2237;; value is either nil or a cons (PA . FROM), where PA is the position of the
2238;; enclosing opening paren/brace/bracket which bounds the backwards search (or
2239;; nil when at top level) and FROM is where the backward search started. It
2240;; is reset to nil in `c-invalidate-state-cache'.
2241
2242
2243;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2244;; Lowish level functions/macros which work directly on `c-state-cache', or a
2245;; list of like structure.
2246(defmacro c-state-cache-top-lparen (&optional cache)
2247 ;; Return the address of the top left brace/bracket/paren recorded in CACHE
2248 ;; (default `c-state-cache') (or nil).
2249 (let ((cash (or cache 'c-state-cache)))
2250 `(if (consp (car ,cash))
2251 (caar ,cash)
2252 (car ,cash))))
2253
2254(defmacro c-state-cache-top-paren (&optional cache)
2255 ;; Return the address of the latest brace/bracket/paren (whether left or
2256 ;; right) recorded in CACHE (default `c-state-cache') or nil.
2257 (let ((cash (or cache 'c-state-cache)))
2258 `(if (consp (car ,cash))
2259 (cdar ,cash)
2260 (car ,cash))))
2261
2262(defmacro c-state-cache-after-top-paren (&optional cache)
2263 ;; Return the position just after the latest brace/bracket/paren (whether
2264 ;; left or right) recorded in CACHE (default `c-state-cache') or nil.
2265 (let ((cash (or cache 'c-state-cache)))
2266 `(if (consp (car ,cash))
2267 (cdar ,cash)
2268 (and (car ,cash)
2269 (1+ (car ,cash))))))
2270
2271(defun c-get-cache-scan-pos (here)
2272 ;; From the state-cache, determine the buffer position from which we might
2273 ;; scan forward to HERE to update this cache. This position will be just
2274 ;; after a paren/brace/bracket recorded in the cache, if possible, otherwise
2275 ;; return the earliest position in the accessible region which isn't within
2276 ;; a literal. If the visible portion of the buffer is entirely within a
2277 ;; literal, return NIL.
2278 (let ((c c-state-cache) elt)
2279 ;(while (>= (or (c-state-cache-top-lparen c) 1) here)
2280 (while (and c
2281 (>= (c-state-cache-top-lparen c) here))
2282 (setq c (cdr c)))
2283
2284 (setq elt (car c))
2285 (cond
2286 ((consp elt)
2287 (if (> (cdr elt) here)
2288 (1+ (car elt))
2289 (cdr elt)))
2290 (elt (1+ elt))
2291 ((<= (c-state-get-min-scan-pos) here)
2292 (c-state-get-min-scan-pos))
2293 (t nil))))
2294
2295;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2296;; Variables which keep track of preprocessor constructs.
2297(defvar c-state-old-cpp-beg nil)
2298(make-variable-buffer-local 'c-state-old-cpp-beg)
2299(defvar c-state-old-cpp-end nil)
2300(make-variable-buffer-local 'c-state-old-cpp-end)
2301;; These are the limits of the macro containing point at the previous call of
2302;; `c-parse-state', or nil.
2303
2304;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
c7015153 2305;; Defuns which analyze the buffer, yet don't change `c-state-cache'.
0ec1d2c5
AM
2306(defun c-get-fallback-scan-pos (here)
2307 ;; Return a start position for building `c-state-cache' from
2308 ;; scratch. This will be at the top level, 2 defuns back.
0386b551
AM
2309 (save-excursion
2310 ;; Go back 2 bods, but ignore any bogus positions returned by
2311 ;; beginning-of-defun (i.e. open paren in column zero).
2312 (goto-char here)
2313 (let ((cnt 2))
2314 (while (not (or (bobp) (zerop cnt)))
0ec1d2c5 2315 (c-beginning-of-defun-1) ; Pure elisp BOD.
0386b551
AM
2316 (if (eq (char-after) ?\{)
2317 (setq cnt (1- cnt)))))
2318 (point)))
2319
c0209c2c
AM
2320(defun c-state-balance-parens-backwards (here- here+ top)
2321 ;; Return the position of the opening paren/brace/bracket before HERE- which
2322 ;; matches the outermost close p/b/b between HERE+ and TOP. Except when
2323 ;; there's a macro, HERE- and HERE+ are the same. Like this:
0ec1d2c5 2324 ;;
c0209c2c
AM
2325 ;; ............................................
2326 ;; | |
2327 ;; ( [ ( .........#macro.. ) ( ) ] )
2328 ;; ^ ^ ^ ^
2329 ;; | | | |
2330 ;; return HERE- HERE+ TOP
0ec1d2c5
AM
2331 ;;
2332 ;; If there aren't enough opening paren/brace/brackets, return the position
c0209c2c 2333 ;; of the outermost one found, or HERE- if there are none. If there are no
da6062e6 2334 ;; closing p/b/bs between HERE+ and TOP, return HERE-. HERE-/+ and TOP
c0209c2c
AM
2335 ;; must not be inside literals. Only the accessible portion of the buffer
2336 ;; will be scanned.
2337
2338 ;; PART 1: scan from `here+' up to `top', accumulating ")"s which enclose
2339 ;; `here'. Go round the next loop each time we pass over such a ")". These
2340 ;; probably match "("s before `here-'.
0ec1d2c5
AM
2341 (let (pos pa ren+1 lonely-rens)
2342 (save-excursion
2343 (save-restriction
2344 (narrow-to-region (point-min) top) ; This can move point, sometimes.
c0209c2c 2345 (setq pos here+)
0ec1d2c5
AM
2346 (c-safe
2347 (while
2348 (setq ren+1 (scan-lists pos 1 1)) ; might signal
2349 (setq lonely-rens (cons ren+1 lonely-rens)
2350 pos ren+1)))))
2351
c0209c2c 2352 ;; PART 2: Scan back before `here-' searching for the "("s
0ec1d2c5
AM
2353 ;; matching/mismatching the ")"s found above. We only need to direct the
2354 ;; caller to scan when we've encountered unmatched right parens.
c0209c2c
AM
2355 (setq pos here-)
2356 (when lonely-rens
2357 (c-safe
2358 (while
2359 (and lonely-rens ; actual values aren't used.
2360 (setq pa (scan-lists pos -1 1)))
2361 (setq pos pa)
2362 (setq lonely-rens (cdr lonely-rens)))))
2363 pos))
0ec1d2c5
AM
2364
2365(defun c-parse-state-get-strategy (here good-pos)
2366 ;; Determine the scanning strategy for adjusting `c-parse-state', attempting
53964682 2367 ;; to minimize the amount of scanning. HERE is the pertinent position in
0ec1d2c5
AM
2368 ;; the buffer, GOOD-POS is a position where `c-state-cache' (possibly with
2369 ;; its head trimmed) is known to be good, or nil if there is no such
2370 ;; position.
2371 ;;
2372 ;; The return value is a list, one of the following:
2373 ;;
2374 ;; o - ('forward CACHE-POS START-POINT) - scan forward from START-POINT,
2375 ;; which is not less than CACHE-POS.
2376 ;; o - ('backward CACHE-POS nil) - scan backwards (from HERE).
2377 ;; o - ('BOD nil START-POINT) - scan forwards from START-POINT, which is at the
2378 ;; top level.
2379 ;; o - ('IN-LIT nil nil) - point is inside the literal containing point-min.
2380 ;; , where CACHE-POS is the highest position recorded in `c-state-cache' at
2381 ;; or below HERE.
2382 (let ((cache-pos (c-get-cache-scan-pos here)) ; highest position below HERE in cache (or 1)
2383 BOD-pos ; position of 2nd BOD before HERE.
2384 strategy ; 'forward, 'backward, 'BOD, or 'IN-LIT.
2385 start-point
2386 how-far) ; putative scanning distance.
2387 (setq good-pos (or good-pos (c-state-get-min-scan-pos)))
2388 (cond
2389 ((< here (c-state-get-min-scan-pos))
2390 (setq strategy 'IN-LIT
2391 start-point nil
2392 cache-pos nil
2393 how-far 0))
2394 ((<= good-pos here)
2395 (setq strategy 'forward
2396 start-point (max good-pos cache-pos)
2397 how-far (- here start-point)))
2398 ((< (- good-pos here) (- here cache-pos)) ; FIXME!!! ; apply some sort of weighting.
2399 (setq strategy 'backward
2400 how-far (- good-pos here)))
2401 (t
2402 (setq strategy 'forward
2403 how-far (- here cache-pos)
2404 start-point cache-pos)))
2405
2406 ;; Might we be better off starting from the top level, two defuns back,
2407 ;; instead?
2408 (when (> how-far c-state-cache-too-far)
2409 (setq BOD-pos (c-get-fallback-scan-pos here)) ; somewhat EXPENSIVE!!!
2410 (if (< (- here BOD-pos) how-far)
2411 (setq strategy 'BOD
2412 start-point BOD-pos)))
2413
2414 (list
2415 strategy
2416 (and (memq strategy '(forward backward)) cache-pos)
2417 (and (memq strategy '(forward BOD)) start-point))))
2418
2419
2420;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2421;; Routines which change `c-state-cache' and associated values.
2422(defun c-renarrow-state-cache ()
2423 ;; The region (more precisely, point-min) has changed since we
2424 ;; calculated `c-state-cache'. Amend `c-state-cache' accordingly.
2425 (if (< (point-min) c-state-point-min)
2426 ;; If point-min has MOVED BACKWARDS then we drop the state completely.
2427 ;; It would be possible to do a better job here and recalculate the top
2428 ;; only.
2429 (progn
2430 (c-state-mark-point-min-literal)
2431 (setq c-state-cache nil
2432 c-state-cache-good-pos c-state-min-scan-pos
2433 c-state-brace-pair-desert nil))
2434
2435 ;; point-min has MOVED FORWARD.
2436
2437 ;; Is the new point-min inside a (different) literal?
2438 (unless (and c-state-point-min-lit-start ; at prev. point-min
2439 (< (point-min) (c-state-get-min-scan-pos)))
2440 (c-state-mark-point-min-literal))
2441
2442 ;; Cut off a bit of the tail from `c-state-cache'.
2443 (let ((ptr (cons nil c-state-cache))
2444 pa)
2445 (while (and (setq pa (c-state-cache-top-lparen (cdr ptr)))
2446 (>= pa (point-min)))
2447 (setq ptr (cdr ptr)))
2448
2449 (when (consp ptr)
2450 (if (eq (cdr ptr) c-state-cache)
2451 (setq c-state-cache nil
2452 c-state-cache-good-pos c-state-min-scan-pos)
2453 (setcdr ptr nil)
2454 (setq c-state-cache-good-pos (1+ (c-state-cache-top-lparen))))
2455 )))
2456
2457 (setq c-state-point-min (point-min)))
2458
2459(defun c-append-lower-brace-pair-to-state-cache (from &optional upper-lim)
2460 ;; If there is a brace pair preceding FROM in the buffer (not necessarily
2461 ;; immediately preceding), push a cons onto `c-state-cache' to represent it.
2462 ;; FROM must not be inside a literal. If UPPER-LIM is non-nil, we append
2463 ;; the highest brace pair whose "}" is below UPPER-LIM.
2464 ;;
2465 ;; Return non-nil when this has been done.
2466 ;;
2467 ;; This routine should be fast. Since it can get called a LOT, we maintain
2468 ;; `c-state-brace-pair-desert', a small cache of "failures", such that we
2469 ;; reduce the time wasted in repeated fruitless searches in brace deserts.
2470 (save-excursion
2471 (save-restriction
2472 (let ((bra from) ce ; Positions of "{" and "}".
2473 new-cons
2474 (cache-pos (c-state-cache-top-lparen)) ; might be nil.
2475 (macro-start-or-from
2476 (progn (goto-char from)
2477 (c-beginning-of-macro)
2478 (point))))
2479 (or upper-lim (setq upper-lim from))
2480
2481 ;; If we're essentially repeating a fruitless search, just give up.
2482 (unless (and c-state-brace-pair-desert
2483 (eq cache-pos (car c-state-brace-pair-desert))
2484 (<= from (cdr c-state-brace-pair-desert)))
2485 ;; Only search what we absolutely need to:
2486 (if (and c-state-brace-pair-desert
cd5495ff 2487 (eq cache-pos (car c-state-brace-pair-desert)))
0ec1d2c5
AM
2488 (narrow-to-region (cdr c-state-brace-pair-desert) (point-max)))
2489
2490 ;; In the next pair of nested loops, the inner one moves back past a
2491 ;; pair of (mis-)matching parens or brackets; the outer one moves
2492 ;; back over a sequence of unmatched close brace/paren/bracket each
2493 ;; time round.
2494 (while
2495 (progn
2496 (c-safe
2497 (while
2498 (and (setq ce (scan-lists bra -1 -1)) ; back past )/]/}; might signal
2499 (setq bra (scan-lists ce -1 1)) ; back past (/[/{; might signal
2500 (or (> ce upper-lim)
2501 (not (eq (char-after bra) ?\{))
2502 (and (goto-char bra)
2503 (c-beginning-of-macro)
2504 (< (point) macro-start-or-from))))))
2505 (and ce (< ce bra)))
2506 (setq bra ce)) ; If we just backed over an unbalanced closing
2507 ; brace, ignore it.
2508
2509 (if (and ce (< bra ce) (eq (char-after bra) ?\{))
2510 ;; We've found the desired brace-pair.
2511 (progn
2512 (setq new-cons (cons bra (1+ ce)))
2513 (cond
2514 ((consp (car c-state-cache))
2515 (setcar c-state-cache new-cons))
2516 ((and (numberp (car c-state-cache)) ; probably never happens
2517 (< ce (car c-state-cache)))
2518 (setcdr c-state-cache
2519 (cons new-cons (cdr c-state-cache))))
2520 (t (setq c-state-cache (cons new-cons c-state-cache)))))
2521
2522 ;; We haven't found a brace pair. Record this.
2523 (setq c-state-brace-pair-desert (cons cache-pos from))))))))
2524
2525(defsubst c-state-push-any-brace-pair (bra+1 macro-start-or-here)
2526 ;; If BRA+1 is nil, do nothing. Otherwise, BRA+1 is the buffer position
2527 ;; following a {, and that brace has a (mis-)matching } (or ]), and we
2528 ;; "push" "a" brace pair onto `c-state-cache'.
2529 ;;
2530 ;; Here "push" means overwrite the top element if it's itself a brace-pair,
2531 ;; otherwise push it normally.
2532 ;;
2533 ;; The brace pair we push is normally the one surrounding BRA+1, but if the
2534 ;; latter is inside a macro, not being a macro containing
2535 ;; MACRO-START-OR-HERE, we scan backwards through the buffer for a non-macro
2536 ;; base pair. This latter case is assumed to be rare.
2537 ;;
2538 ;; Note: POINT is not preserved in this routine.
2539 (if bra+1
2540 (if (or (> bra+1 macro-start-or-here)
2541 (progn (goto-char bra+1)
2542 (not (c-beginning-of-macro))))
2543 (setq c-state-cache
2544 (cons (cons (1- bra+1)
2545 (scan-lists bra+1 1 1))
2546 (if (consp (car c-state-cache))
2547 (cdr c-state-cache)
2548 c-state-cache)))
2549 ;; N.B. This defsubst codes one method for the simple, normal case,
2550 ;; and a more sophisticated, slower way for the general case. Don't
2551 ;; eliminate this defsubst - it's a speed optimisation.
2552 (c-append-lower-brace-pair-to-state-cache (1- bra+1)))))
2553
2554(defun c-append-to-state-cache (from)
2555 ;; Scan the buffer from FROM to (point-max), adding elements into
2556 ;; `c-state-cache' for braces etc. Return a candidate for
2557 ;; `c-state-cache-good-pos'.
2558 ;;
2559 ;; FROM must be after the latest brace/paren/bracket in `c-state-cache', if
2560 ;; any. Typically, it is immediately after it. It must not be inside a
2561 ;; literal.
2562 (let ((here-bol (c-point 'bol (point-max)))
2563 (macro-start-or-here
2564 (save-excursion (goto-char (point-max))
2565 (if (c-beginning-of-macro)
2566 (point)
2567 (point-max))))
2568 pa+1 ; pos just after an opening PAren (or brace).
2569 (ren+1 from) ; usually a pos just after an closing paREN etc.
2570 ; Is actually the pos. to scan for a (/{/[ from,
2571 ; which sometimes is after a silly )/}/].
2572 paren+1 ; Pos after some opening or closing paren.
2573 paren+1s ; A list of `paren+1's; used to determine a
2574 ; good-pos.
2575 bra+1 ce+1 ; just after L/R bra-ces.
2576 bra+1s ; list of OLD values of bra+1.
2577 mstart) ; start of a macro.
2578
2579 (save-excursion
2580 ;; Each time round the following loop, we enter a succesively deeper
2581 ;; level of brace/paren nesting. (Except sometimes we "continue at
2582 ;; the existing level".) `pa+1' is a pos inside an opening
2583 ;; brace/paren/bracket, usually just after it.
2584 (while
2585 (progn
2586 ;; Each time round the next loop moves forward over an opening then
2587 ;; a closing brace/bracket/paren. This loop is white hot, so it
2588 ;; plays ugly tricks to go fast. DON'T PUT ANYTHING INTO THIS
2589 ;; LOOP WHICH ISN'T ABSOLUTELY NECESSARY!!! It terminates when a
2590 ;; call of `scan-lists' signals an error, which happens when there
2591 ;; are no more b/b/p's to scan.
2592 (c-safe
2593 (while t
2594 (setq pa+1 (scan-lists ren+1 1 -1) ; Into (/{/[; might signal
2595 paren+1s (cons pa+1 paren+1s))
2596 (setq ren+1 (scan-lists pa+1 1 1)) ; Out of )/}/]; might signal
2597 (if (and (eq (char-before pa+1) ?{)) ; Check for a macro later.
2598 (setq bra+1 pa+1))
2599 (setcar paren+1s ren+1)))
2600
2601 (if (and pa+1 (> pa+1 ren+1))
2602 ;; We've just entered a deeper nesting level.
2603 (progn
2604 ;; Insert the brace pair (if present) and the single open
2605 ;; paren/brace/bracket into `c-state-cache' It cannot be
2606 ;; inside a macro, except one around point, because of what
2607 ;; `c-neutralize-syntax-in-CPP' has done.
2608 (c-state-push-any-brace-pair bra+1 macro-start-or-here)
2609 ;; Insert the opening brace/bracket/paren position.
2610 (setq c-state-cache (cons (1- pa+1) c-state-cache))
2611 ;; Clear admin stuff for the next more nested part of the scan.
2612 (setq ren+1 pa+1 pa+1 nil bra+1 nil bra+1s nil)
2613 t) ; Carry on the loop
2614
2615 ;; All open p/b/b's at this nesting level, if any, have probably
2616 ;; been closed by matching/mismatching ones. We're probably
2617 ;; finished - we just need to check for having found an
2618 ;; unmatched )/}/], which we ignore. Such a )/}/] can't be in a
2619 ;; macro, due the action of `c-neutralize-syntax-in-CPP'.
2620 (c-safe (setq ren+1 (scan-lists ren+1 1 1)))))) ; acts as loop control.
2621
2622 ;; Record the final, innermost, brace-pair if there is one.
2623 (c-state-push-any-brace-pair bra+1 macro-start-or-here)
2624
2625 ;; Determine a good pos
2626 (while (and (setq paren+1 (car paren+1s))
2627 (> (if (> paren+1 macro-start-or-here)
2628 paren+1
2629 (goto-char paren+1)
2630 (setq mstart (and (c-beginning-of-macro)
2631 (point)))
2632 (or mstart paren+1))
2633 here-bol))
2634 (setq paren+1s (cdr paren+1s)))
2635 (cond
2636 ((and paren+1 mstart)
2637 (min paren+1 mstart))
2638 (paren+1)
2639 (t from)))))
2640
2641(defun c-remove-stale-state-cache (good-pos pps-point)
2642 ;; Remove stale entries from the `c-cache-state', i.e. those which will
2643 ;; not be in it when it is amended for position (point-max).
2644 ;; Additionally, the "outermost" open-brace entry before (point-max)
2645 ;; will be converted to a cons if the matching close-brace is scanned.
2646 ;;
2647 ;; GOOD-POS is a "maximal" "safe position" - there must be no open
2648 ;; parens/braces/brackets between GOOD-POS and (point-max).
2649 ;;
2650 ;; As a second thing, calculate the result of parse-partial-sexp at
2651 ;; PPS-POINT, w.r.t. GOOD-POS. The motivation here is that
2652 ;; `c-state-cache-good-pos' may become PPS-POINT, but the caller may need to
2653 ;; adjust it to get outside a string/comment. (Sorry about this! The code
2654 ;; needs to be FAST).
2655 ;;
2656 ;; Return a list (GOOD-POS SCAN-BACK-POS PPS-STATE), where
2657 ;; o - GOOD-POS is a position where the new value `c-state-cache' is known
2658 ;; to be good (we aim for this to be as high as possible);
2659 ;; o - SCAN-BACK-POS, if not nil, indicates there may be a brace pair
2660 ;; preceding POS which needs to be recorded in `c-state-cache'. It is a
2661 ;; position to scan backwards from.
2662 ;; o - PPS-STATE is the parse-partial-sexp state at PPS-POINT.
2663 (save-restriction
2664 (narrow-to-region 1 (point-max))
2665 (save-excursion
eb2f6eeb 2666 (let* ((in-macro-start ; start of macro containing (point-max) or nil.
0ec1d2c5
AM
2667 (save-excursion
2668 (goto-char (point-max))
2669 (and (c-beginning-of-macro)
2670 (point))))
2671 (good-pos-actual-macro-start ; Start of macro containing good-pos
2672 ; or nil
2673 (and (< good-pos (point-max))
2674 (save-excursion
2675 (goto-char good-pos)
2676 (and (c-beginning-of-macro)
2677 (point)))))
2678 (good-pos-actual-macro-end ; End of this macro, (maybe
2679 ; (point-max)), or nil.
2680 (and good-pos-actual-macro-start
2681 (save-excursion
2682 (goto-char good-pos-actual-macro-start)
2683 (c-end-of-macro)
2684 (point))))
2685 pps-state ; Will be 9 or 10 elements long.
2686 pos
2687 upper-lim ; ,beyond which `c-state-cache' entries are removed
2688 scan-back-pos
2689 pair-beg pps-point-state target-depth)
2690
2691 ;; Remove entries beyond (point-max). Also remove any entries inside
2692 ;; a macro, unless (point-max) is in the same macro.
2693 (setq upper-lim
2694 (if (or (null c-state-old-cpp-beg)
2695 (and (> (point-max) c-state-old-cpp-beg)
2696 (< (point-max) c-state-old-cpp-end)))
2697 (point-max)
2698 (min (point-max) c-state-old-cpp-beg)))
657071fc 2699 (while (and c-state-cache (>= (c-state-cache-top-lparen) upper-lim))
0ec1d2c5
AM
2700 (setq c-state-cache (cdr c-state-cache)))
2701 ;; If `upper-lim' is inside the last recorded brace pair, remove its
2702 ;; RBrace and indicate we'll need to search backwards for a previous
2703 ;; brace pair.
2704 (when (and c-state-cache
2705 (consp (car c-state-cache))
2706 (> (cdar c-state-cache) upper-lim))
2707 (setcar c-state-cache (caar c-state-cache))
2708 (setq scan-back-pos (car c-state-cache)))
2709
2710 ;; The next loop jumps forward out of a nested level of parens each
2711 ;; time round; the corresponding elements in `c-state-cache' are
2712 ;; removed. `pos' is just after the brace-pair or the open paren at
2713 ;; (car c-state-cache). There can be no open parens/braces/brackets
2714 ;; between `good-pos'/`good-pos-actual-macro-start' and (point-max),
2715 ;; due to the interface spec to this function.
dd21b621 2716 (setq pos (if (and good-pos-actual-macro-end
eb2f6eeb
AM
2717 (not (eq good-pos-actual-macro-start
2718 in-macro-start)))
0ec1d2c5
AM
2719 (1+ good-pos-actual-macro-end) ; get outside the macro as
2720 ; marked by a `category' text property.
2721 good-pos))
2722 (goto-char pos)
2723 (while (and c-state-cache
2724 (< (point) (point-max)))
2725 (cond
2726 ((null pps-state) ; first time through
2727 (setq target-depth -1))
2728 ((eq (car pps-state) target-depth) ; found closing ),},]
2729 (setq target-depth (1- (car pps-state))))
2730 ;; Do nothing when we've merely reached pps-point.
2731 )
2732
2733 ;; Scan!
2734 (setq pps-state
2735 (parse-partial-sexp
2736 (point) (if (< (point) pps-point) pps-point (point-max))
2737 target-depth
2738 nil pps-state))
2739
2740 (if (= (point) pps-point)
2741 (setq pps-point-state pps-state))
2742
2743 (when (eq (car pps-state) target-depth)
2744 (setq pos (point)) ; POS is now just after an R-paren/brace.
2745 (cond
2746 ((and (consp (car c-state-cache))
2747 (eq (point) (cdar c-state-cache)))
2748 ;; We've just moved out of the paren pair containing the brace-pair
2749 ;; at (car c-state-cache). `pair-beg' is where the open paren is,
2750 ;; and is potentially where the open brace of a cons in
2751 ;; c-state-cache will be.
2752 (setq pair-beg (car-safe (cdr c-state-cache))
2753 c-state-cache (cdr-safe (cdr c-state-cache)))) ; remove {}pair + containing Lparen.
2754 ((numberp (car c-state-cache))
2755 (setq pair-beg (car c-state-cache)
2756 c-state-cache (cdr c-state-cache))) ; remove this
2757 ; containing Lparen
2758 ((numberp (cadr c-state-cache))
2759 (setq pair-beg (cadr c-state-cache)
2760 c-state-cache (cddr c-state-cache))) ; Remove a paren pair
2761 ; together with enclosed brace pair.
2762 ;; (t nil) ; Ignore an unmated Rparen.
2763 )))
2764
2765 (if (< (point) pps-point)
2766 (setq pps-state (parse-partial-sexp (point) pps-point
2767 nil nil ; TARGETDEPTH, STOPBEFORE
2768 pps-state)))
2769
2770 ;; If the last paren pair we moved out of was actually a brace pair,
2771 ;; insert it into `c-state-cache'.
2772 (when (and pair-beg (eq (char-after pair-beg) ?{))
2773 (if (consp (car-safe c-state-cache))
2774 (setq c-state-cache (cdr c-state-cache)))
2775 (setq c-state-cache (cons (cons pair-beg pos)
2776 c-state-cache)))
2777
2778 (list pos scan-back-pos pps-state)))))
2779
2780(defun c-remove-stale-state-cache-backwards (here cache-pos)
2781 ;; Strip stale elements of `c-state-cache' by moving backwards through the
2782 ;; buffer, and inform the caller of the scenario detected.
2783 ;;
2784 ;; HERE is the position we're setting `c-state-cache' for.
2785 ;; CACHE-POS is just after the latest recorded position in `c-state-cache'
2786 ;; before HERE, or a position at or near point-min which isn't in a
2787 ;; literal.
2788 ;;
2789 ;; This function must only be called only when (> `c-state-cache-good-pos'
2790 ;; HERE). Usually the gap between CACHE-POS and HERE is large. It is thus
53964682 2791 ;; optimised to eliminate (or minimize) scanning between these two
0ec1d2c5
AM
2792 ;; positions.
2793 ;;
2794 ;; Return a three element list (GOOD-POS SCAN-BACK-POS FWD-FLAG), where:
2795 ;; o - GOOD-POS is a "good position", where `c-state-cache' is valid, or
2796 ;; could become so after missing elements are inserted into
2797 ;; `c-state-cache'. This is JUST AFTER an opening or closing
2798 ;; brace/paren/bracket which is already in `c-state-cache' or just before
2799 ;; one otherwise. exceptionally (when there's no such b/p/b handy) the BOL
2800 ;; before `here''s line, or the start of the literal containing it.
2801 ;; o - SCAN-BACK-POS, if non-nil, indicates there may be a brace pair
2802 ;; preceding POS which isn't recorded in `c-state-cache'. It is a position
2803 ;; to scan backwards from.
2804 ;; o - FWD-FLAG, if non-nil, indicates there may be parens/braces between
2805 ;; POS and HERE which aren't recorded in `c-state-cache'.
2806 ;;
2807 ;; The comments in this defun use "paren" to mean parenthesis or square
2808 ;; bracket (as contrasted with a brace), and "(" and ")" likewise.
2809 ;;
2810 ;; . {..} (..) (..) ( .. { } ) (...) ( .... . ..)
2811 ;; | | | | | |
2812 ;; CP E here D C good
2813 (let ((pos c-state-cache-good-pos)
2814 pa ren ; positions of "(" and ")"
2815 dropped-cons ; whether the last element dropped from `c-state-cache'
2816 ; was a cons (representing a brace-pair)
2817 good-pos ; see above.
2818 lit ; (START . END) of a literal containing some point.
2819 here-lit-start here-lit-end ; bounds of literal containing `here'
2820 ; or `here' itself.
c0209c2c 2821 here- here+ ; start/end of macro around HERE, or HERE
0ec1d2c5 2822 (here-bol (c-point 'bol here))
c400c4d7 2823 (too-far-back (max (- here c-state-cache-too-far) (point-min))))
0ec1d2c5
AM
2824
2825 ;; Remove completely irrelevant entries from `c-state-cache'.
2826 (while (and c-state-cache
2827 (>= (setq pa (c-state-cache-top-lparen)) here))
2828 (setq dropped-cons (consp (car c-state-cache)))
2829 (setq c-state-cache (cdr c-state-cache))
2830 (setq pos pa))
2831 ;; At this stage, (> pos here);
2832 ;; (< (c-state-cache-top-lparen) here) (or is nil).
2833
c0209c2c
AM
2834 (cond
2835 ((and (consp (car c-state-cache))
2836 (> (cdar c-state-cache) here))
2837 ;; CASE 1: The top of the cache is a brace pair which now encloses
2838 ;; `here'. As good-pos, return the address. of the "{". Since we've no
2839 ;; knowledge of what's inside these braces, we have no alternative but
2840 ;; to direct the caller to scan the buffer from the opening brace.
2841 (setq pos (caar c-state-cache))
2842 (setcar c-state-cache pos)
2843 (list (1+ pos) pos t)) ; return value. We've just converted a brace pair
2844 ; entry into a { entry, so the caller needs to
2845 ; search for a brace pair before the {.
2846
2847 ;; `here' might be inside a literal. Check for this.
2848 ((progn
2849 (setq lit (c-state-literal-at here)
2850 here-lit-start (or (car lit) here)
2851 here-lit-end (or (cdr lit) here))
2852 ;; Has `here' just "newly entered" a macro?
2853 (save-excursion
2854 (goto-char here-lit-start)
2855 (if (and (c-beginning-of-macro)
2856 (or (null c-state-old-cpp-beg)
2857 (not (= (point) c-state-old-cpp-beg))))
2858 (progn
2859 (setq here- (point))
2860 (c-end-of-macro)
2861 (setq here+ (point)))
2862 (setq here- here-lit-start
2863 here+ here-lit-end)))
2864
2865 ;; `here' might be nested inside any depth of parens (or brackets but
2866 ;; not braces). Scan backwards to find the outermost such opening
2867 ;; paren, if there is one. This will be the scan position to return.
2868 (save-restriction
2869 (narrow-to-region cache-pos (point-max))
2870 (setq pos (c-state-balance-parens-backwards here- here+ pos)))
2871 nil)) ; for the cond
2872
2873 ((< pos here-lit-start)
2874 ;; CASE 2: Address of outermost ( or [ which now encloses `here', but
2875 ;; didn't enclose the (previous) `c-state-cache-good-pos'. If there is
2876 ;; a brace pair preceding this, it will already be in `c-state-cache',
2877 ;; unless there was a brace pair after it, i.e. there'll only be one to
2878 ;; scan for if we've just deleted one.
2879 (list pos (and dropped-cons pos) t)) ; Return value.
2880
2881 ;; `here' isn't enclosed in a (previously unrecorded) bracket/paren.
2882 ;; Further forward scanning isn't needed, but we still need to find a
2883 ;; GOOD-POS. Step out of all enclosing "("s on HERE's line.
2884 ((progn
0ec1d2c5
AM
2885 (save-restriction
2886 (narrow-to-region here-bol (point-max))
2887 (setq pos here-lit-start)
2888 (c-safe (while (setq pa (scan-lists pos -1 1))
2889 (setq pos pa)))) ; might signal
c0209c2c
AM
2890 nil)) ; for the cond
2891
2892 ((setq ren (c-safe-scan-lists pos -1 -1 too-far-back))
2893 ;; CASE 3: After a }/)/] before `here''s BOL.
2894 (list (1+ ren) (and dropped-cons pos) nil)) ; Return value
2895
2896 (t
2897 ;; CASE 4; Best of a bad job: BOL before `here-bol', or beginning of
2898 ;; literal containing it.
2899 (setq good-pos (c-state-lit-beg (c-point 'bopl here-bol)))
2900 (list good-pos (and dropped-cons good-pos) nil)))))
0ec1d2c5
AM
2901
2902
2903;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2904;; Externally visible routines.
2905
2906(defun c-state-cache-init ()
2907 (setq c-state-cache nil
2908 c-state-cache-good-pos 1
2909 c-state-nonlit-pos-cache nil
2910 c-state-nonlit-pos-cache-limit 1
2911 c-state-brace-pair-desert nil
2912 c-state-point-min 1
2913 c-state-point-min-lit-type nil
2914 c-state-point-min-lit-start nil
2915 c-state-min-scan-pos 1
2916 c-state-old-cpp-beg nil
2917 c-state-old-cpp-end nil)
2918 (c-state-mark-point-min-literal))
2919
e84efb70
AM
2920;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2921;; Debugging routines to dump `c-state-cache' in a "replayable" form.
2922;; (defmacro c-sc-de (elt) ; "c-state-cache-dump-element"
2923;; `(format ,(concat "(setq " (symbol-name elt) " %s) ") ,elt))
2924;; (defmacro c-sc-qde (elt) ; "c-state-cache-quote-dump-element"
2925;; `(format ,(concat "(setq " (symbol-name elt) " '%s) ") ,elt))
2926;; (defun c-state-dump ()
2927;; ;; For debugging.
2928;; ;(message
2929;; (concat
2930;; (c-sc-qde c-state-cache)
2931;; (c-sc-de c-state-cache-good-pos)
2932;; (c-sc-qde c-state-nonlit-pos-cache)
2933;; (c-sc-de c-state-nonlit-pos-cache-limit)
2934;; (c-sc-qde c-state-brace-pair-desert)
2935;; (c-sc-de c-state-point-min)
2936;; (c-sc-de c-state-point-min-lit-type)
2937;; (c-sc-de c-state-point-min-lit-start)
2938;; (c-sc-de c-state-min-scan-pos)
2939;; (c-sc-de c-state-old-cpp-beg)
2940;; (c-sc-de c-state-old-cpp-end)))
2941;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2942
0ec1d2c5
AM
2943(defun c-invalidate-state-cache-1 (here)
2944 ;; Invalidate all info on `c-state-cache' that applies to the buffer at HERE
2945 ;; or higher and set `c-state-cache-good-pos' accordingly. The cache is
2946 ;; left in a consistent state.
2947 ;;
2948 ;; This is much like `c-whack-state-after', but it never changes a paren
2949 ;; pair element into an open paren element. Doing that would mean that the
2950 ;; new open paren wouldn't have the required preceding paren pair element.
785eecbb 2951 ;;
0ec1d2c5
AM
2952 ;; This function is called from c-after-change.
2953
2954 ;; The cache of non-literals:
2955 (if (< here c-state-nonlit-pos-cache-limit)
2956 (setq c-state-nonlit-pos-cache-limit here))
2957
2958 ;; `c-state-cache':
2959 ;; Case 1: if `here' is in a literal containing point-min, everything
2960 ;; becomes (or is already) nil.
2961 (if (or (null c-state-cache-good-pos)
2962 (< here (c-state-get-min-scan-pos)))
2963 (setq c-state-cache nil
2964 c-state-cache-good-pos nil
2965 c-state-min-scan-pos nil)
2966
c400c4d7
SM
2967 ;; Truncate `c-state-cache' and set `c-state-cache-good-pos' to a value
2968 ;; below `here'. To maintain its consistency, we may need to insert a new
2969 ;; brace pair.
0ec1d2c5
AM
2970 (let ((here-bol (c-point 'bol here))
2971 too-high-pa ; recorded {/(/[ next above here, or nil.
2972 dropped-cons ; was the last removed element a brace pair?
2973 pa)
2974 ;; The easy bit - knock over-the-top bits off `c-state-cache'.
2975 (while (and c-state-cache
2976 (>= (setq pa (c-state-cache-top-paren)) here))
2977 (setq dropped-cons (consp (car c-state-cache))
2978 too-high-pa (c-state-cache-top-lparen)
2979 c-state-cache (cdr c-state-cache)))
2980
2981 ;; Do we need to add in an earlier brace pair, having lopped one off?
2982 (if (and dropped-cons
2983 (< too-high-pa (+ here c-state-cache-too-far)))
2984 (c-append-lower-brace-pair-to-state-cache too-high-pa here-bol))
2985 (setq c-state-cache-good-pos (or (c-state-cache-after-top-paren)
2986 (c-state-get-min-scan-pos)))))
2987
2988 ;; The brace-pair desert marker:
2989 (when (car c-state-brace-pair-desert)
2990 (if (< here (car c-state-brace-pair-desert))
2991 (setq c-state-brace-pair-desert nil)
2992 (if (< here (cdr c-state-brace-pair-desert))
2993 (setcdr c-state-brace-pair-desert here)))))
2994
2995(defun c-parse-state-1 ()
2996 ;; Find and record all noteworthy parens between some good point earlier in
2997 ;; the file and point. That good point is at least the beginning of the
2998 ;; top-level construct we are in, or the beginning of the preceding
2999 ;; top-level construct if we aren't in one.
3000 ;;
3001 ;; The returned value is a list of the noteworthy parens with the last one
3002 ;; first. If an element in the list is an integer, it's the position of an
3003 ;; open paren (of any type) which has not been closed before the point. If
3004 ;; an element is a cons, it gives the position of a closed BRACE paren
3005 ;; pair[*]; the car is the start brace position and the cdr is the position
3006 ;; following the closing brace. Only the last closed brace paren pair
3007 ;; before each open paren and before the point is recorded, and thus the
3008 ;; state never contains two cons elements in succession. When a close brace
3009 ;; has no matching open brace (e.g., the matching brace is outside the
3010 ;; visible region), it is not represented in the returned value.
3011 ;;
3012 ;; [*] N.B. The close "brace" might be a mismatching close bracket or paren.
3013 ;; This defun explicitly treats mismatching parens/braces/brackets as
3014 ;; matching. It is the open brace which makes it a "brace" pair.
3015 ;;
3016 ;; If POINT is within a macro, open parens and brace pairs within
3017 ;; THIS macro MIGHT be recorded. This depends on whether their
3018 ;; syntactic properties have been suppressed by
3019 ;; `c-neutralize-syntax-in-CPP'. This might need fixing (2008-12-11).
d9e94c22
MS
3020 ;;
3021 ;; Currently no characters which are given paren syntax with the
3022 ;; syntax-table property are recorded, i.e. angle bracket arglist
3023 ;; parens are never present here. Note that this might change.
3024 ;;
0386b551 3025 ;; BUG: This function doesn't cope entirely well with unbalanced
0ec1d2c5
AM
3026 ;; parens in macros. (2008-12-11: this has probably been resolved
3027 ;; by the function `c-neutralize-syntax-in-CPP'.) E.g. in the
3028 ;; following case the brace before the macro isn't balanced with the
3029 ;; one after it:
0386b551
AM
3030 ;;
3031 ;; {
3032 ;; #define X {
3033 ;; }
3034 ;;
0ec1d2c5
AM
3035 ;; Note to maintainers: this function DOES get called with point
3036 ;; within comments and strings, so don't assume it doesn't!
3037 ;;
0386b551 3038 ;; This function might do hidden buffer changes.
0ec1d2c5
AM
3039 (let* ((here (point))
3040 (here-bopl (c-point 'bopl))
3041 strategy ; 'forward, 'backward etc..
3042 ;; Candidate positions to start scanning from:
3043 cache-pos ; highest position below HERE already existing in
3044 ; cache (or 1).
3045 good-pos
3046 start-point
3047 bopl-state
3048 res
3049 scan-backward-pos scan-forward-p) ; used for 'backward.
3050 ;; If POINT-MIN has changed, adjust the cache
3051 (unless (= (point-min) c-state-point-min)
3052 (c-renarrow-state-cache))
3053
3054 ;; Strategy?
3055 (setq res (c-parse-state-get-strategy here c-state-cache-good-pos)
3056 strategy (car res)
3057 cache-pos (cadr res)
3058 start-point (nth 2 res))
3059
3060 (when (eq strategy 'BOD)
3061 (setq c-state-cache nil
3062 c-state-cache-good-pos start-point))
3063
3064 ;; SCAN!
3065 (save-restriction
3066 (cond
3067 ((memq strategy '(forward BOD))
0386b551 3068 (narrow-to-region (point-min) here)
0ec1d2c5
AM
3069 (setq res (c-remove-stale-state-cache start-point here-bopl))
3070 (setq cache-pos (car res)
3071 scan-backward-pos (cadr res)
3072 bopl-state (car (cddr res))) ; will be nil if (< here-bopl
3073 ; start-point)
3074 (if scan-backward-pos
3075 (c-append-lower-brace-pair-to-state-cache scan-backward-pos))
3076 (setq good-pos
3077 (c-append-to-state-cache cache-pos))
3078 (setq c-state-cache-good-pos
3079 (if (and bopl-state
3080 (< good-pos (- here c-state-cache-too-far)))
3081 (c-state-cache-non-literal-place here-bopl bopl-state)
3082 good-pos)))
3083
3084 ((eq strategy 'backward)
3085 (setq res (c-remove-stale-state-cache-backwards here cache-pos)
3086 good-pos (car res)
3087 scan-backward-pos (cadr res)
3088 scan-forward-p (car (cddr res)))
3089 (if scan-backward-pos
3090 (c-append-lower-brace-pair-to-state-cache
3091 scan-backward-pos))
3092 (setq c-state-cache-good-pos
3093 (if scan-forward-p
3094 (progn (narrow-to-region (point-min) here)
3095 (c-append-to-state-cache good-pos))
3096
3097 (c-get-cache-scan-pos good-pos))))
3098
3099 (t ; (eq strategy 'IN-LIT)
3100 (setq c-state-cache nil
3101 c-state-cache-good-pos nil)))))
3102
3103 c-state-cache)
3104
3105(defun c-invalidate-state-cache (here)
3106 ;; This is a wrapper over `c-invalidate-state-cache-1'.
3107 ;;
3108 ;; It suppresses the syntactic effect of the < and > (template) brackets and
3109 ;; of all parens in preprocessor constructs, except for any such construct
3110 ;; containing point. We can then call `c-invalidate-state-cache-1' without
3111 ;; worrying further about macros and template delimiters.
3112 (c-with-<->-as-parens-suppressed
3e8f7d91
AM
3113 (if (and c-state-old-cpp-beg
3114 (< c-state-old-cpp-beg here))
0ec1d2c5 3115 (c-with-all-but-one-cpps-commented-out
3e8f7d91
AM
3116 c-state-old-cpp-beg
3117 (min c-state-old-cpp-end here)
0ec1d2c5
AM
3118 (c-invalidate-state-cache-1 here))
3119 (c-with-cpps-commented-out
3120 (c-invalidate-state-cache-1 here)))))
3121
3122(defun c-parse-state ()
3123 ;; This is a wrapper over `c-parse-state-1'. See that function for a
3124 ;; description of the functionality and return value.
3125 ;;
3126 ;; It suppresses the syntactic effect of the < and > (template) brackets and
3127 ;; of all parens in preprocessor constructs, except for any such construct
3128 ;; containing point. We can then call `c-parse-state-1' without worrying
3129 ;; further about macros and template delimiters.
3130 (let (here-cpp-beg here-cpp-end)
3131 (save-excursion
3132 (when (c-beginning-of-macro)
3133 (setq here-cpp-beg (point))
3134 (unless
3135 (> (setq here-cpp-end (c-syntactic-end-of-macro))
3136 here-cpp-beg)
3137 (setq here-cpp-beg nil here-cpp-end nil))))
3138 ;; FIXME!!! Put in a `condition-case' here to protect the integrity of the
3139 ;; subsystem.
3140 (prog1
3141 (c-with-<->-as-parens-suppressed
3142 (if (and here-cpp-beg (> here-cpp-end here-cpp-beg))
3143 (c-with-all-but-one-cpps-commented-out
3144 here-cpp-beg here-cpp-end
3145 (c-parse-state-1))
3146 (c-with-cpps-commented-out
3147 (c-parse-state-1))))
3e8f7d91
AM
3148 (setq c-state-old-cpp-beg (and here-cpp-beg (copy-marker here-cpp-beg t))
3149 c-state-old-cpp-end (and here-cpp-end (copy-marker here-cpp-end t)))
3150 )))
0ec1d2c5
AM
3151
3152;; Debug tool to catch cache inconsistencies. This is called from
3153;; 000tests.el.
a66cd3ee
MS
3154(defvar c-debug-parse-state nil)
3155(unless (fboundp 'c-real-parse-state)
3156 (fset 'c-real-parse-state (symbol-function 'c-parse-state)))
3157(cc-bytecomp-defun c-real-parse-state)
cd5495ff
AM
3158
3159(defvar c-parse-state-state nil)
3160(defun c-record-parse-state-state ()
3161 (setq c-parse-state-state
3162 (mapcar
3163 (lambda (arg)
3164 (cons arg (symbol-value arg)))
3165 '(c-state-cache
3166 c-state-cache-good-pos
3167 c-state-nonlit-pos-cache
3168 c-state-nonlit-pos-cache-limit
3169 c-state-brace-pair-desert
3170 c-state-point-min
3171 c-state-point-min-lit-type
3172 c-state-point-min-lit-start
3173 c-state-min-scan-pos
3174 c-state-old-cpp-beg
3175 c-state-old-cpp-end))))
3176(defun c-replay-parse-state-state ()
3177 (message
3178 (concat "(setq "
3179 (mapconcat
3180 (lambda (arg)
3181 (format "%s %s%s" (car arg) (if (atom (cdr arg)) "" "'") (cdr arg)))
3182 c-parse-state-state " ")
3183 ")")))
3184
a66cd3ee 3185(defun c-debug-parse-state ()
0ec1d2c5 3186 (let ((here (point)) (res1 (c-real-parse-state)) res2)
0386b551 3187 (let ((c-state-cache nil)
0ec1d2c5
AM
3188 (c-state-cache-good-pos 1)
3189 (c-state-nonlit-pos-cache nil)
3190 (c-state-nonlit-pos-cache-limit 1)
3191 (c-state-brace-pair-desert nil)
3192 (c-state-point-min 1)
3193 (c-state-point-min-lit-type nil)
3194 (c-state-point-min-lit-start nil)
3195 (c-state-min-scan-pos 1)
3196 (c-state-old-cpp-beg nil)
3197 (c-state-old-cpp-end nil))
a66cd3ee
MS
3198 (setq res2 (c-real-parse-state)))
3199 (unless (equal res1 res2)
0386b551
AM
3200 ;; The cache can actually go further back due to the ad-hoc way
3201 ;; the first paren is found, so try to whack off a bit of its
3202 ;; start before complaining.
cd5495ff
AM
3203 ;; (save-excursion
3204 ;; (goto-char (or (c-least-enclosing-brace res2) (point)))
3205 ;; (c-beginning-of-defun-1)
3206 ;; (while (not (or (bobp) (eq (char-after) ?{)))
3207 ;; (c-beginning-of-defun-1))
3208 ;; (unless (equal (c-whack-state-before (point) res1) res2)
3209 ;; (message (concat "c-parse-state inconsistency at %s: "
3210 ;; "using cache: %s, from scratch: %s")
3211 ;; here res1 res2)))
3212 (message (concat "c-parse-state inconsistency at %s: "
3213 "using cache: %s, from scratch: %s")
3214 here res1 res2)
3215 (message "Old state:")
3216 (c-replay-parse-state-state))
3217 (c-record-parse-state-state)
a66cd3ee 3218 res1))
0ec1d2c5 3219
a66cd3ee
MS
3220(defun c-toggle-parse-state-debug (&optional arg)
3221 (interactive "P")
3222 (setq c-debug-parse-state (c-calculate-state arg c-debug-parse-state))
3223 (fset 'c-parse-state (symbol-function (if c-debug-parse-state
3224 'c-debug-parse-state
3225 'c-real-parse-state)))
3226 (c-keep-region-active))
0386b551
AM
3227(when c-debug-parse-state
3228 (c-toggle-parse-state-debug 1))
a66cd3ee 3229
0ec1d2c5 3230\f
d9e94c22
MS
3231(defun c-whack-state-before (bufpos paren-state)
3232 ;; Whack off any state information from PAREN-STATE which lies
3233 ;; before BUFPOS. Not destructive on PAREN-STATE.
d9e94c22
MS
3234 (let* ((newstate (list nil))
3235 (ptr newstate)
3236 car)
3237 (while paren-state
3238 (setq car (car paren-state)
3239 paren-state (cdr paren-state))
3240 (if (< (if (consp car) (car car) car) bufpos)
3241 (setq paren-state nil)
3242 (setcdr ptr (list car))
3243 (setq ptr (cdr ptr))))
3244 (cdr newstate)))
3245
3246(defun c-whack-state-after (bufpos paren-state)
3247 ;; Whack off any state information from PAREN-STATE which lies at or
3248 ;; after BUFPOS. Not destructive on PAREN-STATE.
d9e94c22
MS
3249 (catch 'done
3250 (while paren-state
3251 (let ((car (car paren-state)))
3252 (if (consp car)
3253 ;; just check the car, because in a balanced brace
3254 ;; expression, it must be impossible for the corresponding
3255 ;; close brace to be before point, but the open brace to
3256 ;; be after.
3257 (if (<= bufpos (car car))
3258 nil ; whack it off
3259 (if (< bufpos (cdr car))
3260 ;; its possible that the open brace is before
3261 ;; bufpos, but the close brace is after. In that
3262 ;; case, convert this to a non-cons element. The
3263 ;; rest of the state is before bufpos, so we're
3264 ;; done.
3265 (throw 'done (cons (car car) (cdr paren-state)))
3266 ;; we know that both the open and close braces are
3267 ;; before bufpos, so we also know that everything else
3268 ;; on state is before bufpos.
3269 (throw 'done paren-state)))
3270 (if (<= bufpos car)
3271 nil ; whack it off
3272 ;; it's before bufpos, so everything else should too.
3273 (throw 'done paren-state)))
3274 (setq paren-state (cdr paren-state)))
3275 nil)))
3276
3277(defun c-most-enclosing-brace (paren-state &optional bufpos)
3278 ;; Return the bufpos of the innermost enclosing open paren before
0386b551 3279 ;; bufpos, or nil if none was found.
d9e94c22
MS
3280 (let (enclosingp)
3281 (or bufpos (setq bufpos 134217727))
3282 (while paren-state
3283 (setq enclosingp (car paren-state)
3284 paren-state (cdr paren-state))
3285 (if (or (consp enclosingp)
3286 (>= enclosingp bufpos))
3287 (setq enclosingp nil)
d9e94c22
MS
3288 (setq paren-state nil)))
3289 enclosingp))
3290
0386b551
AM
3291(defun c-least-enclosing-brace (paren-state)
3292 ;; Return the bufpos of the outermost enclosing open paren, or nil
3293 ;; if none was found.
d9e94c22 3294 (let (pos elem)
d9e94c22
MS
3295 (while paren-state
3296 (setq elem (car paren-state)
3297 paren-state (cdr paren-state))
0386b551
AM
3298 (if (integerp elem)
3299 (setq pos elem)))
d9e94c22
MS
3300 pos))
3301
3302(defun c-safe-position (bufpos paren-state)
0386b551
AM
3303 ;; Return the closest "safe" position recorded on PAREN-STATE that
3304 ;; is higher up than BUFPOS. Return nil if PAREN-STATE doesn't
3305 ;; contain any. Return nil if BUFPOS is nil, which is useful to
3306 ;; find the closest limit before a given limit that might be nil.
d9e94c22 3307 ;;
0386b551
AM
3308 ;; A "safe" position is a position at or after a recorded open
3309 ;; paren, or after a recorded close paren. The returned position is
3310 ;; thus either the first position after a close brace, or the first
3311 ;; position after an enclosing paren, or at the enclosing paren in
3312 ;; case BUFPOS is immediately after it.
d9e94c22
MS
3313 (when bufpos
3314 (let (elem)
3315 (catch 'done
3316 (while paren-state
3317 (setq elem (car paren-state))
3318 (if (consp elem)
3319 (cond ((< (cdr elem) bufpos)
3320 (throw 'done (cdr elem)))
3321 ((< (car elem) bufpos)
3322 ;; See below.
3323 (throw 'done (min (1+ (car elem)) bufpos))))
3324 (if (< elem bufpos)
3325 ;; elem is the position at and not after the opening paren, so
3326 ;; we can go forward one more step unless it's equal to
3327 ;; bufpos. This is useful in some cases avoid an extra paren
3328 ;; level between the safe position and bufpos.
3329 (throw 'done (min (1+ elem) bufpos))))
3330 (setq paren-state (cdr paren-state)))))))
3331
3332(defun c-beginning-of-syntax ()
3333 ;; This is used for `font-lock-beginning-of-syntax-function'. It
3334 ;; goes to the closest previous point that is known to be outside
3335 ;; any string literal or comment. `c-state-cache' is used if it has
3336 ;; a position in the vicinity.
3337 (let* ((paren-state c-state-cache)
3338 elem
3339
3340 (pos (catch 'done
3341 ;; Note: Similar code in `c-safe-position'. The
3342 ;; difference is that we accept a safe position at
3343 ;; the point and don't bother to go forward past open
3344 ;; parens.
3345 (while paren-state
3346 (setq elem (car paren-state))
3347 (if (consp elem)
3348 (cond ((<= (cdr elem) (point))
3349 (throw 'done (cdr elem)))
3350 ((<= (car elem) (point))
3351 (throw 'done (car elem))))
3352 (if (<= elem (point))
3353 (throw 'done elem)))
3354 (setq paren-state (cdr paren-state)))
3355 (point-min))))
3356
3357 (if (> pos (- (point) 4000))
3358 (goto-char pos)
3359 ;; The position is far back. Try `c-beginning-of-defun-1'
3360 ;; (although we can't be entirely sure it will go to a position
3361 ;; outside a comment or string in current emacsen). FIXME:
3362 ;; Consult `syntax-ppss' here.
3363 (c-beginning-of-defun-1)
3364 (if (< (point) pos)
3365 (goto-char pos)))))
3366
3367\f
3368;; Tools for scanning identifiers and other tokens.
3369
3370(defun c-on-identifier ()
3371 "Return non-nil if the point is on or directly after an identifier.
3372Keywords are recognized and not considered identifiers. If an
3373identifier is detected, the returned value is its starting position.
0386b551
AM
3374If an identifier ends at the point and another begins at it \(can only
3375happen in Pike) then the point for the preceding one is returned.
d9e94c22 3376
0386b551
AM
3377Note that this function might do hidden buffer changes. See the
3378comment at the start of cc-engine.el for more info."
3379
3380 ;; FIXME: Shouldn't this function handle "operator" in C++?
d9e94c22
MS
3381
3382 (save-excursion
0386b551
AM
3383 (skip-syntax-backward "w_")
3384
3385 (or
3386
3387 ;; Check for a normal (non-keyword) identifier.
3388 (and (looking-at c-symbol-start)
3389 (not (looking-at c-keywords-regexp))
3390 (point))
3391
3392 (when (c-major-mode-is 'pike-mode)
3393 ;; Handle the `<operator> syntax in Pike.
3394 (let ((pos (point)))
3395 (skip-chars-backward "-!%&*+/<=>^|~[]()")
3396 (and (if (< (skip-chars-backward "`") 0)
3397 t
3398 (goto-char pos)
3399 (eq (char-after) ?\`))
3400 (looking-at c-symbol-key)
3401 (>= (match-end 0) pos)
3402 (point))))
3403
3404 ;; Handle the "operator +" syntax in C++.
3405 (when (and c-overloadable-operators-regexp
3406 (= (c-backward-token-2 0) 0))
3407
3408 (cond ((and (looking-at c-overloadable-operators-regexp)
51c9af45 3409 (or (not c-opt-op-identifier-prefix)
0386b551 3410 (and (= (c-backward-token-2 1) 0)
51c9af45 3411 (looking-at c-opt-op-identifier-prefix))))
0386b551
AM
3412 (point))
3413
3414 ((save-excursion
51c9af45
AM
3415 (and c-opt-op-identifier-prefix
3416 (looking-at c-opt-op-identifier-prefix)
0386b551
AM
3417 (= (c-forward-token-2 1) 0)
3418 (looking-at c-overloadable-operators-regexp)))
3419 (point))))
3420
3421 )))
d9e94c22
MS
3422
3423(defsubst c-simple-skip-symbol-backward ()
3424 ;; If the point is at the end of a symbol then skip backward to the
3425 ;; beginning of it. Don't move otherwise. Return non-nil if point
3426 ;; moved.
0386b551
AM
3427 ;;
3428 ;; This function might do hidden buffer changes.
d9e94c22
MS
3429 (or (< (skip-syntax-backward "w_") 0)
3430 (and (c-major-mode-is 'pike-mode)
3431 ;; Handle the `<operator> syntax in Pike.
3432 (let ((pos (point)))
2a15eb73 3433 (if (and (< (skip-chars-backward "-!%&*+/<=>^|~[]()") 0)
d9e94c22
MS
3434 (< (skip-chars-backward "`") 0)
3435 (looking-at c-symbol-key)
3436 (>= (match-end 0) pos))
3437 t
3438 (goto-char pos)
3439 nil)))))
3440
0386b551 3441(defun c-beginning-of-current-token (&optional back-limit)
d9e94c22
MS
3442 ;; Move to the beginning of the current token. Do not move if not
3443 ;; in the middle of one. BACK-LIMIT may be used to bound the
3444 ;; backward search; if given it's assumed to be at the boundary
a85fd6da 3445 ;; between two tokens. Return non-nil if the point is moved, nil
580fba94 3446 ;; otherwise.
0386b551
AM
3447 ;;
3448 ;; This function might do hidden buffer changes.
d9e94c22 3449 (let ((start (point)))
580fba94
AM
3450 (if (looking-at "\\w\\|\\s_")
3451 (skip-syntax-backward "w_" back-limit)
3452 (when (< (skip-syntax-backward ".()" back-limit) 0)
3453 (while (let ((pos (or (and (looking-at c-nonsymbol-token-regexp)
3454 (match-end 0))
3455 ;; `c-nonsymbol-token-regexp' should always match
3456 ;; since we've skipped backward over punctuator
3457 ;; or paren syntax, but consume one char in case
3458 ;; it doesn't so that we don't leave point before
3459 ;; some earlier incorrect token.
3460 (1+ (point)))))
3461 (if (<= pos start)
3462 (goto-char pos))))))
3463 (< (point) start)))
d9e94c22 3464
ff959bab 3465(defun c-end-of-current-token (&optional back-limit)
d9e94c22
MS
3466 ;; Move to the end of the current token. Do not move if not in the
3467 ;; middle of one. BACK-LIMIT may be used to bound the backward
3468 ;; search; if given it's assumed to be at the boundary between two
ff959bab 3469 ;; tokens. Return non-nil if the point is moved, nil otherwise.
0386b551
AM
3470 ;;
3471 ;; This function might do hidden buffer changes.
d9e94c22
MS
3472 (let ((start (point)))
3473 (cond ((< (skip-syntax-backward "w_" (1- start)) 0)
3474 (skip-syntax-forward "w_"))
3475 ((< (skip-syntax-backward ".()" back-limit) 0)
3476 (while (progn
3477 (if (looking-at c-nonsymbol-token-regexp)
3478 (goto-char (match-end 0))
3479 ;; `c-nonsymbol-token-regexp' should always match since
3480 ;; we've skipped backward over punctuator or paren
3481 ;; syntax, but move forward in case it doesn't so that
3482 ;; we don't leave point earlier than we started with.
3483 (forward-char))
ff959bab
MS
3484 (< (point) start)))))
3485 (> (point) start)))
d9e94c22
MS
3486
3487(defconst c-jump-syntax-balanced
3488 (if (memq 'gen-string-delim c-emacs-features)
3489 "\\w\\|\\s_\\|\\s\(\\|\\s\)\\|\\s\"\\|\\s|"
3490 "\\w\\|\\s_\\|\\s\(\\|\\s\)\\|\\s\""))
3491
3492(defconst c-jump-syntax-unbalanced
3493 (if (memq 'gen-string-delim c-emacs-features)
3494 "\\w\\|\\s_\\|\\s\"\\|\\s|"
3495 "\\w\\|\\s_\\|\\s\""))
3496
3497(defun c-forward-token-2 (&optional count balanced limit)
3498 "Move forward by tokens.
3499A token is defined as all symbols and identifiers which aren't
3500syntactic whitespace \(note that multicharacter tokens like \"==\" are
3501treated properly). Point is always either left at the beginning of a
3502token or not moved at all. COUNT specifies the number of tokens to
3503move; a negative COUNT moves in the opposite direction. A COUNT of 0
3504moves to the next token beginning only if not already at one. If
3505BALANCED is true, move over balanced parens, otherwise move into them.
3506Also, if BALANCED is true, never move out of an enclosing paren.
3507
3508LIMIT sets the limit for the movement and defaults to the point limit.
3509The case when LIMIT is set in the middle of a token, comment or macro
3510is handled correctly, i.e. the point won't be left there.
3511
3512Return the number of tokens left to move \(positive or negative). If
3513BALANCED is true, a move over a balanced paren counts as one. Note
3514that if COUNT is 0 and no appropriate token beginning is found, 1 will
3515be returned. Thus, a return value of 0 guarantees that point is at
3516the requested position and a return value less \(without signs) than
0386b551
AM
3517COUNT guarantees that point is at the beginning of some token.
3518
3519Note that this function might do hidden buffer changes. See the
3520comment at the start of cc-engine.el for more info."
d9e94c22
MS
3521
3522 (or count (setq count 1))
3523 (if (< count 0)
3524 (- (c-backward-token-2 (- count) balanced limit))
3525
3526 (let ((jump-syntax (if balanced
3527 c-jump-syntax-balanced
3528 c-jump-syntax-unbalanced))
3529 (last (point))
3530 (prev (point)))
3531
3532 (if (zerop count)
3533 ;; If count is zero we should jump if in the middle of a token.
3534 (c-end-of-current-token))
3535
3536 (save-restriction
3537 (if limit (narrow-to-region (point-min) limit))
3538 (if (/= (point)
3539 (progn (c-forward-syntactic-ws) (point)))
3540 ;; Skip whitespace. Count this as a move if we did in
3541 ;; fact move.
3542 (setq count (max (1- count) 0)))
3543
3544 (if (eobp)
3545 ;; Moved out of bounds. Make sure the returned count isn't zero.
3546 (progn
3547 (if (zerop count) (setq count 1))
3548 (goto-char last))
3549
3550 ;; Use `condition-case' to avoid having the limit tests
3551 ;; inside the loop.
3552 (condition-case nil
3553 (while (and
3554 (> count 0)
3555 (progn
3556 (setq last (point))
3557 (cond ((looking-at jump-syntax)
3558 (goto-char (scan-sexps (point) 1))
3559 t)
3560 ((looking-at c-nonsymbol-token-regexp)
3561 (goto-char (match-end 0))
3562 t)
3563 ;; `c-nonsymbol-token-regexp' above should always
3564 ;; match if there are correct tokens. Try to
3565 ;; widen to see if the limit was set in the
3566 ;; middle of one, else fall back to treating
3567 ;; the offending thing as a one character token.
3568 ((and limit
3569 (save-restriction
3570 (widen)
3571 (looking-at c-nonsymbol-token-regexp)))
3572 nil)
3573 (t
3574 (forward-char)
3575 t))))
3576 (c-forward-syntactic-ws)
3577 (setq prev last
3578 count (1- count)))
3579 (error (goto-char last)))
3580
3581 (when (eobp)
3582 (goto-char prev)
3583 (setq count (1+ count)))))
3584
3585 count)))
3586
3587(defun c-backward-token-2 (&optional count balanced limit)
3588 "Move backward by tokens.
3589See `c-forward-token-2' for details."
3590
3591 (or count (setq count 1))
3592 (if (< count 0)
3593 (- (c-forward-token-2 (- count) balanced limit))
3594
3595 (or limit (setq limit (point-min)))
3596 (let ((jump-syntax (if balanced
3597 c-jump-syntax-balanced
3598 c-jump-syntax-unbalanced))
3599 (last (point)))
3600
3601 (if (zerop count)
3602 ;; The count is zero so try to skip to the beginning of the
3603 ;; current token.
3604 (if (> (point)
3605 (progn (c-beginning-of-current-token) (point)))
3606 (if (< (point) limit)
3607 ;; The limit is inside the same token, so return 1.
3608 (setq count 1))
3609
3610 ;; We're not in the middle of a token. If there's
3611 ;; whitespace after the point then we must move backward,
3612 ;; so set count to 1 in that case.
3613 (and (looking-at c-syntactic-ws-start)
3614 ;; If we're looking at a '#' that might start a cpp
3615 ;; directive then we have to do a more elaborate check.
3616 (or (/= (char-after) ?#)
3617 (not c-opt-cpp-prefix)
3618 (save-excursion
3619 (and (= (point)
3620 (progn (beginning-of-line)
3621 (looking-at "[ \t]*")
3622 (match-end 0)))
3623 (or (bobp)
3624 (progn (backward-char)
3625 (not (eq (char-before) ?\\)))))))
3626 (setq count 1))))
3627
3628 ;; Use `condition-case' to avoid having to check for buffer
3629 ;; limits in `backward-char', `scan-sexps' and `goto-char' below.
3630 (condition-case nil
3631 (while (and
3632 (> count 0)
3633 (progn
3634 (c-backward-syntactic-ws)
3635 (backward-char)
3636 (if (looking-at jump-syntax)
3637 (goto-char (scan-sexps (1+ (point)) -1))
3638 ;; This can be very inefficient if there's a long
3639 ;; sequence of operator tokens without any separation.
3640 ;; That doesn't happen in practice, anyway.
3641 (c-beginning-of-current-token))
3642 (>= (point) limit)))
3643 (setq last (point)
3644 count (1- count)))
3645 (error (goto-char last)))
3646
3647 (if (< (point) limit)
3648 (goto-char last))
3649
3650 count)))
3651
3652(defun c-forward-token-1 (&optional count balanced limit)
3653 "Like `c-forward-token-2' but doesn't treat multicharacter operator
3654tokens like \"==\" as single tokens, i.e. all sequences of symbol
3655characters are jumped over character by character. This function is
3656for compatibility only; it's only a wrapper over `c-forward-token-2'."
3657 (let ((c-nonsymbol-token-regexp "\\s.\\|\\s\(\\|\\s\)"))
3658 (c-forward-token-2 count balanced limit)))
3659
3660(defun c-backward-token-1 (&optional count balanced limit)
3661 "Like `c-backward-token-2' but doesn't treat multicharacter operator
3662tokens like \"==\" as single tokens, i.e. all sequences of symbol
3663characters are jumped over character by character. This function is
3664for compatibility only; it's only a wrapper over `c-backward-token-2'."
3665 (let ((c-nonsymbol-token-regexp "\\s.\\|\\s\(\\|\\s\)"))
3666 (c-backward-token-2 count balanced limit)))
3667
3668\f
3669;; Tools for doing searches restricted to syntactically relevant text.
3670
3671(defun c-syntactic-re-search-forward (regexp &optional bound noerror
3672 paren-level not-inside-token
3673 lookbehind-submatch)
3674 "Like `re-search-forward', but only report matches that are found
3675in syntactically significant text. I.e. matches in comments, macros
3676or string literals are ignored. The start point is assumed to be
3677outside any comment, macro or string literal, or else the content of
3678that region is taken as syntactically significant text.
3679
3680If PAREN-LEVEL is non-nil, an additional restriction is added to
2a15eb73
MS
3681ignore matches in nested paren sexps. The search will also not go
3682outside the current list sexp, which has the effect that if the point
3683should be moved to BOUND when no match is found \(i.e. NOERROR is
3684neither nil nor t), then it will be at the closing paren if the end of
3685the current list sexp is encountered first.
d9e94c22
MS
3686
3687If NOT-INSIDE-TOKEN is non-nil, matches in the middle of tokens are
3688ignored. Things like multicharacter operators and special symbols
3689\(e.g. \"`()\" in Pike) are handled but currently not floating point
3690constants.
3691
3692If LOOKBEHIND-SUBMATCH is non-nil, it's taken as a number of a
3693subexpression in REGEXP. The end of that submatch is used as the
3694position to check for syntactic significance. If LOOKBEHIND-SUBMATCH
3695isn't used or if that subexpression didn't match then the start
3696position of the whole match is used instead. The \"look behind\"
3697subexpression is never tested before the starting position, so it
3698might be a good idea to include \\=\\= as a match alternative in it.
3699
3700Optimization note: Matches might be missed if the \"look behind\"
2a15eb73 3701subexpression can match the end of nonwhite syntactic whitespace,
d9e94c22 3702i.e. the end of comments or cpp directives. This since the function
2a15eb73
MS
3703skips over such things before resuming the search. It's on the other
3704hand not safe to assume that the \"look behind\" subexpression never
3705matches syntactic whitespace.
3706
3707Bug: Unbalanced parens inside cpp directives are currently not handled
3708correctly \(i.e. they don't get ignored as they should) when
0386b551
AM
3709PAREN-LEVEL is set.
3710
3711Note that this function might do hidden buffer changes. See the
3712comment at the start of cc-engine.el for more info."
d9e94c22
MS
3713
3714 (or bound (setq bound (point-max)))
3715 (if paren-level (setq paren-level -1))
3716
3717 ;;(message "c-syntactic-re-search-forward %s %s %S" (point) bound regexp)
3718
3719 (let ((start (point))
2a15eb73
MS
3720 tmp
3721 ;; Start position for the last search.
3722 search-pos
3723 ;; The `parse-partial-sexp' state between the start position
3724 ;; and the point.
3725 state
3726 ;; The current position after the last state update. The next
3727 ;; `parse-partial-sexp' continues from here.
3728 (state-pos (point))
3729 ;; The position at which to check the state and the state
3730 ;; there. This is separate from `state-pos' since we might
3731 ;; need to back up before doing the next search round.
3732 check-pos check-state
3733 ;; Last position known to end a token.
d9e94c22 3734 (last-token-end-pos (point-min))
2a15eb73
MS
3735 ;; Set when a valid match is found.
3736 found)
d9e94c22
MS
3737
3738 (condition-case err
3739 (while
3740 (and
2a15eb73
MS
3741 (progn
3742 (setq search-pos (point))
3743 (re-search-forward regexp bound noerror))
d9e94c22
MS
3744
3745 (progn
2a15eb73
MS
3746 (setq state (parse-partial-sexp
3747 state-pos (match-beginning 0) paren-level nil state)
3748 state-pos (point))
d9e94c22 3749 (if (setq check-pos (and lookbehind-submatch
2a15eb73
MS
3750 (or (not paren-level)
3751 (>= (car state) 0))
d9e94c22
MS
3752 (match-end lookbehind-submatch)))
3753 (setq check-state (parse-partial-sexp
2a15eb73
MS
3754 state-pos check-pos paren-level nil state))
3755 (setq check-pos state-pos
d9e94c22
MS
3756 check-state state))
3757
2a15eb73
MS
3758 ;; NOTE: If we got a look behind subexpression and get
3759 ;; an insignificant match in something that isn't
d9e94c22
MS
3760 ;; syntactic whitespace (i.e. strings or in nested
3761 ;; parentheses), then we can never skip more than a
2a15eb73
MS
3762 ;; single character from the match start position
3763 ;; (i.e. `state-pos' here) before continuing the
3764 ;; search. That since the look behind subexpression
3765 ;; might match the end of the insignificant region in
3766 ;; the next search.
d9e94c22
MS
3767
3768 (cond
d9e94c22
MS
3769 ((elt check-state 7)
3770 ;; Match inside a line comment. Skip to eol. Use
3771 ;; `re-search-forward' instead of `skip-chars-forward' to get
3772 ;; the right bound behavior.
3773 (re-search-forward "[\n\r]" bound noerror))
3774
3775 ((elt check-state 4)
3776 ;; Match inside a block comment. Skip to the '*/'.
3777 (search-forward "*/" bound noerror))
3778
3779 ((and (not (elt check-state 5))
3780 (eq (char-before check-pos) ?/)
2a15eb73 3781 (not (c-get-char-property (1- check-pos) 'syntax-table))
d9e94c22
MS
3782 (memq (char-after check-pos) '(?/ ?*)))
3783 ;; Match in the middle of the opener of a block or line
3784 ;; comment.
3785 (if (= (char-after check-pos) ?/)
3786 (re-search-forward "[\n\r]" bound noerror)
3787 (search-forward "*/" bound noerror)))
3788
2a15eb73
MS
3789 ;; The last `parse-partial-sexp' above might have
3790 ;; stopped short of the real check position if the end
3791 ;; of the current sexp was encountered in paren-level
3792 ;; mode. The checks above are always false in that
3793 ;; case, and since they can do better skipping in
3794 ;; lookbehind-submatch mode, we do them before
3795 ;; checking the paren level.
3796
3797 ((and paren-level
3798 (/= (setq tmp (car check-state)) 0))
3799 ;; Check the paren level first since we're short of the
3800 ;; syntactic checking position if the end of the
3801 ;; current sexp was encountered by `parse-partial-sexp'.
3802 (if (> tmp 0)
3803
3804 ;; Inside a nested paren sexp.
3805 (if lookbehind-submatch
3806 ;; See the NOTE above.
3807 (progn (goto-char state-pos) t)
3808 ;; Skip out of the paren quickly.
3809 (setq state (parse-partial-sexp state-pos bound 0 nil state)
3810 state-pos (point)))
3811
3812 ;; Have exited the current paren sexp.
3813 (if noerror
3814 (progn
3815 ;; The last `parse-partial-sexp' call above
3816 ;; has left us just after the closing paren
3817 ;; in this case, so we can modify the bound
3818 ;; to leave the point at the right position
3819 ;; upon return.
3820 (setq bound (1- (point)))
3821 nil)
3822 (signal 'search-failed (list regexp)))))
3823
3824 ((setq tmp (elt check-state 3))
3825 ;; Match inside a string.
3826 (if (or lookbehind-submatch
3827 (not (integerp tmp)))
3828 ;; See the NOTE above.
3829 (progn (goto-char state-pos) t)
3830 ;; Skip to the end of the string before continuing.
3831 (let ((ender (make-string 1 tmp)) (continue t))
3832 (while (if (search-forward ender bound noerror)
3833 (progn
3834 (setq state (parse-partial-sexp
3835 state-pos (point) nil nil state)
3836 state-pos (point))
3837 (elt state 3))
3838 (setq continue nil)))
3839 continue)))
d9e94c22
MS
3840
3841 ((save-excursion
3842 (save-match-data
3843 (c-beginning-of-macro start)))
3844 ;; Match inside a macro. Skip to the end of it.
3845 (c-end-of-macro)
3846 (cond ((<= (point) bound) t)
3847 (noerror nil)
2a15eb73 3848 (t (signal 'search-failed (list regexp)))))
d9e94c22 3849
2a15eb73
MS
3850 ((and not-inside-token
3851 (or (< check-pos last-token-end-pos)
3852 (< check-pos
3853 (save-excursion
3854 (goto-char check-pos)
3855 (save-match-data
3856 (c-end-of-current-token last-token-end-pos))
3857 (setq last-token-end-pos (point))))))
3858 ;; Inside a token.
3859 (if lookbehind-submatch
3860 ;; See the NOTE above.
3861 (goto-char state-pos)
3862 (goto-char (min last-token-end-pos bound))))
d9e94c22
MS
3863
3864 (t
3865 ;; A real match.
3866 (setq found t)
2a15eb73
MS
3867 nil)))
3868
3869 ;; Should loop to search again, but take care to avoid
3870 ;; looping on the same spot.
3871 (or (/= search-pos (point))
3872 (if (= (point) bound)
3873 (if noerror
3874 nil
3875 (signal 'search-failed (list regexp)))
3876 (forward-char)
3877 t))))
d9e94c22
MS
3878
3879 (error
3880 (goto-char start)
3881 (signal (car err) (cdr err))))
3882
2a15eb73 3883 ;;(message "c-syntactic-re-search-forward done %s" (or (match-end 0) (point)))
d9e94c22
MS
3884
3885 (if found
3886 (progn
2a15eb73
MS
3887 (goto-char (match-end 0))
3888 (match-end 0))
d9e94c22
MS
3889
3890 ;; Search failed. Set point as appropriate.
2a15eb73
MS
3891 (if (eq noerror t)
3892 (goto-char start)
3893 (goto-char bound))
d9e94c22
MS
3894 nil)))
3895
47641aac
GM
3896(defvar safe-pos-list) ; bound in c-syntactic-skip-backward
3897
d0fcee66
AM
3898(defsubst c-ssb-lit-begin ()
3899 ;; Return the start of the literal point is in, or nil.
3900 ;; We read and write the variables `safe-pos', `safe-pos-list', `state'
3901 ;; bound in the caller.
3902
3903 ;; Use `parse-partial-sexp' from a safe position down to the point to check
3904 ;; if it's outside comments and strings.
3905 (save-excursion
3906 (let ((pos (point)) safe-pos state pps-end-pos)
3907 ;; Pick a safe position as close to the point as possible.
3908 ;;
3909 ;; FIXME: Consult `syntax-ppss' here if our cache doesn't give a good
3910 ;; position.
47641aac 3911
d0fcee66
AM
3912 (while (and safe-pos-list
3913 (> (car safe-pos-list) (point)))
3914 (setq safe-pos-list (cdr safe-pos-list)))
3915 (unless (setq safe-pos (car-safe safe-pos-list))
3916 (setq safe-pos (max (or (c-safe-position
3917 (point) (or c-state-cache
3918 (c-parse-state)))
3919 0)
3920 (point-min))
3921 safe-pos-list (list safe-pos)))
3922
3923 ;; Cache positions along the way to use if we have to back up more. We
3924 ;; cache every closing paren on the same level. If the paren cache is
3925 ;; relevant in this region then we're typically already on the same
3926 ;; level as the target position. Note that we might cache positions
3927 ;; after opening parens in case safe-pos is in a nested list. That's
3928 ;; both uncommon and harmless.
3929 (while (progn
3930 (setq state (parse-partial-sexp
3931 safe-pos pos 0))
3932 (< (point) pos))
3933 (setq safe-pos (point)
3934 safe-pos-list (cons safe-pos safe-pos-list)))
3935
3936 ;; If the state contains the start of the containing sexp we cache that
3937 ;; position too, so that parse-partial-sexp in the next run has a bigger
3938 ;; chance of starting at the same level as the target position and thus
3939 ;; will get more good safe positions into the list.
3940 (if (elt state 1)
3941 (setq safe-pos (1+ (elt state 1))
3942 safe-pos-list (cons safe-pos safe-pos-list)))
3943
3944 (if (or (elt state 3) (elt state 4))
3945 ;; Inside string or comment. Continue search at the
3946 ;; beginning of it.
3947 (elt state 8)))))
3948
0386b551 3949(defun c-syntactic-skip-backward (skip-chars &optional limit paren-level)
d9e94c22
MS
3950 "Like `skip-chars-backward' but only look at syntactically relevant chars,
3951i.e. don't stop at positions inside syntactic whitespace or string
3952literals. Preprocessor directives are also ignored, with the exception
3953of the one that the point starts within, if any. If LIMIT is given,
0386b551
AM
3954it's assumed to be at a syntactically relevant position.
3955
3956If PAREN-LEVEL is non-nil, the function won't stop in nested paren
3957sexps, and the search will also not go outside the current paren sexp.
3958However, if LIMIT or the buffer limit is reached inside a nested paren
3959then the point will be left at the limit.
3960
3961Non-nil is returned if the point moved, nil otherwise.
3962
3963Note that this function might do hidden buffer changes. See the
3964comment at the start of cc-engine.el for more info."
d9e94c22
MS
3965
3966 (let ((start (point))
d0fcee66 3967 state-2
d9e94c22
MS
3968 ;; A list of syntactically relevant positions in descending
3969 ;; order. It's used to avoid scanning repeatedly over
3970 ;; potentially large regions with `parse-partial-sexp' to verify
d0fcee66 3971 ;; each position. Used in `c-ssb-lit-begin'
d9e94c22
MS
3972 safe-pos-list
3973 ;; The result from `c-beginning-of-macro' at the start position or the
3974 ;; start position itself if it isn't within a macro. Evaluated on
3975 ;; demand.
0386b551
AM
3976 start-macro-beg
3977 ;; The earliest position after the current one with the same paren
3978 ;; level. Used only when `paren-level' is set.
d0fcee66 3979 lit-beg
0386b551 3980 (paren-level-pos (point)))
d9e94c22 3981
d0fcee66
AM
3982 (while
3983 (progn
3984 ;; The next loop "tries" to find the end point each time round,
3985 ;; loops when it hasn't succeeded.
3986 (while
3987 (and
3988 (< (skip-chars-backward skip-chars limit) 0)
3989
3990 (let ((pos (point)) state-2 pps-end-pos)
3991
3992 (cond
3993 ;; Don't stop inside a literal
3994 ((setq lit-beg (c-ssb-lit-begin))
3995 (goto-char lit-beg)
3996 t)
3997
3998 ((and paren-level
3999 (save-excursion
4000 (setq state-2 (parse-partial-sexp
4001 pos paren-level-pos -1)
4002 pps-end-pos (point))
4003 (/= (car state-2) 0)))
4004 ;; Not at the right level.
4005
4006 (if (and (< (car state-2) 0)
4007 ;; We stop above if we go out of a paren.
4008 ;; Now check whether it precedes or is
4009 ;; nested in the starting sexp.
4010 (save-excursion
4011 (setq state-2
4012 (parse-partial-sexp
4013 pps-end-pos paren-level-pos
4014 nil nil state-2))
4015 (< (car state-2) 0)))
4016
4017 ;; We've stopped short of the starting position
4018 ;; so the hit was inside a nested list. Go up
4019 ;; until we are at the right level.
4020 (condition-case nil
4021 (progn
4022 (goto-char (scan-lists pos -1
4023 (- (car state-2))))
4024 (setq paren-level-pos (point))
4025 (if (and limit (>= limit paren-level-pos))
0386b551 4026 (progn
d0fcee66
AM
4027 (goto-char limit)
4028 nil)
4029 t))
4030 (error
4031 (goto-char (or limit (point-min)))
4032 nil))
4033
4034 ;; The hit was outside the list at the start
4035 ;; position. Go to the start of the list and exit.
4036 (goto-char (1+ (elt state-2 1)))
4037 nil))
4038
4039 ((c-beginning-of-macro limit)
4040 ;; Inside a macro.
4041 (if (< (point)
4042 (or start-macro-beg
4043 (setq start-macro-beg
4044 (save-excursion
4045 (goto-char start)
4046 (c-beginning-of-macro limit)
4047 (point)))))
4048 t
4049
4050 ;; It's inside the same macro we started in so it's
4051 ;; a relevant match.
4052 (goto-char pos)
4053 nil))))))
91af3942 4054
d0fcee66
AM
4055 (> (point)
4056 (progn
4057 ;; Skip syntactic ws afterwards so that we don't stop at the
4058 ;; end of a comment if `skip-chars' is something like "^/".
4059 (c-backward-syntactic-ws)
4060 (point)))))
d9e94c22 4061
0386b551
AM
4062 ;; We might want to extend this with more useful return values in
4063 ;; the future.
4064 (/= (point) start)))
4065
4066;; The following is an alternative implementation of
4067;; `c-syntactic-skip-backward' that uses backward movement to keep
4068;; track of the syntactic context. It turned out to be generally
4069;; slower than the one above which uses forward checks from earlier
4070;; safe positions.
4071;;
4072;;(defconst c-ssb-stop-re
4073;; ;; The regexp matching chars `c-syntactic-skip-backward' needs to
4074;; ;; stop at to avoid going into comments and literals.
4075;; (concat
4076;; ;; Match comment end syntax and string literal syntax. Also match
4077;; ;; '/' for block comment endings (not covered by comment end
4078;; ;; syntax).
4079;; "\\s>\\|/\\|\\s\""
4080;; (if (memq 'gen-string-delim c-emacs-features)
4081;; "\\|\\s|"
4082;; "")
4083;; (if (memq 'gen-comment-delim c-emacs-features)
4084;; "\\|\\s!"
4085;; "")))
4086;;
4087;;(defconst c-ssb-stop-paren-re
4088;; ;; Like `c-ssb-stop-re' but also stops at paren chars.
4089;; (concat c-ssb-stop-re "\\|\\s(\\|\\s)"))
4090;;
4091;;(defconst c-ssb-sexp-end-re
4092;; ;; Regexp matching the ending syntax of a complex sexp.
4093;; (concat c-string-limit-regexp "\\|\\s)"))
4094;;
4095;;(defun c-syntactic-skip-backward (skip-chars &optional limit paren-level)
4096;; "Like `skip-chars-backward' but only look at syntactically relevant chars,
4097;;i.e. don't stop at positions inside syntactic whitespace or string
4098;;literals. Preprocessor directives are also ignored. However, if the
4099;;point is within a comment, string literal or preprocessor directory to
4100;;begin with, its contents is treated as syntactically relevant chars.
4101;;If LIMIT is given, it limits the backward search and the point will be
4102;;left there if no earlier position is found.
4103;;
4104;;If PAREN-LEVEL is non-nil, the function won't stop in nested paren
4105;;sexps, and the search will also not go outside the current paren sexp.
4106;;However, if LIMIT or the buffer limit is reached inside a nested paren
4107;;then the point will be left at the limit.
4108;;
4109;;Non-nil is returned if the point moved, nil otherwise.
4110;;
4111;;Note that this function might do hidden buffer changes. See the
4112;;comment at the start of cc-engine.el for more info."
4113;;
4114;; (save-restriction
4115;; (when limit
4116;; (narrow-to-region limit (point-max)))
4117;;
4118;; (let ((start (point)))
4119;; (catch 'done
4120;; (while (let ((last-pos (point))
4121;; (stop-pos (progn
4122;; (skip-chars-backward skip-chars)
4123;; (point))))
4124;;
4125;; ;; Skip back over the same region as
4126;; ;; `skip-chars-backward' above, but keep to
4127;; ;; syntactically relevant positions.
4128;; (goto-char last-pos)
4129;; (while (and
4130;; ;; `re-search-backward' with a single char regexp
4131;; ;; should be fast.
4132;; (re-search-backward
4133;; (if paren-level c-ssb-stop-paren-re c-ssb-stop-re)
4134;; stop-pos 'move)
4135;;
4136;; (progn
4137;; (cond
4138;; ((looking-at "\\s(")
4139;; ;; `paren-level' is set and we've found the
4140;; ;; start of the containing paren.
4141;; (forward-char)
4142;; (throw 'done t))
4143;;
4144;; ((looking-at c-ssb-sexp-end-re)
4145;; ;; We're at the end of a string literal or paren
4146;; ;; sexp (if `paren-level' is set).
4147;; (forward-char)
4148;; (condition-case nil
4149;; (c-backward-sexp)
4150;; (error
4151;; (goto-char limit)
4152;; (throw 'done t))))
4153;;
4154;; (t
4155;; (forward-char)
4156;; ;; At the end of some syntactic ws or possibly
4157;; ;; after a plain '/' operator.
4158;; (let ((pos (point)))
4159;; (c-backward-syntactic-ws)
4160;; (if (= pos (point))
4161;; ;; Was a plain '/' operator. Go past it.
4162;; (backward-char)))))
4163;;
4164;; (> (point) stop-pos))))
4165;;
4166;; ;; Now the point is either at `stop-pos' or at some
4167;; ;; position further back if `stop-pos' was at a
4168;; ;; syntactically irrelevant place.
4169;;
4170;; ;; Skip additional syntactic ws so that we don't stop
4171;; ;; at the end of a comment if `skip-chars' is
4172;; ;; something like "^/".
4173;; (c-backward-syntactic-ws)
4174;;
4175;; (< (point) stop-pos))))
4176;;
4177;; ;; We might want to extend this with more useful return values
4178;; ;; in the future.
4179;; (/= (point) start))))
d9e94c22
MS
4180
4181\f
4182;; Tools for handling comments and string literals.
4183
4184(defun c-slow-in-literal (&optional lim detect-cpp)
4185 "Return the type of literal point is in, if any.
4186The return value is `c' if in a C-style comment, `c++' if in a C++
4187style comment, `string' if in a string literal, `pound' if DETECT-CPP
4188is non-nil and in a preprocessor line, or nil if somewhere else.
4189Optional LIM is used as the backward limit of the search. If omitted,
4190or nil, `c-beginning-of-defun' is used.
4191
4192The last point calculated is cached if the cache is enabled, i.e. if
4193`c-in-literal-cache' is bound to a two element vector.
4194
0386b551
AM
4195Note that this function might do hidden buffer changes. See the
4196comment at the start of cc-engine.el for more info."
4197
d9e94c22
MS
4198 (if (and (vectorp c-in-literal-cache)
4199 (= (point) (aref c-in-literal-cache 0)))
4200 (aref c-in-literal-cache 1)
4201 (let ((rtn (save-excursion
4202 (let* ((pos (point))
4203 (lim (or lim (progn
4204 (c-beginning-of-syntax)
4205 (point))))
4206 (state (parse-partial-sexp lim pos)))
4207 (cond
4208 ((elt state 3) 'string)
4209 ((elt state 4) (if (elt state 7) 'c++ 'c))
4210 ((and detect-cpp (c-beginning-of-macro lim)) 'pound)
4211 (t nil))))))
4212 ;; cache this result if the cache is enabled
4213 (if (not c-in-literal-cache)
4214 (setq c-in-literal-cache (vector (point) rtn)))
4215 rtn)))
4216
4217;; XEmacs has a built-in function that should make this much quicker.
4218;; I don't think we even need the cache, which makes our lives more
4219;; complicated anyway. In this case, lim is only used to detect
4220;; cpp directives.
4221;;
0d26e0b6 4222;; Note that there is a bug in XEmacs's buffer-syntactic-context when used in
d9e94c22 4223;; conjunction with syntax-table-properties. The bug is present in, e.g.,
0d26e0b6 4224;; XEmacs 21.4.4. It manifested itself thus:
d9e94c22
MS
4225;;
4226;; Starting with an empty AWK Mode buffer, type
4227;; /regexp/ {<C-j>
4228;; Point gets wrongly left at column 0, rather than being indented to tab-width.
4229;;
4230;; AWK Mode is designed such that when the first / is typed, it gets the
4231;; syntax-table property "string fence". When the second / is typed, BOTH /s
4232;; are given the s-t property "string". However, buffer-syntactic-context
4233;; fails to take account of the change of the s-t property on the opening / to
4234;; "string", and reports that the { is within a string started by the second /.
4235;;
4236;; The workaround for this is for the AWK Mode initialisation to switch the
4237;; defalias for c-in-literal to c-slow-in-literal. This will slow down other
0d26e0b6 4238;; cc-modes in XEmacs whenever an awk-buffer has been initialised.
b414f371 4239;;
d9e94c22
MS
4240;; (Alan Mackenzie, 2003/4/30).
4241
4242(defun c-fast-in-literal (&optional lim detect-cpp)
0386b551 4243 ;; This function might do hidden buffer changes.
d9e94c22
MS
4244 (let ((context (buffer-syntactic-context)))
4245 (cond
4246 ((eq context 'string) 'string)
4247 ((eq context 'comment) 'c++)
4248 ((eq context 'block-comment) 'c)
4249 ((and detect-cpp (save-excursion (c-beginning-of-macro lim))) 'pound))))
4250
4251(defalias 'c-in-literal
4252 (if (fboundp 'buffer-syntactic-context)
7bfc3fdb 4253 'c-fast-in-literal ; XEmacs
d9e94c22
MS
4254 'c-slow-in-literal)) ; GNU Emacs
4255
4256;; The defalias above isn't enough to shut up the byte compiler.
4257(cc-bytecomp-defun c-in-literal)
4258
4259(defun c-literal-limits (&optional lim near not-in-delimiter)
4260 "Return a cons of the beginning and end positions of the comment or
4261string surrounding point (including both delimiters), or nil if point
4262isn't in one. If LIM is non-nil, it's used as the \"safe\" position
4263to start parsing from. If NEAR is non-nil, then the limits of any
4264literal next to point is returned. \"Next to\" means there's only
4265spaces and tabs between point and the literal. The search for such a
4266literal is done first in forward direction. If NOT-IN-DELIMITER is
4267non-nil, the case when point is inside a starting delimiter won't be
a85fd6da 4268recognized. This only has effect for comments which have starting
d9e94c22
MS
4269delimiters with more than one character.
4270
0386b551
AM
4271Note that this function might do hidden buffer changes. See the
4272comment at the start of cc-engine.el for more info."
d9e94c22
MS
4273
4274 (save-excursion
4275 (let* ((pos (point))
4276 (lim (or lim (progn
4277 (c-beginning-of-syntax)
4278 (point))))
4279 (state (parse-partial-sexp lim pos)))
4280
0386b551
AM
4281 (cond ((elt state 3) ; String.
4282 (goto-char (elt state 8))
d9e94c22
MS
4283 (cons (point) (or (c-safe (c-forward-sexp 1) (point))
4284 (point-max))))
4285
0386b551
AM
4286 ((elt state 4) ; Comment.
4287 (goto-char (elt state 8))
d9e94c22
MS
4288 (cons (point) (progn (c-forward-single-comment) (point))))
4289
4290 ((and (not not-in-delimiter)
4291 (not (elt state 5))
4292 (eq (char-before) ?/)
4293 (looking-at "[/*]"))
4294 ;; We're standing in a comment starter.
4295 (backward-char 1)
4296 (cons (point) (progn (c-forward-single-comment) (point))))
4297
4298 (near
4299 (goto-char pos)
4300
4301 ;; Search forward for a literal.
4302 (skip-chars-forward " \t")
4303
4304 (cond
4305 ((looking-at c-string-limit-regexp) ; String.
4306 (cons (point) (or (c-safe (c-forward-sexp 1) (point))
4307 (point-max))))
4308
4309 ((looking-at c-comment-start-regexp) ; Line or block comment.
4310 (cons (point) (progn (c-forward-single-comment) (point))))
4311
4312 (t
4313 ;; Search backward.
4314 (skip-chars-backward " \t")
4315
4316 (let ((end (point)) beg)
4317 (cond
4318 ((save-excursion
4319 (< (skip-syntax-backward c-string-syntax) 0)) ; String.
4320 (setq beg (c-safe (c-backward-sexp 1) (point))))
4321
4322 ((and (c-safe (forward-char -2) t)
4323 (looking-at "*/"))
4324 ;; Block comment. Due to the nature of line
4325 ;; comments, they will always be covered by the
4326 ;; normal case above.
4327 (goto-char end)
4328 (c-backward-single-comment)
4329 ;; If LIM is bogus, beg will be bogus.
4330 (setq beg (point))))
4331
4332 (if beg (cons beg end))))))
4333 ))))
4334
0386b551
AM
4335;; In case external callers use this; it did have a docstring.
4336(defalias 'c-literal-limits-fast 'c-literal-limits)
d9e94c22 4337
0386b551
AM
4338(defun c-collect-line-comments (range)
4339 "If the argument is a cons of two buffer positions (such as returned by
4340`c-literal-limits'), and that range contains a C++ style line comment,
4341then an extended range is returned that contains all adjacent line
4342comments (i.e. all comments that starts in the same column with no
4343empty lines or non-whitespace characters between them). Otherwise the
4344argument is returned.
d9e94c22 4345
0386b551
AM
4346Note that this function might do hidden buffer changes. See the
4347comment at the start of cc-engine.el for more info."
d9e94c22
MS
4348
4349 (save-excursion
0386b551
AM
4350 (condition-case nil
4351 (if (and (consp range) (progn
4352 (goto-char (car range))
4353 (looking-at c-line-comment-starter)))
b414f371 4354 (let ((col (current-column))
0386b551
AM
4355 (beg (point))
4356 (bopl (c-point 'bopl))
4357 (end (cdr range)))
4358 ;; Got to take care in the backward direction to handle
4359 ;; comments which are preceded by code.
4360 (while (and (c-backward-single-comment)
4361 (>= (point) bopl)
4362 (looking-at c-line-comment-starter)
4363 (= col (current-column)))
4364 (setq beg (point)
4365 bopl (c-point 'bopl)))
4366 (goto-char end)
4367 (while (and (progn (skip-chars-forward " \t")
4368 (looking-at c-line-comment-starter))
4369 (= col (current-column))
4370 (prog1 (zerop (forward-line 1))
4371 (setq end (point)))))
4372 (cons beg end))
4373 range)
4374 (error range))))
d9e94c22
MS
4375
4376(defun c-literal-type (range)
4377 "Convenience function that given the result of `c-literal-limits',
a85fd6da
AM
4378returns nil or the type of literal that the range surrounds, one
4379of the symbols 'c, 'c++ or 'string. It's much faster than using
4380`c-in-literal' and is intended to be used when you need both the
4381type of a literal and its limits.
d9e94c22 4382
0386b551
AM
4383Note that this function might do hidden buffer changes. See the
4384comment at the start of cc-engine.el for more info."
4385
d9e94c22
MS
4386 (if (consp range)
4387 (save-excursion
4388 (goto-char (car range))
4389 (cond ((looking-at c-string-limit-regexp) 'string)
4390 ((or (looking-at "//") ; c++ line comment
4391 (and (looking-at "\\s<") ; comment starter
4392 (looking-at "#"))) ; awk comment.
4393 'c++)
4394 (t 'c))) ; Assuming the range is valid.
4395 range))
4396
4397\f
4398;; `c-find-decl-spots' and accompanying stuff.
4399
4400;; Variables used in `c-find-decl-spots' to cache the search done for
4401;; the first declaration in the last call. When that function starts,
4402;; it needs to back up over syntactic whitespace to look at the last
4403;; token before the region being searched. That can sometimes cause
4404;; moves back and forth over a quite large region of comments and
4405;; macros, which would be repeated for each changed character when
4406;; we're called during fontification, since font-lock refontifies the
4407;; current line for each change. Thus it's worthwhile to cache the
4408;; first match.
4409;;
4410;; `c-find-decl-syntactic-pos' is a syntactically relevant position in
4411;; the syntactic whitespace less or equal to some start position.
4412;; There's no cached value if it's nil.
4413;;
4414;; `c-find-decl-match-pos' is the match position if
4415;; `c-find-decl-prefix-search' matched before the syntactic whitespace
4416;; at `c-find-decl-syntactic-pos', or nil if there's no such match.
4417(defvar c-find-decl-syntactic-pos nil)
4418(make-variable-buffer-local 'c-find-decl-syntactic-pos)
4419(defvar c-find-decl-match-pos nil)
4420(make-variable-buffer-local 'c-find-decl-match-pos)
4421
4422(defsubst c-invalidate-find-decl-cache (change-min-pos)
4423 (and c-find-decl-syntactic-pos
4424 (< change-min-pos c-find-decl-syntactic-pos)
4425 (setq c-find-decl-syntactic-pos nil)))
4426
4427; (defface c-debug-decl-spot-face
4428; '((t (:background "Turquoise")))
4429; "Debug face to mark the spots where `c-find-decl-spots' stopped.")
4430; (defface c-debug-decl-sws-face
4431; '((t (:background "Khaki")))
4432; "Debug face to mark the syntactic whitespace between the declaration
4433; spots and the preceding token end.")
4434
4435(defmacro c-debug-put-decl-spot-faces (match-pos decl-pos)
4436 (when (facep 'c-debug-decl-spot-face)
0386b551 4437 `(c-save-buffer-state ((match-pos ,match-pos) (decl-pos ,decl-pos))
d9e94c22
MS
4438 (c-debug-add-face (max match-pos (point-min)) decl-pos
4439 'c-debug-decl-sws-face)
4440 (c-debug-add-face decl-pos (min (1+ decl-pos) (point-max))
4441 'c-debug-decl-spot-face))))
4442(defmacro c-debug-remove-decl-spot-faces (beg end)
4443 (when (facep 'c-debug-decl-spot-face)
0386b551 4444 `(c-save-buffer-state ()
d9e94c22
MS
4445 (c-debug-remove-face ,beg ,end 'c-debug-decl-spot-face)
4446 (c-debug-remove-face ,beg ,end 'c-debug-decl-sws-face))))
4447
4448(defmacro c-find-decl-prefix-search ()
4449 ;; Macro used inside `c-find-decl-spots'. It ought to be a defun,
4450 ;; but it contains lots of free variables that refer to things
4451 ;; inside `c-find-decl-spots'. The point is left at `cfd-match-pos'
4452 ;; if there is a match, otherwise at `cfd-limit'.
0386b551
AM
4453 ;;
4454 ;; This macro might do hidden buffer changes.
d9e94c22
MS
4455
4456 '(progn
4457 ;; Find the next property match position if we haven't got one already.
4458 (unless cfd-prop-match
4459 (save-excursion
4460 (while (progn
4461 (goto-char (next-single-property-change
4462 (point) 'c-type nil cfd-limit))
4463 (and (< (point) cfd-limit)
4464 (not (eq (c-get-char-property (1- (point)) 'c-type)
4465 'c-decl-end)))))
4466 (setq cfd-prop-match (point))))
4467
0386b551
AM
4468 ;; Find the next `c-decl-prefix-or-start-re' match if we haven't
4469 ;; got one already.
d9e94c22 4470 (unless cfd-re-match
0386b551
AM
4471
4472 (if (> cfd-re-match-end (point))
4473 (goto-char cfd-re-match-end))
4474
4475 (while (if (setq cfd-re-match-end
4476 (re-search-forward c-decl-prefix-or-start-re
4477 cfd-limit 'move))
4478
4479 ;; Match. Check if it's inside a comment or string literal.
4480 (c-got-face-at
4481 (if (setq cfd-re-match (match-end 1))
4482 ;; Matched the end of a token preceding a decl spot.
4483 (progn
4484 (goto-char cfd-re-match)
4485 (1- cfd-re-match))
4486 ;; Matched a token that start a decl spot.
4487 (goto-char (match-beginning 0))
4488 (point))
4489 c-literal-faces)
4490
4491 ;; No match. Finish up and exit the loop.
4492 (setq cfd-re-match cfd-limit)
4493 nil)
4494
4495 ;; Skip out of comments and string literals.
d9e94c22
MS
4496 (while (progn
4497 (goto-char (next-single-property-change
0386b551 4498 (point) 'face nil cfd-limit))
d9e94c22 4499 (and (< (point) cfd-limit)
0386b551
AM
4500 (c-got-face-at (point) c-literal-faces)))))
4501
4502 ;; If we matched at the decl start, we have to back up over the
4503 ;; preceding syntactic ws to set `cfd-match-pos' and to catch
4504 ;; any decl spots in the syntactic ws.
d9e94c22 4505 (unless cfd-re-match
0386b551
AM
4506 (c-backward-syntactic-ws)
4507 (setq cfd-re-match (point))))
d9e94c22
MS
4508
4509 ;; Choose whichever match is closer to the start.
4510 (if (< cfd-re-match cfd-prop-match)
4511 (setq cfd-match-pos cfd-re-match
4512 cfd-re-match nil)
4513 (setq cfd-match-pos cfd-prop-match
4514 cfd-prop-match nil))
4515
4516 (goto-char cfd-match-pos)
4517
4518 (when (< cfd-match-pos cfd-limit)
4519 ;; Skip forward past comments only so we don't skip macros.
4520 (c-forward-comments)
4521 ;; Set the position to continue at. We can avoid going over
4522 ;; the comments skipped above a second time, but it's possible
4523 ;; that the comment skipping has taken us past `cfd-prop-match'
4524 ;; since the property might be used inside comments.
4525 (setq cfd-continue-pos (if cfd-prop-match
4526 (min cfd-prop-match (point))
4527 (point))))))
4528
4529(defun c-find-decl-spots (cfd-limit cfd-decl-re cfd-face-checklist cfd-fun)
0386b551 4530 ;; Call CFD-FUN for each possible spot for a declaration, cast or
a85fd6da
AM
4531 ;; label from the point to CFD-LIMIT.
4532 ;;
4533 ;; CFD-FUN is called with point at the start of the spot. It's
4534 ;; passed two arguments: The first is the end position of the token
4535 ;; preceding the spot, or 0 for the implicit match at bob. The
4536 ;; second is a flag that is t when the match is inside a macro. If
4537 ;; CFD-FUN adds `c-decl-end' properties somewhere below the current
4538 ;; spot, it should return non-nil to ensure that the next search
4539 ;; will find them.
0386b551 4540 ;;
a85fd6da 4541 ;; Such a spot is:
0386b551
AM
4542 ;; o The first token after bob.
4543 ;; o The first token after the end of submatch 1 in
4544 ;; `c-decl-prefix-or-start-re' when that submatch matches.
4545 ;; o The start of each `c-decl-prefix-or-start-re' match when
4546 ;; submatch 1 doesn't match.
5a89f0a7 4547 ;; o The first token after the end of each occurrence of the
0386b551
AM
4548 ;; `c-type' text property with the value `c-decl-end', provided
4549 ;; `c-type-decl-end-used' is set.
4550 ;;
4551 ;; Only a spot that match CFD-DECL-RE and whose face is in the
4552 ;; CFD-FACE-CHECKLIST list causes CFD-FUN to be called. The face
4553 ;; check is disabled if CFD-FACE-CHECKLIST is nil.
d9e94c22
MS
4554 ;;
4555 ;; If the match is inside a macro then the buffer is narrowed to the
4556 ;; end of it, so that CFD-FUN can investigate the following tokens
4557 ;; without matching something that begins inside a macro and ends
4558 ;; outside it. It's to avoid this work that the CFD-DECL-RE and
4559 ;; CFD-FACE-CHECKLIST checks exist.
4560 ;;
0386b551
AM
4561 ;; The spots are visited approximately in order from top to bottom.
4562 ;; It's however the positions where `c-decl-prefix-or-start-re'
4563 ;; matches and where `c-decl-end' properties are found that are in
4564 ;; order. Since the spots often are at the following token, they
4565 ;; might be visited out of order insofar as more spots are reported
4566 ;; later on within the syntactic whitespace between the match
4567 ;; positions and their spots.
4568 ;;
4569 ;; It's assumed that comments and strings are fontified in the
d9e94c22
MS
4570 ;; searched range.
4571 ;;
4572 ;; This is mainly used in fontification, and so has an elaborate
4573 ;; cache to handle repeated calls from the same start position; see
4574 ;; the variables above.
4575 ;;
4576 ;; All variables in this function begin with `cfd-' to avoid name
4577 ;; collision with the (dynamically bound) variables used in CFD-FUN.
0386b551
AM
4578 ;;
4579 ;; This function might do hidden buffer changes.
d9e94c22 4580
0386b551
AM
4581 (let ((cfd-start-pos (point))
4582 (cfd-buffer-end (point-max))
4583 ;; The end of the token preceding the decl spot last found
4584 ;; with `c-decl-prefix-or-start-re'. `cfd-limit' if there's
4585 ;; no match.
d9e94c22 4586 cfd-re-match
0386b551
AM
4587 ;; The end position of the last `c-decl-prefix-or-start-re'
4588 ;; match. If this is greater than `cfd-continue-pos', the
4589 ;; next regexp search is started here instead.
4590 (cfd-re-match-end (point-min))
4591 ;; The end of the last `c-decl-end' found by
4592 ;; `c-find-decl-prefix-search'. `cfd-limit' if there's no
4593 ;; match. If searching for the property isn't needed then we
4594 ;; disable it by setting it to `cfd-limit' directly.
d9e94c22 4595 (cfd-prop-match (unless c-type-decl-end-used cfd-limit))
0386b551
AM
4596 ;; The end of the token preceding the decl spot last found by
4597 ;; `c-find-decl-prefix-search'. 0 for the implicit match at
4598 ;; bob. `cfd-limit' if there's no match. In other words,
4599 ;; this is the minimum of `cfd-re-match' and `cfd-prop-match'.
d9e94c22
MS
4600 (cfd-match-pos cfd-limit)
4601 ;; The position to continue searching at.
4602 cfd-continue-pos
4603 ;; The position of the last "real" token we've stopped at.
4604 ;; This can be greater than `cfd-continue-pos' when we get
4605 ;; hits inside macros or at `c-decl-end' positions inside
4606 ;; comments.
4607 (cfd-token-pos 0)
4608 ;; The end position of the last entered macro.
4609 (cfd-macro-end 0))
4610
4611 ;; Initialize by finding a syntactically relevant start position
0386b551
AM
4612 ;; before the point, and do the first `c-decl-prefix-or-start-re'
4613 ;; search unless we're at bob.
d9e94c22 4614
0386b551 4615 (let (start-in-literal start-in-macro syntactic-pos)
d9e94c22
MS
4616 ;; Must back up a bit since we look for the end of the previous
4617 ;; statement or declaration, which is earlier than the first
4618 ;; returned match.
4619
0386b551
AM
4620 (cond
4621 ;; First we need to move to a syntactically relevant position.
4622 ;; Begin by backing out of comment or string literals.
4623 ((and
4624 (when (c-got-face-at (point) c-literal-faces)
4625 ;; Try to use the faces to back up to the start of the
4626 ;; literal. FIXME: What if the point is on a declaration
4627 ;; inside a comment?
4628 (while (and (not (bobp))
4629 (c-got-face-at (1- (point)) c-literal-faces))
4630 (goto-char (previous-single-property-change
4631 (point) 'face nil (point-min))))
4632
4633 ;; XEmacs doesn't fontify the quotes surrounding string
4634 ;; literals.
4635 (and (featurep 'xemacs)
4636 (eq (get-text-property (point) 'face)
4637 'font-lock-string-face)
4638 (not (bobp))
4639 (progn (backward-char)
4640 (not (looking-at c-string-limit-regexp)))
4641 (forward-char))
4642
4643 ;; Don't trust the literal to contain only literal faces
4644 ;; (the font lock package might not have fontified the
4645 ;; start of it at all, for instance) so check that we have
4646 ;; arrived at something that looks like a start or else
4647 ;; resort to `c-literal-limits'.
4648 (unless (looking-at c-literal-start-regexp)
4649 (let ((range (c-literal-limits)))
4650 (if range (goto-char (car range)))))
4651
4652 (setq start-in-literal (point)))
4653
4654 ;; The start is in a literal. If the limit is in the same
4655 ;; one we don't have to find a syntactic position etc. We
4656 ;; only check that if the limit is at or before bonl to save
4657 ;; time; it covers the by far most common case when font-lock
4658 ;; refontifies the current line only.
4659 (<= cfd-limit (c-point 'bonl cfd-start-pos))
4660 (save-excursion
4661 (goto-char cfd-start-pos)
4662 (while (progn
4663 (goto-char (next-single-property-change
4664 (point) 'face nil cfd-limit))
4665 (and (< (point) cfd-limit)
4666 (c-got-face-at (point) c-literal-faces))))
4667 (= (point) cfd-limit)))
4668
4669 ;; Completely inside a literal. Set up variables to trig the
4670 ;; (< cfd-continue-pos cfd-start-pos) case below and it'll
4671 ;; find a suitable start position.
4672 (setq cfd-continue-pos start-in-literal))
4673
4674 ;; Check if the region might be completely inside a macro, to
4675 ;; optimize that like the completely-inside-literal above.
4676 ((save-excursion
4677 (and (= (forward-line 1) 0)
4678 (bolp) ; forward-line has funny behavior at eob.
4679 (>= (point) cfd-limit)
4680 (progn (backward-char)
4681 (eq (char-before) ?\\))))
4682 ;; (Maybe) completely inside a macro. Only need to trig the
4683 ;; (< cfd-continue-pos cfd-start-pos) case below to make it
4684 ;; set things up.
4685 (setq cfd-continue-pos (1- cfd-start-pos)
4686 start-in-macro t))
d9e94c22 4687
0386b551
AM
4688 (t
4689 ;; Back out of any macro so we don't miss any declaration
4690 ;; that could follow after it.
4691 (when (c-beginning-of-macro)
4692 (setq start-in-macro t))
4693
4694 ;; Now we're at a proper syntactically relevant position so we
4695 ;; can use the cache. But first clear it if it applied
4696 ;; further down.
4697 (c-invalidate-find-decl-cache cfd-start-pos)
4698
4699 (setq syntactic-pos (point))
4700 (unless (eq syntactic-pos c-find-decl-syntactic-pos)
4701 ;; Don't have to do this if the cache is relevant here,
4702 ;; typically if the same line is refontified again. If
4703 ;; we're just some syntactic whitespace further down we can
4704 ;; still use the cache to limit the skipping.
4705 (c-backward-syntactic-ws c-find-decl-syntactic-pos))
4706
4707 ;; If we hit `c-find-decl-syntactic-pos' and
4708 ;; `c-find-decl-match-pos' is set then we install the cached
4709 ;; values. If we hit `c-find-decl-syntactic-pos' and
4710 ;; `c-find-decl-match-pos' is nil then we know there's no decl
4711 ;; prefix in the whitespace before `c-find-decl-syntactic-pos'
4712 ;; and so we can continue the search from this point. If we
4713 ;; didn't hit `c-find-decl-syntactic-pos' then we're now in
4714 ;; the right spot to begin searching anyway.
4715 (if (and (eq (point) c-find-decl-syntactic-pos)
4716 c-find-decl-match-pos)
d9e94c22
MS
4717 (setq cfd-match-pos c-find-decl-match-pos
4718 cfd-continue-pos syntactic-pos)
0386b551
AM
4719
4720 (setq c-find-decl-syntactic-pos syntactic-pos)
4721
4722 (when (if (bobp)
4723 ;; Always consider bob a match to get the first
4724 ;; declaration in the file. Do this separately instead of
4725 ;; letting `c-decl-prefix-or-start-re' match bob, so that
4726 ;; regexp always can consume at least one character to
4727 ;; ensure that we won't get stuck in an infinite loop.
4728 (setq cfd-re-match 0)
4729 (backward-char)
4730 (c-beginning-of-current-token)
4731 (< (point) cfd-limit))
4732 ;; Do an initial search now. In the bob case above it's
4733 ;; only done to search for a `c-decl-end' spot.
4734 (c-find-decl-prefix-search))
4735
4736 (setq c-find-decl-match-pos (and (< cfd-match-pos cfd-start-pos)
4737 cfd-match-pos)))))
4738
4739 ;; Advance `cfd-continue-pos' if it's before the start position.
4740 ;; The closest continue position that might have effect at or
4741 ;; after the start depends on what we started in. This also
4742 ;; finds a suitable start position in the special cases when the
4743 ;; region is completely within a literal or macro.
4744 (when (and cfd-continue-pos (< cfd-continue-pos cfd-start-pos))
4745
4746 (cond
4747 (start-in-macro
4748 ;; If we're in a macro then it's the closest preceding token
4749 ;; in the macro. Check this before `start-in-literal',
4750 ;; since if we're inside a literal in a macro, the preceding
4751 ;; token is earlier than any `c-decl-end' spot inside the
4752 ;; literal (comment).
4753 (goto-char (or start-in-literal cfd-start-pos))
4754 ;; The only syntactic ws in macros are comments.
d9e94c22 4755 (c-backward-comments)
0386b551
AM
4756 (backward-char)
4757 (c-beginning-of-current-token))
4758
4759 (start-in-literal
4760 ;; If we're in a comment it can only be the closest
4761 ;; preceding `c-decl-end' position within that comment, if
4762 ;; any. Go back to the beginning of such a property so that
4763 ;; `c-find-decl-prefix-search' will find the end of it.
4764 ;; (Can't stop at the end and install it directly on
4765 ;; `cfd-prop-match' since that variable might be cleared
4766 ;; after `cfd-fun' below.)
4767 ;;
4768 ;; Note that if the literal is a string then the property
4769 ;; search will simply skip to the beginning of it right
4770 ;; away.
4771 (if (not c-type-decl-end-used)
4772 (goto-char start-in-literal)
4773 (goto-char cfd-start-pos)
4774 (while (progn
4775 (goto-char (previous-single-property-change
4776 (point) 'c-type nil start-in-literal))
4777 (and (> (point) start-in-literal)
4778 (not (eq (c-get-char-property (point) 'c-type)
4779 'c-decl-end))))))
4780
4781 (when (= (point) start-in-literal)
4782 ;; Didn't find any property inside the comment, so we can
4783 ;; skip it entirely. (This won't skip past a string, but
4784 ;; that'll be handled quickly by the next
4785 ;; `c-find-decl-prefix-search' anyway.)
4786 (c-forward-single-comment)
4787 (if (> (point) cfd-limit)
4788 (goto-char cfd-limit))))
d9e94c22 4789
0386b551
AM
4790 (t
4791 ;; If we started in normal code, the only match that might
4792 ;; apply before the start is what we already got in
4793 ;; `cfd-match-pos' so we can continue at the start position.
4794 ;; (Note that we don't get here if the first match is below
4795 ;; it.)
4796 (goto-char cfd-start-pos)))
4797
4798 ;; Delete found matches if they are before our new continue
4799 ;; position, so that `c-find-decl-prefix-search' won't back up
4800 ;; to them later on.
4801 (setq cfd-continue-pos (point))
4802 (when (and cfd-re-match (< cfd-re-match cfd-continue-pos))
4803 (setq cfd-re-match nil))
4804 (when (and cfd-prop-match (< cfd-prop-match cfd-continue-pos))
4805 (setq cfd-prop-match nil)))
4806
4807 (if syntactic-pos
4808 ;; This is the normal case and we got a proper syntactic
4809 ;; position. If there's a match then it's always outside
4810 ;; macros and comments, so advance to the next token and set
4811 ;; `cfd-token-pos'. The loop below will later go back using
4812 ;; `cfd-continue-pos' to fix declarations inside the
4813 ;; syntactic ws.
4814 (when (and cfd-match-pos (< cfd-match-pos syntactic-pos))
4815 (goto-char syntactic-pos)
4816 (c-forward-syntactic-ws)
4817 (and cfd-continue-pos
4818 (< cfd-continue-pos (point))
4819 (setq cfd-token-pos (point))))
4820
4821 ;; Have one of the special cases when the region is completely
4822 ;; within a literal or macro. `cfd-continue-pos' is set to a
4823 ;; good start position for the search, so do it.
4824 (c-find-decl-prefix-search)))
d9e94c22 4825
51c9af45 4826 ;; Now loop. Round what? (ACM, 2006/7/5). We already got the first match.
d9e94c22
MS
4827
4828 (while (progn
4829 (while (and
4830 (< cfd-match-pos cfd-limit)
4831
4832 (or
4833 ;; Kludge to filter out matches on the "<" that
4834 ;; aren't open parens, for the sake of languages
4835 ;; that got `c-recognize-<>-arglists' set.
4836 (and (eq (char-before cfd-match-pos) ?<)
4837 (not (c-get-char-property (1- cfd-match-pos)
4838 'syntax-table)))
4839
4840 ;; If `cfd-continue-pos' is less or equal to
4841 ;; `cfd-token-pos', we've got a hit inside a macro
4842 ;; that's in the syntactic whitespace before the last
4843 ;; "real" declaration we've checked. If they're equal
4844 ;; we've arrived at the declaration a second time, so
4845 ;; there's nothing to do.
4846 (= cfd-continue-pos cfd-token-pos)
4847
4848 (progn
4849 ;; If `cfd-continue-pos' is less than `cfd-token-pos'
4850 ;; we're still searching for declarations embedded in
4851 ;; the syntactic whitespace. In that case we need
4852 ;; only to skip comments and not macros, since they
4853 ;; can't be nested, and that's already been done in
4854 ;; `c-find-decl-prefix-search'.
4855 (when (> cfd-continue-pos cfd-token-pos)
4856 (c-forward-syntactic-ws)
4857 (setq cfd-token-pos (point)))
4858
4859 ;; Continue if the following token fails the
4860 ;; CFD-DECL-RE and CFD-FACE-CHECKLIST checks.
4861 (when (or (>= (point) cfd-limit)
4862 (not (looking-at cfd-decl-re))
4863 (and cfd-face-checklist
4864 (not (c-got-face-at
4865 (point) cfd-face-checklist))))
4866 (goto-char cfd-continue-pos)
4867 t)))
4868
4869 (< (point) cfd-limit))
4870 (c-find-decl-prefix-search))
4871
4872 (< (point) cfd-limit))
4873
0386b551
AM
4874 (when (and
4875 (>= (point) cfd-start-pos)
d9e94c22 4876
0386b551
AM
4877 (progn
4878 ;; Narrow to the end of the macro if we got a hit inside
4879 ;; one, to avoid recognizing things that start inside the
4880 ;; macro and end outside it.
4881 (when (> cfd-match-pos cfd-macro-end)
4882 ;; Not in the same macro as in the previous round.
4883 (save-excursion
4884 (goto-char cfd-match-pos)
4885 (setq cfd-macro-end
4886 (if (save-excursion (and (c-beginning-of-macro)
4887 (< (point) cfd-match-pos)))
4888 (progn (c-end-of-macro)
4889 (point))
4890 0))))
4891
4892 (if (zerop cfd-macro-end)
4893 t
4894 (if (> cfd-macro-end (point))
4895 (progn (narrow-to-region (point-min) cfd-macro-end)
4896 t)
4897 ;; The matched token was the last thing in the macro,
4898 ;; so the whole match is bogus.
4899 (setq cfd-macro-end 0)
4900 nil))))
d9e94c22
MS
4901
4902 (c-debug-put-decl-spot-faces cfd-match-pos (point))
0386b551
AM
4903 (if (funcall cfd-fun cfd-match-pos (/= cfd-macro-end 0))
4904 (setq cfd-prop-match nil))
d9e94c22
MS
4905
4906 (when (/= cfd-macro-end 0)
4907 ;; Restore limits if we did macro narrowment above.
4908 (narrow-to-region (point-min) cfd-buffer-end)))
4909
4910 (goto-char cfd-continue-pos)
4911 (if (= cfd-continue-pos cfd-limit)
4912 (setq cfd-match-pos cfd-limit)
4913 (c-find-decl-prefix-search)))))
4914
4915\f
4916;; A cache for found types.
4917
4918;; Buffer local variable that contains an obarray with the types we've
4919;; found. If a declaration is recognized somewhere we record the
4920;; fully qualified identifier in it to recognize it as a type
4921;; elsewhere in the file too. This is not accurate since we do not
4922;; bother with the scoping rules of the languages, but in practice the
4923;; same name is seldom used as both a type and something else in a
4924;; file, and we only use this as a last resort in ambiguous cases (see
0386b551
AM
4925;; `c-forward-decl-or-cast-1').
4926;;
580fba94
AM
4927;; Not every type need be in this cache. However, things which have
4928;; ceased to be types must be removed from it.
4929;;
0386b551
AM
4930;; Template types in C++ are added here too but with the template
4931;; arglist replaced with "<>" in references or "<" for the one in the
4932;; primary type. E.g. the type "Foo<A,B>::Bar<C>" is stored as
4933;; "Foo<>::Bar<". This avoids storing very long strings (since C++
4934;; template specs can be fairly sized programs in themselves) and
4935;; improves the hit ratio (it's a type regardless of the template
4936;; args; it's just not the same type, but we're only interested in
4937;; recognizing types, not telling distinct types apart). Note that
4938;; template types in references are added here too; from the example
4939;; above there will also be an entry "Foo<".
d9e94c22
MS
4940(defvar c-found-types nil)
4941(make-variable-buffer-local 'c-found-types)
4942
4943(defsubst c-clear-found-types ()
4944 ;; Clears `c-found-types'.
d9e94c22
MS
4945 (setq c-found-types (make-vector 53 0)))
4946
4947(defun c-add-type (from to)
4948 ;; Add the given region as a type in `c-found-types'. If the region
4949 ;; doesn't match an existing type but there is a type which is equal
4950 ;; to the given one except that the last character is missing, then
4951 ;; the shorter type is removed. That's done to avoid adding all
4952 ;; prefixes of a type as it's being entered and font locked. This
4953 ;; doesn't cover cases like when characters are removed from a type
4954 ;; or added in the middle. We'd need the position of point when the
4955 ;; font locking is invoked to solve this well.
0386b551
AM
4956 ;;
4957 ;; This function might do hidden buffer changes.
4958 (let ((type (c-syntactic-content from to c-recognize-<>-arglists)))
4959 (unless (intern-soft type c-found-types)
4960 (unintern (substring type 0 -1) c-found-types)
4961 (intern type c-found-types))))
d9e94c22 4962
580fba94
AM
4963(defun c-unfind-type (name)
4964 ;; Remove the "NAME" from c-found-types, if present.
4965 (unintern name c-found-types))
4966
d9e94c22
MS
4967(defsubst c-check-type (from to)
4968 ;; Return non-nil if the given region contains a type in
4969 ;; `c-found-types'.
0386b551
AM
4970 ;;
4971 ;; This function might do hidden buffer changes.
4972 (intern-soft (c-syntactic-content from to c-recognize-<>-arglists)
4973 c-found-types))
d9e94c22
MS
4974
4975(defun c-list-found-types ()
4976 ;; Return all the types in `c-found-types' as a sorted list of
4977 ;; strings.
4978 (let (type-list)
4979 (mapatoms (lambda (type)
4980 (setq type-list (cons (symbol-name type)
4981 type-list)))
4982 c-found-types)
4983 (sort type-list 'string-lessp)))
a66cd3ee 4984
2f42c75f
DN
4985;; Shut up the byte compiler.
4986(defvar c-maybe-stale-found-type)
4987
580fba94
AM
4988(defun c-trim-found-types (beg end old-len)
4989 ;; An after change function which, in conjunction with the info in
4990 ;; c-maybe-stale-found-type (set in c-before-change), removes a type
4991 ;; from `c-found-types', should this type have become stale. For
4992 ;; example, this happens to "foo" when "foo \n bar();" becomes
4993 ;; "foo(); \n bar();". Such stale types, if not removed, foul up
4994 ;; the fontification.
b414f371 4995 ;;
580fba94
AM
4996 ;; Have we, perhaps, added non-ws characters to the front/back of a found
4997 ;; type?
4998 (when (> end beg)
4999 (save-excursion
5000 (when (< end (point-max))
5001 (goto-char end)
5002 (if (and (c-beginning-of-current-token) ; only moves when we started in the middle
5003 (progn (goto-char end)
5004 (c-end-of-current-token)))
5005 (c-unfind-type (buffer-substring-no-properties
5006 end (point)))))
5007 (when (> beg (point-min))
5008 (goto-char beg)
5009 (if (and (c-end-of-current-token) ; only moves when we started in the middle
5010 (progn (goto-char beg)
5011 (c-beginning-of-current-token)))
5012 (c-unfind-type (buffer-substring-no-properties
5013 (point) beg))))))
b414f371 5014
580fba94
AM
5015 (if c-maybe-stale-found-type ; e.g. (c-decl-id-start "foo" 97 107 " (* ooka) " "o")
5016 (cond
5017 ;; Changing the amount of (already existing) whitespace - don't do anything.
5018 ((and (c-partial-ws-p beg end)
5019 (or (= beg end) ; removal of WS
580fba94
AM
5020 (string-match "^[ \t\n\r\f\v]*$" (nth 5 c-maybe-stale-found-type)))))
5021
5022 ;; The syntactic relationship which defined a "found type" has been
5023 ;; destroyed.
5024 ((eq (car c-maybe-stale-found-type) 'c-decl-id-start)
5025 (c-unfind-type (cadr c-maybe-stale-found-type)))
5026;; ((eq (car c-maybe-stale-found-type) 'c-decl-type-start) FIXME!!!
5027 )))
5028
d9e94c22 5029\f
dd969a56
AM
5030;; Setting and removing syntax properties on < and > in languages (C++
5031;; and Java) where they can be template/generic delimiters as well as
5032;; their normal meaning of "less/greater than".
5033
5034;; Normally, < and > have syntax 'punctuation'. When they are found to
5035;; be delimiters, they are marked as such with the category properties
5036;; c-<-as-paren-syntax, c->-as-paren-syntax respectively.
5037
5038;; STRATEGY:
5039;;
5040;; It is impossible to determine with certainty whether a <..> pair in
5041;; C++ is two comparison operators or is template delimiters, unless
5042;; one duplicates a lot of a C++ compiler. For example, the following
5043;; code fragment:
5044;;
5045;; foo (a < b, c > d) ;
5046;;
5047;; could be a function call with two integer parameters (each a
5048;; relational expression), or it could be a constructor for class foo
5049;; taking one parameter d of templated type "a < b, c >". They are
5050;; somewhat easier to distinguish in Java.
5051;;
5052;; The strategy now (2010-01) adopted is to mark and unmark < and
5053;; > IN MATCHING PAIRS ONLY. [Previously, they were marked
5054;; individually when their context so indicated. This gave rise to
53964682 5055;; intractable problems when one of a matching pair was deleted, or
dd969a56
AM
5056;; pulled into a literal.]
5057;;
5058;; At each buffer change, the syntax-table properties are removed in a
5059;; before-change function and reapplied, when needed, in an
5060;; after-change function. It is far more important that the
5061;; properties get removed when they they are spurious than that they
5062;; be present when wanted.
5063;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
5064(defun c-clear-<-pair-props (&optional pos)
5065 ;; POS (default point) is at a < character. If it is marked with
5066 ;; open paren syntax-table text property, remove the property,
5067 ;; together with the close paren property on the matching > (if
5068 ;; any).
5069 (save-excursion
5070 (if pos
5071 (goto-char pos)
5072 (setq pos (point)))
5073 (when (equal (c-get-char-property (point) 'syntax-table)
5074 c-<-as-paren-syntax)
5075 (with-syntax-table c-no-parens-syntax-table ; ignore unbalanced [,{,(,..
5076 (c-go-list-forward))
5077 (when (equal (c-get-char-property (1- (point)) 'syntax-table)
5078 c->-as-paren-syntax) ; should always be true.
8a249abc
AM
5079 (c-clear-char-property (1- (point)) 'category))
5080 (c-clear-char-property pos 'category))))
dd969a56
AM
5081
5082(defun c-clear->-pair-props (&optional pos)
5083 ;; POS (default point) is at a > character. If it is marked with
5084 ;; close paren syntax-table property, remove the property, together
5085 ;; with the open paren property on the matching < (if any).
5086 (save-excursion
5087 (if pos
5088 (goto-char pos)
5089 (setq pos (point)))
5090 (when (equal (c-get-char-property (point) 'syntax-table)
5091 c->-as-paren-syntax)
5092 (with-syntax-table c-no-parens-syntax-table ; ignore unbalanced [,{,(,..
5093 (c-go-up-list-backward))
5094 (when (equal (c-get-char-property (point) 'syntax-table)
5095 c-<-as-paren-syntax) ; should always be true.
8a249abc
AM
5096 (c-clear-char-property (point) 'category))
5097 (c-clear-char-property pos 'category))))
dd969a56
AM
5098
5099(defun c-clear-<>-pair-props (&optional pos)
5100 ;; POS (default point) is at a < or > character. If it has an
5101 ;; open/close paren syntax-table property, remove this property both
5102 ;; from the current character and its partner (which will also be
5103 ;; thusly marked).
5104 (cond
5105 ((eq (char-after) ?\<)
5106 (c-clear-<-pair-props pos))
5107 ((eq (char-after) ?\>)
5108 (c-clear->-pair-props pos))
5109 (t (c-benign-error
5110 "c-clear-<>-pair-props called from wrong position"))))
5111
5112(defun c-clear-<-pair-props-if-match-after (lim &optional pos)
5113 ;; POS (default point) is at a < character. If it is both marked
5114 ;; with open/close paren syntax-table property, and has a matching >
5115 ;; (also marked) which is after LIM, remove the property both from
43a91810
AM
5116 ;; the current > and its partner. Return t when this happens, nil
5117 ;; when it doesn't.
dd969a56
AM
5118 (save-excursion
5119 (if pos
5120 (goto-char pos)
5121 (setq pos (point)))
5122 (when (equal (c-get-char-property (point) 'syntax-table)
5123 c-<-as-paren-syntax)
5124 (with-syntax-table c-no-parens-syntax-table ; ignore unbalanced [,{,(,..
5125 (c-go-list-forward))
5126 (when (and (>= (point) lim)
5127 (equal (c-get-char-property (1- (point)) 'syntax-table)
5128 c->-as-paren-syntax)) ; should always be true.
5129 (c-unmark-<->-as-paren (1- (point)))
43a91810
AM
5130 (c-unmark-<->-as-paren pos))
5131 t)))
dd969a56
AM
5132
5133(defun c-clear->-pair-props-if-match-before (lim &optional pos)
5134 ;; POS (default point) is at a > character. If it is both marked
5135 ;; with open/close paren syntax-table property, and has a matching <
5136 ;; (also marked) which is before LIM, remove the property both from
43a91810
AM
5137 ;; the current < and its partner. Return t when this happens, nil
5138 ;; when it doesn't.
dd969a56
AM
5139 (save-excursion
5140 (if pos
5141 (goto-char pos)
5142 (setq pos (point)))
5143 (when (equal (c-get-char-property (point) 'syntax-table)
5144 c->-as-paren-syntax)
5145 (with-syntax-table c-no-parens-syntax-table ; ignore unbalanced [,{,(,..
5146 (c-go-up-list-backward))
5147 (when (and (<= (point) lim)
5148 (equal (c-get-char-property (point) 'syntax-table)
5149 c-<-as-paren-syntax)) ; should always be true.
5150 (c-unmark-<->-as-paren (point))
43a91810
AM
5151 (c-unmark-<->-as-paren pos))
5152 t)))
dd969a56 5153
f3b554af
GM
5154;; Set by c-common-init in cc-mode.el.
5155(defvar c-new-BEG)
5156(defvar c-new-END)
5157
dd969a56
AM
5158(defun c-before-change-check-<>-operators (beg end)
5159 ;; Unmark certain pairs of "< .... >" which are currently marked as
5160 ;; template/generic delimiters. (This marking is via syntax-table
5161 ;; text properties).
5162 ;;
5163 ;; These pairs are those which are in the current "statement" (i.e.,
5164 ;; the region between the {, }, or ; before BEG and the one after
5165 ;; END), and which enclose any part of the interval (BEG END).
5166 ;;
5167 ;; Note that in C++ (?and Java), template/generic parens cannot
5168 ;; enclose a brace or semicolon, so we use these as bounds on the
5169 ;; region we must work on.
5170 ;;
5171 ;; This function is called from before-change-functions (via
5172 ;; c-get-state-before-change-functions). Thus the buffer is widened,
5173 ;; and point is undefined, both at entry and exit.
5174 ;;
5175 ;; FIXME!!! This routine ignores the possibility of macros entirely.
5176 ;; 2010-01-29.
5177 (save-excursion
5178 (let ((beg-lit-limits (progn (goto-char beg) (c-literal-limits)))
43a91810
AM
5179 (end-lit-limits (progn (goto-char end) (c-literal-limits)))
5180 new-beg new-end need-new-beg need-new-end)
dd969a56
AM
5181 ;; Locate the barrier before the changed region
5182 (goto-char (if beg-lit-limits (car beg-lit-limits) beg))
5183 (c-syntactic-skip-backward "^;{}" (max (- beg 2048) (point-min)))
43a91810 5184 (setq new-beg (point))
dd969a56
AM
5185
5186 ;; Remove the syntax-table properties from each pertinent <...> pair.
5187 ;; Firsly, the ones with the < before beg and > after beg.
5188 (while (c-search-forward-char-property 'category 'c-<-as-paren-syntax beg)
43a91810
AM
5189 (if (c-clear-<-pair-props-if-match-after beg (1- (point)))
5190 (setq need-new-beg t)))
dd969a56
AM
5191
5192 ;; Locate the barrier after END.
5193 (goto-char (if end-lit-limits (cdr end-lit-limits) end))
5194 (c-syntactic-re-search-forward "[;{}]"
5195 (min (+ end 2048) (point-max)) 'end)
43a91810 5196 (setq new-end (point))
dd969a56
AM
5197
5198 ;; Remove syntax-table properties from the remaining pertinent <...>
5199 ;; pairs, those with a > after end and < before end.
5200 (while (c-search-backward-char-property 'category 'c->-as-paren-syntax end)
43a91810
AM
5201 (if (c-clear->-pair-props-if-match-before end)
5202 (setq need-new-end t)))
5203
5204 ;; Extend the fontification region, if needed.
5205 (when need-new-beg
5206 (goto-char new-beg)
5207 (c-forward-syntactic-ws)
5208 (and (< (point) c-new-BEG) (setq c-new-BEG (point))))
5209
5210 (when need-new-end
5211 (and (> new-end c-new-END) (setq c-new-END new-end))))))
dd969a56
AM
5212
5213
d9e94c22 5214
0386b551
AM
5215(defun c-after-change-check-<>-operators (beg end)
5216 ;; This is called from `after-change-functions' when
5217 ;; c-recognize-<>-arglists' is set. It ensures that no "<" or ">"
5218 ;; chars with paren syntax become part of another operator like "<<"
5219 ;; or ">=".
5220 ;;
5221 ;; This function might do hidden buffer changes.
5222
cb694ab7
AM
5223 (save-excursion
5224 (goto-char beg)
5225 (when (or (looking-at "[<>]")
5226 (< (skip-chars-backward "<>") 0))
5227
0386b551 5228 (goto-char beg)
cb694ab7
AM
5229 (c-beginning-of-current-token)
5230 (when (and (< (point) beg)
5231 (looking-at c-<>-multichar-token-regexp)
5232 (< beg (setq beg (match-end 0))))
5233 (while (progn (skip-chars-forward "^<>" beg)
5234 (< (point) beg))
dd969a56 5235 (c-clear-<>-pair-props)
cb694ab7
AM
5236 (forward-char))))
5237
5238 (when (< beg end)
5239 (goto-char end)
0386b551
AM
5240 (when (or (looking-at "[<>]")
5241 (< (skip-chars-backward "<>") 0))
5242
cb694ab7 5243 (goto-char end)
0386b551 5244 (c-beginning-of-current-token)
cb694ab7 5245 (when (and (< (point) end)
0386b551 5246 (looking-at c-<>-multichar-token-regexp)
cb694ab7
AM
5247 (< end (setq end (match-end 0))))
5248 (while (progn (skip-chars-forward "^<>" end)
5249 (< (point) end))
dd969a56 5250 (c-clear-<>-pair-props)
cb694ab7 5251 (forward-char)))))))
d9e94c22 5252
dd969a56
AM
5253
5254\f
5255;; Handling of small scale constructs like types and names.
5256
d9e94c22
MS
5257;; Dynamically bound variable that instructs `c-forward-type' to also
5258;; treat possible types (i.e. those that it normally returns 'maybe or
5259;; 'found for) as actual types (and always return 'found for them).
5260;; This means that it records them in `c-record-type-identifiers' if
5261;; that is set, and that it adds them to `c-found-types'.
5262(defvar c-promote-possible-types nil)
5263
0386b551
AM
5264;; Dynamically bound variable that instructs `c-forward-<>-arglist' to
5265;; mark up successfully parsed arglists with paren syntax properties on
5266;; the surrounding angle brackets and with `c-<>-arg-sep' in the
5267;; `c-type' property of each argument separating comma.
5268;;
5269;; Setting this variable also makes `c-forward-<>-arglist' recurse into
5270;; all arglists for side effects (i.e. recording types), otherwise it
5271;; exploits any existing paren syntax properties to quickly jump to the
5272;; end of already parsed arglists.
5273;;
5274;; Marking up the arglists is not the default since doing that correctly
5275;; depends on a proper value for `c-restricted-<>-arglists'.
5276(defvar c-parse-and-markup-<>-arglists nil)
5277
d9e94c22 5278;; Dynamically bound variable that instructs `c-forward-<>-arglist' to
037558bf
MS
5279;; not accept arglists that contain binary operators.
5280;;
5281;; This is primarily used to handle C++ template arglists. C++
5282;; disambiguates them by checking whether the preceding name is a
5283;; template or not. We can't do that, so we assume it is a template
5284;; if it can be parsed as one. That usually works well since
5285;; comparison expressions on the forms "a < b > c" or "a < b, c > d"
5286;; in almost all cases would be pointless.
5287;;
5288;; However, in function arglists, e.g. in "foo (a < b, c > d)", we
5289;; should let the comma separate the function arguments instead. And
5290;; in a context where the value of the expression is taken, e.g. in
5291;; "if (a < b || c > d)", it's probably not a template.
5292(defvar c-restricted-<>-arglists nil)
d9e94c22 5293
0386b551
AM
5294;; Dynamically bound variables that instructs
5295;; `c-forward-keyword-clause', `c-forward-<>-arglist',
5296;; `c-forward-name', `c-forward-type', `c-forward-decl-or-cast-1', and
5297;; `c-forward-label' to record the ranges of all the type and
5298;; reference identifiers they encounter. They will build lists on
5299;; these variables where each element is a cons of the buffer
5300;; positions surrounding each identifier. This recording is only
5301;; activated when `c-record-type-identifiers' is non-nil.
d9e94c22
MS
5302;;
5303;; All known types that can't be identifiers are recorded, and also
5304;; other possible types if `c-promote-possible-types' is set.
5305;; Recording is however disabled inside angle bracket arglists that
5306;; are encountered inside names and other angle bracket arglists.
0386b551 5307;; Such occurrences are taken care of by `c-font-lock-<>-arglists'
d9e94c22
MS
5308;; instead.
5309;;
5310;; Only the names in C++ template style references (e.g. "tmpl" in
5311;; "tmpl<a,b>::foo") are recorded as references, other references
5312;; aren't handled here.
0386b551
AM
5313;;
5314;; `c-forward-label' records the label identifier(s) on
5315;; `c-record-ref-identifiers'.
d9e94c22
MS
5316(defvar c-record-type-identifiers nil)
5317(defvar c-record-ref-identifiers nil)
5318
0386b551
AM
5319;; This variable will receive a cons cell of the range of the last
5320;; single identifier symbol stepped over by `c-forward-name' if it's
5321;; successful. This is the range that should be put on one of the
5322;; record lists above by the caller. It's assigned nil if there's no
5323;; such symbol in the name.
d9e94c22
MS
5324(defvar c-last-identifier-range nil)
5325
5326(defmacro c-record-type-id (range)
5327 (if (eq (car-safe range) 'cons)
5328 ;; Always true.
5329 `(setq c-record-type-identifiers
5330 (cons ,range c-record-type-identifiers))
5331 `(let ((range ,range))
5332 (if range
5333 (setq c-record-type-identifiers
5334 (cons range c-record-type-identifiers))))))
5335
5336(defmacro c-record-ref-id (range)
5337 (if (eq (car-safe range) 'cons)
5338 ;; Always true.
5339 `(setq c-record-ref-identifiers
5340 (cons ,range c-record-ref-identifiers))
5341 `(let ((range ,range))
5342 (if range
5343 (setq c-record-ref-identifiers
5344 (cons range c-record-ref-identifiers))))))
5345
5346;; Dynamically bound variable that instructs `c-forward-type' to
5347;; record the ranges of types that only are found. Behaves otherwise
5348;; like `c-record-type-identifiers'.
5349(defvar c-record-found-types nil)
5350
5351(defmacro c-forward-keyword-prefixed-id (type)
5352 ;; Used internally in `c-forward-keyword-clause' to move forward
5353 ;; over a type (if TYPE is 'type) or a name (otherwise) which
5354 ;; possibly is prefixed by keywords and their associated clauses.
5355 ;; Try with a type/name first to not trip up on those that begin
5356 ;; with a keyword. Return t if a known or found type is moved
5357 ;; over. The point is clobbered if nil is returned. If range
5358 ;; recording is enabled, the identifier is recorded on as a type
5359 ;; if TYPE is 'type or as a reference if TYPE is 'ref.
0386b551
AM
5360 ;;
5361 ;; This macro might do hidden buffer changes.
d9e94c22
MS
5362 `(let (res)
5363 (while (if (setq res ,(if (eq type 'type)
5364 `(c-forward-type)
5365 `(c-forward-name)))
5366 nil
5367 (and (looking-at c-keywords-regexp)
0386b551 5368 (c-forward-keyword-clause 1))))
d9e94c22
MS
5369 (when (memq res '(t known found prefix))
5370 ,(when (eq type 'ref)
5371 `(when c-record-type-identifiers
5372 (c-record-ref-id c-last-identifier-range)))
5373 t)))
5374
0386b551 5375(defmacro c-forward-id-comma-list (type update-safe-pos)
d9e94c22
MS
5376 ;; Used internally in `c-forward-keyword-clause' to move forward
5377 ;; over a comma separated list of types or names using
5378 ;; `c-forward-keyword-prefixed-id'.
0386b551
AM
5379 ;;
5380 ;; This macro might do hidden buffer changes.
d9e94c22 5381 `(while (and (progn
0386b551
AM
5382 ,(when update-safe-pos
5383 `(setq safe-pos (point)))
d9e94c22
MS
5384 (eq (char-after) ?,))
5385 (progn
5386 (forward-char)
5387 (c-forward-syntactic-ws)
5388 (c-forward-keyword-prefixed-id ,type)))))
5389
0386b551
AM
5390(defun c-forward-keyword-clause (match)
5391 ;; Submatch MATCH in the current match data is assumed to surround a
5392 ;; token. If it's a keyword, move over it and any immediately
5393 ;; following clauses associated with it, stopping at the start of
5394 ;; the next token. t is returned in that case, otherwise the point
d9e94c22
MS
5395 ;; stays and nil is returned. The kind of clauses that are
5396 ;; recognized are those specified by `c-type-list-kwds',
5397 ;; `c-ref-list-kwds', `c-colon-type-list-kwds',
5398 ;; `c-paren-nontype-kwds', `c-paren-type-kwds', `c-<>-type-kwds',
5399 ;; and `c-<>-arglist-kwds'.
0386b551
AM
5400 ;;
5401 ;; This function records identifier ranges on
5402 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
5403 ;; `c-record-type-identifiers' is non-nil.
5404 ;;
5405 ;; Note that for `c-colon-type-list-kwds', which doesn't necessary
5406 ;; apply directly after the keyword, the type list is moved over
5407 ;; only when there is no unaccounted token before it (i.e. a token
5408 ;; that isn't moved over due to some other keyword list). The
5409 ;; identifier ranges in the list are still recorded if that should
5410 ;; be done, though.
5411 ;;
5412 ;; This function might do hidden buffer changes.
5413
5414 (let ((kwd-sym (c-keyword-sym (match-string match))) safe-pos pos
5415 ;; The call to `c-forward-<>-arglist' below is made after
5416 ;; `c-<>-sexp-kwds' keywords, so we're certain they actually
5417 ;; are angle bracket arglists and `c-restricted-<>-arglists'
5418 ;; should therefore be nil.
5419 (c-parse-and-markup-<>-arglists t)
5420 c-restricted-<>-arglists)
d9e94c22 5421
d9e94c22 5422 (when kwd-sym
0386b551 5423 (goto-char (match-end match))
d9e94c22
MS
5424 (c-forward-syntactic-ws)
5425 (setq safe-pos (point))
a66cd3ee 5426
d9e94c22
MS
5427 (cond
5428 ((and (c-keyword-member kwd-sym 'c-type-list-kwds)
5429 (c-forward-keyword-prefixed-id type))
5430 ;; There's a type directly after a keyword in `c-type-list-kwds'.
0386b551 5431 (c-forward-id-comma-list type t))
d9e94c22
MS
5432
5433 ((and (c-keyword-member kwd-sym 'c-ref-list-kwds)
5434 (c-forward-keyword-prefixed-id ref))
5435 ;; There's a name directly after a keyword in `c-ref-list-kwds'.
0386b551 5436 (c-forward-id-comma-list ref t))
d9e94c22
MS
5437
5438 ((and (c-keyword-member kwd-sym 'c-paren-any-kwds)
5439 (eq (char-after) ?\())
5440 ;; There's an open paren after a keyword in `c-paren-any-kwds'.
5441
5442 (forward-char)
5443 (when (and (setq pos (c-up-list-forward))
5444 (eq (char-before pos) ?\)))
5445 (when (and c-record-type-identifiers
5446 (c-keyword-member kwd-sym 'c-paren-type-kwds))
5447 ;; Use `c-forward-type' on every identifier we can find
5448 ;; inside the paren, to record the types.
5449 (while (c-syntactic-re-search-forward c-symbol-start pos t)
5450 (goto-char (match-beginning 0))
5451 (unless (c-forward-type)
5452 (looking-at c-symbol-key) ; Always matches.
5453 (goto-char (match-end 0)))))
5454
5455 (goto-char pos)
5456 (c-forward-syntactic-ws)
5457 (setq safe-pos (point))))
5458
5459 ((and (c-keyword-member kwd-sym 'c-<>-sexp-kwds)
5460 (eq (char-after) ?<)
0386b551 5461 (c-forward-<>-arglist (c-keyword-member kwd-sym 'c-<>-type-kwds)))
d9e94c22
MS
5462 (c-forward-syntactic-ws)
5463 (setq safe-pos (point)))
5464
5465 ((and (c-keyword-member kwd-sym 'c-nonsymbol-sexp-kwds)
449a2b0d
MS
5466 (not (looking-at c-symbol-start))
5467 (c-safe (c-forward-sexp) t))
d9e94c22
MS
5468 (c-forward-syntactic-ws)
5469 (setq safe-pos (point))))
5470
0386b551
AM
5471 (when (c-keyword-member kwd-sym 'c-colon-type-list-kwds)
5472 (if (eq (char-after) ?:)
5473 ;; If we are at the colon already, we move over the type
5474 ;; list after it.
5475 (progn
5476 (forward-char)
5477 (c-forward-syntactic-ws)
5478 (when (c-forward-keyword-prefixed-id type)
5479 (c-forward-id-comma-list type t)))
5480 ;; Not at the colon, so stop here. But the identifier
5481 ;; ranges in the type list later on should still be
5482 ;; recorded.
5483 (and c-record-type-identifiers
5484 (progn
5485 ;; If a keyword matched both one of the types above and
5486 ;; this one, we match `c-colon-type-list-re' after the
5487 ;; clause matched above.
5488 (goto-char safe-pos)
5489 (looking-at c-colon-type-list-re))
5490 (progn
5491 (goto-char (match-end 0))
5492 (c-forward-syntactic-ws)
5493 (c-forward-keyword-prefixed-id type))
5494 ;; There's a type after the `c-colon-type-list-re' match
5495 ;; after a keyword in `c-colon-type-list-kwds'.
5496 (c-forward-id-comma-list type nil))))
d9e94c22
MS
5497
5498 (goto-char safe-pos)
5499 t)))
5500
f3b554af
GM
5501;; cc-mode requires cc-fonts.
5502(declare-function c-fontify-recorded-types-and-refs "cc-fonts" ())
5503
0386b551
AM
5504(defun c-forward-<>-arglist (all-types)
5505 ;; The point is assumed to be at a "<". Try to treat it as the open
b4dc7d98 5506 ;; paren of an angle bracket arglist and move forward to the
0386b551
AM
5507 ;; corresponding ">". If successful, the point is left after the
5508 ;; ">" and t is returned, otherwise the point isn't moved and nil is
d9e94c22
MS
5509 ;; returned. If ALL-TYPES is t then all encountered arguments in
5510 ;; the arglist that might be types are treated as found types.
5511 ;;
0386b551
AM
5512 ;; The variable `c-parse-and-markup-<>-arglists' controls how this
5513 ;; function handles text properties on the angle brackets and argument
5514 ;; separating commas.
d9e94c22 5515 ;;
0386b551
AM
5516 ;; `c-restricted-<>-arglists' controls how lenient the template
5517 ;; arglist recognition should be.
5518 ;;
5519 ;; This function records identifier ranges on
5520 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
5521 ;; `c-record-type-identifiers' is non-nil.
5522 ;;
5523 ;; This function might do hidden buffer changes.
d9e94c22
MS
5524
5525 (let ((start (point))
5526 ;; If `c-record-type-identifiers' is set then activate
5527 ;; recording of any found types that constitute an argument in
5528 ;; the arglist.
abfc152b 5529 (c-record-found-types (if c-record-type-identifiers t)))
d9e94c22
MS
5530 (if (catch 'angle-bracket-arglist-escape
5531 (setq c-record-found-types
0386b551 5532 (c-forward-<>-arglist-recur all-types)))
d9e94c22
MS
5533 (progn
5534 (when (consp c-record-found-types)
5535 (setq c-record-type-identifiers
5536 ;; `nconc' doesn't mind that the tail of
5537 ;; `c-record-found-types' is t.
5538 (nconc c-record-found-types c-record-type-identifiers)))
452ea855 5539 (if (c-major-mode-is 'java-mode) (c-fontify-recorded-types-and-refs))
d9e94c22
MS
5540 t)
5541
5542 (goto-char start)
a66cd3ee 5543 nil)))
785eecbb 5544
0386b551 5545(defun c-forward-<>-arglist-recur (all-types)
d9e94c22 5546 ;; Recursive part of `c-forward-<>-arglist'.
0386b551
AM
5547 ;;
5548 ;; This function might do hidden buffer changes.
d9e94c22
MS
5549
5550 (let ((start (point)) res pos tmp
5551 ;; Cover this so that any recorded found type ranges are
5552 ;; automatically lost if it turns out to not be an angle
5553 ;; bracket arglist. It's propagated through the return value
5554 ;; on successful completion.
5555 (c-record-found-types c-record-found-types)
5556 ;; List that collects the positions after the argument
5557 ;; separating ',' in the arglist.
5558 arg-start-pos)
0386b551
AM
5559 ;; If the '<' has paren open syntax then we've marked it as an angle
5560 ;; bracket arglist before, so skip to the end.
5561 (if (and (not c-parse-and-markup-<>-arglists)
5562 (c-get-char-property (point) 'syntax-table))
5563
5564 (progn
5565 (forward-char)
5566 (if (and (c-go-up-list-forward)
5567 (eq (char-before) ?>))
5568 t
0386b551
AM
5569 ;; Got unmatched paren angle brackets. We don't clear the paren
5570 ;; syntax properties and retry, on the basis that it's very
5571 ;; unlikely that paren angle brackets become operators by code
5572 ;; manipulation. It's far more likely that it doesn't match due
5573 ;; to narrowing or some temporary change.
5574 (goto-char start)
5575 nil))
d9e94c22 5576
abfc152b 5577 (forward-char) ; Forward over the opening '<'.
452ea855 5578
d9e94c22 5579 (unless (looking-at c-<-op-cont-regexp)
abfc152b
AM
5580 ;; go forward one non-alphanumeric character (group) per iteration of
5581 ;; this loop.
22c3ce97 5582 (while (and
d9e94c22 5583 (progn
22c3ce97
AM
5584 (c-forward-syntactic-ws)
5585 (let ((orig-record-found-types c-record-found-types))
5586 (when (or (and c-record-type-identifiers all-types)
5587 (c-major-mode-is 'java-mode))
5588 ;; All encountered identifiers are types, so set the
5589 ;; promote flag and parse the type.
5590 (progn
5591 (c-forward-syntactic-ws)
5592 (if (looking-at "\\?")
5593 (forward-char)
5594 (when (looking-at c-identifier-start)
5595 (let ((c-promote-possible-types t)
5596 (c-record-found-types t))
5597 (c-forward-type))))
5598
5599 (c-forward-syntactic-ws)
5600
5601 (when (or (looking-at "extends")
5602 (looking-at "super"))
5603 (forward-word)
5604 (c-forward-syntactic-ws)
5605 (let ((c-promote-possible-types t)
5606 (c-record-found-types t))
5607 (c-forward-type)
5608 (c-forward-syntactic-ws))))))
5609
abfc152b 5610 (setq pos (point)) ; e.g. first token inside the '<'
22c3ce97 5611
a4ee83cc
AM
5612 ;; Note: These regexps exploit the match order in \| so
5613 ;; that "<>" is matched by "<" rather than "[^>:-]>".
5614 (c-syntactic-re-search-forward
5615 ;; Stop on ',', '|', '&', '+' and '-' to catch
5616 ;; common binary operators that could be between
5617 ;; two comparison expressions "a<b" and "c>d".
5618 "[<;{},|+&-]\\|[>)]"
5619 nil t t))
22c3ce97
AM
5620
5621 (cond
5622 ((eq (char-before) ?>)
d9e94c22
MS
5623 ;; Either an operator starting with '>' or the end of
5624 ;; the angle bracket arglist.
5625
0386b551 5626 (if (looking-at c->-op-cont-regexp)
d9e94c22 5627 (progn
0386b551
AM
5628 (goto-char (match-end 0))
5629 t) ; Continue the loop.
d9e94c22 5630
0386b551
AM
5631 ;; The angle bracket arglist is finished.
5632 (when c-parse-and-markup-<>-arglists
d9e94c22 5633 (while arg-start-pos
0386b551
AM
5634 (c-put-c-type-property (1- (car arg-start-pos))
5635 'c-<>-arg-sep)
d9e94c22
MS
5636 (setq arg-start-pos (cdr arg-start-pos)))
5637 (c-mark-<-as-paren start)
0386b551
AM
5638 (c-mark->-as-paren (1- (point))))
5639 (setq res t)
5640 nil)) ; Exit the loop.
d9e94c22
MS
5641
5642 ((eq (char-before) ?<)
5643 ;; Either an operator starting with '<' or a nested arglist.
d9e94c22
MS
5644 (setq pos (point))
5645 (let (id-start id-end subres keyword-match)
abfc152b
AM
5646 (cond
5647 ;; The '<' begins a multi-char operator.
5648 ((looking-at c-<-op-cont-regexp)
5649 (setq tmp (match-end 0))
5650 (goto-char (match-end 0)))
5651 ;; We're at a nested <.....>
5652 ((progn
5653 (setq tmp pos)
5654 (backward-char) ; to the '<'
5655 (and
5656 (save-excursion
5657 ;; There's always an identifier before an angle
5658 ;; bracket arglist, or a keyword in `c-<>-type-kwds'
5659 ;; or `c-<>-arglist-kwds'.
5660 (c-backward-syntactic-ws)
5661 (setq id-end (point))
5662 (c-simple-skip-symbol-backward)
5663 (when (or (setq keyword-match
5664 (looking-at c-opt-<>-sexp-key))
5665 (not (looking-at c-keywords-regexp)))
5666 (setq id-start (point))))
5667 (setq subres
5668 (let ((c-promote-possible-types t)
5669 (c-record-found-types t))
5670 (c-forward-<>-arglist-recur
5671 (and keyword-match
5672 (c-keyword-member
5673 (c-keyword-sym (match-string 1))
5674 'c-<>-type-kwds)))))))
d9e94c22
MS
5675
5676 ;; It was an angle bracket arglist.
5677 (setq c-record-found-types subres)
5678
5679 ;; Record the identifier before the template as a type
5680 ;; or reference depending on whether the arglist is last
5681 ;; in a qualified identifier.
5682 (when (and c-record-type-identifiers
5683 (not keyword-match))
5684 (if (and c-opt-identifier-concat-key
5685 (progn
5686 (c-forward-syntactic-ws)
5687 (looking-at c-opt-identifier-concat-key)))
5688 (c-record-ref-id (cons id-start id-end))
abfc152b
AM
5689 (c-record-type-id (cons id-start id-end)))))
5690
5691 ;; At a "less than" operator.
5692 (t
5693 (forward-char)
5694 )))
5695 t) ; carry on looping.
22c3ce97
AM
5696
5697 ((and (not c-restricted-<>-arglists)
5698 (or (and (eq (char-before) ?&)
5699 (not (eq (char-after) ?&)))
5700 (eq (char-before) ?,)))
5701 ;; Just another argument. Record the position. The
5702 ;; type check stuff that made us stop at it is at
5703 ;; the top of the loop.
d9e94c22
MS
5704 (setq arg-start-pos (cons (point) arg-start-pos)))
5705
5706 (t
5707 ;; Got a character that can't be in an angle bracket
5708 ;; arglist argument. Abort using `throw', since
5709 ;; it's useless to try to find a surrounding arglist
5710 ;; if we're nested.
5711 (throw 'angle-bracket-arglist-escape nil))))))
d9e94c22
MS
5712 (if res
5713 (or c-record-found-types t)))))
5714
0386b551
AM
5715(defun c-backward-<>-arglist (all-types &optional limit)
5716 ;; The point is assumed to be directly after a ">". Try to treat it
5717 ;; as the close paren of an angle bracket arglist and move back to
5718 ;; the corresponding "<". If successful, the point is left at
5719 ;; the "<" and t is returned, otherwise the point isn't moved and
5720 ;; nil is returned. ALL-TYPES is passed on to
5721 ;; `c-forward-<>-arglist'.
5722 ;;
5723 ;; If the optional LIMIT is given, it bounds the backward search.
5724 ;; It's then assumed to be at a syntactically relevant position.
5725 ;;
5726 ;; This is a wrapper around `c-forward-<>-arglist'. See that
5727 ;; function for more details.
5728
5729 (let ((start (point)))
5730 (backward-char)
5731 (if (and (not c-parse-and-markup-<>-arglists)
5732 (c-get-char-property (point) 'syntax-table))
5733
5734 (if (and (c-go-up-list-backward)
5735 (eq (char-after) ?<))
5736 t
5737 ;; See corresponding note in `c-forward-<>-arglist'.
5738 (goto-char start)
5739 nil)
5740
51c9af45 5741 (while (progn
0386b551
AM
5742 (c-syntactic-skip-backward "^<;{}" limit t)
5743
51c9af45
AM
5744 (and
5745 (if (eq (char-before) ?<)
5746 t
5747 ;; Stopped at bob or a char that isn't allowed in an
5748 ;; arglist, so we've failed.
5749 (goto-char start)
5750 nil)
0386b551 5751
51c9af45
AM
5752 (if (> (point)
5753 (progn (c-beginning-of-current-token)
5754 (point)))
5755 ;; If we moved then the "<" was part of some
5756 ;; multicharacter token.
5757 t
0386b551 5758
51c9af45
AM
5759 (backward-char)
5760 (let ((beg-pos (point)))
5761 (if (c-forward-<>-arglist all-types)
5762 (cond ((= (point) start)
5763 ;; Matched the arglist. Break the while.
5764 (goto-char beg-pos)
5765 nil)
5766 ((> (point) start)
5767 ;; We started from a non-paren ">" inside an
5768 ;; arglist.
5769 (goto-char start)
5770 nil)
5771 (t
5772 ;; Matched a shorter arglist. Can be a nested
5773 ;; one so continue looking.
5774 (goto-char beg-pos)
5775 t))
5776 t))))))
0386b551
AM
5777
5778 (/= (point) start))))
5779
d9e94c22
MS
5780(defun c-forward-name ()
5781 ;; Move forward over a complete name if at the beginning of one,
e15f8aaa
AM
5782 ;; stopping at the next following token. A keyword, as such,
5783 ;; doesn't count as a name. If the point is not at something that
5784 ;; is recognized as a name then it stays put.
5785 ;;
5786 ;; A name could be something as simple as "foo" in C or something as
d9e94c22
MS
5787 ;; complex as "X<Y<class A<int>::B, BIT_MAX >> b>, ::operator<> ::
5788 ;; Z<(a>b)> :: operator const X<&foo>::T Q::G<unsigned short
5789 ;; int>::*volatile const" in C++ (this function is actually little
5790 ;; more than a `looking-at' call in all modes except those that,
e15f8aaa
AM
5791 ;; like C++, have `c-recognize-<>-arglists' set).
5792 ;;
5793 ;; Return
5794 ;; o - nil if no name is found;
5795 ;; o - 'template if it's an identifier ending with an angle bracket
5796 ;; arglist;
5797 ;; o - 'operator of it's an operator identifier;
5798 ;; o - t if it's some other kind of name.
0386b551
AM
5799 ;;
5800 ;; This function records identifier ranges on
5801 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
5802 ;; `c-record-type-identifiers' is non-nil.
5803 ;;
5804 ;; This function might do hidden buffer changes.
d9e94c22 5805
0386b551 5806 (let ((pos (point)) (start (point)) res id-start id-end
d9e94c22
MS
5807 ;; Turn off `c-promote-possible-types' here since we might
5808 ;; call `c-forward-<>-arglist' and we don't want it to promote
5809 ;; every suspect thing in the arglist to a type. We're
5810 ;; typically called from `c-forward-type' in this case, and
5811 ;; the caller only wants the top level type that it finds to
5812 ;; be promoted.
5813 c-promote-possible-types)
5814 (while
5815 (and
5816 (looking-at c-identifier-key)
5817
5818 (progn
5819 ;; Check for keyword. We go to the last symbol in
5820 ;; `c-identifier-key' first.
0386b551
AM
5821 (goto-char (setq id-end (match-end 0)))
5822 (c-simple-skip-symbol-backward)
5823 (setq id-start (point))
d9e94c22
MS
5824
5825 (if (looking-at c-keywords-regexp)
5826 (when (and (c-major-mode-is 'c++-mode)
5827 (looking-at
5828 (cc-eval-when-compile
5829 (concat "\\(operator\\|\\(template\\)\\)"
5830 "\\(" (c-lang-const c-nonsymbol-key c++)
5831 "\\|$\\)")))
5832 (if (match-beginning 2)
5833 ;; "template" is only valid inside an
5834 ;; identifier if preceded by "::".
5835 (save-excursion
5836 (c-backward-syntactic-ws)
5837 (and (c-safe (backward-char 2) t)
5838 (looking-at "::")))
5839 t))
5840
5841 ;; Handle a C++ operator or template identifier.
5842 (goto-char id-end)
5843 (c-forward-syntactic-ws)
5844 (cond ((eq (char-before id-end) ?e)
5845 ;; Got "... ::template".
5846 (let ((subres (c-forward-name)))
5847 (when subres
5848 (setq pos (point)
5849 res subres))))
5850
5851 ((looking-at c-identifier-start)
5852 ;; Got a cast operator.
5853 (when (c-forward-type)
5854 (setq pos (point)
5855 res 'operator)
5856 ;; Now we should match a sequence of either
5857 ;; '*', '&' or a name followed by ":: *",
5858 ;; where each can be followed by a sequence
5859 ;; of `c-opt-type-modifier-key'.
5860 (while (cond ((looking-at "[*&]")
5861 (goto-char (match-end 0))
5862 t)
5863 ((looking-at c-identifier-start)
5864 (and (c-forward-name)
5865 (looking-at "::")
5866 (progn
5867 (goto-char (match-end 0))
5868 (c-forward-syntactic-ws)
5869 (eq (char-after) ?*))
5870 (progn
5871 (forward-char)
5872 t))))
5873 (while (progn
5874 (c-forward-syntactic-ws)
5875 (setq pos (point))
5876 (looking-at c-opt-type-modifier-key))
5877 (goto-char (match-end 1))))))
5878
5879 ((looking-at c-overloadable-operators-regexp)
5880 ;; Got some other operator.
0386b551
AM
5881 (setq c-last-identifier-range
5882 (cons (point) (match-end 0)))
d9e94c22
MS
5883 (goto-char (match-end 0))
5884 (c-forward-syntactic-ws)
5885 (setq pos (point)
5886 res 'operator)))
5887
5888 nil)
5889
0386b551
AM
5890 ;; `id-start' is equal to `id-end' if we've jumped over
5891 ;; an identifier that doesn't end with a symbol token.
5892 ;; That can occur e.g. for Java import directives on the
5893 ;; form "foo.bar.*".
5894 (when (and id-start (/= id-start id-end))
d9e94c22
MS
5895 (setq c-last-identifier-range
5896 (cons id-start id-end)))
5897 (goto-char id-end)
5898 (c-forward-syntactic-ws)
5899 (setq pos (point)
5900 res t)))
5901
5902 (progn
5903 (goto-char pos)
5904 (when (or c-opt-identifier-concat-key
5905 c-recognize-<>-arglists)
5906
5907 (cond
5908 ((and c-opt-identifier-concat-key
5909 (looking-at c-opt-identifier-concat-key))
5910 ;; Got a concatenated identifier. This handles the
5911 ;; cases with tricky syntactic whitespace that aren't
5912 ;; covered in `c-identifier-key'.
5913 (goto-char (match-end 0))
5914 (c-forward-syntactic-ws)
5915 t)
5916
5917 ((and c-recognize-<>-arglists
5918 (eq (char-after) ?<))
5919 ;; Maybe an angle bracket arglist.
452ea855
AM
5920 (when (let ((c-record-type-identifiers t)
5921 (c-record-found-types t))
0386b551
AM
5922 (c-forward-<>-arglist nil))
5923
5924 (c-add-type start (1+ pos))
d9e94c22 5925 (c-forward-syntactic-ws)
0386b551
AM
5926 (setq pos (point)
5927 c-last-identifier-range nil)
5928
d9e94c22
MS
5929 (if (and c-opt-identifier-concat-key
5930 (looking-at c-opt-identifier-concat-key))
0386b551 5931
d9e94c22
MS
5932 ;; Continue if there's an identifier concatenation
5933 ;; operator after the template argument.
5934 (progn
0386b551
AM
5935 (when (and c-record-type-identifiers id-start)
5936 (c-record-ref-id (cons id-start id-end)))
d9e94c22
MS
5937 (forward-char 2)
5938 (c-forward-syntactic-ws)
5939 t)
0386b551
AM
5940
5941 (when (and c-record-type-identifiers id-start)
5942 (c-record-type-id (cons id-start id-end)))
d9e94c22
MS
5943 (setq res 'template)
5944 nil)))
5945 )))))
5946
5947 (goto-char pos)
5948 res))
5949
e15f8aaa 5950(defun c-forward-type (&optional brace-block-too)
d9e94c22 5951 ;; Move forward over a type spec if at the beginning of one,
e15f8aaa
AM
5952 ;; stopping at the next following token. The keyword "typedef"
5953 ;; isn't part of a type spec here.
5954 ;;
5955 ;; BRACE-BLOCK-TOO, when non-nil, means move over the brace block in
5956 ;; constructs like "struct foo {...} bar ;" or "struct {...} bar;".
5957 ;; The current (2009-03-10) intention is to convert all uses of
5958 ;; `c-forward-type' to call with this parameter set, then to
5959 ;; eliminate it.
5960 ;;
5961 ;; Return
5962 ;; o - t if it's a known type that can't be a name or other
5963 ;; expression;
5964 ;; o - 'known if it's an otherwise known type (according to
5965 ;; `*-font-lock-extra-types');
5966 ;; o - 'prefix if it's a known prefix of a type;
5967 ;; o - 'found if it's a type that matches one in `c-found-types';
53964682 5968 ;; o - 'maybe if it's an identifier that might be a type; or
e15f8aaa
AM
5969 ;; o - nil if it can't be a type (the point isn't moved then).
5970 ;;
5971 ;; The point is assumed to be at the beginning of a token.
d9e94c22
MS
5972 ;;
5973 ;; Note that this function doesn't skip past the brace definition
5974 ;; that might be considered part of the type, e.g.
5975 ;; "enum {a, b, c} foo".
0386b551
AM
5976 ;;
5977 ;; This function records identifier ranges on
5978 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
5979 ;; `c-record-type-identifiers' is non-nil.
5980 ;;
5981 ;; This function might do hidden buffer changes.
0e5cf2b8
AM
5982 (when (and c-recognize-<>-arglists
5983 (looking-at "<"))
452ea855
AM
5984 (c-forward-<>-arglist t)
5985 (c-forward-syntactic-ws))
0386b551
AM
5986
5987 (let ((start (point)) pos res name-res id-start id-end id-range)
d9e94c22
MS
5988
5989 ;; Skip leading type modifiers. If any are found we know it's a
5990 ;; prefix of a type.
e15f8aaa 5991 (when c-opt-type-modifier-key ; e.g. "const" "volatile", but NOT "typedef"
d9e94c22
MS
5992 (while (looking-at c-opt-type-modifier-key)
5993 (goto-char (match-end 1))
5994 (c-forward-syntactic-ws)
5995 (setq res 'prefix)))
5996
5997 (cond
e15f8aaa
AM
5998 ((looking-at c-type-prefix-key) ; e.g. "struct", "class", but NOT
5999 ; "typedef".
d9e94c22
MS
6000 (goto-char (match-end 1))
6001 (c-forward-syntactic-ws)
6002 (setq pos (point))
e15f8aaa
AM
6003
6004 (setq name-res (c-forward-name))
6005 (setq res (not (null name-res)))
6006 (when (eq name-res t)
6007 ;; In many languages the name can be used without the
6008 ;; prefix, so we add it to `c-found-types'.
6009 (c-add-type pos (point))
6010 (when (and c-record-type-identifiers
6011 c-last-identifier-range)
6012 (c-record-type-id c-last-identifier-range)))
6013 (when (and brace-block-too
6014 (memq res '(t nil))
6015 (eq (char-after) ?\{)
6016 (save-excursion
6017 (c-safe
6018 (progn (c-forward-sexp)
6019 (c-forward-syntactic-ws)
6020 (setq pos (point))))))
6021 (goto-char pos)
6022 (setq res t))
6023 (unless res (goto-char start))) ; invalid syntax
d9e94c22
MS
6024
6025 ((progn
6026 (setq pos nil)
6027 (if (looking-at c-identifier-start)
6028 (save-excursion
6029 (setq id-start (point)
0386b551
AM
6030 name-res (c-forward-name))
6031 (when name-res
d9e94c22
MS
6032 (setq id-end (point)
6033 id-range c-last-identifier-range))))
6034 (and (cond ((looking-at c-primitive-type-key)
6035 (setq res t))
6036 ((c-with-syntax-table c-identifier-syntax-table
6037 (looking-at c-known-type-key))
6038 (setq res 'known)))
6039 (or (not id-end)
6040 (>= (save-excursion
6041 (save-match-data
6042 (goto-char (match-end 1))
6043 (c-forward-syntactic-ws)
6044 (setq pos (point))))
6045 id-end)
6046 (setq res nil))))
6047 ;; Looking at a primitive or known type identifier. We've
6048 ;; checked for a name first so that we don't go here if the
6049 ;; known type match only is a prefix of another name.
6050
6051 (setq id-end (match-end 1))
6052
6053 (when (and c-record-type-identifiers
6054 (or c-promote-possible-types (eq res t)))
6055 (c-record-type-id (cons (match-beginning 1) (match-end 1))))
6056
6057 (if (and c-opt-type-component-key
6058 (save-match-data
6059 (looking-at c-opt-type-component-key)))
6060 ;; There might be more keywords for the type.
6061 (let (safe-pos)
0386b551 6062 (c-forward-keyword-clause 1)
d9e94c22
MS
6063 (while (progn
6064 (setq safe-pos (point))
6065 (looking-at c-opt-type-component-key))
6066 (when (and c-record-type-identifiers
6067 (looking-at c-primitive-type-key))
6068 (c-record-type-id (cons (match-beginning 1)
6069 (match-end 1))))
0386b551 6070 (c-forward-keyword-clause 1))
d9e94c22
MS
6071 (if (looking-at c-primitive-type-key)
6072 (progn
6073 (when c-record-type-identifiers
6074 (c-record-type-id (cons (match-beginning 1)
6075 (match-end 1))))
0386b551 6076 (c-forward-keyword-clause 1)
d9e94c22
MS
6077 (setq res t))
6078 (goto-char safe-pos)
6079 (setq res 'prefix)))
0386b551 6080 (unless (save-match-data (c-forward-keyword-clause 1))
d9e94c22
MS
6081 (if pos
6082 (goto-char pos)
6083 (goto-char (match-end 1))
6084 (c-forward-syntactic-ws)))))
6085
0386b551
AM
6086 (name-res
6087 (cond ((eq name-res t)
d9e94c22
MS
6088 ;; A normal identifier.
6089 (goto-char id-end)
6090 (if (or res c-promote-possible-types)
6091 (progn
6092 (c-add-type id-start id-end)
0386b551 6093 (when (and c-record-type-identifiers id-range)
d9e94c22
MS
6094 (c-record-type-id id-range))
6095 (unless res
6096 (setq res 'found)))
6097 (setq res (if (c-check-type id-start id-end)
6098 ;; It's an identifier that has been used as
6099 ;; a type somewhere else.
6100 'found
6101 ;; It's an identifier that might be a type.
6102 'maybe))))
0386b551 6103 ((eq name-res 'template)
d9e94c22
MS
6104 ;; A template is a type.
6105 (goto-char id-end)
6106 (setq res t))
6107 (t
6108 ;; Otherwise it's an operator identifier, which is not a type.
6109 (goto-char start)
6110 (setq res nil)))))
6111
6112 (when res
e15f8aaa 6113 ;; Skip trailing type modifiers. If any are found we know it's
d9e94c22
MS
6114 ;; a type.
6115 (when c-opt-type-modifier-key
e15f8aaa 6116 (while (looking-at c-opt-type-modifier-key) ; e.g. "const", "volatile"
d9e94c22
MS
6117 (goto-char (match-end 1))
6118 (c-forward-syntactic-ws)
6119 (setq res t)))
d9e94c22
MS
6120 ;; Step over any type suffix operator. Do not let the existence
6121 ;; of these alter the classification of the found type, since
6122 ;; these operators typically are allowed in normal expressions
6123 ;; too.
6124 (when c-opt-type-suffix-key
6125 (while (looking-at c-opt-type-suffix-key)
6126 (goto-char (match-end 1))
6127 (c-forward-syntactic-ws)))
6128
e15f8aaa 6129 (when c-opt-type-concat-key ; Only/mainly for pike.
0386b551
AM
6130 ;; Look for a trailing operator that concatenates the type
6131 ;; with a following one, and if so step past that one through
6132 ;; a recursive call. Note that we don't record concatenated
6133 ;; types in `c-found-types' - it's the component types that
6134 ;; are recorded when appropriate.
d9e94c22
MS
6135 (setq pos (point))
6136 (let* ((c-promote-possible-types (or (memq res '(t known))
6137 c-promote-possible-types))
6138 ;; If we can't promote then set `c-record-found-types' so that
6139 ;; we can merge in the types from the second part afterwards if
6140 ;; it turns out to be a known type there.
6141 (c-record-found-types (and c-record-type-identifiers
0386b551
AM
6142 (not c-promote-possible-types)))
6143 subres)
d9e94c22
MS
6144 (if (and (looking-at c-opt-type-concat-key)
6145
6146 (progn
6147 (goto-char (match-end 1))
6148 (c-forward-syntactic-ws)
0386b551 6149 (setq subres (c-forward-type))))
d9e94c22
MS
6150
6151 (progn
6152 ;; If either operand certainly is a type then both are, but we
6153 ;; don't let the existence of the operator itself promote two
6154 ;; uncertain types to a certain one.
6155 (cond ((eq res t))
0386b551
AM
6156 ((eq subres t)
6157 (unless (eq name-res 'template)
6158 (c-add-type id-start id-end))
6159 (when (and c-record-type-identifiers id-range)
d9e94c22
MS
6160 (c-record-type-id id-range))
6161 (setq res t))
2a15eb73 6162 ((eq res 'known))
0386b551 6163 ((eq subres 'known)
2a15eb73 6164 (setq res 'known))
d9e94c22 6165 ((eq res 'found))
0386b551 6166 ((eq subres 'found)
d9e94c22
MS
6167 (setq res 'found))
6168 (t
6169 (setq res 'maybe)))
6170
6171 (when (and (eq res t)
6172 (consp c-record-found-types))
6173 ;; Merge in the ranges of any types found by the second
6174 ;; `c-forward-type'.
6175 (setq c-record-type-identifiers
6176 ;; `nconc' doesn't mind that the tail of
6177 ;; `c-record-found-types' is t.
6178 (nconc c-record-found-types
6179 c-record-type-identifiers))))
6180
6181 (goto-char pos))))
6182
6183 (when (and c-record-found-types (memq res '(known found)) id-range)
6184 (setq c-record-found-types
6185 (cons id-range c-record-found-types))))
6186
6187 ;;(message "c-forward-type %s -> %s: %s" start (point) res)
6188
6189 res))
6190
452ea855
AM
6191(defun c-forward-annotation ()
6192 ;; Used for Java code only at the moment. Assumes point is on the
6193 ;; @, moves forward an annotation. returns nil if there is no
6194 ;; annotation at point.
6195 (and (looking-at "@")
6196 (progn (forward-char) t)
6197 (c-forward-type)
6198 (progn (c-forward-syntactic-ws) t)
6199 (if (looking-at "(")
6200 (c-go-list-forward)
6201 t)))
6202
785eecbb 6203\f
d9e94c22
MS
6204;; Handling of large scale constructs like statements and declarations.
6205
0386b551
AM
6206;; Macro used inside `c-forward-decl-or-cast-1'. It ought to be a
6207;; defsubst or perhaps even a defun, but it contains lots of free
6208;; variables that refer to things inside `c-forward-decl-or-cast-1'.
6209(defmacro c-fdoc-shift-type-backward (&optional short)
6210 ;; `c-forward-decl-or-cast-1' can consume an arbitrary length list
6211 ;; of types when parsing a declaration, which means that it
6212 ;; sometimes consumes the identifier in the declaration as a type.
6213 ;; This is used to "backtrack" and make the last type be treated as
6214 ;; an identifier instead.
6215 `(progn
6216 ,(unless short
6217 ;; These identifiers are bound only in the inner let.
6218 '(setq identifier-type at-type
6219 identifier-start type-start
6220 got-parens nil
6221 got-identifier t
6222 got-suffix t
6223 got-suffix-after-parens id-start
6224 paren-depth 0))
6225
6226 (if (setq at-type (if (eq backup-at-type 'prefix)
6227 t
6228 backup-at-type))
6229 (setq type-start backup-type-start
6230 id-start backup-id-start)
6231 (setq type-start start-pos
6232 id-start start-pos))
6233
6234 ;; When these flags already are set we've found specifiers that
6235 ;; unconditionally signal these attributes - backtracking doesn't
6236 ;; change that. So keep them set in that case.
6237 (or at-type-decl
6238 (setq at-type-decl backup-at-type-decl))
6239 (or maybe-typeless
6240 (setq maybe-typeless backup-maybe-typeless))
6241
6242 ,(unless short
6243 ;; This identifier is bound only in the inner let.
6244 '(setq start id-start))))
6245
6246(defun c-forward-decl-or-cast-1 (preceding-token-end context last-cast-end)
6247 ;; Move forward over a declaration or a cast if at the start of one.
6248 ;; The point is assumed to be at the start of some token. Nil is
6249 ;; returned if no declaration or cast is recognized, and the point
6250 ;; is clobbered in that case.
6251 ;;
6252 ;; If a declaration is parsed:
6253 ;;
6254 ;; The point is left at the first token after the first complete
6255 ;; declarator, if there is one. The return value is a cons where
022d0cf4
AM
6256 ;; the car is the position of the first token in the declarator. (See
6257 ;; below for the cdr.)
0386b551
AM
6258 ;; Some examples:
6259 ;;
6260 ;; void foo (int a, char *b) stuff ...
6261 ;; car ^ ^ point
6262 ;; float (*a)[], b;
6263 ;; car ^ ^ point
6264 ;; unsigned int a = c_style_initializer, b;
6265 ;; car ^ ^ point
6266 ;; unsigned int a (cplusplus_style_initializer), b;
6267 ;; car ^ ^ point (might change)
6268 ;; class Foo : public Bar {}
6269 ;; car ^ ^ point
6270 ;; class PikeClass (int a, string b) stuff ...
6271 ;; car ^ ^ point
6272 ;; enum bool;
6273 ;; car ^ ^ point
6274 ;; enum bool flag;
6275 ;; car ^ ^ point
6276 ;; void cplusplus_function (int x) throw (Bad);
6277 ;; car ^ ^ point
6278 ;; Foo::Foo (int b) : Base (b) {}
6279 ;; car ^ ^ point
91af3942 6280 ;;
e15f8aaa
AM
6281 ;; The cdr of the return value is non-nil when a
6282 ;; `c-typedef-decl-kwds' specifier is found in the declaration.
6283 ;; Specifically it is a dotted pair (A . B) where B is t when a
6284 ;; `c-typedef-kwds' ("typedef") is present, and A is t when some
6285 ;; other `c-typedef-decl-kwds' (e.g. class, struct, enum)
6286 ;; specifier is present. I.e., (some of) the declared
6287 ;; identifier(s) are types.
91af3942 6288 ;;
0386b551
AM
6289 ;; If a cast is parsed:
6290 ;;
6291 ;; The point is left at the first token after the closing paren of
6292 ;; the cast. The return value is `cast'. Note that the start
6293 ;; position must be at the first token inside the cast parenthesis
6294 ;; to recognize it.
6295 ;;
6296 ;; PRECEDING-TOKEN-END is the first position after the preceding
6297 ;; token, i.e. on the other side of the syntactic ws from the point.
6298 ;; Use a value less than or equal to (point-min) if the point is at
6299 ;; the first token in (the visible part of) the buffer.
6300 ;;
6301 ;; CONTEXT is a symbol that describes the context at the point:
022d0cf4 6302 ;; 'decl In a comma-separated declaration context (typically
0386b551
AM
6303 ;; inside a function declaration arglist).
6304 ;; '<> In an angle bracket arglist.
6305 ;; 'arglist Some other type of arglist.
a85fd6da
AM
6306 ;; nil Some other context or unknown context. Includes
6307 ;; within the parens of an if, for, ... construct.
0386b551
AM
6308 ;;
6309 ;; LAST-CAST-END is the first token after the closing paren of a
6310 ;; preceding cast, or nil if none is known. If
6311 ;; `c-forward-decl-or-cast-1' is used in succession, it should be
6312 ;; the position after the closest preceding call where a cast was
6313 ;; matched. In that case it's used to discover chains of casts like
6314 ;; "(a) (b) c".
6315 ;;
6316 ;; This function records identifier ranges on
6317 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
6318 ;; `c-record-type-identifiers' is non-nil.
6319 ;;
6320 ;; This function might do hidden buffer changes.
6321
6322 (let (;; `start-pos' is used below to point to the start of the
6323 ;; first type, i.e. after any leading specifiers. It might
6324 ;; also point at the beginning of the preceding syntactic
6325 ;; whitespace.
6326 (start-pos (point))
6327 ;; Set to the result of `c-forward-type'.
6328 at-type
6329 ;; The position of the first token in what we currently
6330 ;; believe is the type in the declaration or cast, after any
6331 ;; specifiers and their associated clauses.
6332 type-start
6333 ;; The position of the first token in what we currently
6334 ;; believe is the declarator for the first identifier. Set
6335 ;; when the type is found, and moved forward over any
6336 ;; `c-decl-hangon-kwds' and their associated clauses that
6337 ;; occurs after the type.
6338 id-start
6339 ;; These store `at-type', `type-start' and `id-start' of the
6340 ;; identifier before the one in those variables. The previous
6341 ;; identifier might turn out to be the real type in a
6342 ;; declaration if the last one has to be the declarator in it.
6343 ;; If `backup-at-type' is nil then the other variables have
6344 ;; undefined values.
6345 backup-at-type backup-type-start backup-id-start
e15f8aaa
AM
6346 ;; Set if we've found a specifier (apart from "typedef") that makes
6347 ;; the defined identifier(s) types.
0386b551 6348 at-type-decl
e15f8aaa
AM
6349 ;; Set if we've a "typedef" keyword.
6350 at-typedef
0386b551
AM
6351 ;; Set if we've found a specifier that can start a declaration
6352 ;; where there's no type.
6353 maybe-typeless
6354 ;; If a specifier is found that also can be a type prefix,
6355 ;; these flags are set instead of those above. If we need to
6356 ;; back up an identifier, they are copied to the real flag
6357 ;; variables. Thus they only take effect if we fail to
6358 ;; interpret it as a type.
6359 backup-at-type-decl backup-maybe-typeless
6360 ;; Whether we've found a declaration or a cast. We might know
6361 ;; this before we've found the type in it. It's 'ids if we've
6362 ;; found two consecutive identifiers (usually a sure sign, but
6363 ;; we should allow that in labels too), and t if we've found a
6364 ;; specifier keyword (a 100% sure sign).
6365 at-decl-or-cast
6366 ;; Set when we need to back up to parse this as a declaration
6367 ;; but not as a cast.
6368 backup-if-not-cast
6369 ;; For casts, the return position.
6370 cast-end
6371 ;; Save `c-record-type-identifiers' and
6372 ;; `c-record-ref-identifiers' since ranges are recorded
6373 ;; speculatively and should be thrown away if it turns out
6374 ;; that it isn't a declaration or cast.
6375 (save-rec-type-ids c-record-type-identifiers)
6376 (save-rec-ref-ids c-record-ref-identifiers))
6377
452ea855
AM
6378 (while (c-forward-annotation)
6379 (c-forward-syntactic-ws))
6380
0386b551
AM
6381 ;; Check for a type. Unknown symbols are treated as possible
6382 ;; types, but they could also be specifiers disguised through
6383 ;; macros like __INLINE__, so we recognize both types and known
6384 ;; specifiers after them too.
6385 (while
6386 (let* ((start (point)) kwd-sym kwd-clause-end found-type)
6387
6388 ;; Look for a specifier keyword clause.
e3715033
AM
6389 (when (or (looking-at c-prefix-spec-kwds-re)
6390 (and (c-major-mode-is 'java-mode)
6391 (looking-at "@[A-Za-z0-9]+")))
e15f8aaa
AM
6392 (if (looking-at c-typedef-key)
6393 (setq at-typedef t))
0386b551
AM
6394 (setq kwd-sym (c-keyword-sym (match-string 1)))
6395 (save-excursion
6396 (c-forward-keyword-clause 1)
6397 (setq kwd-clause-end (point))))
6398
e15f8aaa 6399 (when (setq found-type (c-forward-type t)) ; brace-block-too
0386b551
AM
6400 ;; Found a known or possible type or a prefix of a known type.
6401
6402 (when at-type
6403 ;; Got two identifiers with nothing but whitespace
6404 ;; between them. That can only happen in declarations.
6405 (setq at-decl-or-cast 'ids)
6406
6407 (when (eq at-type 'found)
6408 ;; If the previous identifier is a found type we
6409 ;; record it as a real one; it might be some sort of
6410 ;; alias for a prefix like "unsigned".
6411 (save-excursion
6412 (goto-char type-start)
6413 (let ((c-promote-possible-types t))
6414 (c-forward-type)))))
6415
6416 (setq backup-at-type at-type
6417 backup-type-start type-start
6418 backup-id-start id-start
6419 at-type found-type
6420 type-start start
6421 id-start (point)
6422 ;; The previous ambiguous specifier/type turned out
6423 ;; to be a type since we've parsed another one after
6424 ;; it, so clear these backup flags.
6425 backup-at-type-decl nil
6426 backup-maybe-typeless nil))
6427
6428 (if kwd-sym
6429 (progn
6430 ;; Handle known specifier keywords and
6431 ;; `c-decl-hangon-kwds' which can occur after known
6432 ;; types.
6433
6434 (if (c-keyword-member kwd-sym 'c-decl-hangon-kwds)
6435 ;; It's a hang-on keyword that can occur anywhere.
6436 (progn
6437 (setq at-decl-or-cast t)
6438 (if at-type
6439 ;; Move the identifier start position if
6440 ;; we've passed a type.
6441 (setq id-start kwd-clause-end)
6442 ;; Otherwise treat this as a specifier and
6443 ;; move the fallback position.
6444 (setq start-pos kwd-clause-end))
6445 (goto-char kwd-clause-end))
6446
6447 ;; It's an ordinary specifier so we know that
6448 ;; anything before this can't be the type.
6449 (setq backup-at-type nil
6450 start-pos kwd-clause-end)
6451
6452 (if found-type
6453 ;; It's ambiguous whether this keyword is a
6454 ;; specifier or a type prefix, so set the backup
6455 ;; flags. (It's assumed that `c-forward-type'
6456 ;; moved further than `c-forward-keyword-clause'.)
6457 (progn
6458 (when (c-keyword-member kwd-sym 'c-typedef-decl-kwds)
6459 (setq backup-at-type-decl t))
6460 (when (c-keyword-member kwd-sym 'c-typeless-decl-kwds)
6461 (setq backup-maybe-typeless t)))
6462
6463 (when (c-keyword-member kwd-sym 'c-typedef-decl-kwds)
e15f8aaa
AM
6464 ;; This test only happens after we've scanned a type.
6465 ;; So, with valid syntax, kwd-sym can't be 'typedef.
0386b551
AM
6466 (setq at-type-decl t))
6467 (when (c-keyword-member kwd-sym 'c-typeless-decl-kwds)
6468 (setq maybe-typeless t))
6469
6470 ;; Haven't matched a type so it's an umambiguous
6471 ;; specifier keyword and we know we're in a
6472 ;; declaration.
6473 (setq at-decl-or-cast t)
6474
6475 (goto-char kwd-clause-end))))
6476
6477 ;; If the type isn't known we continue so that we'll jump
6478 ;; over all specifiers and type identifiers. The reason
6479 ;; to do this for a known type prefix is to make things
6480 ;; like "unsigned INT16" work.
6481 (and found-type (not (eq found-type t))))))
6482
6483 (cond
6484 ((eq at-type t)
6485 ;; If a known type was found, we still need to skip over any
6486 ;; hangon keyword clauses after it. Otherwise it has already
6487 ;; been done in the loop above.
6488 (while (looking-at c-decl-hangon-key)
6489 (c-forward-keyword-clause 1))
6490 (setq id-start (point)))
6491
6492 ((eq at-type 'prefix)
6493 ;; A prefix type is itself a primitive type when it's not
6494 ;; followed by another type.
6495 (setq at-type t))
6496
6497 ((not at-type)
6498 ;; Got no type but set things up to continue anyway to handle
6499 ;; the various cases when a declaration doesn't start with a
6500 ;; type.
6501 (setq id-start start-pos))
6502
6503 ((and (eq at-type 'maybe)
6504 (c-major-mode-is 'c++-mode))
6505 ;; If it's C++ then check if the last "type" ends on the form
6506 ;; "foo::foo" or "foo::~foo", i.e. if it's the name of a
6507 ;; (con|de)structor.
6508 (save-excursion
6509 (let (name end-2 end-1)
6510 (goto-char id-start)
6511 (c-backward-syntactic-ws)
6512 (setq end-2 (point))
6513 (when (and
6514 (c-simple-skip-symbol-backward)
6515 (progn
6516 (setq name
6517 (buffer-substring-no-properties (point) end-2))
6518 ;; Cheating in the handling of syntactic ws below.
6519 (< (skip-chars-backward ":~ \t\n\r\v\f") 0))
6520 (progn
6521 (setq end-1 (point))
6522 (c-simple-skip-symbol-backward))
6523 (>= (point) type-start)
6524 (equal (buffer-substring-no-properties (point) end-1)
6525 name))
6526 ;; It is a (con|de)structor name. In that case the
6527 ;; declaration is typeless so zap out any preceding
6528 ;; identifier(s) that we might have taken as types.
6529 (goto-char type-start)
6530 (setq at-type nil
6531 backup-at-type nil
6532 id-start type-start))))))
6533
6534 ;; Check for and step over a type decl expression after the thing
6535 ;; that is or might be a type. This can't be skipped since we
6536 ;; need the correct end position of the declarator for
6537 ;; `max-type-decl-end-*'.
6538 (let ((start (point)) (paren-depth 0) pos
6539 ;; True if there's a non-open-paren match of
6540 ;; `c-type-decl-prefix-key'.
6541 got-prefix
6542 ;; True if the declarator is surrounded by a parenthesis pair.
6543 got-parens
6544 ;; True if there is an identifier in the declarator.
6545 got-identifier
6546 ;; True if there's a non-close-paren match of
6547 ;; `c-type-decl-suffix-key'.
6548 got-suffix
6549 ;; True if there's a prefix match outside the outermost
6550 ;; paren pair that surrounds the declarator.
6551 got-prefix-before-parens
1379f2c5 6552 ;; True if there's a suffix match outside the outermost
0386b551
AM
6553 ;; paren pair that surrounds the declarator. The value is
6554 ;; the position of the first suffix match.
6555 got-suffix-after-parens
6556 ;; True if we've parsed the type decl to a token that is
6557 ;; known to end declarations in this context.
6558 at-decl-end
6559 ;; The earlier values of `at-type' and `type-start' if we've
6560 ;; shifted the type backwards.
6561 identifier-type identifier-start
6562 ;; If `c-parse-and-markup-<>-arglists' is set we need to
6563 ;; turn it off during the name skipping below to avoid
6564 ;; getting `c-type' properties that might be bogus. That
6565 ;; can happen since we don't know if
6566 ;; `c-restricted-<>-arglists' will be correct inside the
6567 ;; arglist paren that gets entered.
6568 c-parse-and-markup-<>-arglists)
6569
6570 (goto-char id-start)
6571
6572 ;; Skip over type decl prefix operators. (Note similar code in
6573 ;; `c-font-lock-declarators'.)
6574 (while (and (looking-at c-type-decl-prefix-key)
6575 (if (and (c-major-mode-is 'c++-mode)
05842630 6576 (match-beginning 3))
0386b551
AM
6577 ;; If the second submatch matches in C++ then
6578 ;; we're looking at an identifier that's a
6579 ;; prefix only if it specifies a member pointer.
6580 (when (setq got-identifier (c-forward-name))
6581 (if (looking-at "\\(::\\)")
6582 ;; We only check for a trailing "::" and
6583 ;; let the "*" that should follow be
6584 ;; matched in the next round.
6585 (progn (setq got-identifier nil) t)
6586 ;; It turned out to be the real identifier,
6587 ;; so stop.
6588 nil))
6589 t))
6590
6591 (if (eq (char-after) ?\()
6592 (progn
6593 (setq paren-depth (1+ paren-depth))
6594 (forward-char))
6595 (unless got-prefix-before-parens
6596 (setq got-prefix-before-parens (= paren-depth 0)))
6597 (setq got-prefix t)
6598 (goto-char (match-end 1)))
6599 (c-forward-syntactic-ws))
6600
6601 (setq got-parens (> paren-depth 0))
6602
6603 ;; Skip over an identifier.
6604 (or got-identifier
6605 (and (looking-at c-identifier-start)
6606 (setq got-identifier (c-forward-name))))
6607
6608 ;; Skip over type decl suffix operators.
6609 (while (if (looking-at c-type-decl-suffix-key)
6610
6611 (if (eq (char-after) ?\))
6612 (when (> paren-depth 0)
6613 (setq paren-depth (1- paren-depth))
6614 (forward-char)
6615 t)
6616 (when (if (save-match-data (looking-at "\\s\("))
6617 (c-safe (c-forward-sexp 1) t)
6618 (goto-char (match-end 1))
6619 t)
6620 (when (and (not got-suffix-after-parens)
6621 (= paren-depth 0))
6622 (setq got-suffix-after-parens (match-beginning 0)))
6623 (setq got-suffix t)))
6624
6625 ;; No suffix matched. We might have matched the
6626 ;; identifier as a type and the open paren of a
6627 ;; function arglist as a type decl prefix. In that
6628 ;; case we should "backtrack": Reinterpret the last
6629 ;; type as the identifier, move out of the arglist and
6630 ;; continue searching for suffix operators.
6631 ;;
6632 ;; Do this even if there's no preceding type, to cope
6633 ;; with old style function declarations in K&R C,
6634 ;; (con|de)structors in C++ and `c-typeless-decl-kwds'
6635 ;; style declarations. That isn't applicable in an
6636 ;; arglist context, though.
6637 (when (and (= paren-depth 1)
6638 (not got-prefix-before-parens)
6639 (not (eq at-type t))
6640 (or backup-at-type
6641 maybe-typeless
6642 backup-maybe-typeless
6643 (when c-recognize-typeless-decls
6644 (not context)))
6645 (setq pos (c-up-list-forward (point)))
6646 (eq (char-before pos) ?\)))
6647 (c-fdoc-shift-type-backward)
6648 (goto-char pos)
6649 t))
6650
6651 (c-forward-syntactic-ws))
6652
6653 (when (and (or maybe-typeless backup-maybe-typeless)
6654 (not got-identifier)
6655 (not got-prefix)
6656 at-type)
6657 ;; Have found no identifier but `c-typeless-decl-kwds' has
6658 ;; matched so we know we're inside a declaration. The
6659 ;; preceding type must be the identifier instead.
6660 (c-fdoc-shift-type-backward))
6661
6662 (setq
6663 at-decl-or-cast
6664 (catch 'at-decl-or-cast
6665
a85fd6da 6666 ;; CASE 1
0386b551
AM
6667 (when (> paren-depth 0)
6668 ;; Encountered something inside parens that isn't matched by
6669 ;; the `c-type-decl-*' regexps, so it's not a type decl
6670 ;; expression. Try to skip out to the same paren depth to
6671 ;; not confuse the cast check below.
6672 (c-safe (goto-char (scan-lists (point) 1 paren-depth)))
6673 ;; If we've found a specifier keyword then it's a
6674 ;; declaration regardless.
6675 (throw 'at-decl-or-cast (eq at-decl-or-cast t)))
6676
6677 (setq at-decl-end
6678 (looking-at (cond ((eq context '<>) "[,>]")
6679 (context "[,\)]")
6680 (t "[,;]"))))
6681
6682 ;; Now we've collected info about various characteristics of
6683 ;; the construct we're looking at. Below follows a decision
6684 ;; tree based on that. It's ordered to check more certain
6685 ;; signs before less certain ones.
6686
6687 (if got-identifier
6688 (progn
6689
a85fd6da 6690 ;; CASE 2
0386b551
AM
6691 (when (and (or at-type maybe-typeless)
6692 (not (or got-prefix got-parens)))
6693 ;; Got another identifier directly after the type, so it's a
6694 ;; declaration.
6695 (throw 'at-decl-or-cast t))
6696
6697 (when (and got-parens
6698 (not got-prefix)
6699 (not got-suffix-after-parens)
6700 (or backup-at-type
6701 maybe-typeless
6702 backup-maybe-typeless))
6703 ;; Got a declaration of the form "foo bar (gnu);" where we've
6704 ;; recognized "bar" as the type and "gnu" as the declarator.
6705 ;; In this case it's however more likely that "bar" is the
6706 ;; declarator and "gnu" a function argument or initializer (if
6707 ;; `c-recognize-paren-inits' is set), since the parens around
6708 ;; "gnu" would be superfluous if it's a declarator. Shift the
6709 ;; type one step backward.
6710 (c-fdoc-shift-type-backward)))
6711
6712 ;; Found no identifier.
6713
6714 (if backup-at-type
6715 (progn
6716
b248a85d
AM
6717
6718 ;; CASE 3
6719 (when (= (point) start)
6720 ;; Got a plain list of identifiers. If a colon follows it's
6721 ;; a valid label, or maybe a bitfield. Otherwise the last
6722 ;; one probably is the declared identifier and we should
6723 ;; back up to the previous type, providing it isn't a cast.
452ea855
AM
6724 (if (and (eq (char-after) ?:)
6725 (not (c-major-mode-is 'java-mode)))
b248a85d
AM
6726 (cond
6727 ;; If we've found a specifier keyword then it's a
6728 ;; declaration regardless.
6729 ((eq at-decl-or-cast t)
6730 (throw 'at-decl-or-cast t))
6731 ((and c-has-bitfields
6732 (eq at-decl-or-cast 'ids)) ; bitfield.
6733 (setq backup-if-not-cast t)
6734 (throw 'at-decl-or-cast t)))
6735
6736 (setq backup-if-not-cast t)
6737 (throw 'at-decl-or-cast t)))
0386b551 6738
a85fd6da 6739 ;; CASE 4
0386b551
AM
6740 (when (and got-suffix
6741 (not got-prefix)
6742 (not got-parens))
6743 ;; Got a plain list of identifiers followed by some suffix.
6744 ;; If this isn't a cast then the last identifier probably is
6745 ;; the declared one and we should back up to the previous
6746 ;; type.
6747 (setq backup-if-not-cast t)
6748 (throw 'at-decl-or-cast t)))
6749
a85fd6da 6750 ;; CASE 5
0386b551
AM
6751 (when (eq at-type t)
6752 ;; If the type is known we know that there can't be any
6753 ;; identifier somewhere else, and it's only in declarations in
6754 ;; e.g. function prototypes and in casts that the identifier may
6755 ;; be left out.
6756 (throw 'at-decl-or-cast t))
6757
6758 (when (= (point) start)
6759 ;; Only got a single identifier (parsed as a type so far).
a85fd6da 6760 ;; CASE 6
0386b551
AM
6761 (if (and
6762 ;; Check that the identifier isn't at the start of an
6763 ;; expression.
6764 at-decl-end
6765 (cond
6766 ((eq context 'decl)
6767 ;; Inside an arglist that contains declarations. If K&R
6768 ;; style declarations and parenthesis style initializers
6769 ;; aren't allowed then the single identifier must be a
6770 ;; type, else we require that it's known or found
6771 ;; (primitive types are handled above).
6772 (or (and (not c-recognize-knr-p)
6773 (not c-recognize-paren-inits))
6774 (memq at-type '(known found))))
6775 ((eq context '<>)
6776 ;; Inside a template arglist. Accept known and found
6777 ;; types; other identifiers could just as well be
6778 ;; constants in C++.
6779 (memq at-type '(known found)))))
6780 (throw 'at-decl-or-cast t)
a85fd6da 6781 ;; CASE 7
0386b551
AM
6782 ;; Can't be a valid declaration or cast, but if we've found a
6783 ;; specifier it can't be anything else either, so treat it as
6784 ;; an invalid/unfinished declaration or cast.
6785 (throw 'at-decl-or-cast at-decl-or-cast))))
6786
6787 (if (and got-parens
6788 (not got-prefix)
6789 (not context)
6790 (not (eq at-type t))
6791 (or backup-at-type
6792 maybe-typeless
6793 backup-maybe-typeless
6794 (when c-recognize-typeless-decls
6795 (or (not got-suffix)
6796 (not (looking-at
6797 c-after-suffixed-type-maybe-decl-key))))))
6798 ;; Got an empty paren pair and a preceding type that probably
6799 ;; really is the identifier. Shift the type backwards to make
6800 ;; the last one the identifier. This is analogous to the
6801 ;; "backtracking" done inside the `c-type-decl-suffix-key' loop
6802 ;; above.
6803 ;;
6804 ;; Exception: In addition to the conditions in that
6805 ;; "backtracking" code, do not shift backward if we're not
6806 ;; looking at either `c-after-suffixed-type-decl-key' or "[;,]".
6807 ;; Since there's no preceding type, the shift would mean that
6808 ;; the declaration is typeless. But if the regexp doesn't match
6809 ;; then we will simply fall through in the tests below and not
6810 ;; recognize it at all, so it's better to try it as an abstract
6811 ;; declarator instead.
6812 (c-fdoc-shift-type-backward)
6813
6814 ;; Still no identifier.
a85fd6da 6815 ;; CASE 8
0386b551
AM
6816 (when (and got-prefix (or got-parens got-suffix))
6817 ;; Require `got-prefix' together with either `got-parens' or
6818 ;; `got-suffix' to recognize it as an abstract declarator:
6819 ;; `got-parens' only is probably an empty function call.
6820 ;; `got-suffix' only can build an ordinary expression together
6821 ;; with the preceding identifier which we've taken as a type.
6822 ;; We could actually accept on `got-prefix' only, but that can
6823 ;; easily occur temporarily while writing an expression so we
6824 ;; avoid that case anyway. We could do a better job if we knew
6825 ;; the point when the fontification was invoked.
6826 (throw 'at-decl-or-cast t))
6827
a85fd6da 6828 ;; CASE 9
0386b551
AM
6829 (when (and at-type
6830 (not got-prefix)
6831 (not got-parens)
6832 got-suffix-after-parens
6833 (eq (char-after got-suffix-after-parens) ?\())
6834 ;; Got a type, no declarator but a paren suffix. I.e. it's a
6835 ;; normal function call afterall (or perhaps a C++ style object
6836 ;; instantiation expression).
6837 (throw 'at-decl-or-cast nil))))
6838
a85fd6da 6839 ;; CASE 10
0386b551
AM
6840 (when at-decl-or-cast
6841 ;; By now we've located the type in the declaration that we know
6842 ;; we're in.
6843 (throw 'at-decl-or-cast t))
6844
a85fd6da 6845 ;; CASE 11
0386b551
AM
6846 (when (and got-identifier
6847 (not context)
6848 (looking-at c-after-suffixed-type-decl-key)
6849 (if (and got-parens
6850 (not got-prefix)
6851 (not got-suffix)
6852 (not (eq at-type t)))
6853 ;; Shift the type backward in the case that there's a
6854 ;; single identifier inside parens. That can only
6855 ;; occur in K&R style function declarations so it's
6856 ;; more likely that it really is a function call.
6857 ;; Therefore we only do this after
6858 ;; `c-after-suffixed-type-decl-key' has matched.
6859 (progn (c-fdoc-shift-type-backward) t)
6860 got-suffix-after-parens))
6861 ;; A declaration according to `c-after-suffixed-type-decl-key'.
6862 (throw 'at-decl-or-cast t))
6863
a85fd6da 6864 ;; CASE 12
0386b551
AM
6865 (when (and (or got-prefix (not got-parens))
6866 (memq at-type '(t known)))
6867 ;; It's a declaration if a known type precedes it and it can't be a
6868 ;; function call.
6869 (throw 'at-decl-or-cast t))
6870
6871 ;; If we get here we can't tell if this is a type decl or a normal
6872 ;; expression by looking at it alone. (That's under the assumption
6873 ;; that normal expressions always can look like type decl expressions,
6874 ;; which isn't really true but the cases where it doesn't hold are so
6875 ;; uncommon (e.g. some placements of "const" in C++) it's not worth
6876 ;; the effort to look for them.)
6877
6878 (unless (or at-decl-end (looking-at "=[^=]"))
6879 ;; If this is a declaration it should end here or its initializer(*)
6880 ;; should start here, so check for allowed separation tokens. Note
6881 ;; that this rule doesn't work e.g. with a K&R arglist after a
6882 ;; function header.
6883 ;;
6884 ;; *) Don't check for C++ style initializers using parens
6885 ;; since those already have been matched as suffixes.
6886 ;;
6887 ;; If `at-decl-or-cast' is then we've found some other sign that
6888 ;; it's a declaration or cast, so then it's probably an
6889 ;; invalid/unfinished one.
6890 (throw 'at-decl-or-cast at-decl-or-cast))
6891
6892 ;; Below are tests that only should be applied when we're certain to
6893 ;; not have parsed halfway through an expression.
6894
a85fd6da 6895 ;; CASE 14
0386b551
AM
6896 (when (memq at-type '(t known))
6897 ;; The expression starts with a known type so treat it as a
6898 ;; declaration.
6899 (throw 'at-decl-or-cast t))
6900
a85fd6da 6901 ;; CASE 15
0386b551
AM
6902 (when (and (c-major-mode-is 'c++-mode)
6903 ;; In C++ we check if the identifier is a known type, since
6904 ;; (con|de)structors use the class name as identifier.
6905 ;; We've always shifted over the identifier as a type and
6906 ;; then backed up again in this case.
6907 identifier-type
6908 (or (memq identifier-type '(found known))
6909 (and (eq (char-after identifier-start) ?~)
6910 ;; `at-type' probably won't be 'found for
6911 ;; destructors since the "~" is then part of the
6912 ;; type name being checked against the list of
6913 ;; known types, so do a check without that
6914 ;; operator.
6915 (or (save-excursion
6916 (goto-char (1+ identifier-start))
6917 (c-forward-syntactic-ws)
6918 (c-with-syntax-table
6919 c-identifier-syntax-table
6920 (looking-at c-known-type-key)))
6921 (save-excursion
6922 (goto-char (1+ identifier-start))
6923 ;; We have already parsed the type earlier,
6924 ;; so it'd be possible to cache the end
6925 ;; position instead of redoing it here, but
6926 ;; then we'd need to keep track of another
6927 ;; position everywhere.
6928 (c-check-type (point)
6929 (progn (c-forward-type)
6930 (point))))))))
6931 (throw 'at-decl-or-cast t))
6932
6933 (if got-identifier
6934 (progn
a85fd6da 6935 ;; CASE 16
0386b551
AM
6936 (when (and got-prefix-before-parens
6937 at-type
6938 (or at-decl-end (looking-at "=[^=]"))
6939 (not context)
6940 (not got-suffix))
6941 ;; Got something like "foo * bar;". Since we're not inside an
6942 ;; arglist it would be a meaningless expression because the
6943 ;; result isn't used. We therefore choose to recognize it as
6944 ;; a declaration. Do not allow a suffix since it could then
6945 ;; be a function call.
6946 (throw 'at-decl-or-cast t))
6947
a85fd6da 6948 ;; CASE 17
0386b551
AM
6949 (when (and (or got-suffix-after-parens
6950 (looking-at "=[^=]"))
6951 (eq at-type 'found)
6952 (not (eq context 'arglist)))
6953 ;; Got something like "a (*b) (c);" or "a (b) = c;". It could
6954 ;; be an odd expression or it could be a declaration. Treat
6955 ;; it as a declaration if "a" has been used as a type
6956 ;; somewhere else (if it's a known type we won't get here).
6957 (throw 'at-decl-or-cast t)))
6958
a85fd6da 6959 ;; CASE 18
0386b551
AM
6960 (when (and context
6961 (or got-prefix
6962 (and (eq context 'decl)
6963 (not c-recognize-paren-inits)
6964 (or got-parens got-suffix))))
6965 ;; Got a type followed by an abstract declarator. If `got-prefix'
6966 ;; is set it's something like "a *" without anything after it. If
6967 ;; `got-parens' or `got-suffix' is set it's "a()", "a[]", "a()[]",
6968 ;; or similar, which we accept only if the context rules out
6969 ;; expressions.
6970 (throw 'at-decl-or-cast t)))
6971
6972 ;; If we had a complete symbol table here (which rules out
6973 ;; `c-found-types') we should return t due to the disambiguation rule
6974 ;; (in at least C++) that anything that can be parsed as a declaration
6975 ;; is a declaration. Now we're being more defensive and prefer to
6976 ;; highlight things like "foo (bar);" as a declaration only if we're
6977 ;; inside an arglist that contains declarations.
6978 (eq context 'decl))))
6979
6980 ;; The point is now after the type decl expression.
6981
6982 (cond
6983 ;; Check for a cast.
6984 ((save-excursion
6985 (and
6986 c-cast-parens
6987
6988 ;; Should be the first type/identifier in a cast paren.
6989 (> preceding-token-end (point-min))
6990 (memq (char-before preceding-token-end) c-cast-parens)
6991
6992 ;; The closing paren should follow.
6993 (progn
6994 (c-forward-syntactic-ws)
6995 (looking-at "\\s\)"))
6996
6997 ;; There should be a primary expression after it.
6998 (let (pos)
6999 (forward-char)
7000 (c-forward-syntactic-ws)
7001 (setq cast-end (point))
7002 (and (looking-at c-primary-expr-regexp)
7003 (progn
7004 (setq pos (match-end 0))
7005 (or
7006 ;; Check if the expression begins with a prefix keyword.
7007 (match-beginning 2)
7008 (if (match-beginning 1)
7009 ;; Expression begins with an ambiguous operator. Treat
7010 ;; it as a cast if it's a type decl or if we've
7011 ;; recognized the type somewhere else.
7012 (or at-decl-or-cast
7013 (memq at-type '(t known found)))
7014 ;; Unless it's a keyword, it's the beginning of a primary
7015 ;; expression.
7016 (not (looking-at c-keywords-regexp)))))
7017 ;; If `c-primary-expr-regexp' matched a nonsymbol token, check
7018 ;; that it matched a whole one so that we don't e.g. confuse
7019 ;; the operator '-' with '->'. It's ok if it matches further,
7020 ;; though, since it e.g. can match the float '.5' while the
7021 ;; operator regexp only matches '.'.
7022 (or (not (looking-at c-nonsymbol-token-regexp))
7023 (<= (match-end 0) pos))))
7024
7025 ;; There should either be a cast before it or something that isn't an
7026 ;; identifier or close paren.
7027 (> preceding-token-end (point-min))
7028 (progn
7029 (goto-char (1- preceding-token-end))
7030 (or (eq (point) last-cast-end)
7031 (progn
7032 (c-backward-syntactic-ws)
7033 (if (< (skip-syntax-backward "w_") 0)
7034 ;; It's a symbol. Accept it only if it's one of the
7035 ;; keywords that can precede an expression (without
7036 ;; surrounding parens).
7037 (looking-at c-simple-stmt-key)
7038 (and
7039 ;; Check that it isn't a close paren (block close is ok,
7040 ;; though).
7041 (not (memq (char-before) '(?\) ?\])))
7042 ;; Check that it isn't a nonsymbol identifier.
7043 (not (c-on-identifier)))))))))
7044
7045 ;; Handle the cast.
7046 (when (and c-record-type-identifiers at-type (not (eq at-type t)))
7047 (let ((c-promote-possible-types t))
7048 (goto-char type-start)
7049 (c-forward-type)))
7050
7051 (goto-char cast-end)
7052 'cast)
7053
7054 (at-decl-or-cast
7055 ;; We're at a declaration. Highlight the type and the following
7056 ;; declarators.
7057
7058 (when backup-if-not-cast
7059 (c-fdoc-shift-type-backward t))
7060
7061 (when (and (eq context 'decl) (looking-at ","))
7062 ;; Make sure to propagate the `c-decl-arg-start' property to
7063 ;; the next argument if it's set in this one, to cope with
7064 ;; interactive refontification.
7065 (c-put-c-type-property (point) 'c-decl-arg-start))
7066
7067 (when (and c-record-type-identifiers at-type (not (eq at-type t)))
7068 (let ((c-promote-possible-types t))
7069 (save-excursion
7070 (goto-char type-start)
7071 (c-forward-type))))
7072
e15f8aaa
AM
7073 (cons id-start
7074 (and (or at-type-decl at-typedef)
7075 (cons at-type-decl at-typedef))))
0386b551
AM
7076
7077 (t
7078 ;; False alarm. Restore the recorded ranges.
7079 (setq c-record-type-identifiers save-rec-type-ids
7080 c-record-ref-identifiers save-rec-ref-ids)
7081 nil))))
7082
7083(defun c-forward-label (&optional assume-markup preceding-token-end limit)
51c9af45 7084 ;; Assuming that point is at the beginning of a token, check if it starts a
1379f2c5
AM
7085 ;; label and if so move over it and return non-nil (t in default situations,
7086 ;; specific symbols (see below) for interesting situations), otherwise don't
7087 ;; move and return nil. "Label" here means "most things with a colon".
51c9af45
AM
7088 ;;
7089 ;; More precisely, a "label" is regarded as one of:
1379f2c5
AM
7090 ;; (i) a goto target like "foo:" - returns the symbol `goto-target';
7091 ;; (ii) A case label - either the entire construct "case FOO:", or just the
7092 ;; bare "case", should the colon be missing. We return t;
7093 ;; (iii) a keyword which needs a colon, like "default:" or "private:"; We
7094 ;; return t;
51c9af45 7095 ;; (iv) One of QT's "extended" C++ variants of
1379f2c5
AM
7096 ;; "private:"/"protected:"/"public:"/"more:" looking like "public slots:".
7097 ;; Returns the symbol `qt-2kwds-colon'.
7098 ;; (v) QT's construct "signals:". Returns the symbol `qt-1kwd-colon'.
82ba65cf 7099 ;; (vi) One of the keywords matched by `c-opt-extra-label-key' (without any
51c9af45 7100 ;; colon). Currently (2006-03), this applies only to Objective C's
1379f2c5 7101 ;; keywords "@private", "@protected", and "@public". Returns t.
51c9af45
AM
7102 ;;
7103 ;; One of the things which will NOT be recognised as a label is a bit-field
7104 ;; element of a struct, something like "int foo:5".
7105 ;;
7106 ;; The end of the label is taken to be just after the colon, or the end of
7107 ;; the first submatch in `c-opt-extra-label-key'. The point is directly
7108 ;; after the end on return. The terminating char gets marked with
7109 ;; `c-decl-end' to improve recognition of the following declaration or
7110 ;; statement.
0386b551
AM
7111 ;;
7112 ;; If ASSUME-MARKUP is non-nil, it's assumed that the preceding
51c9af45 7113 ;; label, if any, has already been marked up like that.
0386b551
AM
7114 ;;
7115 ;; If PRECEDING-TOKEN-END is given, it should be the first position
7116 ;; after the preceding token, i.e. on the other side of the
7117 ;; syntactic ws from the point. Use a value less than or equal to
7118 ;; (point-min) if the point is at the first token in (the visible
7119 ;; part of) the buffer.
7120 ;;
7121 ;; The optional LIMIT limits the forward scan for the colon.
7122 ;;
7123 ;; This function records the ranges of the label symbols on
7124 ;; `c-record-ref-identifiers' if `c-record-type-identifiers' (!) is
7125 ;; non-nil.
7126 ;;
7127 ;; This function might do hidden buffer changes.
7128
51c9af45 7129 (let ((start (point))
1379f2c5 7130 label-end
51c9af45 7131 qt-symbol-idx
1379f2c5 7132 macro-start ; if we're in one.
f412a567
AM
7133 label-type
7134 kwd)
0386b551 7135 (cond
b414f371 7136 ;; "case" or "default" (Doesn't apply to AWK).
0386b551
AM
7137 ((looking-at c-label-kwds-regexp)
7138 (let ((kwd-end (match-end 1)))
7139 ;; Record only the keyword itself for fontification, since in
7140 ;; case labels the following is a constant expression and not
7141 ;; a label.
7142 (when c-record-type-identifiers
7143 (c-record-ref-id (cons (match-beginning 1) kwd-end)))
7144
7145 ;; Find the label end.
7146 (goto-char kwd-end)
1379f2c5
AM
7147 (setq label-type
7148 (if (and (c-syntactic-re-search-forward
7149 ;; Stop on chars that aren't allowed in expressions,
7150 ;; and on operator chars that would be meaningless
7151 ;; there. FIXME: This doesn't cope with ?: operators.
7152 "[;{=,@]\\|\\(\\=\\|[^:]\\):\\([^:]\\|\\'\\)"
7153 limit t t nil 1)
7154 (match-beginning 2))
7155
7156 (progn ; there's a proper :
7157 (goto-char (match-beginning 2)) ; just after the :
7158 (c-put-c-type-property (1- (point)) 'c-decl-end)
7159 t)
7160
7161 ;; It's an unfinished label. We consider the keyword enough
7162 ;; to recognize it as a label, so that it gets fontified.
7163 ;; Leave the point at the end of it, but don't put any
7164 ;; `c-decl-end' marker.
7165 (goto-char kwd-end)
7166 t))))
0386b551 7167
51c9af45 7168 ;; @private, @protected, @public, in Objective C, or similar.
0386b551
AM
7169 ((and c-opt-extra-label-key
7170 (looking-at c-opt-extra-label-key))
7171 ;; For a `c-opt-extra-label-key' match, we record the whole
7172 ;; thing for fontification. That's to get the leading '@' in
7173 ;; Objective-C protection labels fontified.
7174 (goto-char (match-end 1))
7175 (when c-record-type-identifiers
7176 (c-record-ref-id (cons (match-beginning 1) (point))))
7177 (c-put-c-type-property (1- (point)) 'c-decl-end)
1379f2c5 7178 (setq label-type t))
0386b551 7179
51c9af45
AM
7180 ;; All other cases of labels.
7181 ((and c-recognize-colon-labels ; nil for AWK and IDL, otherwise t.
0386b551
AM
7182
7183 ;; A colon label must have something before the colon.
7184 (not (eq (char-after) ?:))
7185
7186 ;; Check that we're not after a token that can't precede a label.
7187 (or
7188 ;; Trivially succeeds when there's no preceding token.
536610a4 7189 ;; Succeeds when we're at a virtual semicolon.
0386b551
AM
7190 (if preceding-token-end
7191 (<= preceding-token-end (point-min))
7192 (save-excursion
7193 (c-backward-syntactic-ws)
7194 (setq preceding-token-end (point))
536610a4
AM
7195 (or (bobp)
7196 (c-at-vsemi-p))))
0386b551
AM
7197
7198 ;; Check if we're after a label, if we're after a closing
7199 ;; paren that belong to statement, and with
7200 ;; `c-label-prefix-re'. It's done in different order
7201 ;; depending on `assume-markup' since the checks have
7202 ;; different expensiveness.
7203 (if assume-markup
7204 (or
7205 (eq (c-get-char-property (1- preceding-token-end) 'c-type)
7206 'c-decl-end)
7207
7208 (save-excursion
7209 (goto-char (1- preceding-token-end))
7210 (c-beginning-of-current-token)
51c9af45
AM
7211 (or (looking-at c-label-prefix-re)
7212 (looking-at c-block-stmt-1-key)))
0386b551
AM
7213
7214 (and (eq (char-before preceding-token-end) ?\))
7215 (c-after-conditional)))
7216
7217 (or
7218 (save-excursion
7219 (goto-char (1- preceding-token-end))
7220 (c-beginning-of-current-token)
51c9af45
AM
7221 (or (looking-at c-label-prefix-re)
7222 (looking-at c-block-stmt-1-key)))
0386b551
AM
7223
7224 (cond
7225 ((eq (char-before preceding-token-end) ?\))
7226 (c-after-conditional))
7227
7228 ((eq (char-before preceding-token-end) ?:)
7229 ;; Might be after another label, so check it recursively.
51c9af45
AM
7230 (save-restriction
7231 (save-excursion
7232 (goto-char (1- preceding-token-end))
7233 ;; Essentially the same as the
7234 ;; `c-syntactic-re-search-forward' regexp below.
7235 (setq macro-start
7236 (save-excursion (and (c-beginning-of-macro)
7237 (point))))
7238 (if macro-start (narrow-to-region macro-start (point-max)))
7239 (c-syntactic-skip-backward "^-]:?;}=*/%&|,<>!@+" nil t)
7240 ;; Note: the following should work instead of the
7241 ;; narrow-to-region above. Investigate why not,
7242 ;; sometime. ACM, 2006-03-31.
7243 ;; (c-syntactic-skip-backward "^-]:?;}=*/%&|,<>!@+"
7244 ;; macro-start t)
7245 (let ((pte (point))
7246 ;; If the caller turned on recording for us,
7247 ;; it shouldn't apply when we check the
7248 ;; preceding label.
7249 c-record-type-identifiers)
7250 ;; A label can't start at a cpp directive. Check for
7251 ;; this, since c-forward-syntactic-ws would foul up on it.
7252 (unless (and c-opt-cpp-prefix (looking-at c-opt-cpp-prefix))
7253 (c-forward-syntactic-ws)
7254 (c-forward-label nil pte start))))))))))
7255
1379f2c5 7256 ;; Point is still at the beginning of the possible label construct.
b414f371 7257 ;;
51c9af45
AM
7258 ;; Check that the next nonsymbol token is ":", or that we're in one
7259 ;; of QT's "slots" declarations. Allow '(' for the sake of macro
7260 ;; arguments. FIXME: Should build this regexp from the language
7261 ;; constants.
1379f2c5
AM
7262 (cond
7263 ;; public: protected: private:
7264 ((and
7265 (c-major-mode-is 'c++-mode)
7266 (search-forward-regexp
7267 "\\=p\\(r\\(ivate\\|otected\\)\\|ublic\\)\\>[^_]" nil t)
7268 (progn (backward-char)
7269 (c-forward-syntactic-ws limit)
7270 (looking-at ":\\([^:]\\|\\'\\)"))) ; A single colon.
7271 (forward-char)
7272 (setq label-type t))
7273 ;; QT double keyword like "protected slots:" or goto target.
7274 ((progn (goto-char start) nil))
7275 ((when (c-syntactic-re-search-forward
7276 "[ \t\n[:?;{=*/%&|,<>!@+-]" limit t t) ; not at EOB
7277 (backward-char)
7278 (setq label-end (point))
7279 (setq qt-symbol-idx
7280 (and (c-major-mode-is 'c++-mode)
7281 (string-match
7282 "\\(p\\(r\\(ivate\\|otected\\)\\|ublic\\)\\|more\\)\\>"
b414f371 7283 (buffer-substring start (point)))))
1379f2c5
AM
7284 (c-forward-syntactic-ws limit)
7285 (cond
7286 ((looking-at ":\\([^:]\\|\\'\\)") ; A single colon.
7287 (forward-char)
7288 (setq label-type
f412a567
AM
7289 (if (or (string= "signals" ; Special QT macro
7290 (setq kwd (buffer-substring-no-properties start label-end)))
7291 (string= "Q_SIGNALS" kwd))
1379f2c5
AM
7292 'qt-1kwd-colon
7293 'goto-target)))
7294 ((and qt-symbol-idx
f412a567 7295 (search-forward-regexp "\\=\\(slots\\|Q_SLOTS\\)\\>" limit t)
1379f2c5
AM
7296 (progn (c-forward-syntactic-ws limit)
7297 (looking-at ":\\([^:]\\|\\'\\)"))) ; A single colon
7298 (forward-char)
7299 (setq label-type 'qt-2kwds-colon)))))))
0386b551
AM
7300
7301 (save-restriction
7302 (narrow-to-region start (point))
7303
7304 ;; Check that `c-nonlabel-token-key' doesn't match anywhere.
7305 (catch 'check-label
7306 (goto-char start)
7307 (while (progn
7308 (when (looking-at c-nonlabel-token-key)
7309 (goto-char start)
1379f2c5 7310 (setq label-type nil)
0386b551
AM
7311 (throw 'check-label nil))
7312 (and (c-safe (c-forward-sexp)
7313 (c-forward-syntactic-ws)
7314 t)
7315 (not (eobp)))))
7316
7317 ;; Record the identifiers in the label for fontification, unless
7318 ;; it begins with `c-label-kwds' in which case the following
7319 ;; identifiers are part of a (constant) expression that
7320 ;; shouldn't be fontified.
7321 (when (and c-record-type-identifiers
7322 (progn (goto-char start)
7323 (not (looking-at c-label-kwds-regexp))))
7324 (while (c-syntactic-re-search-forward c-symbol-key nil t)
7325 (c-record-ref-id (cons (match-beginning 0)
7326 (match-end 0)))))
7327
7328 (c-put-c-type-property (1- (point-max)) 'c-decl-end)
82ba65cf 7329 (goto-char (point-max)))))
0386b551
AM
7330
7331 (t
7332 ;; Not a label.
1379f2c5
AM
7333 (goto-char start)))
7334 label-type))
0386b551
AM
7335
7336(defun c-forward-objc-directive ()
7337 ;; Assuming the point is at the beginning of a token, try to move
7338 ;; forward to the end of the Objective-C directive that starts
7339 ;; there. Return t if a directive was fully recognized, otherwise
7340 ;; the point is moved as far as one could be successfully parsed and
7341 ;; nil is returned.
7342 ;;
7343 ;; This function records identifier ranges on
7344 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
7345 ;; `c-record-type-identifiers' is non-nil.
7346 ;;
7347 ;; This function might do hidden buffer changes.
7348
7349 (let ((start (point))
7350 start-char
7351 (c-promote-possible-types t)
7352 ;; Turn off recognition of angle bracket arglists while parsing
7353 ;; types here since the protocol reference list might then be
7354 ;; considered part of the preceding name or superclass-name.
7355 c-recognize-<>-arglists)
7356
7357 (if (or
7358 (when (looking-at
7359 (eval-when-compile
7360 (c-make-keywords-re t
7361 (append (c-lang-const c-protection-kwds objc)
7362 '("@end"))
7363 'objc-mode)))
7364 (goto-char (match-end 1))
7365 t)
7366
7367 (and
7368 (looking-at
7369 (eval-when-compile
7370 (c-make-keywords-re t
7371 '("@interface" "@implementation" "@protocol")
7372 'objc-mode)))
7373
7374 ;; Handle the name of the class itself.
7375 (progn
cb694ab7
AM
7376; (c-forward-token-2) ; 2006/1/13 This doesn't move if the token's
7377; at EOB.
7378 (goto-char (match-end 0))
7379 (c-skip-ws-forward)
0386b551
AM
7380 (c-forward-type))
7381
7382 (catch 'break
7383 ;; Look for ": superclass-name" or "( category-name )".
7384 (when (looking-at "[:\(]")
7385 (setq start-char (char-after))
7386 (forward-char)
7387 (c-forward-syntactic-ws)
7388 (unless (c-forward-type) (throw 'break nil))
7389 (when (eq start-char ?\()
7390 (unless (eq (char-after) ?\)) (throw 'break nil))
7391 (forward-char)
7392 (c-forward-syntactic-ws)))
7393
7394 ;; Look for a protocol reference list.
7395 (if (eq (char-after) ?<)
7396 (let ((c-recognize-<>-arglists t)
7397 (c-parse-and-markup-<>-arglists t)
7398 c-restricted-<>-arglists)
7399 (c-forward-<>-arglist t))
7400 t))))
7401
7402 (progn
7403 (c-backward-syntactic-ws)
7404 (c-clear-c-type-property start (1- (point)) 'c-decl-end)
7405 (c-put-c-type-property (1- (point)) 'c-decl-end)
7406 t)
7407
7408 (c-clear-c-type-property start (point) 'c-decl-end)
7409 nil)))
7410
785eecbb
RS
7411(defun c-beginning-of-inheritance-list (&optional lim)
7412 ;; Go to the first non-whitespace after the colon that starts a
7413 ;; multiple inheritance introduction. Optional LIM is the farthest
7414 ;; back we should search.
0386b551
AM
7415 ;;
7416 ;; This function might do hidden buffer changes.
7417 (c-with-syntax-table c++-template-syntax-table
7418 (c-backward-token-2 0 t lim)
7419 (while (and (or (looking-at c-symbol-start)
7420 (looking-at "[<,]\\|::"))
7421 (zerop (c-backward-token-2 1 t lim))))))
785eecbb 7422
785eecbb
RS
7423(defun c-in-method-def-p ()
7424 ;; Return nil if we aren't in a method definition, otherwise the
7425 ;; position of the initial [+-].
0386b551
AM
7426 ;;
7427 ;; This function might do hidden buffer changes.
785eecbb
RS
7428 (save-excursion
7429 (beginning-of-line)
a66cd3ee
MS
7430 (and c-opt-method-key
7431 (looking-at c-opt-method-key)
785eecbb
RS
7432 (point))
7433 ))
7434
a66cd3ee
MS
7435;; Contributed by Kevin Ryde <user42@zip.com.au>.
7436(defun c-in-gcc-asm-p ()
7437 ;; Return non-nil if point is within a gcc \"asm\" block.
7438 ;;
7439 ;; This should be called with point inside an argument list.
7440 ;;
7441 ;; Only one level of enclosing parentheses is considered, so for
7442 ;; instance `nil' is returned when in a function call within an asm
7443 ;; operand.
0386b551
AM
7444 ;;
7445 ;; This function might do hidden buffer changes.
a66cd3ee
MS
7446
7447 (and c-opt-asm-stmt-key
7448 (save-excursion
7449 (beginning-of-line)
7450 (backward-up-list 1)
7451 (c-beginning-of-statement-1 (point-min) nil t)
7452 (looking-at c-opt-asm-stmt-key))))
7453
abb7e5cf 7454(defun c-at-toplevel-p ()
a85fd6da
AM
7455 "Return a determination as to whether point is \"at the top level\".
7456Informally, \"at the top level\" is anywhere where you can write
7457a function.
7458
7459More precisely, being at the top-level means that point is either
7460outside any enclosing block (such as a function definition), or
7461directly inside a class, namespace or other block that contains
7462another declaration level.
abb7e5cf
SM
7463
7464If point is not at the top-level (e.g. it is inside a method
7465definition), then nil is returned. Otherwise, if point is at a
7466top-level not enclosed within a class definition, t is returned.
7467Otherwise, a 2-vector is returned where the zeroth element is the
7468buffer position of the start of the class declaration, and the first
7469element is the buffer position of the enclosing class's opening
0386b551
AM
7470brace.
7471
7472Note that this function might do hidden buffer changes. See the
7473comment at the start of cc-engine.el for more info."
a66cd3ee
MS
7474 (let ((paren-state (c-parse-state)))
7475 (or (not (c-most-enclosing-brace paren-state))
7476 (c-search-uplist-for-classkey paren-state))))
7477
d9e94c22 7478(defun c-just-after-func-arglist-p (&optional lim)
0386b551
AM
7479 ;; Return non-nil if the point is in the region after the argument
7480 ;; list of a function and its opening brace (or semicolon in case it
7481 ;; got no body). If there are K&R style argument declarations in
7482 ;; that region, the point has to be inside the first one for this
7483 ;; function to recognize it.
a66cd3ee 7484 ;;
0386b551
AM
7485 ;; If successful, the point is moved to the first token after the
7486 ;; function header (see `c-forward-decl-or-cast-1' for details) and
7487 ;; the position of the opening paren of the function arglist is
7488 ;; returned.
7489 ;;
7490 ;; The point is clobbered if not successful.
7491 ;;
7492 ;; LIM is used as bound for backward buffer searches.
7493 ;;
7494 ;; This function might do hidden buffer changes.
7495
7496 (let ((beg (point)) end id-start)
7497 (and
7498 (eq (c-beginning-of-statement-1 lim) 'same)
7499
7500 (not (or (c-major-mode-is 'objc-mode)
7501 (c-forward-objc-directive)))
7502
7503 (setq id-start
7504 (car-safe (c-forward-decl-or-cast-1 (c-point 'bosws) nil nil)))
7505 (< id-start beg)
7506
7507 ;; There should not be a '=' or ',' between beg and the
7508 ;; start of the declaration since that means we were in the
7509 ;; "expression part" of the declaration.
7510 (or (> (point) beg)
7511 (not (looking-at "[=,]")))
7512
7513 (save-excursion
7514 ;; Check that there's an arglist paren in the
7515 ;; declaration.
7516 (goto-char id-start)
7517 (cond ((eq (char-after) ?\()
7518 ;; The declarator is a paren expression, so skip past it
7519 ;; so that we don't get stuck on that instead of the
7520 ;; function arglist.
7521 (c-forward-sexp))
51c9af45
AM
7522 ((and c-opt-op-identifier-prefix
7523 (looking-at c-opt-op-identifier-prefix))
0386b551
AM
7524 ;; Don't trip up on "operator ()".
7525 (c-forward-token-2 2 t)))
7526 (and (< (point) beg)
7527 (c-syntactic-re-search-forward "(" beg t t)
7528 (1- (point)))))))
785eecbb 7529
a66cd3ee
MS
7530(defun c-in-knr-argdecl (&optional lim)
7531 ;; Return the position of the first argument declaration if point is
7532 ;; inside a K&R style argument declaration list, nil otherwise.
7533 ;; `c-recognize-knr-p' is not checked. If LIM is non-nil, it's a
7534 ;; position that bounds the backward search for the argument list.
7535 ;;
9cf17ef1
AM
7536 ;; Point must be within a possible K&R region, e.g. just before a top-level
7537 ;; "{". It must be outside of parens and brackets. The test can return
7538 ;; false positives otherwise.
0386b551
AM
7539 ;;
7540 ;; This function might do hidden buffer changes.
d9e94c22 7541
a66cd3ee
MS
7542 (save-excursion
7543 (save-restriction
9cf17ef1
AM
7544 ;; If we're in a macro, our search range is restricted to it. Narrow to
7545 ;; the searchable range.
7546 (let* ((macro-start (c-query-macro-start))
7547 (lim (max (or lim (point-min)) (or macro-start (point-min))))
6b6481ed
AM
7548 before-lparen after-rparen
7549 (pp-count-out 20)) ; Max number of paren/brace constructs before we give up
9cf17ef1
AM
7550 (narrow-to-region lim (c-point 'eol))
7551
7552 ;; Search backwards for the defun's argument list. We give up if we
7553 ;; encounter a "}" (end of a previous defun) or BOB.
7554 ;;
7555 ;; The criterion for a paren structure being the arg list is:
7556 ;; o - there is non-WS stuff after it but before any "{"; AND
7557 ;; o - the token after it isn't a ";" AND
7558 ;; o - it is preceded by either an identifier (the function name) or
7559 ;; a macro expansion like "DEFUN (...)"; AND
7560 ;; o - its content is a non-empty comma-separated list of identifiers
7561 ;; (an empty arg list won't have a knr region).
7562 ;;
7563 ;; The following snippet illustrates these rules:
7564 ;; int foo (bar, baz, yuk)
7565 ;; int bar [] ;
7566 ;; int (*baz) (my_type) ;
7567 ;; int (*) (void) (*yuk) (void) ;
7568 ;; {
7569
7570 (catch 'knr
6b6481ed
AM
7571 (while (> pp-count-out 0) ; go back one paren/bracket pair each time.
7572 (setq pp-count-out (1- pp-count-out))
10489fcb 7573 (c-syntactic-skip-backward "^)]}")
9cf17ef1
AM
7574 (cond ((eq (char-before) ?\))
7575 (setq after-rparen (point)))
10489fcb
AM
7576 ((eq (char-before) ?\])
7577 (setq after-rparen nil))
7578 (t ; either } (hit previous defun) or no more parens/brackets
7579 (throw 'knr nil)))
9cf17ef1
AM
7580
7581 (if after-rparen
7582 ;; We're inside a paren. Could it be our argument list....?
7583 (if
7584 (and
7585 (progn
7586 (goto-char after-rparen)
7587 (unless (c-go-list-backward) (throw 'knr nil)) ;
7588 ;; FIXME!!! What about macros between the parens? 2007/01/20
7589 (setq before-lparen (point)))
d9e94c22 7590
9cf17ef1
AM
7591 ;; It can't be the arg list if next token is ; or {
7592 (progn (goto-char after-rparen)
7593 (c-forward-syntactic-ws)
7594 (not (memq (char-after) '(?\; ?\{))))
d9e94c22 7595
9cf17ef1
AM
7596 ;; Is the thing preceding the list an identifier (the
7597 ;; function name), or a macro expansion?
7598 (progn
7599 (goto-char before-lparen)
7600 (eq (c-backward-token-2) 0)
7601 (or (c-on-identifier)
7602 (and (eq (char-after) ?\))
7603 (c-go-up-list-backward)
7604 (eq (c-backward-token-2) 0)
7605 (c-on-identifier))))
7606
7607 ;; Have we got a non-empty list of comma-separated
7608 ;; identifiers?
7609 (progn
7610 (goto-char before-lparen)
7611 (c-forward-token-2) ; to first token inside parens
7612 (and
7613 (c-on-identifier)
7614 (c-forward-token-2)
7615 (catch 'id-list
7616 (while (eq (char-after) ?\,)
7617 (c-forward-token-2)
7618 (unless (c-on-identifier) (throw 'id-list nil))
7619 (c-forward-token-2))
7620 (eq (char-after) ?\))))))
7621
7622 ;; ...Yes. We've identified the function's argument list.
7623 (throw 'knr
7624 (progn (goto-char after-rparen)
7625 (c-forward-syntactic-ws)
7626 (point)))
7627
7628 ;; ...No. The current parens aren't the function's arg list.
7629 (goto-char before-lparen))
7630
7631 (or (c-go-list-backward) ; backwards over [ .... ]
7632 (throw 'knr nil)))))))))
785eecbb
RS
7633
7634(defun c-skip-conditional ()
7635 ;; skip forward over conditional at point, including any predicate
7636 ;; statements in parentheses. No error checking is performed.
0386b551
AM
7637 ;;
7638 ;; This function might do hidden buffer changes.
0ec8351b
BW
7639 (c-forward-sexp (cond
7640 ;; else if()
a66cd3ee
MS
7641 ((looking-at (concat "\\<else"
7642 "\\([ \t\n]\\|\\\\\n\\)+"
7643 "if\\>\\([^_]\\|$\\)"))
7644 3)
0ec8351b 7645 ;; do, else, try, finally
a66cd3ee
MS
7646 ((looking-at (concat "\\<\\("
7647 "do\\|else\\|try\\|finally"
7648 "\\)\\>\\([^_]\\|$\\)"))
130c507e 7649 1)
ce8c7486 7650 ;; for, if, while, switch, catch, synchronized, foreach
0ec8351b 7651 (t 2))))
785eecbb 7652
a66cd3ee
MS
7653(defun c-after-conditional (&optional lim)
7654 ;; If looking at the token after a conditional then return the
7655 ;; position of its start, otherwise return nil.
0386b551
AM
7656 ;;
7657 ;; This function might do hidden buffer changes.
a66cd3ee 7658 (save-excursion
d9e94c22 7659 (and (zerop (c-backward-token-2 1 t lim))
a66cd3ee
MS
7660 (or (looking-at c-block-stmt-1-key)
7661 (and (eq (char-after) ?\()
d9e94c22 7662 (zerop (c-backward-token-2 1 t lim))
a66cd3ee
MS
7663 (looking-at c-block-stmt-2-key)))
7664 (point))))
7665
0386b551
AM
7666(defun c-after-special-operator-id (&optional lim)
7667 ;; If the point is after an operator identifier that isn't handled
7668 ;; like an ordinary symbol (i.e. like "operator =" in C++) then the
7669 ;; position of the start of that identifier is returned. nil is
7670 ;; returned otherwise. The point may be anywhere in the syntactic
7671 ;; whitespace after the last token of the operator identifier.
7672 ;;
7673 ;; This function might do hidden buffer changes.
7674 (save-excursion
7675 (and c-overloadable-operators-regexp
7676 (zerop (c-backward-token-2 1 nil lim))
7677 (looking-at c-overloadable-operators-regexp)
51c9af45 7678 (or (not c-opt-op-identifier-prefix)
0386b551
AM
7679 (and
7680 (zerop (c-backward-token-2 1 nil lim))
51c9af45 7681 (looking-at c-opt-op-identifier-prefix)))
0386b551
AM
7682 (point))))
7683
a66cd3ee
MS
7684(defsubst c-backward-to-block-anchor (&optional lim)
7685 ;; Assuming point is at a brace that opens a statement block of some
7686 ;; kind, move to the proper anchor point for that block. It might
7687 ;; need to be adjusted further by c-add-stmt-syntax, but the
7688 ;; position at return is suitable as start position for that
7689 ;; function.
0386b551
AM
7690 ;;
7691 ;; This function might do hidden buffer changes.
a66cd3ee
MS
7692 (unless (= (point) (c-point 'boi))
7693 (let ((start (c-after-conditional lim)))
7694 (if start
7695 (goto-char start)))))
7696
037558bf 7697(defsubst c-backward-to-decl-anchor (&optional lim)
a66cd3ee
MS
7698 ;; Assuming point is at a brace that opens the block of a top level
7699 ;; declaration of some kind, move to the proper anchor point for
7700 ;; that block.
0386b551
AM
7701 ;;
7702 ;; This function might do hidden buffer changes.
a66cd3ee 7703 (unless (= (point) (c-point 'boi))
037558bf 7704 (c-beginning-of-statement-1 lim)))
a66cd3ee 7705
ff959bab 7706(defun c-search-decl-header-end ()
a66cd3ee
MS
7707 ;; Search forward for the end of the "header" of the current
7708 ;; declaration. That's the position where the definition body
7709 ;; starts, or the first variable initializer, or the ending
7710 ;; semicolon. I.e. search forward for the closest following
7711 ;; (syntactically relevant) '{', '=' or ';' token. Point is left
7712 ;; _after_ the first found token, or at point-max if none is found.
0386b551
AM
7713 ;;
7714 ;; This function might do hidden buffer changes.
ff959bab
MS
7715
7716 (let ((base (point)))
7717 (if (c-major-mode-is 'c++-mode)
7718
7719 ;; In C++ we need to take special care to handle operator
7720 ;; tokens and those pesky template brackets.
7721 (while (and
7722 (c-syntactic-re-search-forward "[;{<=]" nil 'move t t)
7723 (or
7724 (c-end-of-current-token base)
7725 ;; Handle operator identifiers, i.e. ignore any
7726 ;; operator token preceded by "operator".
7727 (save-excursion
7728 (and (c-safe (c-backward-sexp) t)
51c9af45 7729 (looking-at c-opt-op-identifier-prefix)))
ff959bab
MS
7730 (and (eq (char-before) ?<)
7731 (c-with-syntax-table c++-template-syntax-table
7732 (if (c-safe (goto-char (c-up-list-forward (point))))
7733 t
7734 (goto-char (point-max))
7735 nil)))))
7736 (setq base (point)))
7737
7738 (while (and
7739 (c-syntactic-re-search-forward "[;{=]" nil 'move t t)
7740 (c-end-of-current-token base))
7741 (setq base (point))))))
a66cd3ee
MS
7742
7743(defun c-beginning-of-decl-1 (&optional lim)
7744 ;; Go to the beginning of the current declaration, or the beginning
7745 ;; of the previous one if already at the start of it. Point won't
0386b551 7746 ;; be moved out of any surrounding paren. Return a cons cell of the
a66cd3ee
MS
7747 ;; form (MOVE . KNR-POS). MOVE is like the return value from
7748 ;; `c-beginning-of-statement-1'. If point skipped over some K&R
7749 ;; style argument declarations (and they are to be recognized) then
7750 ;; KNR-POS is set to the start of the first such argument
7751 ;; declaration, otherwise KNR-POS is nil. If LIM is non-nil, it's a
7752 ;; position that bounds the backward search.
7753 ;;
7754 ;; NB: Cases where the declaration continues after the block, as in
7755 ;; "struct foo { ... } bar;", are currently recognized as two
7756 ;; declarations, e.g. "struct foo { ... }" and "bar;" in this case.
0386b551
AM
7757 ;;
7758 ;; This function might do hidden buffer changes.
a66cd3ee
MS
7759 (catch 'return
7760 (let* ((start (point))
d9e94c22 7761 (last-stmt-start (point))
0386b551 7762 (move (c-beginning-of-statement-1 lim nil t)))
a66cd3ee 7763
a66cd3ee
MS
7764 ;; `c-beginning-of-statement-1' stops at a block start, but we
7765 ;; want to continue if the block doesn't begin a top level
2a15eb73
MS
7766 ;; construct, i.e. if it isn't preceded by ';', '}', ':', bob,
7767 ;; or an open paren.
d9e94c22 7768 (let ((beg (point)) tentative-move)
51c9af45
AM
7769 ;; Go back one "statement" each time round the loop until we're just
7770 ;; after a ;, }, or :, or at BOB or the start of a macro or start of
7771 ;; an ObjC method. This will move over a multiple declaration whose
7772 ;; components are comma separated.
d9e94c22
MS
7773 (while (and
7774 ;; Must check with c-opt-method-key in ObjC mode.
7775 (not (and c-opt-method-key
7776 (looking-at c-opt-method-key)))
7777 (/= last-stmt-start (point))
7778 (progn
7779 (c-backward-syntactic-ws lim)
7780 (not (memq (char-before) '(?\; ?} ?: nil))))
2a15eb73
MS
7781 (save-excursion
7782 (backward-char)
7783 (not (looking-at "\\s(")))
d9e94c22
MS
7784 ;; Check that we don't move from the first thing in a
7785 ;; macro to its header.
7786 (not (eq (setq tentative-move
0386b551 7787 (c-beginning-of-statement-1 lim nil t))
d9e94c22
MS
7788 'macro)))
7789 (setq last-stmt-start beg
7790 beg (point)
7791 move tentative-move))
7792 (goto-char beg))
7793
7794 (when c-recognize-knr-p
7795 (let ((fallback-pos (point)) knr-argdecl-start)
7796 ;; Handle K&R argdecls. Back up after the "statement" jumped
7797 ;; over by `c-beginning-of-statement-1', unless it was the
7798 ;; function body, in which case we're sitting on the opening
7799 ;; brace now. Then test if we're in a K&R argdecl region and
7800 ;; that we started at the other side of the first argdecl in
7801 ;; it.
7802 (unless (eq (char-after) ?{)
7803 (goto-char last-stmt-start))
7804 (if (and (setq knr-argdecl-start (c-in-knr-argdecl lim))
7805 (< knr-argdecl-start start)
7806 (progn
7807 (goto-char knr-argdecl-start)
0386b551 7808 (not (eq (c-beginning-of-statement-1 lim nil t) 'macro))))
d9e94c22
MS
7809 (throw 'return
7810 (cons (if (eq (char-after fallback-pos) ?{)
7811 'previous
7812 'same)
7813 knr-argdecl-start))
7814 (goto-char fallback-pos))))
7815
51c9af45
AM
7816 ;; `c-beginning-of-statement-1' counts each brace block as a separate
7817 ;; statement, so the result will be 'previous if we've moved over any.
7818 ;; So change our result back to 'same if necessary.
7819 ;;
7820 ;; If they were brace list initializers we might not have moved over a
7821 ;; declaration boundary though, so change it to 'same if we've moved
7822 ;; past a '=' before '{', but not ';'. (This ought to be integrated
7823 ;; into `c-beginning-of-statement-1', so we avoid this extra pass which
7824 ;; potentially can search over a large amount of text.). Take special
7825 ;; pains not to get mislead by C++'s "operator=", and the like.
d9e94c22
MS
7826 (if (and (eq move 'previous)
7827 (c-with-syntax-table (if (c-major-mode-is 'c++-mode)
7828 c++-template-syntax-table
7829 (syntax-table))
7830 (save-excursion
51c9af45
AM
7831 (and
7832 (progn
7833 (while ; keep going back to "[;={"s until we either find
7834 ; no more, or get to one which isn't an "operator ="
7835 (and (c-syntactic-re-search-forward "[;={]" start t t t)
7836 (eq (char-before) ?=)
7837 c-overloadable-operators-regexp
7838 c-opt-op-identifier-prefix
7839 (save-excursion
7840 (eq (c-backward-token-2) 0)
7841 (looking-at c-overloadable-operators-regexp)
7842 (eq (c-backward-token-2) 0)
7843 (looking-at c-opt-op-identifier-prefix))))
7844 (eq (char-before) ?=))
7845 (c-syntactic-re-search-forward "[;{]" start t t)
7846 (eq (char-before) ?{)
7847 (c-safe (goto-char (c-up-list-forward (point))) t)
7848 (not (c-syntactic-re-search-forward ";" start t t))))))
d9e94c22
MS
7849 (cons 'same nil)
7850 (cons move nil)))))
a66cd3ee
MS
7851
7852(defun c-end-of-decl-1 ()
7853 ;; Assuming point is at the start of a declaration (as detected by
7854 ;; e.g. `c-beginning-of-decl-1'), go to the end of it. Unlike
7855 ;; `c-beginning-of-decl-1', this function handles the case when a
7856 ;; block is followed by identifiers in e.g. struct declarations in C
7857 ;; or C++. If a proper end was found then t is returned, otherwise
7858 ;; point is moved as far as possible within the current sexp and nil
7859 ;; is returned. This function doesn't handle macros; use
7860 ;; `c-end-of-macro' instead in those cases.
0386b551
AM
7861 ;;
7862 ;; This function might do hidden buffer changes.
ce8c7486 7863 (let ((start (point))
a66cd3ee
MS
7864 (decl-syntax-table (if (c-major-mode-is 'c++-mode)
7865 c++-template-syntax-table
7866 (syntax-table))))
7867 (catch 'return
7868 (c-search-decl-header-end)
7869
7870 (when (and c-recognize-knr-p
7871 (eq (char-before) ?\;)
7872 (c-in-knr-argdecl start))
7873 ;; Stopped at the ';' in a K&R argdecl section which is
7874 ;; detected using the same criteria as in
7875 ;; `c-beginning-of-decl-1'. Move to the following block
7876 ;; start.
d9e94c22 7877 (c-syntactic-re-search-forward "{" nil 'move t))
a66cd3ee
MS
7878
7879 (when (eq (char-before) ?{)
7880 ;; Encountered a block in the declaration. Jump over it.
7881 (condition-case nil
7882 (goto-char (c-up-list-forward (point)))
d9e94c22
MS
7883 (error (goto-char (point-max))
7884 (throw 'return nil)))
a66cd3ee
MS
7885 (if (or (not c-opt-block-decls-with-vars-key)
7886 (save-excursion
7887 (c-with-syntax-table decl-syntax-table
7888 (let ((lim (point)))
7889 (goto-char start)
b3cf7e18
MS
7890 (not (and
7891 ;; Check for `c-opt-block-decls-with-vars-key'
7892 ;; before the first paren.
7893 (c-syntactic-re-search-forward
d9e94c22 7894 (concat "[;=\(\[{]\\|\\("
b3cf7e18
MS
7895 c-opt-block-decls-with-vars-key
7896 "\\)")
d9e94c22 7897 lim t t t)
b3cf7e18
MS
7898 (match-beginning 1)
7899 (not (eq (char-before) ?_))
d9e94c22
MS
7900 ;; Check that the first following paren is
7901 ;; the block.
7902 (c-syntactic-re-search-forward "[;=\(\[{]"
7903 lim t t t)
b3cf7e18 7904 (eq (char-before) ?{)))))))
a66cd3ee
MS
7905 ;; The declaration doesn't have any of the
7906 ;; `c-opt-block-decls-with-vars' keywords in the
7907 ;; beginning, so it ends here at the end of the block.
7908 (throw 'return t)))
7909
7910 (c-with-syntax-table decl-syntax-table
7911 (while (progn
7912 (if (eq (char-before) ?\;)
7913 (throw 'return t))
d9e94c22 7914 (c-syntactic-re-search-forward ";" nil 'move t))))
a66cd3ee 7915 nil)))
ce8c7486 7916
0386b551
AM
7917(defun c-looking-at-decl-block (containing-sexp goto-start &optional limit)
7918 ;; Assuming the point is at an open brace, check if it starts a
7919 ;; block that contains another declaration level, i.e. that isn't a
7920 ;; statement block or a brace list, and if so return non-nil.
7921 ;;
7922 ;; If the check is successful, the return value is the start of the
7923 ;; keyword that tells what kind of construct it is, i.e. typically
7924 ;; what `c-decl-block-key' matched. Also, if GOTO-START is set then
7925 ;; the point will be at the start of the construct, before any
7926 ;; leading specifiers, otherwise it's at the returned position.
7927 ;;
7928 ;; The point is clobbered if the check is unsuccessful.
7929 ;;
7930 ;; CONTAINING-SEXP is the position of the open of the surrounding
7931 ;; paren, or nil if none.
7932 ;;
7933 ;; The optional LIMIT limits the backward search for the start of
7934 ;; the construct. It's assumed to be at a syntactically relevant
7935 ;; position.
7936 ;;
7937 ;; If any template arglists are found in the searched region before
7938 ;; the open brace, they get marked with paren syntax.
7939 ;;
7940 ;; This function might do hidden buffer changes.
7941
7942 (let ((open-brace (point)) kwd-start first-specifier-pos)
7943 (c-syntactic-skip-backward c-block-prefix-charset limit t)
7944
7945 (when (and c-recognize-<>-arglists
7946 (eq (char-before) ?>))
7947 ;; Could be at the end of a template arglist.
7948 (let ((c-parse-and-markup-<>-arglists t)
7949 (c-disallow-comma-in-<>-arglists
7950 (and containing-sexp
7951 (not (eq (char-after containing-sexp) ?{)))))
7952 (while (and
7953 (c-backward-<>-arglist nil limit)
7954 (progn
7955 (c-syntactic-skip-backward c-block-prefix-charset limit t)
7956 (eq (char-before) ?>))))))
7957
7958 ;; Note: Can't get bogus hits inside template arglists below since they
7959 ;; have gotten paren syntax above.
7960 (when (and
7961 ;; If `goto-start' is set we begin by searching for the
7962 ;; first possible position of a leading specifier list.
7963 ;; The `c-decl-block-key' search continues from there since
7964 ;; we know it can't match earlier.
7965 (if goto-start
7966 (when (c-syntactic-re-search-forward c-symbol-start
7967 open-brace t t)
7968 (goto-char (setq first-specifier-pos (match-beginning 0)))
7969 t)
7970 t)
7971
7972 (cond
7973 ((c-syntactic-re-search-forward c-decl-block-key open-brace t t t)
7974 (goto-char (setq kwd-start (match-beginning 0)))
7975 (or
7976
7977 ;; Found a keyword that can't be a type?
7978 (match-beginning 1)
7979
7980 ;; Can be a type too, in which case it's the return type of a
7981 ;; function (under the assumption that no declaration level
7982 ;; block construct starts with a type).
7983 (not (c-forward-type))
7984
7985 ;; Jumped over a type, but it could be a declaration keyword
7986 ;; followed by the declared identifier that we've jumped over
7987 ;; instead (e.g. in "class Foo {"). If it indeed is a type
7988 ;; then we should be at the declarator now, so check for a
7989 ;; valid declarator start.
7990 ;;
7991 ;; Note: This doesn't cope with the case when a declared
7992 ;; identifier is followed by e.g. '(' in a language where '('
7993 ;; also might be part of a declarator expression. Currently
7994 ;; there's no such language.
7995 (not (or (looking-at c-symbol-start)
7996 (looking-at c-type-decl-prefix-key)))))
7997
7998 ;; In Pike a list of modifiers may be followed by a brace
7999 ;; to make them apply to many identifiers. Note that the
8000 ;; match data will be empty on return in this case.
8001 ((and (c-major-mode-is 'pike-mode)
8002 (progn
8003 (goto-char open-brace)
8004 (= (c-backward-token-2) 0))
8005 (looking-at c-specifier-key)
8006 ;; Use this variant to avoid yet another special regexp.
8007 (c-keyword-member (c-keyword-sym (match-string 1))
8008 'c-modifier-kwds))
8009 (setq kwd-start (point))
8010 t)))
8011
8012 ;; Got a match.
8013
8014 (if goto-start
8015 ;; Back up over any preceding specifiers and their clauses
8016 ;; by going forward from `first-specifier-pos', which is the
8017 ;; earliest possible position where the specifier list can
8018 ;; start.
8019 (progn
8020 (goto-char first-specifier-pos)
8021
8022 (while (< (point) kwd-start)
8023 (if (looking-at c-symbol-key)
8024 ;; Accept any plain symbol token on the ground that
8025 ;; it's a specifier masked through a macro (just
8026 ;; like `c-forward-decl-or-cast-1' skip forward over
8027 ;; such tokens).
8028 ;;
8029 ;; Could be more restrictive wrt invalid keywords,
8030 ;; but that'd only occur in invalid code so there's
8031 ;; no use spending effort on it.
8032 (let ((end (match-end 0)))
8033 (unless (c-forward-keyword-clause 0)
8034 (goto-char end)
8035 (c-forward-syntactic-ws)))
8036
8037 ;; Can't parse a declaration preamble and is still
8038 ;; before `kwd-start'. That means `first-specifier-pos'
8039 ;; was in some earlier construct. Search again.
8040 (if (c-syntactic-re-search-forward c-symbol-start
8041 kwd-start 'move t)
8042 (goto-char (setq first-specifier-pos (match-beginning 0)))
8043 ;; Got no preamble before the block declaration keyword.
8044 (setq first-specifier-pos kwd-start))))
8045
8046 (goto-char first-specifier-pos))
8047 (goto-char kwd-start))
8048
8049 kwd-start)))
ce8c7486 8050
a66cd3ee 8051(defun c-search-uplist-for-classkey (paren-state)
0386b551
AM
8052 ;; Check if the closest containing paren sexp is a declaration
8053 ;; block, returning a 2 element vector in that case. Aref 0
8054 ;; contains the bufpos at boi of the class key line, and aref 1
8055 ;; contains the bufpos of the open brace. This function is an
8056 ;; obsolete wrapper for `c-looking-at-decl-block'.
8057 ;;
8058 ;; This function might do hidden buffer changes.
8059 (let ((open-paren-pos (c-most-enclosing-brace paren-state)))
8060 (when open-paren-pos
8061 (save-excursion
8062 (goto-char open-paren-pos)
8063 (when (and (eq (char-after) ?{)
8064 (c-looking-at-decl-block
8065 (c-safe-position open-paren-pos paren-state)
8066 nil))
8067 (back-to-indentation)
8068 (vector (point) open-paren-pos))))))
785eecbb 8069
020716e1
AM
8070(defmacro c-pull-open-brace (ps)
8071 ;; Pull the next open brace from PS (which has the form of paren-state),
8072 ;; skipping over any brace pairs. Returns NIL when PS is exhausted.
8073 `(progn
8074 (while (consp (car ,ps))
8075 (setq ,ps (cdr ,ps)))
8076 (prog1 (car ,ps)
8077 (setq ,ps (cdr ,ps)))))
8078
8079(defun c-most-enclosing-decl-block (paren-state)
8080 ;; Return the buffer position of the most enclosing decl-block brace (in the
8081 ;; sense of c-looking-at-decl-block) in the PAREN-STATE structure, or nil if
8082 ;; none was found.
8083 (let* ((open-brace (c-pull-open-brace paren-state))
8084 (next-open-brace (c-pull-open-brace paren-state)))
8085 (while (and open-brace
8086 (save-excursion
8087 (goto-char open-brace)
8088 (not (c-looking-at-decl-block next-open-brace nil))))
8089 (setq open-brace next-open-brace
8090 next-open-brace (c-pull-open-brace paren-state)))
8091 open-brace))
8092
a66cd3ee 8093(defun c-inside-bracelist-p (containing-sexp paren-state)
785eecbb
RS
8094 ;; return the buffer position of the beginning of the brace list
8095 ;; statement if we're inside a brace list, otherwise return nil.
8096 ;; CONTAINING-SEXP is the buffer pos of the innermost containing
0386b551 8097 ;; paren. PAREN-STATE is the remainder of the state of enclosing
130c507e 8098 ;; braces
785eecbb
RS
8099 ;;
8100 ;; N.B.: This algorithm can potentially get confused by cpp macros
0386b551 8101 ;; placed in inconvenient locations. It's a trade-off we make for
785eecbb 8102 ;; speed.
0386b551
AM
8103 ;;
8104 ;; This function might do hidden buffer changes.
785eecbb 8105 (or
d9e94c22 8106 ;; This will pick up brace list declarations.
b2acd789
RS
8107 (c-safe
8108 (save-excursion
8109 (goto-char containing-sexp)
0ec8351b 8110 (c-forward-sexp -1)
b2acd789 8111 (let (bracepos)
d9e94c22 8112 (if (and (or (looking-at c-brace-list-key)
0ec8351b 8113 (progn (c-forward-sexp -1)
d9e94c22 8114 (looking-at c-brace-list-key)))
a66cd3ee 8115 (setq bracepos (c-down-list-forward (point)))
b2acd789
RS
8116 (not (c-crosses-statement-barrier-p (point)
8117 (- bracepos 2))))
8118 (point)))))
785eecbb
RS
8119 ;; this will pick up array/aggregate init lists, even if they are nested.
8120 (save-excursion
0ec8351b
BW
8121 (let ((class-key
8122 ;; Pike can have class definitions anywhere, so we must
8123 ;; check for the class key here.
8124 (and (c-major-mode-is 'pike-mode)
a66cd3ee
MS
8125 c-decl-block-key))
8126 bufpos braceassignp lim next-containing)
785eecbb
RS
8127 (while (and (not bufpos)
8128 containing-sexp)
a66cd3ee
MS
8129 (when paren-state
8130 (if (consp (car paren-state))
8131 (setq lim (cdr (car paren-state))
8132 paren-state (cdr paren-state))
8133 (setq lim (car paren-state)))
8134 (when paren-state
8135 (setq next-containing (car paren-state)
8136 paren-state (cdr paren-state))))
785eecbb 8137 (goto-char containing-sexp)
a66cd3ee
MS
8138 (if (c-looking-at-inexpr-block next-containing next-containing)
8139 ;; We're in an in-expression block of some kind. Do not
8140 ;; check nesting. We deliberately set the limit to the
8141 ;; containing sexp, so that c-looking-at-inexpr-block
8142 ;; doesn't check for an identifier before it.
0ec8351b
BW
8143 (setq containing-sexp nil)
8144 ;; see if the open brace is preceded by = or [...] in
8145 ;; this statement, but watch out for operator=
a66cd3ee 8146 (setq braceassignp 'dontknow)
d9e94c22 8147 (c-backward-token-2 1 t lim)
6393fef2 8148 ;; Checks to do only on the first sexp before the brace.
d9e94c22 8149 (when (and c-opt-inexpr-brace-list-key
6393fef2
RS
8150 (eq (char-after) ?\[))
8151 ;; In Java, an initialization brace list may follow
8152 ;; directly after "new Foo[]", so check for a "new"
8153 ;; earlier.
8154 (while (eq braceassignp 'dontknow)
8155 (setq braceassignp
d9e94c22
MS
8156 (cond ((/= (c-backward-token-2 1 t lim) 0) nil)
8157 ((looking-at c-opt-inexpr-brace-list-key) t)
6393fef2
RS
8158 ((looking-at "\\sw\\|\\s_\\|[.[]")
8159 ;; Carry on looking if this is an
8160 ;; identifier (may contain "." in Java)
8161 ;; or another "[]" sexp.
8162 'dontknow)
8163 (t nil)))))
8164 ;; Checks to do on all sexps before the brace, up to the
8165 ;; beginning of the statement.
8166 (while (eq braceassignp 'dontknow)
0ec8351b
BW
8167 (cond ((eq (char-after) ?\;)
8168 (setq braceassignp nil))
8169 ((and class-key
8170 (looking-at class-key))
8171 (setq braceassignp nil))
8172 ((eq (char-after) ?=)
8173 ;; We've seen a =, but must check earlier tokens so
8174 ;; that it isn't something that should be ignored.
8175 (setq braceassignp 'maybe)
8176 (while (and (eq braceassignp 'maybe)
d9e94c22 8177 (zerop (c-backward-token-2 1 t lim)))
0ec8351b
BW
8178 (setq braceassignp
8179 (cond
8180 ;; Check for operator =
51c9af45
AM
8181 ((and c-opt-op-identifier-prefix
8182 (looking-at c-opt-op-identifier-prefix))
0386b551 8183 nil)
130c507e
GM
8184 ;; Check for `<opchar>= in Pike.
8185 ((and (c-major-mode-is 'pike-mode)
8186 (or (eq (char-after) ?`)
8187 ;; Special case for Pikes
8188 ;; `[]=, since '[' is not in
8189 ;; the punctuation class.
8190 (and (eq (char-after) ?\[)
8191 (eq (char-before) ?`))))
8192 nil)
0ec8351b
BW
8193 ((looking-at "\\s.") 'maybe)
8194 ;; make sure we're not in a C++ template
8195 ;; argument assignment
a66cd3ee
MS
8196 ((and
8197 (c-major-mode-is 'c++-mode)
8198 (save-excursion
8199 (let ((here (point))
8200 (pos< (progn
8201 (skip-chars-backward "^<>")
8202 (point))))
8203 (and (eq (char-before) ?<)
8204 (not (c-crosses-statement-barrier-p
8205 pos< here))
8206 (not (c-in-literal))
8207 ))))
0ec8351b 8208 nil)
6393fef2
RS
8209 (t t))))))
8210 (if (and (eq braceassignp 'dontknow)
d9e94c22 8211 (/= (c-backward-token-2 1 t lim) 0))
6393fef2
RS
8212 (setq braceassignp nil)))
8213 (if (not braceassignp)
0ec8351b
BW
8214 (if (eq (char-after) ?\;)
8215 ;; Brace lists can't contain a semicolon, so we're done.
8216 (setq containing-sexp nil)
a66cd3ee
MS
8217 ;; Go up one level.
8218 (setq containing-sexp next-containing
8219 lim nil
8220 next-containing nil))
0ec8351b
BW
8221 ;; we've hit the beginning of the aggregate list
8222 (c-beginning-of-statement-1
a66cd3ee 8223 (c-most-enclosing-brace paren-state))
0ec8351b 8224 (setq bufpos (point))))
a66cd3ee 8225 )
785eecbb
RS
8226 bufpos))
8227 ))
8228
0ec8351b
BW
8229(defun c-looking-at-special-brace-list (&optional lim)
8230