* progmodes/cc-engine.el (c-forward-decl-or-cast-1): Use
[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 16923.
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
708 ;; `c-nonlabel-token-key'. It's `start' if no valid label
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
712 ;; Symbol just scanned back over (e.g. 'while or 'boundary).
713 ;; See above.
714 sym
715 ;; Current state in the automaton. See above.
716 state
717 ;; Current saved positions. See above.
718 saved-pos
719 ;; Stack of conses (state . saved-pos).
720 stack
721 ;; Regexp which matches "for", "if", etc.
722 (cond-key (or c-opt-block-stmt-key
a66cd3ee 723 "\\<\\>")) ; Matches nothing.
0386b551
AM
724 ;; Return value.
725 (ret 'same)
726 ;; Positions of the last three sexps or bounds we've stopped at.
727 tok ptok pptok)
a66cd3ee
MS
728
729 (save-restriction
730 (if lim (narrow-to-region lim (point-max)))
731
732 (if (save-excursion
733 (and (c-beginning-of-macro)
734 (/= (point) start)))
735 (setq macro-start (point)))
736
d9e94c22 737 ;; Try to skip back over unary operator characters, to register
a66cd3ee
MS
738 ;; that we've moved.
739 (while (progn
740 (setq pos (point))
0386b551
AM
741 (c-backward-syntactic-ws)
742 ;; Protect post-++/-- operators just before a virtual semicolon.
743 (and (not (c-at-vsemi-p))
744 (/= (skip-chars-backward "-+!*&~@`#") 0))))
d9e94c22
MS
745
746 ;; Skip back over any semicolon here. If it was a bare semicolon, we're
0386b551 747 ;; done. Later on we ignore the boundaries for statements that don't
d9e94c22
MS
748 ;; contain any sexp. The only thing that is affected is that the error
749 ;; checking is a little less strict, and we really don't bother.
a66cd3ee
MS
750 (if (and (memq (char-before) delims)
751 (progn (forward-char -1)
752 (setq saved (point))
0386b551 753 (c-backward-syntactic-ws)
a66cd3ee
MS
754 (or (memq (char-before) delims)
755 (memq (char-before) '(?: nil))
d9e94c22 756 (eq (char-syntax (char-before)) ?\()
0386b551 757 (c-at-vsemi-p))))
a66cd3ee
MS
758 (setq ret 'previous
759 pos saved)
760
761 ;; Begin at start and not pos to detect macros if we stand
762 ;; directly after the #.
763 (goto-char start)
764 (if (looking-at "\\<\\|\\W")
765 ;; Record this as the first token if not starting inside it.
766 (setq tok start))
767
d9e94c22
MS
768 ;; The following while loop goes back one sexp (balanced parens,
769 ;; etc. with contents, or symbol or suchlike) each iteration. This
770 ;; movement is accomplished with a call to scan-sexps approx 130 lines
771 ;; below.
a66cd3ee
MS
772 (while
773 (catch 'loop ;; Throw nil to break, non-nil to continue.
774 (cond
a66cd3ee 775 ((save-excursion
0386b551 776 (and macro-start ; Always NIL for AWK.
a66cd3ee
MS
777 (progn (skip-chars-backward " \t")
778 (eq (char-before) ?#))
779 (progn (setq saved (1- (point)))
780 (beginning-of-line)
781 (not (eq (char-before (1- (point))) ?\\)))
d9e94c22 782 (looking-at c-opt-cpp-start)
a66cd3ee
MS
783 (progn (skip-chars-forward " \t")
784 (eq (point) saved))))
785 (goto-char saved)
786 (if (and (c-forward-to-cpp-define-body)
787 (progn (c-forward-syntactic-ws start)
788 (< (point) start)))
789 ;; Stop at the first token in the content of the macro.
790 (setq pos (point)
791 ignore-labels t) ; Avoid the label check on exit.
792 (setq pos saved
793 ret 'macro
794 ignore-labels t))
795 (throw 'loop nil))
796
d9e94c22
MS
797 ;; Do a round through the automaton if we've just passed a
798 ;; statement boundary or passed a "while"-like token.
a66cd3ee
MS
799 ((or sym
800 (and (looking-at cond-key)
801 (setq sym (intern (match-string 1)))))
802
803 (when (and (< pos start) (null stack))
804 (throw 'loop nil))
805
d9e94c22
MS
806 ;; The PDA state handling.
807 ;;
037558bf 808 ;; Refer to the description of the PDA in the opening
d9e94c22
MS
809 ;; comments. In the following OR form, the first leaf
810 ;; attempts to handles one of the specific actions detailed
811 ;; (e.g., finding token "if" whilst in state `else-boundary').
812 ;; We drop through to the second leaf (which handles common
813 ;; state) if no specific handler is found in the first cond.
814 ;; If a parsing error is detected (e.g. an "else" with no
815 ;; preceding "if"), we throw to the enclosing catch.
816 ;;
817 ;; Note that the (eq state 'else) means
818 ;; "we've just passed an else", NOT "we're looking for an
819 ;; else".
a66cd3ee
MS
820 (or (cond
821 ((eq state 'else)
822 (if (eq sym 'boundary)
823 (setq state 'else-boundary)
824 (c-bos-report-error)
825 (c-bos-pop-state-and-retry)))
826
827 ((eq state 'else-boundary)
828 (cond ((eq sym 'if)
829 (c-bos-pop-state (setq ret 'beginning)))
830 ((eq sym 'boundary)
831 (c-bos-report-error)
832 (c-bos-pop-state))))
833
834 ((eq state 'while)
835 (if (and (eq sym 'boundary)
836 ;; Since this can cause backtracking we do a
837 ;; little more careful analysis to avoid it:
838 ;; If there's a label in front of the while
839 ;; it can't be part of a do-while.
840 (not after-labels-pos))
841 (progn (c-bos-save-pos)
842 (setq state 'while-boundary))
d9e94c22 843 (c-bos-pop-state-and-retry))) ; Can't be a do-while
a66cd3ee
MS
844
845 ((eq state 'while-boundary)
846 (cond ((eq sym 'do)
847 (c-bos-pop-state (setq ret 'beginning)))
d9e94c22
MS
848 ((eq sym 'boundary) ; isn't a do-while
849 (c-bos-restore-pos) ; the position of the while
850 (c-bos-pop-state)))) ; no longer searching for do.
a66cd3ee
MS
851
852 ((eq state 'catch)
853 (if (eq sym 'boundary)
854 (setq state 'catch-boundary)
855 (c-bos-report-error)
856 (c-bos-pop-state-and-retry)))
857
858 ((eq state 'catch-boundary)
859 (cond
860 ((eq sym 'try)
861 (c-bos-pop-state (setq ret 'beginning)))
862 ((eq sym 'catch)
863 (setq state 'catch))
864 ((eq sym 'boundary)
865 (c-bos-report-error)
866 (c-bos-pop-state)))))
867
d9e94c22
MS
868 ;; This is state common. We get here when the previous
869 ;; cond statement found no particular state handler.
a66cd3ee 870 (cond ((eq sym 'boundary)
d9e94c22
MS
871 ;; If we have a boundary at the start
872 ;; position we push a frame to go to the
873 ;; previous statement.
874 (if (>= pos start)
875 (c-bos-push-state)
876 (c-bos-pop-state)))
a66cd3ee
MS
877 ((eq sym 'else)
878 (c-bos-push-state)
879 (c-bos-save-error-info 'if 'else)
880 (setq state 'else))
881 ((eq sym 'while)
0386b551
AM
882 ;; Is this a real while, or a do-while?
883 ;; The next `when' triggers unless we are SURE that
884 ;; the `while' is not the tailend of a `do-while'.
a66cd3ee 885 (when (or (not pptok)
d9e94c22 886 (memq (char-after pptok) delims)
0386b551
AM
887 ;; The following kludge is to prevent
888 ;; infinite recursion when called from
889 ;; c-awk-after-if-for-while-condition-p,
890 ;; or the like.
891 (and (eq (point) start)
892 (c-vsemi-status-unknown-p))
893 (c-at-vsemi-p pptok))
a66cd3ee
MS
894 ;; Since this can cause backtracking we do a
895 ;; little more careful analysis to avoid it: If
0386b551
AM
896 ;; the while isn't followed by a (possibly
897 ;; virtual) semicolon it can't be a do-while.
a66cd3ee
MS
898 (c-bos-push-state)
899 (setq state 'while)))
900 ((memq sym '(catch finally))
901 (c-bos-push-state)
902 (c-bos-save-error-info 'try sym)
903 (setq state 'catch))))
904
905 (when c-maybe-labelp
906 ;; We're either past a statement boundary or at the
907 ;; start of a statement, so throw away any label data
908 ;; for the previous one.
909 (setq after-labels-pos nil
910 last-label-pos nil
911 c-maybe-labelp nil))))
912
d9e94c22
MS
913 ;; Step to the previous sexp, but not if we crossed a
914 ;; boundary, since that doesn't consume an sexp.
a66cd3ee
MS
915 (if (eq sym 'boundary)
916 (setq ret 'previous)
d9e94c22
MS
917
918 ;; HERE IS THE SINGLE PLACE INSIDE THE PDA LOOP WHERE WE MOVE
0386b551
AM
919 ;; BACKWARDS THROUGH THE SOURCE.
920
921 ;; This is typically fast with the caching done by
922 ;; c-(backward|forward)-sws.
923 (c-backward-syntactic-ws)
924
925 (let ((before-sws-pos (point))
926 ;; Set as long as we have to continue jumping by sexps.
927 ;; It's the position to use as end in the next round.
928 sexp-loop-continue-pos
929 ;; The end position of the area to search for statement
930 ;; barriers in this round.
931 (sexp-loop-end-pos pos))
932
cb694ab7 933 ;; The following while goes back one sexp per iteration.
0386b551
AM
934 (while
935 (progn
936 (unless (c-safe (c-backward-sexp) t)
937 ;; Give up if we hit an unbalanced block. Since the
938 ;; stack won't be empty the code below will report a
939 ;; suitable error.
a66cd3ee 940 (throw 'loop nil))
0386b551
AM
941
942 ;; Check if the sexp movement crossed a statement or
943 ;; declaration boundary. But first modify the point
944 ;; so that `c-crosses-statement-barrier-p' only looks
945 ;; at the non-sexp chars following the sexp.
946 (save-excursion
947 (when (setq
948 boundary-pos
949 (cond
950 ((if macro-start
951 nil
952 (save-excursion
953 (when (c-beginning-of-macro)
954 ;; Set continuation position in case
955 ;; `c-crosses-statement-barrier-p'
956 ;; doesn't detect anything below.
957 (setq sexp-loop-continue-pos (point)))))
958 ;; If the sexp movement took us into a
959 ;; macro then there were only some non-sexp
960 ;; chars after it. Skip out of the macro
961 ;; to analyze them but not the non-sexp
962 ;; chars that might be inside the macro.
963 (c-end-of-macro)
964 (c-crosses-statement-barrier-p
965 (point) sexp-loop-end-pos))
966
967 ((and
968 (eq (char-after) ?{)
969 (not (c-looking-at-inexpr-block lim nil t)))
970 ;; Passed a block sexp. That's a boundary
971 ;; alright.
972 (point))
973
974 ((looking-at "\\s\(")
975 ;; Passed some other paren. Only analyze
976 ;; the non-sexp chars after it.
977 (goto-char (1+ (c-down-list-backward
978 before-sws-pos)))
979 ;; We're at a valid token start position
980 ;; (outside the `save-excursion') if
981 ;; `c-crosses-statement-barrier-p' failed.
982 (c-crosses-statement-barrier-p
983 (point) sexp-loop-end-pos))
984
985 (t
986 ;; Passed a symbol sexp or line
987 ;; continuation. It doesn't matter that
988 ;; it's included in the analyzed region.
989 (if (c-crosses-statement-barrier-p
990 (point) sexp-loop-end-pos)
991 t
992 ;; If it was a line continuation then we
993 ;; have to continue looping.
994 (if (looking-at "\\\\$")
995 (setq sexp-loop-continue-pos (point)))
996 nil))))
997
998 (setq pptok ptok
999 ptok tok
1000 tok boundary-pos
1001 sym 'boundary)
1002 ;; Like a C "continue". Analyze the next sexp.
1003 (throw 'loop t)))
1004
d28e7f28 1005 sexp-loop-continue-pos) ; End of "go back a sexp" loop condition.
0386b551
AM
1006 (goto-char sexp-loop-continue-pos)
1007 (setq sexp-loop-end-pos sexp-loop-continue-pos
1008 sexp-loop-continue-pos nil))))
a66cd3ee
MS
1009
1010 ;; ObjC method def?
1011 (when (and c-opt-method-key
1012 (setq saved (c-in-method-def-p)))
1013 (setq pos saved
1014 ignore-labels t) ; Avoid the label check on exit.
1015 (throw 'loop nil))
1016
0386b551
AM
1017 ;; Handle labels.
1018 (unless (eq ignore-labels t)
1019 (when (numberp c-maybe-labelp)
cb694ab7
AM
1020 ;; `c-crosses-statement-barrier-p' has found a colon, so we
1021 ;; might be in a label now. Have we got a real label
1022 ;; (including a case label) or something like C++'s "public:"?
d28e7f28
AM
1023 ;; A case label might use an expression rather than a token.
1024 (setq after-case:-pos (or tok start))
1025 (if (looking-at c-nonlabel-token-key) ; e.g. "while" or "'a'"
1026 (setq c-maybe-labelp nil)
1027 (if after-labels-pos ; Have we already encountered a label?
1028 (if (not last-label-pos)
1029 (setq last-label-pos (or tok start)))
1030 (setq after-labels-pos (or tok start)))
1031 (setq c-maybe-labelp t
1032 label-good-pos nil))) ; bogus "label"
cb694ab7
AM
1033
1034 (when (and (not label-good-pos) ; i.e. no invalid "label"'s yet
1035 ; been found.
1036 (looking-at c-nonlabel-token-key)) ; e.g. "while :"
0386b551
AM
1037 ;; We're in a potential label and it's the first
1038 ;; time we've found something that isn't allowed in
1039 ;; one.
1040 (setq label-good-pos (or tok start))))
1041
1042 ;; We've moved back by a sexp, so update the token positions.
a66cd3ee
MS
1043 (setq sym nil
1044 pptok ptok
1045 ptok tok
1046 tok (point)
d9e94c22 1047 pos tok))) ; Not nil (for the while loop).
a66cd3ee
MS
1048
1049 ;; If the stack isn't empty there might be errors to report.
1050 (while stack
1051 (if (and (vectorp saved-pos) (eq (length saved-pos) 3))
1052 (c-bos-report-error))
1053 (setq saved-pos (cdr (car stack))
1054 stack (cdr stack)))
1055
1056 (when (and (eq ret 'same)
1057 (not (memq sym '(boundary ignore nil))))
1058 ;; Need to investigate closer whether we've crossed
1059 ;; between a substatement and its containing statement.
1060 (if (setq saved (if (looking-at c-block-stmt-1-key)
1061 ptok
1062 pptok))
1063 (cond ((> start saved) (setq pos saved))
1064 ((= start saved) (setq ret 'up)))))
1065
0386b551
AM
1066 (when (and (not ignore-labels)
1067 (eq c-maybe-labelp t)
d9e94c22 1068 (not (eq ret 'beginning))
0386b551
AM
1069 after-labels-pos
1070 (or (not label-good-pos)
1071 (<= label-good-pos pos)
1072 (progn
1073 (goto-char (if (and last-label-pos
1074 (< last-label-pos start))
1075 last-label-pos
1076 pos))
1077 (looking-at c-label-kwds-regexp))))
a66cd3ee
MS
1078 ;; We're in a label. Maybe we should step to the statement
1079 ;; after it.
1080 (if (< after-labels-pos start)
1081 (setq pos after-labels-pos)
1082 (setq ret 'label)
0386b551
AM
1083 (if (and last-label-pos (< last-label-pos start))
1084 ;; Might have jumped over several labels. Go to the last one.
a66cd3ee
MS
1085 (setq pos last-label-pos)))))
1086
d28e7f28 1087 ;; Have we got "case <expression>:"?
a66cd3ee 1088 (goto-char pos)
d28e7f28
AM
1089 (when (and after-case:-pos
1090 (not (eq ret 'beginning))
1091 (looking-at c-case-kwds-regexp))
1092 (if (< after-case:-pos start)
0000ee90
AM
1093 (setq pos after-case:-pos))
1094 (if (eq ret 'same)
1095 (setq ret 'label)))
d28e7f28
AM
1096
1097 ;; Skip over the unary operators that can start the statement.
a66cd3ee 1098 (while (progn
0386b551
AM
1099 (c-backward-syntactic-ws)
1100 ;; protect AWK post-inc/decrement operators, etc.
1101 (and (not (c-at-vsemi-p (point)))
1102 (/= (skip-chars-backward "-+!*&~@`#") 0)))
a66cd3ee
MS
1103 (setq pos (point)))
1104 (goto-char pos)
1105 ret)))
785eecbb 1106
785eecbb 1107(defun c-crosses-statement-barrier-p (from to)
a66cd3ee
MS
1108 "Return non-nil if buffer positions FROM to TO cross one or more
1109statement or declaration boundaries. The returned value is actually
d9e94c22
MS
1110the position of the earliest boundary char. FROM must not be within
1111a string or comment.
a66cd3ee
MS
1112
1113The variable `c-maybe-labelp' is set to the position of the first `:' that
1114might start a label (i.e. not part of `::' and not preceded by `?'). If a
0386b551
AM
1115single `?' is found, then `c-maybe-labelp' is cleared.
1116
1117For AWK, a statement which is terminated by an EOL (not a \; or a }) is
1118regarded as having a \"virtual semicolon\" immediately after the last token on
b414f371 1119the line. If this virtual semicolon is _at_ from, the function recognizes it.
0386b551
AM
1120
1121Note that this function might do hidden buffer changes. See the
1122comment at the start of cc-engine.el for more info."
a66cd3ee
MS
1123 (let ((skip-chars c-stmt-delim-chars)
1124 lit-range)
1125 (save-excursion
1126 (catch 'done
1127 (goto-char from)
1128 (while (progn (skip-chars-forward skip-chars to)
785eecbb 1129 (< (point) to))
0386b551
AM
1130 (cond
1131 ((setq lit-range (c-literal-limits from)) ; Have we landed in a string/comment?
1132 (goto-char (cdr lit-range)))
1133 ((eq (char-after) ?:)
1134 (forward-char)
1135 (if (and (eq (char-after) ?:)
1136 (< (point) to))
1137 ;; Ignore scope operators.
1138 (forward-char)
1139 (setq c-maybe-labelp (1- (point)))))
1140 ((eq (char-after) ??)
1141 ;; A question mark. Can't be a label, so stop
1142 ;; looking for more : and ?.
1143 (setq c-maybe-labelp nil
1144 skip-chars (substring c-stmt-delim-chars 0 -2)))
1145 ((memq (char-after) '(?# ?\n ?\r)) ; A virtual semicolon?
1146 (if (and (eq (char-before) ?\\) (memq (char-after) '(?\n ?\r)))
1147 (backward-char))
1148 (skip-chars-backward " \t" from)
1149 (if (c-at-vsemi-p)
1150 (throw 'done (point))
1151 (forward-line)))
1152 (t (throw 'done (point)))))
1153 ;; In trailing space after an as yet undetected virtual semicolon?
1154 (c-backward-syntactic-ws from)
1155 (if (and (< (point) to)
1156 (c-at-vsemi-p))
1157 (point)
1158 nil)))))
1159
1160(defun c-at-statement-start-p ()
1161 "Return non-nil if the point is at the first token in a statement
1162or somewhere in the syntactic whitespace before it.
1163
1164A \"statement\" here is not restricted to those inside code blocks.
1165Any kind of declaration-like construct that occur outside function
1166bodies is also considered a \"statement\".
1167
1168Note that this function might do hidden buffer changes. See the
1169comment at the start of cc-engine.el for more info."
1170
1171 (save-excursion
1172 (let ((end (point))
1173 c-maybe-labelp)
1174 (c-syntactic-skip-backward (substring c-stmt-delim-chars 1) nil t)
1175 (or (bobp)
1176 (eq (char-before) ?})
1177 (and (eq (char-before) ?{)
1178 (not (and c-special-brace-lists
1179 (progn (backward-char)
1180 (c-looking-at-special-brace-list)))))
1181 (c-crosses-statement-barrier-p (point) end)))))
1182
1183(defun c-at-expression-start-p ()
1184 "Return non-nil if the point is at the first token in an expression or
1185statement, or somewhere in the syntactic whitespace before it.
1186
1187An \"expression\" here is a bit different from the normal language
1188grammar sense: It's any sequence of expression tokens except commas,
1189unless they are enclosed inside parentheses of some kind. Also, an
1190expression never continues past an enclosing parenthesis, but it might
1191contain parenthesis pairs of any sort except braces.
1192
1193Since expressions never cross statement boundaries, this function also
1194recognizes statement beginnings, just like `c-at-statement-start-p'.
1195
1196Note that this function might do hidden buffer changes. See the
1197comment at the start of cc-engine.el for more info."
1198
1199 (save-excursion
1200 (let ((end (point))
1201 (c-stmt-delim-chars c-stmt-delim-chars-with-comma)
1202 c-maybe-labelp)
1203 (c-syntactic-skip-backward (substring c-stmt-delim-chars 1) nil t)
1204 (or (bobp)
1205 (memq (char-before) '(?{ ?}))
1206 (save-excursion (backward-char)
1207 (looking-at "\\s("))
1208 (c-crosses-statement-barrier-p (point) end)))))
785eecbb
RS
1209
1210\f
d9e94c22
MS
1211;; A set of functions that covers various idiosyncrasies in
1212;; implementations of `forward-comment'.
1213
1214;; Note: Some emacsen considers incorrectly that any line comment
1215;; ending with a backslash continues to the next line. I can't think
1216;; of any way to work around that in a reliable way without changing
1217;; the buffer, though. Suggestions welcome. ;) (No, temporarily
1218;; changing the syntax for backslash doesn't work since we must treat
1219;; escapes in string literals correctly.)
1220
1221(defun c-forward-single-comment ()
1222 "Move forward past whitespace and the closest following comment, if any.
1223Return t if a comment was found, nil otherwise. In either case, the
1224point is moved past the following whitespace. Line continuations,
1225i.e. a backslashes followed by line breaks, are treated as whitespace.
1226The line breaks that end line comments are considered to be the
1227comment enders, so the point will be put on the beginning of the next
1228line if it moved past a line comment.
1229
1230This function does not do any hidden buffer changes."
1231
1232 (let ((start (point)))
1233 (when (looking-at "\\([ \t\n\r\f\v]\\|\\\\[\n\r]\\)+")
1234 (goto-char (match-end 0)))
1235
1236 (when (forward-comment 1)
1237 (if (eobp)
1238 ;; Some emacsen (e.g. XEmacs 21) return t when moving
1239 ;; forwards at eob.
1240 nil
1241
1242 ;; Emacs includes the ending newline in a b-style (c++)
1243 ;; comment, but XEmacs doesn't. We depend on the Emacs
1244 ;; behavior (which also is symmetric).
1245 (if (and (eolp) (elt (parse-partial-sexp start (point)) 7))
1246 (condition-case nil (forward-char 1)))
1247
1248 t))))
1249
1250(defsubst c-forward-comments ()
1251 "Move forward past all following whitespace and comments.
1252Line continuations, i.e. a backslashes followed by line breaks, are
1253treated as whitespace.
1254
0386b551
AM
1255Note that this function might do hidden buffer changes. See the
1256comment at the start of cc-engine.el for more info."
d9e94c22
MS
1257
1258 (while (or
1259 ;; If forward-comment in at least XEmacs 21 is given a large
1260 ;; positive value, it'll loop all the way through if it hits
1261 ;; eob.
1262 (and (forward-comment 5)
1263 ;; Some emacsen (e.g. XEmacs 21) return t when moving
1264 ;; forwards at eob.
1265 (not (eobp)))
1266
1267 (when (looking-at "\\\\[\n\r]")
1268 (forward-char 2)
1269 t))))
1270
1271(defun c-backward-single-comment ()
1272 "Move backward past whitespace and the closest preceding comment, if any.
1273Return t if a comment was found, nil otherwise. In either case, the
1274point is moved past the preceding whitespace. Line continuations,
1275i.e. a backslashes followed by line breaks, are treated as whitespace.
1276The line breaks that end line comments are considered to be the
1277comment enders, so the point cannot be at the end of the same line to
1278move over a line comment.
1279
1280This function does not do any hidden buffer changes."
1281
1282 (let ((start (point)))
1283 ;; When we got newline terminated comments, forward-comment in all
1284 ;; supported emacsen so far will stop at eol of each line not
1285 ;; ending with a comment when moving backwards. This corrects for
1286 ;; that, and at the same time handles line continuations.
1287 (while (progn
1288 (skip-chars-backward " \t\n\r\f\v")
1289 (and (looking-at "[\n\r]")
0386b551 1290 (eq (char-before) ?\\)))
d9e94c22
MS
1291 (backward-char))
1292
1293 (if (bobp)
1294 ;; Some emacsen (e.g. Emacs 19.34) return t when moving
1295 ;; backwards at bob.
1296 nil
1297
1298 ;; Leave point after the closest following newline if we've
1299 ;; backed up over any above, since forward-comment won't move
1300 ;; backward over a line comment if point is at the end of the
1301 ;; same line.
1302 (re-search-forward "\\=\\s *[\n\r]" start t)
1303
1304 (if (if (forward-comment -1)
1305 (if (eolp)
1306 ;; If forward-comment above succeeded and we're at eol
1307 ;; then the newline we moved over above didn't end a
1308 ;; line comment, so we give it another go.
1309 (forward-comment -1)
1310 t))
1311
1312 ;; Emacs <= 20 and XEmacs move back over the closer of a
1313 ;; block comment that lacks an opener.
1314 (if (looking-at "\\*/")
1315 (progn (forward-char 2) nil)
1316 t)))))
1317
1318(defsubst c-backward-comments ()
1319 "Move backward past all preceding whitespace and comments.
1320Line continuations, i.e. a backslashes followed by line breaks, are
1321treated as whitespace. The line breaks that end line comments are
1322considered to be the comment enders, so the point cannot be at the end
0386b551
AM
1323of the same line to move over a line comment. Unlike
1324c-backward-syntactic-ws, this function doesn't move back over
1325preprocessor directives.
d9e94c22 1326
0386b551
AM
1327Note that this function might do hidden buffer changes. See the
1328comment at the start of cc-engine.el for more info."
d9e94c22
MS
1329
1330 (let ((start (point)))
1331 (while (and
0386b551 1332 ;; `forward-comment' in some emacsen (e.g. XEmacs 21.4)
d9e94c22
MS
1333 ;; return t when moving backwards at bob.
1334 (not (bobp))
1335
1336 (if (forward-comment -1)
1337 (if (looking-at "\\*/")
1338 ;; Emacs <= 20 and XEmacs move back over the
1339 ;; closer of a block comment that lacks an opener.
1340 (progn (forward-char 2) nil)
1341 t)
1342
1343 ;; XEmacs treats line continuations as whitespace but
1344 ;; only in the backward direction, which seems a bit
1345 ;; odd. Anyway, this is necessary for Emacs.
1346 (when (and (looking-at "[\n\r]")
1347 (eq (char-before) ?\\)
1348 (< (point) start))
1349 (backward-char)
1350 t))))))
1351
1352\f
d9e94c22 1353;; Tools for skipping over syntactic whitespace.
a66cd3ee 1354
d9e94c22
MS
1355;; The following functions use text properties to cache searches over
1356;; large regions of syntactic whitespace. It works as follows:
1357;;
1358;; o If a syntactic whitespace region contains anything but simple
1359;; whitespace (i.e. space, tab and line breaks), the text property
1360;; `c-in-sws' is put over it. At places where we have stopped
1361;; within that region there's also a `c-is-sws' text property.
1362;; That since there typically are nested whitespace inside that
1363;; must be handled separately, e.g. whitespace inside a comment or
1364;; cpp directive. Thus, from one point with `c-is-sws' it's safe
1365;; to jump to another point with that property within the same
1366;; `c-in-sws' region. It can be likened to a ladder where
1367;; `c-in-sws' marks the bars and `c-is-sws' the rungs.
1368;;
1369;; o The `c-is-sws' property is put on the simple whitespace chars at
1370;; a "rung position" and also maybe on the first following char.
1371;; As many characters as can be conveniently found in this range
1372;; are marked, but no assumption can be made that the whole range
1373;; is marked (it could be clobbered by later changes, for
1374;; instance).
1375;;
1376;; Note that some part of the beginning of a sequence of simple
1377;; whitespace might be part of the end of a preceding line comment
1378;; or cpp directive and must not be considered part of the "rung".
1379;; Such whitespace is some amount of horizontal whitespace followed
1380;; by a newline. In the case of cpp directives it could also be
1381;; two newlines with horizontal whitespace between them.
1382;;
1383;; The reason to include the first following char is to cope with
1384;; "rung positions" that doesn't have any ordinary whitespace. If
1385;; `c-is-sws' is put on a token character it does not have
1386;; `c-in-sws' set simultaneously. That's the only case when that
1387;; can occur, and the reason for not extending the `c-in-sws'
1388;; region to cover it is that the `c-in-sws' region could then be
1389;; accidentally merged with a following one if the token is only
1390;; one character long.
1391;;
1392;; o On buffer changes the `c-in-sws' and `c-is-sws' properties are
1393;; removed in the changed region. If the change was inside
1394;; syntactic whitespace that means that the "ladder" is broken, but
1395;; a later call to `c-forward-sws' or `c-backward-sws' will use the
1396;; parts on either side and use an ordinary search only to "repair"
1397;; the gap.
1398;;
1399;; Special care needs to be taken if a region is removed: If there
1400;; are `c-in-sws' on both sides of it which do not connect inside
1401;; the region then they can't be joined. If e.g. a marked macro is
1402;; broken, syntactic whitespace inside the new text might be
1403;; marked. If those marks would become connected with the old
1404;; `c-in-sws' range around the macro then we could get a ladder
1405;; with one end outside the macro and the other at some whitespace
1406;; within it.
1407;;
1408;; The main motivation for this system is to increase the speed in
1409;; skipping over the large whitespace regions that can occur at the
1410;; top level in e.g. header files that contain a lot of comments and
1411;; cpp directives. For small comments inside code it's probably
1412;; slower than using `forward-comment' straightforwardly, but speed is
1413;; not a significant factor there anyway.
1414
1415; (defface c-debug-is-sws-face
1416; '((t (:background "GreenYellow")))
1417; "Debug face to mark the `c-is-sws' property.")
1418; (defface c-debug-in-sws-face
1419; '((t (:underline t)))
1420; "Debug face to mark the `c-in-sws' property.")
1421
1422; (defun c-debug-put-sws-faces ()
1423; ;; Put the sws debug faces on all the `c-is-sws' and `c-in-sws'
1424; ;; properties in the buffer.
1425; (interactive)
1426; (save-excursion
0386b551 1427; (c-save-buffer-state (in-face)
d9e94c22
MS
1428; (goto-char (point-min))
1429; (setq in-face (if (get-text-property (point) 'c-is-sws)
1430; (point)))
1431; (while (progn
1432; (goto-char (next-single-property-change
1433; (point) 'c-is-sws nil (point-max)))
1434; (if in-face
1435; (progn
1436; (c-debug-add-face in-face (point) 'c-debug-is-sws-face)
1437; (setq in-face nil))
1438; (setq in-face (point)))
1439; (not (eobp))))
1440; (goto-char (point-min))
1441; (setq in-face (if (get-text-property (point) 'c-in-sws)
1442; (point)))
1443; (while (progn
1444; (goto-char (next-single-property-change
1445; (point) 'c-in-sws nil (point-max)))
1446; (if in-face
1447; (progn
1448; (c-debug-add-face in-face (point) 'c-debug-in-sws-face)
1449; (setq in-face nil))
1450; (setq in-face (point)))
1451; (not (eobp)))))))
1452
1453(defmacro c-debug-sws-msg (&rest args)
1454 ;;`(message ,@args)
1455 )
1456
1457(defmacro c-put-is-sws (beg end)
0386b551 1458 ;; This macro does a hidden buffer change.
d9e94c22
MS
1459 `(let ((beg ,beg) (end ,end))
1460 (put-text-property beg end 'c-is-sws t)
1461 ,@(when (facep 'c-debug-is-sws-face)
1462 `((c-debug-add-face beg end 'c-debug-is-sws-face)))))
1463
1464(defmacro c-put-in-sws (beg end)
0386b551 1465 ;; This macro does a hidden buffer change.
d9e94c22
MS
1466 `(let ((beg ,beg) (end ,end))
1467 (put-text-property beg end 'c-in-sws t)
1468 ,@(when (facep 'c-debug-is-sws-face)
1469 `((c-debug-add-face beg end 'c-debug-in-sws-face)))))
1470
1471(defmacro c-remove-is-sws (beg end)
0386b551 1472 ;; This macro does a hidden buffer change.
d9e94c22
MS
1473 `(let ((beg ,beg) (end ,end))
1474 (remove-text-properties beg end '(c-is-sws nil))
1475 ,@(when (facep 'c-debug-is-sws-face)
1476 `((c-debug-remove-face beg end 'c-debug-is-sws-face)))))
1477
1478(defmacro c-remove-in-sws (beg end)
0386b551 1479 ;; This macro does a hidden buffer change.
d9e94c22
MS
1480 `(let ((beg ,beg) (end ,end))
1481 (remove-text-properties beg end '(c-in-sws nil))
1482 ,@(when (facep 'c-debug-is-sws-face)
1483 `((c-debug-remove-face beg end 'c-debug-in-sws-face)))))
1484
1485(defmacro c-remove-is-and-in-sws (beg end)
0386b551 1486 ;; This macro does a hidden buffer change.
d9e94c22
MS
1487 `(let ((beg ,beg) (end ,end))
1488 (remove-text-properties beg end '(c-is-sws nil c-in-sws nil))
1489 ,@(when (facep 'c-debug-is-sws-face)
1490 `((c-debug-remove-face beg end 'c-debug-is-sws-face)
1491 (c-debug-remove-face beg end 'c-debug-in-sws-face)))))
1492
1493(defsubst c-invalidate-sws-region-after (beg end)
1494 ;; Called from `after-change-functions'. Note that if
1495 ;; `c-forward-sws' or `c-backward-sws' are used outside
1496 ;; `c-save-buffer-state' or similar then this will remove the cache
1497 ;; properties right after they're added.
0386b551
AM
1498 ;;
1499 ;; This function does hidden buffer changes.
d9e94c22
MS
1500
1501 (save-excursion
1502 ;; Adjust the end to remove the properties in any following simple
1503 ;; ws up to and including the next line break, if there is any
1504 ;; after the changed region. This is necessary e.g. when a rung
1505 ;; marked empty line is converted to a line comment by inserting
1506 ;; "//" before the line break. In that case the line break would
1507 ;; keep the rung mark which could make a later `c-backward-sws'
1508 ;; move into the line comment instead of over it.
1509 (goto-char end)
1510 (skip-chars-forward " \t\f\v")
1511 (when (and (eolp) (not (eobp)))
1512 (setq end (1+ (point)))))
1513
1514 (when (and (= beg end)
1515 (get-text-property beg 'c-in-sws)
2a15eb73 1516 (> beg (point-min))
d9e94c22
MS
1517 (get-text-property (1- beg) 'c-in-sws))
1518 ;; Ensure that an `c-in-sws' range gets broken. Note that it isn't
1519 ;; safe to keep a range that was continuous before the change. E.g:
1520 ;;
1521 ;; #define foo
1522 ;; \
1523 ;; bar
1524 ;;
1525 ;; There can be a "ladder" between "#" and "b". Now, if the newline
1526 ;; after "foo" is removed then "bar" will become part of the cpp
1527 ;; directive instead of a syntactically relevant token. In that
1528 ;; case there's no longer syntactic ws from "#" to "b".
1529 (setq beg (1- beg)))
1530
1531 (c-debug-sws-msg "c-invalidate-sws-region-after [%s..%s]" beg end)
1532 (c-remove-is-and-in-sws beg end))
1533
1534(defun c-forward-sws ()
1535 ;; Used by `c-forward-syntactic-ws' to implement the unbounded search.
0386b551
AM
1536 ;;
1537 ;; This function might do hidden buffer changes.
d9e94c22
MS
1538
1539 (let (;; `rung-pos' is set to a position as early as possible in the
1540 ;; unmarked part of the simple ws region.
1541 (rung-pos (point)) next-rung-pos rung-end-pos last-put-in-sws-pos
1542 rung-is-marked next-rung-is-marked simple-ws-end
1543 ;; `safe-start' is set when it's safe to cache the start position.
1544 ;; It's not set if we've initially skipped over comments and line
1545 ;; continuations since we might have gone out through the end of a
1546 ;; macro then. This provision makes `c-forward-sws' not populate the
1547 ;; cache in the majority of cases, but otoh is `c-backward-sws' by far
1548 ;; more common.
1549 safe-start)
1550
1551 ;; Skip simple ws and do a quick check on the following character to see
1552 ;; if it's anything that can't start syntactic ws, so we can bail out
1553 ;; early in the majority of cases when there just are a few ws chars.
1554 (skip-chars-forward " \t\n\r\f\v")
1555 (when (looking-at c-syntactic-ws-start)
1556
1557 (setq rung-end-pos (min (1+ (point)) (point-max)))
1558 (if (setq rung-is-marked (text-property-any rung-pos rung-end-pos
1559 'c-is-sws t))
1560 ;; Find the last rung position to avoid setting properties in all
1561 ;; the cases when the marked rung is complete.
1562 ;; (`next-single-property-change' is certain to move at least one
1563 ;; step forward.)
1564 (setq rung-pos (1- (next-single-property-change
1565 rung-is-marked 'c-is-sws nil rung-end-pos)))
1566 ;; Got no marked rung here. Since the simple ws might have started
1567 ;; inside a line comment or cpp directive we must set `rung-pos' as
1568 ;; high as possible.
1569 (setq rung-pos (point)))
1570
1571 (while
1572 (progn
1573 (while
1574 (when (and rung-is-marked
1575 (get-text-property (point) 'c-in-sws))
1576
1577 ;; The following search is the main reason that `c-in-sws'
1578 ;; and `c-is-sws' aren't combined to one property.
1579 (goto-char (next-single-property-change
1580 (point) 'c-in-sws nil (point-max)))
1581 (unless (get-text-property (point) 'c-is-sws)
1582 ;; If the `c-in-sws' region extended past the last
1583 ;; `c-is-sws' char we have to go back a bit.
1584 (or (get-text-property (1- (point)) 'c-is-sws)
1585 (goto-char (previous-single-property-change
1586 (point) 'c-is-sws)))
1587 (backward-char))
1588
1589 (c-debug-sws-msg
1590 "c-forward-sws cached move %s -> %s (max %s)"
1591 rung-pos (point) (point-max))
1592
1593 (setq rung-pos (point))
1594 (and (> (skip-chars-forward " \t\n\r\f\v") 0)
1595 (not (eobp))))
1596
1597 ;; We'll loop here if there is simple ws after the last rung.
1598 ;; That means that there's been some change in it and it's
1599 ;; possible that we've stepped into another ladder, so extend
1600 ;; the previous one to join with it if there is one, and try to
1601 ;; use the cache again.
1602 (c-debug-sws-msg
1603 "c-forward-sws extending rung with [%s..%s] (max %s)"
1604 (1+ rung-pos) (1+ (point)) (point-max))
1605 (unless (get-text-property (point) 'c-is-sws)
1606 ;; Remove any `c-in-sws' property from the last char of
1607 ;; the rung before we mark it with `c-is-sws', so that we
1608 ;; won't connect with the remains of a broken "ladder".
1609 (c-remove-in-sws (point) (1+ (point))))
1610 (c-put-is-sws (1+ rung-pos)
1611 (1+ (point)))
1612 (c-put-in-sws rung-pos
1613 (setq rung-pos (point)
1614 last-put-in-sws-pos rung-pos)))
1615
1616 (setq simple-ws-end (point))
1617 (c-forward-comments)
1618
1619 (cond
1620 ((/= (point) simple-ws-end)
1621 ;; Skipped over comments. Don't cache at eob in case the buffer
1622 ;; is narrowed.
1623 (not (eobp)))
1624
1625 ((save-excursion
1626 (and c-opt-cpp-prefix
1627 (looking-at c-opt-cpp-start)
1628 (progn (skip-chars-backward " \t")
1629 (bolp))
1630 (or (bobp)
1631 (progn (backward-char)
1632 (not (eq (char-before) ?\\))))))
1633 ;; Skip a preprocessor directive.
1634 (end-of-line)
1635 (while (and (eq (char-before) ?\\)
1636 (= (forward-line 1) 0))
1637 (end-of-line))
1638 (forward-line 1)
1639 (setq safe-start t)
1640 ;; Don't cache at eob in case the buffer is narrowed.
1641 (not (eobp)))))
1642
1643 ;; We've searched over a piece of non-white syntactic ws. See if this
1644 ;; can be cached.
1645 (setq next-rung-pos (point))
1646 (skip-chars-forward " \t\n\r\f\v")
1647 (setq rung-end-pos (min (1+ (point)) (point-max)))
1648
1649 (if (or
1650 ;; Cache if we haven't skipped comments only, and if we started
1651 ;; either from a marked rung or from a completely uncached
1652 ;; position.
1653 (and safe-start
1654 (or rung-is-marked
1655 (not (get-text-property simple-ws-end 'c-in-sws))))
1656
1657 ;; See if there's a marked rung in the encountered simple ws. If
1658 ;; so then we can cache, unless `safe-start' is nil. Even then
1659 ;; we need to do this to check if the cache can be used for the
1660 ;; next step.
1661 (and (setq next-rung-is-marked
1662 (text-property-any next-rung-pos rung-end-pos
1663 'c-is-sws t))
1664 safe-start))
b2acd789 1665
0ec8351b 1666 (progn
d9e94c22
MS
1667 (c-debug-sws-msg
1668 "c-forward-sws caching [%s..%s] - [%s..%s] (max %s)"
1669 rung-pos (1+ simple-ws-end) next-rung-pos rung-end-pos
1670 (point-max))
1671
1672 ;; Remove the properties for any nested ws that might be cached.
1673 ;; Only necessary for `c-is-sws' since `c-in-sws' will be set
1674 ;; anyway.
1675 (c-remove-is-sws (1+ simple-ws-end) next-rung-pos)
1676 (unless (and rung-is-marked (= rung-pos simple-ws-end))
1677 (c-put-is-sws rung-pos
1678 (1+ simple-ws-end))
1679 (setq rung-is-marked t))
1680 (c-put-in-sws rung-pos
1681 (setq rung-pos (point)
1682 last-put-in-sws-pos rung-pos))
1683 (unless (get-text-property (1- rung-end-pos) 'c-is-sws)
1684 ;; Remove any `c-in-sws' property from the last char of
1685 ;; the rung before we mark it with `c-is-sws', so that we
1686 ;; won't connect with the remains of a broken "ladder".
1687 (c-remove-in-sws (1- rung-end-pos) rung-end-pos))
1688 (c-put-is-sws next-rung-pos
1689 rung-end-pos))
1690
1691 (c-debug-sws-msg
1692 "c-forward-sws not caching [%s..%s] - [%s..%s] (max %s)"
1693 rung-pos (1+ simple-ws-end) next-rung-pos rung-end-pos
1694 (point-max))
1695
1696 ;; Set `rung-pos' for the next rung. It's the same thing here as
1697 ;; initially, except that the rung position is set as early as
1698 ;; possible since we can't be in the ending ws of a line comment or
1699 ;; cpp directive now.
1700 (if (setq rung-is-marked next-rung-is-marked)
1701 (setq rung-pos (1- (next-single-property-change
1702 rung-is-marked 'c-is-sws nil rung-end-pos)))
1703 (setq rung-pos next-rung-pos))
1704 (setq safe-start t)))
1705
1706 ;; Make sure that the newly marked `c-in-sws' region doesn't connect to
1707 ;; another one after the point (which might occur when editing inside a
1708 ;; comment or macro).
1709 (when (eq last-put-in-sws-pos (point))
1710 (cond ((< last-put-in-sws-pos (point-max))
1711 (c-debug-sws-msg
1712 "c-forward-sws clearing at %s for cache separation"
1713 last-put-in-sws-pos)
1714 (c-remove-in-sws last-put-in-sws-pos
1715 (1+ last-put-in-sws-pos)))
1716 (t
1717 ;; If at eob we have to clear the last character before the end
1718 ;; instead since the buffer might be narrowed and there might
1719 ;; be a `c-in-sws' after (point-max). In this case it's
1720 ;; necessary to clear both properties.
1721 (c-debug-sws-msg
1722 "c-forward-sws clearing thoroughly at %s for cache separation"
1723 (1- last-put-in-sws-pos))
1724 (c-remove-is-and-in-sws (1- last-put-in-sws-pos)
1725 last-put-in-sws-pos))))
1726 )))
b2acd789 1727
d9e94c22
MS
1728(defun c-backward-sws ()
1729 ;; Used by `c-backward-syntactic-ws' to implement the unbounded search.
0386b551
AM
1730 ;;
1731 ;; This function might do hidden buffer changes.
d9e94c22
MS
1732
1733 (let (;; `rung-pos' is set to a position as late as possible in the unmarked
1734 ;; part of the simple ws region.
1735 (rung-pos (point)) next-rung-pos last-put-in-sws-pos
1736 rung-is-marked simple-ws-beg cmt-skip-pos)
1737
1738 ;; Skip simple horizontal ws and do a quick check on the preceding
1739 ;; character to see if it's anying that can't end syntactic ws, so we can
1740 ;; bail out early in the majority of cases when there just are a few ws
1741 ;; chars. Newlines are complicated in the backward direction, so we can't
1742 ;; skip over them.
1743 (skip-chars-backward " \t\f")
1744 (when (and (not (bobp))
1745 (save-excursion
1746 (backward-char)
1747 (looking-at c-syntactic-ws-end)))
1748
1749 ;; Try to find a rung position in the simple ws preceding point, so that
1750 ;; we can get a cache hit even if the last bit of the simple ws has
1751 ;; changed recently.
1752 (setq simple-ws-beg (point))
1753 (skip-chars-backward " \t\n\r\f\v")
1754 (if (setq rung-is-marked (text-property-any
1755 (point) (min (1+ rung-pos) (point-max))
1756 'c-is-sws t))
1757 ;; `rung-pos' will be the earliest marked position, which means that
1758 ;; there might be later unmarked parts in the simple ws region.
1759 ;; It's not worth the effort to fix that; the last part of the
1760 ;; simple ws is also typically edited often, so it could be wasted.
1761 (goto-char (setq rung-pos rung-is-marked))
1762 (goto-char simple-ws-beg))
1763
1764 (while
1765 (progn
1766 (while
1767 (when (and rung-is-marked
1768 (not (bobp))
1769 (get-text-property (1- (point)) 'c-in-sws))
1770
1771 ;; The following search is the main reason that `c-in-sws'
1772 ;; and `c-is-sws' aren't combined to one property.
1773 (goto-char (previous-single-property-change
1774 (point) 'c-in-sws nil (point-min)))
1775 (unless (get-text-property (point) 'c-is-sws)
1776 ;; If the `c-in-sws' region extended past the first
1777 ;; `c-is-sws' char we have to go forward a bit.
1778 (goto-char (next-single-property-change
1779 (point) 'c-is-sws)))
1780
1781 (c-debug-sws-msg
1782 "c-backward-sws cached move %s <- %s (min %s)"
1783 (point) rung-pos (point-min))
1784
1785 (setq rung-pos (point))
1786 (if (and (< (min (skip-chars-backward " \t\f\v")
1787 (progn
1788 (setq simple-ws-beg (point))
1789 (skip-chars-backward " \t\n\r\f\v")))
1790 0)
1791 (setq rung-is-marked
1792 (text-property-any (point) rung-pos
1793 'c-is-sws t)))
1794 t
1795 (goto-char simple-ws-beg)
1796 nil))
1797
1798 ;; We'll loop here if there is simple ws before the first rung.
1799 ;; That means that there's been some change in it and it's
1800 ;; possible that we've stepped into another ladder, so extend
1801 ;; the previous one to join with it if there is one, and try to
1802 ;; use the cache again.
1803 (c-debug-sws-msg
1804 "c-backward-sws extending rung with [%s..%s] (min %s)"
1805 rung-is-marked rung-pos (point-min))
1806 (unless (get-text-property (1- rung-pos) 'c-is-sws)
1807 ;; Remove any `c-in-sws' property from the last char of
1808 ;; the rung before we mark it with `c-is-sws', so that we
1809 ;; won't connect with the remains of a broken "ladder".
1810 (c-remove-in-sws (1- rung-pos) rung-pos))
1811 (c-put-is-sws rung-is-marked
1812 rung-pos)
1813 (c-put-in-sws rung-is-marked
1814 (1- rung-pos))
1815 (setq rung-pos rung-is-marked
1816 last-put-in-sws-pos rung-pos))
1817
1818 (c-backward-comments)
1819 (setq cmt-skip-pos (point))
a66cd3ee 1820
d9e94c22
MS
1821 (cond
1822 ((and c-opt-cpp-prefix
1823 (/= cmt-skip-pos simple-ws-beg)
1824 (c-beginning-of-macro))
1825 ;; Inside a cpp directive. See if it should be skipped over.
1826 (let ((cpp-beg (point)))
1827
1828 ;; Move back over all line continuations in the region skipped
1829 ;; over by `c-backward-comments'. If we go past it then we
1830 ;; started inside the cpp directive.
1831 (goto-char simple-ws-beg)
1832 (beginning-of-line)
1833 (while (and (> (point) cmt-skip-pos)
1834 (progn (backward-char)
1835 (eq (char-before) ?\\)))
1836 (beginning-of-line))
1837
1838 (if (< (point) cmt-skip-pos)
1839 ;; Don't move past the cpp directive if we began inside
1840 ;; it. Note that the position at the end of the last line
1841 ;; of the macro is also considered to be within it.
1842 (progn (goto-char cmt-skip-pos)
1843 nil)
1844
1845 ;; It's worthwhile to spend a little bit of effort on finding
1846 ;; the end of the macro, to get a good `simple-ws-beg'
1847 ;; position for the cache. Note that `c-backward-comments'
1848 ;; could have stepped over some comments before going into
1849 ;; the macro, and then `simple-ws-beg' must be kept on the
1850 ;; same side of those comments.
1851 (goto-char simple-ws-beg)
1852 (skip-chars-backward " \t\n\r\f\v")
1853 (if (eq (char-before) ?\\)
1854 (forward-char))
1855 (forward-line 1)
1856 (if (< (point) simple-ws-beg)
1857 ;; Might happen if comments after the macro were skipped
1858 ;; over.
1859 (setq simple-ws-beg (point)))
1860
1861 (goto-char cpp-beg)
1862 t)))
1863
1864 ((/= (save-excursion
1865 (skip-chars-forward " \t\n\r\f\v" simple-ws-beg)
1866 (setq next-rung-pos (point)))
1867 simple-ws-beg)
1868 ;; Skipped over comments. Must put point at the end of
1869 ;; the simple ws at point since we might be after a line
1870 ;; comment or cpp directive that's been partially
1871 ;; narrowed out, and we can't risk marking the simple ws
1872 ;; at the end of it.
1873 (goto-char next-rung-pos)
1874 t)))
1875
1876 ;; We've searched over a piece of non-white syntactic ws. See if this
1877 ;; can be cached.
1878 (setq next-rung-pos (point))
1879 (skip-chars-backward " \t\f\v")
1880
1881 (if (or
1882 ;; Cache if we started either from a marked rung or from a
1883 ;; completely uncached position.
1884 rung-is-marked
1885 (not (get-text-property (1- simple-ws-beg) 'c-in-sws))
1886
1887 ;; Cache if there's a marked rung in the encountered simple ws.
1888 (save-excursion
1889 (skip-chars-backward " \t\n\r\f\v")
1890 (text-property-any (point) (min (1+ next-rung-pos) (point-max))
1891 'c-is-sws t)))
a66cd3ee 1892
d9e94c22
MS
1893 (progn
1894 (c-debug-sws-msg
1895 "c-backward-sws caching [%s..%s] - [%s..%s] (min %s)"
1896 (point) (1+ next-rung-pos)
1897 simple-ws-beg (min (1+ rung-pos) (point-max))
1898 (point-min))
1899
1900 ;; Remove the properties for any nested ws that might be cached.
1901 ;; Only necessary for `c-is-sws' since `c-in-sws' will be set
1902 ;; anyway.
1903 (c-remove-is-sws (1+ next-rung-pos) simple-ws-beg)
1904 (unless (and rung-is-marked (= simple-ws-beg rung-pos))
1905 (let ((rung-end-pos (min (1+ rung-pos) (point-max))))
1906 (unless (get-text-property (1- rung-end-pos) 'c-is-sws)
1907 ;; Remove any `c-in-sws' property from the last char of
1908 ;; the rung before we mark it with `c-is-sws', so that we
1909 ;; won't connect with the remains of a broken "ladder".
1910 (c-remove-in-sws (1- rung-end-pos) rung-end-pos))
1911 (c-put-is-sws simple-ws-beg
1912 rung-end-pos)
1913 (setq rung-is-marked t)))
1914 (c-put-in-sws (setq simple-ws-beg (point)
1915 last-put-in-sws-pos simple-ws-beg)
1916 rung-pos)
1917 (c-put-is-sws (setq rung-pos simple-ws-beg)
1918 (1+ next-rung-pos)))
1919
1920 (c-debug-sws-msg
1921 "c-backward-sws not caching [%s..%s] - [%s..%s] (min %s)"
1922 (point) (1+ next-rung-pos)
1923 simple-ws-beg (min (1+ rung-pos) (point-max))
1924 (point-min))
1925 (setq rung-pos next-rung-pos
1926 simple-ws-beg (point))
1927 ))
1928
1929 ;; Make sure that the newly marked `c-in-sws' region doesn't connect to
1930 ;; another one before the point (which might occur when editing inside a
1931 ;; comment or macro).
1932 (when (eq last-put-in-sws-pos (point))
1933 (cond ((< (point-min) last-put-in-sws-pos)
1934 (c-debug-sws-msg
1935 "c-backward-sws clearing at %s for cache separation"
1936 (1- last-put-in-sws-pos))
1937 (c-remove-in-sws (1- last-put-in-sws-pos)
1938 last-put-in-sws-pos))
1939 ((> (point-min) 1)
1940 ;; If at bob and the buffer is narrowed, we have to clear the
1941 ;; character we're standing on instead since there might be a
1942 ;; `c-in-sws' before (point-min). In this case it's necessary
1943 ;; to clear both properties.
1944 (c-debug-sws-msg
1945 "c-backward-sws clearing thoroughly at %s for cache separation"
1946 last-put-in-sws-pos)
1947 (c-remove-is-and-in-sws last-put-in-sws-pos
1948 (1+ last-put-in-sws-pos)))))
1949 )))
785eecbb 1950
d9e94c22 1951\f
580fba94
AM
1952;; Other whitespace tools
1953(defun c-partial-ws-p (beg end)
1954 ;; Is the region (beg end) WS, and is there WS (or BOB/EOB) next to the
1955 ;; region? This is a "heuristic" function. .....
b414f371 1956 ;;
c421028a
AM
1957 ;; The motivation for the second bit is to check whether removing this
1958 ;; region would coalesce two symbols.
580fba94
AM
1959 ;;
1960 ;; FIXME!!! This function doesn't check virtual semicolons in any way. Be
1961 ;; careful about using this function for, e.g. AWK. (2007/3/7)
1962 (save-excursion
1963 (let ((end+1 (min (1+ end) (point-max))))
1964 (or (progn (goto-char (max (point-min) (1- beg)))
1965 (c-skip-ws-forward end)
1966 (eq (point) end))
1967 (progn (goto-char beg)
1968 (c-skip-ws-forward end+1)
1969 (eq (point) end+1))))))
1970\f
0386b551 1971;; A system for finding noteworthy parens before the point.
e1c458ae 1972
0ec1d2c5
AM
1973(defconst c-state-cache-too-far 5000)
1974;; A maximum comfortable scanning distance, e.g. between
1975;; `c-state-cache-good-pos' and "HERE" (where we call c-parse-state). When
1976;; this distance is exceeded, we take "emergency meausures", e.g. by clearing
1977;; the cache and starting again from point-min or a beginning of defun. This
1978;; value can be tuned for efficiency or set to a lower value for testing.
1979
d9e94c22
MS
1980(defvar c-state-cache nil)
1981(make-variable-buffer-local 'c-state-cache)
1982;; The state cache used by `c-parse-state' to cut down the amount of
0ec1d2c5
AM
1983;; searching. It's the result from some earlier `c-parse-state' call. See
1984;; `c-parse-state''s doc string for details of its structure.
0386b551 1985;;
d9e94c22
MS
1986;; The use of the cached info is more effective if the next
1987;; `c-parse-state' call is on a line close by the one the cached state
1988;; was made at; the cache can actually slow down a little if the
1989;; cached state was made very far back in the buffer. The cache is
1990;; most effective if `c-parse-state' is used on each line while moving
1991;; forward.
e1c458ae 1992
0386b551
AM
1993(defvar c-state-cache-good-pos 1)
1994(make-variable-buffer-local 'c-state-cache-good-pos)
0ec1d2c5
AM
1995;; This is a position where `c-state-cache' is known to be correct, or
1996;; nil (see below). It's a position inside one of the recorded unclosed
1997;; parens or the top level, but not further nested inside any literal or
1998;; subparen that is closed before the last recorded position.
0386b551
AM
1999;;
2000;; The exact position is chosen to try to be close to yet earlier than
2001;; the position where `c-state-cache' will be called next. Right now
2002;; the heuristic is to set it to the position after the last found
2003;; closing paren (of any type) before the line on which
2004;; `c-parse-state' was called. That is chosen primarily to work well
2005;; with refontification of the current line.
0ec1d2c5
AM
2006;;
2007;; 2009-07-28: When `c-state-point-min' and the last position where
2008;; `c-parse-state' or for which `c-invalidate-state-cache' was called, are
2009;; both in the same literal, there is no such "good position", and
2010;; c-state-cache-good-pos is then nil. This is the ONLY circumstance in which
2011;; it can be nil. In this case, `c-state-point-min-literal' will be non-nil.
2012;;
2013;; 2009-06-12: In a brace desert, c-state-cache-good-pos may also be in
2014;; the middle of the desert, as long as it is not within a brace pair
2015;; recorded in `c-state-cache' or a paren/bracket pair.
2016
2017
2018;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2019;; We maintain a simple cache of positions which aren't in a literal, so as to
2020;; speed up testing for non-literality.
2021(defconst c-state-nonlit-pos-interval 10000)
2022;; The approximate interval between entries in `c-state-nonlit-pos-cache'.
2023
2024(defvar c-state-nonlit-pos-cache nil)
2025(make-variable-buffer-local 'c-state-nonlit-pos-cache)
e84efb70
AM
2026;; A list of buffer positions which are known not to be in a literal or a cpp
2027;; construct. This is ordered with higher positions at the front of the list.
2028;; Only those which are less than `c-state-nonlit-pos-cache-limit' are valid.
0ec1d2c5
AM
2029
2030(defvar c-state-nonlit-pos-cache-limit 1)
2031(make-variable-buffer-local 'c-state-nonlit-pos-cache-limit)
2032;; An upper limit on valid entries in `c-state-nonlit-pos-cache'. This is
2033;; reduced by buffer changes, and increased by invocations of
2034;; `c-state-literal-at'.
2035
2036(defsubst c-state-pp-to-literal (from to)
2037 ;; Do a parse-partial-sexp from FROM to TO, returning the bounds of any
2038 ;; literal at TO as a cons, otherwise NIL.
2039 ;; FROM must not be in a literal, and the buffer should already be wide
2040 ;; enough.
2041 (save-excursion
2042 (let ((s (parse-partial-sexp from to)))
2043 (when (or (nth 3 s) (nth 4 s)) ; in a string or comment
2044 (parse-partial-sexp (point) (point-max)
2045 nil ; TARGETDEPTH
2046 nil ; STOPBEFORE
2047 s ; OLDSTATE
2048 'syntax-table) ; stop at end of literal
2049 (cons (nth 8 s) (point))))))
2050
2051(defun c-state-literal-at (here)
2052 ;; If position HERE is inside a literal, return (START . END), the
2053 ;; boundaries of the literal (which may be outside the accessible bit of the
2054 ;; buffer). Otherwise, return nil.
2055 ;;
2056 ;; This function is almost the same as `c-literal-limits'. It differs in
2057 ;; that it is a lower level function, and that it rigourously follows the
2058 ;; syntax from BOB, whereas `c-literal-limits' uses a "local" safe position.
e84efb70
AM
2059 ;;
2060 ;; NOTE: This function manipulates `c-state-nonlit-pos-cache'. This cache
2061 ;; MAY NOT contain any positions within macros, since macros are frequently
2062 ;; turned into comments by use of the `c-cpp-delimiter' category properties.
2063 ;; We cannot rely on this mechanism whilst determining a cache pos since
2064 ;; this function is also called from outwith `c-parse-state'.
0ec1d2c5
AM
2065 (save-restriction
2066 (widen)
2067 (save-excursion
2068 (let ((c c-state-nonlit-pos-cache)
2069 pos npos lit)
2070 ;; Trim the cache to take account of buffer changes.
2071 (while (and c (> (car c) c-state-nonlit-pos-cache-limit))
2072 (setq c (cdr c)))
2073 (setq c-state-nonlit-pos-cache c)
2074
2075 (while (and c (> (car c) here))
2076 (setq c (cdr c)))
2077 (setq pos (or (car c) (point-min)))
2078
2079 (while (<= (setq npos (+ pos c-state-nonlit-pos-interval))
2080 here)
2081 (setq lit (c-state-pp-to-literal pos npos))
2082 (setq pos (or (cdr lit) npos)) ; end of literal containing npos.
e84efb70
AM
2083 (goto-char pos)
2084 (when (and (c-beginning-of-macro) (/= (point) pos))
2085 (c-syntactic-end-of-macro)
2086 (or (eobp) (forward-char))
2087 (setq pos (point)))
0ec1d2c5
AM
2088 (setq c-state-nonlit-pos-cache (cons pos c-state-nonlit-pos-cache)))
2089
2090 (if (> pos c-state-nonlit-pos-cache-limit)
2091 (setq c-state-nonlit-pos-cache-limit pos))
2092 (if (< pos here)
2093 (setq lit (c-state-pp-to-literal pos here)))
2094 lit))))
2095
2096(defsubst c-state-lit-beg (pos)
2097 ;; Return the start of the literal containing POS, or POS itself.
2098 (or (car (c-state-literal-at pos))
2099 pos))
2100
2101(defsubst c-state-cache-non-literal-place (pos state)
2102 ;; Return a position outside of a string/comment at or before POS.
2103 ;; STATE is the parse-partial-sexp state at POS.
2104 (if (or (nth 3 state) ; in a string?
2105 (nth 4 state)) ; in a comment?
2106 (nth 8 state)
2107 pos))
2108
2109
2110;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2111;; Stuff to do with point-min, and coping with any literal there.
2112(defvar c-state-point-min 1)
2113(make-variable-buffer-local 'c-state-point-min)
2114;; This is (point-min) when `c-state-cache' was last calculated. A change of
2115;; narrowing is likely to affect the parens that are visible before the point.
2116
2117(defvar c-state-point-min-lit-type nil)
2118(make-variable-buffer-local 'c-state-point-min-lit-type)
2119(defvar c-state-point-min-lit-start nil)
2120(make-variable-buffer-local 'c-state-point-min-lit-start)
2121;; These two variables define the literal, if any, containing point-min.
2122;; Their values are, respectively, 'string, c, or c++, and the start of the
2123;; literal. If there's no literal there, they're both nil.
2124
2125(defvar c-state-min-scan-pos 1)
2126(make-variable-buffer-local 'c-state-min-scan-pos)
2127;; This is the earliest buffer-pos from which scanning can be done. It is
2128;; either the end of the literal containing point-min, or point-min itself.
2129;; It becomes nil if the buffer is changed earlier than this point.
2130(defun c-state-get-min-scan-pos ()
2131 ;; Return the lowest valid scanning pos. This will be the end of the
2132 ;; literal enclosing point-min, or point-min itself.
2133 (or c-state-min-scan-pos
2134 (save-restriction
2135 (save-excursion
2136 (widen)
2137 (goto-char c-state-point-min-lit-start)
2138 (if (eq c-state-point-min-lit-type 'string)
2139 (forward-sexp)
2140 (forward-comment 1))
2141 (setq c-state-min-scan-pos (point))))))
2142
2143(defun c-state-mark-point-min-literal ()
2144 ;; Determine the properties of any literal containing POINT-MIN, setting the
2145 ;; variables `c-state-point-min-lit-type', `c-state-point-min-lit-start',
2146 ;; and `c-state-min-scan-pos' accordingly. The return value is meaningless.
2147 (let ((p-min (point-min))
2148 lit)
2149 (save-restriction
2150 (widen)
2151 (setq lit (c-state-literal-at p-min))
2152 (if lit
2153 (setq c-state-point-min-lit-type
2154 (save-excursion
2155 (goto-char (car lit))
2156 (cond
2157 ((looking-at c-block-comment-start-regexp) 'c)
2158 ((looking-at c-line-comment-starter) 'c++)
2159 (t 'string)))
2160 c-state-point-min-lit-start (car lit)
2161 c-state-min-scan-pos (cdr lit))
2162 (setq c-state-point-min-lit-type nil
2163 c-state-point-min-lit-start nil
2164 c-state-min-scan-pos p-min)))))
2165
2166
2167;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2168;; A variable which signals a brace dessert - helpful for reducing the number
2169;; of fruitless backward scans.
2170(defvar c-state-brace-pair-desert nil)
2171(make-variable-buffer-local 'c-state-brace-pair-desert)
e84efb70 2172;; Used only in `c-append-lower-brace-pair-to-state-cache'. It is set when
0ec1d2c5
AM
2173;; that defun has searched backwards for a brace pair and not found one. Its
2174;; value is either nil or a cons (PA . FROM), where PA is the position of the
2175;; enclosing opening paren/brace/bracket which bounds the backwards search (or
2176;; nil when at top level) and FROM is where the backward search started. It
2177;; is reset to nil in `c-invalidate-state-cache'.
2178
2179
2180;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2181;; Lowish level functions/macros which work directly on `c-state-cache', or a
2182;; list of like structure.
2183(defmacro c-state-cache-top-lparen (&optional cache)
2184 ;; Return the address of the top left brace/bracket/paren recorded in CACHE
2185 ;; (default `c-state-cache') (or nil).
2186 (let ((cash (or cache 'c-state-cache)))
2187 `(if (consp (car ,cash))
2188 (caar ,cash)
2189 (car ,cash))))
2190
2191(defmacro c-state-cache-top-paren (&optional cache)
2192 ;; Return the address of the latest brace/bracket/paren (whether left or
2193 ;; right) recorded in CACHE (default `c-state-cache') or nil.
2194 (let ((cash (or cache 'c-state-cache)))
2195 `(if (consp (car ,cash))
2196 (cdar ,cash)
2197 (car ,cash))))
2198
2199(defmacro c-state-cache-after-top-paren (&optional cache)
2200 ;; Return the position just after the latest brace/bracket/paren (whether
2201 ;; left or right) recorded in CACHE (default `c-state-cache') or nil.
2202 (let ((cash (or cache 'c-state-cache)))
2203 `(if (consp (car ,cash))
2204 (cdar ,cash)
2205 (and (car ,cash)
2206 (1+ (car ,cash))))))
2207
2208(defun c-get-cache-scan-pos (here)
2209 ;; From the state-cache, determine the buffer position from which we might
2210 ;; scan forward to HERE to update this cache. This position will be just
2211 ;; after a paren/brace/bracket recorded in the cache, if possible, otherwise
2212 ;; return the earliest position in the accessible region which isn't within
2213 ;; a literal. If the visible portion of the buffer is entirely within a
2214 ;; literal, return NIL.
2215 (let ((c c-state-cache) elt)
2216 ;(while (>= (or (c-state-cache-top-lparen c) 1) here)
2217 (while (and c
2218 (>= (c-state-cache-top-lparen c) here))
2219 (setq c (cdr c)))
2220
2221 (setq elt (car c))
2222 (cond
2223 ((consp elt)
2224 (if (> (cdr elt) here)
2225 (1+ (car elt))
2226 (cdr elt)))
2227 (elt (1+ elt))
2228 ((<= (c-state-get-min-scan-pos) here)
2229 (c-state-get-min-scan-pos))
2230 (t nil))))
2231
2232;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2233;; Variables which keep track of preprocessor constructs.
2234(defvar c-state-old-cpp-beg nil)
2235(make-variable-buffer-local 'c-state-old-cpp-beg)
2236(defvar c-state-old-cpp-end nil)
2237(make-variable-buffer-local 'c-state-old-cpp-end)
2238;; These are the limits of the macro containing point at the previous call of
2239;; `c-parse-state', or nil.
2240
2241;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2242;; Defuns which analyse the buffer, yet don't change `c-state-cache'.
2243(defun c-get-fallback-scan-pos (here)
2244 ;; Return a start position for building `c-state-cache' from
2245 ;; scratch. This will be at the top level, 2 defuns back.
0386b551
AM
2246 (save-excursion
2247 ;; Go back 2 bods, but ignore any bogus positions returned by
2248 ;; beginning-of-defun (i.e. open paren in column zero).
2249 (goto-char here)
2250 (let ((cnt 2))
2251 (while (not (or (bobp) (zerop cnt)))
0ec1d2c5 2252 (c-beginning-of-defun-1) ; Pure elisp BOD.
0386b551
AM
2253 (if (eq (char-after) ?\{)
2254 (setq cnt (1- cnt)))))
2255 (point)))
2256
c0209c2c
AM
2257(defun c-state-balance-parens-backwards (here- here+ top)
2258 ;; Return the position of the opening paren/brace/bracket before HERE- which
2259 ;; matches the outermost close p/b/b between HERE+ and TOP. Except when
2260 ;; there's a macro, HERE- and HERE+ are the same. Like this:
0ec1d2c5 2261 ;;
c0209c2c
AM
2262 ;; ............................................
2263 ;; | |
2264 ;; ( [ ( .........#macro.. ) ( ) ] )
2265 ;; ^ ^ ^ ^
2266 ;; | | | |
2267 ;; return HERE- HERE+ TOP
0ec1d2c5
AM
2268 ;;
2269 ;; If there aren't enough opening paren/brace/brackets, return the position
c0209c2c
AM
2270 ;; of the outermost one found, or HERE- if there are none. If there are no
2271 ;; closeing p/b/bs between HERE+ and TOP, return HERE-. HERE-/+ and TOP
2272 ;; must not be inside literals. Only the accessible portion of the buffer
2273 ;; will be scanned.
2274
2275 ;; PART 1: scan from `here+' up to `top', accumulating ")"s which enclose
2276 ;; `here'. Go round the next loop each time we pass over such a ")". These
2277 ;; probably match "("s before `here-'.
0ec1d2c5
AM
2278 (let (pos pa ren+1 lonely-rens)
2279 (save-excursion
2280 (save-restriction
2281 (narrow-to-region (point-min) top) ; This can move point, sometimes.
c0209c2c 2282 (setq pos here+)
0ec1d2c5
AM
2283 (c-safe
2284 (while
2285 (setq ren+1 (scan-lists pos 1 1)) ; might signal
2286 (setq lonely-rens (cons ren+1 lonely-rens)
2287 pos ren+1)))))
2288
c0209c2c 2289 ;; PART 2: Scan back before `here-' searching for the "("s
0ec1d2c5
AM
2290 ;; matching/mismatching the ")"s found above. We only need to direct the
2291 ;; caller to scan when we've encountered unmatched right parens.
c0209c2c
AM
2292 (setq pos here-)
2293 (when lonely-rens
2294 (c-safe
2295 (while
2296 (and lonely-rens ; actual values aren't used.
2297 (setq pa (scan-lists pos -1 1)))
2298 (setq pos pa)
2299 (setq lonely-rens (cdr lonely-rens)))))
2300 pos))
0ec1d2c5
AM
2301
2302(defun c-parse-state-get-strategy (here good-pos)
2303 ;; Determine the scanning strategy for adjusting `c-parse-state', attempting
2304 ;; to minimise the amount of scanning. HERE is the pertinent position in
2305 ;; the buffer, GOOD-POS is a position where `c-state-cache' (possibly with
2306 ;; its head trimmed) is known to be good, or nil if there is no such
2307 ;; position.
2308 ;;
2309 ;; The return value is a list, one of the following:
2310 ;;
2311 ;; o - ('forward CACHE-POS START-POINT) - scan forward from START-POINT,
2312 ;; which is not less than CACHE-POS.
2313 ;; o - ('backward CACHE-POS nil) - scan backwards (from HERE).
2314 ;; o - ('BOD nil START-POINT) - scan forwards from START-POINT, which is at the
2315 ;; top level.
2316 ;; o - ('IN-LIT nil nil) - point is inside the literal containing point-min.
2317 ;; , where CACHE-POS is the highest position recorded in `c-state-cache' at
2318 ;; or below HERE.
2319 (let ((cache-pos (c-get-cache-scan-pos here)) ; highest position below HERE in cache (or 1)
2320 BOD-pos ; position of 2nd BOD before HERE.
2321 strategy ; 'forward, 'backward, 'BOD, or 'IN-LIT.
2322 start-point
2323 how-far) ; putative scanning distance.
2324 (setq good-pos (or good-pos (c-state-get-min-scan-pos)))
2325 (cond
2326 ((< here (c-state-get-min-scan-pos))
2327 (setq strategy 'IN-LIT
2328 start-point nil
2329 cache-pos nil
2330 how-far 0))
2331 ((<= good-pos here)
2332 (setq strategy 'forward
2333 start-point (max good-pos cache-pos)
2334 how-far (- here start-point)))
2335 ((< (- good-pos here) (- here cache-pos)) ; FIXME!!! ; apply some sort of weighting.
2336 (setq strategy 'backward
2337 how-far (- good-pos here)))
2338 (t
2339 (setq strategy 'forward
2340 how-far (- here cache-pos)
2341 start-point cache-pos)))
2342
2343 ;; Might we be better off starting from the top level, two defuns back,
2344 ;; instead?
2345 (when (> how-far c-state-cache-too-far)
2346 (setq BOD-pos (c-get-fallback-scan-pos here)) ; somewhat EXPENSIVE!!!
2347 (if (< (- here BOD-pos) how-far)
2348 (setq strategy 'BOD
2349 start-point BOD-pos)))
2350
2351 (list
2352 strategy
2353 (and (memq strategy '(forward backward)) cache-pos)
2354 (and (memq strategy '(forward BOD)) start-point))))
2355
2356
2357;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2358;; Routines which change `c-state-cache' and associated values.
2359(defun c-renarrow-state-cache ()
2360 ;; The region (more precisely, point-min) has changed since we
2361 ;; calculated `c-state-cache'. Amend `c-state-cache' accordingly.
2362 (if (< (point-min) c-state-point-min)
2363 ;; If point-min has MOVED BACKWARDS then we drop the state completely.
2364 ;; It would be possible to do a better job here and recalculate the top
2365 ;; only.
2366 (progn
2367 (c-state-mark-point-min-literal)
2368 (setq c-state-cache nil
2369 c-state-cache-good-pos c-state-min-scan-pos
2370 c-state-brace-pair-desert nil))
2371
2372 ;; point-min has MOVED FORWARD.
2373
2374 ;; Is the new point-min inside a (different) literal?
2375 (unless (and c-state-point-min-lit-start ; at prev. point-min
2376 (< (point-min) (c-state-get-min-scan-pos)))
2377 (c-state-mark-point-min-literal))
2378
2379 ;; Cut off a bit of the tail from `c-state-cache'.
2380 (let ((ptr (cons nil c-state-cache))
2381 pa)
2382 (while (and (setq pa (c-state-cache-top-lparen (cdr ptr)))
2383 (>= pa (point-min)))
2384 (setq ptr (cdr ptr)))
2385
2386 (when (consp ptr)
2387 (if (eq (cdr ptr) c-state-cache)
2388 (setq c-state-cache nil
2389 c-state-cache-good-pos c-state-min-scan-pos)
2390 (setcdr ptr nil)
2391 (setq c-state-cache-good-pos (1+ (c-state-cache-top-lparen))))
2392 )))
2393
2394 (setq c-state-point-min (point-min)))
2395
2396(defun c-append-lower-brace-pair-to-state-cache (from &optional upper-lim)
2397 ;; If there is a brace pair preceding FROM in the buffer (not necessarily
2398 ;; immediately preceding), push a cons onto `c-state-cache' to represent it.
2399 ;; FROM must not be inside a literal. If UPPER-LIM is non-nil, we append
2400 ;; the highest brace pair whose "}" is below UPPER-LIM.
2401 ;;
2402 ;; Return non-nil when this has been done.
2403 ;;
2404 ;; This routine should be fast. Since it can get called a LOT, we maintain
2405 ;; `c-state-brace-pair-desert', a small cache of "failures", such that we
2406 ;; reduce the time wasted in repeated fruitless searches in brace deserts.
2407 (save-excursion
2408 (save-restriction
2409 (let ((bra from) ce ; Positions of "{" and "}".
2410 new-cons
2411 (cache-pos (c-state-cache-top-lparen)) ; might be nil.
2412 (macro-start-or-from
2413 (progn (goto-char from)
2414 (c-beginning-of-macro)
2415 (point))))
2416 (or upper-lim (setq upper-lim from))
2417
2418 ;; If we're essentially repeating a fruitless search, just give up.
2419 (unless (and c-state-brace-pair-desert
2420 (eq cache-pos (car c-state-brace-pair-desert))
2421 (<= from (cdr c-state-brace-pair-desert)))
2422 ;; Only search what we absolutely need to:
2423 (if (and c-state-brace-pair-desert
2424 (> from (cdr c-state-brace-pair-desert)))
2425 (narrow-to-region (cdr c-state-brace-pair-desert) (point-max)))
2426
2427 ;; In the next pair of nested loops, the inner one moves back past a
2428 ;; pair of (mis-)matching parens or brackets; the outer one moves
2429 ;; back over a sequence of unmatched close brace/paren/bracket each
2430 ;; time round.
2431 (while
2432 (progn
2433 (c-safe
2434 (while
2435 (and (setq ce (scan-lists bra -1 -1)) ; back past )/]/}; might signal
2436 (setq bra (scan-lists ce -1 1)) ; back past (/[/{; might signal
2437 (or (> ce upper-lim)
2438 (not (eq (char-after bra) ?\{))
2439 (and (goto-char bra)
2440 (c-beginning-of-macro)
2441 (< (point) macro-start-or-from))))))
2442 (and ce (< ce bra)))
2443 (setq bra ce)) ; If we just backed over an unbalanced closing
2444 ; brace, ignore it.
2445
2446 (if (and ce (< bra ce) (eq (char-after bra) ?\{))
2447 ;; We've found the desired brace-pair.
2448 (progn
2449 (setq new-cons (cons bra (1+ ce)))
2450 (cond
2451 ((consp (car c-state-cache))
2452 (setcar c-state-cache new-cons))
2453 ((and (numberp (car c-state-cache)) ; probably never happens
2454 (< ce (car c-state-cache)))
2455 (setcdr c-state-cache
2456 (cons new-cons (cdr c-state-cache))))
2457 (t (setq c-state-cache (cons new-cons c-state-cache)))))
2458
2459 ;; We haven't found a brace pair. Record this.
2460 (setq c-state-brace-pair-desert (cons cache-pos from))))))))
2461
2462(defsubst c-state-push-any-brace-pair (bra+1 macro-start-or-here)
2463 ;; If BRA+1 is nil, do nothing. Otherwise, BRA+1 is the buffer position
2464 ;; following a {, and that brace has a (mis-)matching } (or ]), and we
2465 ;; "push" "a" brace pair onto `c-state-cache'.
2466 ;;
2467 ;; Here "push" means overwrite the top element if it's itself a brace-pair,
2468 ;; otherwise push it normally.
2469 ;;
2470 ;; The brace pair we push is normally the one surrounding BRA+1, but if the
2471 ;; latter is inside a macro, not being a macro containing
2472 ;; MACRO-START-OR-HERE, we scan backwards through the buffer for a non-macro
2473 ;; base pair. This latter case is assumed to be rare.
2474 ;;
2475 ;; Note: POINT is not preserved in this routine.
2476 (if bra+1
2477 (if (or (> bra+1 macro-start-or-here)
2478 (progn (goto-char bra+1)
2479 (not (c-beginning-of-macro))))
2480 (setq c-state-cache
2481 (cons (cons (1- bra+1)
2482 (scan-lists bra+1 1 1))
2483 (if (consp (car c-state-cache))
2484 (cdr c-state-cache)
2485 c-state-cache)))
2486 ;; N.B. This defsubst codes one method for the simple, normal case,
2487 ;; and a more sophisticated, slower way for the general case. Don't
2488 ;; eliminate this defsubst - it's a speed optimisation.
2489 (c-append-lower-brace-pair-to-state-cache (1- bra+1)))))
2490
2491(defun c-append-to-state-cache (from)
2492 ;; Scan the buffer from FROM to (point-max), adding elements into
2493 ;; `c-state-cache' for braces etc. Return a candidate for
2494 ;; `c-state-cache-good-pos'.
2495 ;;
2496 ;; FROM must be after the latest brace/paren/bracket in `c-state-cache', if
2497 ;; any. Typically, it is immediately after it. It must not be inside a
2498 ;; literal.
2499 (let ((here-bol (c-point 'bol (point-max)))
2500 (macro-start-or-here
2501 (save-excursion (goto-char (point-max))
2502 (if (c-beginning-of-macro)
2503 (point)
2504 (point-max))))
2505 pa+1 ; pos just after an opening PAren (or brace).
2506 (ren+1 from) ; usually a pos just after an closing paREN etc.
2507 ; Is actually the pos. to scan for a (/{/[ from,
2508 ; which sometimes is after a silly )/}/].
2509 paren+1 ; Pos after some opening or closing paren.
2510 paren+1s ; A list of `paren+1's; used to determine a
2511 ; good-pos.
2512 bra+1 ce+1 ; just after L/R bra-ces.
2513 bra+1s ; list of OLD values of bra+1.
2514 mstart) ; start of a macro.
2515
2516 (save-excursion
2517 ;; Each time round the following loop, we enter a succesively deeper
2518 ;; level of brace/paren nesting. (Except sometimes we "continue at
2519 ;; the existing level".) `pa+1' is a pos inside an opening
2520 ;; brace/paren/bracket, usually just after it.
2521 (while
2522 (progn
2523 ;; Each time round the next loop moves forward over an opening then
2524 ;; a closing brace/bracket/paren. This loop is white hot, so it
2525 ;; plays ugly tricks to go fast. DON'T PUT ANYTHING INTO THIS
2526 ;; LOOP WHICH ISN'T ABSOLUTELY NECESSARY!!! It terminates when a
2527 ;; call of `scan-lists' signals an error, which happens when there
2528 ;; are no more b/b/p's to scan.
2529 (c-safe
2530 (while t
2531 (setq pa+1 (scan-lists ren+1 1 -1) ; Into (/{/[; might signal
2532 paren+1s (cons pa+1 paren+1s))
2533 (setq ren+1 (scan-lists pa+1 1 1)) ; Out of )/}/]; might signal
2534 (if (and (eq (char-before pa+1) ?{)) ; Check for a macro later.
2535 (setq bra+1 pa+1))
2536 (setcar paren+1s ren+1)))
2537
2538 (if (and pa+1 (> pa+1 ren+1))
2539 ;; We've just entered a deeper nesting level.
2540 (progn
2541 ;; Insert the brace pair (if present) and the single open
2542 ;; paren/brace/bracket into `c-state-cache' It cannot be
2543 ;; inside a macro, except one around point, because of what
2544 ;; `c-neutralize-syntax-in-CPP' has done.
2545 (c-state-push-any-brace-pair bra+1 macro-start-or-here)
2546 ;; Insert the opening brace/bracket/paren position.
2547 (setq c-state-cache (cons (1- pa+1) c-state-cache))
2548 ;; Clear admin stuff for the next more nested part of the scan.
2549 (setq ren+1 pa+1 pa+1 nil bra+1 nil bra+1s nil)
2550 t) ; Carry on the loop
2551
2552 ;; All open p/b/b's at this nesting level, if any, have probably
2553 ;; been closed by matching/mismatching ones. We're probably
2554 ;; finished - we just need to check for having found an
2555 ;; unmatched )/}/], which we ignore. Such a )/}/] can't be in a
2556 ;; macro, due the action of `c-neutralize-syntax-in-CPP'.
2557 (c-safe (setq ren+1 (scan-lists ren+1 1 1)))))) ; acts as loop control.
2558
2559 ;; Record the final, innermost, brace-pair if there is one.
2560 (c-state-push-any-brace-pair bra+1 macro-start-or-here)
2561
2562 ;; Determine a good pos
2563 (while (and (setq paren+1 (car paren+1s))
2564 (> (if (> paren+1 macro-start-or-here)
2565 paren+1
2566 (goto-char paren+1)
2567 (setq mstart (and (c-beginning-of-macro)
2568 (point)))
2569 (or mstart paren+1))
2570 here-bol))
2571 (setq paren+1s (cdr paren+1s)))
2572 (cond
2573 ((and paren+1 mstart)
2574 (min paren+1 mstart))
2575 (paren+1)
2576 (t from)))))
2577
2578(defun c-remove-stale-state-cache (good-pos pps-point)
2579 ;; Remove stale entries from the `c-cache-state', i.e. those which will
2580 ;; not be in it when it is amended for position (point-max).
2581 ;; Additionally, the "outermost" open-brace entry before (point-max)
2582 ;; will be converted to a cons if the matching close-brace is scanned.
2583 ;;
2584 ;; GOOD-POS is a "maximal" "safe position" - there must be no open
2585 ;; parens/braces/brackets between GOOD-POS and (point-max).
2586 ;;
2587 ;; As a second thing, calculate the result of parse-partial-sexp at
2588 ;; PPS-POINT, w.r.t. GOOD-POS. The motivation here is that
2589 ;; `c-state-cache-good-pos' may become PPS-POINT, but the caller may need to
2590 ;; adjust it to get outside a string/comment. (Sorry about this! The code
2591 ;; needs to be FAST).
2592 ;;
2593 ;; Return a list (GOOD-POS SCAN-BACK-POS PPS-STATE), where
2594 ;; o - GOOD-POS is a position where the new value `c-state-cache' is known
2595 ;; to be good (we aim for this to be as high as possible);
2596 ;; o - SCAN-BACK-POS, if not nil, indicates there may be a brace pair
2597 ;; preceding POS which needs to be recorded in `c-state-cache'. It is a
2598 ;; position to scan backwards from.
2599 ;; o - PPS-STATE is the parse-partial-sexp state at PPS-POINT.
2600 (save-restriction
2601 (narrow-to-region 1 (point-max))
2602 (save-excursion
eb2f6eeb 2603 (let* ((in-macro-start ; start of macro containing (point-max) or nil.
0ec1d2c5
AM
2604 (save-excursion
2605 (goto-char (point-max))
2606 (and (c-beginning-of-macro)
2607 (point))))
2608 (good-pos-actual-macro-start ; Start of macro containing good-pos
2609 ; or nil
2610 (and (< good-pos (point-max))
2611 (save-excursion
2612 (goto-char good-pos)
2613 (and (c-beginning-of-macro)
2614 (point)))))
2615 (good-pos-actual-macro-end ; End of this macro, (maybe
2616 ; (point-max)), or nil.
2617 (and good-pos-actual-macro-start
2618 (save-excursion
2619 (goto-char good-pos-actual-macro-start)
2620 (c-end-of-macro)
2621 (point))))
2622 pps-state ; Will be 9 or 10 elements long.
2623 pos
2624 upper-lim ; ,beyond which `c-state-cache' entries are removed
2625 scan-back-pos
2626 pair-beg pps-point-state target-depth)
2627
2628 ;; Remove entries beyond (point-max). Also remove any entries inside
2629 ;; a macro, unless (point-max) is in the same macro.
2630 (setq upper-lim
2631 (if (or (null c-state-old-cpp-beg)
2632 (and (> (point-max) c-state-old-cpp-beg)
2633 (< (point-max) c-state-old-cpp-end)))
2634 (point-max)
2635 (min (point-max) c-state-old-cpp-beg)))
657071fc 2636 (while (and c-state-cache (>= (c-state-cache-top-lparen) upper-lim))
0ec1d2c5
AM
2637 (setq c-state-cache (cdr c-state-cache)))
2638 ;; If `upper-lim' is inside the last recorded brace pair, remove its
2639 ;; RBrace and indicate we'll need to search backwards for a previous
2640 ;; brace pair.
2641 (when (and c-state-cache
2642 (consp (car c-state-cache))
2643 (> (cdar c-state-cache) upper-lim))
2644 (setcar c-state-cache (caar c-state-cache))
2645 (setq scan-back-pos (car c-state-cache)))
2646
2647 ;; The next loop jumps forward out of a nested level of parens each
2648 ;; time round; the corresponding elements in `c-state-cache' are
2649 ;; removed. `pos' is just after the brace-pair or the open paren at
2650 ;; (car c-state-cache). There can be no open parens/braces/brackets
2651 ;; between `good-pos'/`good-pos-actual-macro-start' and (point-max),
2652 ;; due to the interface spec to this function.
dd21b621 2653 (setq pos (if (and good-pos-actual-macro-end
eb2f6eeb
AM
2654 (not (eq good-pos-actual-macro-start
2655 in-macro-start)))
0ec1d2c5
AM
2656 (1+ good-pos-actual-macro-end) ; get outside the macro as
2657 ; marked by a `category' text property.
2658 good-pos))
2659 (goto-char pos)
2660 (while (and c-state-cache
2661 (< (point) (point-max)))
2662 (cond
2663 ((null pps-state) ; first time through
2664 (setq target-depth -1))
2665 ((eq (car pps-state) target-depth) ; found closing ),},]
2666 (setq target-depth (1- (car pps-state))))
2667 ;; Do nothing when we've merely reached pps-point.
2668 )
2669
2670 ;; Scan!
2671 (setq pps-state
2672 (parse-partial-sexp
2673 (point) (if (< (point) pps-point) pps-point (point-max))
2674 target-depth
2675 nil pps-state))
2676
2677 (if (= (point) pps-point)
2678 (setq pps-point-state pps-state))
2679
2680 (when (eq (car pps-state) target-depth)
2681 (setq pos (point)) ; POS is now just after an R-paren/brace.
2682 (cond
2683 ((and (consp (car c-state-cache))
2684 (eq (point) (cdar c-state-cache)))
2685 ;; We've just moved out of the paren pair containing the brace-pair
2686 ;; at (car c-state-cache). `pair-beg' is where the open paren is,
2687 ;; and is potentially where the open brace of a cons in
2688 ;; c-state-cache will be.
2689 (setq pair-beg (car-safe (cdr c-state-cache))
2690 c-state-cache (cdr-safe (cdr c-state-cache)))) ; remove {}pair + containing Lparen.
2691 ((numberp (car c-state-cache))
2692 (setq pair-beg (car c-state-cache)
2693 c-state-cache (cdr c-state-cache))) ; remove this
2694 ; containing Lparen
2695 ((numberp (cadr c-state-cache))
2696 (setq pair-beg (cadr c-state-cache)
2697 c-state-cache (cddr c-state-cache))) ; Remove a paren pair
2698 ; together with enclosed brace pair.
2699 ;; (t nil) ; Ignore an unmated Rparen.
2700 )))
2701
2702 (if (< (point) pps-point)
2703 (setq pps-state (parse-partial-sexp (point) pps-point
2704 nil nil ; TARGETDEPTH, STOPBEFORE
2705 pps-state)))
2706
2707 ;; If the last paren pair we moved out of was actually a brace pair,
2708 ;; insert it into `c-state-cache'.
2709 (when (and pair-beg (eq (char-after pair-beg) ?{))
2710 (if (consp (car-safe c-state-cache))
2711 (setq c-state-cache (cdr c-state-cache)))
2712 (setq c-state-cache (cons (cons pair-beg pos)
2713 c-state-cache)))
2714
2715 (list pos scan-back-pos pps-state)))))
2716
2717(defun c-remove-stale-state-cache-backwards (here cache-pos)
2718 ;; Strip stale elements of `c-state-cache' by moving backwards through the
2719 ;; buffer, and inform the caller of the scenario detected.
2720 ;;
2721 ;; HERE is the position we're setting `c-state-cache' for.
2722 ;; CACHE-POS is just after the latest recorded position in `c-state-cache'
2723 ;; before HERE, or a position at or near point-min which isn't in a
2724 ;; literal.
2725 ;;
2726 ;; This function must only be called only when (> `c-state-cache-good-pos'
2727 ;; HERE). Usually the gap between CACHE-POS and HERE is large. It is thus
2728 ;; optimised to eliminate (or minimise) scanning between these two
2729 ;; positions.
2730 ;;
2731 ;; Return a three element list (GOOD-POS SCAN-BACK-POS FWD-FLAG), where:
2732 ;; o - GOOD-POS is a "good position", where `c-state-cache' is valid, or
2733 ;; could become so after missing elements are inserted into
2734 ;; `c-state-cache'. This is JUST AFTER an opening or closing
2735 ;; brace/paren/bracket which is already in `c-state-cache' or just before
2736 ;; one otherwise. exceptionally (when there's no such b/p/b handy) the BOL
2737 ;; before `here''s line, or the start of the literal containing it.
2738 ;; o - SCAN-BACK-POS, if non-nil, indicates there may be a brace pair
2739 ;; preceding POS which isn't recorded in `c-state-cache'. It is a position
2740 ;; to scan backwards from.
2741 ;; o - FWD-FLAG, if non-nil, indicates there may be parens/braces between
2742 ;; POS and HERE which aren't recorded in `c-state-cache'.
2743 ;;
2744 ;; The comments in this defun use "paren" to mean parenthesis or square
2745 ;; bracket (as contrasted with a brace), and "(" and ")" likewise.
2746 ;;
2747 ;; . {..} (..) (..) ( .. { } ) (...) ( .... . ..)
2748 ;; | | | | | |
2749 ;; CP E here D C good
2750 (let ((pos c-state-cache-good-pos)
2751 pa ren ; positions of "(" and ")"
2752 dropped-cons ; whether the last element dropped from `c-state-cache'
2753 ; was a cons (representing a brace-pair)
2754 good-pos ; see above.
2755 lit ; (START . END) of a literal containing some point.
2756 here-lit-start here-lit-end ; bounds of literal containing `here'
2757 ; or `here' itself.
c0209c2c 2758 here- here+ ; start/end of macro around HERE, or HERE
0ec1d2c5
AM
2759 (here-bol (c-point 'bol here))
2760 (too-far-back (max (- here c-state-cache-too-far) 1)))
2761
2762 ;; Remove completely irrelevant entries from `c-state-cache'.
2763 (while (and c-state-cache
2764 (>= (setq pa (c-state-cache-top-lparen)) here))
2765 (setq dropped-cons (consp (car c-state-cache)))
2766 (setq c-state-cache (cdr c-state-cache))
2767 (setq pos pa))
2768 ;; At this stage, (> pos here);
2769 ;; (< (c-state-cache-top-lparen) here) (or is nil).
2770
c0209c2c
AM
2771 (cond
2772 ((and (consp (car c-state-cache))
2773 (> (cdar c-state-cache) here))
2774 ;; CASE 1: The top of the cache is a brace pair which now encloses
2775 ;; `here'. As good-pos, return the address. of the "{". Since we've no
2776 ;; knowledge of what's inside these braces, we have no alternative but
2777 ;; to direct the caller to scan the buffer from the opening brace.
2778 (setq pos (caar c-state-cache))
2779 (setcar c-state-cache pos)
2780 (list (1+ pos) pos t)) ; return value. We've just converted a brace pair
2781 ; entry into a { entry, so the caller needs to
2782 ; search for a brace pair before the {.
2783
2784 ;; `here' might be inside a literal. Check for this.
2785 ((progn
2786 (setq lit (c-state-literal-at here)
2787 here-lit-start (or (car lit) here)
2788 here-lit-end (or (cdr lit) here))
2789 ;; Has `here' just "newly entered" a macro?
2790 (save-excursion
2791 (goto-char here-lit-start)
2792 (if (and (c-beginning-of-macro)
2793 (or (null c-state-old-cpp-beg)
2794 (not (= (point) c-state-old-cpp-beg))))
2795 (progn
2796 (setq here- (point))
2797 (c-end-of-macro)
2798 (setq here+ (point)))
2799 (setq here- here-lit-start
2800 here+ here-lit-end)))
2801
2802 ;; `here' might be nested inside any depth of parens (or brackets but
2803 ;; not braces). Scan backwards to find the outermost such opening
2804 ;; paren, if there is one. This will be the scan position to return.
2805 (save-restriction
2806 (narrow-to-region cache-pos (point-max))
2807 (setq pos (c-state-balance-parens-backwards here- here+ pos)))
2808 nil)) ; for the cond
2809
2810 ((< pos here-lit-start)
2811 ;; CASE 2: Address of outermost ( or [ which now encloses `here', but
2812 ;; didn't enclose the (previous) `c-state-cache-good-pos'. If there is
2813 ;; a brace pair preceding this, it will already be in `c-state-cache',
2814 ;; unless there was a brace pair after it, i.e. there'll only be one to
2815 ;; scan for if we've just deleted one.
2816 (list pos (and dropped-cons pos) t)) ; Return value.
2817
2818 ;; `here' isn't enclosed in a (previously unrecorded) bracket/paren.
2819 ;; Further forward scanning isn't needed, but we still need to find a
2820 ;; GOOD-POS. Step out of all enclosing "("s on HERE's line.
2821 ((progn
0ec1d2c5
AM
2822 (save-restriction
2823 (narrow-to-region here-bol (point-max))
2824 (setq pos here-lit-start)
2825 (c-safe (while (setq pa (scan-lists pos -1 1))
2826 (setq pos pa)))) ; might signal
c0209c2c
AM
2827 nil)) ; for the cond
2828
2829 ((setq ren (c-safe-scan-lists pos -1 -1 too-far-back))
2830 ;; CASE 3: After a }/)/] before `here''s BOL.
2831 (list (1+ ren) (and dropped-cons pos) nil)) ; Return value
2832
2833 (t
2834 ;; CASE 4; Best of a bad job: BOL before `here-bol', or beginning of
2835 ;; literal containing it.
2836 (setq good-pos (c-state-lit-beg (c-point 'bopl here-bol)))
2837 (list good-pos (and dropped-cons good-pos) nil)))))
0ec1d2c5
AM
2838
2839
2840;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2841;; Externally visible routines.
2842
2843(defun c-state-cache-init ()
2844 (setq c-state-cache nil
2845 c-state-cache-good-pos 1
2846 c-state-nonlit-pos-cache nil
2847 c-state-nonlit-pos-cache-limit 1
2848 c-state-brace-pair-desert nil
2849 c-state-point-min 1
2850 c-state-point-min-lit-type nil
2851 c-state-point-min-lit-start nil
2852 c-state-min-scan-pos 1
2853 c-state-old-cpp-beg nil
2854 c-state-old-cpp-end nil)
2855 (c-state-mark-point-min-literal))
2856
e84efb70
AM
2857;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2858;; Debugging routines to dump `c-state-cache' in a "replayable" form.
2859;; (defmacro c-sc-de (elt) ; "c-state-cache-dump-element"
2860;; `(format ,(concat "(setq " (symbol-name elt) " %s) ") ,elt))
2861;; (defmacro c-sc-qde (elt) ; "c-state-cache-quote-dump-element"
2862;; `(format ,(concat "(setq " (symbol-name elt) " '%s) ") ,elt))
2863;; (defun c-state-dump ()
2864;; ;; For debugging.
2865;; ;(message
2866;; (concat
2867;; (c-sc-qde c-state-cache)
2868;; (c-sc-de c-state-cache-good-pos)
2869;; (c-sc-qde c-state-nonlit-pos-cache)
2870;; (c-sc-de c-state-nonlit-pos-cache-limit)
2871;; (c-sc-qde c-state-brace-pair-desert)
2872;; (c-sc-de c-state-point-min)
2873;; (c-sc-de c-state-point-min-lit-type)
2874;; (c-sc-de c-state-point-min-lit-start)
2875;; (c-sc-de c-state-min-scan-pos)
2876;; (c-sc-de c-state-old-cpp-beg)
2877;; (c-sc-de c-state-old-cpp-end)))
2878;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2879
0ec1d2c5
AM
2880(defun c-invalidate-state-cache-1 (here)
2881 ;; Invalidate all info on `c-state-cache' that applies to the buffer at HERE
2882 ;; or higher and set `c-state-cache-good-pos' accordingly. The cache is
2883 ;; left in a consistent state.
2884 ;;
2885 ;; This is much like `c-whack-state-after', but it never changes a paren
2886 ;; pair element into an open paren element. Doing that would mean that the
2887 ;; new open paren wouldn't have the required preceding paren pair element.
785eecbb 2888 ;;
0ec1d2c5
AM
2889 ;; This function is called from c-after-change.
2890
2891 ;; The cache of non-literals:
2892 (if (< here c-state-nonlit-pos-cache-limit)
2893 (setq c-state-nonlit-pos-cache-limit here))
2894
2895 ;; `c-state-cache':
2896 ;; Case 1: if `here' is in a literal containing point-min, everything
2897 ;; becomes (or is already) nil.
2898 (if (or (null c-state-cache-good-pos)
2899 (< here (c-state-get-min-scan-pos)))
2900 (setq c-state-cache nil
2901 c-state-cache-good-pos nil
2902 c-state-min-scan-pos nil)
2903
2904;;; Truncate `c-state-cache' and set `c-state-cache-good-pos' to a value below
2905;;; `here'. To maintain its consistency, we may need to insert a new brace
2906;;; pair.
2907 (let ((here-bol (c-point 'bol here))
2908 too-high-pa ; recorded {/(/[ next above here, or nil.
2909 dropped-cons ; was the last removed element a brace pair?
2910 pa)
2911 ;; The easy bit - knock over-the-top bits off `c-state-cache'.
2912 (while (and c-state-cache
2913 (>= (setq pa (c-state-cache-top-paren)) here))
2914 (setq dropped-cons (consp (car c-state-cache))
2915 too-high-pa (c-state-cache-top-lparen)
2916 c-state-cache (cdr c-state-cache)))
2917
2918 ;; Do we need to add in an earlier brace pair, having lopped one off?
2919 (if (and dropped-cons
2920 (< too-high-pa (+ here c-state-cache-too-far)))
2921 (c-append-lower-brace-pair-to-state-cache too-high-pa here-bol))
2922 (setq c-state-cache-good-pos (or (c-state-cache-after-top-paren)
2923 (c-state-get-min-scan-pos)))))
2924
2925 ;; The brace-pair desert marker:
2926 (when (car c-state-brace-pair-desert)
2927 (if (< here (car c-state-brace-pair-desert))
2928 (setq c-state-brace-pair-desert nil)
2929 (if (< here (cdr c-state-brace-pair-desert))
2930 (setcdr c-state-brace-pair-desert here)))))
2931
2932(defun c-parse-state-1 ()
2933 ;; Find and record all noteworthy parens between some good point earlier in
2934 ;; the file and point. That good point is at least the beginning of the
2935 ;; top-level construct we are in, or the beginning of the preceding
2936 ;; top-level construct if we aren't in one.
2937 ;;
2938 ;; The returned value is a list of the noteworthy parens with the last one
2939 ;; first. If an element in the list is an integer, it's the position of an
2940 ;; open paren (of any type) which has not been closed before the point. If
2941 ;; an element is a cons, it gives the position of a closed BRACE paren
2942 ;; pair[*]; the car is the start brace position and the cdr is the position
2943 ;; following the closing brace. Only the last closed brace paren pair
2944 ;; before each open paren and before the point is recorded, and thus the
2945 ;; state never contains two cons elements in succession. When a close brace
2946 ;; has no matching open brace (e.g., the matching brace is outside the
2947 ;; visible region), it is not represented in the returned value.
2948 ;;
2949 ;; [*] N.B. The close "brace" might be a mismatching close bracket or paren.
2950 ;; This defun explicitly treats mismatching parens/braces/brackets as
2951 ;; matching. It is the open brace which makes it a "brace" pair.
2952 ;;
2953 ;; If POINT is within a macro, open parens and brace pairs within
2954 ;; THIS macro MIGHT be recorded. This depends on whether their
2955 ;; syntactic properties have been suppressed by
2956 ;; `c-neutralize-syntax-in-CPP'. This might need fixing (2008-12-11).
d9e94c22
MS
2957 ;;
2958 ;; Currently no characters which are given paren syntax with the
2959 ;; syntax-table property are recorded, i.e. angle bracket arglist
2960 ;; parens are never present here. Note that this might change.
2961 ;;
0386b551 2962 ;; BUG: This function doesn't cope entirely well with unbalanced
0ec1d2c5
AM
2963 ;; parens in macros. (2008-12-11: this has probably been resolved
2964 ;; by the function `c-neutralize-syntax-in-CPP'.) E.g. in the
2965 ;; following case the brace before the macro isn't balanced with the
2966 ;; one after it:
0386b551
AM
2967 ;;
2968 ;; {
2969 ;; #define X {
2970 ;; }
2971 ;;
0ec1d2c5
AM
2972 ;; Note to maintainers: this function DOES get called with point
2973 ;; within comments and strings, so don't assume it doesn't!
2974 ;;
0386b551 2975 ;; This function might do hidden buffer changes.
0ec1d2c5
AM
2976 (let* ((here (point))
2977 (here-bopl (c-point 'bopl))
2978 strategy ; 'forward, 'backward etc..
2979 ;; Candidate positions to start scanning from:
2980 cache-pos ; highest position below HERE already existing in
2981 ; cache (or 1).
2982 good-pos
2983 start-point
2984 bopl-state
2985 res
2986 scan-backward-pos scan-forward-p) ; used for 'backward.
2987 ;; If POINT-MIN has changed, adjust the cache
2988 (unless (= (point-min) c-state-point-min)
2989 (c-renarrow-state-cache))
2990
2991 ;; Strategy?
2992 (setq res (c-parse-state-get-strategy here c-state-cache-good-pos)
2993 strategy (car res)
2994 cache-pos (cadr res)
2995 start-point (nth 2 res))
2996
2997 (when (eq strategy 'BOD)
2998 (setq c-state-cache nil
2999 c-state-cache-good-pos start-point))
3000
3001 ;; SCAN!
3002 (save-restriction
3003 (cond
3004 ((memq strategy '(forward BOD))
0386b551 3005 (narrow-to-region (point-min) here)
0ec1d2c5
AM
3006 (setq res (c-remove-stale-state-cache start-point here-bopl))
3007 (setq cache-pos (car res)
3008 scan-backward-pos (cadr res)
3009 bopl-state (car (cddr res))) ; will be nil if (< here-bopl
3010 ; start-point)
3011 (if scan-backward-pos
3012 (c-append-lower-brace-pair-to-state-cache scan-backward-pos))
3013 (setq good-pos
3014 (c-append-to-state-cache cache-pos))
3015 (setq c-state-cache-good-pos
3016 (if (and bopl-state
3017 (< good-pos (- here c-state-cache-too-far)))
3018 (c-state-cache-non-literal-place here-bopl bopl-state)
3019 good-pos)))
3020
3021 ((eq strategy 'backward)
3022 (setq res (c-remove-stale-state-cache-backwards here cache-pos)
3023 good-pos (car res)
3024 scan-backward-pos (cadr res)
3025 scan-forward-p (car (cddr res)))
3026 (if scan-backward-pos
3027 (c-append-lower-brace-pair-to-state-cache
3028 scan-backward-pos))
3029 (setq c-state-cache-good-pos
3030 (if scan-forward-p
3031 (progn (narrow-to-region (point-min) here)
3032 (c-append-to-state-cache good-pos))
3033
3034 (c-get-cache-scan-pos good-pos))))
3035
3036 (t ; (eq strategy 'IN-LIT)
3037 (setq c-state-cache nil
3038 c-state-cache-good-pos nil)))))
3039
3040 c-state-cache)
3041
3042(defun c-invalidate-state-cache (here)
3043 ;; This is a wrapper over `c-invalidate-state-cache-1'.
3044 ;;
3045 ;; It suppresses the syntactic effect of the < and > (template) brackets and
3046 ;; of all parens in preprocessor constructs, except for any such construct
3047 ;; containing point. We can then call `c-invalidate-state-cache-1' without
3048 ;; worrying further about macros and template delimiters.
3049 (c-with-<->-as-parens-suppressed
3e8f7d91
AM
3050 (if (and c-state-old-cpp-beg
3051 (< c-state-old-cpp-beg here))
0ec1d2c5 3052 (c-with-all-but-one-cpps-commented-out
3e8f7d91
AM
3053 c-state-old-cpp-beg
3054 (min c-state-old-cpp-end here)
0ec1d2c5
AM
3055 (c-invalidate-state-cache-1 here))
3056 (c-with-cpps-commented-out
3057 (c-invalidate-state-cache-1 here)))))
3058
3059(defun c-parse-state ()
3060 ;; This is a wrapper over `c-parse-state-1'. See that function for a
3061 ;; description of the functionality and return value.
3062 ;;
3063 ;; It suppresses the syntactic effect of the < and > (template) brackets and
3064 ;; of all parens in preprocessor constructs, except for any such construct
3065 ;; containing point. We can then call `c-parse-state-1' without worrying
3066 ;; further about macros and template delimiters.
3067 (let (here-cpp-beg here-cpp-end)
3068 (save-excursion
3069 (when (c-beginning-of-macro)
3070 (setq here-cpp-beg (point))
3071 (unless
3072 (> (setq here-cpp-end (c-syntactic-end-of-macro))
3073 here-cpp-beg)
3074 (setq here-cpp-beg nil here-cpp-end nil))))
3075 ;; FIXME!!! Put in a `condition-case' here to protect the integrity of the
3076 ;; subsystem.
3077 (prog1
3078 (c-with-<->-as-parens-suppressed
3079 (if (and here-cpp-beg (> here-cpp-end here-cpp-beg))
3080 (c-with-all-but-one-cpps-commented-out
3081 here-cpp-beg here-cpp-end
3082 (c-parse-state-1))
3083 (c-with-cpps-commented-out
3084 (c-parse-state-1))))
3e8f7d91
AM
3085 (setq c-state-old-cpp-beg (and here-cpp-beg (copy-marker here-cpp-beg t))
3086 c-state-old-cpp-end (and here-cpp-end (copy-marker here-cpp-end t)))
3087 )))
0ec1d2c5
AM
3088
3089;; Debug tool to catch cache inconsistencies. This is called from
3090;; 000tests.el.
a66cd3ee
MS
3091(defvar c-debug-parse-state nil)
3092(unless (fboundp 'c-real-parse-state)
3093 (fset 'c-real-parse-state (symbol-function 'c-parse-state)))
3094(cc-bytecomp-defun c-real-parse-state)
3095(defun c-debug-parse-state ()
0ec1d2c5 3096 (let ((here (point)) (res1 (c-real-parse-state)) res2)
0386b551 3097 (let ((c-state-cache nil)
0ec1d2c5
AM
3098 (c-state-cache-good-pos 1)
3099 (c-state-nonlit-pos-cache nil)
3100 (c-state-nonlit-pos-cache-limit 1)
3101 (c-state-brace-pair-desert nil)
3102 (c-state-point-min 1)
3103 (c-state-point-min-lit-type nil)
3104 (c-state-point-min-lit-start nil)
3105 (c-state-min-scan-pos 1)
3106 (c-state-old-cpp-beg nil)
3107 (c-state-old-cpp-end nil))
a66cd3ee
MS
3108 (setq res2 (c-real-parse-state)))
3109 (unless (equal res1 res2)
0386b551
AM
3110 ;; The cache can actually go further back due to the ad-hoc way
3111 ;; the first paren is found, so try to whack off a bit of its
3112 ;; start before complaining.
3113 (save-excursion
3114 (goto-char (or (c-least-enclosing-brace res2) (point)))
3115 (c-beginning-of-defun-1)
3116 (while (not (or (bobp) (eq (char-after) ?{)))
3117 (c-beginning-of-defun-1))
3118 (unless (equal (c-whack-state-before (point) res1) res2)
0ec1d2c5 3119 (message (concat "c-parse-state inconsistency at %s: "
0386b551 3120 "using cache: %s, from scratch: %s")
0ec1d2c5 3121 here res1 res2))))
a66cd3ee 3122 res1))
0ec1d2c5 3123
a66cd3ee
MS
3124(defun c-toggle-parse-state-debug (&optional arg)
3125 (interactive "P")
3126 (setq c-debug-parse-state (c-calculate-state arg c-debug-parse-state))
3127 (fset 'c-parse-state (symbol-function (if c-debug-parse-state
3128 'c-debug-parse-state
3129 'c-real-parse-state)))
3130 (c-keep-region-active))
0386b551
AM
3131(when c-debug-parse-state
3132 (c-toggle-parse-state-debug 1))
a66cd3ee 3133
0ec1d2c5 3134\f
d9e94c22
MS
3135(defun c-whack-state-before (bufpos paren-state)
3136 ;; Whack off any state information from PAREN-STATE which lies
3137 ;; before BUFPOS. Not destructive on PAREN-STATE.
d9e94c22
MS
3138 (let* ((newstate (list nil))
3139 (ptr newstate)
3140 car)
3141 (while paren-state
3142 (setq car (car paren-state)
3143 paren-state (cdr paren-state))
3144 (if (< (if (consp car) (car car) car) bufpos)
3145 (setq paren-state nil)
3146 (setcdr ptr (list car))
3147 (setq ptr (cdr ptr))))
3148 (cdr newstate)))
3149
3150(defun c-whack-state-after (bufpos paren-state)
3151 ;; Whack off any state information from PAREN-STATE which lies at or
3152 ;; after BUFPOS. Not destructive on PAREN-STATE.
d9e94c22
MS
3153 (catch 'done
3154 (while paren-state
3155 (let ((car (car paren-state)))
3156 (if (consp car)
3157 ;; just check the car, because in a balanced brace
3158 ;; expression, it must be impossible for the corresponding
3159 ;; close brace to be before point, but the open brace to
3160 ;; be after.
3161 (if (<= bufpos (car car))
3162 nil ; whack it off
3163 (if (< bufpos (cdr car))
3164 ;; its possible that the open brace is before
3165 ;; bufpos, but the close brace is after. In that
3166 ;; case, convert this to a non-cons element. The
3167 ;; rest of the state is before bufpos, so we're
3168 ;; done.
3169 (throw 'done (cons (car car) (cdr paren-state)))
3170 ;; we know that both the open and close braces are
3171 ;; before bufpos, so we also know that everything else
3172 ;; on state is before bufpos.
3173 (throw 'done paren-state)))
3174 (if (<= bufpos car)
3175 nil ; whack it off
3176 ;; it's before bufpos, so everything else should too.
3177 (throw 'done paren-state)))
3178 (setq paren-state (cdr paren-state)))
3179 nil)))
3180
3181(defun c-most-enclosing-brace (paren-state &optional bufpos)
3182 ;; Return the bufpos of the innermost enclosing open paren before
0386b551 3183 ;; bufpos, or nil if none was found.
d9e94c22
MS
3184 (let (enclosingp)
3185 (or bufpos (setq bufpos 134217727))
3186 (while paren-state
3187 (setq enclosingp (car paren-state)
3188 paren-state (cdr paren-state))
3189 (if (or (consp enclosingp)
3190 (>= enclosingp bufpos))
3191 (setq enclosingp nil)
d9e94c22
MS
3192 (setq paren-state nil)))
3193 enclosingp))
3194
0386b551
AM
3195(defun c-least-enclosing-brace (paren-state)
3196 ;; Return the bufpos of the outermost enclosing open paren, or nil
3197 ;; if none was found.
d9e94c22 3198 (let (pos elem)
d9e94c22
MS
3199 (while paren-state
3200 (setq elem (car paren-state)
3201 paren-state (cdr paren-state))
0386b551
AM
3202 (if (integerp elem)
3203 (setq pos elem)))
d9e94c22
MS
3204 pos))
3205
3206(defun c-safe-position (bufpos paren-state)
0386b551
AM
3207 ;; Return the closest "safe" position recorded on PAREN-STATE that
3208 ;; is higher up than BUFPOS. Return nil if PAREN-STATE doesn't
3209 ;; contain any. Return nil if BUFPOS is nil, which is useful to
3210 ;; find the closest limit before a given limit that might be nil.
d9e94c22 3211 ;;
0386b551
AM
3212 ;; A "safe" position is a position at or after a recorded open
3213 ;; paren, or after a recorded close paren. The returned position is
3214 ;; thus either the first position after a close brace, or the first
3215 ;; position after an enclosing paren, or at the enclosing paren in
3216 ;; case BUFPOS is immediately after it.
d9e94c22
MS
3217 (when bufpos
3218 (let (elem)
3219 (catch 'done
3220 (while paren-state
3221 (setq elem (car paren-state))
3222 (if (consp elem)
3223 (cond ((< (cdr elem) bufpos)
3224 (throw 'done (cdr elem)))
3225 ((< (car elem) bufpos)
3226 ;; See below.
3227 (throw 'done (min (1+ (car elem)) bufpos))))
3228 (if (< elem bufpos)
3229 ;; elem is the position at and not after the opening paren, so
3230 ;; we can go forward one more step unless it's equal to
3231 ;; bufpos. This is useful in some cases avoid an extra paren
3232 ;; level between the safe position and bufpos.
3233 (throw 'done (min (1+ elem) bufpos))))
3234 (setq paren-state (cdr paren-state)))))))
3235
3236(defun c-beginning-of-syntax ()
3237 ;; This is used for `font-lock-beginning-of-syntax-function'. It
3238 ;; goes to the closest previous point that is known to be outside
3239 ;; any string literal or comment. `c-state-cache' is used if it has
3240 ;; a position in the vicinity.
3241 (let* ((paren-state c-state-cache)
3242 elem
3243
3244 (pos (catch 'done
3245 ;; Note: Similar code in `c-safe-position'. The
3246 ;; difference is that we accept a safe position at
3247 ;; the point and don't bother to go forward past open
3248 ;; parens.
3249 (while paren-state
3250 (setq elem (car paren-state))
3251 (if (consp elem)
3252 (cond ((<= (cdr elem) (point))
3253 (throw 'done (cdr elem)))
3254 ((<= (car elem) (point))
3255 (throw 'done (car elem))))
3256 (if (<= elem (point))
3257 (throw 'done elem)))
3258 (setq paren-state (cdr paren-state)))
3259 (point-min))))
3260
3261 (if (> pos (- (point) 4000))
3262 (goto-char pos)
3263 ;; The position is far back. Try `c-beginning-of-defun-1'
3264 ;; (although we can't be entirely sure it will go to a position
3265 ;; outside a comment or string in current emacsen). FIXME:
3266 ;; Consult `syntax-ppss' here.
3267 (c-beginning-of-defun-1)
3268 (if (< (point) pos)
3269 (goto-char pos)))))
3270
3271\f
3272;; Tools for scanning identifiers and other tokens.
3273
3274(defun c-on-identifier ()
3275 "Return non-nil if the point is on or directly after an identifier.
3276Keywords are recognized and not considered identifiers. If an
3277identifier is detected, the returned value is its starting position.
0386b551
AM
3278If an identifier ends at the point and another begins at it \(can only
3279happen in Pike) then the point for the preceding one is returned.
d9e94c22 3280
0386b551
AM
3281Note that this function might do hidden buffer changes. See the
3282comment at the start of cc-engine.el for more info."
3283
3284 ;; FIXME: Shouldn't this function handle "operator" in C++?
d9e94c22
MS
3285
3286 (save-excursion
0386b551
AM
3287 (skip-syntax-backward "w_")
3288
3289 (or
3290
3291 ;; Check for a normal (non-keyword) identifier.
3292 (and (looking-at c-symbol-start)
3293 (not (looking-at c-keywords-regexp))
3294 (point))
3295
3296 (when (c-major-mode-is 'pike-mode)
3297 ;; Handle the `<operator> syntax in Pike.
3298 (let ((pos (point)))
3299 (skip-chars-backward "-!%&*+/<=>^|~[]()")
3300 (and (if (< (skip-chars-backward "`") 0)
3301 t
3302 (goto-char pos)
3303 (eq (char-after) ?\`))
3304 (looking-at c-symbol-key)
3305 (>= (match-end 0) pos)
3306 (point))))
3307
3308 ;; Handle the "operator +" syntax in C++.
3309 (when (and c-overloadable-operators-regexp
3310 (= (c-backward-token-2 0) 0))
3311
3312 (cond ((and (looking-at c-overloadable-operators-regexp)
51c9af45 3313 (or (not c-opt-op-identifier-prefix)
0386b551 3314 (and (= (c-backward-token-2 1) 0)
51c9af45 3315 (looking-at c-opt-op-identifier-prefix))))
0386b551
AM
3316 (point))
3317
3318 ((save-excursion
51c9af45
AM
3319 (and c-opt-op-identifier-prefix
3320 (looking-at c-opt-op-identifier-prefix)
0386b551
AM
3321 (= (c-forward-token-2 1) 0)
3322 (looking-at c-overloadable-operators-regexp)))
3323 (point))))
3324
3325 )))
d9e94c22
MS
3326
3327(defsubst c-simple-skip-symbol-backward ()
3328 ;; If the point is at the end of a symbol then skip backward to the
3329 ;; beginning of it. Don't move otherwise. Return non-nil if point
3330 ;; moved.
0386b551
AM
3331 ;;
3332 ;; This function might do hidden buffer changes.
d9e94c22
MS
3333 (or (< (skip-syntax-backward "w_") 0)
3334 (and (c-major-mode-is 'pike-mode)
3335 ;; Handle the `<operator> syntax in Pike.
3336 (let ((pos (point)))
2a15eb73 3337 (if (and (< (skip-chars-backward "-!%&*+/<=>^|~[]()") 0)
d9e94c22
MS
3338 (< (skip-chars-backward "`") 0)
3339 (looking-at c-symbol-key)
3340 (>= (match-end 0) pos))
3341 t
3342 (goto-char pos)
3343 nil)))))
3344
0386b551 3345(defun c-beginning-of-current-token (&optional back-limit)
d9e94c22
MS
3346 ;; Move to the beginning of the current token. Do not move if not
3347 ;; in the middle of one. BACK-LIMIT may be used to bound the
3348 ;; backward search; if given it's assumed to be at the boundary
a85fd6da 3349 ;; between two tokens. Return non-nil if the point is moved, nil
580fba94 3350 ;; otherwise.
0386b551
AM
3351 ;;
3352 ;; This function might do hidden buffer changes.
d9e94c22 3353 (let ((start (point)))
580fba94
AM
3354 (if (looking-at "\\w\\|\\s_")
3355 (skip-syntax-backward "w_" back-limit)
3356 (when (< (skip-syntax-backward ".()" back-limit) 0)
3357 (while (let ((pos (or (and (looking-at c-nonsymbol-token-regexp)
3358 (match-end 0))
3359 ;; `c-nonsymbol-token-regexp' should always match
3360 ;; since we've skipped backward over punctuator
3361 ;; or paren syntax, but consume one char in case
3362 ;; it doesn't so that we don't leave point before
3363 ;; some earlier incorrect token.
3364 (1+ (point)))))
3365 (if (<= pos start)
3366 (goto-char pos))))))
3367 (< (point) start)))
d9e94c22 3368
ff959bab 3369(defun c-end-of-current-token (&optional back-limit)
d9e94c22
MS
3370 ;; Move to the end of the current token. Do not move if not in the
3371 ;; middle of one. BACK-LIMIT may be used to bound the backward
3372 ;; search; if given it's assumed to be at the boundary between two
ff959bab 3373 ;; tokens. Return non-nil if the point is moved, nil otherwise.
0386b551
AM
3374 ;;
3375 ;; This function might do hidden buffer changes.
d9e94c22
MS
3376 (let ((start (point)))
3377 (cond ((< (skip-syntax-backward "w_" (1- start)) 0)
3378 (skip-syntax-forward "w_"))
3379 ((< (skip-syntax-backward ".()" back-limit) 0)
3380 (while (progn
3381 (if (looking-at c-nonsymbol-token-regexp)
3382 (goto-char (match-end 0))
3383 ;; `c-nonsymbol-token-regexp' should always match since
3384 ;; we've skipped backward over punctuator or paren
3385 ;; syntax, but move forward in case it doesn't so that
3386 ;; we don't leave point earlier than we started with.
3387 (forward-char))
ff959bab
MS
3388 (< (point) start)))))
3389 (> (point) start)))
d9e94c22
MS
3390
3391(defconst c-jump-syntax-balanced
3392 (if (memq 'gen-string-delim c-emacs-features)
3393 "\\w\\|\\s_\\|\\s\(\\|\\s\)\\|\\s\"\\|\\s|"
3394 "\\w\\|\\s_\\|\\s\(\\|\\s\)\\|\\s\""))
3395
3396(defconst c-jump-syntax-unbalanced
3397 (if (memq 'gen-string-delim c-emacs-features)
3398 "\\w\\|\\s_\\|\\s\"\\|\\s|"
3399 "\\w\\|\\s_\\|\\s\""))
3400
3401(defun c-forward-token-2 (&optional count balanced limit)
3402 "Move forward by tokens.
3403A token is defined as all symbols and identifiers which aren't
3404syntactic whitespace \(note that multicharacter tokens like \"==\" are
3405treated properly). Point is always either left at the beginning of a
3406token or not moved at all. COUNT specifies the number of tokens to
3407move; a negative COUNT moves in the opposite direction. A COUNT of 0
3408moves to the next token beginning only if not already at one. If
3409BALANCED is true, move over balanced parens, otherwise move into them.
3410Also, if BALANCED is true, never move out of an enclosing paren.
3411
3412LIMIT sets the limit for the movement and defaults to the point limit.
3413The case when LIMIT is set in the middle of a token, comment or macro
3414is handled correctly, i.e. the point won't be left there.
3415
3416Return the number of tokens left to move \(positive or negative). If
3417BALANCED is true, a move over a balanced paren counts as one. Note
3418that if COUNT is 0 and no appropriate token beginning is found, 1 will
3419be returned. Thus, a return value of 0 guarantees that point is at
3420the requested position and a return value less \(without signs) than
0386b551
AM
3421COUNT guarantees that point is at the beginning of some token.
3422
3423Note that this function might do hidden buffer changes. See the
3424comment at the start of cc-engine.el for more info."
d9e94c22
MS
3425
3426 (or count (setq count 1))
3427 (if (< count 0)
3428 (- (c-backward-token-2 (- count) balanced limit))
3429
3430 (let ((jump-syntax (if balanced
3431 c-jump-syntax-balanced
3432 c-jump-syntax-unbalanced))
3433 (last (point))
3434 (prev (point)))
3435
3436 (if (zerop count)
3437 ;; If count is zero we should jump if in the middle of a token.
3438 (c-end-of-current-token))
3439
3440 (save-restriction
3441 (if limit (narrow-to-region (point-min) limit))
3442 (if (/= (point)
3443 (progn (c-forward-syntactic-ws) (point)))
3444 ;; Skip whitespace. Count this as a move if we did in
3445 ;; fact move.
3446 (setq count (max (1- count) 0)))
3447
3448 (if (eobp)
3449 ;; Moved out of bounds. Make sure the returned count isn't zero.
3450 (progn
3451 (if (zerop count) (setq count 1))
3452 (goto-char last))
3453
3454 ;; Use `condition-case' to avoid having the limit tests
3455 ;; inside the loop.
3456 (condition-case nil
3457 (while (and
3458 (> count 0)
3459 (progn
3460 (setq last (point))
3461 (cond ((looking-at jump-syntax)
3462 (goto-char (scan-sexps (point) 1))
3463 t)
3464 ((looking-at c-nonsymbol-token-regexp)
3465 (goto-char (match-end 0))
3466 t)
3467 ;; `c-nonsymbol-token-regexp' above should always
3468 ;; match if there are correct tokens. Try to
3469 ;; widen to see if the limit was set in the
3470 ;; middle of one, else fall back to treating
3471 ;; the offending thing as a one character token.
3472 ((and limit
3473 (save-restriction
3474 (widen)
3475 (looking-at c-nonsymbol-token-regexp)))
3476 nil)
3477 (t
3478 (forward-char)
3479 t))))
3480 (c-forward-syntactic-ws)
3481 (setq prev last
3482 count (1- count)))
3483 (error (goto-char last)))
3484
3485 (when (eobp)
3486 (goto-char prev)
3487 (setq count (1+ count)))))
3488
3489 count)))
3490
3491(defun c-backward-token-2 (&optional count balanced limit)
3492 "Move backward by tokens.
3493See `c-forward-token-2' for details."
3494
3495 (or count (setq count 1))
3496 (if (< count 0)
3497 (- (c-forward-token-2 (- count) balanced limit))
3498
3499 (or limit (setq limit (point-min)))
3500 (let ((jump-syntax (if balanced
3501 c-jump-syntax-balanced
3502 c-jump-syntax-unbalanced))
3503 (last (point)))
3504
3505 (if (zerop count)
3506 ;; The count is zero so try to skip to the beginning of the
3507 ;; current token.
3508 (if (> (point)
3509 (progn (c-beginning-of-current-token) (point)))
3510 (if (< (point) limit)
3511 ;; The limit is inside the same token, so return 1.
3512 (setq count 1))
3513
3514 ;; We're not in the middle of a token. If there's
3515 ;; whitespace after the point then we must move backward,
3516 ;; so set count to 1 in that case.
3517 (and (looking-at c-syntactic-ws-start)
3518 ;; If we're looking at a '#' that might start a cpp
3519 ;; directive then we have to do a more elaborate check.
3520 (or (/= (char-after) ?#)
3521 (not c-opt-cpp-prefix)
3522 (save-excursion
3523 (and (= (point)
3524 (progn (beginning-of-line)
3525 (looking-at "[ \t]*")
3526 (match-end 0)))
3527 (or (bobp)
3528 (progn (backward-char)
3529 (not (eq (char-before) ?\\)))))))
3530 (setq count 1))))
3531
3532 ;; Use `condition-case' to avoid having to check for buffer
3533 ;; limits in `backward-char', `scan-sexps' and `goto-char' below.
3534 (condition-case nil
3535 (while (and
3536 (> count 0)
3537 (progn
3538 (c-backward-syntactic-ws)
3539 (backward-char)
3540 (if (looking-at jump-syntax)
3541 (goto-char (scan-sexps (1+ (point)) -1))
3542 ;; This can be very inefficient if there's a long
3543 ;; sequence of operator tokens without any separation.
3544 ;; That doesn't happen in practice, anyway.
3545 (c-beginning-of-current-token))
3546 (>= (point) limit)))
3547 (setq last (point)
3548 count (1- count)))
3549 (error (goto-char last)))
3550
3551 (if (< (point) limit)
3552 (goto-char last))
3553
3554 count)))
3555
3556(defun c-forward-token-1 (&optional count balanced limit)
3557 "Like `c-forward-token-2' but doesn't treat multicharacter operator
3558tokens like \"==\" as single tokens, i.e. all sequences of symbol
3559characters are jumped over character by character. This function is
3560for compatibility only; it's only a wrapper over `c-forward-token-2'."
3561 (let ((c-nonsymbol-token-regexp "\\s.\\|\\s\(\\|\\s\)"))
3562 (c-forward-token-2 count balanced limit)))
3563
3564(defun c-backward-token-1 (&optional count balanced limit)
3565 "Like `c-backward-token-2' but doesn't treat multicharacter operator
3566tokens like \"==\" as single tokens, i.e. all sequences of symbol
3567characters are jumped over character by character. This function is
3568for compatibility only; it's only a wrapper over `c-backward-token-2'."
3569 (let ((c-nonsymbol-token-regexp "\\s.\\|\\s\(\\|\\s\)"))
3570 (c-backward-token-2 count balanced limit)))
3571
3572\f
3573;; Tools for doing searches restricted to syntactically relevant text.
3574
3575(defun c-syntactic-re-search-forward (regexp &optional bound noerror
3576 paren-level not-inside-token
3577 lookbehind-submatch)
3578 "Like `re-search-forward', but only report matches that are found
3579in syntactically significant text. I.e. matches in comments, macros
3580or string literals are ignored. The start point is assumed to be
3581outside any comment, macro or string literal, or else the content of
3582that region is taken as syntactically significant text.
3583
3584If PAREN-LEVEL is non-nil, an additional restriction is added to
2a15eb73
MS
3585ignore matches in nested paren sexps. The search will also not go
3586outside the current list sexp, which has the effect that if the point
3587should be moved to BOUND when no match is found \(i.e. NOERROR is
3588neither nil nor t), then it will be at the closing paren if the end of
3589the current list sexp is encountered first.
d9e94c22
MS
3590
3591If NOT-INSIDE-TOKEN is non-nil, matches in the middle of tokens are
3592ignored. Things like multicharacter operators and special symbols
3593\(e.g. \"`()\" in Pike) are handled but currently not floating point
3594constants.
3595
3596If LOOKBEHIND-SUBMATCH is non-nil, it's taken as a number of a
3597subexpression in REGEXP. The end of that submatch is used as the
3598position to check for syntactic significance. If LOOKBEHIND-SUBMATCH
3599isn't used or if that subexpression didn't match then the start
3600position of the whole match is used instead. The \"look behind\"
3601subexpression is never tested before the starting position, so it
3602might be a good idea to include \\=\\= as a match alternative in it.
3603
3604Optimization note: Matches might be missed if the \"look behind\"
2a15eb73 3605subexpression can match the end of nonwhite syntactic whitespace,
d9e94c22 3606i.e. the end of comments or cpp directives. This since the function
2a15eb73
MS
3607skips over such things before resuming the search. It's on the other
3608hand not safe to assume that the \"look behind\" subexpression never
3609matches syntactic whitespace.
3610
3611Bug: Unbalanced parens inside cpp directives are currently not handled
3612correctly \(i.e. they don't get ignored as they should) when
0386b551
AM
3613PAREN-LEVEL is set.
3614
3615Note that this function might do hidden buffer changes. See the
3616comment at the start of cc-engine.el for more info."
d9e94c22
MS
3617
3618 (or bound (setq bound (point-max)))
3619 (if paren-level (setq paren-level -1))
3620
3621 ;;(message "c-syntactic-re-search-forward %s %s %S" (point) bound regexp)
3622
3623 (let ((start (point))
2a15eb73
MS
3624 tmp
3625 ;; Start position for the last search.
3626 search-pos
3627 ;; The `parse-partial-sexp' state between the start position
3628 ;; and the point.
3629 state
3630 ;; The current position after the last state update. The next
3631 ;; `parse-partial-sexp' continues from here.
3632 (state-pos (point))
3633 ;; The position at which to check the state and the state
3634 ;; there. This is separate from `state-pos' since we might
3635 ;; need to back up before doing the next search round.
3636 check-pos check-state
3637 ;; Last position known to end a token.
d9e94c22 3638 (last-token-end-pos (point-min))
2a15eb73
MS
3639 ;; Set when a valid match is found.
3640 found)
d9e94c22
MS
3641
3642 (condition-case err
3643 (while
3644 (and
2a15eb73
MS
3645 (progn
3646 (setq search-pos (point))
3647 (re-search-forward regexp bound noerror))
d9e94c22
MS
3648
3649 (progn
2a15eb73
MS
3650 (setq state (parse-partial-sexp
3651 state-pos (match-beginning 0) paren-level nil state)
3652 state-pos (point))
d9e94c22 3653 (if (setq check-pos (and lookbehind-submatch
2a15eb73
MS
3654 (or (not paren-level)
3655 (>= (car state) 0))
d9e94c22
MS
3656 (match-end lookbehind-submatch)))
3657 (setq check-state (parse-partial-sexp
2a15eb73
MS
3658 state-pos check-pos paren-level nil state))
3659 (setq check-pos state-pos
d9e94c22
MS
3660 check-state state))
3661
2a15eb73
MS
3662 ;; NOTE: If we got a look behind subexpression and get
3663 ;; an insignificant match in something that isn't
d9e94c22
MS
3664 ;; syntactic whitespace (i.e. strings or in nested
3665 ;; parentheses), then we can never skip more than a
2a15eb73
MS
3666 ;; single character from the match start position
3667 ;; (i.e. `state-pos' here) before continuing the
3668 ;; search. That since the look behind subexpression
3669 ;; might match the end of the insignificant region in
3670 ;; the next search.
d9e94c22
MS
3671
3672 (cond
d9e94c22
MS
3673 ((elt check-state 7)
3674 ;; Match inside a line comment. Skip to eol. Use
3675 ;; `re-search-forward' instead of `skip-chars-forward' to get
3676 ;; the right bound behavior.
3677 (re-search-forward "[\n\r]" bound noerror))
3678
3679 ((elt check-state 4)
3680 ;; Match inside a block comment. Skip to the '*/'.
3681 (search-forward "*/" bound noerror))
3682
3683 ((and (not (elt check-state 5))
3684 (eq (char-before check-pos) ?/)
2a15eb73 3685 (not (c-get-char-property (1- check-pos) 'syntax-table))
d9e94c22
MS
3686 (memq (char-after check-pos) '(?/ ?*)))
3687 ;; Match in the middle of the opener of a block or line
3688 ;; comment.
3689 (if (= (char-after check-pos) ?/)
3690 (re-search-forward "[\n\r]" bound noerror)
3691 (search-forward "*/" bound noerror)))
3692
2a15eb73
MS
3693 ;; The last `parse-partial-sexp' above might have
3694 ;; stopped short of the real check position if the end
3695 ;; of the current sexp was encountered in paren-level
3696 ;; mode. The checks above are always false in that
3697 ;; case, and since they can do better skipping in
3698 ;; lookbehind-submatch mode, we do them before
3699 ;; checking the paren level.
3700
3701 ((and paren-level
3702 (/= (setq tmp (car check-state)) 0))
3703 ;; Check the paren level first since we're short of the
3704 ;; syntactic checking position if the end of the
3705 ;; current sexp was encountered by `parse-partial-sexp'.
3706 (if (> tmp 0)
3707
3708 ;; Inside a nested paren sexp.
3709 (if lookbehind-submatch
3710 ;; See the NOTE above.
3711 (progn (goto-char state-pos) t)
3712 ;; Skip out of the paren quickly.
3713 (setq state (parse-partial-sexp state-pos bound 0 nil state)
3714 state-pos (point)))
3715
3716 ;; Have exited the current paren sexp.
3717 (if noerror
3718 (progn
3719 ;; The last `parse-partial-sexp' call above
3720 ;; has left us just after the closing paren
3721 ;; in this case, so we can modify the bound
3722 ;; to leave the point at the right position
3723 ;; upon return.
3724 (setq bound (1- (point)))
3725 nil)
3726 (signal 'search-failed (list regexp)))))
3727
3728 ((setq tmp (elt check-state 3))
3729 ;; Match inside a string.
3730 (if (or lookbehind-submatch
3731 (not (integerp tmp)))
3732 ;; See the NOTE above.
3733 (progn (goto-char state-pos) t)
3734 ;; Skip to the end of the string before continuing.
3735 (let ((ender (make-string 1 tmp)) (continue t))
3736 (while (if (search-forward ender bound noerror)
3737 (progn
3738 (setq state (parse-partial-sexp
3739 state-pos (point) nil nil state)
3740 state-pos (point))
3741 (elt state 3))
3742 (setq continue nil)))
3743 continue)))
d9e94c22
MS
3744
3745 ((save-excursion
3746 (save-match-data
3747 (c-beginning-of-macro start)))
3748 ;; Match inside a macro. Skip to the end of it.
3749 (c-end-of-macro)
3750 (cond ((<= (point) bound) t)
3751 (noerror nil)
2a15eb73 3752 (t (signal 'search-failed (list regexp)))))
d9e94c22 3753
2a15eb73
MS
3754 ((and not-inside-token
3755 (or (< check-pos last-token-end-pos)
3756 (< check-pos
3757 (save-excursion
3758 (goto-char check-pos)
3759 (save-match-data
3760 (c-end-of-current-token last-token-end-pos))
3761 (setq last-token-end-pos (point))))))
3762 ;; Inside a token.
3763 (if lookbehind-submatch
3764 ;; See the NOTE above.
3765 (goto-char state-pos)
3766 (goto-char (min last-token-end-pos bound))))
d9e94c22
MS
3767
3768 (t
3769 ;; A real match.
3770 (setq found t)
2a15eb73
MS
3771 nil)))
3772
3773 ;; Should loop to search again, but take care to avoid
3774 ;; looping on the same spot.
3775 (or (/= search-pos (point))
3776 (if (= (point) bound)
3777 (if noerror
3778 nil
3779 (signal 'search-failed (list regexp)))
3780 (forward-char)
3781 t))))
d9e94c22
MS
3782
3783 (error
3784 (goto-char start)
3785 (signal (car err) (cdr err))))
3786
2a15eb73 3787 ;;(message "c-syntactic-re-search-forward done %s" (or (match-end 0) (point)))
d9e94c22
MS
3788
3789 (if found
3790 (progn
2a15eb73
MS
3791 (goto-char (match-end 0))
3792 (match-end 0))
d9e94c22
MS
3793
3794 ;; Search failed. Set point as appropriate.
2a15eb73
MS
3795 (if (eq noerror t)
3796 (goto-char start)
3797 (goto-char bound))
d9e94c22
MS
3798 nil)))
3799
47641aac
GM
3800(defvar safe-pos-list) ; bound in c-syntactic-skip-backward
3801
d0fcee66
AM
3802(defsubst c-ssb-lit-begin ()
3803 ;; Return the start of the literal point is in, or nil.
3804 ;; We read and write the variables `safe-pos', `safe-pos-list', `state'
3805 ;; bound in the caller.
3806
3807 ;; Use `parse-partial-sexp' from a safe position down to the point to check
3808 ;; if it's outside comments and strings.
3809 (save-excursion
3810 (let ((pos (point)) safe-pos state pps-end-pos)
3811 ;; Pick a safe position as close to the point as possible.
3812 ;;
3813 ;; FIXME: Consult `syntax-ppss' here if our cache doesn't give a good
3814 ;; position.
47641aac 3815
d0fcee66
AM
3816 (while (and safe-pos-list
3817 (> (car safe-pos-list) (point)))
3818 (setq safe-pos-list (cdr safe-pos-list)))
3819 (unless (setq safe-pos (car-safe safe-pos-list))
3820 (setq safe-pos (max (or (c-safe-position
3821 (point) (or c-state-cache
3822 (c-parse-state)))
3823 0)
3824 (point-min))
3825 safe-pos-list (list safe-pos)))
3826
3827 ;; Cache positions along the way to use if we have to back up more. We
3828 ;; cache every closing paren on the same level. If the paren cache is
3829 ;; relevant in this region then we're typically already on the same
3830 ;; level as the target position. Note that we might cache positions
3831 ;; after opening parens in case safe-pos is in a nested list. That's
3832 ;; both uncommon and harmless.
3833 (while (progn
3834 (setq state (parse-partial-sexp
3835 safe-pos pos 0))
3836 (< (point) pos))
3837 (setq safe-pos (point)
3838 safe-pos-list (cons safe-pos safe-pos-list)))
3839
3840 ;; If the state contains the start of the containing sexp we cache that
3841 ;; position too, so that parse-partial-sexp in the next run has a bigger
3842 ;; chance of starting at the same level as the target position and thus
3843 ;; will get more good safe positions into the list.
3844 (if (elt state 1)
3845 (setq safe-pos (1+ (elt state 1))
3846 safe-pos-list (cons safe-pos safe-pos-list)))
3847
3848 (if (or (elt state 3) (elt state 4))
3849 ;; Inside string or comment. Continue search at the
3850 ;; beginning of it.
3851 (elt state 8)))))
3852
0386b551 3853(defun c-syntactic-skip-backward (skip-chars &optional limit paren-level)
d9e94c22
MS
3854 "Like `skip-chars-backward' but only look at syntactically relevant chars,
3855i.e. don't stop at positions inside syntactic whitespace or string
3856literals. Preprocessor directives are also ignored, with the exception
3857of the one that the point starts within, if any. If LIMIT is given,
0386b551
AM
3858it's assumed to be at a syntactically relevant position.
3859
3860If PAREN-LEVEL is non-nil, the function won't stop in nested paren
3861sexps, and the search will also not go outside the current paren sexp.
3862However, if LIMIT or the buffer limit is reached inside a nested paren
3863then the point will be left at the limit.
3864
3865Non-nil is returned if the point moved, nil otherwise.
3866
3867Note that this function might do hidden buffer changes. See the
3868comment at the start of cc-engine.el for more info."
d9e94c22
MS
3869
3870 (let ((start (point))
d0fcee66 3871 state-2
d9e94c22
MS
3872 ;; A list of syntactically relevant positions in descending
3873 ;; order. It's used to avoid scanning repeatedly over
3874 ;; potentially large regions with `parse-partial-sexp' to verify
d0fcee66 3875 ;; each position. Used in `c-ssb-lit-begin'
d9e94c22
MS
3876 safe-pos-list
3877 ;; The result from `c-beginning-of-macro' at the start position or the
3878 ;; start position itself if it isn't within a macro. Evaluated on
3879 ;; demand.
0386b551
AM
3880 start-macro-beg
3881 ;; The earliest position after the current one with the same paren
3882 ;; level. Used only when `paren-level' is set.
d0fcee66 3883 lit-beg
0386b551 3884 (paren-level-pos (point)))
d9e94c22 3885
d0fcee66
AM
3886 (while
3887 (progn
3888 ;; The next loop "tries" to find the end point each time round,
3889 ;; loops when it hasn't succeeded.
3890 (while
3891 (and
3892 (< (skip-chars-backward skip-chars limit) 0)
3893
3894 (let ((pos (point)) state-2 pps-end-pos)
3895
3896 (cond
3897 ;; Don't stop inside a literal
3898 ((setq lit-beg (c-ssb-lit-begin))
3899 (goto-char lit-beg)
3900 t)
3901
3902 ((and paren-level
3903 (save-excursion
3904 (setq state-2 (parse-partial-sexp
3905 pos paren-level-pos -1)
3906 pps-end-pos (point))
3907 (/= (car state-2) 0)))
3908 ;; Not at the right level.
3909
3910 (if (and (< (car state-2) 0)
3911 ;; We stop above if we go out of a paren.
3912 ;; Now check whether it precedes or is
3913 ;; nested in the starting sexp.
3914 (save-excursion
3915 (setq state-2
3916 (parse-partial-sexp
3917 pps-end-pos paren-level-pos
3918 nil nil state-2))
3919 (< (car state-2) 0)))
3920
3921 ;; We've stopped short of the starting position
3922 ;; so the hit was inside a nested list. Go up
3923 ;; until we are at the right level.
3924 (condition-case nil
3925 (progn
3926 (goto-char (scan-lists pos -1
3927 (- (car state-2))))
3928 (setq paren-level-pos (point))
3929 (if (and limit (>= limit paren-level-pos))
0386b551 3930 (progn
d0fcee66
AM
3931 (goto-char limit)
3932 nil)
3933 t))
3934 (error
3935 (goto-char (or limit (point-min)))
3936 nil))
3937
3938 ;; The hit was outside the list at the start
3939 ;; position. Go to the start of the list and exit.
3940 (goto-char (1+ (elt state-2 1)))
3941 nil))
3942
3943 ((c-beginning-of-macro limit)
3944 ;; Inside a macro.
3945 (if (< (point)
3946 (or start-macro-beg
3947 (setq start-macro-beg
3948 (save-excursion
3949 (goto-char start)
3950 (c-beginning-of-macro limit)
3951 (point)))))
3952 t
3953
3954 ;; It's inside the same macro we started in so it's
3955 ;; a relevant match.
3956 (goto-char pos)
3957 nil))))))
3958
3959 (> (point)
3960 (progn
3961 ;; Skip syntactic ws afterwards so that we don't stop at the
3962 ;; end of a comment if `skip-chars' is something like "^/".
3963 (c-backward-syntactic-ws)
3964 (point)))))
d9e94c22 3965
0386b551
AM
3966 ;; We might want to extend this with more useful return values in
3967 ;; the future.
3968 (/= (point) start)))
3969
3970;; The following is an alternative implementation of
3971;; `c-syntactic-skip-backward' that uses backward movement to keep
3972;; track of the syntactic context. It turned out to be generally
3973;; slower than the one above which uses forward checks from earlier
3974;; safe positions.
3975;;
3976;;(defconst c-ssb-stop-re
3977;; ;; The regexp matching chars `c-syntactic-skip-backward' needs to
3978;; ;; stop at to avoid going into comments and literals.
3979;; (concat
3980;; ;; Match comment end syntax and string literal syntax. Also match
3981;; ;; '/' for block comment endings (not covered by comment end
3982;; ;; syntax).
3983;; "\\s>\\|/\\|\\s\""
3984;; (if (memq 'gen-string-delim c-emacs-features)
3985;; "\\|\\s|"
3986;; "")
3987;; (if (memq 'gen-comment-delim c-emacs-features)
3988;; "\\|\\s!"
3989;; "")))
3990;;
3991;;(defconst c-ssb-stop-paren-re
3992;; ;; Like `c-ssb-stop-re' but also stops at paren chars.
3993;; (concat c-ssb-stop-re "\\|\\s(\\|\\s)"))
3994;;
3995;;(defconst c-ssb-sexp-end-re
3996;; ;; Regexp matching the ending syntax of a complex sexp.
3997;; (concat c-string-limit-regexp "\\|\\s)"))
3998;;
3999;;(defun c-syntactic-skip-backward (skip-chars &optional limit paren-level)
4000;; "Like `skip-chars-backward' but only look at syntactically relevant chars,
4001;;i.e. don't stop at positions inside syntactic whitespace or string
4002;;literals. Preprocessor directives are also ignored. However, if the
4003;;point is within a comment, string literal or preprocessor directory to
4004;;begin with, its contents is treated as syntactically relevant chars.
4005;;If LIMIT is given, it limits the backward search and the point will be
4006;;left there if no earlier position is found.
4007;;
4008;;If PAREN-LEVEL is non-nil, the function won't stop in nested paren
4009;;sexps, and the search will also not go outside the current paren sexp.
4010;;However, if LIMIT or the buffer limit is reached inside a nested paren
4011;;then the point will be left at the limit.
4012;;
4013;;Non-nil is returned if the point moved, nil otherwise.
4014;;
4015;;Note that this function might do hidden buffer changes. See the
4016;;comment at the start of cc-engine.el for more info."
4017;;
4018;; (save-restriction
4019;; (when limit
4020;; (narrow-to-region limit (point-max)))
4021;;
4022;; (let ((start (point)))
4023;; (catch 'done
4024;; (while (let ((last-pos (point))
4025;; (stop-pos (progn
4026;; (skip-chars-backward skip-chars)
4027;; (point))))
4028;;
4029;; ;; Skip back over the same region as
4030;; ;; `skip-chars-backward' above, but keep to
4031;; ;; syntactically relevant positions.
4032;; (goto-char last-pos)
4033;; (while (and
4034;; ;; `re-search-backward' with a single char regexp
4035;; ;; should be fast.
4036;; (re-search-backward
4037;; (if paren-level c-ssb-stop-paren-re c-ssb-stop-re)
4038;; stop-pos 'move)
4039;;
4040;; (progn
4041;; (cond
4042;; ((looking-at "\\s(")
4043;; ;; `paren-level' is set and we've found the
4044;; ;; start of the containing paren.
4045;; (forward-char)
4046;; (throw 'done t))
4047;;
4048;; ((looking-at c-ssb-sexp-end-re)
4049;; ;; We're at the end of a string literal or paren
4050;; ;; sexp (if `paren-level' is set).
4051;; (forward-char)
4052;; (condition-case nil
4053;; (c-backward-sexp)
4054;; (error
4055;; (goto-char limit)
4056;; (throw 'done t))))
4057;;
4058;; (t
4059;; (forward-char)
4060;; ;; At the end of some syntactic ws or possibly
4061;; ;; after a plain '/' operator.
4062;; (let ((pos (point)))
4063;; (c-backward-syntactic-ws)
4064;; (if (= pos (point))
4065;; ;; Was a plain '/' operator. Go past it.
4066;; (backward-char)))))
4067;;
4068;; (> (point) stop-pos))))
4069;;
4070;; ;; Now the point is either at `stop-pos' or at some
4071;; ;; position further back if `stop-pos' was at a
4072;; ;; syntactically irrelevant place.
4073;;
4074;; ;; Skip additional syntactic ws so that we don't stop
4075;; ;; at the end of a comment if `skip-chars' is
4076;; ;; something like "^/".
4077;; (c-backward-syntactic-ws)
4078;;
4079;; (< (point) stop-pos))))
4080;;
4081;; ;; We might want to extend this with more useful return values
4082;; ;; in the future.
4083;; (/= (point) start))))
d9e94c22
MS
4084
4085\f
4086;; Tools for handling comments and string literals.
4087
4088(defun c-slow-in-literal (&optional lim detect-cpp)
4089 "Return the type of literal point is in, if any.
4090The return value is `c' if in a C-style comment, `c++' if in a C++
4091style comment, `string' if in a string literal, `pound' if DETECT-CPP
4092is non-nil and in a preprocessor line, or nil if somewhere else.
4093Optional LIM is used as the backward limit of the search. If omitted,
4094or nil, `c-beginning-of-defun' is used.
4095
4096The last point calculated is cached if the cache is enabled, i.e. if
4097`c-in-literal-cache' is bound to a two element vector.
4098
0386b551
AM
4099Note that this function might do hidden buffer changes. See the
4100comment at the start of cc-engine.el for more info."
4101
d9e94c22
MS
4102 (if (and (vectorp c-in-literal-cache)
4103 (= (point) (aref c-in-literal-cache 0)))
4104 (aref c-in-literal-cache 1)
4105 (let ((rtn (save-excursion
4106 (let* ((pos (point))
4107 (lim (or lim (progn
4108 (c-beginning-of-syntax)
4109 (point))))
4110 (state (parse-partial-sexp lim pos)))
4111 (cond
4112 ((elt state 3) 'string)
4113 ((elt state 4) (if (elt state 7) 'c++ 'c))
4114 ((and detect-cpp (c-beginning-of-macro lim)) 'pound)
4115 (t nil))))))
4116 ;; cache this result if the cache is enabled
4117 (if (not c-in-literal-cache)
4118 (setq c-in-literal-cache (vector (point) rtn)))
4119 rtn)))
4120
4121;; XEmacs has a built-in function that should make this much quicker.
4122;; I don't think we even need the cache, which makes our lives more
4123;; complicated anyway. In this case, lim is only used to detect
4124;; cpp directives.
4125;;
4126;; Note that there is a bug in Xemacs's buffer-syntactic-context when used in
4127;; conjunction with syntax-table-properties. The bug is present in, e.g.,
4128;; Xemacs 21.4.4. It manifested itself thus:
4129;;
4130;; Starting with an empty AWK Mode buffer, type
4131;; /regexp/ {<C-j>
4132;; Point gets wrongly left at column 0, rather than being indented to tab-width.
4133;;
4134;; AWK Mode is designed such that when the first / is typed, it gets the
4135;; syntax-table property "string fence". When the second / is typed, BOTH /s
4136;; are given the s-t property "string". However, buffer-syntactic-context
4137;; fails to take account of the change of the s-t property on the opening / to
4138;; "string", and reports that the { is within a string started by the second /.
4139;;
4140;; The workaround for this is for the AWK Mode initialisation to switch the
4141;; defalias for c-in-literal to c-slow-in-literal. This will slow down other
4142;; cc-modes in Xemacs whenever an awk-buffer has been initialised.
b414f371 4143;;
d9e94c22
MS
4144;; (Alan Mackenzie, 2003/4/30).
4145
4146(defun c-fast-in-literal (&optional lim detect-cpp)
0386b551 4147 ;; This function might do hidden buffer changes.
d9e94c22
MS
4148 (let ((context (buffer-syntactic-context)))
4149 (cond
4150 ((eq context 'string) 'string)
4151 ((eq context 'comment) 'c++)
4152 ((eq context 'block-comment) 'c)
4153 ((and detect-cpp (save-excursion (c-beginning-of-macro lim))) 'pound))))
4154
4155(defalias 'c-in-literal
4156 (if (fboundp 'buffer-syntactic-context)
7bfc3fdb 4157 'c-fast-in-literal ; XEmacs
d9e94c22
MS
4158 'c-slow-in-literal)) ; GNU Emacs
4159
4160;; The defalias above isn't enough to shut up the byte compiler.
4161(cc-bytecomp-defun c-in-literal)
4162
4163(defun c-literal-limits (&optional lim near not-in-delimiter)
4164 "Return a cons of the beginning and end positions of the comment or
4165string surrounding point (including both delimiters), or nil if point
4166isn't in one. If LIM is non-nil, it's used as the \"safe\" position
4167to start parsing from. If NEAR is non-nil, then the limits of any
4168literal next to point is returned. \"Next to\" means there's only
4169spaces and tabs between point and the literal. The search for such a
4170literal is done first in forward direction. If NOT-IN-DELIMITER is
4171non-nil, the case when point is inside a starting delimiter won't be
a85fd6da 4172recognized. This only has effect for comments which have starting
d9e94c22
MS
4173delimiters with more than one character.
4174
0386b551
AM
4175Note that this function might do hidden buffer changes. See the
4176comment at the start of cc-engine.el for more info."
d9e94c22
MS
4177
4178 (save-excursion
4179 (let* ((pos (point))
4180 (lim (or lim (progn
4181 (c-beginning-of-syntax)
4182 (point))))
4183 (state (parse-partial-sexp lim pos)))
4184
0386b551
AM
4185 (cond ((elt state 3) ; String.
4186 (goto-char (elt state 8))
d9e94c22
MS
4187 (cons (point) (or (c-safe (c-forward-sexp 1) (point))
4188 (point-max))))
4189
0386b551
AM
4190 ((elt state 4) ; Comment.
4191 (goto-char (elt state 8))
d9e94c22
MS
4192 (cons (point) (progn (c-forward-single-comment) (point))))
4193
4194 ((and (not not-in-delimiter)
4195 (not (elt state 5))
4196 (eq (char-before) ?/)
4197 (looking-at "[/*]"))
4198 ;; We're standing in a comment starter.
4199 (backward-char 1)
4200 (cons (point) (progn (c-forward-single-comment) (point))))
4201
4202 (near
4203 (goto-char pos)
4204
4205 ;; Search forward for a literal.
4206 (skip-chars-forward " \t")
4207
4208 (cond
4209 ((looking-at c-string-limit-regexp) ; String.
4210 (cons (point) (or (c-safe (c-forward-sexp 1) (point))
4211 (point-max))))
4212
4213 ((looking-at c-comment-start-regexp) ; Line or block comment.
4214 (cons (point) (progn (c-forward-single-comment) (point))))
4215
4216 (t
4217 ;; Search backward.
4218 (skip-chars-backward " \t")
4219
4220 (let ((end (point)) beg)
4221 (cond
4222 ((save-excursion
4223 (< (skip-syntax-backward c-string-syntax) 0)) ; String.
4224 (setq beg (c-safe (c-backward-sexp 1) (point))))
4225
4226 ((and (c-safe (forward-char -2) t)
4227 (looking-at "*/"))
4228 ;; Block comment. Due to the nature of line
4229 ;; comments, they will always be covered by the
4230 ;; normal case above.
4231 (goto-char end)
4232 (c-backward-single-comment)
4233 ;; If LIM is bogus, beg will be bogus.
4234 (setq beg (point))))
4235
4236 (if beg (cons beg end))))))
4237 ))))
4238
0386b551
AM
4239;; In case external callers use this; it did have a docstring.
4240(defalias 'c-literal-limits-fast 'c-literal-limits)
d9e94c22 4241
0386b551
AM
4242(defun c-collect-line-comments (range)
4243 "If the argument is a cons of two buffer positions (such as returned by
4244`c-literal-limits'), and that range contains a C++ style line comment,
4245then an extended range is returned that contains all adjacent line
4246comments (i.e. all comments that starts in the same column with no
4247empty lines or non-whitespace characters between them). Otherwise the
4248argument is returned.
d9e94c22 4249
0386b551
AM
4250Note that this function might do hidden buffer changes. See the
4251comment at the start of cc-engine.el for more info."
d9e94c22
MS
4252
4253 (save-excursion
0386b551
AM
4254 (condition-case nil
4255 (if (and (consp range) (progn
4256 (goto-char (car range))
4257 (looking-at c-line-comment-starter)))
b414f371 4258 (let ((col (current-column))
0386b551
AM
4259 (beg (point))
4260 (bopl (c-point 'bopl))
4261 (end (cdr range)))
4262 ;; Got to take care in the backward direction to handle
4263 ;; comments which are preceded by code.
4264 (while (and (c-backward-single-comment)
4265 (>= (point) bopl)
4266 (looking-at c-line-comment-starter)
4267 (= col (current-column)))
4268 (setq beg (point)
4269 bopl (c-point 'bopl)))
4270 (goto-char end)
4271 (while (and (progn (skip-chars-forward " \t")
4272 (looking-at c-line-comment-starter))
4273 (= col (current-column))
4274 (prog1 (zerop (forward-line 1))
4275 (setq end (point)))))
4276 (cons beg end))
4277 range)
4278 (error range))))
d9e94c22
MS
4279
4280(defun c-literal-type (range)
4281 "Convenience function that given the result of `c-literal-limits',
a85fd6da
AM
4282returns nil or the type of literal that the range surrounds, one
4283of the symbols 'c, 'c++ or 'string. It's much faster than using
4284`c-in-literal' and is intended to be used when you need both the
4285type of a literal and its limits.
d9e94c22 4286
0386b551
AM
4287Note that this function might do hidden buffer changes. See the
4288comment at the start of cc-engine.el for more info."
4289
d9e94c22
MS
4290 (if (consp range)
4291 (save-excursion
4292 (goto-char (car range))
4293 (cond ((looking-at c-string-limit-regexp) 'string)
4294 ((or (looking-at "//") ; c++ line comment
4295 (and (looking-at "\\s<") ; comment starter
4296 (looking-at "#"))) ; awk comment.
4297 'c++)
4298 (t 'c))) ; Assuming the range is valid.
4299 range))
4300
4301\f
4302;; `c-find-decl-spots' and accompanying stuff.
4303
4304;; Variables used in `c-find-decl-spots' to cache the search done for
4305;; the first declaration in the last call. When that function starts,
4306;; it needs to back up over syntactic whitespace to look at the last
4307;; token before the region being searched. That can sometimes cause
4308;; moves back and forth over a quite large region of comments and
4309;; macros, which would be repeated for each changed character when
4310;; we're called during fontification, since font-lock refontifies the
4311;; current line for each change. Thus it's worthwhile to cache the
4312;; first match.
4313;;
4314;; `c-find-decl-syntactic-pos' is a syntactically relevant position in
4315;; the syntactic whitespace less or equal to some start position.
4316;; There's no cached value if it's nil.
4317;;
4318;; `c-find-decl-match-pos' is the match position if
4319;; `c-find-decl-prefix-search' matched before the syntactic whitespace
4320;; at `c-find-decl-syntactic-pos', or nil if there's no such match.
4321(defvar c-find-decl-syntactic-pos nil)
4322(make-variable-buffer-local 'c-find-decl-syntactic-pos)
4323(defvar c-find-decl-match-pos nil)
4324(make-variable-buffer-local 'c-find-decl-match-pos)
4325
4326(defsubst c-invalidate-find-decl-cache (change-min-pos)
4327 (and c-find-decl-syntactic-pos
4328 (< change-min-pos c-find-decl-syntactic-pos)
4329 (setq c-find-decl-syntactic-pos nil)))
4330
4331; (defface c-debug-decl-spot-face
4332; '((t (:background "Turquoise")))
4333; "Debug face to mark the spots where `c-find-decl-spots' stopped.")
4334; (defface c-debug-decl-sws-face
4335; '((t (:background "Khaki")))
4336; "Debug face to mark the syntactic whitespace between the declaration
4337; spots and the preceding token end.")
4338
4339(defmacro c-debug-put-decl-spot-faces (match-pos decl-pos)
4340 (when (facep 'c-debug-decl-spot-face)
0386b551 4341 `(c-save-buffer-state ((match-pos ,match-pos) (decl-pos ,decl-pos))
d9e94c22
MS
4342 (c-debug-add-face (max match-pos (point-min)) decl-pos
4343 'c-debug-decl-sws-face)
4344 (c-debug-add-face decl-pos (min (1+ decl-pos) (point-max))
4345 'c-debug-decl-spot-face))))
4346(defmacro c-debug-remove-decl-spot-faces (beg end)
4347 (when (facep 'c-debug-decl-spot-face)
0386b551 4348 `(c-save-buffer-state ()
d9e94c22
MS
4349 (c-debug-remove-face ,beg ,end 'c-debug-decl-spot-face)
4350 (c-debug-remove-face ,beg ,end 'c-debug-decl-sws-face))))
4351
4352(defmacro c-find-decl-prefix-search ()
4353 ;; Macro used inside `c-find-decl-spots'. It ought to be a defun,
4354 ;; but it contains lots of free variables that refer to things
4355 ;; inside `c-find-decl-spots'. The point is left at `cfd-match-pos'
4356 ;; if there is a match, otherwise at `cfd-limit'.
0386b551
AM
4357 ;;
4358 ;; This macro might do hidden buffer changes.
d9e94c22
MS
4359
4360 '(progn
4361 ;; Find the next property match position if we haven't got one already.
4362 (unless cfd-prop-match
4363 (save-excursion
4364 (while (progn
4365 (goto-char (next-single-property-change
4366 (point) 'c-type nil cfd-limit))
4367 (and (< (point) cfd-limit)
4368 (not (eq (c-get-char-property (1- (point)) 'c-type)
4369 'c-decl-end)))))
4370 (setq cfd-prop-match (point))))
4371
0386b551
AM
4372 ;; Find the next `c-decl-prefix-or-start-re' match if we haven't
4373 ;; got one already.
d9e94c22 4374 (unless cfd-re-match
0386b551
AM
4375
4376 (if (> cfd-re-match-end (point))
4377 (goto-char cfd-re-match-end))
4378
4379 (while (if (setq cfd-re-match-end
4380 (re-search-forward c-decl-prefix-or-start-re
4381 cfd-limit 'move))
4382
4383 ;; Match. Check if it's inside a comment or string literal.
4384 (c-got-face-at
4385 (if (setq cfd-re-match (match-end 1))
4386 ;; Matched the end of a token preceding a decl spot.
4387 (progn
4388 (goto-char cfd-re-match)
4389 (1- cfd-re-match))
4390 ;; Matched a token that start a decl spot.
4391 (goto-char (match-beginning 0))
4392 (point))
4393 c-literal-faces)
4394
4395 ;; No match. Finish up and exit the loop.
4396 (setq cfd-re-match cfd-limit)
4397 nil)
4398
4399 ;; Skip out of comments and string literals.
d9e94c22
MS
4400 (while (progn
4401 (goto-char (next-single-property-change
0386b551 4402 (point) 'face nil cfd-limit))
d9e94c22 4403 (and (< (point) cfd-limit)
0386b551
AM
4404 (c-got-face-at (point) c-literal-faces)))))
4405
4406 ;; If we matched at the decl start, we have to back up over the
4407 ;; preceding syntactic ws to set `cfd-match-pos' and to catch
4408 ;; any decl spots in the syntactic ws.
d9e94c22 4409 (unless cfd-re-match
0386b551
AM
4410 (c-backward-syntactic-ws)
4411 (setq cfd-re-match (point))))
d9e94c22
MS
4412
4413 ;; Choose whichever match is closer to the start.
4414 (if (< cfd-re-match cfd-prop-match)
4415 (setq cfd-match-pos cfd-re-match
4416 cfd-re-match nil)
4417 (setq cfd-match-pos cfd-prop-match
4418 cfd-prop-match nil))
4419
4420 (goto-char cfd-match-pos)
4421
4422 (when (< cfd-match-pos cfd-limit)
4423 ;; Skip forward past comments only so we don't skip macros.
4424 (c-forward-comments)
4425 ;; Set the position to continue at. We can avoid going over
4426 ;; the comments skipped above a second time, but it's possible
4427 ;; that the comment skipping has taken us past `cfd-prop-match'
4428 ;; since the property might be used inside comments.
4429 (setq cfd-continue-pos (if cfd-prop-match
4430 (min cfd-prop-match (point))
4431 (point))))))
4432
4433(defun c-find-decl-spots (cfd-limit cfd-decl-re cfd-face-checklist cfd-fun)
0386b551 4434 ;; Call CFD-FUN for each possible spot for a declaration, cast or
a85fd6da
AM
4435 ;; label from the point to CFD-LIMIT.
4436 ;;
4437 ;; CFD-FUN is called with point at the start of the spot. It's
4438 ;; passed two arguments: The first is the end position of the token
4439 ;; preceding the spot, or 0 for the implicit match at bob. The
4440 ;; second is a flag that is t when the match is inside a macro. If
4441 ;; CFD-FUN adds `c-decl-end' properties somewhere below the current
4442 ;; spot, it should return non-nil to ensure that the next search
4443 ;; will find them.
0386b551 4444 ;;
a85fd6da 4445 ;; Such a spot is:
0386b551
AM
4446 ;; o The first token after bob.
4447 ;; o The first token after the end of submatch 1 in
4448 ;; `c-decl-prefix-or-start-re' when that submatch matches.
4449 ;; o The start of each `c-decl-prefix-or-start-re' match when
4450 ;; submatch 1 doesn't match.
5a89f0a7 4451 ;; o The first token after the end of each occurrence of the
0386b551
AM
4452 ;; `c-type' text property with the value `c-decl-end', provided
4453 ;; `c-type-decl-end-used' is set.
4454 ;;
4455 ;; Only a spot that match CFD-DECL-RE and whose face is in the
4456 ;; CFD-FACE-CHECKLIST list causes CFD-FUN to be called. The face
4457 ;; check is disabled if CFD-FACE-CHECKLIST is nil.
d9e94c22
MS
4458 ;;
4459 ;; If the match is inside a macro then the buffer is narrowed to the
4460 ;; end of it, so that CFD-FUN can investigate the following tokens
4461 ;; without matching something that begins inside a macro and ends
4462 ;; outside it. It's to avoid this work that the CFD-DECL-RE and
4463 ;; CFD-FACE-CHECKLIST checks exist.
4464 ;;
0386b551
AM
4465 ;; The spots are visited approximately in order from top to bottom.
4466 ;; It's however the positions where `c-decl-prefix-or-start-re'
4467 ;; matches and where `c-decl-end' properties are found that are in
4468 ;; order. Since the spots often are at the following token, they
4469 ;; might be visited out of order insofar as more spots are reported
4470 ;; later on within the syntactic whitespace between the match
4471 ;; positions and their spots.
4472 ;;
4473 ;; It's assumed that comments and strings are fontified in the
d9e94c22
MS
4474 ;; searched range.
4475 ;;
4476 ;; This is mainly used in fontification, and so has an elaborate
4477 ;; cache to handle repeated calls from the same start position; see
4478 ;; the variables above.
4479 ;;
4480 ;; All variables in this function begin with `cfd-' to avoid name
4481 ;; collision with the (dynamically bound) variables used in CFD-FUN.
0386b551
AM
4482 ;;
4483 ;; This function might do hidden buffer changes.
d9e94c22 4484
0386b551
AM
4485 (let ((cfd-start-pos (point))
4486 (cfd-buffer-end (point-max))
4487 ;; The end of the token preceding the decl spot last found
4488 ;; with `c-decl-prefix-or-start-re'. `cfd-limit' if there's
4489 ;; no match.
d9e94c22 4490 cfd-re-match
0386b551
AM
4491 ;; The end position of the last `c-decl-prefix-or-start-re'
4492 ;; match. If this is greater than `cfd-continue-pos', the
4493 ;; next regexp search is started here instead.
4494 (cfd-re-match-end (point-min))
4495 ;; The end of the last `c-decl-end' found by
4496 ;; `c-find-decl-prefix-search'. `cfd-limit' if there's no
4497 ;; match. If searching for the property isn't needed then we
4498 ;; disable it by setting it to `cfd-limit' directly.
d9e94c22 4499 (cfd-prop-match (unless c-type-decl-end-used cfd-limit))
0386b551
AM
4500 ;; The end of the token preceding the decl spot last found by
4501 ;; `c-find-decl-prefix-search'. 0 for the implicit match at
4502 ;; bob. `cfd-limit' if there's no match. In other words,
4503 ;; this is the minimum of `cfd-re-match' and `cfd-prop-match'.
d9e94c22
MS
4504 (cfd-match-pos cfd-limit)
4505 ;; The position to continue searching at.
4506 cfd-continue-pos
4507 ;; The position of the last "real" token we've stopped at.
4508 ;; This can be greater than `cfd-continue-pos' when we get
4509 ;; hits inside macros or at `c-decl-end' positions inside
4510 ;; comments.
4511 (cfd-token-pos 0)
4512 ;; The end position of the last entered macro.
4513 (cfd-macro-end 0))
4514
4515 ;; Initialize by finding a syntactically relevant start position
0386b551
AM
4516 ;; before the point, and do the first `c-decl-prefix-or-start-re'
4517 ;; search unless we're at bob.
d9e94c22 4518
0386b551 4519 (let (start-in-literal start-in-macro syntactic-pos)
d9e94c22
MS
4520 ;; Must back up a bit since we look for the end of the previous
4521 ;; statement or declaration, which is earlier than the first
4522 ;; returned match.
4523
0386b551
AM
4524 (cond
4525 ;; First we need to move to a syntactically relevant position.
4526 ;; Begin by backing out of comment or string literals.
4527 ((and
4528 (when (c-got-face-at (point) c-literal-faces)
4529 ;; Try to use the faces to back up to the start of the
4530 ;; literal. FIXME: What if the point is on a declaration
4531 ;; inside a comment?
4532 (while (and (not (bobp))
4533 (c-got-face-at (1- (point)) c-literal-faces))
4534 (goto-char (previous-single-property-change
4535 (point) 'face nil (point-min))))
4536
4537 ;; XEmacs doesn't fontify the quotes surrounding string
4538 ;; literals.
4539 (and (featurep 'xemacs)
4540 (eq (get-text-property (point) 'face)
4541 'font-lock-string-face)
4542 (not (bobp))
4543 (progn (backward-char)
4544 (not (looking-at c-string-limit-regexp)))
4545 (forward-char))
4546
4547 ;; Don't trust the literal to contain only literal faces
4548 ;; (the font lock package might not have fontified the
4549 ;; start of it at all, for instance) so check that we have
4550 ;; arrived at something that looks like a start or else
4551 ;; resort to `c-literal-limits'.
4552 (unless (looking-at c-literal-start-regexp)
4553 (let ((range (c-literal-limits)))
4554 (if range (goto-char (car range)))))
4555
4556 (setq start-in-literal (point)))
4557
4558 ;; The start is in a literal. If the limit is in the same
4559 ;; one we don't have to find a syntactic position etc. We
4560 ;; only check that if the limit is at or before bonl to save
4561 ;; time; it covers the by far most common case when font-lock
4562 ;; refontifies the current line only.
4563 (<= cfd-limit (c-point 'bonl cfd-start-pos))
4564 (save-excursion
4565 (goto-char cfd-start-pos)
4566 (while (progn
4567 (goto-char (next-single-property-change
4568 (point) 'face nil cfd-limit))
4569 (and (< (point) cfd-limit)
4570 (c-got-face-at (point) c-literal-faces))))
4571 (= (point) cfd-limit)))
4572
4573 ;; Completely inside a literal. Set up variables to trig the
4574 ;; (< cfd-continue-pos cfd-start-pos) case below and it'll
4575 ;; find a suitable start position.
4576 (setq cfd-continue-pos start-in-literal))
4577
4578 ;; Check if the region might be completely inside a macro, to
4579 ;; optimize that like the completely-inside-literal above.
4580 ((save-excursion
4581 (and (= (forward-line 1) 0)
4582 (bolp) ; forward-line has funny behavior at eob.
4583 (>= (point) cfd-limit)
4584 (progn (backward-char)
4585 (eq (char-before) ?\\))))
4586 ;; (Maybe) completely inside a macro. Only need to trig the
4587 ;; (< cfd-continue-pos cfd-start-pos) case below to make it
4588 ;; set things up.
4589 (setq cfd-continue-pos (1- cfd-start-pos)
4590 start-in-macro t))
d9e94c22 4591
0386b551
AM
4592 (t
4593 ;; Back out of any macro so we don't miss any declaration
4594 ;; that could follow after it.
4595 (when (c-beginning-of-macro)
4596 (setq start-in-macro t))
4597
4598 ;; Now we're at a proper syntactically relevant position so we
4599 ;; can use the cache. But first clear it if it applied
4600 ;; further down.
4601 (c-invalidate-find-decl-cache cfd-start-pos)
4602
4603 (setq syntactic-pos (point))
4604 (unless (eq syntactic-pos c-find-decl-syntactic-pos)
4605 ;; Don't have to do this if the cache is relevant here,
4606 ;; typically if the same line is refontified again. If
4607 ;; we're just some syntactic whitespace further down we can
4608 ;; still use the cache to limit the skipping.
4609 (c-backward-syntactic-ws c-find-decl-syntactic-pos))
4610
4611 ;; If we hit `c-find-decl-syntactic-pos' and
4612 ;; `c-find-decl-match-pos' is set then we install the cached
4613 ;; values. If we hit `c-find-decl-syntactic-pos' and
4614 ;; `c-find-decl-match-pos' is nil then we know there's no decl
4615 ;; prefix in the whitespace before `c-find-decl-syntactic-pos'
4616 ;; and so we can continue the search from this point. If we
4617 ;; didn't hit `c-find-decl-syntactic-pos' then we're now in
4618 ;; the right spot to begin searching anyway.
4619 (if (and (eq (point) c-find-decl-syntactic-pos)
4620 c-find-decl-match-pos)
d9e94c22
MS
4621 (setq cfd-match-pos c-find-decl-match-pos
4622 cfd-continue-pos syntactic-pos)
0386b551
AM
4623
4624 (setq c-find-decl-syntactic-pos syntactic-pos)
4625
4626 (when (if (bobp)
4627 ;; Always consider bob a match to get the first
4628 ;; declaration in the file. Do this separately instead of
4629 ;; letting `c-decl-prefix-or-start-re' match bob, so that
4630 ;; regexp always can consume at least one character to
4631 ;; ensure that we won't get stuck in an infinite loop.
4632 (setq cfd-re-match 0)
4633 (backward-char)
4634 (c-beginning-of-current-token)
4635 (< (point) cfd-limit))
4636 ;; Do an initial search now. In the bob case above it's
4637 ;; only done to search for a `c-decl-end' spot.
4638 (c-find-decl-prefix-search))
4639
4640 (setq c-find-decl-match-pos (and (< cfd-match-pos cfd-start-pos)
4641 cfd-match-pos)))))
4642
4643 ;; Advance `cfd-continue-pos' if it's before the start position.
4644 ;; The closest continue position that might have effect at or
4645 ;; after the start depends on what we started in. This also
4646 ;; finds a suitable start position in the special cases when the
4647 ;; region is completely within a literal or macro.
4648 (when (and cfd-continue-pos (< cfd-continue-pos cfd-start-pos))
4649
4650 (cond
4651 (start-in-macro
4652 ;; If we're in a macro then it's the closest preceding token
4653 ;; in the macro. Check this before `start-in-literal',
4654 ;; since if we're inside a literal in a macro, the preceding
4655 ;; token is earlier than any `c-decl-end' spot inside the
4656 ;; literal (comment).
4657 (goto-char (or start-in-literal cfd-start-pos))
4658 ;; The only syntactic ws in macros are comments.
d9e94c22 4659 (c-backward-comments)
0386b551
AM
4660 (backward-char)
4661 (c-beginning-of-current-token))
4662
4663 (start-in-literal
4664 ;; If we're in a comment it can only be the closest
4665 ;; preceding `c-decl-end' position within that comment, if
4666 ;; any. Go back to the beginning of such a property so that
4667 ;; `c-find-decl-prefix-search' will find the end of it.
4668 ;; (Can't stop at the end and install it directly on
4669 ;; `cfd-prop-match' since that variable might be cleared
4670 ;; after `cfd-fun' below.)
4671 ;;
4672 ;; Note that if the literal is a string then the property
4673 ;; search will simply skip to the beginning of it right
4674 ;; away.
4675 (if (not c-type-decl-end-used)
4676 (goto-char start-in-literal)
4677 (goto-char cfd-start-pos)
4678 (while (progn
4679 (goto-char (previous-single-property-change
4680 (point) 'c-type nil start-in-literal))
4681 (and (> (point) start-in-literal)
4682 (not (eq (c-get-char-property (point) 'c-type)
4683 'c-decl-end))))))
4684
4685 (when (= (point) start-in-literal)
4686 ;; Didn't find any property inside the comment, so we can
4687 ;; skip it entirely. (This won't skip past a string, but
4688 ;; that'll be handled quickly by the next
4689 ;; `c-find-decl-prefix-search' anyway.)
4690 (c-forward-single-comment)
4691 (if (> (point) cfd-limit)
4692 (goto-char cfd-limit))))
d9e94c22 4693
0386b551
AM
4694 (t
4695 ;; If we started in normal code, the only match that might
4696 ;; apply before the start is what we already got in
4697 ;; `cfd-match-pos' so we can continue at the start position.
4698 ;; (Note that we don't get here if the first match is below
4699 ;; it.)
4700 (goto-char cfd-start-pos)))
4701
4702 ;; Delete found matches if they are before our new continue
4703 ;; position, so that `c-find-decl-prefix-search' won't back up
4704 ;; to them later on.
4705 (setq cfd-continue-pos (point))
4706 (when (and cfd-re-match (< cfd-re-match cfd-continue-pos))
4707 (setq cfd-re-match nil))
4708 (when (and cfd-prop-match (< cfd-prop-match cfd-continue-pos))
4709 (setq cfd-prop-match nil)))
4710
4711 (if syntactic-pos
4712 ;; This is the normal case and we got a proper syntactic
4713 ;; position. If there's a match then it's always outside
4714 ;; macros and comments, so advance to the next token and set
4715 ;; `cfd-token-pos'. The loop below will later go back using
4716 ;; `cfd-continue-pos' to fix declarations inside the
4717 ;; syntactic ws.
4718 (when (and cfd-match-pos (< cfd-match-pos syntactic-pos))
4719 (goto-char syntactic-pos)
4720 (c-forward-syntactic-ws)
4721 (and cfd-continue-pos
4722 (< cfd-continue-pos (point))
4723 (setq cfd-token-pos (point))))
4724
4725 ;; Have one of the special cases when the region is completely
4726 ;; within a literal or macro. `cfd-continue-pos' is set to a
4727 ;; good start position for the search, so do it.
4728 (c-find-decl-prefix-search)))
d9e94c22 4729
51c9af45 4730 ;; Now loop. Round what? (ACM, 2006/7/5). We already got the first match.
d9e94c22
MS
4731
4732 (while (progn
4733 (while (and
4734 (< cfd-match-pos cfd-limit)
4735
4736 (or
4737 ;; Kludge to filter out matches on the "<" that
4738 ;; aren't open parens, for the sake of languages
4739 ;; that got `c-recognize-<>-arglists' set.
4740 (and (eq (char-before cfd-match-pos) ?<)
4741 (not (c-get-char-property (1- cfd-match-pos)
4742 'syntax-table)))
4743
4744 ;; If `cfd-continue-pos' is less or equal to
4745 ;; `cfd-token-pos', we've got a hit inside a macro
4746 ;; that's in the syntactic whitespace before the last
4747 ;; "real" declaration we've checked. If they're equal
4748 ;; we've arrived at the declaration a second time, so
4749 ;; there's nothing to do.
4750 (= cfd-continue-pos cfd-token-pos)
4751
4752 (progn
4753 ;; If `cfd-continue-pos' is less than `cfd-token-pos'
4754 ;; we're still searching for declarations embedded in
4755 ;; the syntactic whitespace. In that case we need
4756 ;; only to skip comments and not macros, since they
4757 ;; can't be nested, and that's already been done in
4758 ;; `c-find-decl-prefix-search'.
4759 (when (> cfd-continue-pos cfd-token-pos)
4760 (c-forward-syntactic-ws)
4761 (setq cfd-token-pos (point)))
4762
4763 ;; Continue if the following token fails the
4764 ;; CFD-DECL-RE and CFD-FACE-CHECKLIST checks.
4765 (when (or (>= (point) cfd-limit)
4766 (not (looking-at cfd-decl-re))
4767 (and cfd-face-checklist
4768 (not (c-got-face-at
4769 (point) cfd-face-checklist))))
4770 (goto-char cfd-continue-pos)
4771 t)))
4772
4773 (< (point) cfd-limit))
4774 (c-find-decl-prefix-search))
4775
4776 (< (point) cfd-limit))
4777
0386b551
AM
4778 (when (and
4779 (>= (point) cfd-start-pos)
d9e94c22 4780
0386b551
AM
4781 (progn
4782 ;; Narrow to the end of the macro if we got a hit inside
4783 ;; one, to avoid recognizing things that start inside the
4784 ;; macro and end outside it.
4785 (when (> cfd-match-pos cfd-macro-end)
4786 ;; Not in the same macro as in the previous round.
4787 (save-excursion
4788 (goto-char cfd-match-pos)
4789 (setq cfd-macro-end
4790 (if (save-excursion (and (c-beginning-of-macro)
4791 (< (point) cfd-match-pos)))
4792 (progn (c-end-of-macro)
4793 (point))
4794 0))))
4795
4796 (if (zerop cfd-macro-end)
4797 t
4798 (if (> cfd-macro-end (point))
4799 (progn (narrow-to-region (point-min) cfd-macro-end)
4800 t)
4801 ;; The matched token was the last thing in the macro,
4802 ;; so the whole match is bogus.
4803 (setq cfd-macro-end 0)
4804 nil))))
d9e94c22
MS
4805
4806 (c-debug-put-decl-spot-faces cfd-match-pos (point))
0386b551
AM
4807 (if (funcall cfd-fun cfd-match-pos (/= cfd-macro-end 0))
4808 (setq cfd-prop-match nil))
d9e94c22
MS
4809
4810 (when (/= cfd-macro-end 0)
4811 ;; Restore limits if we did macro narrowment above.
4812 (narrow-to-region (point-min) cfd-buffer-end)))
4813
4814 (goto-char cfd-continue-pos)
4815 (if (= cfd-continue-pos cfd-limit)
4816 (setq cfd-match-pos cfd-limit)
4817 (c-find-decl-prefix-search)))))
4818
4819\f
4820;; A cache for found types.
4821
4822;; Buffer local variable that contains an obarray with the types we've
4823;; found. If a declaration is recognized somewhere we record the
4824;; fully qualified identifier in it to recognize it as a type
4825;; elsewhere in the file too. This is not accurate since we do not
4826;; bother with the scoping rules of the languages, but in practice the
4827;; same name is seldom used as both a type and something else in a
4828;; file, and we only use this as a last resort in ambiguous cases (see
0386b551
AM
4829;; `c-forward-decl-or-cast-1').
4830;;
580fba94
AM
4831;; Not every type need be in this cache. However, things which have
4832;; ceased to be types must be removed from it.
4833;;
0386b551
AM
4834;; Template types in C++ are added here too but with the template
4835;; arglist replaced with "<>" in references or "<" for the one in the
4836;; primary type. E.g. the type "Foo<A,B>::Bar<C>" is stored as
4837;; "Foo<>::Bar<". This avoids storing very long strings (since C++
4838;; template specs can be fairly sized programs in themselves) and
4839;; improves the hit ratio (it's a type regardless of the template
4840;; args; it's just not the same type, but we're only interested in
4841;; recognizing types, not telling distinct types apart). Note that
4842;; template types in references are added here too; from the example
4843;; above there will also be an entry "Foo<".
d9e94c22
MS
4844(defvar c-found-types nil)
4845(make-variable-buffer-local 'c-found-types)
4846
4847(defsubst c-clear-found-types ()
4848 ;; Clears `c-found-types'.
d9e94c22
MS
4849 (setq c-found-types (make-vector 53 0)))
4850
4851(defun c-add-type (from to)
4852 ;; Add the given region as a type in `c-found-types'. If the region
4853 ;; doesn't match an existing type but there is a type which is equal
4854 ;; to the given one except that the last character is missing, then
4855 ;; the shorter type is removed. That's done to avoid adding all
4856 ;; prefixes of a type as it's being entered and font locked. This
4857 ;; doesn't cover cases like when characters are removed from a type
4858 ;; or added in the middle. We'd need the position of point when the
4859 ;; font locking is invoked to solve this well.
0386b551
AM
4860 ;;
4861 ;; This function might do hidden buffer changes.
4862 (let ((type (c-syntactic-content from to c-recognize-<>-arglists)))
4863 (unless (intern-soft type c-found-types)
4864 (unintern (substring type 0 -1) c-found-types)
4865 (intern type c-found-types))))
d9e94c22 4866
580fba94
AM
4867(defun c-unfind-type (name)
4868 ;; Remove the "NAME" from c-found-types, if present.
4869 (unintern name c-found-types))
4870
d9e94c22
MS
4871(defsubst c-check-type (from to)
4872 ;; Return non-nil if the given region contains a type in
4873 ;; `c-found-types'.
0386b551
AM
4874 ;;
4875 ;; This function might do hidden buffer changes.
4876 (intern-soft (c-syntactic-content from to c-recognize-<>-arglists)
4877 c-found-types))
d9e94c22
MS
4878
4879(defun c-list-found-types ()
4880 ;; Return all the types in `c-found-types' as a sorted list of
4881 ;; strings.
4882 (let (type-list)
4883 (mapatoms (lambda (type)
4884 (setq type-list (cons (symbol-name type)
4885 type-list)))
4886 c-found-types)
4887 (sort type-list 'string-lessp)))
a66cd3ee 4888
2f42c75f
DN
4889;; Shut up the byte compiler.
4890(defvar c-maybe-stale-found-type)
4891
580fba94
AM
4892(defun c-trim-found-types (beg end old-len)
4893 ;; An after change function which, in conjunction with the info in
4894 ;; c-maybe-stale-found-type (set in c-before-change), removes a type
4895 ;; from `c-found-types', should this type have become stale. For
4896 ;; example, this happens to "foo" when "foo \n bar();" becomes
4897 ;; "foo(); \n bar();". Such stale types, if not removed, foul up
4898 ;; the fontification.
b414f371 4899 ;;
580fba94
AM
4900 ;; Have we, perhaps, added non-ws characters to the front/back of a found
4901 ;; type?
4902 (when (> end beg)
4903 (save-excursion
4904 (when (< end (point-max))
4905 (goto-char end)
4906 (if (and (c-beginning-of-current-token) ; only moves when we started in the middle
4907 (progn (goto-char end)
4908 (c-end-of-current-token)))
4909 (c-unfind-type (buffer-substring-no-properties
4910 end (point)))))
4911 (when (> beg (point-min))
4912 (goto-char beg)
4913 (if (and (c-end-of-current-token) ; only moves when we started in the middle
4914 (progn (goto-char beg)
4915 (c-beginning-of-current-token)))
4916 (c-unfind-type (buffer-substring-no-properties
4917 (point) beg))))))
b414f371 4918
580fba94
AM
4919 (if c-maybe-stale-found-type ; e.g. (c-decl-id-start "foo" 97 107 " (* ooka) " "o")
4920 (cond
4921 ;; Changing the amount of (already existing) whitespace - don't do anything.
4922 ((and (c-partial-ws-p beg end)
4923 (or (= beg end) ; removal of WS
580fba94
AM
4924 (string-match "^[ \t\n\r\f\v]*$" (nth 5 c-maybe-stale-found-type)))))
4925
4926 ;; The syntactic relationship which defined a "found type" has been
4927 ;; destroyed.
4928 ((eq (car c-maybe-stale-found-type) 'c-decl-id-start)
4929 (c-unfind-type (cadr c-maybe-stale-found-type)))
4930;; ((eq (car c-maybe-stale-found-type) 'c-decl-type-start) FIXME!!!
4931 )))
4932
d9e94c22 4933\f
dd969a56
AM
4934;; Setting and removing syntax properties on < and > in languages (C++
4935;; and Java) where they can be template/generic delimiters as well as
4936;; their normal meaning of "less/greater than".
4937
4938;; Normally, < and > have syntax 'punctuation'. When they are found to
4939;; be delimiters, they are marked as such with the category properties
4940;; c-<-as-paren-syntax, c->-as-paren-syntax respectively.
4941
4942;; STRATEGY:
4943;;
4944;; It is impossible to determine with certainty whether a <..> pair in
4945;; C++ is two comparison operators or is template delimiters, unless
4946;; one duplicates a lot of a C++ compiler. For example, the following
4947;; code fragment:
4948;;
4949;; foo (a < b, c > d) ;
4950;;
4951;; could be a function call with two integer parameters (each a
4952;; relational expression), or it could be a constructor for class foo
4953;; taking one parameter d of templated type "a < b, c >". They are
4954;; somewhat easier to distinguish in Java.
4955;;
4956;; The strategy now (2010-01) adopted is to mark and unmark < and
4957;; > IN MATCHING PAIRS ONLY. [Previously, they were marked
4958;; individually when their context so indicated. This gave rise to
4959;; intractible problems when one of a matching pair was deleted, or
4960;; pulled into a literal.]
4961;;
4962;; At each buffer change, the syntax-table properties are removed in a
4963;; before-change function and reapplied, when needed, in an
4964;; after-change function. It is far more important that the
4965;; properties get removed when they they are spurious than that they
4966;; be present when wanted.
4967;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
4968(defun c-clear-<-pair-props (&optional pos)
4969 ;; POS (default point) is at a < character. If it is marked with
4970 ;; open paren syntax-table text property, remove the property,
4971 ;; together with the close paren property on the matching > (if
4972 ;; any).
4973 (save-excursion
4974 (if pos
4975 (goto-char pos)
4976 (setq pos (point)))
4977 (when (equal (c-get-char-property (point) 'syntax-table)
4978 c-<-as-paren-syntax)
4979 (with-syntax-table c-no-parens-syntax-table ; ignore unbalanced [,{,(,..
4980 (c-go-list-forward))
4981 (when (equal (c-get-char-property (1- (point)) 'syntax-table)
4982 c->-as-paren-syntax) ; should always be true.
8a249abc
AM
4983 (c-clear-char-property (1- (point)) 'category))
4984 (c-clear-char-property pos 'category))))
dd969a56
AM
4985
4986(defun c-clear->-pair-props (&optional pos)
4987 ;; POS (default point) is at a > character. If it is marked with
4988 ;; close paren syntax-table property, remove the property, together
4989 ;; with the open paren property on the matching < (if any).
4990 (save-excursion
4991 (if pos
4992 (goto-char pos)
4993 (setq pos (point)))
4994 (when (equal (c-get-char-property (point) 'syntax-table)
4995 c->-as-paren-syntax)
4996 (with-syntax-table c-no-parens-syntax-table ; ignore unbalanced [,{,(,..
4997 (c-go-up-list-backward))
4998 (when (equal (c-get-char-property (point) 'syntax-table)
4999 c-<-as-paren-syntax) ; should always be true.
8a249abc
AM
5000 (c-clear-char-property (point) 'category))
5001 (c-clear-char-property pos 'category))))
dd969a56
AM
5002
5003(defun c-clear-<>-pair-props (&optional pos)
5004 ;; POS (default point) is at a < or > character. If it has an
5005 ;; open/close paren syntax-table property, remove this property both
5006 ;; from the current character and its partner (which will also be
5007 ;; thusly marked).
5008 (cond
5009 ((eq (char-after) ?\<)
5010 (c-clear-<-pair-props pos))
5011 ((eq (char-after) ?\>)
5012 (c-clear->-pair-props pos))
5013 (t (c-benign-error
5014 "c-clear-<>-pair-props called from wrong position"))))
5015
5016(defun c-clear-<-pair-props-if-match-after (lim &optional pos)
5017 ;; POS (default point) is at a < character. If it is both marked
5018 ;; with open/close paren syntax-table property, and has a matching >
5019 ;; (also marked) which is after LIM, remove the property both from
43a91810
AM
5020 ;; the current > and its partner. Return t when this happens, nil
5021 ;; when it doesn't.
dd969a56
AM
5022 (save-excursion
5023 (if pos
5024 (goto-char pos)
5025 (setq pos (point)))
5026 (when (equal (c-get-char-property (point) 'syntax-table)
5027 c-<-as-paren-syntax)
5028 (with-syntax-table c-no-parens-syntax-table ; ignore unbalanced [,{,(,..
5029 (c-go-list-forward))
5030 (when (and (>= (point) lim)
5031 (equal (c-get-char-property (1- (point)) 'syntax-table)
5032 c->-as-paren-syntax)) ; should always be true.
5033 (c-unmark-<->-as-paren (1- (point)))
43a91810
AM
5034 (c-unmark-<->-as-paren pos))
5035 t)))
dd969a56
AM
5036
5037(defun c-clear->-pair-props-if-match-before (lim &optional pos)
5038 ;; POS (default point) is at a > character. If it is both marked
5039 ;; with open/close paren syntax-table property, and has a matching <
5040 ;; (also marked) which is before LIM, remove the property both from
43a91810
AM
5041 ;; the current < and its partner. Return t when this happens, nil
5042 ;; when it doesn't.
dd969a56
AM
5043 (save-excursion
5044 (if pos
5045 (goto-char pos)
5046 (setq pos (point)))
5047 (when (equal (c-get-char-property (point) 'syntax-table)
5048 c->-as-paren-syntax)
5049 (with-syntax-table c-no-parens-syntax-table ; ignore unbalanced [,{,(,..
5050 (c-go-up-list-backward))
5051 (when (and (<= (point) lim)
5052 (equal (c-get-char-property (point) 'syntax-table)
5053 c-<-as-paren-syntax)) ; should always be true.
5054 (c-unmark-<->-as-paren (point))
43a91810
AM
5055 (c-unmark-<->-as-paren pos))
5056 t)))
dd969a56 5057
f3b554af
GM
5058;; Set by c-common-init in cc-mode.el.
5059(defvar c-new-BEG)
5060(defvar c-new-END)
5061
dd969a56
AM
5062(defun c-before-change-check-<>-operators (beg end)
5063 ;; Unmark certain pairs of "< .... >" which are currently marked as
5064 ;; template/generic delimiters. (This marking is via syntax-table
5065 ;; text properties).
5066 ;;
5067 ;; These pairs are those which are in the current "statement" (i.e.,
5068 ;; the region between the {, }, or ; before BEG and the one after
5069 ;; END), and which enclose any part of the interval (BEG END).
5070 ;;
5071 ;; Note that in C++ (?and Java), template/generic parens cannot
5072 ;; enclose a brace or semicolon, so we use these as bounds on the
5073 ;; region we must work on.
5074 ;;
5075 ;; This function is called from before-change-functions (via
5076 ;; c-get-state-before-change-functions). Thus the buffer is widened,
5077 ;; and point is undefined, both at entry and exit.
5078 ;;
5079 ;; FIXME!!! This routine ignores the possibility of macros entirely.
5080 ;; 2010-01-29.
5081 (save-excursion
5082 (let ((beg-lit-limits (progn (goto-char beg) (c-literal-limits)))
43a91810
AM
5083 (end-lit-limits (progn (goto-char end) (c-literal-limits)))
5084 new-beg new-end need-new-beg need-new-end)
dd969a56
AM
5085 ;; Locate the barrier before the changed region
5086 (goto-char (if beg-lit-limits (car beg-lit-limits) beg))
5087 (c-syntactic-skip-backward "^;{}" (max (- beg 2048) (point-min)))
43a91810 5088 (setq new-beg (point))
dd969a56
AM
5089
5090 ;; Remove the syntax-table properties from each pertinent <...> pair.
5091 ;; Firsly, the ones with the < before beg and > after beg.
5092 (while (c-search-forward-char-property 'category 'c-<-as-paren-syntax beg)
43a91810
AM
5093 (if (c-clear-<-pair-props-if-match-after beg (1- (point)))
5094 (setq need-new-beg t)))
dd969a56
AM
5095
5096 ;; Locate the barrier after END.
5097 (goto-char (if end-lit-limits (cdr end-lit-limits) end))
5098 (c-syntactic-re-search-forward "[;{}]"
5099 (min (+ end 2048) (point-max)) 'end)
43a91810 5100 (setq new-end (point))
dd969a56
AM
5101
5102 ;; Remove syntax-table properties from the remaining pertinent <...>
5103 ;; pairs, those with a > after end and < before end.
5104 (while (c-search-backward-char-property 'category 'c->-as-paren-syntax end)
43a91810
AM
5105 (if (c-clear->-pair-props-if-match-before end)
5106 (setq need-new-end t)))
5107
5108 ;; Extend the fontification region, if needed.
5109 (when need-new-beg
5110 (goto-char new-beg)
5111 (c-forward-syntactic-ws)
5112 (and (< (point) c-new-BEG) (setq c-new-BEG (point))))
5113
5114 (when need-new-end
5115 (and (> new-end c-new-END) (setq c-new-END new-end))))))
dd969a56
AM
5116
5117
d9e94c22 5118
0386b551
AM
5119(defun c-after-change-check-<>-operators (beg end)
5120 ;; This is called from `after-change-functions' when
5121 ;; c-recognize-<>-arglists' is set. It ensures that no "<" or ">"
5122 ;; chars with paren syntax become part of another operator like "<<"
5123 ;; or ">=".
5124 ;;
5125 ;; This function might do hidden buffer changes.
5126
cb694ab7
AM
5127 (save-excursion
5128 (goto-char beg)
5129 (when (or (looking-at "[<>]")
5130 (< (skip-chars-backward "<>") 0))
5131
0386b551 5132 (goto-char beg)
cb694ab7
AM
5133 (c-beginning-of-current-token)
5134 (when (and (< (point) beg)
5135 (looking-at c-<>-multichar-token-regexp)
5136 (< beg (setq beg (match-end 0))))
5137 (while (progn (skip-chars-forward "^<>" beg)
5138 (< (point) beg))
dd969a56 5139 (c-clear-<>-pair-props)
cb694ab7
AM
5140 (forward-char))))
5141
5142 (when (< beg end)
5143 (goto-char end)
0386b551
AM
5144 (when (or (looking-at "[<>]")
5145 (< (skip-chars-backward "<>") 0))
5146
cb694ab7 5147 (goto-char end)
0386b551 5148 (c-beginning-of-current-token)
cb694ab7 5149 (when (and (< (point) end)
0386b551 5150 (looking-at c-<>-multichar-token-regexp)
cb694ab7
AM
5151 (< end (setq end (match-end 0))))
5152 (while (progn (skip-chars-forward "^<>" end)
5153 (< (point) end))
dd969a56 5154 (c-clear-<>-pair-props)
cb694ab7 5155 (forward-char)))))))
d9e94c22 5156
dd969a56
AM
5157
5158\f
5159;; Handling of small scale constructs like types and names.
5160
d9e94c22
MS
5161;; Dynamically bound variable that instructs `c-forward-type' to also
5162;; treat possible types (i.e. those that it normally returns 'maybe or
5163;; 'found for) as actual types (and always return 'found for them).
5164;; This means that it records them in `c-record-type-identifiers' if
5165;; that is set, and that it adds them to `c-found-types'.
5166(defvar c-promote-possible-types nil)
5167
0386b551
AM
5168;; Dynamically bound variable that instructs `c-forward-<>-arglist' to
5169;; mark up successfully parsed arglists with paren syntax properties on
5170;; the surrounding angle brackets and with `c-<>-arg-sep' in the
5171;; `c-type' property of each argument separating comma.
5172;;
5173;; Setting this variable also makes `c-forward-<>-arglist' recurse into
5174;; all arglists for side effects (i.e. recording types), otherwise it
5175;; exploits any existing paren syntax properties to quickly jump to the
5176;; end of already parsed arglists.
5177;;
5178;; Marking up the arglists is not the default since doing that correctly
5179;; depends on a proper value for `c-restricted-<>-arglists'.
5180(defvar c-parse-and-markup-<>-arglists nil)
5181
d9e94c22 5182;; Dynamically bound variable that instructs `c-forward-<>-arglist' to
037558bf
MS
5183;; not accept arglists that contain binary operators.
5184;;
5185;; This is primarily used to handle C++ template arglists. C++
5186;; disambiguates them by checking whether the preceding name is a
5187;; template or not. We can't do that, so we assume it is a template
5188;; if it can be parsed as one. That usually works well since
5189;; comparison expressions on the forms "a < b > c" or "a < b, c > d"
5190;; in almost all cases would be pointless.
5191;;
5192;; However, in function arglists, e.g. in "foo (a < b, c > d)", we
5193;; should let the comma separate the function arguments instead. And
5194;; in a context where the value of the expression is taken, e.g. in
5195;; "if (a < b || c > d)", it's probably not a template.
5196(defvar c-restricted-<>-arglists nil)
d9e94c22 5197
0386b551
AM
5198;; Dynamically bound variables that instructs
5199;; `c-forward-keyword-clause', `c-forward-<>-arglist',
5200;; `c-forward-name', `c-forward-type', `c-forward-decl-or-cast-1', and
5201;; `c-forward-label' to record the ranges of all the type and
5202;; reference identifiers they encounter. They will build lists on
5203;; these variables where each element is a cons of the buffer
5204;; positions surrounding each identifier. This recording is only
5205;; activated when `c-record-type-identifiers' is non-nil.
d9e94c22
MS
5206;;
5207;; All known types that can't be identifiers are recorded, and also
5208;; other possible types if `c-promote-possible-types' is set.
5209;; Recording is however disabled inside angle bracket arglists that
5210;; are encountered inside names and other angle bracket arglists.
0386b551 5211;; Such occurrences are taken care of by `c-font-lock-<>-arglists'
d9e94c22
MS
5212;; instead.
5213;;
5214;; Only the names in C++ template style references (e.g. "tmpl" in
5215;; "tmpl<a,b>::foo") are recorded as references, other references
5216;; aren't handled here.
0386b551
AM
5217;;
5218;; `c-forward-label' records the label identifier(s) on
5219;; `c-record-ref-identifiers'.
d9e94c22
MS
5220(defvar c-record-type-identifiers nil)
5221(defvar c-record-ref-identifiers nil)
5222
0386b551
AM
5223;; This variable will receive a cons cell of the range of the last
5224;; single identifier symbol stepped over by `c-forward-name' if it's
5225;; successful. This is the range that should be put on one of the
5226;; record lists above by the caller. It's assigned nil if there's no
5227;; such symbol in the name.
d9e94c22
MS
5228(defvar c-last-identifier-range nil)
5229
5230(defmacro c-record-type-id (range)
5231 (if (eq (car-safe range) 'cons)
5232 ;; Always true.
5233 `(setq c-record-type-identifiers
5234 (cons ,range c-record-type-identifiers))
5235 `(let ((range ,range))
5236 (if range
5237 (setq c-record-type-identifiers
5238 (cons range c-record-type-identifiers))))))
5239
5240(defmacro c-record-ref-id (range)
5241 (if (eq (car-safe range) 'cons)
5242 ;; Always true.
5243 `(setq c-record-ref-identifiers
5244 (cons ,range c-record-ref-identifiers))
5245 `(let ((range ,range))
5246 (if range
5247 (setq c-record-ref-identifiers
5248 (cons range c-record-ref-identifiers))))))
5249
5250;; Dynamically bound variable that instructs `c-forward-type' to
5251;; record the ranges of types that only are found. Behaves otherwise
5252;; like `c-record-type-identifiers'.
5253(defvar c-record-found-types nil)
5254
5255(defmacro c-forward-keyword-prefixed-id (type)
5256 ;; Used internally in `c-forward-keyword-clause' to move forward
5257 ;; over a type (if TYPE is 'type) or a name (otherwise) which
5258 ;; possibly is prefixed by keywords and their associated clauses.
5259 ;; Try with a type/name first to not trip up on those that begin
5260 ;; with a keyword. Return t if a known or found type is moved
5261 ;; over. The point is clobbered if nil is returned. If range
5262 ;; recording is enabled, the identifier is recorded on as a type
5263 ;; if TYPE is 'type or as a reference if TYPE is 'ref.
0386b551
AM
5264 ;;
5265 ;; This macro might do hidden buffer changes.
d9e94c22
MS
5266 `(let (res)
5267 (while (if (setq res ,(if (eq type 'type)
5268 `(c-forward-type)
5269 `(c-forward-name)))
5270 nil
5271 (and (looking-at c-keywords-regexp)
0386b551 5272 (c-forward-keyword-clause 1))))
d9e94c22
MS
5273 (when (memq res '(t known found prefix))
5274 ,(when (eq type 'ref)
5275 `(when c-record-type-identifiers
5276 (c-record-ref-id c-last-identifier-range)))
5277 t)))
5278
0386b551 5279(defmacro c-forward-id-comma-list (type update-safe-pos)
d9e94c22
MS
5280 ;; Used internally in `c-forward-keyword-clause' to move forward
5281 ;; over a comma separated list of types or names using
5282 ;; `c-forward-keyword-prefixed-id'.
0386b551
AM
5283 ;;
5284 ;; This macro might do hidden buffer changes.
d9e94c22 5285 `(while (and (progn
0386b551
AM
5286 ,(when update-safe-pos
5287 `(setq safe-pos (point)))
d9e94c22
MS
5288 (eq (char-after) ?,))
5289 (progn
5290 (forward-char)
5291 (c-forward-syntactic-ws)
5292 (c-forward-keyword-prefixed-id ,type)))))
5293
0386b551
AM
5294(defun c-forward-keyword-clause (match)
5295 ;; Submatch MATCH in the current match data is assumed to surround a
5296 ;; token. If it's a keyword, move over it and any immediately
5297 ;; following clauses associated with it, stopping at the start of
5298 ;; the next token. t is returned in that case, otherwise the point
d9e94c22
MS
5299 ;; stays and nil is returned. The kind of clauses that are
5300 ;; recognized are those specified by `c-type-list-kwds',
5301 ;; `c-ref-list-kwds', `c-colon-type-list-kwds',
5302 ;; `c-paren-nontype-kwds', `c-paren-type-kwds', `c-<>-type-kwds',
5303 ;; and `c-<>-arglist-kwds'.
0386b551
AM
5304 ;;
5305 ;; This function records identifier ranges on
5306 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
5307 ;; `c-record-type-identifiers' is non-nil.
5308 ;;
5309 ;; Note that for `c-colon-type-list-kwds', which doesn't necessary
5310 ;; apply directly after the keyword, the type list is moved over
5311 ;; only when there is no unaccounted token before it (i.e. a token
5312 ;; that isn't moved over due to some other keyword list). The
5313 ;; identifier ranges in the list are still recorded if that should
5314 ;; be done, though.
5315 ;;
5316 ;; This function might do hidden buffer changes.
5317
5318 (let ((kwd-sym (c-keyword-sym (match-string match))) safe-pos pos
5319 ;; The call to `c-forward-<>-arglist' below is made after
5320 ;; `c-<>-sexp-kwds' keywords, so we're certain they actually
5321 ;; are angle bracket arglists and `c-restricted-<>-arglists'
5322 ;; should therefore be nil.
5323 (c-parse-and-markup-<>-arglists t)
5324 c-restricted-<>-arglists)
d9e94c22 5325
d9e94c22 5326 (when kwd-sym
0386b551 5327 (goto-char (match-end match))
d9e94c22
MS
5328 (c-forward-syntactic-ws)
5329 (setq safe-pos (point))
a66cd3ee 5330
d9e94c22
MS
5331 (cond
5332 ((and (c-keyword-member kwd-sym 'c-type-list-kwds)
5333 (c-forward-keyword-prefixed-id type))
5334 ;; There's a type directly after a keyword in `c-type-list-kwds'.
0386b551 5335 (c-forward-id-comma-list type t))
d9e94c22
MS
5336
5337 ((and (c-keyword-member kwd-sym 'c-ref-list-kwds)
5338 (c-forward-keyword-prefixed-id ref))
5339 ;; There's a name directly after a keyword in `c-ref-list-kwds'.
0386b551 5340 (c-forward-id-comma-list ref t))
d9e94c22
MS
5341
5342 ((and (c-keyword-member kwd-sym 'c-paren-any-kwds)
5343 (eq (char-after) ?\())
5344 ;; There's an open paren after a keyword in `c-paren-any-kwds'.
5345
5346 (forward-char)
5347 (when (and (setq pos (c-up-list-forward))
5348 (eq (char-before pos) ?\)))
5349 (when (and c-record-type-identifiers
5350 (c-keyword-member kwd-sym 'c-paren-type-kwds))
5351 ;; Use `c-forward-type' on every identifier we can find
5352 ;; inside the paren, to record the types.
5353 (while (c-syntactic-re-search-forward c-symbol-start pos t)
5354 (goto-char (match-beginning 0))
5355 (unless (c-forward-type)
5356 (looking-at c-symbol-key) ; Always matches.
5357 (goto-char (match-end 0)))))
5358
5359 (goto-char pos)
5360 (c-forward-syntactic-ws)
5361 (setq safe-pos (point))))
5362
5363 ((and (c-keyword-member kwd-sym 'c-<>-sexp-kwds)
5364 (eq (char-after) ?<)
0386b551 5365 (c-forward-<>-arglist (c-keyword-member kwd-sym 'c-<>-type-kwds)))
d9e94c22
MS
5366 (c-forward-syntactic-ws)
5367 (setq safe-pos (point)))
5368
5369 ((and (c-keyword-member kwd-sym 'c-nonsymbol-sexp-kwds)
449a2b0d
MS
5370 (not (looking-at c-symbol-start))
5371 (c-safe (c-forward-sexp) t))
d9e94c22
MS
5372 (c-forward-syntactic-ws)
5373 (setq safe-pos (point))))
5374
0386b551
AM
5375 (when (c-keyword-member kwd-sym 'c-colon-type-list-kwds)
5376 (if (eq (char-after) ?:)
5377 ;; If we are at the colon already, we move over the type
5378 ;; list after it.
5379 (progn
5380 (forward-char)
5381 (c-forward-syntactic-ws)
5382 (when (c-forward-keyword-prefixed-id type)
5383 (c-forward-id-comma-list type t)))
5384 ;; Not at the colon, so stop here. But the identifier
5385 ;; ranges in the type list later on should still be
5386 ;; recorded.
5387 (and c-record-type-identifiers
5388 (progn
5389 ;; If a keyword matched both one of the types above and
5390 ;; this one, we match `c-colon-type-list-re' after the
5391 ;; clause matched above.
5392 (goto-char safe-pos)
5393 (looking-at c-colon-type-list-re))
5394 (progn
5395 (goto-char (match-end 0))
5396 (c-forward-syntactic-ws)
5397 (c-forward-keyword-prefixed-id type))
5398 ;; There's a type after the `c-colon-type-list-re' match
5399 ;; after a keyword in `c-colon-type-list-kwds'.
5400 (c-forward-id-comma-list type nil))))
d9e94c22
MS
5401
5402 (goto-char safe-pos)
5403 t)))
5404
f3b554af
GM
5405;; cc-mode requires cc-fonts.
5406(declare-function c-fontify-recorded-types-and-refs "cc-fonts" ())
5407
0386b551
AM
5408(defun c-forward-<>-arglist (all-types)
5409 ;; The point is assumed to be at a "<". Try to treat it as the open
b4dc7d98 5410 ;; paren of an angle bracket arglist and move forward to the
0386b551
AM
5411 ;; corresponding ">". If successful, the point is left after the
5412 ;; ">" and t is returned, otherwise the point isn't moved and nil is
d9e94c22
MS
5413 ;; returned. If ALL-TYPES is t then all encountered arguments in
5414 ;; the arglist that might be types are treated as found types.
5415 ;;
0386b551
AM
5416 ;; The variable `c-parse-and-markup-<>-arglists' controls how this
5417 ;; function handles text properties on the angle brackets and argument
5418 ;; separating commas.
d9e94c22 5419 ;;
0386b551
AM
5420 ;; `c-restricted-<>-arglists' controls how lenient the template
5421 ;; arglist recognition should be.
5422 ;;
5423 ;; This function records identifier ranges on
5424 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
5425 ;; `c-record-type-identifiers' is non-nil.
5426 ;;
5427 ;; This function might do hidden buffer changes.
d9e94c22
MS
5428
5429 (let ((start (point))
5430 ;; If `c-record-type-identifiers' is set then activate
5431 ;; recording of any found types that constitute an argument in
5432 ;; the arglist.
abfc152b 5433 (c-record-found-types (if c-record-type-identifiers t)))
d9e94c22
MS
5434 (if (catch 'angle-bracket-arglist-escape
5435 (setq c-record-found-types
0386b551 5436 (c-forward-<>-arglist-recur all-types)))
d9e94c22
MS
5437 (progn
5438 (when (consp c-record-found-types)
5439 (setq c-record-type-identifiers
5440 ;; `nconc' doesn't mind that the tail of
5441 ;; `c-record-found-types' is t.
5442 (nconc c-record-found-types c-record-type-identifiers)))
452ea855 5443 (if (c-major-mode-is 'java-mode) (c-fontify-recorded-types-and-refs))
d9e94c22
MS
5444 t)
5445
5446 (goto-char start)
a66cd3ee 5447 nil)))
785eecbb 5448
0386b551 5449(defun c-forward-<>-arglist-recur (all-types)
d9e94c22 5450 ;; Recursive part of `c-forward-<>-arglist'.
0386b551
AM
5451 ;;
5452 ;; This function might do hidden buffer changes.
d9e94c22
MS
5453
5454 (let ((start (point)) res pos tmp
5455 ;; Cover this so that any recorded found type ranges are
5456 ;; automatically lost if it turns out to not be an angle
5457 ;; bracket arglist. It's propagated through the return value
5458 ;; on successful completion.
5459 (c-record-found-types c-record-found-types)
5460 ;; List that collects the positions after the argument
5461 ;; separating ',' in the arglist.
5462 arg-start-pos)
0386b551
AM
5463 ;; If the '<' has paren open syntax then we've marked it as an angle
5464 ;; bracket arglist before, so skip to the end.
5465 (if (and (not c-parse-and-markup-<>-arglists)
5466 (c-get-char-property (point) 'syntax-table))
5467
5468 (progn
5469 (forward-char)
5470 (if (and (c-go-up-list-forward)
5471 (eq (char-before) ?>))
5472 t
0386b551
AM
5473 ;; Got unmatched paren angle brackets. We don't clear the paren
5474 ;; syntax properties and retry, on the basis that it's very
5475 ;; unlikely that paren angle brackets become operators by code
5476 ;; manipulation. It's far more likely that it doesn't match due
5477 ;; to narrowing or some temporary change.
5478 (goto-char start)
5479 nil))
d9e94c22 5480
abfc152b 5481 (forward-char) ; Forward over the opening '<'.
452ea855 5482
d9e94c22 5483 (unless (looking-at c-<-op-cont-regexp)
abfc152b
AM
5484 ;; go forward one non-alphanumeric character (group) per iteration of
5485 ;; this loop.
22c3ce97 5486 (while (and
d9e94c22 5487 (progn
22c3ce97
AM
5488 (c-forward-syntactic-ws)
5489 (let ((orig-record-found-types c-record-found-types))
5490 (when (or (and c-record-type-identifiers all-types)
5491 (c-major-mode-is 'java-mode))
5492 ;; All encountered identifiers are types, so set the
5493 ;; promote flag and parse the type.
5494 (progn
5495 (c-forward-syntactic-ws)
5496 (if (looking-at "\\?")
5497 (forward-char)
5498 (when (looking-at c-identifier-start)
5499 (let ((c-promote-possible-types t)
5500 (c-record-found-types t))
5501 (c-forward-type))))
5502
5503 (c-forward-syntactic-ws)
5504
5505 (when (or (looking-at "extends")
5506 (looking-at "super"))
5507 (forward-word)
5508 (c-forward-syntactic-ws)
5509 (let ((c-promote-possible-types t)
5510 (c-record-found-types t))
5511 (c-forward-type)
5512 (c-forward-syntactic-ws))))))
5513
abfc152b 5514 (setq pos (point)) ; e.g. first token inside the '<'
22c3ce97 5515
a4ee83cc
AM
5516 ;; Note: These regexps exploit the match order in \| so
5517 ;; that "<>" is matched by "<" rather than "[^>:-]>".
5518 (c-syntactic-re-search-forward
5519 ;; Stop on ',', '|', '&', '+' and '-' to catch
5520 ;; common binary operators that could be between
5521 ;; two comparison expressions "a<b" and "c>d".
5522 "[<;{},|+&-]\\|[>)]"
5523 nil t t))
22c3ce97
AM
5524
5525 (cond
5526 ((eq (char-before) ?>)
d9e94c22
MS
5527 ;; Either an operator starting with '>' or the end of
5528 ;; the angle bracket arglist.
5529
0386b551 5530 (if (looking-at c->-op-cont-regexp)
d9e94c22 5531 (progn
0386b551
AM
5532 (goto-char (match-end 0))
5533 t) ; Continue the loop.
d9e94c22 5534
0386b551
AM
5535 ;; The angle bracket arglist is finished.
5536 (when c-parse-and-markup-<>-arglists
d9e94c22 5537 (while arg-start-pos
0386b551
AM
5538 (c-put-c-type-property (1- (car arg-start-pos))
5539 'c-<>-arg-sep)
d9e94c22
MS
5540 (setq arg-start-pos (cdr arg-start-pos)))
5541 (c-mark-<-as-paren start)
0386b551
AM
5542 (c-mark->-as-paren (1- (point))))
5543 (setq res t)
5544 nil)) ; Exit the loop.
d9e94c22
MS
5545
5546 ((eq (char-before) ?<)
5547 ;; Either an operator starting with '<' or a nested arglist.
d9e94c22
MS
5548 (setq pos (point))
5549 (let (id-start id-end subres keyword-match)
abfc152b
AM
5550 (cond
5551 ;; The '<' begins a multi-char operator.
5552 ((looking-at c-<-op-cont-regexp)
5553 (setq tmp (match-end 0))
5554 (goto-char (match-end 0)))
5555 ;; We're at a nested <.....>
5556 ((progn
5557 (setq tmp pos)
5558 (backward-char) ; to the '<'
5559 (and
5560 (save-excursion
5561 ;; There's always an identifier before an angle
5562 ;; bracket arglist, or a keyword in `c-<>-type-kwds'
5563 ;; or `c-<>-arglist-kwds'.
5564 (c-backward-syntactic-ws)
5565 (setq id-end (point))
5566 (c-simple-skip-symbol-backward)
5567 (when (or (setq keyword-match
5568 (looking-at c-opt-<>-sexp-key))
5569 (not (looking-at c-keywords-regexp)))
5570 (setq id-start (point))))
5571 (setq subres
5572 (let ((c-promote-possible-types t)
5573 (c-record-found-types t))
5574 (c-forward-<>-arglist-recur
5575 (and keyword-match
5576 (c-keyword-member
5577 (c-keyword-sym (match-string 1))
5578 'c-<>-type-kwds)))))))
d9e94c22
MS
5579
5580 ;; It was an angle bracket arglist.
5581 (setq c-record-found-types subres)
5582
5583 ;; Record the identifier before the template as a type
5584 ;; or reference depending on whether the arglist is last
5585 ;; in a qualified identifier.
5586 (when (and c-record-type-identifiers
5587 (not keyword-match))
5588 (if (and c-opt-identifier-concat-key
5589 (progn
5590 (c-forward-syntactic-ws)
5591 (looking-at c-opt-identifier-concat-key)))
5592 (c-record-ref-id (cons id-start id-end))
abfc152b
AM
5593 (c-record-type-id (cons id-start id-end)))))
5594
5595 ;; At a "less than" operator.
5596 (t
5597 (forward-char)
5598 )))
5599 t) ; carry on looping.
22c3ce97
AM
5600
5601 ((and (not c-restricted-<>-arglists)
5602 (or (and (eq (char-before) ?&)
5603 (not (eq (char-after) ?&)))
5604 (eq (char-before) ?,)))
5605 ;; Just another argument. Record the position. The
5606 ;; type check stuff that made us stop at it is at
5607 ;; the top of the loop.
d9e94c22
MS
5608 (setq arg-start-pos (cons (point) arg-start-pos)))
5609
5610 (t
5611 ;; Got a character that can't be in an angle bracket
5612 ;; arglist argument. Abort using `throw', since
5613 ;; it's useless to try to find a surrounding arglist
5614 ;; if we're nested.
5615 (throw 'angle-bracket-arglist-escape nil))))))
d9e94c22
MS
5616 (if res
5617 (or c-record-found-types t)))))
5618
0386b551
AM
5619(defun c-backward-<>-arglist (all-types &optional limit)
5620 ;; The point is assumed to be directly after a ">". Try to treat it
5621 ;; as the close paren of an angle bracket arglist and move back to
5622 ;; the corresponding "<". If successful, the point is left at
5623 ;; the "<" and t is returned, otherwise the point isn't moved and
5624 ;; nil is returned. ALL-TYPES is passed on to
5625 ;; `c-forward-<>-arglist'.
5626 ;;
5627 ;; If the optional LIMIT is given, it bounds the backward search.
5628 ;; It's then assumed to be at a syntactically relevant position.
5629 ;;
5630 ;; This is a wrapper around `c-forward-<>-arglist'. See that
5631 ;; function for more details.
5632
5633 (let ((start (point)))
5634 (backward-char)
5635 (if (and (not c-parse-and-markup-<>-arglists)
5636 (c-get-char-property (point) 'syntax-table))
5637
5638 (if (and (c-go-up-list-backward)
5639 (eq (char-after) ?<))
5640 t
5641 ;; See corresponding note in `c-forward-<>-arglist'.
5642 (goto-char start)
5643 nil)
5644
51c9af45 5645 (while (progn
0386b551
AM
5646 (c-syntactic-skip-backward "^<;{}" limit t)
5647
51c9af45
AM
5648 (and
5649 (if (eq (char-before) ?<)
5650 t
5651 ;; Stopped at bob or a char that isn't allowed in an
5652 ;; arglist, so we've failed.
5653 (goto-char start)
5654 nil)
0386b551 5655
51c9af45
AM
5656 (if (> (point)
5657 (progn (c-beginning-of-current-token)
5658 (point)))
5659 ;; If we moved then the "<" was part of some
5660 ;; multicharacter token.
5661 t
0386b551 5662
51c9af45
AM
5663 (backward-char)
5664 (let ((beg-pos (point)))
5665 (if (c-forward-<>-arglist all-types)
5666 (cond ((= (point) start)
5667 ;; Matched the arglist. Break the while.
5668 (goto-char beg-pos)
5669 nil)
5670 ((> (point) start)
5671 ;; We started from a non-paren ">" inside an
5672 ;; arglist.
5673 (goto-char start)
5674 nil)
5675 (t
5676 ;; Matched a shorter arglist. Can be a nested
5677 ;; one so continue looking.
5678 (goto-char beg-pos)
5679 t))
5680 t))))))
0386b551
AM
5681
5682 (/= (point) start))))
5683
d9e94c22
MS
5684(defun c-forward-name ()
5685 ;; Move forward over a complete name if at the beginning of one,
e15f8aaa
AM
5686 ;; stopping at the next following token. A keyword, as such,
5687 ;; doesn't count as a name. If the point is not at something that
5688 ;; is recognized as a name then it stays put.
5689 ;;
5690 ;; A name could be something as simple as "foo" in C or something as
d9e94c22
MS
5691 ;; complex as "X<Y<class A<int>::B, BIT_MAX >> b>, ::operator<> ::
5692 ;; Z<(a>b)> :: operator const X<&foo>::T Q::G<unsigned short
5693 ;; int>::*volatile const" in C++ (this function is actually little
5694 ;; more than a `looking-at' call in all modes except those that,
e15f8aaa
AM
5695 ;; like C++, have `c-recognize-<>-arglists' set).
5696 ;;
5697 ;; Return
5698 ;; o - nil if no name is found;
5699 ;; o - 'template if it's an identifier ending with an angle bracket
5700 ;; arglist;
5701 ;; o - 'operator of it's an operator identifier;
5702 ;; o - t if it's some other kind of name.
0386b551
AM
5703 ;;
5704 ;; This function records identifier ranges on
5705 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
5706 ;; `c-record-type-identifiers' is non-nil.
5707 ;;
5708 ;; This function might do hidden buffer changes.
d9e94c22 5709
0386b551 5710 (let ((pos (point)) (start (point)) res id-start id-end
d9e94c22
MS
5711 ;; Turn off `c-promote-possible-types' here since we might
5712 ;; call `c-forward-<>-arglist' and we don't want it to promote
5713 ;; every suspect thing in the arglist to a type. We're
5714 ;; typically called from `c-forward-type' in this case, and
5715 ;; the caller only wants the top level type that it finds to
5716 ;; be promoted.
5717 c-promote-possible-types)
5718 (while
5719 (and
5720 (looking-at c-identifier-key)
5721
5722 (progn
5723 ;; Check for keyword. We go to the last symbol in
5724 ;; `c-identifier-key' first.
0386b551
AM
5725 (goto-char (setq id-end (match-end 0)))
5726 (c-simple-skip-symbol-backward)
5727 (setq id-start (point))
d9e94c22
MS
5728
5729 (if (looking-at c-keywords-regexp)
5730 (when (and (c-major-mode-is 'c++-mode)
5731 (looking-at
5732 (cc-eval-when-compile
5733 (concat "\\(operator\\|\\(template\\)\\)"
5734 "\\(" (c-lang-const c-nonsymbol-key c++)
5735 "\\|$\\)")))
5736 (if (match-beginning 2)
5737 ;; "template" is only valid inside an
5738 ;; identifier if preceded by "::".
5739 (save-excursion
5740 (c-backward-syntactic-ws)
5741 (and (c-safe (backward-char 2) t)
5742 (looking-at "::")))
5743 t))
5744
5745 ;; Handle a C++ operator or template identifier.
5746 (goto-char id-end)
5747 (c-forward-syntactic-ws)
5748 (cond ((eq (char-before id-end) ?e)
5749 ;; Got "... ::template".
5750 (let ((subres (c-forward-name)))
5751 (when subres
5752 (setq pos (point)
5753 res subres))))
5754
5755 ((looking-at c-identifier-start)
5756 ;; Got a cast operator.
5757 (when (c-forward-type)
5758 (setq pos (point)
5759 res 'operator)
5760 ;; Now we should match a sequence of either
5761 ;; '*', '&' or a name followed by ":: *",
5762 ;; where each can be followed by a sequence
5763 ;; of `c-opt-type-modifier-key'.
5764 (while (cond ((looking-at "[*&]")
5765 (goto-char (match-end 0))
5766 t)
5767 ((looking-at c-identifier-start)
5768 (and (c-forward-name)
5769 (looking-at "::")
5770 (progn
5771 (goto-char (match-end 0))
5772 (c-forward-syntactic-ws)
5773 (eq (char-after) ?*))
5774 (progn
5775 (forward-char)
5776 t))))
5777 (while (progn
5778 (c-forward-syntactic-ws)
5779 (setq pos (point))
5780 (looking-at c-opt-type-modifier-key))
5781 (goto-char (match-end 1))))))
5782
5783 ((looking-at c-overloadable-operators-regexp)
5784 ;; Got some other operator.
0386b551
AM
5785 (setq c-last-identifier-range
5786 (cons (point) (match-end 0)))
d9e94c22
MS
5787 (goto-char (match-end 0))
5788 (c-forward-syntactic-ws)
5789 (setq pos (point)
5790 res 'operator)))
5791
5792 nil)
5793
0386b551
AM
5794 ;; `id-start' is equal to `id-end' if we've jumped over
5795 ;; an identifier that doesn't end with a symbol token.
5796 ;; That can occur e.g. for Java import directives on the
5797 ;; form "foo.bar.*".
5798 (when (and id-start (/= id-start id-end))
d9e94c22
MS
5799 (setq c-last-identifier-range
5800 (cons id-start id-end)))
5801 (goto-char id-end)
5802 (c-forward-syntactic-ws)
5803 (setq pos (point)
5804 res t)))
5805
5806 (progn
5807 (goto-char pos)
5808 (when (or c-opt-identifier-concat-key
5809 c-recognize-<>-arglists)
5810
5811 (cond
5812 ((and c-opt-identifier-concat-key
5813 (looking-at c-opt-identifier-concat-key))
5814 ;; Got a concatenated identifier. This handles the
5815 ;; cases with tricky syntactic whitespace that aren't
5816 ;; covered in `c-identifier-key'.
5817 (goto-char (match-end 0))
5818 (c-forward-syntactic-ws)
5819 t)
5820
5821 ((and c-recognize-<>-arglists
5822 (eq (char-after) ?<))
5823 ;; Maybe an angle bracket arglist.
452ea855
AM
5824 (when (let ((c-record-type-identifiers t)
5825 (c-record-found-types t))
0386b551
AM
5826 (c-forward-<>-arglist nil))
5827
5828 (c-add-type start (1+ pos))
d9e94c22 5829 (c-forward-syntactic-ws)
0386b551
AM
5830 (setq pos (point)
5831 c-last-identifier-range nil)
5832
d9e94c22
MS
5833 (if (and c-opt-identifier-concat-key
5834 (looking-at c-opt-identifier-concat-key))
0386b551 5835
d9e94c22
MS
5836 ;; Continue if there's an identifier concatenation
5837 ;; operator after the template argument.
5838 (progn
0386b551
AM
5839 (when (and c-record-type-identifiers id-start)
5840 (c-record-ref-id (cons id-start id-end)))
d9e94c22
MS
5841 (forward-char 2)
5842 (c-forward-syntactic-ws)
5843 t)
0386b551
AM
5844
5845 (when (and c-record-type-identifiers id-start)
5846 (c-record-type-id (cons id-start id-end)))
d9e94c22
MS
5847 (setq res 'template)
5848 nil)))
5849 )))))
5850
5851 (goto-char pos)
5852 res))
5853
e15f8aaa 5854(defun c-forward-type (&optional brace-block-too)
d9e94c22 5855 ;; Move forward over a type spec if at the beginning of one,
e15f8aaa
AM
5856 ;; stopping at the next following token. The keyword "typedef"
5857 ;; isn't part of a type spec here.
5858 ;;
5859 ;; BRACE-BLOCK-TOO, when non-nil, means move over the brace block in
5860 ;; constructs like "struct foo {...} bar ;" or "struct {...} bar;".
5861 ;; The current (2009-03-10) intention is to convert all uses of
5862 ;; `c-forward-type' to call with this parameter set, then to
5863 ;; eliminate it.
5864 ;;
5865 ;; Return
5866 ;; o - t if it's a known type that can't be a name or other
5867 ;; expression;
5868 ;; o - 'known if it's an otherwise known type (according to
5869 ;; `*-font-lock-extra-types');
5870 ;; o - 'prefix if it's a known prefix of a type;
5871 ;; o - 'found if it's a type that matches one in `c-found-types';
5872 ;; o - 'maybe if it's an identfier that might be a type; or
5873 ;; o - nil if it can't be a type (the point isn't moved then).
5874 ;;
5875 ;; The point is assumed to be at the beginning of a token.
d9e94c22
MS
5876 ;;
5877 ;; Note that this function doesn't skip past the brace definition
5878 ;; that might be considered part of the type, e.g.
5879 ;; "enum {a, b, c} foo".
0386b551
AM
5880 ;;
5881 ;; This function records identifier ranges on
5882 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
5883 ;; `c-record-type-identifiers' is non-nil.
5884 ;;
5885 ;; This function might do hidden buffer changes.
0e5cf2b8
AM
5886 (when (and c-recognize-<>-arglists
5887 (looking-at "<"))
452ea855
AM
5888 (c-forward-<>-arglist t)
5889 (c-forward-syntactic-ws))
0386b551
AM
5890
5891 (let ((start (point)) pos res name-res id-start id-end id-range)
d9e94c22
MS
5892
5893 ;; Skip leading type modifiers. If any are found we know it's a
5894 ;; prefix of a type.
e15f8aaa 5895 (when c-opt-type-modifier-key ; e.g. "const" "volatile", but NOT "typedef"
d9e94c22
MS
5896 (while (looking-at c-opt-type-modifier-key)
5897 (goto-char (match-end 1))
5898 (c-forward-syntactic-ws)
5899 (setq res 'prefix)))
5900
5901 (cond
e15f8aaa
AM
5902 ((looking-at c-type-prefix-key) ; e.g. "struct", "class", but NOT
5903 ; "typedef".
d9e94c22
MS
5904 (goto-char (match-end 1))
5905 (c-forward-syntactic-ws)
5906 (setq pos (point))
e15f8aaa
AM
5907
5908 (setq name-res (c-forward-name))
5909 (setq res (not (null name-res)))
5910 (when (eq name-res t)
5911 ;; In many languages the name can be used without the
5912 ;; prefix, so we add it to `c-found-types'.
5913 (c-add-type pos (point))
5914 (when (and c-record-type-identifiers
5915 c-last-identifier-range)
5916 (c-record-type-id c-last-identifier-range)))
5917 (when (and brace-block-too
5918 (memq res '(t nil))
5919 (eq (char-after) ?\{)
5920 (save-excursion
5921 (c-safe
5922 (progn (c-forward-sexp)
5923 (c-forward-syntactic-ws)
5924 (setq pos (point))))))
5925 (goto-char pos)
5926 (setq res t))
5927 (unless res (goto-char start))) ; invalid syntax
d9e94c22
MS
5928
5929 ((progn
5930 (setq pos nil)
5931 (if (looking-at c-identifier-start)
5932 (save-excursion
5933 (setq id-start (point)
0386b551
AM
5934 name-res (c-forward-name))
5935 (when name-res
d9e94c22
MS
5936 (setq id-end (point)
5937 id-range c-last-identifier-range))))
5938 (and (cond ((looking-at c-primitive-type-key)
5939 (setq res t))
5940 ((c-with-syntax-table c-identifier-syntax-table
5941 (looking-at c-known-type-key))
5942 (setq res 'known)))
5943 (or (not id-end)
5944 (>= (save-excursion
5945 (save-match-data
5946 (goto-char (match-end 1))
5947 (c-forward-syntactic-ws)
5948 (setq pos (point))))
5949 id-end)
5950 (setq res nil))))
5951 ;; Looking at a primitive or known type identifier. We've
5952 ;; checked for a name first so that we don't go here if the
5953 ;; known type match only is a prefix of another name.
5954
5955 (setq id-end (match-end 1))
5956
5957 (when (and c-record-type-identifiers
5958 (or c-promote-possible-types (eq res t)))
5959 (c-record-type-id (cons (match-beginning 1) (match-end 1))))
5960
5961 (if (and c-opt-type-component-key
5962 (save-match-data
5963 (looking-at c-opt-type-component-key)))
5964 ;; There might be more keywords for the type.
5965 (let (safe-pos)
0386b551 5966 (c-forward-keyword-clause 1)
d9e94c22
MS
5967 (while (progn
5968 (setq safe-pos (point))
5969 (looking-at c-opt-type-component-key))
5970 (when (and c-record-type-identifiers
5971 (looking-at c-primitive-type-key))
5972 (c-record-type-id (cons (match-beginning 1)
5973 (match-end 1))))
0386b551 5974 (c-forward-keyword-clause 1))
d9e94c22
MS
5975 (if (looking-at c-primitive-type-key)
5976 (progn
5977 (when c-record-type-identifiers
5978 (c-record-type-id (cons (match-beginning 1)
5979 (match-end 1))))
0386b551 5980 (c-forward-keyword-clause 1)
d9e94c22
MS
5981 (setq res t))
5982 (goto-char safe-pos)
5983 (setq res 'prefix)))
0386b551 5984 (unless (save-match-data (c-forward-keyword-clause 1))
d9e94c22
MS
5985 (if pos
5986 (goto-char pos)
5987 (goto-char (match-end 1))
5988 (c-forward-syntactic-ws)))))
5989
0386b551
AM
5990 (name-res
5991 (cond ((eq name-res t)
d9e94c22
MS
5992 ;; A normal identifier.
5993 (goto-char id-end)
5994 (if (or res c-promote-possible-types)
5995 (progn
5996 (c-add-type id-start id-end)
0386b551 5997 (when (and c-record-type-identifiers id-range)
d9e94c22
MS
5998 (c-record-type-id id-range))
5999 (unless res
6000 (setq res 'found)))
6001 (setq res (if (c-check-type id-start id-end)
6002 ;; It's an identifier that has been used as
6003 ;; a type somewhere else.
6004 'found
6005 ;; It's an identifier that might be a type.
6006 'maybe))))
0386b551 6007 ((eq name-res 'template)
d9e94c22
MS
6008 ;; A template is a type.
6009 (goto-char id-end)
6010 (setq res t))
6011 (t
6012 ;; Otherwise it's an operator identifier, which is not a type.
6013 (goto-char start)
6014 (setq res nil)))))
6015
6016 (when res
e15f8aaa 6017 ;; Skip trailing type modifiers. If any are found we know it's
d9e94c22
MS
6018 ;; a type.
6019 (when c-opt-type-modifier-key
e15f8aaa 6020 (while (looking-at c-opt-type-modifier-key) ; e.g. "const", "volatile"
d9e94c22
MS
6021 (goto-char (match-end 1))
6022 (c-forward-syntactic-ws)
6023 (setq res t)))
d9e94c22
MS
6024 ;; Step over any type suffix operator. Do not let the existence
6025 ;; of these alter the classification of the found type, since
6026 ;; these operators typically are allowed in normal expressions
6027 ;; too.
6028 (when c-opt-type-suffix-key
6029 (while (looking-at c-opt-type-suffix-key)
6030 (goto-char (match-end 1))
6031 (c-forward-syntactic-ws)))
6032
e15f8aaa 6033 (when c-opt-type-concat-key ; Only/mainly for pike.
0386b551
AM
6034 ;; Look for a trailing operator that concatenates the type
6035 ;; with a following one, and if so step past that one through
6036 ;; a recursive call. Note that we don't record concatenated
6037 ;; types in `c-found-types' - it's the component types that
6038 ;; are recorded when appropriate.
d9e94c22
MS
6039 (setq pos (point))
6040 (let* ((c-promote-possible-types (or (memq res '(t known))
6041 c-promote-possible-types))
6042 ;; If we can't promote then set `c-record-found-types' so that
6043 ;; we can merge in the types from the second part afterwards if
6044 ;; it turns out to be a known type there.
6045 (c-record-found-types (and c-record-type-identifiers
0386b551
AM
6046 (not c-promote-possible-types)))
6047 subres)
d9e94c22
MS
6048 (if (and (looking-at c-opt-type-concat-key)
6049
6050 (progn
6051 (goto-char (match-end 1))
6052 (c-forward-syntactic-ws)
0386b551 6053 (setq subres (c-forward-type))))
d9e94c22
MS
6054
6055 (progn
6056 ;; If either operand certainly is a type then both are, but we
6057 ;; don't let the existence of the operator itself promote two
6058 ;; uncertain types to a certain one.
6059 (cond ((eq res t))
0386b551
AM
6060 ((eq subres t)
6061 (unless (eq name-res 'template)
6062 (c-add-type id-start id-end))
6063 (when (and c-record-type-identifiers id-range)
d9e94c22
MS
6064 (c-record-type-id id-range))
6065 (setq res t))
2a15eb73 6066 ((eq res 'known))
0386b551 6067 ((eq subres 'known)
2a15eb73 6068 (setq res 'known))
d9e94c22 6069 ((eq res 'found))
0386b551 6070 ((eq subres 'found)
d9e94c22
MS
6071 (setq res 'found))
6072 (t
6073 (setq res 'maybe)))
6074
6075 (when (and (eq res t)
6076 (consp c-record-found-types))
6077 ;; Merge in the ranges of any types found by the second
6078 ;; `c-forward-type'.
6079 (setq c-record-type-identifiers
6080 ;; `nconc' doesn't mind that the tail of
6081 ;; `c-record-found-types' is t.
6082 (nconc c-record-found-types
6083 c-record-type-identifiers))))
6084
6085 (goto-char pos))))
6086
6087 (when (and c-record-found-types (memq res '(known found)) id-range)
6088 (setq c-record-found-types
6089 (cons id-range c-record-found-types))))
6090
6091 ;;(message "c-forward-type %s -> %s: %s" start (point) res)
6092
6093 res))
6094
452ea855
AM
6095(defun c-forward-annotation ()
6096 ;; Used for Java code only at the moment. Assumes point is on the
6097 ;; @, moves forward an annotation. returns nil if there is no
6098 ;; annotation at point.
6099 (and (looking-at "@")
6100 (progn (forward-char) t)
6101 (c-forward-type)
6102 (progn (c-forward-syntactic-ws) t)
6103 (if (looking-at "(")
6104 (c-go-list-forward)
6105 t)))
6106
785eecbb 6107\f
d9e94c22
MS
6108;; Handling of large scale constructs like statements and declarations.
6109
0386b551
AM
6110;; Macro used inside `c-forward-decl-or-cast-1'. It ought to be a
6111;; defsubst or perhaps even a defun, but it contains lots of free
6112;; variables that refer to things inside `c-forward-decl-or-cast-1'.
6113(defmacro c-fdoc-shift-type-backward (&optional short)
6114 ;; `c-forward-decl-or-cast-1' can consume an arbitrary length list
6115 ;; of types when parsing a declaration, which means that it
6116 ;; sometimes consumes the identifier in the declaration as a type.
6117 ;; This is used to "backtrack" and make the last type be treated as
6118 ;; an identifier instead.
6119 `(progn
6120 ,(unless short
6121 ;; These identifiers are bound only in the inner let.
6122 '(setq identifier-type at-type
6123 identifier-start type-start
6124 got-parens nil
6125 got-identifier t
6126 got-suffix t
6127 got-suffix-after-parens id-start
6128 paren-depth 0))
6129
6130 (if (setq at-type (if (eq backup-at-type 'prefix)
6131 t
6132 backup-at-type))
6133 (setq type-start backup-type-start
6134 id-start backup-id-start)
6135 (setq type-start start-pos
6136 id-start start-pos))
6137
6138 ;; When these flags already are set we've found specifiers that
6139 ;; unconditionally signal these attributes - backtracking doesn't
6140 ;; change that. So keep them set in that case.
6141 (or at-type-decl
6142 (setq at-type-decl backup-at-type-decl))
6143 (or maybe-typeless
6144 (setq maybe-typeless backup-maybe-typeless))
6145
6146 ,(unless short
6147 ;; This identifier is bound only in the inner let.
6148 '(setq start id-start))))
6149
6150(defun c-forward-decl-or-cast-1 (preceding-token-end context last-cast-end)
6151 ;; Move forward over a declaration or a cast if at the start of one.
6152 ;; The point is assumed to be at the start of some token. Nil is
6153 ;; returned if no declaration or cast is recognized, and the point
6154 ;; is clobbered in that case.
6155 ;;
6156 ;; If a declaration is parsed:
6157 ;;
6158 ;; The point is left at the first token after the first complete
6159 ;; declarator, if there is one. The return value is a cons where
022d0cf4
AM
6160 ;; the car is the position of the first token in the declarator. (See
6161 ;; below for the cdr.)
0386b551
AM
6162 ;; Some examples:
6163 ;;
6164 ;; void foo (int a, char *b) stuff ...
6165 ;; car ^ ^ point
6166 ;; float (*a)[], b;
6167 ;; car ^ ^ point
6168 ;; unsigned int a = c_style_initializer, b;
6169 ;; car ^ ^ point
6170 ;; unsigned int a (cplusplus_style_initializer), b;
6171 ;; car ^ ^ point (might change)
6172 ;; class Foo : public Bar {}
6173 ;; car ^ ^ point
6174 ;; class PikeClass (int a, string b) stuff ...
6175 ;; car ^ ^ point
6176 ;; enum bool;
6177 ;; car ^ ^ point
6178 ;; enum bool flag;
6179 ;; car ^ ^ point
6180 ;; void cplusplus_function (int x) throw (Bad);
6181 ;; car ^ ^ point
6182 ;; Foo::Foo (int b) : Base (b) {}
6183 ;; car ^ ^ point
e15f8aaa
AM
6184 ;;
6185 ;; The cdr of the return value is non-nil when a
6186 ;; `c-typedef-decl-kwds' specifier is found in the declaration.
6187 ;; Specifically it is a dotted pair (A . B) where B is t when a
6188 ;; `c-typedef-kwds' ("typedef") is present, and A is t when some
6189 ;; other `c-typedef-decl-kwds' (e.g. class, struct, enum)
6190 ;; specifier is present. I.e., (some of) the declared
6191 ;; identifier(s) are types.
6192 ;;
0386b551
AM
6193 ;; If a cast is parsed:
6194 ;;
6195 ;; The point is left at the first token after the closing paren of
6196 ;; the cast. The return value is `cast'. Note that the start
6197 ;; position must be at the first token inside the cast parenthesis
6198 ;; to recognize it.
6199 ;;
6200 ;; PRECEDING-TOKEN-END is the first position after the preceding
6201 ;; token, i.e. on the other side of the syntactic ws from the point.
6202 ;; Use a value less than or equal to (point-min) if the point is at
6203 ;; the first token in (the visible part of) the buffer.
6204 ;;
6205 ;; CONTEXT is a symbol that describes the context at the point:
022d0cf4 6206 ;; 'decl In a comma-separated declaration context (typically
0386b551
AM
6207 ;; inside a function declaration arglist).
6208 ;; '<> In an angle bracket arglist.
6209 ;; 'arglist Some other type of arglist.
a85fd6da
AM
6210 ;; nil Some other context or unknown context. Includes
6211 ;; within the parens of an if, for, ... construct.
0386b551
AM
6212 ;;
6213 ;; LAST-CAST-END is the first token after the closing paren of a
6214 ;; preceding cast, or nil if none is known. If
6215 ;; `c-forward-decl-or-cast-1' is used in succession, it should be
6216 ;; the position after the closest preceding call where a cast was
6217 ;; matched. In that case it's used to discover chains of casts like
6218 ;; "(a) (b) c".
6219 ;;
6220 ;; This function records identifier ranges on
6221 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
6222 ;; `c-record-type-identifiers' is non-nil.
6223 ;;
6224 ;; This function might do hidden buffer changes.
6225
6226 (let (;; `start-pos' is used below to point to the start of the
6227 ;; first type, i.e. after any leading specifiers. It might
6228 ;; also point at the beginning of the preceding syntactic
6229 ;; whitespace.
6230 (start-pos (point))
6231 ;; Set to the result of `c-forward-type'.
6232 at-type
6233 ;; The position of the first token in what we currently
6234 ;; believe is the type in the declaration or cast, after any
6235 ;; specifiers and their associated clauses.
6236 type-start
6237 ;; The position of the first token in what we currently
6238 ;; believe is the declarator for the first identifier. Set
6239 ;; when the type is found, and moved forward over any
6240 ;; `c-decl-hangon-kwds' and their associated clauses that
6241 ;; occurs after the type.
6242 id-start
6243 ;; These store `at-type', `type-start' and `id-start' of the
6244 ;; identifier before the one in those variables. The previous
6245 ;; identifier might turn out to be the real type in a
6246 ;; declaration if the last one has to be the declarator in it.
6247 ;; If `backup-at-type' is nil then the other variables have
6248 ;; undefined values.
6249 backup-at-type backup-type-start backup-id-start
e15f8aaa
AM
6250 ;; Set if we've found a specifier (apart from "typedef") that makes
6251 ;; the defined identifier(s) types.
0386b551 6252 at-type-decl
e15f8aaa
AM
6253 ;; Set if we've a "typedef" keyword.
6254 at-typedef
0386b551
AM
6255 ;; Set if we've found a specifier that can start a declaration
6256 ;; where there's no type.
6257 maybe-typeless
6258 ;; If a specifier is found that also can be a type prefix,
6259 ;; these flags are set instead of those above. If we need to
6260 ;; back up an identifier, they are copied to the real flag
6261 ;; variables. Thus they only take effect if we fail to
6262 ;; interpret it as a type.
6263 backup-at-type-decl backup-maybe-typeless
6264 ;; Whether we've found a declaration or a cast. We might know
6265 ;; this before we've found the type in it. It's 'ids if we've
6266 ;; found two consecutive identifiers (usually a sure sign, but
6267 ;; we should allow that in labels too), and t if we've found a
6268 ;; specifier keyword (a 100% sure sign).
6269 at-decl-or-cast
6270 ;; Set when we need to back up to parse this as a declaration
6271 ;; but not as a cast.
6272 backup-if-not-cast
6273 ;; For casts, the return position.
6274 cast-end
6275 ;; Save `c-record-type-identifiers' and
6276 ;; `c-record-ref-identifiers' since ranges are recorded
6277 ;; speculatively and should be thrown away if it turns out
6278 ;; that it isn't a declaration or cast.
6279 (save-rec-type-ids c-record-type-identifiers)
6280 (save-rec-ref-ids c-record-ref-identifiers))
6281
452ea855
AM
6282 (while (c-forward-annotation)
6283 (c-forward-syntactic-ws))
6284
0386b551
AM
6285 ;; Check for a type. Unknown symbols are treated as possible
6286 ;; types, but they could also be specifiers disguised through
6287 ;; macros like __INLINE__, so we recognize both types and known
6288 ;; specifiers after them too.
6289 (while
6290 (let* ((start (point)) kwd-sym kwd-clause-end found-type)
6291
6292 ;; Look for a specifier keyword clause.
6293 (when (looking-at c-prefix-spec-kwds-re)
e15f8aaa
AM
6294 (if (looking-at c-typedef-key)
6295 (setq at-typedef t))
0386b551
AM
6296 (setq kwd-sym (c-keyword-sym (match-string 1)))
6297 (save-excursion
6298 (c-forward-keyword-clause 1)
6299 (setq kwd-clause-end (point))))
6300
e15f8aaa 6301 (when (setq found-type (c-forward-type t)) ; brace-block-too
0386b551
AM
6302 ;; Found a known or possible type or a prefix of a known type.
6303
6304 (when at-type
6305 ;; Got two identifiers with nothing but whitespace
6306 ;; between them. That can only happen in declarations.
6307 (setq at-decl-or-cast 'ids)
6308
6309 (when (eq at-type 'found)
6310 ;; If the previous identifier is a found type we
6311 ;; record it as a real one; it might be some sort of
6312 ;; alias for a prefix like "unsigned".
6313 (save-excursion
6314 (goto-char type-start)
6315 (let ((c-promote-possible-types t))
6316 (c-forward-type)))))
6317
6318 (setq backup-at-type at-type
6319 backup-type-start type-start
6320 backup-id-start id-start
6321 at-type found-type
6322 type-start start
6323 id-start (point)
6324 ;; The previous ambiguous specifier/type turned out
6325 ;; to be a type since we've parsed another one after
6326 ;; it, so clear these backup flags.
6327 backup-at-type-decl nil
6328 backup-maybe-typeless nil))
6329
6330 (if kwd-sym
6331 (progn
6332 ;; Handle known specifier keywords and
6333 ;; `c-decl-hangon-kwds' which can occur after known
6334 ;; types.
6335
6336 (if (c-keyword-member kwd-sym 'c-decl-hangon-kwds)
6337 ;; It's a hang-on keyword that can occur anywhere.
6338 (progn
6339 (setq at-decl-or-cast t)
6340 (if at-type
6341 ;; Move the identifier start position if
6342 ;; we've passed a type.
6343 (setq id-start kwd-clause-end)
6344 ;; Otherwise treat this as a specifier and
6345 ;; move the fallback position.
6346 (setq start-pos kwd-clause-end))
6347 (goto-char kwd-clause-end))
6348
6349 ;; It's an ordinary specifier so we know that
6350 ;; anything before this can't be the type.
6351 (setq backup-at-type nil
6352 start-pos kwd-clause-end)
6353
6354 (if found-type
6355 ;; It's ambiguous whether this keyword is a
6356 ;; specifier or a type prefix, so set the backup
6357 ;; flags. (It's assumed that `c-forward-type'
6358 ;; moved further than `c-forward-keyword-clause'.)
6359 (progn
6360 (when (c-keyword-member kwd-sym 'c-typedef-decl-kwds)
6361 (setq backup-at-type-decl t))
6362 (when (c-keyword-member kwd-sym 'c-typeless-decl-kwds)
6363 (setq backup-maybe-typeless t)))
6364
6365 (when (c-keyword-member kwd-sym 'c-typedef-decl-kwds)
e15f8aaa
AM
6366 ;; This test only happens after we've scanned a type.
6367 ;; So, with valid syntax, kwd-sym can't be 'typedef.
0386b551
AM
6368 (setq at-type-decl t))
6369 (when (c-keyword-member kwd-sym 'c-typeless-decl-kwds)
6370 (setq maybe-typeless t))
6371
6372 ;; Haven't matched a type so it's an umambiguous
6373 ;; specifier keyword and we know we're in a
6374 ;; declaration.
6375 (setq at-decl-or-cast t)
6376
6377 (goto-char kwd-clause-end))))
6378
6379 ;; If the type isn't known we continue so that we'll jump
6380 ;; over all specifiers and type identifiers. The reason
6381 ;; to do this for a known type prefix is to make things
6382 ;; like "unsigned INT16" work.
6383 (and found-type (not (eq found-type t))))))
6384
6385 (cond
6386 ((eq at-type t)
6387 ;; If a known type was found, we still need to skip over any
6388 ;; hangon keyword clauses after it. Otherwise it has already
6389 ;; been done in the loop above.
6390 (while (looking-at c-decl-hangon-key)
6391 (c-forward-keyword-clause 1))
6392 (setq id-start (point)))
6393
6394 ((eq at-type 'prefix)
6395 ;; A prefix type is itself a primitive type when it's not
6396 ;; followed by another type.
6397 (setq at-type t))
6398
6399 ((not at-type)
6400 ;; Got no type but set things up to continue anyway to handle
6401 ;; the various cases when a declaration doesn't start with a
6402 ;; type.
6403 (setq id-start start-pos))
6404
6405 ((and (eq at-type 'maybe)
6406 (c-major-mode-is 'c++-mode))
6407 ;; If it's C++ then check if the last "type" ends on the form
6408 ;; "foo::foo" or "foo::~foo", i.e. if it's the name of a
6409 ;; (con|de)structor.
6410 (save-excursion
6411 (let (name end-2 end-1)
6412 (goto-char id-start)
6413 (c-backward-syntactic-ws)
6414 (setq end-2 (point))
6415 (when (and
6416 (c-simple-skip-symbol-backward)
6417 (progn
6418 (setq name
6419 (buffer-substring-no-properties (point) end-2))
6420 ;; Cheating in the handling of syntactic ws below.
6421 (< (skip-chars-backward ":~ \t\n\r\v\f") 0))
6422 (progn
6423 (setq end-1 (point))
6424 (c-simple-skip-symbol-backward))
6425 (>= (point) type-start)
6426 (equal (buffer-substring-no-properties (point) end-1)
6427 name))
6428 ;; It is a (con|de)structor name. In that case the
6429 ;; declaration is typeless so zap out any preceding
6430 ;; identifier(s) that we might have taken as types.
6431 (goto-char type-start)
6432 (setq at-type nil
6433 backup-at-type nil
6434 id-start type-start))))))
6435
6436 ;; Check for and step over a type decl expression after the thing
6437 ;; that is or might be a type. This can't be skipped since we
6438 ;; need the correct end position of the declarator for
6439 ;; `max-type-decl-end-*'.
6440 (let ((start (point)) (paren-depth 0) pos
6441 ;; True if there's a non-open-paren match of
6442 ;; `c-type-decl-prefix-key'.
6443 got-prefix
6444 ;; True if the declarator is surrounded by a parenthesis pair.
6445 got-parens
6446 ;; True if there is an identifier in the declarator.
6447 got-identifier
6448 ;; True if there's a non-close-paren match of
6449 ;; `c-type-decl-suffix-key'.
6450 got-suffix
6451 ;; True if there's a prefix match outside the outermost
6452 ;; paren pair that surrounds the declarator.
6453 got-prefix-before-parens
1379f2c5 6454 ;; True if there's a suffix match outside the outermost
0386b551
AM
6455 ;; paren pair that surrounds the declarator. The value is
6456 ;; the position of the first suffix match.
6457 got-suffix-after-parens
6458 ;; True if we've parsed the type decl to a token that is
6459 ;; known to end declarations in this context.
6460 at-decl-end
6461 ;; The earlier values of `at-type' and `type-start' if we've
6462 ;; shifted the type backwards.
6463 identifier-type identifier-start
6464 ;; If `c-parse-and-markup-<>-arglists' is set we need to
6465 ;; turn it off during the name skipping below to avoid
6466 ;; getting `c-type' properties that might be bogus. That
6467 ;; can happen since we don't know if
6468 ;; `c-restricted-<>-arglists' will be correct inside the
6469 ;; arglist paren that gets entered.
6470 c-parse-and-markup-<>-arglists)
6471
6472 (goto-char id-start)
6473
6474 ;; Skip over type decl prefix operators. (Note similar code in
6475 ;; `c-font-lock-declarators'.)
6476 (while (and (looking-at c-type-decl-prefix-key)
6477 (if (and (c-major-mode-is 'c++-mode)
05842630 6478 (match-beginning 3))
0386b551
AM
6479 ;; If the second submatch matches in C++ then
6480 ;; we're looking at an identifier that's a
6481 ;; prefix only if it specifies a member pointer.
6482 (when (setq got-identifier (c-forward-name))
6483 (if (looking-at "\\(::\\)")
6484 ;; We only check for a trailing "::" and
6485 ;; let the "*" that should follow be
6486 ;; matched in the next round.
6487 (progn (setq got-identifier nil) t)
6488 ;; It turned out to be the real identifier,
6489 ;; so stop.
6490 nil))
6491 t))
6492
6493 (if (eq (char-after) ?\()
6494 (progn
6495 (setq paren-depth (1+ paren-depth))
6496 (forward-char))
6497 (unless got-prefix-before-parens
6498 (setq got-prefix-before-parens (= paren-depth 0)))
6499 (setq got-prefix t)
6500 (goto-char (match-end 1)))
6501 (c-forward-syntactic-ws))
6502
6503 (setq got-parens (> paren-depth 0))
6504
6505 ;; Skip over an identifier.
6506 (or got-identifier
6507 (and (looking-at c-identifier-start)
6508 (setq got-identifier (c-forward-name))))
6509
6510 ;; Skip over type decl suffix operators.
6511 (while (if (looking-at c-type-decl-suffix-key)
6512
6513 (if (eq (char-after) ?\))
6514 (when (> paren-depth 0)
6515 (setq paren-depth (1- paren-depth))
6516 (forward-char)
6517 t)
6518 (when (if (save-match-data (looking-at "\\s\("))
6519 (c-safe (c-forward-sexp 1) t)
6520 (goto-char (match-end 1))
6521 t)
6522 (when (and (not got-suffix-after-parens)
6523 (= paren-depth 0))
6524 (setq got-suffix-after-parens (match-beginning 0)))
6525 (setq got-suffix t)))
6526
6527 ;; No suffix matched. We might have matched the
6528 ;; identifier as a type and the open paren of a
6529 ;; function arglist as a type decl prefix. In that
6530 ;; case we should "backtrack": Reinterpret the last
6531 ;; type as the identifier, move out of the arglist and
6532 ;; continue searching for suffix operators.
6533 ;;
6534 ;; Do this even if there's no preceding type, to cope
6535 ;; with old style function declarations in K&R C,
6536 ;; (con|de)structors in C++ and `c-typeless-decl-kwds'
6537 ;; style declarations. That isn't applicable in an
6538 ;; arglist context, though.
6539 (when (and (= paren-depth 1)
6540 (not got-prefix-before-parens)
6541 (not (eq at-type t))
6542 (or backup-at-type
6543 maybe-typeless
6544 backup-maybe-typeless
6545 (when c-recognize-typeless-decls
6546 (not context)))
6547 (setq pos (c-up-list-forward (point)))
6548 (eq (char-before pos) ?\)))
6549 (c-fdoc-shift-type-backward)
6550 (goto-char pos)
6551 t))
6552
6553 (c-forward-syntactic-ws))
6554
6555 (when (and (or maybe-typeless backup-maybe-typeless)
6556 (not got-identifier)
6557 (not got-prefix)
6558 at-type)
6559 ;; Have found no identifier but `c-typeless-decl-kwds' has
6560 ;; matched so we know we're inside a declaration. The
6561 ;; preceding type must be the identifier instead.
6562 (c-fdoc-shift-type-backward))
6563
6564 (setq
6565 at-decl-or-cast
6566 (catch 'at-decl-or-cast
6567
a85fd6da 6568 ;; CASE 1
0386b551
AM
6569 (when (> paren-depth 0)
6570 ;; Encountered something inside parens that isn't matched by
6571 ;; the `c-type-decl-*' regexps, so it's not a type decl
6572 ;; expression. Try to skip out to the same paren depth to
6573 ;; not confuse the cast check below.
6574 (c-safe (goto-char (scan-lists (point) 1 paren-depth)))
6575 ;; If we've found a specifier keyword then it's a
6576 ;; declaration regardless.
6577 (throw 'at-decl-or-cast (eq at-decl-or-cast t)))
6578
6579 (setq at-decl-end
6580 (looking-at (cond ((eq context '<>) "[,>]")
6581 (context "[,\)]")
6582 (t "[,;]"))))
6583
6584 ;; Now we've collected info about various characteristics of
6585 ;; the construct we're looking at. Below follows a decision
6586 ;; tree based on that. It's ordered to check more certain
6587 ;; signs before less certain ones.
6588
6589 (if got-identifier
6590 (progn
6591
a85fd6da 6592 ;; CASE 2
0386b551
AM
6593 (when (and (or at-type maybe-typeless)
6594 (not (or got-prefix got-parens)))
6595 ;; Got another identifier directly after the type, so it's a
6596 ;; declaration.
6597 (throw 'at-decl-or-cast t))
6598
6599 (when (and got-parens
6600 (not got-prefix)
6601 (not got-suffix-after-parens)
6602 (or backup-at-type
6603 maybe-typeless
6604 backup-maybe-typeless))
6605 ;; Got a declaration of the form "foo bar (gnu);" where we've
6606 ;; recognized "bar" as the type and "gnu" as the declarator.
6607 ;; In this case it's however more likely that "bar" is the
6608 ;; declarator and "gnu" a function argument or initializer (if
6609 ;; `c-recognize-paren-inits' is set), since the parens around
6610 ;; "gnu" would be superfluous if it's a declarator. Shift the
6611 ;; type one step backward.
6612 (c-fdoc-shift-type-backward)))
6613
6614 ;; Found no identifier.
6615
6616 (if backup-at-type
6617 (progn
6618
a85fd6da 6619 ;; CASE 3
0386b551
AM
6620 (when (= (point) start)
6621 ;; Got a plain list of identifiers. If a colon follows it's
452ea855
AM
6622 ;; a valid label. Otherwise the last one probably is the
6623 ;; declared identifier and we should back up to the previous
6624 ;; type, providing it isn't a cast.
6625 (if (and (eq (char-after) ?:)
6626 (not (c-major-mode-is 'java-mode)))
6627 ;; If we've found a specifier keyword then it's a
6628 ;; declaration regardless.
6629 (throw 'at-decl-or-cast (eq at-decl-or-cast t))
0386b551
AM
6630 (setq backup-if-not-cast t)
6631 (throw 'at-decl-or-cast t)))
6632
a85fd6da 6633 ;; CASE 4
0386b551
AM
6634 (when (and got-suffix
6635 (not got-prefix)
6636 (not got-parens))
6637 ;; Got a plain list of identifiers followed by some suffix.
6638 ;; If this isn't a cast then the last identifier probably is
6639 ;; the declared one and we should back up to the previous
6640 ;; type.
6641 (setq backup-if-not-cast t)
6642 (throw 'at-decl-or-cast t)))
6643
a85fd6da 6644 ;; CASE 5
0386b551
AM
6645 (when (eq at-type t)
6646 ;; If the type is known we know that there can't be any
6647 ;; identifier somewhere else, and it's only in declarations in
6648 ;; e.g. function prototypes and in casts that the identifier may
6649 ;; be left out.
6650 (throw 'at-decl-or-cast t))
6651
6652 (when (= (point) start)
6653 ;; Only got a single identifier (parsed as a type so far).
a85fd6da 6654 ;; CASE 6
0386b551
AM
6655 (if (and
6656 ;; Check that the identifier isn't at the start of an
6657 ;; expression.
6658 at-decl-end
6659 (cond
6660 ((eq context 'decl)
6661 ;; Inside an arglist that contains declarations. If K&R
6662 ;; style declarations and parenthesis style initializers
6663 ;; aren't allowed then the single identifier must be a
6664 ;; type, else we require that it's known or found
6665 ;; (primitive types are handled above).
6666 (or (and (not c-recognize-knr-p)
6667 (not c-recognize-paren-inits))
6668 (memq at-type '(known found))))
6669 ((eq context '<>)
6670 ;; Inside a template arglist. Accept known and found
6671 ;; types; other identifiers could just as well be
6672 ;; constants in C++.
6673 (memq at-type '(known found)))))
6674 (throw 'at-decl-or-cast t)
a85fd6da 6675 ;; CASE 7
0386b551
AM
6676 ;; Can't be a valid declaration or cast, but if we've found a
6677 ;; specifier it can't be anything else either, so treat it as
6678 ;; an invalid/unfinished declaration or cast.
6679 (throw 'at-decl-or-cast at-decl-or-cast))))
6680
6681 (if (and got-parens
6682 (not got-prefix)
6683 (not context)
6684 (not (eq at-type t))
6685 (or backup-at-type
6686 maybe-typeless
6687 backup-maybe-typeless
6688 (when c-recognize-typeless-decls
6689 (or (not got-suffix)
6690 (not (looking-at
6691 c-after-suffixed-type-maybe-decl-key))))))
6692 ;; Got an empty paren pair and a preceding type that probably
6693 ;; really is the identifier. Shift the type backwards to make
6694 ;; the last one the identifier. This is analogous to the
6695 ;; "backtracking" done inside the `c-type-decl-suffix-key' loop
6696 ;; above.
6697 ;;
6698 ;; Exception: In addition to the conditions in that
6699 ;; "backtracking" code, do not shift backward if we're not
6700 ;; looking at either `c-after-suffixed-type-decl-key' or "[;,]".
6701 ;; Since there's no preceding type, the shift would mean that
6702 ;; the declaration is typeless. But if the regexp doesn't match
6703 ;; then we will simply fall through in the tests below and not
6704 ;; recognize it at all, so it's better to try it as an abstract
6705 ;; declarator instead.
6706 (c-fdoc-shift-type-backward)
6707
6708 ;; Still no identifier.
a85fd6da 6709 ;; CASE 8
0386b551
AM
6710 (when (and got-prefix (or got-parens got-suffix))
6711 ;; Require `got-prefix' together with either `got-parens' or
6712 ;; `got-suffix' to recognize it as an abstract declarator:
6713 ;; `got-parens' only is probably an empty function call.
6714 ;; `got-suffix' only can build an ordinary expression together
6715 ;; with the preceding identifier which we've taken as a type.
6716 ;; We could actually accept on `got-prefix' only, but that can
6717 ;; easily occur temporarily while writing an expression so we
6718 ;; avoid that case anyway. We could do a better job if we knew
6719 ;; the point when the fontification was invoked.
6720 (throw 'at-decl-or-cast t))
6721
a85fd6da 6722 ;; CASE 9
0386b551
AM
6723 (when (and at-type
6724 (not got-prefix)
6725 (not got-parens)
6726 got-suffix-after-parens
6727 (eq (char-after got-suffix-after-parens) ?\())
6728 ;; Got a type, no declarator but a paren suffix. I.e. it's a
6729 ;; normal function call afterall (or perhaps a C++ style object
6730 ;; instantiation expression).
6731 (throw 'at-decl-or-cast nil))))
6732
a85fd6da 6733 ;; CASE 10
0386b551
AM
6734 (when at-decl-or-cast
6735 ;; By now we've located the type in the declaration that we know
6736 ;; we're in.
6737 (throw 'at-decl-or-cast t))
6738
a85fd6da 6739 ;; CASE 11
0386b551
AM
6740 (when (and got-identifier
6741 (not context)
6742 (looking-at c-after-suffixed-type-decl-key)
6743 (if (and got-parens
6744 (not got-prefix)
6745 (not got-suffix)
6746 (not (eq at-type t)))
6747 ;; Shift the type backward in the case that there's a
6748 ;; single identifier inside parens. That can only
6749 ;; occur in K&R style function declarations so it's
6750 ;; more likely that it really is a function call.
6751 ;; Therefore we only do this after
6752 ;; `c-after-suffixed-type-decl-key' has matched.
6753 (progn (c-fdoc-shift-type-backward) t)
6754 got-suffix-after-parens))
6755 ;; A declaration according to `c-after-suffixed-type-decl-key'.
6756 (throw 'at-decl-or-cast t))
6757
a85fd6da 6758 ;; CASE 12
0386b551
AM
6759 (when (and (or got-prefix (not got-parens))
6760 (memq at-type '(t known)))
6761 ;; It's a declaration if a known type precedes it and it can't be a
6762 ;; function call.
6763 (throw 'at-decl-or-cast t))
6764
6765 ;; If we get here we can't tell if this is a type decl or a normal
6766 ;; expression by looking at it alone. (That's under the assumption
6767 ;; that normal expressions always can look like type decl expressions,
6768 ;; which isn't really true but the cases where it doesn't hold are so
6769 ;; uncommon (e.g. some placements of "const" in C++) it's not worth
6770 ;; the effort to look for them.)
6771
6772 (unless (or at-decl-end (looking-at "=[^=]"))
6773 ;; If this is a declaration it should end here or its initializer(*)
6774 ;; should start here, so check for allowed separation tokens. Note
6775 ;; that this rule doesn't work e.g. with a K&R arglist after a
6776 ;; function header.
6777 ;;
6778 ;; *) Don't check for C++ style initializers using parens
6779 ;; since those already have been matched as suffixes.
6780 ;;
6781 ;; If `at-decl-or-cast' is then we've found some other sign that
6782 ;; it's a declaration or cast, so then it's probably an
6783 ;; invalid/unfinished one.
6784 (throw 'at-decl-or-cast at-decl-or-cast))
6785
6786 ;; Below are tests that only should be applied when we're certain to
6787 ;; not have parsed halfway through an expression.
6788
a85fd6da 6789 ;; CASE 14
0386b551
AM
6790 (when (memq at-type '(t known))
6791 ;; The expression starts with a known type so treat it as a
6792 ;; declaration.
6793 (throw 'at-decl-or-cast t))
6794
a85fd6da 6795 ;; CASE 15
0386b551
AM
6796 (when (and (c-major-mode-is 'c++-mode)
6797 ;; In C++ we check if the identifier is a known type, since
6798 ;; (con|de)structors use the class name as identifier.
6799 ;; We've always shifted over the identifier as a type and
6800 ;; then backed up again in this case.
6801 identifier-type
6802 (or (memq identifier-type '(found known))
6803 (and (eq (char-after identifier-start) ?~)
6804 ;; `at-type' probably won't be 'found for
6805 ;; destructors since the "~" is then part of the
6806 ;; type name being checked against the list of
6807 ;; known types, so do a check without that
6808 ;; operator.
6809 (or (save-excursion
6810 (goto-char (1+ identifier-start))
6811 (c-forward-syntactic-ws)
6812 (c-with-syntax-table
6813 c-identifier-syntax-table
6814 (looking-at c-known-type-key)))
6815 (save-excursion
6816 (goto-char (1+ identifier-start))
6817 ;; We have already parsed the type earlier,
6818 ;; so it'd be possible to cache the end
6819 ;; position instead of redoing it here, but
6820 ;; then we'd need to keep track of another
6821 ;; position everywhere.
6822 (c-check-type (point)
6823 (progn (c-forward-type)
6824 (point))))))))
6825 (throw 'at-decl-or-cast t))
6826
6827 (if got-identifier
6828 (progn
a85fd6da 6829 ;; CASE 16
0386b551
AM
6830 (when (and got-prefix-before-parens
6831 at-type
6832 (or at-decl-end (looking-at "=[^=]"))
6833 (not context)
6834 (not got-suffix))
6835 ;; Got something like "foo * bar;". Since we're not inside an
6836 ;; arglist it would be a meaningless expression because the
6837 ;; result isn't used. We therefore choose to recognize it as
6838 ;; a declaration. Do not allow a suffix since it could then
6839 ;; be a function call.
6840 (throw 'at-decl-or-cast t))
6841
a85fd6da 6842 ;; CASE 17
0386b551
AM
6843 (when (and (or got-suffix-after-parens
6844 (looking-at "=[^=]"))
6845 (eq at-type 'found)
6846 (not (eq context 'arglist)))
6847 ;; Got something like "a (*b) (c);" or "a (b) = c;". It could
6848 ;; be an odd expression or it could be a declaration. Treat
6849 ;; it as a declaration if "a" has been used as a type
6850 ;; somewhere else (if it's a known type we won't get here).
6851 (throw 'at-decl-or-cast t)))
6852
a85fd6da 6853 ;; CASE 18
0386b551
AM
6854 (when (and context
6855 (or got-prefix
6856 (and (eq context 'decl)
6857 (not c-recognize-paren-inits)
6858 (or got-parens got-suffix))))
6859 ;; Got a type followed by an abstract declarator. If `got-prefix'
6860 ;; is set it's something like "a *" without anything after it. If
6861 ;; `got-parens' or `got-suffix' is set it's "a()", "a[]", "a()[]",
6862 ;; or similar, which we accept only if the context rules out
6863 ;; expressions.
6864 (throw 'at-decl-or-cast t)))
6865
6866 ;; If we had a complete symbol table here (which rules out
6867 ;; `c-found-types') we should return t due to the disambiguation rule
6868 ;; (in at least C++) that anything that can be parsed as a declaration
6869 ;; is a declaration. Now we're being more defensive and prefer to
6870 ;; highlight things like "foo (bar);" as a declaration only if we're
6871 ;; inside an arglist that contains declarations.
6872 (eq context 'decl))))
6873
6874 ;; The point is now after the type decl expression.
6875
6876 (cond
6877 ;; Check for a cast.
6878 ((save-excursion
6879 (and
6880 c-cast-parens
6881
6882 ;; Should be the first type/identifier in a cast paren.
6883 (> preceding-token-end (point-min))
6884 (memq (char-before preceding-token-end) c-cast-parens)
6885
6886 ;; The closing paren should follow.
6887 (progn
6888 (c-forward-syntactic-ws)
6889 (looking-at "\\s\)"))
6890
6891 ;; There should be a primary expression after it.
6892 (let (pos)
6893 (forward-char)
6894 (c-forward-syntactic-ws)
6895 (setq cast-end (point))
6896 (and (looking-at c-primary-expr-regexp)
6897 (progn
6898 (setq pos (match-end 0))
6899 (or
6900 ;; Check if the expression begins with a prefix keyword.
6901 (match-beginning 2)
6902 (if (match-beginning 1)
6903 ;; Expression begins with an ambiguous operator. Treat
6904 ;; it as a cast if it's a type decl or if we've
6905 ;; recognized the type somewhere else.
6906 (or at-decl-or-cast
6907 (memq at-type '(t known found)))
6908 ;; Unless it's a keyword, it's the beginning of a primary
6909 ;; expression.
6910 (not (looking-at c-keywords-regexp)))))
6911 ;; If `c-primary-expr-regexp' matched a nonsymbol token, check
6912 ;; that it matched a whole one so that we don't e.g. confuse
6913 ;; the operator '-' with '->'. It's ok if it matches further,
6914 ;; though, since it e.g. can match the float '.5' while the
6915 ;; operator regexp only matches '.'.
6916 (or (not (looking-at c-nonsymbol-token-regexp))
6917 (<= (match-end 0) pos))))
6918
6919 ;; There should either be a cast before it or something that isn't an
6920 ;; identifier or close paren.
6921 (> preceding-token-end (point-min))
6922 (progn
6923 (goto-char (1- preceding-token-end))
6924 (or (eq (point) last-cast-end)
6925 (progn
6926 (c-backward-syntactic-ws)
6927 (if (< (skip-syntax-backward "w_") 0)
6928 ;; It's a symbol. Accept it only if it's one of the
6929 ;; keywords that can precede an expression (without
6930 ;; surrounding parens).
6931 (looking-at c-simple-stmt-key)
6932 (and
6933 ;; Check that it isn't a close paren (block close is ok,
6934 ;; though).
6935 (not (memq (char-before) '(?\) ?\])))
6936 ;; Check that it isn't a nonsymbol identifier.
6937 (not (c-on-identifier)))))))))
6938
6939 ;; Handle the cast.
6940 (when (and c-record-type-identifiers at-type (not (eq at-type t)))
6941 (let ((c-promote-possible-types t))
6942 (goto-char type-start)
6943 (c-forward-type)))
6944
6945 (goto-char cast-end)
6946 'cast)
6947
6948 (at-decl-or-cast
6949 ;; We're at a declaration. Highlight the type and the following
6950 ;; declarators.
6951
6952 (when backup-if-not-cast
6953 (c-fdoc-shift-type-backward t))
6954
6955 (when (and (eq context 'decl) (looking-at ","))
6956 ;; Make sure to propagate the `c-decl-arg-start' property to
6957 ;; the next argument if it's set in this one, to cope with
6958 ;; interactive refontification.
6959 (c-put-c-type-property (point) 'c-decl-arg-start))
6960
6961 (when (and c-record-type-identifiers at-type (not (eq at-type t)))
6962 (let ((c-promote-possible-types t))
6963 (save-excursion
6964 (goto-char type-start)
6965 (c-forward-type))))
6966
e15f8aaa
AM
6967 (cons id-start
6968 (and (or at-type-decl at-typedef)
6969 (cons at-type-decl at-typedef))))
0386b551
AM
6970
6971 (t
6972 ;; False alarm. Restore the recorded ranges.
6973 (setq c-record-type-identifiers save-rec-type-ids
6974 c-record-ref-identifiers save-rec-ref-ids)
6975 nil))))
6976
6977(defun c-forward-label (&optional assume-markup preceding-token-end limit)
51c9af45 6978 ;; Assuming that point is at the beginning of a token, check if it starts a
1379f2c5
AM
6979 ;; label and if so move over it and return non-nil (t in default situations,
6980 ;; specific symbols (see below) for interesting situations), otherwise don't
6981 ;; move and return nil. "Label" here means "most things with a colon".
51c9af45
AM
6982 ;;
6983 ;; More precisely, a "label" is regarded as one of:
1379f2c5
AM
6984 ;; (i) a goto target like "foo:" - returns the symbol `goto-target';
6985 ;; (ii) A case label - either the entire construct "case FOO:", or just the
6986 ;; bare "case", should the colon be missing. We return t;
6987 ;; (iii) a keyword which needs a colon, like "default:" or "private:"; We
6988 ;; return t;
51c9af45 6989 ;; (iv) One of QT's "extended" C++ variants of
1379f2c5
AM
6990 ;; "private:"/"protected:"/"public:"/"more:" looking like "public slots:".
6991 ;; Returns the symbol `qt-2kwds-colon'.
6992 ;; (v) QT's construct "signals:". Returns the symbol `qt-1kwd-colon'.
82ba65cf 6993 ;; (vi) One of the keywords matched by `c-opt-extra-label-key' (without any
51c9af45 6994 ;; colon). Currently (2006-03), this applies only to Objective C's
1379f2c5 6995 ;; keywords "@private", "@protected", and "@public". Returns t.
51c9af45
AM
6996 ;;
6997 ;; One of the things which will NOT be recognised as a label is a bit-field
6998 ;; element of a struct, something like "int foo:5".
6999 ;;
7000 ;; The end of the label is taken to be just after the colon, or the end of
7001 ;; the first submatch in `c-opt-extra-label-key'. The point is directly
7002 ;; after the end on return. The terminating char gets marked with
7003 ;; `c-decl-end' to improve recognition of the following declaration or
7004 ;; statement.
0386b551
AM
7005 ;;
7006 ;; If ASSUME-MARKUP is non-nil, it's assumed that the preceding
51c9af45 7007 ;; label, if any, has already been marked up like that.
0386b551
AM
7008 ;;
7009 ;; If PRECEDING-TOKEN-END is given, it should be the first position
7010 ;; after the preceding token, i.e. on the other side of the
7011 ;; syntactic ws from the point. Use a value less than or equal to
7012 ;; (point-min) if the point is at the first token in (the visible
7013 ;; part of) the buffer.
7014 ;;
7015 ;; The optional LIMIT limits the forward scan for the colon.
7016 ;;
7017 ;; This function records the ranges of the label symbols on
7018 ;; `c-record-ref-identifiers' if `c-record-type-identifiers' (!) is
7019 ;; non-nil.
7020 ;;
7021 ;; This function might do hidden buffer changes.
7022
51c9af45 7023 (let ((start (point))
1379f2c5 7024 label-end
51c9af45 7025 qt-symbol-idx
1379f2c5 7026 macro-start ; if we're in one.
f412a567
AM
7027 label-type
7028 kwd)
0386b551 7029 (cond
b414f371 7030 ;; "case" or "default" (Doesn't apply to AWK).
0386b551
AM
7031 ((looking-at c-label-kwds-regexp)
7032 (let ((kwd-end (match-end 1)))
7033 ;; Record only the keyword itself for fontification, since in
7034 ;; case labels the following is a constant expression and not
7035 ;; a label.
7036 (when c-record-type-identifiers
7037 (c-record-ref-id (cons (match-beginning 1) kwd-end)))
7038
7039 ;; Find the label end.
7040 (goto-char kwd-end)
1379f2c5
AM
7041 (setq label-type
7042 (if (and (c-syntactic-re-search-forward
7043 ;; Stop on chars that aren't allowed in expressions,
7044 ;; and on operator chars that would be meaningless
7045 ;; there. FIXME: This doesn't cope with ?: operators.
7046 "[;{=,@]\\|\\(\\=\\|[^:]\\):\\([^:]\\|\\'\\)"
7047 limit t t nil 1)
7048 (match-beginning 2))
7049
7050 (progn ; there's a proper :
7051 (goto-char (match-beginning 2)) ; just after the :
7052 (c-put-c-type-property (1- (point)) 'c-decl-end)
7053 t)
7054
7055 ;; It's an unfinished label. We consider the keyword enough
7056 ;; to recognize it as a label, so that it gets fontified.
7057 ;; Leave the point at the end of it, but don't put any
7058 ;; `c-decl-end' marker.
7059 (goto-char kwd-end)
7060 t))))
0386b551 7061
51c9af45 7062 ;; @private, @protected, @public, in Objective C, or similar.
0386b551
AM
7063 ((and c-opt-extra-label-key
7064 (looking-at c-opt-extra-label-key))
7065 ;; For a `c-opt-extra-label-key' match, we record the whole
7066 ;; thing for fontification. That's to get the leading '@' in
7067 ;; Objective-C protection labels fontified.
7068 (goto-char (match-end 1))
7069 (when c-record-type-identifiers
7070 (c-record-ref-id (cons (match-beginning 1) (point))))
7071 (c-put-c-type-property (1- (point)) 'c-decl-end)
1379f2c5 7072 (setq label-type t))
0386b551 7073
51c9af45
AM
7074 ;; All other cases of labels.
7075 ((and c-recognize-colon-labels ; nil for AWK and IDL, otherwise t.
0386b551
AM
7076
7077 ;; A colon label must have something before the colon.
7078 (not (eq (char-after) ?:))
7079
7080 ;; Check that we're not after a token that can't precede a label.
7081 (or
7082 ;; Trivially succeeds when there's no preceding token.
7083 (if preceding-token-end
7084 (<= preceding-token-end (point-min))
7085 (save-excursion
7086 (c-backward-syntactic-ws)
7087 (setq preceding-token-end (point))
7088 (bobp)))
7089
7090 ;; Check if we're after a label, if we're after a closing
7091 ;; paren that belong to statement, and with
7092 ;; `c-label-prefix-re'. It's done in different order
7093 ;; depending on `assume-markup' since the checks have
7094 ;; different expensiveness.
7095 (if assume-markup
7096 (or
7097 (eq (c-get-char-property (1- preceding-token-end) 'c-type)
7098 'c-decl-end)
7099
7100 (save-excursion
7101 (goto-char (1- preceding-token-end))
7102 (c-beginning-of-current-token)
51c9af45
AM
7103 (or (looking-at c-label-prefix-re)
7104 (looking-at c-block-stmt-1-key)))
0386b551
AM
7105
7106 (and (eq (char-before preceding-token-end) ?\))
7107 (c-after-conditional)))
7108
7109 (or
7110 (save-excursion
7111 (goto-char (1- preceding-token-end))
7112 (c-beginning-of-current-token)
51c9af45
AM
7113 (or (looking-at c-label-prefix-re)
7114 (looking-at c-block-stmt-1-key)))
0386b551
AM
7115
7116 (cond
7117 ((eq (char-before preceding-token-end) ?\))
7118 (c-after-conditional))
7119
7120 ((eq (char-before preceding-token-end) ?:)
7121 ;; Might be after another label, so check it recursively.
51c9af45
AM
7122 (save-restriction
7123 (save-excursion
7124 (goto-char (1- preceding-token-end))
7125 ;; Essentially the same as the
7126 ;; `c-syntactic-re-search-forward' regexp below.
7127 (setq macro-start
7128 (save-excursion (and (c-beginning-of-macro)
7129 (point))))
7130 (if macro-start (narrow-to-region macro-start (point-max)))
7131 (c-syntactic-skip-backward "^-]:?;}=*/%&|,<>!@+" nil t)
7132 ;; Note: the following should work instead of the
7133 ;; narrow-to-region above. Investigate why not,
7134 ;; sometime. ACM, 2006-03-31.
7135 ;; (c-syntactic-skip-backward "^-]:?;}=*/%&|,<>!@+"
7136 ;; macro-start t)
7137 (let ((pte (point))
7138 ;; If the caller turned on recording for us,
7139 ;; it shouldn't apply when we check the
7140 ;; preceding label.
7141 c-record-type-identifiers)
7142 ;; A label can't start at a cpp directive. Check for
7143 ;; this, since c-forward-syntactic-ws would foul up on it.
7144 (unless (and c-opt-cpp-prefix (looking-at c-opt-cpp-prefix))
7145 (c-forward-syntactic-ws)
7146 (c-forward-label nil pte start))))))))))
7147
1379f2c5 7148 ;; Point is still at the beginning of the possible label construct.
b414f371 7149 ;;
51c9af45
AM
7150 ;; Check that the next nonsymbol token is ":", or that we're in one
7151 ;; of QT's "slots" declarations. Allow '(' for the sake of macro
7152 ;; arguments. FIXME: Should build this regexp from the language
7153 ;; constants.
1379f2c5
AM
7154 (cond
7155 ;; public: protected: private:
7156 ((and
7157 (c-major-mode-is 'c++-mode)
7158 (search-forward-regexp
7159 "\\=p\\(r\\(ivate\\|otected\\)\\|ublic\\)\\>[^_]" nil t)
7160 (progn (backward-char)
7161 (c-forward-syntactic-ws limit)
7162 (looking-at ":\\([^:]\\|\\'\\)"))) ; A single colon.
7163 (forward-char)
7164 (setq label-type t))
7165 ;; QT double keyword like "protected slots:" or goto target.
7166 ((progn (goto-char start) nil))
7167 ((when (c-syntactic-re-search-forward
7168 "[ \t\n[:?;{=*/%&|,<>!@+-]" limit t t) ; not at EOB
7169 (backward-char)
7170 (setq label-end (point))
7171 (setq qt-symbol-idx
7172 (and (c-major-mode-is 'c++-mode)
7173 (string-match
7174 "\\(p\\(r\\(ivate\\|otected\\)\\|ublic\\)\\|more\\)\\>"
b414f371 7175 (buffer-substring start (point)))))
1379f2c5
AM
7176 (c-forward-syntactic-ws limit)
7177 (cond
7178 ((looking-at ":\\([^:]\\|\\'\\)") ; A single colon.
7179 (forward-char)
7180 (setq label-type
f412a567
AM
7181 (if (or (string= "signals" ; Special QT macro
7182 (setq kwd (buffer-substring-no-properties start label-end)))
7183 (string= "Q_SIGNALS" kwd))
1379f2c5
AM
7184 'qt-1kwd-colon
7185 'goto-target)))
7186 ((and qt-symbol-idx
f412a567 7187 (search-forward-regexp "\\=\\(slots\\|Q_SLOTS\\)\\>" limit t)
1379f2c5
AM
7188 (progn (c-forward-syntactic-ws limit)
7189 (looking-at ":\\([^:]\\|\\'\\)"))) ; A single colon
7190 (forward-char)
7191 (setq label-type 'qt-2kwds-colon)))))))
0386b551
AM
7192
7193 (save-restriction
7194 (narrow-to-region start (point))
7195
7196 ;; Check that `c-nonlabel-token-key' doesn't match anywhere.
7197 (catch 'check-label
7198 (goto-char start)
7199 (while (progn
7200 (when (looking-at c-nonlabel-token-key)
7201 (goto-char start)
1379f2c5 7202 (setq label-type nil)
0386b551
AM
7203 (throw 'check-label nil))
7204 (and (c-safe (c-forward-sexp)
7205 (c-forward-syntactic-ws)
7206 t)
7207 (not (eobp)))))
7208
7209 ;; Record the identifiers in the label for fontification, unless
7210 ;; it begins with `c-label-kwds' in which case the following
7211 ;; identifiers are part of a (constant) expression that
7212 ;; shouldn't be fontified.
7213 (when (and c-record-type-identifiers
7214 (progn (goto-char start)
7215 (not (looking-at c-label-kwds-regexp))))
7216 (while (c-syntactic-re-search-forward c-symbol-key nil t)
7217 (c-record-ref-id (cons (match-beginning 0)
7218 (match-end 0)))))
7219
7220 (c-put-c-type-property (1- (point-max)) 'c-decl-end)
82ba65cf 7221 (goto-char (point-max)))))
0386b551
AM
7222
7223 (t
7224 ;; Not a label.
1379f2c5
AM
7225 (goto-char start)))
7226 label-type))
0386b551
AM
7227
7228(defun c-forward-objc-directive ()
7229 ;; Assuming the point is at the beginning of a token, try to move
7230 ;; forward to the end of the Objective-C directive that starts
7231 ;; there. Return t if a directive was fully recognized, otherwise
7232 ;; the point is moved as far as one could be successfully parsed and
7233 ;; nil is returned.
7234 ;;
7235 ;; This function records identifier ranges on
7236 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
7237 ;; `c-record-type-identifiers' is non-nil.
7238 ;;
7239 ;; This function might do hidden buffer changes.
7240
7241 (let ((start (point))
7242 start-char
7243 (c-promote-possible-types t)
7244 ;; Turn off recognition of angle bracket arglists while parsing
7245 ;; types here since the protocol reference list might then be
7246 ;; considered part of the preceding name or superclass-name.
7247 c-recognize-<>-arglists)
7248
7249 (if (or
7250 (when (looking-at
7251 (eval-when-compile
7252 (c-make-keywords-re t
7253 (append (c-lang-const c-protection-kwds objc)
7254 '("@end"))
7255 'objc-mode)))
7256 (goto-char (match-end 1))
7257 t)
7258
7259 (and
7260 (looking-at
7261 (eval-when-compile
7262 (c-make-keywords-re t
7263 '("@interface" "@implementation" "@protocol")
7264 'objc-mode)))
7265
7266 ;; Handle the name of the class itself.
7267 (progn
cb694ab7
AM
7268; (c-forward-token-2) ; 2006/1/13 This doesn't move if the token's
7269; at EOB.
7270 (goto-char (match-end 0))
7271 (c-skip-ws-forward)
0386b551
AM
7272 (c-forward-type))
7273
7274 (catch 'break
7275 ;; Look for ": superclass-name" or "( category-name )".
7276 (when (looking-at "[:\(]")
7277 (setq start-char (char-after))
7278 (forward-char)
7279 (c-forward-syntactic-ws)
7280 (unless (c-forward-type) (throw 'break nil))
7281 (when (eq start-char ?\()
7282 (unless (eq (char-after) ?\)) (throw 'break nil))
7283 (forward-char)
7284 (c-forward-syntactic-ws)))
7285
7286 ;; Look for a protocol reference list.
7287 (if (eq (char-after) ?<)
7288 (let ((c-recognize-<>-arglists t)
7289 (c-parse-and-markup-<>-arglists t)
7290 c-restricted-<>-arglists)
7291 (c-forward-<>-arglist t))
7292 t))))
7293
7294 (progn
7295 (c-backward-syntactic-ws)
7296 (c-clear-c-type-property start (1- (point)) 'c-decl-end)
7297 (c-put-c-type-property (1- (point)) 'c-decl-end)
7298 t)
7299
7300 (c-clear-c-type-property start (point) 'c-decl-end)
7301 nil)))
7302
785eecbb
RS
7303(defun c-beginning-of-inheritance-list (&optional lim)
7304 ;; Go to the first non-whitespace after the colon that starts a
7305 ;; multiple inheritance introduction. Optional LIM is the farthest
7306 ;; back we should search.
0386b551
AM
7307 ;;
7308 ;; This function might do hidden buffer changes.
7309 (c-with-syntax-table c++-template-syntax-table
7310 (c-backward-token-2 0 t lim)
7311 (while (and (or (looking-at c-symbol-start)
7312 (looking-at "[<,]\\|::"))
7313 (zerop (c-backward-token-2 1 t lim))))))
785eecbb 7314
785eecbb
RS
7315(defun c-in-method-def-p ()
7316 ;; Return nil if we aren't in a method definition, otherwise the
7317 ;; position of the initial [+-].
0386b551
AM
7318 ;;
7319 ;; This function might do hidden buffer changes.
785eecbb
RS
7320 (save-excursion
7321 (beginning-of-line)
a66cd3ee
MS
7322 (and c-opt-method-key
7323 (looking-at c-opt-method-key)
785eecbb
RS
7324 (point))
7325 ))
7326
a66cd3ee
MS
7327;; Contributed by Kevin Ryde <user42@zip.com.au>.
7328(defun c-in-gcc-asm-p ()
7329 ;; Return non-nil if point is within a gcc \"asm\" block.
7330 ;;
7331 ;; This should be called with point inside an argument list.
7332 ;;
7333 ;; Only one level of enclosing parentheses is considered, so for
7334 ;; instance `nil' is returned when in a function call within an asm
7335 ;; operand.
0386b551
AM
7336 ;;
7337 ;; This function might do hidden buffer changes.
a66cd3ee
MS
7338
7339 (and c-opt-asm-stmt-key
7340 (save-excursion
7341 (beginning-of-line)
7342 (backward-up-list 1)
7343 (c-beginning-of-statement-1 (point-min) nil t)
7344 (looking-at c-opt-asm-stmt-key))))
7345
abb7e5cf 7346(defun c-at-toplevel-p ()
a85fd6da
AM
7347 "Return a determination as to whether point is \"at the top level\".
7348Informally, \"at the top level\" is anywhere where you can write
7349a function.
7350
7351More precisely, being at the top-level means that point is either
7352outside any enclosing block (such as a function definition), or
7353directly inside a class, namespace or other block that contains
7354another declaration level.
abb7e5cf
SM
7355
7356If point is not at the top-level (e.g. it is inside a method
7357definition), then nil is returned. Otherwise, if point is at a
7358top-level not enclosed within a class definition, t is returned.
7359Otherwise, a 2-vector is returned where the zeroth element is the
7360buffer position of the start of the class declaration, and the first
7361element is the buffer position of the enclosing class's opening
0386b551
AM
7362brace.
7363
7364Note that this function might do hidden buffer changes. See the
7365comment at the start of cc-engine.el for more info."
a66cd3ee
MS
7366 (let ((paren-state (c-parse-state)))
7367 (or (not (c-most-enclosing-brace paren-state))
7368 (c-search-uplist-for-classkey paren-state))))
7369
d9e94c22 7370(defun c-just-after-func-arglist-p (&optional lim)
0386b551
AM
7371 ;; Return non-nil if the point is in the region after the argument
7372 ;; list of a function and its opening brace (or semicolon in case it
7373 ;; got no body). If there are K&R style argument declarations in
7374 ;; that region, the point has to be inside the first one for this
7375 ;; function to recognize it.
a66cd3ee 7376 ;;
0386b551
AM
7377 ;; If successful, the point is moved to the first token after the
7378 ;; function header (see `c-forward-decl-or-cast-1' for details) and
7379 ;; the position of the opening paren of the function arglist is
7380 ;; returned.
7381 ;;
7382 ;; The point is clobbered if not successful.
7383 ;;
7384 ;; LIM is used as bound for backward buffer searches.
7385 ;;
7386 ;; This function might do hidden buffer changes.
7387
7388 (let ((beg (point)) end id-start)
7389 (and
7390 (eq (c-beginning-of-statement-1 lim) 'same)
7391
7392 (not (or (c-major-mode-is 'objc-mode)
7393 (c-forward-objc-directive)))
7394
7395 (setq id-start
7396 (car-safe (c-forward-decl-or-cast-1 (c-point 'bosws) nil nil)))
7397 (< id-start beg)
7398
7399 ;; There should not be a '=' or ',' between beg and the
7400 ;; start of the declaration since that means we were in the
7401 ;; "expression part" of the declaration.
7402 (or (> (point) beg)
7403 (not (looking-at "[=,]")))
7404
7405 (save-excursion
7406 ;; Check that there's an arglist paren in the
7407 ;; declaration.
7408 (goto-char id-start)
7409 (cond ((eq (char-after) ?\()
7410 ;; The declarator is a paren expression, so skip past it
7411 ;; so that we don't get stuck on that instead of the
7412 ;; function arglist.
7413 (c-forward-sexp))
51c9af45
AM
7414 ((and c-opt-op-identifier-prefix
7415 (looking-at c-opt-op-identifier-prefix))
0386b551
AM
7416 ;; Don't trip up on "operator ()".
7417 (c-forward-token-2 2 t)))
7418 (and (< (point) beg)
7419 (c-syntactic-re-search-forward "(" beg t t)
7420 (1- (point)))))))
785eecbb 7421
a66cd3ee
MS
7422(defun c-in-knr-argdecl (&optional lim)
7423 ;; Return the position of the first argument declaration if point is
7424 ;; inside a K&R style argument declaration list, nil otherwise.
7425 ;; `c-recognize-knr-p' is not checked. If LIM is non-nil, it's a
7426 ;; position that bounds the backward search for the argument list.
7427 ;;
9cf17ef1
AM
7428 ;; Point must be within a possible K&R region, e.g. just before a top-level
7429 ;; "{". It must be outside of parens and brackets. The test can return
7430 ;; false positives otherwise.
0386b551
AM
7431 ;;
7432 ;; This function might do hidden buffer changes.
d9e94c22 7433
a66cd3ee
MS
7434 (save-excursion
7435 (save-restriction
9cf17ef1
AM
7436 ;; If we're in a macro, our search range is restricted to it. Narrow to
7437 ;; the searchable range.
7438 (let* ((macro-start (c-query-macro-start))
7439 (lim (max (or lim (point-min)) (or macro-start (point-min))))
6b6481ed
AM
7440 before-lparen after-rparen
7441 (pp-count-out 20)) ; Max number of paren/brace constructs before we give up
9cf17ef1
AM
7442 (narrow-to-region lim (c-point 'eol))
7443
7444 ;; Search backwards for the defun's argument list. We give up if we
7445 ;; encounter a "}" (end of a previous defun) or BOB.
7446 ;;
7447 ;; The criterion for a paren structure being the arg list is:
7448 ;; o - there is non-WS stuff after it but before any "{"; AND
7449 ;; o - the token after it isn't a ";" AND
7450 ;; o - it is preceded by either an identifier (the function name) or
7451 ;; a macro expansion like "DEFUN (...)"; AND
7452 ;; o - its content is a non-empty comma-separated list of identifiers
7453 ;; (an empty arg list won't have a knr region).
7454 ;;
7455 ;; The following snippet illustrates these rules:
7456 ;; int foo (bar, baz, yuk)
7457 ;; int bar [] ;
7458 ;; int (*baz) (my_type) ;
7459 ;; int (*) (void) (*yuk) (void) ;
7460 ;; {
7461
7462 (catch 'knr
6b6481ed
AM
7463 (while (> pp-count-out 0) ; go back one paren/bracket pair each time.
7464 (setq pp-count-out (1- pp-count-out))
10489fcb 7465 (c-syntactic-skip-backward "^)]}")
9cf17ef1
AM
7466 (cond ((eq (char-before) ?\))
7467 (setq after-rparen (point)))
10489fcb
AM
7468 ((eq (char-before) ?\])
7469 (setq after-rparen nil))
7470 (t ; either } (hit previous defun) or no more parens/brackets
7471 (throw 'knr nil)))
9cf17ef1
AM
7472
7473 (if after-rparen
7474 ;; We're inside a paren. Could it be our argument list....?
7475 (if
7476 (and
7477 (progn
7478 (goto-char after-rparen)
7479 (unless (c-go-list-backward) (throw 'knr nil)) ;
7480 ;; FIXME!!! What about macros between the parens? 2007/01/20
7481 (setq before-lparen (point)))
d9e94c22 7482
9cf17ef1
AM
7483 ;; It can't be the arg list if next token is ; or {
7484 (progn (goto-char after-rparen)
7485 (c-forward-syntactic-ws)
7486 (not (memq (char-after) '(?\; ?\{))))
d9e94c22 7487
9cf17ef1
AM
7488 ;; Is the thing preceding the list an identifier (the
7489 ;; function name), or a macro expansion?
7490 (progn
7491 (goto-char before-lparen)
7492 (eq (c-backward-token-2) 0)
7493 (or (c-on-identifier)
7494 (and (eq (char-after) ?\))
7495 (c-go-up-list-backward)
7496 (eq (c-backward-token-2) 0)
7497 (c-on-identifier))))
7498
7499 ;; Have we got a non-empty list of comma-separated
7500 ;; identifiers?
7501 (progn
7502 (goto-char before-lparen)
7503 (c-forward-token-2) ; to first token inside parens
7504 (and
7505 (c-on-identifier)
7506 (c-forward-token-2)
7507 (catch 'id-list
7508 (while (eq (char-after) ?\,)
7509 (c-forward-token-2)
7510 (unless (c-on-identifier) (throw 'id-list nil))
7511 (c-forward-token-2))
7512 (eq (char-after) ?\))))))
7513
7514 ;; ...Yes. We've identified the function's argument list.
7515 (throw 'knr
7516 (progn (goto-char after-rparen)
7517 (c-forward-syntactic-ws)
7518 (point)))
7519
7520 ;; ...No. The current parens aren't the function's arg list.
7521 (goto-char before-lparen))
7522
7523 (or (c-go-list-backward) ; backwards over [ .... ]
7524 (throw 'knr nil)))))))))
785eecbb
RS
7525
7526(defun c-skip-conditional ()
7527 ;; skip forward over conditional at point, including any predicate
7528 ;; statements in parentheses. No error checking is performed.
0386b551
AM
7529 ;;
7530 ;; This function might do hidden buffer changes.
0ec8351b
BW
7531 (c-forward-sexp (cond
7532 ;; else if()
a66cd3ee
MS
7533 ((looking-at (concat "\\<else"
7534 "\\([ \t\n]\\|\\\\\n\\)+"
7535 "if\\>\\([^_]\\|$\\)"))
7536 3)
0ec8351b 7537 ;; do, else, try, finally
a66cd3ee
MS
7538 ((looking-at (concat "\\<\\("
7539 "do\\|else\\|try\\|finally"
7540 "\\)\\>\\([^_]\\|$\\)"))
130c507e 7541 1)
ce8c7486 7542 ;; for, if, while, switch, catch, synchronized, foreach
0ec8351b 7543 (t 2))))
785eecbb 7544
a66cd3ee
MS
7545(defun c-after-conditional (&optional lim)
7546 ;; If looking at the token after a conditional then return the
7547 ;; position of its start, otherwise return nil.
0386b551
AM
7548 ;;
7549 ;; This function might do hidden buffer changes.
a66cd3ee 7550 (save-excursion
d9e94c22 7551 (and (zerop (c-backward-token-2 1 t lim))
a66cd3ee
MS
7552 (or (looking-at c-block-stmt-1-key)
7553 (and (eq (char-after) ?\()
d9e94c22 7554 (zerop (c-backward-token-2 1 t lim))
a66cd3ee
MS
7555 (looking-at c-block-stmt-2-key)))
7556 (point))))
7557
0386b551
AM
7558(defun c-after-special-operator-id (&optional lim)
7559 ;; If the point is after an operator identifier that isn't handled
7560 ;; like an ordinary symbol (i.e. like "operator =" in C++) then the
7561 ;; position of the start of that identifier is returned. nil is
7562 ;; returned otherwise. The point may be anywhere in the syntactic
7563 ;; whitespace after the last token of the operator identifier.
7564 ;;
7565 ;; This function might do hidden buffer changes.
7566 (save-excursion
7567 (and c-overloadable-operators-regexp
7568 (zerop (c-backward-token-2 1 nil lim))
7569 (looking-at c-overloadable-operators-regexp)
51c9af45 7570 (or (not c-opt-op-identifier-prefix)
0386b551
AM
7571 (and
7572 (zerop (c-backward-token-2 1 nil lim))
51c9af45 7573 (looking-at c-opt-op-identifier-prefix)))
0386b551
AM
7574 (point))))
7575
a66cd3ee
MS
7576(defsubst c-backward-to-block-anchor (&optional lim)
7577 ;; Assuming point is at a brace that opens a statement block of some
7578 ;; kind, move to the proper anchor point for that block. It might
7579 ;; need to be adjusted further by c-add-stmt-syntax, but the
7580 ;; position at return is suitable as start position for that
7581 ;; function.
0386b551
AM
7582 ;;
7583 ;; This function might do hidden buffer changes.
a66cd3ee
MS
7584 (unless (= (point) (c-point 'boi))
7585 (let ((start (c-after-conditional lim)))
7586 (if start
7587 (goto-char start)))))
7588
037558bf 7589(defsubst c-backward-to-decl-anchor (&optional lim)
a66cd3ee
MS
7590 ;; Assuming point is at a brace that opens the block of a top level
7591 ;; declaration of some kind, move to the proper anchor point for
7592 ;; that block.
0386b551
AM
7593 ;;
7594 ;; This function might do hidden buffer changes.
a66cd3ee 7595 (unless (= (point) (c-point 'boi))
037558bf 7596 (c-beginning-of-statement-1 lim)))
a66cd3ee 7597
ff959bab 7598(defun c-search-decl-header-end ()
a66cd3ee
MS
7599 ;; Search forward for the end of the "header" of the current
7600 ;; declaration. That's the position where the definition body
7601 ;; starts, or the first variable initializer, or the ending
7602 ;; semicolon. I.e. search forward for the closest following
7603 ;; (syntactically relevant) '{', '=' or ';' token. Point is left
7604 ;; _after_ the first found token, or at point-max if none is found.
0386b551
AM
7605 ;;
7606 ;; This function might do hidden buffer changes.
ff959bab
MS
7607
7608 (let ((base (point)))
7609 (if (c-major-mode-is 'c++-mode)
7610
7611 ;; In C++ we need to take special care to handle operator
7612 ;; tokens and those pesky template brackets.
7613 (while (and
7614 (c-syntactic-re-search-forward "[;{<=]" nil 'move t t)
7615 (or
7616 (c-end-of-current-token base)
7617 ;; Handle operator identifiers, i.e. ignore any
7618 ;; operator token preceded by "operator".
7619 (save-excursion
7620 (and (c-safe (c-backward-sexp) t)
51c9af45 7621 (looking-at c-opt-op-identifier-prefix)))
ff959bab
MS
7622 (and (eq (char-before) ?<)
7623 (c-with-syntax-table c++-template-syntax-table
7624 (if (c-safe (goto-char (c-up-list-forward (point))))
7625 t
7626 (goto-char (point-max))
7627 nil)))))
7628 (setq base (point)))
7629
7630 (while (and
7631 (c-syntactic-re-search-forward "[;{=]" nil 'move t t)
7632 (c-end-of-current-token base))
7633 (setq base (point))))))
a66cd3ee
MS
7634
7635(defun c-beginning-of-decl-1 (&optional lim)
7636 ;; Go to the beginning of the current declaration, or the beginning
7637 ;; of the previous one if already at the start of it. Point won't
0386b551 7638 ;; be moved out of any surrounding paren. Return a cons cell of the
a66cd3ee
MS
7639 ;; form (MOVE . KNR-POS). MOVE is like the return value from
7640 ;; `c-beginning-of-statement-1'. If point skipped over some K&R
7641 ;; style argument declarations (and they are to be recognized) then
7642 ;; KNR-POS is set to the start of the first such argument
7643 ;; declaration, otherwise KNR-POS is nil. If LIM is non-nil, it's a
7644 ;; position that bounds the backward search.
7645 ;;
7646 ;; NB: Cases where the declaration continues after the block, as in
7647 ;; "struct foo { ... } bar;", are currently recognized as two
7648 ;; declarations, e.g. "struct foo { ... }" and "bar;" in this case.
0386b551
AM
7649 ;;
7650 ;; This function might do hidden buffer changes.
a66cd3ee
MS
7651 (catch 'return
7652 (let* ((start (point))
d9e94c22 7653 (last-stmt-start (point))
0386b551 7654 (move (c-beginning-of-statement-1 lim nil t)))
a66cd3ee 7655
a66cd3ee
MS
7656 ;; `c-beginning-of-statement-1' stops at a block start, but we
7657 ;; want to continue if the block doesn't begin a top level
2a15eb73
MS
7658 ;; construct, i.e. if it isn't preceded by ';', '}', ':', bob,
7659 ;; or an open paren.
d9e94c22 7660 (let ((beg (point)) tentative-move)
51c9af45
AM
7661 ;; Go back one "statement" each time round the loop until we're just
7662 ;; after a ;, }, or :, or at BOB or the start of a macro or start of
7663 ;; an ObjC method. This will move over a multiple declaration whose
7664 ;; components are comma separated.
d9e94c22
MS
7665 (while (and
7666 ;; Must check with c-opt-method-key in ObjC mode.
7667 (not (and c-opt-method-key
7668 (looking-at c-opt-method-key)))
7669 (/= last-stmt-start (point))
7670 (progn
7671 (c-backward-syntactic-ws lim)
7672 (not (memq (char-before) '(?\; ?} ?: nil))))
2a15eb73
MS
7673 (save-excursion
7674 (backward-char)
7675 (not (looking-at "\\s(")))
d9e94c22
MS
7676 ;; Check that we don't move from the first thing in a
7677 ;; macro to its header.
7678 (not (eq (setq tentative-move
0386b551 7679 (c-beginning-of-statement-1 lim nil t))
d9e94c22
MS
7680 'macro)))
7681 (setq last-stmt-start beg
7682 beg (point)
7683 move tentative-move))
7684 (goto-char beg))
7685
7686 (when c-recognize-knr-p
7687 (let ((fallback-pos (point)) knr-argdecl-start)
7688 ;; Handle K&R argdecls. Back up after the "statement" jumped
7689 ;; over by `c-beginning-of-statement-1', unless it was the
7690 ;; function body, in which case we're sitting on the opening
7691 ;; brace now. Then test if we're in a K&R argdecl region and
7692 ;; that we started at the other side of the first argdecl in
7693 ;; it.
7694 (unless (eq (char-after) ?{)
7695 (goto-char last-stmt-start))
7696 (if (and (setq knr-argdecl-start (c-in-knr-argdecl lim))
7697 (< knr-argdecl-start start)
7698 (progn
7699 (goto-char knr-argdecl-start)
0386b551 7700 (not (eq (c-beginning-of-statement-1 lim nil t) 'macro))))
d9e94c22
MS
7701 (throw 'return
7702 (cons (if (eq (char-after fallback-pos) ?{)
7703 'previous
7704 'same)
7705 knr-argdecl-start))
7706 (goto-char fallback-pos))))
7707
51c9af45
AM
7708 ;; `c-beginning-of-statement-1' counts each brace block as a separate
7709 ;; statement, so the result will be 'previous if we've moved over any.
7710 ;; So change our result back to 'same if necessary.
7711 ;;
7712 ;; If they were brace list initializers we might not have moved over a
7713 ;; declaration boundary though, so change it to 'same if we've moved
7714 ;; past a '=' before '{', but not ';'. (This ought to be integrated
7715 ;; into `c-beginning-of-statement-1', so we avoid this extra pass which
7716 ;; potentially can search over a large amount of text.). Take special
7717 ;; pains not to get mislead by C++'s "operator=", and the like.
d9e94c22
MS
7718 (if (and (eq move 'previous)
7719 (c-with-syntax-table (if (c-major-mode-is 'c++-mode)
7720 c++-template-syntax-table
7721 (syntax-table))
7722 (save-excursion
51c9af45
AM
7723 (and
7724 (progn
7725 (while ; keep going back to "[;={"s until we either find
7726 ; no more, or get to one which isn't an "operator ="
7727 (and (c-syntactic-re-search-forward "[;={]" start t t t)
7728 (eq (char-before) ?=)
7729 c-overloadable-operators-regexp
7730 c-opt-op-identifier-prefix
7731 (save-excursion
7732 (eq (c-backward-token-2) 0)
7733 (looking-at c-overloadable-operators-regexp)
7734 (eq (c-backward-token-2) 0)
7735 (looking-at c-opt-op-identifier-prefix))))
7736 (eq (char-before) ?=))
7737 (c-syntactic-re-search-forward "[;{]" start t t)
7738 (eq (char-before) ?{)
7739 (c-safe (goto-char (c-up-list-forward (point))) t)
7740 (not (c-syntactic-re-search-forward ";" start t t))))))
d9e94c22
MS
7741 (cons 'same nil)
7742 (cons move nil)))))
a66cd3ee
MS
7743
7744(defun c-end-of-decl-1 ()
7745 ;; Assuming point is at the start of a declaration (as detected by
7746 ;; e.g. `c-beginning-of-decl-1'), go to the end of it. Unlike
7747 ;; `c-beginning-of-decl-1', this function handles the case when a
7748 ;; block is followed by identifiers in e.g. struct declarations in C
7749 ;; or C++. If a proper end was found then t is returned, otherwise
7750 ;; point is moved as far as possible within the current sexp and nil
7751 ;; is returned. This function doesn't handle macros; use
7752 ;; `c-end-of-macro' instead in those cases.
0386b551
AM
7753 ;;
7754 ;; This function might do hidden buffer changes.
ce8c7486 7755 (let ((start (point))
a66cd3ee
MS
7756 (decl-syntax-table (if (c-major-mode-is 'c++-mode)
7757 c++-template-syntax-table
7758 (syntax-table))))
7759 (catch 'return
7760 (c-search-decl-header-end)
7761
7762 (when (and c-recognize-knr-p
7763 (eq (char-before) ?\;)
7764 (c-in-knr-argdecl start))
7765 ;; Stopped at the ';' in a K&R argdecl section which is
7766 ;; detected using the same criteria as in
7767 ;; `c-beginning-of-decl-1'. Move to the following block
7768 ;; start.
d9e94c22 7769 (c-syntactic-re-search-forward "{" nil 'move t))
a66cd3ee
MS
7770
7771 (when (eq (char-before) ?{)
7772 ;; Encountered a block in the declaration. Jump over it.
7773 (condition-case nil
7774 (goto-char (c-up-list-forward (point)))
d9e94c22
MS
7775 (error (goto-char (point-max))
7776 (throw 'return nil)))
a66cd3ee
MS
7777 (if (or (not c-opt-block-decls-with-vars-key)
7778 (save-excursion
7779 (c-with-syntax-table decl-syntax-table
7780 (let ((lim (point)))
7781 (goto-char start)
b3cf7e18
MS
7782 (not (and
7783 ;; Check for `c-opt-block-decls-with-vars-key'
7784 ;; before the first paren.
7785 (c-syntactic-re-search-forward
d9e94c22 7786 (concat "[;=\(\[{]\\|\\("
b3cf7e18
MS
7787 c-opt-block-decls-with-vars-key
7788 "\\)")
d9e94c22 7789 lim t t t)
b3cf7e18
MS
7790 (match-beginning 1)
7791 (not (eq (char-before) ?_))
d9e94c22
MS
7792 ;; Check that the first following paren is
7793 ;; the block.
7794 (c-syntactic-re-search-forward "[;=\(\[{]"
7795 lim t t t)
b3cf7e18 7796 (eq (char-before) ?{)))))))
a66cd3ee
MS
7797 ;; The declaration doesn't have any of the
7798 ;; `c-opt-block-decls-with-vars' keywords in the
7799 ;; beginning, so it ends here at the end of the block.
7800 (throw 'return t)))
7801
7802 (c-with-syntax-table decl-syntax-table
7803 (while (progn
7804 (if (eq (char-before) ?\;)
7805 (throw 'return t))
d9e94c22 7806 (c-syntactic-re-search-forward ";" nil 'move t))))
a66cd3ee 7807 nil)))
ce8c7486 7808
0386b551
AM
7809(defun c-looking-at-decl-block (containing-sexp goto-start &optional limit)
7810 ;; Assuming the point is at an open brace, check if it starts a
7811 ;; block that contains another declaration level, i.e. that isn't a
7812 ;; statement block or a brace list, and if so return non-nil.
7813 ;;
7814 ;; If the check is successful, the return value is the start of the
7815 ;; keyword that tells what kind of construct it is, i.e. typically
7816 ;; what `c-decl-block-key' matched. Also, if GOTO-START is set then
7817 ;; the point will be at the start of the construct, before any
7818 ;; leading specifiers, otherwise it's at the returned position.
7819 ;;
7820 ;; The point is clobbered if the check is unsuccessful.
7821 ;;
7822 ;; CONTAINING-SEXP is the position of the open of the surrounding
7823 ;; paren, or nil if none.
7824 ;;
7825 ;; The optional LIMIT limits the backward search for the start of
7826 ;; the construct. It's assumed to be at a syntactically relevant
7827 ;; position.
7828 ;;
7829 ;; If any template arglists are found in the searched region before
7830 ;; the open brace, they get marked with paren syntax.
7831 ;;
7832 ;; This function might do hidden buffer changes.
7833
7834 (let ((open-brace (point)) kwd-start first-specifier-pos)
7835 (c-syntactic-skip-backward c-block-prefix-charset limit t)
7836
7837 (when (and c-recognize-<>-arglists
7838 (eq (char-before) ?>))
7839 ;; Could be at the end of a template arglist.
7840 (let ((c-parse-and-markup-<>-arglists t)
7841 (c-disallow-comma-in-<>-arglists
7842 (and containing-sexp
7843 (not (eq (char-after containing-sexp) ?{)))))
7844 (while (and
7845 (c-backward-<>-arglist nil limit)
7846 (progn
7847 (c-syntactic-skip-backward c-block-prefix-charset limit t)
7848 (eq (char-before) ?>))))))
7849
7850 ;; Note: Can't get bogus hits inside template arglists below since they
7851 ;; have gotten paren syntax above.
7852 (when (and
7853 ;; If `goto-start' is set we begin by searching for the
7854 ;; first possible position of a leading specifier list.
7855 ;; The `c-decl-block-key' search continues from there since
7856 ;; we know it can't match earlier.
7857 (if goto-start
7858 (when (c-syntactic-re-search-forward c-symbol-start
7859 open-brace t t)
7860 (goto-char (setq first-specifier-pos (match-beginning 0)))
7861 t)
7862 t)
7863
7864 (cond
7865 ((c-syntactic-re-search-forward c-decl-block-key open-brace t t t)
7866 (goto-char (setq kwd-start (match-beginning 0)))
7867 (or
7868
7869 ;; Found a keyword that can't be a type?
7870 (match-beginning 1)
7871
7872 ;; Can be a type too, in which case it's the return type of a
7873 ;; function (under the assumption that no declaration level
7874 ;; block construct starts with a type).
7875 (not (c-forward-type))
7876
7877 ;; Jumped over a type, but it could be a declaration keyword
7878 ;; followed by the declared identifier that we've jumped over
7879 ;; instead (e.g. in "class Foo {"). If it indeed is a type
7880 ;; then we should be at the declarator now, so check for a
7881 ;; valid declarator start.
7882 ;;
7883 ;; Note: This doesn't cope with the case when a declared
7884 ;; identifier is followed by e.g. '(' in a language where '('
7885 ;; also might be part of a declarator expression. Currently
7886 ;; there's no such language.
7887 (not (or (looking-at c-symbol-start)
7888 (looking-at c-type-decl-prefix-key)))))
7889
7890 ;; In Pike a list of modifiers may be followed by a brace
7891 ;; to make them apply to many identifiers. Note that the
7892 ;; match data will be empty on return in this case.
7893 ((and (c-major-mode-is 'pike-mode)
7894 (progn
7895 (goto-char open-brace)
7896 (= (c-backward-token-2) 0))
7897 (looking-at c-specifier-key)
7898 ;; Use this variant to avoid yet another special regexp.
7899 (c-keyword-member (c-keyword-sym (match-string 1))
7900 'c-modifier-kwds))
7901 (setq kwd-start (point))
7902 t)))
7903
7904 ;; Got a match.
7905
7906 (if goto-start
7907 ;; Back up over any preceding specifiers and their clauses
7908 ;; by going forward from `first-specifier-pos', which is the
7909 ;; earliest possible position where the specifier list can
7910 ;; start.
7911 (progn
7912 (goto-char first-specifier-pos)
7913
7914 (while (< (point) kwd-start)
7915 (if (looking-at c-symbol-key)
7916 ;; Accept any plain symbol token on the ground that
7917 ;; it's a specifier masked through a macro (just
7918 ;; like `c-forward-decl-or-cast-1' skip forward over
7919 ;; such tokens).
7920 ;;
7921 ;; Could be more restrictive wrt invalid keywords,
7922 ;; but that'd only occur in invalid code so there's
7923 ;; no use spending effort on it.
7924 (let ((end (match-end 0)))
7925 (unless (c-forward-keyword-clause 0)
7926 (goto-char end)
7927 (c-forward-syntactic-ws)))
7928
7929 ;; Can't parse a declaration preamble and is still
7930 ;; before `kwd-start'. That means `first-specifier-pos'
7931 ;; was in some earlier construct. Search again.
7932 (if (c-syntactic-re-search-forward c-symbol-start
7933 kwd-start 'move t)
7934 (goto-char (setq first-specifier-pos (match-beginning 0)))
7935 ;; Got no preamble before the block declaration keyword.
7936 (setq first-specifier-pos kwd-start))))
7937
7938 (goto-char first-specifier-pos))
7939 (goto-char kwd-start))
7940
7941 kwd-start)))
ce8c7486 7942
a66cd3ee 7943(defun c-search-uplist-for-classkey (paren-state)
0386b551
AM
7944 ;; Check if the closest containing paren sexp is a declaration
7945 ;; block, returning a 2 element vector in that case. Aref 0
7946 ;; contains the bufpos at boi of the class key line, and aref 1
7947 ;; contains the bufpos of the open brace. This function is an
7948 ;; obsolete wrapper for `c-looking-at-decl-block'.
7949 ;;
7950 ;; This function might do hidden buffer changes.
7951 (let ((open-paren-pos (c-most-enclosing-brace paren-state)))
7952 (when open-paren-pos
7953 (save-excursion
7954 (goto-char open-paren-pos)
7955 (when (and (eq (char-after) ?{)
7956 (c-looking-at-decl-block
7957 (c-safe-position open-paren-pos paren-state)
7958 nil))
7959 (back-to-indentation)
7960 (vector (point) open-paren-pos))))))
785eecbb 7961
a66cd3ee 7962(defun c-inside-bracelist-p (containing-sexp paren-state)
785eecbb
RS
7963 ;; return the buffer position of the beginning of the brace list
7964 ;; statement if we're inside a brace list, otherwise return nil.
7965 ;; CONTAINING-SEXP is the buffer pos of the innermost containing
0386b551 7966 ;; paren. PAREN-STATE is the remainder of the state of enclosing
130c507e 7967 ;; braces
785eecbb
RS
7968 ;;
7969 ;; N.B.: This algorithm can potentially get confused by cpp macros
0386b551 7970 ;; placed in inconvenient locations. It's a trade-off we make for
785eecbb 7971 ;; speed.
0386b551
AM
7972 ;;
7973 ;; This function might do hidden buffer changes.
785eecbb 7974 (or
d9e94c22 7975 ;; This will pick up brace list declarations.
b2acd789
RS
7976 (c-safe
7977 (save-excursion
7978 (goto-char containing-sexp)
0ec8351b 7979 (c-forward-sexp -1)
b2acd789 7980 (let (bracepos)
d9e94c22 7981 (if (and (or (looking-at c-brace-list-key)
0ec8351b 7982 (progn (c-forward-sexp -1)
d9e94c22 7983 (looking-at c-brace-list-key)))
a66cd3ee 7984 (setq bracepos (c-down-list-forward (point)))
b2acd789
RS
7985 (not (c-crosses-statement-barrier-p (point)
7986 (- bracepos 2))))
7987 (point)))))
785eecbb
RS
7988 ;; this will pick up array/aggregate init lists, even if they are nested.
7989 (save-excursion
0ec8351b
BW
7990 (let ((class-key
7991 ;; Pike can have class definitions anywhere, so we must
7992 ;; check for the class key here.
7993 (and (c-major-mode-is 'pike-mode)
a66cd3ee
MS
7994 c-decl-block-key))
7995 bufpos braceassignp lim next-containing)
785eecbb
RS
7996 (while (and (not bufpos)
7997 containing-sexp)
a66cd3ee
MS
7998 (when paren-state
7999 (if (consp (car paren-state))
8000 (setq lim (cdr (car paren-state))
8001 paren-state (cdr paren-state))
8002 (setq lim (car paren-state)))
8003 (when paren-state
8004 (setq next-containing (car paren-state)
8005 paren-state (cdr paren-state))))
785eecbb 8006 (goto-char containing-sexp)
a66cd3ee
MS
8007 (if (c-looking-at-inexpr-block next-containing next-containing)
8008 ;; We're in an in-expression block of some kind. Do not
8009 ;; check nesting. We deliberately set the limit to the
8010 ;; containing sexp, so that c-looking-at-inexpr-block
8011 ;; doesn't check for an identifier before it.
0ec8351b
BW
8012 (setq containing-sexp nil)
8013 ;; see if the open brace is preceded by = or [...] in
8014 ;; this statement, but watch out for operator=
a66cd3ee 8015 (setq braceassignp 'dontknow)
d9e94c22 8016 (c-backward-token-2 1 t lim)
6393fef2 8017 ;; Checks to do only on the first sexp before the brace.
d9e94c22 8018 (when (and c-opt-inexpr-brace-list-key
6393fef2
RS
8019 (eq (char-after) ?\[))
8020 ;; In Java, an initialization brace list may follow
8021 ;; directly after "new Foo[]", so check for a "new"
8022 ;; earlier.
8023 (while (eq braceassignp 'dontknow)
8024 (setq braceassignp
d9e94c22
MS
8025 (cond ((/= (c-backward-token-2 1 t lim) 0) nil)
8026 ((looking-at c-opt-inexpr-brace-list-key) t)
6393fef2
RS
8027 ((looking-at "\\sw\\|\\s_\\|[.[]")
8028 ;; Carry on looking if this is an
8029 ;; identifier (may contain "." in Java)
8030 ;; or another "[]" sexp.
8031 'dontknow)
8032 (t nil)))))
8033 ;; Checks to do on all sexps before the brace, up to the
8034 ;; beginning of the statement.
8035 (while (eq braceassignp 'dontknow)
0ec8351b
BW
8036 (cond ((eq (char-after) ?\;)
8037 (setq braceassignp nil))
8038 ((and class-key
8039 (looking-at class-key))
8040 (setq braceassignp nil))
8041 ((eq (char-after) ?=)
8042 ;; We've seen a =, but must check earlier tokens so
8043 ;; that it isn't something that should be ignored.
8044 (setq braceassignp 'maybe)
8045 (while (and (eq braceassignp 'maybe)
d9e94c22 8046 (zerop (c-backward-token-2 1 t lim)))
0ec8351b
BW
8047 (setq braceassignp
8048 (cond
8049 ;; Check for operator =
51c9af45
AM
8050 ((and c-opt-op-identifier-prefix
8051 (looking-at c-opt-op-identifier-prefix))
0386b551 8052 nil)
130c507e
GM
8053 ;; Check for `<opchar>= in Pike.
8054 ((and (c-major-mode-is 'pike-mode)
8055 (or (eq (char-after) ?`)
8056 ;; Special case for Pikes
8057 ;; `[]=, since '[' is not in
8058 ;; the punctuation class.
8059 (and (eq (char-after) ?\[)
8060 (eq (char-before) ?`))))
8061 nil)
0ec8351b
BW
8062 ((looking-at "\\s.") 'maybe)
8063 ;; make sure we're not in a C++ template
8064 ;; argument assignment
a66cd3ee
MS
8065 ((and
8066 (c-major-mode-is 'c++-mode)
8067 (save-excursion
8068 (let ((here (point))
8069 (pos< (progn
8070 (skip-chars-backward "^<>")
8071 (point))))
8072 (and (eq (char-before) ?<)
8073 (not (c-crosses-statement-barrier-p
8074 pos< here))
8075 (not (c-in-literal))
8076 ))))
0ec8351b 8077 nil)
6393fef2
RS
8078 (t t))))))
8079 (if (and (eq braceassignp 'dontknow)
d9e94c22 8080 (/= (c-backward-token-2 1 t lim) 0))
6393fef2
RS
8081 (setq braceassignp nil)))
8082 (if (not braceassignp)
0ec8351b
BW
8083 (if (eq (char-after) ?\;)
8084 ;; Brace lists can't contain a semicolon, so we're done.
8085 (setq containing-sexp nil)
a66cd3ee
MS
8086 ;; Go up one level.
8087 (setq containing-sexp next-containing
8088 lim nil
8089 next-containing nil))
0ec8351b
BW
8090 ;; we've hit the beginning of the aggregate list
8091 (c-beginning-of-statement-1
a66cd3ee 8092 (c-most-enclosing-brace paren-state))
0ec8351b 8093 (setq bufpos (point))))
a66cd3ee 8094 )
785eecbb
RS
8095 bufpos))
8096 ))
8097
0ec8351b
BW
8098(defun c-looking-at-special-brace-list (&optional lim)
8099