Merge from emacs-24; up to 2012-12-22T19:09:52Z!rgm@gnu.org
[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 18137.
CommitLineData
785eecbb
RS
1;;; cc-engine.el --- core syntax guessing engine for CC mode
2
ab422c4d 3;; Copyright (C) 1985, 1987, 1992-2013 Free Software Foundation, Inc.
785eecbb 4
e309f66c
AM
5;; Authors: 2001- Alan Mackenzie
6;; 1998- Martin Stjernholm
d9e94c22 7;; 1992-1999 Barry A. Warsaw
5858f68c
GM
8;; 1987 Dave Detlefs
9;; 1987 Stewart Clamen
785eecbb 10;; 1985 Richard M. Stallman
0ec8351b 11;; Maintainer: bug-cc-mode@gnu.org
785eecbb 12;; Created: 22-Apr-1997 (split from cc-mode.el)
bd78fa1d
CY
13;; Keywords: c languages
14;; Package: cc-mode
785eecbb
RS
15
16;; This file is part of GNU Emacs.
17
b1fc2b50 18;; GNU Emacs is free software: you can redistribute it and/or modify
785eecbb 19;; it under the terms of the GNU General Public License as published by
b1fc2b50
GM
20;; the Free Software Foundation, either version 3 of the License, or
21;; (at your option) any later version.
785eecbb
RS
22
23;; GNU Emacs is distributed in the hope that it will be useful,
24;; but WITHOUT ANY WARRANTY; without even the implied warranty of
25;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
26;; GNU General Public License for more details.
27
28;; You should have received a copy of the GNU General Public License
b1fc2b50 29;; along with GNU Emacs. If not, see <http://www.gnu.org/licenses/>.
785eecbb 30
3afbc435
PJ
31;;; Commentary:
32
a66cd3ee
MS
33;; The functions which have docstring documentation can be considered
34;; part of an API which other packages can use in CC Mode buffers.
35;; Otoh, undocumented functions and functions with the documentation
36;; in comments are considered purely internal and can change semantics
37;; or even disappear in the future.
38;;
39;; (This policy applies to CC Mode as a whole, not just this file. It
40;; probably also applies to many other Emacs packages, but here it's
41;; clearly spelled out.)
42
d9e94c22
MS
43;; Hidden buffer changes
44;;
45;; Various functions in CC Mode use text properties for caching and
46;; syntactic markup purposes, and those of them that might modify such
0386b551
AM
47;; properties but still don't modify the buffer in a visible way are
48;; said to do "hidden buffer changes". They should be used within
49;; `c-save-buffer-state' or a similar function that saves and restores
50;; buffer modifiedness, disables buffer change hooks, etc.
d9e94c22 51;;
0386b551
AM
52;; Interactive functions are assumed to not do hidden buffer changes,
53;; except in the specific parts of them that do real changes.
d9e94c22 54;;
0386b551
AM
55;; Lineup functions are assumed to do hidden buffer changes. They
56;; must not do real changes, though.
d9e94c22 57;;
0386b551
AM
58;; All other functions that do hidden buffer changes have that noted
59;; in their doc string or comment.
60;;
61;; The intention with this system is to avoid wrapping every leaf
62;; function that do hidden buffer changes inside
63;; `c-save-buffer-state'. It should be used as near the top of the
64;; interactive functions as possible.
65;;
66;; Functions called during font locking are allowed to do hidden
67;; buffer changes since the font-lock package run them in a context
68;; similar to `c-save-buffer-state' (in fact, that function is heavily
69;; inspired by `save-buffer-state' in the font-lock package).
d9e94c22
MS
70
71;; Use of text properties
72;;
73;; CC Mode uses several text properties internally to mark up various
74;; positions, e.g. to improve speed and to eliminate glitches in
75;; interactive refontification.
76;;
7bfc3fdb
MS
77;; Note: This doc is for internal use only. Other packages should not
78;; assume that these text properties are used as described here.
79;;
0ec1d2c5
AM
80;; 'category
81;; Used for "indirection". With its help, some other property can
82;; be cheaply and easily switched on or off everywhere it occurs.
83;;
d9e94c22 84;; 'syntax-table
c8018ede
AM
85;; Used to modify the syntax of some characters. It is used to
86;; mark the "<" and ">" of angle bracket parens with paren syntax, and
87;; to "hide" obtrusive characters in preprocessor lines.
d9e94c22
MS
88;;
89;; This property is used on single characters and is therefore
90;; always treated as front and rear nonsticky (or start and end open
91;; in XEmacs vocabulary). It's therefore installed on
92;; `text-property-default-nonsticky' if that variable exists (Emacs
93;; >= 21).
94;;
95;; 'c-is-sws and 'c-in-sws
96;; Used by `c-forward-syntactic-ws' and `c-backward-syntactic-ws' to
97;; speed them up. See the comment blurb before `c-put-is-sws'
98;; below for further details.
99;;
100;; 'c-type
101;; This property is used on single characters to mark positions with
0386b551
AM
102;; special syntactic relevance of various sorts. Its primary use is
103;; to avoid glitches when multiline constructs are refontified
d9e94c22
MS
104;; interactively (on font lock decoration level 3). It's cleared in
105;; a region before it's fontified and is then put on relevant chars
106;; in that region as they are encountered during the fontification.
107;; The value specifies the kind of position:
108;;
109;; 'c-decl-arg-start
110;; Put on the last char of the token preceding each declaration
111;; inside a declaration style arglist (typically in a function
112;; prototype).
113;;
114;; 'c-decl-end
115;; Put on the last char of the token preceding a declaration.
116;; This is used in cases where declaration boundaries can't be
117;; recognized simply by looking for a token like ";" or "}".
118;; `c-type-decl-end-used' must be set if this is used (see also
119;; `c-find-decl-spots').
120;;
121;; 'c-<>-arg-sep
122;; Put on the commas that separate arguments in angle bracket
123;; arglists like C++ template arglists.
124;;
125;; 'c-decl-id-start and 'c-decl-type-start
126;; Put on the last char of the token preceding each declarator
127;; in the declarator list of a declaration. They are also used
128;; between the identifiers cases like enum declarations.
129;; 'c-decl-type-start is used when the declarators are types,
130;; 'c-decl-id-start otherwise.
131;;
132;; 'c-awk-NL-prop
133;; Used in AWK mode to mark the various kinds of newlines. See
134;; cc-awk.el.
135
3afbc435
PJ
136;;; Code:
137
0ec8351b 138(eval-when-compile
51f606de 139 (let ((load-path
130c507e
GM
140 (if (and (boundp 'byte-compile-dest-file)
141 (stringp byte-compile-dest-file))
142 (cons (file-name-directory byte-compile-dest-file) load-path)
51f606de 143 load-path)))
d9e94c22 144 (load "cc-bytecomp" nil t)))
130c507e
GM
145
146(cc-require 'cc-defs)
d9e94c22 147(cc-require-when-compile 'cc-langs)
130c507e 148(cc-require 'cc-vars)
d9e94c22 149
130c507e
GM
150;; Silence the compiler.
151(cc-bytecomp-defun buffer-syntactic-context) ; XEmacs
0ec8351b 152
51f606de 153\f
d9e94c22
MS
154;; Make declarations for all the `c-lang-defvar' variables in cc-langs.
155
156(defmacro c-declare-lang-variables ()
157 `(progn
485fe977
RS
158 ,@(apply 'nconc
159 (mapcar (lambda (init)
160 `(,(if (elt init 2)
161 `(defvar ,(car init) nil ,(elt init 2))
162 `(defvar ,(car init) nil))
163 (make-variable-buffer-local ',(car init))))
164 (cdr c-lang-variable-inits)))))
d9e94c22
MS
165(c-declare-lang-variables)
166
167\f
168;;; Internal state variables.
169
170;; Internal state of hungry delete key feature
171(defvar c-hungry-delete-key nil)
172(make-variable-buffer-local 'c-hungry-delete-key)
173
0386b551
AM
174;; The electric flag (toggled by `c-toggle-electric-state').
175;; If t, electric actions (like automatic reindentation, and (if
176;; c-auto-newline is also set) auto newlining) will happen when an electric
177;; key like `{' is pressed (or an electric keyword like `else').
178(defvar c-electric-flag t)
179(make-variable-buffer-local 'c-electric-flag)
180
d9e94c22
MS
181;; Internal state of auto newline feature.
182(defvar c-auto-newline nil)
183(make-variable-buffer-local 'c-auto-newline)
184
0386b551 185;; Included in the mode line to indicate the active submodes.
cb694ab7
AM
186;; (defvar c-submode-indicators nil)
187;; (make-variable-buffer-local 'c-submode-indicators)
d9e94c22 188
a66cd3ee
MS
189(defun c-calculate-state (arg prevstate)
190 ;; Calculate the new state of PREVSTATE, t or nil, based on arg. If
191 ;; arg is nil or zero, toggle the state. If arg is negative, turn
192 ;; the state off, and if arg is positive, turn the state on
193 (if (or (not arg)
194 (zerop (setq arg (prefix-numeric-value arg))))
195 (not prevstate)
196 (> arg 0)))
197
d9e94c22 198\f
037558bf
MS
199;; Basic handling of preprocessor directives.
200
201;; This is a dynamically bound cache used together with
202;; `c-query-macro-start' and `c-query-and-set-macro-start'. It only
203;; works as long as point doesn't cross a macro boundary.
204(defvar c-macro-start 'unknown)
205
206(defsubst c-query-and-set-macro-start ()
037558bf
MS
207 (if (symbolp c-macro-start)
208 (setq c-macro-start (save-excursion
0386b551
AM
209 (c-save-buffer-state ()
210 (and (c-beginning-of-macro)
211 (point)))))
037558bf
MS
212 c-macro-start))
213
214(defsubst c-query-macro-start ()
037558bf
MS
215 (if (symbolp c-macro-start)
216 (save-excursion
0386b551
AM
217 (c-save-buffer-state ()
218 (and (c-beginning-of-macro)
219 (point))))
037558bf
MS
220 c-macro-start))
221
56d093a9
AM
222;; One element macro cache to cope with continual movement within very large
223;; CPP macros.
224(defvar c-macro-cache nil)
225(make-variable-buffer-local 'c-macro-cache)
226;; Nil or cons of the bounds of the most recent CPP form probed by
227;; `c-beginning-of-macro', `c-end-of-macro' or `c-syntactic-end-of-macro'.
228;; The cdr will be nil if we know only the start of the CPP form.
229(defvar c-macro-cache-start-pos nil)
230(make-variable-buffer-local 'c-macro-cache-start-pos)
231;; The starting position from where we determined `c-macro-cache'.
232(defvar c-macro-cache-syntactic nil)
233(make-variable-buffer-local 'c-macro-cache-syntactic)
234;; non-nil iff `c-macro-cache' has both elements set AND the cdr is at a
235;; syntactic end of macro, not merely an apparent one.
236
237(defun c-invalidate-macro-cache (beg end)
238 ;; Called from a before-change function. If the change region is before or
f003f294 239 ;; in the macro characterized by `c-macro-cache' etc., nullify it
56d093a9
AM
240 ;; appropriately. BEG and END are the standard before-change-functions
241 ;; parameters. END isn't used.
242 (cond
243 ((null c-macro-cache))
244 ((< beg (car c-macro-cache))
245 (setq c-macro-cache nil
246 c-macro-cache-start-pos nil
247 c-macro-cache-syntactic nil))
248 ((and (cdr c-macro-cache)
249 (< beg (cdr c-macro-cache)))
250 (setcdr c-macro-cache nil)
251 (setq c-macro-cache-start-pos beg
252 c-macro-cache-syntactic nil))))
253
037558bf
MS
254(defun c-beginning-of-macro (&optional lim)
255 "Go to the beginning of a preprocessor directive.
256Leave point at the beginning of the directive and return t if in one,
257otherwise return nil and leave point unchanged.
258
0386b551
AM
259Note that this function might do hidden buffer changes. See the
260comment at the start of cc-engine.el for more info."
56d093a9
AM
261 (let ((here (point)))
262 (when c-opt-cpp-prefix
263 (if (and (car c-macro-cache)
264 (>= (point) (car c-macro-cache))
265 (or (and (cdr c-macro-cache)
266 (<= (point) (cdr c-macro-cache)))
267 (<= (point) c-macro-cache-start-pos)))
268 (unless (< (car c-macro-cache) (or lim (point-min)))
269 (progn (goto-char (max (or lim (point-min)) (car c-macro-cache)))
270 (setq c-macro-cache-start-pos
271 (max c-macro-cache-start-pos here))
272 t))
273 (setq c-macro-cache nil
274 c-macro-cache-start-pos nil
275 c-macro-cache-syntactic nil)
276
277 (save-restriction
278 (if lim (narrow-to-region lim (point-max)))
279 (beginning-of-line)
280 (while (eq (char-before (1- (point))) ?\\)
281 (forward-line -1))
282 (back-to-indentation)
283 (if (and (<= (point) here)
284 (looking-at c-opt-cpp-start))
285 (progn
286 (setq c-macro-cache (cons (point) nil)
287 c-macro-cache-start-pos here)
288 t)
289 (goto-char here)
290 nil))))))
037558bf
MS
291
292(defun c-end-of-macro ()
293 "Go to the end of a preprocessor directive.
0386b551
AM
294More accurately, move the point to the end of the closest following
295line that doesn't end with a line continuation backslash - no check is
296done that the point is inside a cpp directive to begin with.
037558bf 297
0386b551
AM
298Note that this function might do hidden buffer changes. See the
299comment at the start of cc-engine.el for more info."
56d093a9
AM
300 (if (and (cdr c-macro-cache)
301 (<= (point) (cdr c-macro-cache))
302 (>= (point) (car c-macro-cache)))
303 (goto-char (cdr c-macro-cache))
304 (unless (and (car c-macro-cache)
305 (<= (point) c-macro-cache-start-pos)
306 (>= (point) (car c-macro-cache)))
307 (setq c-macro-cache nil
308 c-macro-cache-start-pos nil
309 c-macro-cache-syntactic nil))
310 (while (progn
311 (end-of-line)
312 (when (and (eq (char-before) ?\\)
313 (not (eobp)))
314 (forward-char)
315 t)))
316 (when (car c-macro-cache)
317 (setcdr c-macro-cache (point)))))
037558bf 318
0ec1d2c5
AM
319(defun c-syntactic-end-of-macro ()
320 ;; Go to the end of a CPP directive, or a "safe" pos just before.
321 ;;
322 ;; This is normally the end of the next non-escaped line. A "safe"
323 ;; position is one not within a string or comment. (The EOL on a line
324 ;; comment is NOT "safe").
325 ;;
326 ;; This function must only be called from the beginning of a CPP construct.
327 ;;
328 ;; Note that this function might do hidden buffer changes. See the comment
329 ;; at the start of cc-engine.el for more info.
330 (let* ((here (point))
331 (there (progn (c-end-of-macro) (point)))
56d093a9
AM
332 s)
333 (unless c-macro-cache-syntactic
334 (setq s (parse-partial-sexp here there))
335 (while (and (or (nth 3 s) ; in a string
336 (nth 4 s)) ; in a comment (maybe at end of line comment)
337 (> there here)) ; No infinite loops, please.
338 (setq there (1- (nth 8 s)))
339 (setq s (parse-partial-sexp here there)))
340 (setq c-macro-cache-syntactic (car c-macro-cache)))
0ec1d2c5
AM
341 (point)))
342
51c9af45
AM
343(defun c-forward-over-cpp-define-id ()
344 ;; Assuming point is at the "#" that introduces a preprocessor
345 ;; directive, it's moved forward to the end of the identifier which is
346 ;; "#define"d (or whatever c-opt-cpp-macro-define specifies). Non-nil
347 ;; is returned in this case, in all other cases nil is returned and
348 ;; point isn't moved.
349 ;;
350 ;; This function might do hidden buffer changes.
351 (when (and c-opt-cpp-macro-define-id
352 (looking-at c-opt-cpp-macro-define-id))
353 (goto-char (match-end 0))))
354
037558bf
MS
355(defun c-forward-to-cpp-define-body ()
356 ;; Assuming point is at the "#" that introduces a preprocessor
357 ;; directive, it's moved forward to the start of the definition body
0386b551
AM
358 ;; if it's a "#define" (or whatever c-opt-cpp-macro-define
359 ;; specifies). Non-nil is returned in this case, in all other cases
360 ;; nil is returned and point isn't moved.
361 ;;
362 ;; This function might do hidden buffer changes.
363 (when (and c-opt-cpp-macro-define-start
364 (looking-at c-opt-cpp-macro-define-start)
037558bf
MS
365 (not (= (match-end 0) (c-point 'eol))))
366 (goto-char (match-end 0))))
367
368\f
d9e94c22
MS
369;;; Basic utility functions.
370
0386b551 371(defun c-syntactic-content (from to paren-level)
d9e94c22
MS
372 ;; Return the given region as a string where all syntactic
373 ;; whitespace is removed or, where necessary, replaced with a single
0386b551
AM
374 ;; space. If PAREN-LEVEL is given then all parens in the region are
375 ;; collapsed to "()", "[]" etc.
376 ;;
377 ;; This function might do hidden buffer changes.
378
d9e94c22 379 (save-excursion
0386b551
AM
380 (save-restriction
381 (narrow-to-region from to)
382 (goto-char from)
383 (let* ((parts (list nil)) (tail parts) pos in-paren)
384
385 (while (re-search-forward c-syntactic-ws-start to t)
386 (goto-char (setq pos (match-beginning 0)))
387 (c-forward-syntactic-ws)
388 (if (= (point) pos)
389 (forward-char)
390
391 (when paren-level
392 (save-excursion
393 (setq in-paren (= (car (parse-partial-sexp from pos 1)) 1)
394 pos (point))))
395
396 (if (and (> pos from)
397 (< (point) to)
398 (looking-at "\\w\\|\\s_")
399 (save-excursion
400 (goto-char (1- pos))
401 (looking-at "\\w\\|\\s_")))
402 (progn
403 (setcdr tail (list (buffer-substring-no-properties from pos)
404 " "))
405 (setq tail (cddr tail)))
406 (setcdr tail (list (buffer-substring-no-properties from pos)))
407 (setq tail (cdr tail)))
408
409 (when in-paren
410 (when (= (car (parse-partial-sexp pos to -1)) -1)
411 (setcdr tail (list (buffer-substring-no-properties
412 (1- (point)) (point))))
413 (setq tail (cdr tail))))
414
415 (setq from (point))))
416
417 (setcdr tail (list (buffer-substring-no-properties from to)))
418 (apply 'concat (cdr parts))))))
419
420(defun c-shift-line-indentation (shift-amt)
421 ;; Shift the indentation of the current line with the specified
422 ;; amount (positive inwards). The buffer is modified only if
423 ;; SHIFT-AMT isn't equal to zero.
424 (let ((pos (- (point-max) (point)))
425 (c-macro-start c-macro-start)
426 tmp-char-inserted)
427 (if (zerop shift-amt)
428 nil
429 ;; If we're on an empty line inside a macro, we take the point
430 ;; to be at the current indentation and shift it to the
431 ;; appropriate column. This way we don't treat the extra
432 ;; whitespace out to the line continuation as indentation.
433 (when (and (c-query-and-set-macro-start)
434 (looking-at "[ \t]*\\\\$")
435 (save-excursion
436 (skip-chars-backward " \t")
437 (bolp)))
438 (insert ?x)
439 (backward-char)
440 (setq tmp-char-inserted t))
441 (unwind-protect
442 (let ((col (current-indentation)))
443 (delete-region (c-point 'bol) (c-point 'boi))
444 (beginning-of-line)
445 (indent-to (+ col shift-amt)))
446 (when tmp-char-inserted
447 (delete-char 1))))
448 ;; If initial point was within line's indentation and we're not on
449 ;; a line with a line continuation in a macro, position after the
450 ;; indentation. Else stay at same point in text.
451 (if (and (< (point) (c-point 'boi))
452 (not tmp-char-inserted))
453 (back-to-indentation)
454 (if (> (- (point-max) pos) (point))
455 (goto-char (- (point-max) pos))))))
d9e94c22
MS
456
457(defsubst c-keyword-sym (keyword)
458 ;; Return non-nil if the string KEYWORD is a known keyword. More
459 ;; precisely, the value is the symbol for the keyword in
460 ;; `c-keywords-obarray'.
461 (intern-soft keyword c-keywords-obarray))
462
463(defsubst c-keyword-member (keyword-sym lang-constant)
464 ;; Return non-nil if the symbol KEYWORD-SYM, as returned by
465 ;; `c-keyword-sym', is a member of LANG-CONSTANT, which is the name
466 ;; of a language constant that ends with "-kwds". If KEYWORD-SYM is
467 ;; nil then the result is nil.
468 (get keyword-sym lang-constant))
469
470;; String syntax chars, suitable for skip-syntax-(forward|backward).
471(defconst c-string-syntax (if (memq 'gen-string-delim c-emacs-features)
472 "\"|"
473 "\""))
474
0386b551 475;; Regexp matching string limit syntax.
d9e94c22
MS
476(defconst c-string-limit-regexp (if (memq 'gen-string-delim c-emacs-features)
477 "\\s\"\\|\\s|"
478 "\\s\""))
479
0386b551
AM
480;; Regexp matching WS followed by string limit syntax.
481(defconst c-ws*-string-limit-regexp
482 (concat "[ \t]*\\(" c-string-limit-regexp "\\)"))
483
d9e94c22
MS
484;; Holds formatted error strings for the few cases where parse errors
485;; are reported.
a66cd3ee 486(defvar c-parsing-error nil)
d9e94c22
MS
487(make-variable-buffer-local 'c-parsing-error)
488
489(defun c-echo-parsing-error (&optional quiet)
d9e94c22
MS
490 (when (and c-report-syntactic-errors c-parsing-error (not quiet))
491 (c-benign-error "%s" c-parsing-error))
492 c-parsing-error)
493
494;; Faces given to comments and string literals. This is used in some
495;; situations to speed up recognition; it isn't mandatory that font
496;; locking is in use. This variable is extended with the face in
497;; `c-doc-face-name' when fontification is activated in cc-fonts.el.
1e330469 498(defvar c-literal-faces
0386b551
AM
499 (append '(font-lock-comment-face font-lock-string-face)
500 (when (facep 'font-lock-comment-delimiter-face)
501 ;; New in Emacs 22.
502 '(font-lock-comment-delimiter-face))))
503
504(defsubst c-put-c-type-property (pos value)
505 ;; Put a c-type property with the given value at POS.
506 (c-put-char-property pos 'c-type value))
507
508(defun c-clear-c-type-property (from to value)
5a89f0a7 509 ;; Remove all occurrences of the c-type property that has the given
0386b551
AM
510 ;; value in the region between FROM and TO. VALUE is assumed to not
511 ;; be nil.
512 ;;
513 ;; Note: This assumes that c-type is put on single chars only; it's
514 ;; very inefficient if matching properties cover large regions.
515 (save-excursion
516 (goto-char from)
517 (while (progn
518 (when (eq (get-text-property (point) 'c-type) value)
519 (c-clear-char-property (point) 'c-type))
520 (goto-char (next-single-property-change (point) 'c-type nil to))
521 (< (point) to)))))
037558bf 522
d9e94c22
MS
523\f
524;; Some debug tools to visualize various special positions. This
525;; debug code isn't as portable as the rest of CC Mode.
526
527(cc-bytecomp-defun overlays-in)
528(cc-bytecomp-defun overlay-get)
529(cc-bytecomp-defun overlay-start)
530(cc-bytecomp-defun overlay-end)
531(cc-bytecomp-defun delete-overlay)
532(cc-bytecomp-defun overlay-put)
533(cc-bytecomp-defun make-overlay)
534
535(defun c-debug-add-face (beg end face)
536 (c-save-buffer-state ((overlays (overlays-in beg end)) overlay)
537 (while overlays
538 (setq overlay (car overlays)
539 overlays (cdr overlays))
540 (when (eq (overlay-get overlay 'face) face)
541 (setq beg (min beg (overlay-start overlay))
542 end (max end (overlay-end overlay)))
543 (delete-overlay overlay)))
544 (overlay-put (make-overlay beg end) 'face face)))
545
546(defun c-debug-remove-face (beg end face)
547 (c-save-buffer-state ((overlays (overlays-in beg end)) overlay
548 (ol-beg beg) (ol-end end))
549 (while overlays
550 (setq overlay (car overlays)
551 overlays (cdr overlays))
552 (when (eq (overlay-get overlay 'face) face)
553 (setq ol-beg (min ol-beg (overlay-start overlay))
554 ol-end (max ol-end (overlay-end overlay)))
555 (delete-overlay overlay)))
556 (when (< ol-beg beg)
557 (overlay-put (make-overlay ol-beg beg) 'face face))
558 (when (> ol-end end)
559 (overlay-put (make-overlay end ol-end) 'face face))))
560
561\f
562;; `c-beginning-of-statement-1' and accompanying stuff.
130c507e 563
64001211
RS
564;; KLUDGE ALERT: c-maybe-labelp is used to pass information between
565;; c-crosses-statement-barrier-p and c-beginning-of-statement-1. A
566;; better way should be implemented, but this will at least shut up
567;; the byte compiler.
0386b551 568(defvar c-maybe-labelp)
64001211 569
d9e94c22
MS
570;; New awk-compatible version of c-beginning-of-statement-1, ACM 2002/6/22
571
a66cd3ee
MS
572;; Macros used internally in c-beginning-of-statement-1 for the
573;; automaton actions.
574(defmacro c-bos-push-state ()
575 '(setq stack (cons (cons state saved-pos)
576 stack)))
577(defmacro c-bos-pop-state (&optional do-if-done)
578 `(if (setq state (car (car stack))
579 saved-pos (cdr (car stack))
580 stack (cdr stack))
581 t
582 ,do-if-done
583 (throw 'loop nil)))
584(defmacro c-bos-pop-state-and-retry ()
585 '(throw 'loop (setq state (car (car stack))
586 saved-pos (cdr (car stack))
587 ;; Throw nil if stack is empty, else throw non-nil.
588 stack (cdr stack))))
589(defmacro c-bos-save-pos ()
590 '(setq saved-pos (vector pos tok ptok pptok)))
591(defmacro c-bos-restore-pos ()
592 '(unless (eq (elt saved-pos 0) start)
593 (setq pos (elt saved-pos 0)
594 tok (elt saved-pos 1)
595 ptok (elt saved-pos 2)
596 pptok (elt saved-pos 3))
597 (goto-char pos)
598 (setq sym nil)))
599(defmacro c-bos-save-error-info (missing got)
600 `(setq saved-pos (vector pos ,missing ,got)))
601(defmacro c-bos-report-error ()
602 '(unless noerror
603 (setq c-parsing-error
604 (format "No matching `%s' found for `%s' on line %d"
605 (elt saved-pos 1)
606 (elt saved-pos 2)
607 (1+ (count-lines (point-min)
608 (c-point 'bol (elt saved-pos 0))))))))
609
610(defun c-beginning-of-statement-1 (&optional lim ignore-labels
611 noerror comma-delim)
612 "Move to the start of the current statement or declaration, or to
613the previous one if already at the beginning of one. Only
614statements/declarations on the same level are considered, i.e. don't
615move into or out of sexps (not even normal expression parentheses).
616
5a89f0a7 617If point is already at the earliest statement within braces or parens,
a85fd6da
AM
618this function doesn't move back into any whitespace preceding it; it
619returns 'same in this case.
620
d9e94c22
MS
621Stop at statement continuation tokens like \"else\", \"catch\",
622\"finally\" and the \"while\" in \"do ... while\" if the start point
623is within the continuation. If starting at such a token, move to the
624corresponding statement start. If at the beginning of a statement,
625move to the closest containing statement if there is any. This might
626also stop at a continuation clause.
a66cd3ee 627
0386b551
AM
628Labels are treated as part of the following statements if
629IGNORE-LABELS is non-nil. (FIXME: Doesn't work if we stop at a known
cb694ab7
AM
630statement start keyword.) Otherwise, each label is treated as a
631separate statement.
a66cd3ee 632
cb694ab7
AM
633Macros are ignored \(i.e. skipped over) unless point is within one, in
634which case the content of the macro is treated as normal code. Aside
635from any normal statement starts found in it, stop at the first token
636of the content in the macro, i.e. the expression of an \"#if\" or the
637start of the definition in a \"#define\". Also stop at start of
638macros before leaving them.
a66cd3ee 639
a85fd6da 640Return:
d28e7f28 641'label if stopped at a label or \"case...:\" or \"default:\";
a85fd6da
AM
642'same if stopped at the beginning of the current statement;
643'up if stepped to a containing statement;
644'previous if stepped to a preceding statement;
645'beginning if stepped from a statement continuation clause to
646 its start clause; or
647'macro if stepped to a macro start.
648Note that 'same and not 'label is returned if stopped at the same
649label without crossing the colon character.
a66cd3ee
MS
650
651LIM may be given to limit the search. If the search hits the limit,
652point will be left at the closest following token, or at the start
653position if that is less ('same is returned in this case).
654
655NOERROR turns off error logging to `c-parsing-error'.
656
cb694ab7
AM
657Normally only ';' and virtual semicolons are considered to delimit
658statements, but if COMMA-DELIM is non-nil then ',' is treated
659as a delimiter too.
0386b551
AM
660
661Note that this function might do hidden buffer changes. See the
662comment at the start of cc-engine.el for more info."
a66cd3ee 663
d9e94c22
MS
664 ;; The bulk of this function is a pushdown automaton that looks at statement
665 ;; boundaries and the tokens (such as "while") in c-opt-block-stmt-key. Its
666 ;; purpose is to keep track of nested statements, ensuring that such
5a89f0a7 667 ;; statements are skipped over in their entirety (somewhat akin to what C-M-p
d9e94c22 668 ;; does with nested braces/brackets/parentheses).
a66cd3ee
MS
669 ;;
670 ;; Note: The position of a boundary is the following token.
671 ;;
d9e94c22
MS
672 ;; Beginning with the current token (the one following point), move back one
673 ;; sexp at a time (where a sexp is, more or less, either a token or the
674 ;; entire contents of a brace/bracket/paren pair). Each time a statement
675 ;; boundary is crossed or a "while"-like token is found, update the state of
676 ;; the PDA. Stop at the beginning of a statement when the stack (holding
677 ;; nested statement info) is empty and the position has been moved.
678 ;;
679 ;; The following variables constitute the PDA:
680 ;;
681 ;; sym: This is either the "while"-like token (e.g. 'for) we've just
682 ;; scanned back over, 'boundary if we've just gone back over a
683 ;; statement boundary, or nil otherwise.
684 ;; state: takes one of the values (nil else else-boundary while
685 ;; while-boundary catch catch-boundary).
686 ;; nil means "no "while"-like token yet scanned".
687 ;; 'else, for example, means "just gone back over an else".
688 ;; 'else-boundary means "just gone back over a statement boundary
689 ;; immediately after having gone back over an else".
690 ;; saved-pos: A vector of either saved positions (tok ptok pptok, etc.) or
691 ;; of error reporting information.
692 ;; stack: The stack onto which the PDA pushes its state. Each entry
693 ;; consists of a saved value of state and saved-pos. An entry is
694 ;; pushed when we move back over a "continuation" token (e.g. else)
695 ;; and popped when we encounter the corresponding opening token
696 ;; (e.g. if).
697 ;;
698 ;;
b414f371 699 ;; The following diagram briefly outlines the PDA.
a66cd3ee
MS
700 ;;
701 ;; Common state:
d9e94c22
MS
702 ;; "else": Push state, goto state `else'.
703 ;; "while": Push state, goto state `while'.
704 ;; "catch" or "finally": Push state, goto state `catch'.
705 ;; boundary: Pop state.
a66cd3ee
MS
706 ;; other: Do nothing special.
707 ;;
d9e94c22
MS
708 ;; State `else':
709 ;; boundary: Goto state `else-boundary'.
710 ;; other: Error, pop state, retry token.
711 ;;
712 ;; State `else-boundary':
713 ;; "if": Pop state.
714 ;; boundary: Error, pop state.
715 ;; other: See common state.
716 ;;
717 ;; State `while':
718 ;; boundary: Save position, goto state `while-boundary'.
719 ;; other: Pop state, retry token.
720 ;;
721 ;; State `while-boundary':
722 ;; "do": Pop state.
723 ;; boundary: Restore position if it's not at start, pop state. [*see below]
724 ;; other: See common state.
725 ;;
726 ;; State `catch':
727 ;; boundary: Goto state `catch-boundary'.
728 ;; other: Error, pop state, retry token.
729 ;;
730 ;; State `catch-boundary':
731 ;; "try": Pop state.
732 ;; "catch": Goto state `catch'.
733 ;; boundary: Error, pop state.
734 ;; other: See common state.
735 ;;
736 ;; [*] In the `while-boundary' state, we had pushed a 'while state, and were
737 ;; searching for a "do" which would have opened a do-while. If we didn't
738 ;; find it, we discard the analysis done since the "while", go back to this
739 ;; token in the buffer and restart the scanning there, this time WITHOUT
740 ;; pushing the 'while state onto the stack.
741 ;;
a66cd3ee
MS
742 ;; In addition to the above there is some special handling of labels
743 ;; and macros.
744
745 (let ((case-fold-search nil)
746 (start (point))
747 macro-start
748 (delims (if comma-delim '(?\; ?,) '(?\;)))
749 (c-stmt-delim-chars (if comma-delim
750 c-stmt-delim-chars-with-comma
751 c-stmt-delim-chars))
d28e7f28 752 c-in-literal-cache c-maybe-labelp after-case:-pos saved
0386b551
AM
753 ;; Current position.
754 pos
755 ;; Position of last stmt boundary character (e.g. ;).
756 boundary-pos
757 ;; The position of the last sexp or bound that follows the
758 ;; first found colon, i.e. the start of the nonlabel part of
759 ;; the statement. It's `start' if a colon is found just after
760 ;; the start.
761 after-labels-pos
762 ;; Like `after-labels-pos', but the first such position inside
763 ;; a label, i.e. the start of the last label before the start
764 ;; of the nonlabel part of the statement.
765 last-label-pos
766 ;; The last position where a label is possible provided the
767 ;; statement started there. It's nil as long as no invalid
768 ;; label content has been found (according to
020716e1 769 ;; `c-nonlabel-token-key'). It's `start' if no valid label
0386b551
AM
770 ;; content was found in the label. Note that we might still
771 ;; regard it a label if it starts with `c-label-kwds'.
772 label-good-pos
b248a85d
AM
773 ;; Putative positions of the components of a bitfield declaration,
774 ;; e.g. "int foo : NUM_FOO_BITS ;"
775 bitfield-type-pos bitfield-id-pos bitfield-size-pos
0386b551
AM
776 ;; Symbol just scanned back over (e.g. 'while or 'boundary).
777 ;; See above.
778 sym
779 ;; Current state in the automaton. See above.
780 state
781 ;; Current saved positions. See above.
782 saved-pos
783 ;; Stack of conses (state . saved-pos).
784 stack
785 ;; Regexp which matches "for", "if", etc.
786 (cond-key (or c-opt-block-stmt-key
a66cd3ee 787 "\\<\\>")) ; Matches nothing.
0386b551
AM
788 ;; Return value.
789 (ret 'same)
790 ;; Positions of the last three sexps or bounds we've stopped at.
791 tok ptok pptok)
a66cd3ee
MS
792
793 (save-restriction
794 (if lim (narrow-to-region lim (point-max)))
795
796 (if (save-excursion
797 (and (c-beginning-of-macro)
798 (/= (point) start)))
799 (setq macro-start (point)))
800
d9e94c22 801 ;; Try to skip back over unary operator characters, to register
a66cd3ee
MS
802 ;; that we've moved.
803 (while (progn
804 (setq pos (point))
0386b551
AM
805 (c-backward-syntactic-ws)
806 ;; Protect post-++/-- operators just before a virtual semicolon.
807 (and (not (c-at-vsemi-p))
808 (/= (skip-chars-backward "-+!*&~@`#") 0))))
d9e94c22
MS
809
810 ;; Skip back over any semicolon here. If it was a bare semicolon, we're
0386b551 811 ;; done. Later on we ignore the boundaries for statements that don't
d9e94c22
MS
812 ;; contain any sexp. The only thing that is affected is that the error
813 ;; checking is a little less strict, and we really don't bother.
a66cd3ee
MS
814 (if (and (memq (char-before) delims)
815 (progn (forward-char -1)
816 (setq saved (point))
0386b551 817 (c-backward-syntactic-ws)
a66cd3ee
MS
818 (or (memq (char-before) delims)
819 (memq (char-before) '(?: nil))
d9e94c22 820 (eq (char-syntax (char-before)) ?\()
0386b551 821 (c-at-vsemi-p))))
a66cd3ee
MS
822 (setq ret 'previous
823 pos saved)
824
825 ;; Begin at start and not pos to detect macros if we stand
826 ;; directly after the #.
827 (goto-char start)
828 (if (looking-at "\\<\\|\\W")
829 ;; Record this as the first token if not starting inside it.
830 (setq tok start))
831
b248a85d
AM
832
833 ;; The following while loop goes back one sexp (balanced parens,
834 ;; etc. with contents, or symbol or suchlike) each iteration. This
835 ;; movement is accomplished with a call to c-backward-sexp approx 170
836 ;; lines below.
837 ;;
838 ;; The loop is exited only by throwing nil to the (catch 'loop ...):
839 ;; 1. On reaching the start of a macro;
840 ;; 2. On having passed a stmt boundary with the PDA stack empty;
841 ;; 3. On reaching the start of an Objective C method def;
842 ;; 4. From macro `c-bos-pop-state'; when the stack is empty;
843 ;; 5. From macro `c-bos-pop-state-and-retry' when the stack is empty.
a66cd3ee
MS
844 (while
845 (catch 'loop ;; Throw nil to break, non-nil to continue.
846 (cond
b248a85d 847 ;; Are we in a macro, just after the opening #?
a66cd3ee 848 ((save-excursion
0386b551 849 (and macro-start ; Always NIL for AWK.
a66cd3ee
MS
850 (progn (skip-chars-backward " \t")
851 (eq (char-before) ?#))
852 (progn (setq saved (1- (point)))
853 (beginning-of-line)
854 (not (eq (char-before (1- (point))) ?\\)))
d9e94c22 855 (looking-at c-opt-cpp-start)
a66cd3ee
MS
856 (progn (skip-chars-forward " \t")
857 (eq (point) saved))))
858 (goto-char saved)
859 (if (and (c-forward-to-cpp-define-body)
860 (progn (c-forward-syntactic-ws start)
861 (< (point) start)))
862 ;; Stop at the first token in the content of the macro.
863 (setq pos (point)
864 ignore-labels t) ; Avoid the label check on exit.
865 (setq pos saved
866 ret 'macro
867 ignore-labels t))
b248a85d 868 (throw 'loop nil)) ; 1. Start of macro.
a66cd3ee 869
d9e94c22
MS
870 ;; Do a round through the automaton if we've just passed a
871 ;; statement boundary or passed a "while"-like token.
a66cd3ee
MS
872 ((or sym
873 (and (looking-at cond-key)
874 (setq sym (intern (match-string 1)))))
875
876 (when (and (< pos start) (null stack))
b248a85d 877 (throw 'loop nil)) ; 2. Statement boundary.
a66cd3ee 878
d9e94c22
MS
879 ;; The PDA state handling.
880 ;;
037558bf 881 ;; Refer to the description of the PDA in the opening
d9e94c22
MS
882 ;; comments. In the following OR form, the first leaf
883 ;; attempts to handles one of the specific actions detailed
884 ;; (e.g., finding token "if" whilst in state `else-boundary').
885 ;; We drop through to the second leaf (which handles common
886 ;; state) if no specific handler is found in the first cond.
887 ;; If a parsing error is detected (e.g. an "else" with no
888 ;; preceding "if"), we throw to the enclosing catch.
889 ;;
890 ;; Note that the (eq state 'else) means
891 ;; "we've just passed an else", NOT "we're looking for an
892 ;; else".
a66cd3ee
MS
893 (or (cond
894 ((eq state 'else)
895 (if (eq sym 'boundary)
896 (setq state 'else-boundary)
897 (c-bos-report-error)
898 (c-bos-pop-state-and-retry)))
899
900 ((eq state 'else-boundary)
901 (cond ((eq sym 'if)
902 (c-bos-pop-state (setq ret 'beginning)))
903 ((eq sym 'boundary)
904 (c-bos-report-error)
905 (c-bos-pop-state))))
906
907 ((eq state 'while)
908 (if (and (eq sym 'boundary)
909 ;; Since this can cause backtracking we do a
910 ;; little more careful analysis to avoid it:
911 ;; If there's a label in front of the while
912 ;; it can't be part of a do-while.
913 (not after-labels-pos))
914 (progn (c-bos-save-pos)
915 (setq state 'while-boundary))
d9e94c22 916 (c-bos-pop-state-and-retry))) ; Can't be a do-while
a66cd3ee
MS
917
918 ((eq state 'while-boundary)
919 (cond ((eq sym 'do)
920 (c-bos-pop-state (setq ret 'beginning)))
d9e94c22
MS
921 ((eq sym 'boundary) ; isn't a do-while
922 (c-bos-restore-pos) ; the position of the while
923 (c-bos-pop-state)))) ; no longer searching for do.
a66cd3ee
MS
924
925 ((eq state 'catch)
926 (if (eq sym 'boundary)
927 (setq state 'catch-boundary)
928 (c-bos-report-error)
929 (c-bos-pop-state-and-retry)))
930
931 ((eq state 'catch-boundary)
932 (cond
933 ((eq sym 'try)
934 (c-bos-pop-state (setq ret 'beginning)))
935 ((eq sym 'catch)
936 (setq state 'catch))
937 ((eq sym 'boundary)
938 (c-bos-report-error)
939 (c-bos-pop-state)))))
940
d9e94c22
MS
941 ;; This is state common. We get here when the previous
942 ;; cond statement found no particular state handler.
a66cd3ee 943 (cond ((eq sym 'boundary)
d9e94c22
MS
944 ;; If we have a boundary at the start
945 ;; position we push a frame to go to the
946 ;; previous statement.
947 (if (>= pos start)
948 (c-bos-push-state)
949 (c-bos-pop-state)))
a66cd3ee
MS
950 ((eq sym 'else)
951 (c-bos-push-state)
952 (c-bos-save-error-info 'if 'else)
953 (setq state 'else))
954 ((eq sym 'while)
0386b551
AM
955 ;; Is this a real while, or a do-while?
956 ;; The next `when' triggers unless we are SURE that
40ba43b4 957 ;; the `while' is not the tail end of a `do-while'.
a66cd3ee 958 (when (or (not pptok)
d9e94c22 959 (memq (char-after pptok) delims)
0386b551
AM
960 ;; The following kludge is to prevent
961 ;; infinite recursion when called from
962 ;; c-awk-after-if-for-while-condition-p,
963 ;; or the like.
964 (and (eq (point) start)
965 (c-vsemi-status-unknown-p))
966 (c-at-vsemi-p pptok))
a66cd3ee
MS
967 ;; Since this can cause backtracking we do a
968 ;; little more careful analysis to avoid it: If
0386b551
AM
969 ;; the while isn't followed by a (possibly
970 ;; virtual) semicolon it can't be a do-while.
a66cd3ee
MS
971 (c-bos-push-state)
972 (setq state 'while)))
973 ((memq sym '(catch finally))
974 (c-bos-push-state)
975 (c-bos-save-error-info 'try sym)
976 (setq state 'catch))))
977
978 (when c-maybe-labelp
979 ;; We're either past a statement boundary or at the
980 ;; start of a statement, so throw away any label data
981 ;; for the previous one.
982 (setq after-labels-pos nil
983 last-label-pos nil
984 c-maybe-labelp nil))))
985
d9e94c22
MS
986 ;; Step to the previous sexp, but not if we crossed a
987 ;; boundary, since that doesn't consume an sexp.
a66cd3ee
MS
988 (if (eq sym 'boundary)
989 (setq ret 'previous)
d9e94c22
MS
990
991 ;; HERE IS THE SINGLE PLACE INSIDE THE PDA LOOP WHERE WE MOVE
0386b551
AM
992 ;; BACKWARDS THROUGH THE SOURCE.
993
0386b551 994 (c-backward-syntactic-ws)
0386b551 995 (let ((before-sws-pos (point))
0386b551
AM
996 ;; The end position of the area to search for statement
997 ;; barriers in this round.
b248a85d 998 (maybe-after-boundary-pos pos))
0386b551 999
b248a85d
AM
1000 ;; Go back over exactly one logical sexp, taking proper
1001 ;; account of macros and escaped EOLs.
0386b551
AM
1002 (while
1003 (progn
1004 (unless (c-safe (c-backward-sexp) t)
1005 ;; Give up if we hit an unbalanced block. Since the
1006 ;; stack won't be empty the code below will report a
1007 ;; suitable error.
a66cd3ee 1008 (throw 'loop nil))
b248a85d
AM
1009 (cond
1010 ;; Have we moved into a macro?
1011 ((and (not macro-start)
1012 (c-beginning-of-macro))
1013 ;; Have we crossed a statement boundary? If not,
1014 ;; keep going back until we find one or a "real" sexp.
1015 (and
1016 (save-excursion
1017 (c-end-of-macro)
1018 (not (c-crosses-statement-barrier-p
1019 (point) maybe-after-boundary-pos)))
1020 (setq maybe-after-boundary-pos (point))))
1021 ;; Have we just gone back over an escaped NL? This
1022 ;; doesn't count as a sexp.
1023 ((looking-at "\\\\$")))))
1024
1025 ;; Have we crossed a statement boundary?
1026 (setq boundary-pos
1027 (cond
1028 ;; Are we at a macro beginning?
1029 ((and (not macro-start)
1030 c-opt-cpp-prefix
1031 (looking-at c-opt-cpp-prefix))
1032 (save-excursion
1033 (c-end-of-macro)
1034 (c-crosses-statement-barrier-p
1035 (point) maybe-after-boundary-pos)))
1036 ;; Just gone back over a brace block?
1037 ((and
1038 (eq (char-after) ?{)
1039 (not (c-looking-at-inexpr-block lim nil t)))
1040 (save-excursion
1041 (c-forward-sexp) (point)))
1042 ;; Just gone back over some paren block?
1043 ((looking-at "\\s\(")
1044 (save-excursion
1045 (goto-char (1+ (c-down-list-backward
1046 before-sws-pos)))
1047 (c-crosses-statement-barrier-p
1048 (point) maybe-after-boundary-pos)))
1049 ;; Just gone back over an ordinary symbol of some sort?
1050 (t (c-crosses-statement-barrier-p
1051 (point) maybe-after-boundary-pos))))
91af3942 1052
b248a85d
AM
1053 (when boundary-pos
1054 (setq pptok ptok
1055 ptok tok
1056 tok boundary-pos
1057 sym 'boundary)
1058 ;; Like a C "continue". Analyze the next sexp.
1059 (throw 'loop t))))
a66cd3ee
MS
1060
1061 ;; ObjC method def?
1062 (when (and c-opt-method-key
1063 (setq saved (c-in-method-def-p)))
1064 (setq pos saved
1065 ignore-labels t) ; Avoid the label check on exit.
b248a85d
AM
1066 (throw 'loop nil)) ; 3. ObjC method def.
1067
1068 ;; Might we have a bitfield declaration, "<type> <id> : <size>"?
1069 (if c-has-bitfields
1070 (cond
1071 ;; The : <size> and <id> fields?
1072 ((and (numberp c-maybe-labelp)
1073 (not bitfield-size-pos)
1074 (save-excursion
1075 (goto-char (or tok start))
1076 (not (looking-at c-keywords-regexp)))
1077 (not (looking-at c-keywords-regexp))
1078 (not (c-punctuation-in (point) c-maybe-labelp)))
1079 (setq bitfield-size-pos (or tok start)
1080 bitfield-id-pos (point)))
1081 ;; The <type> field?
1082 ((and bitfield-id-pos
1083 (not bitfield-type-pos))
1084 (if (and (looking-at c-symbol-key) ; Can only be an integer type. :-)
1085 (not (looking-at c-not-primitive-type-keywords-regexp))
1086 (not (c-punctuation-in (point) tok)))
1087 (setq bitfield-type-pos (point))
1088 (setq bitfield-size-pos nil
1089 bitfield-id-pos nil)))))
a66cd3ee 1090
0386b551
AM
1091 ;; Handle labels.
1092 (unless (eq ignore-labels t)
1093 (when (numberp c-maybe-labelp)
cb694ab7
AM
1094 ;; `c-crosses-statement-barrier-p' has found a colon, so we
1095 ;; might be in a label now. Have we got a real label
1096 ;; (including a case label) or something like C++'s "public:"?
d28e7f28
AM
1097 ;; A case label might use an expression rather than a token.
1098 (setq after-case:-pos (or tok start))
020716e1
AM
1099 (if (or (looking-at c-nonlabel-token-key) ; e.g. "while" or "'a'"
1100 ;; Catch C++'s inheritance construct "class foo : bar".
1101 (save-excursion
1102 (and
1103 (c-safe (c-backward-sexp) t)
1104 (looking-at c-nonlabel-token-2-key))))
d28e7f28
AM
1105 (setq c-maybe-labelp nil)
1106 (if after-labels-pos ; Have we already encountered a label?
1107 (if (not last-label-pos)
1108 (setq last-label-pos (or tok start)))
1109 (setq after-labels-pos (or tok start)))
1110 (setq c-maybe-labelp t
1111 label-good-pos nil))) ; bogus "label"
cb694ab7
AM
1112
1113 (when (and (not label-good-pos) ; i.e. no invalid "label"'s yet
1114 ; been found.
1115 (looking-at c-nonlabel-token-key)) ; e.g. "while :"
0386b551
AM
1116 ;; We're in a potential label and it's the first
1117 ;; time we've found something that isn't allowed in
1118 ;; one.
1119 (setq label-good-pos (or tok start))))
1120
1121 ;; We've moved back by a sexp, so update the token positions.
a66cd3ee
MS
1122 (setq sym nil
1123 pptok ptok
1124 ptok tok
1125 tok (point)
b248a85d
AM
1126 pos tok) ; always non-nil
1127 ) ; end of (catch loop ....)
1128 ) ; end of sexp-at-a-time (while ....)
91af3942 1129
a66cd3ee
MS
1130 ;; If the stack isn't empty there might be errors to report.
1131 (while stack
1132 (if (and (vectorp saved-pos) (eq (length saved-pos) 3))
1133 (c-bos-report-error))
1134 (setq saved-pos (cdr (car stack))
1135 stack (cdr stack)))
1136
1137 (when (and (eq ret 'same)
1138 (not (memq sym '(boundary ignore nil))))
1139 ;; Need to investigate closer whether we've crossed
1140 ;; between a substatement and its containing statement.
1141 (if (setq saved (if (looking-at c-block-stmt-1-key)
1142 ptok
1143 pptok))
1144 (cond ((> start saved) (setq pos saved))
1145 ((= start saved) (setq ret 'up)))))
1146
0386b551
AM
1147 (when (and (not ignore-labels)
1148 (eq c-maybe-labelp t)
d9e94c22 1149 (not (eq ret 'beginning))
0386b551 1150 after-labels-pos
b248a85d 1151 (not bitfield-type-pos) ; Bitfields take precedence over labels.
0386b551
AM
1152 (or (not label-good-pos)
1153 (<= label-good-pos pos)
1154 (progn
1155 (goto-char (if (and last-label-pos
1156 (< last-label-pos start))
1157 last-label-pos
1158 pos))
1159 (looking-at c-label-kwds-regexp))))
a66cd3ee
MS
1160 ;; We're in a label. Maybe we should step to the statement
1161 ;; after it.
1162 (if (< after-labels-pos start)
1163 (setq pos after-labels-pos)
1164 (setq ret 'label)
0386b551
AM
1165 (if (and last-label-pos (< last-label-pos start))
1166 ;; Might have jumped over several labels. Go to the last one.
a66cd3ee
MS
1167 (setq pos last-label-pos)))))
1168
d28e7f28 1169 ;; Have we got "case <expression>:"?
a66cd3ee 1170 (goto-char pos)
d28e7f28
AM
1171 (when (and after-case:-pos
1172 (not (eq ret 'beginning))
1173 (looking-at c-case-kwds-regexp))
1174 (if (< after-case:-pos start)
0000ee90
AM
1175 (setq pos after-case:-pos))
1176 (if (eq ret 'same)
1177 (setq ret 'label)))
d28e7f28
AM
1178
1179 ;; Skip over the unary operators that can start the statement.
a66cd3ee 1180 (while (progn
0386b551
AM
1181 (c-backward-syntactic-ws)
1182 ;; protect AWK post-inc/decrement operators, etc.
1183 (and (not (c-at-vsemi-p (point)))
1184 (/= (skip-chars-backward "-+!*&~@`#") 0)))
a66cd3ee
MS
1185 (setq pos (point)))
1186 (goto-char pos)
1187 ret)))
785eecbb 1188
b248a85d
AM
1189(defun c-punctuation-in (from to)
1190 "Return non-nil if there is a non-comment non-macro punctuation character
1191between FROM and TO. FROM must not be in a string or comment. The returned
1192value is the position of the first such character."
1193 (save-excursion
1194 (goto-char from)
1195 (let ((pos (point)))
1196 (while (progn (skip-chars-forward c-symbol-chars to)
1197 (c-forward-syntactic-ws to)
1198 (> (point) pos))
1199 (setq pos (point))))
1200 (and (< (point) to) (point))))
1201
785eecbb 1202(defun c-crosses-statement-barrier-p (from to)
a66cd3ee
MS
1203 "Return non-nil if buffer positions FROM to TO cross one or more
1204statement or declaration boundaries. The returned value is actually
d9e94c22
MS
1205the position of the earliest boundary char. FROM must not be within
1206a string or comment.
a66cd3ee
MS
1207
1208The variable `c-maybe-labelp' is set to the position of the first `:' that
1209might start a label (i.e. not part of `::' and not preceded by `?'). If a
0386b551
AM
1210single `?' is found, then `c-maybe-labelp' is cleared.
1211
1212For AWK, a statement which is terminated by an EOL (not a \; or a }) is
1213regarded as having a \"virtual semicolon\" immediately after the last token on
b414f371 1214the line. If this virtual semicolon is _at_ from, the function recognizes it.
0386b551
AM
1215
1216Note that this function might do hidden buffer changes. See the
1217comment at the start of cc-engine.el for more info."
536610a4
AM
1218 (let* ((skip-chars
1219 ;; If the current language has CPP macros, insert # into skip-chars.
1220 (if c-opt-cpp-symbol
1221 (concat (substring c-stmt-delim-chars 0 1) ; "^"
1222 c-opt-cpp-symbol ; usually "#"
1223 (substring c-stmt-delim-chars 1)) ; e.g. ";{}?:"
1224 c-stmt-delim-chars))
1225 (non-skip-list
1226 (append (substring skip-chars 1) nil)) ; e.g. (?# ?\; ?{ ?} ?? ?:)
1227 lit-range vsemi-pos)
1228 (save-restriction
1229 (widen)
1230 (save-excursion
1231 (catch 'done
1232 (goto-char from)
1233 (while (progn (skip-chars-forward
1234 skip-chars
1235 (min to (c-point 'bonl)))
1236 (< (point) to))
1237 (cond
1238 ;; Virtual semicolon?
1239 ((and (bolp)
1240 (save-excursion
1241 (progn
1242 (if (setq lit-range (c-literal-limits from)) ; Have we landed in a string/comment?
1243 (goto-char (car lit-range)))
1244 (c-backward-syntactic-ws) ; ? put a limit here, maybe?
1245 (setq vsemi-pos (point))
1246 (c-at-vsemi-p))))
1247 (throw 'done vsemi-pos))
1248 ;; In a string/comment?
2e492df3 1249 ((setq lit-range (c-literal-limits from))
536610a4
AM
1250 (goto-char (cdr lit-range)))
1251 ((eq (char-after) ?:)
1252 (forward-char)
1253 (if (and (eq (char-after) ?:)
1254 (< (point) to))
1255 ;; Ignore scope operators.
1256 (forward-char)
1257 (setq c-maybe-labelp (1- (point)))))
1258 ((eq (char-after) ??)
1259 ;; A question mark. Can't be a label, so stop
1260 ;; looking for more : and ?.
1261 (setq c-maybe-labelp nil
1262 skip-chars (substring c-stmt-delim-chars 0 -2)))
1263 ;; At a CPP construct?
1264 ((and c-opt-cpp-symbol (looking-at c-opt-cpp-symbol)
1265 (save-excursion
1266 (forward-line 0)
1267 (looking-at c-opt-cpp-prefix)))
1268 (c-end-of-macro))
1269 ((memq (char-after) non-skip-list)
1270 (throw 'done (point)))))
1271 ;; In trailing space after an as yet undetected virtual semicolon?
1272 (c-backward-syntactic-ws from)
1273 (if (and (< (point) to)
1274 (c-at-vsemi-p))
1275 (point)
1276 nil))))))
0386b551
AM
1277
1278(defun c-at-statement-start-p ()
1279 "Return non-nil if the point is at the first token in a statement
1280or somewhere in the syntactic whitespace before it.
1281
1282A \"statement\" here is not restricted to those inside code blocks.
1283Any kind of declaration-like construct that occur outside function
1284bodies is also considered a \"statement\".
1285
1286Note that this function might do hidden buffer changes. See the
1287comment at the start of cc-engine.el for more info."
1288
1289 (save-excursion
1290 (let ((end (point))
1291 c-maybe-labelp)
1292 (c-syntactic-skip-backward (substring c-stmt-delim-chars 1) nil t)
1293 (or (bobp)
1294 (eq (char-before) ?})
1295 (and (eq (char-before) ?{)
1296 (not (and c-special-brace-lists
1297 (progn (backward-char)
1298 (c-looking-at-special-brace-list)))))
1299 (c-crosses-statement-barrier-p (point) end)))))
1300
1301(defun c-at-expression-start-p ()
1302 "Return non-nil if the point is at the first token in an expression or
1303statement, or somewhere in the syntactic whitespace before it.
1304
1305An \"expression\" here is a bit different from the normal language
1306grammar sense: It's any sequence of expression tokens except commas,
1307unless they are enclosed inside parentheses of some kind. Also, an
1308expression never continues past an enclosing parenthesis, but it might
1309contain parenthesis pairs of any sort except braces.
1310
1311Since expressions never cross statement boundaries, this function also
1312recognizes statement beginnings, just like `c-at-statement-start-p'.
1313
1314Note that this function might do hidden buffer changes. See the
1315comment at the start of cc-engine.el for more info."
1316
1317 (save-excursion
1318 (let ((end (point))
1319 (c-stmt-delim-chars c-stmt-delim-chars-with-comma)
1320 c-maybe-labelp)
1321 (c-syntactic-skip-backward (substring c-stmt-delim-chars 1) nil t)
1322 (or (bobp)
1323 (memq (char-before) '(?{ ?}))
1324 (save-excursion (backward-char)
1325 (looking-at "\\s("))
1326 (c-crosses-statement-barrier-p (point) end)))))
785eecbb
RS
1327
1328\f
d9e94c22
MS
1329;; A set of functions that covers various idiosyncrasies in
1330;; implementations of `forward-comment'.
1331
1332;; Note: Some emacsen considers incorrectly that any line comment
1333;; ending with a backslash continues to the next line. I can't think
1334;; of any way to work around that in a reliable way without changing
1335;; the buffer, though. Suggestions welcome. ;) (No, temporarily
1336;; changing the syntax for backslash doesn't work since we must treat
1337;; escapes in string literals correctly.)
1338
1339(defun c-forward-single-comment ()
1340 "Move forward past whitespace and the closest following comment, if any.
1341Return t if a comment was found, nil otherwise. In either case, the
1342point is moved past the following whitespace. Line continuations,
1343i.e. a backslashes followed by line breaks, are treated as whitespace.
1344The line breaks that end line comments are considered to be the
1345comment enders, so the point will be put on the beginning of the next
1346line if it moved past a line comment.
1347
1348This function does not do any hidden buffer changes."
1349
1350 (let ((start (point)))
1351 (when (looking-at "\\([ \t\n\r\f\v]\\|\\\\[\n\r]\\)+")
1352 (goto-char (match-end 0)))
1353
1354 (when (forward-comment 1)
1355 (if (eobp)
1356 ;; Some emacsen (e.g. XEmacs 21) return t when moving
1357 ;; forwards at eob.
1358 nil
1359
1360 ;; Emacs includes the ending newline in a b-style (c++)
1361 ;; comment, but XEmacs doesn't. We depend on the Emacs
1362 ;; behavior (which also is symmetric).
1363 (if (and (eolp) (elt (parse-partial-sexp start (point)) 7))
1364 (condition-case nil (forward-char 1)))
1365
1366 t))))
1367
1368(defsubst c-forward-comments ()
1369 "Move forward past all following whitespace and comments.
1370Line continuations, i.e. a backslashes followed by line breaks, are
1371treated as whitespace.
1372
0386b551
AM
1373Note that this function might do hidden buffer changes. See the
1374comment at the start of cc-engine.el for more info."
d9e94c22
MS
1375
1376 (while (or
1377 ;; If forward-comment in at least XEmacs 21 is given a large
1378 ;; positive value, it'll loop all the way through if it hits
1379 ;; eob.
1380 (and (forward-comment 5)
1381 ;; Some emacsen (e.g. XEmacs 21) return t when moving
1382 ;; forwards at eob.
1383 (not (eobp)))
1384
1385 (when (looking-at "\\\\[\n\r]")
1386 (forward-char 2)
1387 t))))
1388
1389(defun c-backward-single-comment ()
1390 "Move backward past whitespace and the closest preceding comment, if any.
1391Return t if a comment was found, nil otherwise. In either case, the
1392point is moved past the preceding whitespace. Line continuations,
1393i.e. a backslashes followed by line breaks, are treated as whitespace.
1394The line breaks that end line comments are considered to be the
1395comment enders, so the point cannot be at the end of the same line to
1396move over a line comment.
1397
1398This function does not do any hidden buffer changes."
1399
1400 (let ((start (point)))
1401 ;; When we got newline terminated comments, forward-comment in all
1402 ;; supported emacsen so far will stop at eol of each line not
1403 ;; ending with a comment when moving backwards. This corrects for
1404 ;; that, and at the same time handles line continuations.
1405 (while (progn
1406 (skip-chars-backward " \t\n\r\f\v")
1407 (and (looking-at "[\n\r]")
0386b551 1408 (eq (char-before) ?\\)))
d9e94c22
MS
1409 (backward-char))
1410
1411 (if (bobp)
1412 ;; Some emacsen (e.g. Emacs 19.34) return t when moving
1413 ;; backwards at bob.
1414 nil
1415
1416 ;; Leave point after the closest following newline if we've
1417 ;; backed up over any above, since forward-comment won't move
1418 ;; backward over a line comment if point is at the end of the
1419 ;; same line.
1420 (re-search-forward "\\=\\s *[\n\r]" start t)
1421
bf2c1571 1422 (if (if (let (open-paren-in-column-0-is-defun-start) (forward-comment -1))
d9e94c22
MS
1423 (if (eolp)
1424 ;; If forward-comment above succeeded and we're at eol
1425 ;; then the newline we moved over above didn't end a
1426 ;; line comment, so we give it another go.
bf2c1571
AM
1427 (let (open-paren-in-column-0-is-defun-start)
1428 (forward-comment -1))
d9e94c22
MS
1429 t))
1430
1431 ;; Emacs <= 20 and XEmacs move back over the closer of a
1432 ;; block comment that lacks an opener.
1433 (if (looking-at "\\*/")
1434 (progn (forward-char 2) nil)
1435 t)))))
1436
1437(defsubst c-backward-comments ()
1438 "Move backward past all preceding whitespace and comments.
1439Line continuations, i.e. a backslashes followed by line breaks, are
1440treated as whitespace. The line breaks that end line comments are
1441considered to be the comment enders, so the point cannot be at the end
0386b551
AM
1442of the same line to move over a line comment. Unlike
1443c-backward-syntactic-ws, this function doesn't move back over
1444preprocessor directives.
d9e94c22 1445
0386b551
AM
1446Note that this function might do hidden buffer changes. See the
1447comment at the start of cc-engine.el for more info."
d9e94c22
MS
1448
1449 (let ((start (point)))
1450 (while (and
0386b551 1451 ;; `forward-comment' in some emacsen (e.g. XEmacs 21.4)
d9e94c22
MS
1452 ;; return t when moving backwards at bob.
1453 (not (bobp))
1454
14b8c3d9
AM
1455 (if (let (open-paren-in-column-0-is-defun-start moved-comment)
1456 (while
1457 (and (not (setq moved-comment (forward-comment -1)))
1458 ;; Cope specifically with ^M^J here -
1459 ;; forward-comment sometimes gets stuck after ^Ms,
1460 ;; sometimes after ^M^J.
1461 (or
1462 (when (eq (char-before) ?\r)
1463 (backward-char)
1464 t)
1465 (when (and (eq (char-before) ?\n)
1466 (eq (char-before (1- (point))) ?\r))
1467 (backward-char 2)
1468 t))))
1469 moved-comment)
d9e94c22
MS
1470 (if (looking-at "\\*/")
1471 ;; Emacs <= 20 and XEmacs move back over the
1472 ;; closer of a block comment that lacks an opener.
1473 (progn (forward-char 2) nil)
1474 t)
1475
1476 ;; XEmacs treats line continuations as whitespace but
1477 ;; only in the backward direction, which seems a bit
1478 ;; odd. Anyway, this is necessary for Emacs.
1479 (when (and (looking-at "[\n\r]")
1480 (eq (char-before) ?\\)
1481 (< (point) start))
1482 (backward-char)
1483 t))))))
1484
1485\f
d9e94c22 1486;; Tools for skipping over syntactic whitespace.
a66cd3ee 1487
d9e94c22
MS
1488;; The following functions use text properties to cache searches over
1489;; large regions of syntactic whitespace. It works as follows:
1490;;
1491;; o If a syntactic whitespace region contains anything but simple
1492;; whitespace (i.e. space, tab and line breaks), the text property
1493;; `c-in-sws' is put over it. At places where we have stopped
1494;; within that region there's also a `c-is-sws' text property.
1495;; That since there typically are nested whitespace inside that
1496;; must be handled separately, e.g. whitespace inside a comment or
1497;; cpp directive. Thus, from one point with `c-is-sws' it's safe
1498;; to jump to another point with that property within the same
1499;; `c-in-sws' region. It can be likened to a ladder where
1500;; `c-in-sws' marks the bars and `c-is-sws' the rungs.
1501;;
1502;; o The `c-is-sws' property is put on the simple whitespace chars at
1503;; a "rung position" and also maybe on the first following char.
1504;; As many characters as can be conveniently found in this range
1505;; are marked, but no assumption can be made that the whole range
1506;; is marked (it could be clobbered by later changes, for
1507;; instance).
1508;;
1509;; Note that some part of the beginning of a sequence of simple
1510;; whitespace might be part of the end of a preceding line comment
1511;; or cpp directive and must not be considered part of the "rung".
1512;; Such whitespace is some amount of horizontal whitespace followed
1513;; by a newline. In the case of cpp directives it could also be
1514;; two newlines with horizontal whitespace between them.
1515;;
1516;; The reason to include the first following char is to cope with
1517;; "rung positions" that doesn't have any ordinary whitespace. If
1518;; `c-is-sws' is put on a token character it does not have
1519;; `c-in-sws' set simultaneously. That's the only case when that
1520;; can occur, and the reason for not extending the `c-in-sws'
1521;; region to cover it is that the `c-in-sws' region could then be
1522;; accidentally merged with a following one if the token is only
1523;; one character long.
1524;;
1525;; o On buffer changes the `c-in-sws' and `c-is-sws' properties are
1526;; removed in the changed region. If the change was inside
1527;; syntactic whitespace that means that the "ladder" is broken, but
1528;; a later call to `c-forward-sws' or `c-backward-sws' will use the
1529;; parts on either side and use an ordinary search only to "repair"
1530;; the gap.
1531;;
1532;; Special care needs to be taken if a region is removed: If there
1533;; are `c-in-sws' on both sides of it which do not connect inside
1534;; the region then they can't be joined. If e.g. a marked macro is
1535;; broken, syntactic whitespace inside the new text might be
1536;; marked. If those marks would become connected with the old
1537;; `c-in-sws' range around the macro then we could get a ladder
1538;; with one end outside the macro and the other at some whitespace
1539;; within it.
1540;;
1541;; The main motivation for this system is to increase the speed in
1542;; skipping over the large whitespace regions that can occur at the
1543;; top level in e.g. header files that contain a lot of comments and
1544;; cpp directives. For small comments inside code it's probably
1545;; slower than using `forward-comment' straightforwardly, but speed is
1546;; not a significant factor there anyway.
1547
1548; (defface c-debug-is-sws-face
1549; '((t (:background "GreenYellow")))
1550; "Debug face to mark the `c-is-sws' property.")
1551; (defface c-debug-in-sws-face
1552; '((t (:underline t)))
1553; "Debug face to mark the `c-in-sws' property.")
1554
1555; (defun c-debug-put-sws-faces ()
1556; ;; Put the sws debug faces on all the `c-is-sws' and `c-in-sws'
1557; ;; properties in the buffer.
1558; (interactive)
1559; (save-excursion
0386b551 1560; (c-save-buffer-state (in-face)
d9e94c22
MS
1561; (goto-char (point-min))
1562; (setq in-face (if (get-text-property (point) 'c-is-sws)
1563; (point)))
1564; (while (progn
1565; (goto-char (next-single-property-change
1566; (point) 'c-is-sws nil (point-max)))
1567; (if in-face
1568; (progn
1569; (c-debug-add-face in-face (point) 'c-debug-is-sws-face)
1570; (setq in-face nil))
1571; (setq in-face (point)))
1572; (not (eobp))))
1573; (goto-char (point-min))
1574; (setq in-face (if (get-text-property (point) 'c-in-sws)
1575; (point)))
1576; (while (progn
1577; (goto-char (next-single-property-change
1578; (point) 'c-in-sws nil (point-max)))
1579; (if in-face
1580; (progn
1581; (c-debug-add-face in-face (point) 'c-debug-in-sws-face)
1582; (setq in-face nil))
1583; (setq in-face (point)))
1584; (not (eobp)))))))
1585
1586(defmacro c-debug-sws-msg (&rest args)
1587 ;;`(message ,@args)
1588 )
1589
1590(defmacro c-put-is-sws (beg end)
0386b551 1591 ;; This macro does a hidden buffer change.
d9e94c22
MS
1592 `(let ((beg ,beg) (end ,end))
1593 (put-text-property beg end 'c-is-sws t)
1594 ,@(when (facep 'c-debug-is-sws-face)
1595 `((c-debug-add-face beg end 'c-debug-is-sws-face)))))
1596
1597(defmacro c-put-in-sws (beg end)
0386b551 1598 ;; This macro does a hidden buffer change.
d9e94c22
MS
1599 `(let ((beg ,beg) (end ,end))
1600 (put-text-property beg end 'c-in-sws t)
1601 ,@(when (facep 'c-debug-is-sws-face)
1602 `((c-debug-add-face beg end 'c-debug-in-sws-face)))))
1603
1604(defmacro c-remove-is-sws (beg end)
0386b551 1605 ;; This macro does a hidden buffer change.
d9e94c22
MS
1606 `(let ((beg ,beg) (end ,end))
1607 (remove-text-properties beg end '(c-is-sws nil))
1608 ,@(when (facep 'c-debug-is-sws-face)
1609 `((c-debug-remove-face beg end 'c-debug-is-sws-face)))))
1610
1611(defmacro c-remove-in-sws (beg end)
0386b551 1612 ;; This macro does a hidden buffer change.
d9e94c22
MS
1613 `(let ((beg ,beg) (end ,end))
1614 (remove-text-properties beg end '(c-in-sws nil))
1615 ,@(when (facep 'c-debug-is-sws-face)
1616 `((c-debug-remove-face beg end 'c-debug-in-sws-face)))))
1617
1618(defmacro c-remove-is-and-in-sws (beg end)
0386b551 1619 ;; This macro does a hidden buffer change.
d9e94c22
MS
1620 `(let ((beg ,beg) (end ,end))
1621 (remove-text-properties beg end '(c-is-sws nil c-in-sws nil))
1622 ,@(when (facep 'c-debug-is-sws-face)
1623 `((c-debug-remove-face beg end 'c-debug-is-sws-face)
1624 (c-debug-remove-face beg end 'c-debug-in-sws-face)))))
1625
1626(defsubst c-invalidate-sws-region-after (beg end)
1627 ;; Called from `after-change-functions'. Note that if
1628 ;; `c-forward-sws' or `c-backward-sws' are used outside
1629 ;; `c-save-buffer-state' or similar then this will remove the cache
1630 ;; properties right after they're added.
0386b551
AM
1631 ;;
1632 ;; This function does hidden buffer changes.
d9e94c22
MS
1633
1634 (save-excursion
1635 ;; Adjust the end to remove the properties in any following simple
1636 ;; ws up to and including the next line break, if there is any
1637 ;; after the changed region. This is necessary e.g. when a rung
1638 ;; marked empty line is converted to a line comment by inserting
1639 ;; "//" before the line break. In that case the line break would
1640 ;; keep the rung mark which could make a later `c-backward-sws'
1641 ;; move into the line comment instead of over it.
1642 (goto-char end)
1643 (skip-chars-forward " \t\f\v")
1644 (when (and (eolp) (not (eobp)))
1645 (setq end (1+ (point)))))
1646
1647 (when (and (= beg end)
1648 (get-text-property beg 'c-in-sws)
2a15eb73 1649 (> beg (point-min))
d9e94c22
MS
1650 (get-text-property (1- beg) 'c-in-sws))
1651 ;; Ensure that an `c-in-sws' range gets broken. Note that it isn't
1652 ;; safe to keep a range that was continuous before the change. E.g:
1653 ;;
1654 ;; #define foo
1655 ;; \
1656 ;; bar
1657 ;;
1658 ;; There can be a "ladder" between "#" and "b". Now, if the newline
1659 ;; after "foo" is removed then "bar" will become part of the cpp
1660 ;; directive instead of a syntactically relevant token. In that
1661 ;; case there's no longer syntactic ws from "#" to "b".
1662 (setq beg (1- beg)))
1663
1664 (c-debug-sws-msg "c-invalidate-sws-region-after [%s..%s]" beg end)
1665 (c-remove-is-and-in-sws beg end))
1666
1667(defun c-forward-sws ()
1668 ;; Used by `c-forward-syntactic-ws' to implement the unbounded search.
0386b551
AM
1669 ;;
1670 ;; This function might do hidden buffer changes.
d9e94c22
MS
1671
1672 (let (;; `rung-pos' is set to a position as early as possible in the
1673 ;; unmarked part of the simple ws region.
1674 (rung-pos (point)) next-rung-pos rung-end-pos last-put-in-sws-pos
1675 rung-is-marked next-rung-is-marked simple-ws-end
1676 ;; `safe-start' is set when it's safe to cache the start position.
1677 ;; It's not set if we've initially skipped over comments and line
1678 ;; continuations since we might have gone out through the end of a
1679 ;; macro then. This provision makes `c-forward-sws' not populate the
1680 ;; cache in the majority of cases, but otoh is `c-backward-sws' by far
1681 ;; more common.
1682 safe-start)
1683
1684 ;; Skip simple ws and do a quick check on the following character to see
1685 ;; if it's anything that can't start syntactic ws, so we can bail out
1686 ;; early in the majority of cases when there just are a few ws chars.
1687 (skip-chars-forward " \t\n\r\f\v")
1688 (when (looking-at c-syntactic-ws-start)
1689
1690 (setq rung-end-pos (min (1+ (point)) (point-max)))
1691 (if (setq rung-is-marked (text-property-any rung-pos rung-end-pos
1692 'c-is-sws t))
1693 ;; Find the last rung position to avoid setting properties in all
1694 ;; the cases when the marked rung is complete.
1695 ;; (`next-single-property-change' is certain to move at least one
1696 ;; step forward.)
1697 (setq rung-pos (1- (next-single-property-change
1698 rung-is-marked 'c-is-sws nil rung-end-pos)))
1699 ;; Got no marked rung here. Since the simple ws might have started
1700 ;; inside a line comment or cpp directive we must set `rung-pos' as
1701 ;; high as possible.
1702 (setq rung-pos (point)))
1703
2143fa32 1704 (with-silent-modifications
d9e94c22
MS
1705 (while
1706 (progn
1707 (while
1708 (when (and rung-is-marked
1709 (get-text-property (point) 'c-in-sws))
1710
1711 ;; The following search is the main reason that `c-in-sws'
1712 ;; and `c-is-sws' aren't combined to one property.
1713 (goto-char (next-single-property-change
1714 (point) 'c-in-sws nil (point-max)))
1715 (unless (get-text-property (point) 'c-is-sws)
1716 ;; If the `c-in-sws' region extended past the last
1717 ;; `c-is-sws' char we have to go back a bit.
1718 (or (get-text-property (1- (point)) 'c-is-sws)
1719 (goto-char (previous-single-property-change
1720 (point) 'c-is-sws)))
1721 (backward-char))
1722
1723 (c-debug-sws-msg
1724 "c-forward-sws cached move %s -> %s (max %s)"
1725 rung-pos (point) (point-max))
1726
1727 (setq rung-pos (point))
1728 (and (> (skip-chars-forward " \t\n\r\f\v") 0)
1729 (not (eobp))))
1730
1731 ;; We'll loop here if there is simple ws after the last rung.
1732 ;; That means that there's been some change in it and it's
1733 ;; possible that we've stepped into another ladder, so extend
1734 ;; the previous one to join with it if there is one, and try to
1735 ;; use the cache again.
1736 (c-debug-sws-msg
1737 "c-forward-sws extending rung with [%s..%s] (max %s)"
1738 (1+ rung-pos) (1+ (point)) (point-max))
1739 (unless (get-text-property (point) 'c-is-sws)
1740 ;; Remove any `c-in-sws' property from the last char of
1741 ;; the rung before we mark it with `c-is-sws', so that we
1742 ;; won't connect with the remains of a broken "ladder".
1743 (c-remove-in-sws (point) (1+ (point))))
1744 (c-put-is-sws (1+ rung-pos)
1745 (1+ (point)))
1746 (c-put-in-sws rung-pos
1747 (setq rung-pos (point)
1748 last-put-in-sws-pos rung-pos)))
1749
1750 (setq simple-ws-end (point))
1751 (c-forward-comments)
1752
1753 (cond
1754 ((/= (point) simple-ws-end)
1755 ;; Skipped over comments. Don't cache at eob in case the buffer
1756 ;; is narrowed.
1757 (not (eobp)))
1758
1759 ((save-excursion
1760 (and c-opt-cpp-prefix
1761 (looking-at c-opt-cpp-start)
1762 (progn (skip-chars-backward " \t")
1763 (bolp))
1764 (or (bobp)
1765 (progn (backward-char)
1766 (not (eq (char-before) ?\\))))))
1767 ;; Skip a preprocessor directive.
1768 (end-of-line)
1769 (while (and (eq (char-before) ?\\)
1770 (= (forward-line 1) 0))
1771 (end-of-line))
1772 (forward-line 1)
1773 (setq safe-start t)
1774 ;; Don't cache at eob in case the buffer is narrowed.
1775 (not (eobp)))))
1776
1777 ;; We've searched over a piece of non-white syntactic ws. See if this
1778 ;; can be cached.
1779 (setq next-rung-pos (point))
1780 (skip-chars-forward " \t\n\r\f\v")
1781 (setq rung-end-pos (min (1+ (point)) (point-max)))
1782
1783 (if (or
1784 ;; Cache if we haven't skipped comments only, and if we started
1785 ;; either from a marked rung or from a completely uncached
1786 ;; position.
1787 (and safe-start
1788 (or rung-is-marked
1789 (not (get-text-property simple-ws-end 'c-in-sws))))
1790
1791 ;; See if there's a marked rung in the encountered simple ws. If
1792 ;; so then we can cache, unless `safe-start' is nil. Even then
1793 ;; we need to do this to check if the cache can be used for the
1794 ;; next step.
1795 (and (setq next-rung-is-marked
1796 (text-property-any next-rung-pos rung-end-pos
1797 'c-is-sws t))
1798 safe-start))
b2acd789 1799
0ec8351b 1800 (progn
d9e94c22
MS
1801 (c-debug-sws-msg
1802 "c-forward-sws caching [%s..%s] - [%s..%s] (max %s)"
1803 rung-pos (1+ simple-ws-end) next-rung-pos rung-end-pos
1804 (point-max))
1805
1806 ;; Remove the properties for any nested ws that might be cached.
1807 ;; Only necessary for `c-is-sws' since `c-in-sws' will be set
1808 ;; anyway.
1809 (c-remove-is-sws (1+ simple-ws-end) next-rung-pos)
1810 (unless (and rung-is-marked (= rung-pos simple-ws-end))
1811 (c-put-is-sws rung-pos
1812 (1+ simple-ws-end))
1813 (setq rung-is-marked t))
1814 (c-put-in-sws rung-pos
1815 (setq rung-pos (point)
1816 last-put-in-sws-pos rung-pos))
1817 (unless (get-text-property (1- rung-end-pos) 'c-is-sws)
1818 ;; Remove any `c-in-sws' property from the last char of
1819 ;; the rung before we mark it with `c-is-sws', so that we
1820 ;; won't connect with the remains of a broken "ladder".
1821 (c-remove-in-sws (1- rung-end-pos) rung-end-pos))
1822 (c-put-is-sws next-rung-pos
1823 rung-end-pos))
1824
1825 (c-debug-sws-msg
1826 "c-forward-sws not caching [%s..%s] - [%s..%s] (max %s)"
1827 rung-pos (1+ simple-ws-end) next-rung-pos rung-end-pos
1828 (point-max))
1829
1830 ;; Set `rung-pos' for the next rung. It's the same thing here as
1831 ;; initially, except that the rung position is set as early as
1832 ;; possible since we can't be in the ending ws of a line comment or
1833 ;; cpp directive now.
1834 (if (setq rung-is-marked next-rung-is-marked)
1835 (setq rung-pos (1- (next-single-property-change
1836 rung-is-marked 'c-is-sws nil rung-end-pos)))
1837 (setq rung-pos next-rung-pos))
1838 (setq safe-start t)))
1839
1840 ;; Make sure that the newly marked `c-in-sws' region doesn't connect to
1841 ;; another one after the point (which might occur when editing inside a
1842 ;; comment or macro).
1843 (when (eq last-put-in-sws-pos (point))
1844 (cond ((< last-put-in-sws-pos (point-max))
1845 (c-debug-sws-msg
1846 "c-forward-sws clearing at %s for cache separation"
1847 last-put-in-sws-pos)
1848 (c-remove-in-sws last-put-in-sws-pos
1849 (1+ last-put-in-sws-pos)))
1850 (t
1851 ;; If at eob we have to clear the last character before the end
1852 ;; instead since the buffer might be narrowed and there might
1853 ;; be a `c-in-sws' after (point-max). In this case it's
1854 ;; necessary to clear both properties.
1855 (c-debug-sws-msg
1856 "c-forward-sws clearing thoroughly at %s for cache separation"
1857 (1- last-put-in-sws-pos))
1858 (c-remove-is-and-in-sws (1- last-put-in-sws-pos)
1859 last-put-in-sws-pos))))
2143fa32 1860 ))))
b2acd789 1861
d9e94c22
MS
1862(defun c-backward-sws ()
1863 ;; Used by `c-backward-syntactic-ws' to implement the unbounded search.
0386b551
AM
1864 ;;
1865 ;; This function might do hidden buffer changes.
d9e94c22
MS
1866
1867 (let (;; `rung-pos' is set to a position as late as possible in the unmarked
1868 ;; part of the simple ws region.
1869 (rung-pos (point)) next-rung-pos last-put-in-sws-pos
1870 rung-is-marked simple-ws-beg cmt-skip-pos)
1871
1872 ;; Skip simple horizontal ws and do a quick check on the preceding
91af3942 1873 ;; character to see if it's anything that can't end syntactic ws, so we can
d9e94c22
MS
1874 ;; bail out early in the majority of cases when there just are a few ws
1875 ;; chars. Newlines are complicated in the backward direction, so we can't
1876 ;; skip over them.
1877 (skip-chars-backward " \t\f")
1878 (when (and (not (bobp))
1879 (save-excursion
1880 (backward-char)
1881 (looking-at c-syntactic-ws-end)))
1882
1883 ;; Try to find a rung position in the simple ws preceding point, so that
1884 ;; we can get a cache hit even if the last bit of the simple ws has
1885 ;; changed recently.
1886 (setq simple-ws-beg (point))
1887 (skip-chars-backward " \t\n\r\f\v")
1888 (if (setq rung-is-marked (text-property-any
1889 (point) (min (1+ rung-pos) (point-max))
1890 'c-is-sws t))
1891 ;; `rung-pos' will be the earliest marked position, which means that
1892 ;; there might be later unmarked parts in the simple ws region.
1893 ;; It's not worth the effort to fix that; the last part of the
1894 ;; simple ws is also typically edited often, so it could be wasted.
1895 (goto-char (setq rung-pos rung-is-marked))
1896 (goto-char simple-ws-beg))
1897
2143fa32 1898 (with-silent-modifications
d9e94c22
MS
1899 (while
1900 (progn
1901 (while
1902 (when (and rung-is-marked
1903 (not (bobp))
1904 (get-text-property (1- (point)) 'c-in-sws))
1905
1906 ;; The following search is the main reason that `c-in-sws'
1907 ;; and `c-is-sws' aren't combined to one property.
1908 (goto-char (previous-single-property-change
1909 (point) 'c-in-sws nil (point-min)))
1910 (unless (get-text-property (point) 'c-is-sws)
1911 ;; If the `c-in-sws' region extended past the first
1912 ;; `c-is-sws' char we have to go forward a bit.
1913 (goto-char (next-single-property-change
1914 (point) 'c-is-sws)))
1915
1916 (c-debug-sws-msg
1917 "c-backward-sws cached move %s <- %s (min %s)"
1918 (point) rung-pos (point-min))
1919
1920 (setq rung-pos (point))
1921 (if (and (< (min (skip-chars-backward " \t\f\v")
1922 (progn
1923 (setq simple-ws-beg (point))
1924 (skip-chars-backward " \t\n\r\f\v")))
1925 0)
1926 (setq rung-is-marked
1927 (text-property-any (point) rung-pos
1928 'c-is-sws t)))
1929 t
1930 (goto-char simple-ws-beg)
1931 nil))
1932
1933 ;; We'll loop here if there is simple ws before the first rung.
1934 ;; That means that there's been some change in it and it's
1935 ;; possible that we've stepped into another ladder, so extend
1936 ;; the previous one to join with it if there is one, and try to
1937 ;; use the cache again.
1938 (c-debug-sws-msg
1939 "c-backward-sws extending rung with [%s..%s] (min %s)"
1940 rung-is-marked rung-pos (point-min))
1941 (unless (get-text-property (1- rung-pos) 'c-is-sws)
1942 ;; Remove any `c-in-sws' property from the last char of
1943 ;; the rung before we mark it with `c-is-sws', so that we
1944 ;; won't connect with the remains of a broken "ladder".
1945 (c-remove-in-sws (1- rung-pos) rung-pos))
1946 (c-put-is-sws rung-is-marked
1947 rung-pos)
1948 (c-put-in-sws rung-is-marked
1949 (1- rung-pos))
1950 (setq rung-pos rung-is-marked
1951 last-put-in-sws-pos rung-pos))
1952
1953 (c-backward-comments)
1954 (setq cmt-skip-pos (point))
a66cd3ee 1955
d9e94c22
MS
1956 (cond
1957 ((and c-opt-cpp-prefix
1958 (/= cmt-skip-pos simple-ws-beg)
1959 (c-beginning-of-macro))
1960 ;; Inside a cpp directive. See if it should be skipped over.
1961 (let ((cpp-beg (point)))
1962
1963 ;; Move back over all line continuations in the region skipped
1964 ;; over by `c-backward-comments'. If we go past it then we
1965 ;; started inside the cpp directive.
1966 (goto-char simple-ws-beg)
1967 (beginning-of-line)
1968 (while (and (> (point) cmt-skip-pos)
1969 (progn (backward-char)
1970 (eq (char-before) ?\\)))
1971 (beginning-of-line))
1972
1973 (if (< (point) cmt-skip-pos)
1974 ;; Don't move past the cpp directive if we began inside
1975 ;; it. Note that the position at the end of the last line
1976 ;; of the macro is also considered to be within it.
1977 (progn (goto-char cmt-skip-pos)
1978 nil)
1979
1980 ;; It's worthwhile to spend a little bit of effort on finding
1981 ;; the end of the macro, to get a good `simple-ws-beg'
1982 ;; position for the cache. Note that `c-backward-comments'
1983 ;; could have stepped over some comments before going into
1984 ;; the macro, and then `simple-ws-beg' must be kept on the
1985 ;; same side of those comments.
1986 (goto-char simple-ws-beg)
1987 (skip-chars-backward " \t\n\r\f\v")
1988 (if (eq (char-before) ?\\)
1989 (forward-char))
1990 (forward-line 1)
1991 (if (< (point) simple-ws-beg)
1992 ;; Might happen if comments after the macro were skipped
1993 ;; over.
1994 (setq simple-ws-beg (point)))
1995
1996 (goto-char cpp-beg)
1997 t)))
1998
1999 ((/= (save-excursion
2000 (skip-chars-forward " \t\n\r\f\v" simple-ws-beg)
2001 (setq next-rung-pos (point)))
2002 simple-ws-beg)
2003 ;; Skipped over comments. Must put point at the end of
2004 ;; the simple ws at point since we might be after a line
2005 ;; comment or cpp directive that's been partially
2006 ;; narrowed out, and we can't risk marking the simple ws
2007 ;; at the end of it.
2008 (goto-char next-rung-pos)
2009 t)))
2010
2011 ;; We've searched over a piece of non-white syntactic ws. See if this
2012 ;; can be cached.
2013 (setq next-rung-pos (point))
2014 (skip-chars-backward " \t\f\v")
2015
2016 (if (or
2017 ;; Cache if we started either from a marked rung or from a
2018 ;; completely uncached position.
2019 rung-is-marked
2020 (not (get-text-property (1- simple-ws-beg) 'c-in-sws))
2021
2022 ;; Cache if there's a marked rung in the encountered simple ws.
2023 (save-excursion
2024 (skip-chars-backward " \t\n\r\f\v")
2025 (text-property-any (point) (min (1+ next-rung-pos) (point-max))
2026 'c-is-sws t)))
a66cd3ee 2027
d9e94c22
MS
2028 (progn
2029 (c-debug-sws-msg
2030 "c-backward-sws caching [%s..%s] - [%s..%s] (min %s)"
2031 (point) (1+ next-rung-pos)
2032 simple-ws-beg (min (1+ rung-pos) (point-max))
2033 (point-min))
2034
2035 ;; Remove the properties for any nested ws that might be cached.
2036 ;; Only necessary for `c-is-sws' since `c-in-sws' will be set
2037 ;; anyway.
2038 (c-remove-is-sws (1+ next-rung-pos) simple-ws-beg)
2039 (unless (and rung-is-marked (= simple-ws-beg rung-pos))
2040 (let ((rung-end-pos (min (1+ rung-pos) (point-max))))
2041 (unless (get-text-property (1- rung-end-pos) 'c-is-sws)
2042 ;; Remove any `c-in-sws' property from the last char of
2043 ;; the rung before we mark it with `c-is-sws', so that we
2044 ;; won't connect with the remains of a broken "ladder".
2045 (c-remove-in-sws (1- rung-end-pos) rung-end-pos))
2046 (c-put-is-sws simple-ws-beg
2047 rung-end-pos)
2048 (setq rung-is-marked t)))
2049 (c-put-in-sws (setq simple-ws-beg (point)
2050 last-put-in-sws-pos simple-ws-beg)
2051 rung-pos)
2052 (c-put-is-sws (setq rung-pos simple-ws-beg)
2053 (1+ next-rung-pos)))
2054
2055 (c-debug-sws-msg
2056 "c-backward-sws not caching [%s..%s] - [%s..%s] (min %s)"
2057 (point) (1+ next-rung-pos)
2058 simple-ws-beg (min (1+ rung-pos) (point-max))
2059 (point-min))
2060 (setq rung-pos next-rung-pos
2061 simple-ws-beg (point))
2062 ))
2063
2064 ;; Make sure that the newly marked `c-in-sws' region doesn't connect to
2065 ;; another one before the point (which might occur when editing inside a
2066 ;; comment or macro).
2067 (when (eq last-put-in-sws-pos (point))
2068 (cond ((< (point-min) last-put-in-sws-pos)
2069 (c-debug-sws-msg
2070 "c-backward-sws clearing at %s for cache separation"
2071 (1- last-put-in-sws-pos))
2072 (c-remove-in-sws (1- last-put-in-sws-pos)
2073 last-put-in-sws-pos))
2074 ((> (point-min) 1)
2075 ;; If at bob and the buffer is narrowed, we have to clear the
2076 ;; character we're standing on instead since there might be a
2077 ;; `c-in-sws' before (point-min). In this case it's necessary
2078 ;; to clear both properties.
2079 (c-debug-sws-msg
2080 "c-backward-sws clearing thoroughly at %s for cache separation"
2081 last-put-in-sws-pos)
2082 (c-remove-is-and-in-sws last-put-in-sws-pos
2083 (1+ last-put-in-sws-pos)))))
2143fa32 2084 ))))
785eecbb 2085
d9e94c22 2086\f
580fba94
AM
2087;; Other whitespace tools
2088(defun c-partial-ws-p (beg end)
2089 ;; Is the region (beg end) WS, and is there WS (or BOB/EOB) next to the
2090 ;; region? This is a "heuristic" function. .....
b414f371 2091 ;;
c421028a
AM
2092 ;; The motivation for the second bit is to check whether removing this
2093 ;; region would coalesce two symbols.
580fba94
AM
2094 ;;
2095 ;; FIXME!!! This function doesn't check virtual semicolons in any way. Be
2096 ;; careful about using this function for, e.g. AWK. (2007/3/7)
2097 (save-excursion
2098 (let ((end+1 (min (1+ end) (point-max))))
2099 (or (progn (goto-char (max (point-min) (1- beg)))
2100 (c-skip-ws-forward end)
2101 (eq (point) end))
2102 (progn (goto-char beg)
2103 (c-skip-ws-forward end+1)
2104 (eq (point) end+1))))))
2105\f
0386b551 2106;; A system for finding noteworthy parens before the point.
e1c458ae 2107
0ec1d2c5
AM
2108(defconst c-state-cache-too-far 5000)
2109;; A maximum comfortable scanning distance, e.g. between
2110;; `c-state-cache-good-pos' and "HERE" (where we call c-parse-state). When
0d26e0b6 2111;; this distance is exceeded, we take "emergency measures", e.g. by clearing
0ec1d2c5
AM
2112;; the cache and starting again from point-min or a beginning of defun. This
2113;; value can be tuned for efficiency or set to a lower value for testing.
2114
d9e94c22
MS
2115(defvar c-state-cache nil)
2116(make-variable-buffer-local 'c-state-cache)
2117;; The state cache used by `c-parse-state' to cut down the amount of
0ec1d2c5
AM
2118;; searching. It's the result from some earlier `c-parse-state' call. See
2119;; `c-parse-state''s doc string for details of its structure.
0386b551 2120;;
d9e94c22
MS
2121;; The use of the cached info is more effective if the next
2122;; `c-parse-state' call is on a line close by the one the cached state
2123;; was made at; the cache can actually slow down a little if the
2124;; cached state was made very far back in the buffer. The cache is
2125;; most effective if `c-parse-state' is used on each line while moving
2126;; forward.
e1c458ae 2127
0386b551
AM
2128(defvar c-state-cache-good-pos 1)
2129(make-variable-buffer-local 'c-state-cache-good-pos)
0ec1d2c5
AM
2130;; This is a position where `c-state-cache' is known to be correct, or
2131;; nil (see below). It's a position inside one of the recorded unclosed
2132;; parens or the top level, but not further nested inside any literal or
2133;; subparen that is closed before the last recorded position.
0386b551
AM
2134;;
2135;; The exact position is chosen to try to be close to yet earlier than
2136;; the position where `c-state-cache' will be called next. Right now
2137;; the heuristic is to set it to the position after the last found
2138;; closing paren (of any type) before the line on which
2139;; `c-parse-state' was called. That is chosen primarily to work well
2140;; with refontification of the current line.
0ec1d2c5
AM
2141;;
2142;; 2009-07-28: When `c-state-point-min' and the last position where
2143;; `c-parse-state' or for which `c-invalidate-state-cache' was called, are
2144;; both in the same literal, there is no such "good position", and
2145;; c-state-cache-good-pos is then nil. This is the ONLY circumstance in which
2146;; it can be nil. In this case, `c-state-point-min-literal' will be non-nil.
2147;;
2148;; 2009-06-12: In a brace desert, c-state-cache-good-pos may also be in
2149;; the middle of the desert, as long as it is not within a brace pair
2150;; recorded in `c-state-cache' or a paren/bracket pair.
2151
2152
2153;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2154;; We maintain a simple cache of positions which aren't in a literal, so as to
2155;; speed up testing for non-literality.
9657183b 2156(defconst c-state-nonlit-pos-interval 3000)
0ec1d2c5
AM
2157;; The approximate interval between entries in `c-state-nonlit-pos-cache'.
2158
2159(defvar c-state-nonlit-pos-cache nil)
2160(make-variable-buffer-local 'c-state-nonlit-pos-cache)
e84efb70
AM
2161;; A list of buffer positions which are known not to be in a literal or a cpp
2162;; construct. This is ordered with higher positions at the front of the list.
2163;; Only those which are less than `c-state-nonlit-pos-cache-limit' are valid.
0ec1d2c5
AM
2164
2165(defvar c-state-nonlit-pos-cache-limit 1)
2166(make-variable-buffer-local 'c-state-nonlit-pos-cache-limit)
2167;; An upper limit on valid entries in `c-state-nonlit-pos-cache'. This is
2168;; reduced by buffer changes, and increased by invocations of
2169;; `c-state-literal-at'.
2170
56d093a9
AM
2171(defvar c-state-semi-nonlit-pos-cache nil)
2172(make-variable-buffer-local 'c-state-semi-nonlit-pos-cache)
2173;; A list of buffer positions which are known not to be in a literal. This is
2174;; ordered with higher positions at the front of the list. Only those which
2175;; are less than `c-state-semi-nonlit-pos-cache-limit' are valid.
2176
2177(defvar c-state-semi-nonlit-pos-cache-limit 1)
2178(make-variable-buffer-local 'c-state-semi-nonlit-pos-cache-limit)
2179;; An upper limit on valid entries in `c-state-semi-nonlit-pos-cache'. This is
2180;; reduced by buffer changes, and increased by invocations of
f003f294 2181;; `c-state-literal-at'. FIXME!!!
56d093a9 2182
0ec1d2c5 2183(defsubst c-state-pp-to-literal (from to)
13d49cbb
AM
2184 ;; Do a parse-partial-sexp from FROM to TO, returning either
2185 ;; (STATE TYPE (BEG . END)) if TO is in a literal; or
2186 ;; (STATE) otherwise,
2187 ;; where STATE is the parsing state at TO, TYPE is the type of the literal
2188 ;; (one of 'c, 'c++, 'string) and (BEG . END) is the boundaries of the literal.
2189 ;;
2190 ;; Only elements 3 (in a string), 4 (in a comment), 5 (following a quote),
2191 ;; 7 (comment type) and 8 (start of comment/string) (and possibly 9) of
2192 ;; STATE are valid.
0ec1d2c5 2193 (save-excursion
13d49cbb
AM
2194 (let ((s (parse-partial-sexp from to))
2195 ty)
0ec1d2c5 2196 (when (or (nth 3 s) (nth 4 s)) ; in a string or comment
13d49cbb
AM
2197 (setq ty (cond
2198 ((nth 3 s) 'string)
2199 ((eq (nth 7 s) t) 'c++)
2200 (t 'c)))
0ec1d2c5
AM
2201 (parse-partial-sexp (point) (point-max)
2202 nil ; TARGETDEPTH
2203 nil ; STOPBEFORE
2204 s ; OLDSTATE
13d49cbb
AM
2205 'syntax-table)) ; stop at end of literal
2206 (if ty
2207 `(,s ,ty (,(nth 8 s) . ,(point)))
2208 `(,s)))))
2209
2210(defun c-state-safe-place (here)
2211 ;; Return a buffer position before HERE which is "safe", i.e. outside any
2212 ;; string, comment, or macro.
e84efb70
AM
2213 ;;
2214 ;; NOTE: This function manipulates `c-state-nonlit-pos-cache'. This cache
2215 ;; MAY NOT contain any positions within macros, since macros are frequently
2216 ;; turned into comments by use of the `c-cpp-delimiter' category properties.
2217 ;; We cannot rely on this mechanism whilst determining a cache pos since
2218 ;; this function is also called from outwith `c-parse-state'.
0ec1d2c5
AM
2219 (save-restriction
2220 (widen)
2221 (save-excursion
2222 (let ((c c-state-nonlit-pos-cache)
56d093a9 2223 pos npos high-pos lit macro-beg macro-end)
0ec1d2c5
AM
2224 ;; Trim the cache to take account of buffer changes.
2225 (while (and c (> (car c) c-state-nonlit-pos-cache-limit))
2226 (setq c (cdr c)))
2227 (setq c-state-nonlit-pos-cache c)
2228
2229 (while (and c (> (car c) here))
56d093a9 2230 (setq high-pos (car c))
0ec1d2c5
AM
2231 (setq c (cdr c)))
2232 (setq pos (or (car c) (point-min)))
2233
56d093a9
AM
2234 (unless high-pos
2235 (while
2236 ;; Add an element to `c-state-nonlit-pos-cache' each iteration.
2237 (and
2238 (<= (setq npos (+ pos c-state-nonlit-pos-interval)) here)
2239
2240 ;; Test for being in a literal. If so, go to after it.
2241 (progn
2242 (setq lit (car (cddr (c-state-pp-to-literal pos npos))))
2243 (or (null lit)
2244 (prog1 (<= (cdr lit) here)
2245 (setq npos (cdr lit)))))
2246
2247 ;; Test for being in a macro. If so, go to after it.
2248 (progn
2249 (goto-char npos)
2250 (setq macro-beg
2251 (and (c-beginning-of-macro) (/= (point) npos) (point)))
2252 (when macro-beg
2253 (c-syntactic-end-of-macro)
2254 (or (eobp) (forward-char))
2255 (setq macro-end (point)))
2256 (or (null macro-beg)
2257 (prog1 (<= macro-end here)
2258 (setq npos macro-end)))))
2259
2260 (setq pos npos)
2261 (setq c-state-nonlit-pos-cache (cons pos c-state-nonlit-pos-cache)))
2262 ;; Add one extra element above HERE so as to to avoid the previous
2263 ;; expensive calculation when the next call is close to the current
2264 ;; one. This is especially useful when inside a large macro.
2265 (setq c-state-nonlit-pos-cache (cons npos c-state-nonlit-pos-cache)))
0ec1d2c5
AM
2266
2267 (if (> pos c-state-nonlit-pos-cache-limit)
2268 (setq c-state-nonlit-pos-cache-limit pos))
13d49cbb 2269 pos))))
9858f6c3 2270
56d093a9
AM
2271(defun c-state-semi-safe-place (here)
2272 ;; Return a buffer position before HERE which is "safe", i.e. outside any
2273 ;; string or comment. It may be in a macro.
2274 (save-restriction
2275 (widen)
2276 (save-excursion
2277 (let ((c c-state-semi-nonlit-pos-cache)
2278 pos npos high-pos lit macro-beg macro-end)
2279 ;; Trim the cache to take account of buffer changes.
2280 (while (and c (> (car c) c-state-semi-nonlit-pos-cache-limit))
2281 (setq c (cdr c)))
2282 (setq c-state-semi-nonlit-pos-cache c)
2283
2284 (while (and c (> (car c) here))
2285 (setq high-pos (car c))
2286 (setq c (cdr c)))
2287 (setq pos (or (car c) (point-min)))
2288
2289 (unless high-pos
2290 (while
2291 ;; Add an element to `c-state-semi-nonlit-pos-cache' each iteration.
2292 (and
2293 (<= (setq npos (+ pos c-state-nonlit-pos-interval)) here)
2294
2295 ;; Test for being in a literal. If so, go to after it.
2296 (progn
2297 (setq lit (car (cddr (c-state-pp-to-literal pos npos))))
2298 (or (null lit)
2299 (prog1 (<= (cdr lit) here)
2300 (setq npos (cdr lit))))))
2301
2302 (setq pos npos)
2303 (setq c-state-semi-nonlit-pos-cache
2304 (cons pos c-state-semi-nonlit-pos-cache))))
2305
2306 (if (> pos c-state-semi-nonlit-pos-cache-limit)
2307 (setq c-state-semi-nonlit-pos-cache-limit pos))
2308 pos))))
2309
13d49cbb
AM
2310(defun c-state-literal-at (here)
2311 ;; If position HERE is inside a literal, return (START . END), the
2312 ;; boundaries of the literal (which may be outside the accessible bit of the
2313 ;; buffer). Otherwise, return nil.
2314 ;;
2315 ;; This function is almost the same as `c-literal-limits'. Previously, it
9858f6c3 2316 ;; differed in that it was a lower level function, and that it rigorously
13d49cbb
AM
2317 ;; followed the syntax from BOB. `c-literal-limits' is now (2011-12)
2318 ;; virtually identical to this function.
2319 (save-restriction
2320 (widen)
2321 (save-excursion
2322 (let ((pos (c-state-safe-place here)))
2323 (car (cddr (c-state-pp-to-literal pos here)))))))
0ec1d2c5
AM
2324
2325(defsubst c-state-lit-beg (pos)
2326 ;; Return the start of the literal containing POS, or POS itself.
2327 (or (car (c-state-literal-at pos))
2328 pos))
2329
2330(defsubst c-state-cache-non-literal-place (pos state)
9c4aeabf 2331 ;; Return a position outside of a string/comment/macro at or before POS.
0ec1d2c5 2332 ;; STATE is the parse-partial-sexp state at POS.
9c4aeabf
AM
2333 (let ((res (if (or (nth 3 state) ; in a string?
2334 (nth 4 state)) ; in a comment?
2335 (nth 8 state)
2336 pos)))
2337 (save-excursion
2338 (goto-char res)
2339 (if (c-beginning-of-macro)
2340 (point)
2341 res))))
0ec1d2c5
AM
2342
2343;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2344;; Stuff to do with point-min, and coping with any literal there.
2345(defvar c-state-point-min 1)
2346(make-variable-buffer-local 'c-state-point-min)
2347;; This is (point-min) when `c-state-cache' was last calculated. A change of
2348;; narrowing is likely to affect the parens that are visible before the point.
2349
2350(defvar c-state-point-min-lit-type nil)
2351(make-variable-buffer-local 'c-state-point-min-lit-type)
2352(defvar c-state-point-min-lit-start nil)
2353(make-variable-buffer-local 'c-state-point-min-lit-start)
2354;; These two variables define the literal, if any, containing point-min.
2355;; Their values are, respectively, 'string, c, or c++, and the start of the
2356;; literal. If there's no literal there, they're both nil.
2357
2358(defvar c-state-min-scan-pos 1)
2359(make-variable-buffer-local 'c-state-min-scan-pos)
2360;; This is the earliest buffer-pos from which scanning can be done. It is
2361;; either the end of the literal containing point-min, or point-min itself.
2362;; It becomes nil if the buffer is changed earlier than this point.
2363(defun c-state-get-min-scan-pos ()
2364 ;; Return the lowest valid scanning pos. This will be the end of the
2365 ;; literal enclosing point-min, or point-min itself.
2366 (or c-state-min-scan-pos
2367 (save-restriction
2368 (save-excursion
2369 (widen)
2370 (goto-char c-state-point-min-lit-start)
2371 (if (eq c-state-point-min-lit-type 'string)
2372 (forward-sexp)
2373 (forward-comment 1))
2374 (setq c-state-min-scan-pos (point))))))
2375
2376(defun c-state-mark-point-min-literal ()
2377 ;; Determine the properties of any literal containing POINT-MIN, setting the
2378 ;; variables `c-state-point-min-lit-type', `c-state-point-min-lit-start',
2379 ;; and `c-state-min-scan-pos' accordingly. The return value is meaningless.
2380 (let ((p-min (point-min))
2381 lit)
2382 (save-restriction
2383 (widen)
2384 (setq lit (c-state-literal-at p-min))
2385 (if lit
2386 (setq c-state-point-min-lit-type
2387 (save-excursion
2388 (goto-char (car lit))
2389 (cond
2390 ((looking-at c-block-comment-start-regexp) 'c)
2391 ((looking-at c-line-comment-starter) 'c++)
2392 (t 'string)))
2393 c-state-point-min-lit-start (car lit)
2394 c-state-min-scan-pos (cdr lit))
2395 (setq c-state-point-min-lit-type nil
2396 c-state-point-min-lit-start nil
2397 c-state-min-scan-pos p-min)))))
2398
2399
2400;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2401;; A variable which signals a brace dessert - helpful for reducing the number
2402;; of fruitless backward scans.
2403(defvar c-state-brace-pair-desert nil)
2404(make-variable-buffer-local 'c-state-brace-pair-desert)
e84efb70 2405;; Used only in `c-append-lower-brace-pair-to-state-cache'. It is set when
0ec1d2c5
AM
2406;; that defun has searched backwards for a brace pair and not found one. Its
2407;; value is either nil or a cons (PA . FROM), where PA is the position of the
2408;; enclosing opening paren/brace/bracket which bounds the backwards search (or
2409;; nil when at top level) and FROM is where the backward search started. It
2410;; is reset to nil in `c-invalidate-state-cache'.
2411
2412
2413;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2414;; Lowish level functions/macros which work directly on `c-state-cache', or a
2415;; list of like structure.
2416(defmacro c-state-cache-top-lparen (&optional cache)
2417 ;; Return the address of the top left brace/bracket/paren recorded in CACHE
2418 ;; (default `c-state-cache') (or nil).
2419 (let ((cash (or cache 'c-state-cache)))
2420 `(if (consp (car ,cash))
2421 (caar ,cash)
2422 (car ,cash))))
2423
2424(defmacro c-state-cache-top-paren (&optional cache)
2425 ;; Return the address of the latest brace/bracket/paren (whether left or
2426 ;; right) recorded in CACHE (default `c-state-cache') or nil.
2427 (let ((cash (or cache 'c-state-cache)))
2428 `(if (consp (car ,cash))
2429 (cdar ,cash)
2430 (car ,cash))))
2431
2432(defmacro c-state-cache-after-top-paren (&optional cache)
2433 ;; Return the position just after the latest brace/bracket/paren (whether
2434 ;; left or right) recorded in CACHE (default `c-state-cache') or nil.
2435 (let ((cash (or cache 'c-state-cache)))
2436 `(if (consp (car ,cash))
2437 (cdar ,cash)
2438 (and (car ,cash)
2439 (1+ (car ,cash))))))
2440
2441(defun c-get-cache-scan-pos (here)
2442 ;; From the state-cache, determine the buffer position from which we might
2443 ;; scan forward to HERE to update this cache. This position will be just
2444 ;; after a paren/brace/bracket recorded in the cache, if possible, otherwise
2445 ;; return the earliest position in the accessible region which isn't within
2446 ;; a literal. If the visible portion of the buffer is entirely within a
2447 ;; literal, return NIL.
2448 (let ((c c-state-cache) elt)
2449 ;(while (>= (or (c-state-cache-top-lparen c) 1) here)
2450 (while (and c
2451 (>= (c-state-cache-top-lparen c) here))
2452 (setq c (cdr c)))
2453
2454 (setq elt (car c))
2455 (cond
2456 ((consp elt)
2457 (if (> (cdr elt) here)
2458 (1+ (car elt))
2459 (cdr elt)))
2460 (elt (1+ elt))
2461 ((<= (c-state-get-min-scan-pos) here)
2462 (c-state-get-min-scan-pos))
2463 (t nil))))
2464
2465;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2466;; Variables which keep track of preprocessor constructs.
8cff5feb
AM
2467(defvar c-state-old-cpp-beg-marker nil)
2468(make-variable-buffer-local 'c-state-old-cpp-beg-marker)
0ec1d2c5
AM
2469(defvar c-state-old-cpp-beg nil)
2470(make-variable-buffer-local 'c-state-old-cpp-beg)
8cff5feb
AM
2471(defvar c-state-old-cpp-end-marker nil)
2472(make-variable-buffer-local 'c-state-old-cpp-end-marker)
0ec1d2c5
AM
2473(defvar c-state-old-cpp-end nil)
2474(make-variable-buffer-local 'c-state-old-cpp-end)
2475;; These are the limits of the macro containing point at the previous call of
2476;; `c-parse-state', or nil.
2477
2478;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
c7015153 2479;; Defuns which analyze the buffer, yet don't change `c-state-cache'.
c0209c2c
AM
2480(defun c-state-balance-parens-backwards (here- here+ top)
2481 ;; Return the position of the opening paren/brace/bracket before HERE- which
2482 ;; matches the outermost close p/b/b between HERE+ and TOP. Except when
2483 ;; there's a macro, HERE- and HERE+ are the same. Like this:
0ec1d2c5 2484 ;;
c0209c2c
AM
2485 ;; ............................................
2486 ;; | |
2487 ;; ( [ ( .........#macro.. ) ( ) ] )
2488 ;; ^ ^ ^ ^
2489 ;; | | | |
2490 ;; return HERE- HERE+ TOP
0ec1d2c5
AM
2491 ;;
2492 ;; If there aren't enough opening paren/brace/brackets, return the position
c0209c2c 2493 ;; of the outermost one found, or HERE- if there are none. If there are no
da6062e6 2494 ;; closing p/b/bs between HERE+ and TOP, return HERE-. HERE-/+ and TOP
c0209c2c
AM
2495 ;; must not be inside literals. Only the accessible portion of the buffer
2496 ;; will be scanned.
2497
2498 ;; PART 1: scan from `here+' up to `top', accumulating ")"s which enclose
2499 ;; `here'. Go round the next loop each time we pass over such a ")". These
2500 ;; probably match "("s before `here-'.
0ec1d2c5
AM
2501 (let (pos pa ren+1 lonely-rens)
2502 (save-excursion
2503 (save-restriction
2504 (narrow-to-region (point-min) top) ; This can move point, sometimes.
c0209c2c 2505 (setq pos here+)
0ec1d2c5
AM
2506 (c-safe
2507 (while
2508 (setq ren+1 (scan-lists pos 1 1)) ; might signal
2509 (setq lonely-rens (cons ren+1 lonely-rens)
2510 pos ren+1)))))
2511
c0209c2c 2512 ;; PART 2: Scan back before `here-' searching for the "("s
0ec1d2c5
AM
2513 ;; matching/mismatching the ")"s found above. We only need to direct the
2514 ;; caller to scan when we've encountered unmatched right parens.
c0209c2c
AM
2515 (setq pos here-)
2516 (when lonely-rens
2517 (c-safe
2518 (while
2519 (and lonely-rens ; actual values aren't used.
2520 (setq pa (scan-lists pos -1 1)))
2521 (setq pos pa)
2522 (setq lonely-rens (cdr lonely-rens)))))
2523 pos))
0ec1d2c5
AM
2524
2525(defun c-parse-state-get-strategy (here good-pos)
2526 ;; Determine the scanning strategy for adjusting `c-parse-state', attempting
53964682 2527 ;; to minimize the amount of scanning. HERE is the pertinent position in
0ec1d2c5
AM
2528 ;; the buffer, GOOD-POS is a position where `c-state-cache' (possibly with
2529 ;; its head trimmed) is known to be good, or nil if there is no such
2530 ;; position.
2531 ;;
2532 ;; The return value is a list, one of the following:
2533 ;;
267eacb5
AM
2534 ;; o - ('forward START-POINT) - scan forward from START-POINT,
2535 ;; which is not less than the highest position in `c-state-cache' below here.
2536 ;; o - ('backward nil) - scan backwards (from HERE).
267eacb5 2537 ;; o - ('IN-LIT nil) - point is inside the literal containing point-min.
0ec1d2c5 2538 (let ((cache-pos (c-get-cache-scan-pos here)) ; highest position below HERE in cache (or 1)
d23b0804
AM
2539 strategy ; 'forward, 'backward, or 'IN-LIT.
2540 start-point)
0ec1d2c5
AM
2541 (setq good-pos (or good-pos (c-state-get-min-scan-pos)))
2542 (cond
2543 ((< here (c-state-get-min-scan-pos))
d23b0804 2544 (setq strategy 'IN-LIT))
0ec1d2c5
AM
2545 ((<= good-pos here)
2546 (setq strategy 'forward
d23b0804 2547 start-point (max good-pos cache-pos)))
0ec1d2c5 2548 ((< (- good-pos here) (- here cache-pos)) ; FIXME!!! ; apply some sort of weighting.
d23b0804 2549 (setq strategy 'backward))
0ec1d2c5
AM
2550 (t
2551 (setq strategy 'forward
d23b0804
AM
2552 start-point cache-pos)))
2553 (list strategy (and (eq strategy 'forward) start-point))))
0ec1d2c5
AM
2554
2555
2556;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
2557;; Routines which change `c-state-cache' and associated values.
2558(defun c-renarrow-state-cache ()
2559 ;; The region (more precisely, point-min) has changed since we
2560 ;; calculated `c-state-cache'. Amend `c-state-cache' accordingly.
2561 (if (< (point-min) c-state-point-min)
2562 ;; If point-min has MOVED BACKWARDS then we drop the state completely.
2563 ;; It would be possible to do a better job here and recalculate the top
2564 ;; only.
2565 (progn
2566 (c-state-mark-point-min-literal)
2567 (setq c-state-cache nil
2568 c-state-cache-good-pos c-state-min-scan-pos
2569 c-state-brace-pair-desert nil))
2570
2571 ;; point-min has MOVED FORWARD.
2572
2573 ;; Is the new point-min inside a (different) literal?
2574 (unless (and c-state-point-min-lit-start ; at prev. point-min
2575 (< (point-min) (c-state-get-min-scan-pos)))
2576 (c-state-mark-point-min-literal))
2577
2578 ;; Cut off a bit of the tail from `c-state-cache'.
2579 (let ((ptr (cons nil c-state-cache))
2580 pa)
2581 (while (and (setq pa (c-state-cache-top-lparen (cdr ptr)))
2582 (>= pa (point-min)))
2583 (setq ptr (cdr ptr)))
2584
2585 (when (consp ptr)
2586 (if (eq (cdr ptr) c-state-cache)
2587 (setq c-state-cache nil
2588 c-state-cache-good-pos c-state-min-scan-pos)
2589 (setcdr ptr nil)
2590 (setq c-state-cache-good-pos (1+ (c-state-cache-top-lparen))))
2591 )))
2592
2593 (setq c-state-point-min (point-min)))
2594
d23b0804 2595(defun c-append-lower-brace-pair-to-state-cache (from here &optional upper-lim)
583e23bd
AM
2596 ;; If there is a brace pair preceding FROM in the buffer, at the same level
2597 ;; of nesting (not necessarily immediately preceding), push a cons onto
2598 ;; `c-state-cache' to represent it. FROM must not be inside a literal. If
2599 ;; UPPER-LIM is non-nil, we append the highest brace pair whose "}" is below
2600 ;; UPPER-LIM.
0ec1d2c5
AM
2601 ;;
2602 ;; Return non-nil when this has been done.
2603 ;;
583e23bd
AM
2604 ;; The situation it copes with is this transformation:
2605 ;;
2606 ;; OLD: { (.) {...........}
2607 ;; ^ ^
2608 ;; FROM HERE
2609 ;;
2610 ;; NEW: { {....} (.) {.........
2611 ;; ^ ^ ^
2612 ;; LOWER BRACE PAIR HERE or HERE
2613 ;;
0ec1d2c5
AM
2614 ;; This routine should be fast. Since it can get called a LOT, we maintain
2615 ;; `c-state-brace-pair-desert', a small cache of "failures", such that we
2616 ;; reduce the time wasted in repeated fruitless searches in brace deserts.
2617 (save-excursion
2618 (save-restriction
d23b0804 2619 (let* (new-cons
8cff5feb
AM
2620 (cache-pos (c-state-cache-top-lparen)) ; might be nil.
2621 (macro-start-or-from
2622 (progn (goto-char from)
2623 (c-beginning-of-macro)
2624 (point)))
2625 (bra ; Position of "{".
2626 ;; Don't start scanning in the middle of a CPP construct unless
2627 ;; it contains HERE - these constructs, in Emacs, are "commented
2628 ;; out" with category properties.
2629 (if (eq (c-get-char-property macro-start-or-from 'category)
2630 'c-cpp-delimiter)
2631 macro-start-or-from
2632 from))
2633 ce) ; Position of "}"
0ec1d2c5
AM
2634 (or upper-lim (setq upper-lim from))
2635
2636 ;; If we're essentially repeating a fruitless search, just give up.
2637 (unless (and c-state-brace-pair-desert
2638 (eq cache-pos (car c-state-brace-pair-desert))
b03aabda
AM
2639 (or (null (car c-state-brace-pair-desert))
2640 (> from (car c-state-brace-pair-desert)))
0ec1d2c5 2641 (<= from (cdr c-state-brace-pair-desert)))
b03aabda 2642 ;; DESERT-LIM. Avoid repeated searching through the cached desert.
583e23bd
AM
2643 (let ((desert-lim
2644 (and c-state-brace-pair-desert
2645 (eq cache-pos (car c-state-brace-pair-desert))
b03aabda 2646 (>= from (cdr c-state-brace-pair-desert))
583e23bd
AM
2647 (cdr c-state-brace-pair-desert)))
2648 ;; CACHE-LIM. This limit will be necessary when an opening
2649 ;; paren at `cache-pos' has just had its matching close paren
b03aabda
AM
2650 ;; inserted into the buffer. `cache-pos' continues to be a
2651 ;; search bound, even though the algorithm below would skip
2652 ;; over the new paren pair.
583e23bd
AM
2653 (cache-lim (and cache-pos (< cache-pos from) cache-pos)))
2654 (narrow-to-region
2655 (cond
2656 ((and desert-lim cache-lim)
2657 (max desert-lim cache-lim))
2658 (desert-lim)
2659 (cache-lim)
2660 ((point-min)))
d23b0804
AM
2661 ;; The top limit is EOB to ensure that `bra' is inside the
2662 ;; accessible part of the buffer at the next scan operation.
2663 (1+ (buffer-size))))
0ec1d2c5
AM
2664
2665 ;; In the next pair of nested loops, the inner one moves back past a
2666 ;; pair of (mis-)matching parens or brackets; the outer one moves
2667 ;; back over a sequence of unmatched close brace/paren/bracket each
2668 ;; time round.
2669 (while
2670 (progn
2671 (c-safe
2672 (while
2673 (and (setq ce (scan-lists bra -1 -1)) ; back past )/]/}; might signal
2674 (setq bra (scan-lists ce -1 1)) ; back past (/[/{; might signal
267eacb5
AM
2675 (or (> bra here) ;(> ce here)
2676 (and
2677 (< ce here)
2678 (or (not (eq (char-after bra) ?\{))
2679 (and (goto-char bra)
2680 (c-beginning-of-macro)
2681 (< (point) macro-start-or-from))))))))
0ec1d2c5
AM
2682 (and ce (< ce bra)))
2683 (setq bra ce)) ; If we just backed over an unbalanced closing
2684 ; brace, ignore it.
2685
267eacb5 2686 (if (and ce (< ce here) (< bra ce) (eq (char-after bra) ?\{))
0ec1d2c5
AM
2687 ;; We've found the desired brace-pair.
2688 (progn
2689 (setq new-cons (cons bra (1+ ce)))
2690 (cond
2691 ((consp (car c-state-cache))
2692 (setcar c-state-cache new-cons))
2693 ((and (numberp (car c-state-cache)) ; probably never happens
2694 (< ce (car c-state-cache)))
2695 (setcdr c-state-cache
2696 (cons new-cons (cdr c-state-cache))))
2697 (t (setq c-state-cache (cons new-cons c-state-cache)))))
2698
583e23bd 2699 ;; We haven't found a brace pair. Record this in the cache.
267eacb5
AM
2700 (setq c-state-brace-pair-desert
2701 (cons (if (and ce (< bra ce) (> ce here)) ; {..} straddling HERE?
2702 bra
2703 (point-min))
2704 (min here from)))))))))
0ec1d2c5
AM
2705
2706(defsubst c-state-push-any-brace-pair (bra+1 macro-start-or-here)
2707 ;; If BRA+1 is nil, do nothing. Otherwise, BRA+1 is the buffer position
2708 ;; following a {, and that brace has a (mis-)matching } (or ]), and we
2709 ;; "push" "a" brace pair onto `c-state-cache'.
2710 ;;
2711 ;; Here "push" means overwrite the top element if it's itself a brace-pair,
2712 ;; otherwise push it normally.
2713 ;;
2714 ;; The brace pair we push is normally the one surrounding BRA+1, but if the
2715 ;; latter is inside a macro, not being a macro containing
2716 ;; MACRO-START-OR-HERE, we scan backwards through the buffer for a non-macro
2717 ;; base pair. This latter case is assumed to be rare.
2718 ;;
2719 ;; Note: POINT is not preserved in this routine.
2720 (if bra+1
2721 (if (or (> bra+1 macro-start-or-here)
2722 (progn (goto-char bra+1)
2723 (not (c-beginning-of-macro))))
2724 (setq c-state-cache
2725 (cons (cons (1- bra+1)
2726 (scan-lists bra+1 1 1))
2727 (if (consp (car c-state-cache))
2728 (cdr c-state-cache)
2729 c-state-cache)))
d23b0804 2730 ;; N.B. This defsubst codes one method for the simple, normal case,
0ec1d2c5 2731 ;; and a more sophisticated, slower way for the general case. Don't
e1dbe924 2732 ;; eliminate this defsubst - it's a speed optimization.
d23b0804 2733 (c-append-lower-brace-pair-to-state-cache (1- bra+1) (point-max)))))
0ec1d2c5 2734
d23b0804
AM
2735(defun c-append-to-state-cache (from here)
2736 ;; Scan the buffer from FROM to HERE, adding elements into `c-state-cache'
2737 ;; for braces etc. Return a candidate for `c-state-cache-good-pos'.
0ec1d2c5
AM
2738 ;;
2739 ;; FROM must be after the latest brace/paren/bracket in `c-state-cache', if
2740 ;; any. Typically, it is immediately after it. It must not be inside a
2741 ;; literal.
d23b0804 2742 (let ((here-bol (c-point 'bol here))
0ec1d2c5 2743 (macro-start-or-here
d23b0804 2744 (save-excursion (goto-char here)
0ec1d2c5
AM
2745 (if (c-beginning-of-macro)
2746 (point)
d23b0804 2747 here)))
0ec1d2c5
AM
2748 pa+1 ; pos just after an opening PAren (or brace).
2749 (ren+1 from) ; usually a pos just after an closing paREN etc.
2750 ; Is actually the pos. to scan for a (/{/[ from,
2751 ; which sometimes is after a silly )/}/].
2752 paren+1 ; Pos after some opening or closing paren.
2753 paren+1s ; A list of `paren+1's; used to determine a
2754 ; good-pos.
2755 bra+1 ce+1 ; just after L/R bra-ces.
2756 bra+1s ; list of OLD values of bra+1.
2757 mstart) ; start of a macro.
2758
2759 (save-excursion
d23b0804
AM
2760 (save-restriction
2761 (narrow-to-region (point-min) here)
2762 ;; Each time round the following loop, we enter a successively deeper
2763 ;; level of brace/paren nesting. (Except sometimes we "continue at
2764 ;; the existing level".) `pa+1' is a pos inside an opening
2765 ;; brace/paren/bracket, usually just after it.
2766 (while
2767 (progn
2768 ;; Each time round the next loop moves forward over an opening then
2769 ;; a closing brace/bracket/paren. This loop is white hot, so it
2770 ;; plays ugly tricks to go fast. DON'T PUT ANYTHING INTO THIS
2771 ;; LOOP WHICH ISN'T ABSOLUTELY NECESSARY!!! It terminates when a
2772 ;; call of `scan-lists' signals an error, which happens when there
2773 ;; are no more b/b/p's to scan.
2774 (c-safe
2775 (while t
2776 (setq pa+1 (scan-lists ren+1 1 -1) ; Into (/{/[; might signal
2777 paren+1s (cons pa+1 paren+1s))
2778 (setq ren+1 (scan-lists pa+1 1 1)) ; Out of )/}/]; might signal
2779 (if (and (eq (char-before pa+1) ?{)) ; Check for a macro later.
2780 (setq bra+1 pa+1))
2781 (setcar paren+1s ren+1)))
2782
2783 (if (and pa+1 (> pa+1 ren+1))
2784 ;; We've just entered a deeper nesting level.
2785 (progn
2786 ;; Insert the brace pair (if present) and the single open
2787 ;; paren/brace/bracket into `c-state-cache' It cannot be
2788 ;; inside a macro, except one around point, because of what
2789 ;; `c-neutralize-syntax-in-CPP' has done.
2790 (c-state-push-any-brace-pair bra+1 macro-start-or-here)
2791 ;; Insert the opening brace/bracket/paren position.
2792 (setq c-state-cache (cons (1- pa+1) c-state-cache))
2793 ;; Clear admin stuff for the next more nested part of the scan.
2794 (setq ren+1 pa+1 pa+1 nil bra+1 nil bra+1s nil)
2795 t) ; Carry on the loop
2796
2797 ;; All open p/b/b's at this nesting level, if any, have probably
2798 ;; been closed by matching/mismatching ones. We're probably
2799 ;; finished - we just need to check for having found an
2800 ;; unmatched )/}/], which we ignore. Such a )/}/] can't be in a
2801 ;; macro, due the action of `c-neutralize-syntax-in-CPP'.
2802 (c-safe (setq ren+1 (scan-lists ren+1 1 1)))))) ; acts as loop control.
2803
2804 ;; Record the final, innermost, brace-pair if there is one.
2805 (c-state-push-any-brace-pair bra+1 macro-start-or-here)
2806
2807 ;; Determine a good pos
2808 (while (and (setq paren+1 (car paren+1s))
2809 (> (if (> paren+1 macro-start-or-here)
2810 paren+1
2811 (goto-char paren+1)
2812 (setq mstart (and (c-beginning-of-macro)
2813 (point)))
2814 (or mstart paren+1))
2815 here-bol))
2816 (setq paren+1s (cdr paren+1s)))
2817 (cond
2818 ((and paren+1 mstart)
2819 (min paren+1 mstart))
2820 (paren+1)
2821 (t from))))))
2822
2823(defun c-remove-stale-state-cache (start-point here pps-point)
0ec1d2c5 2824 ;; Remove stale entries from the `c-cache-state', i.e. those which will
d23b0804
AM
2825 ;; not be in it when it is amended for position HERE. Additionally, the
2826 ;; "outermost" open-brace entry before HERE will be converted to a cons if
2827 ;; the matching close-brace is scanned.
0ec1d2c5 2828 ;;
267eacb5 2829 ;; START-POINT is a "maximal" "safe position" - there must be no open
d23b0804 2830 ;; parens/braces/brackets between START-POINT and HERE.
0ec1d2c5
AM
2831 ;;
2832 ;; As a second thing, calculate the result of parse-partial-sexp at
267eacb5 2833 ;; PPS-POINT, w.r.t. START-POINT. The motivation here is that
0ec1d2c5 2834 ;; `c-state-cache-good-pos' may become PPS-POINT, but the caller may need to
267eacb5 2835 ;; adjust it to get outside a string/comment. (Sorry about this! The code
0ec1d2c5
AM
2836 ;; needs to be FAST).
2837 ;;
2838 ;; Return a list (GOOD-POS SCAN-BACK-POS PPS-STATE), where
2839 ;; o - GOOD-POS is a position where the new value `c-state-cache' is known
2840 ;; to be good (we aim for this to be as high as possible);
2841 ;; o - SCAN-BACK-POS, if not nil, indicates there may be a brace pair
2842 ;; preceding POS which needs to be recorded in `c-state-cache'. It is a
267eacb5
AM
2843 ;; position to scan backwards from. It is the position of the "{" of the
2844 ;; last element to be removed from `c-state-cache', when that elt is a
2845 ;; cons, otherwise nil.
0ec1d2c5 2846 ;; o - PPS-STATE is the parse-partial-sexp state at PPS-POINT.
d23b0804
AM
2847 (save-excursion
2848 (save-restriction
2849 (narrow-to-region 1 (point-max))
2850 (let* ((in-macro-start ; start of macro containing HERE or nil.
0ec1d2c5 2851 (save-excursion
d23b0804 2852 (goto-char here)
0ec1d2c5
AM
2853 (and (c-beginning-of-macro)
2854 (point))))
267eacb5
AM
2855 (start-point-actual-macro-start ; Start of macro containing
2856 ; start-point or nil
d23b0804 2857 (and (< start-point here)
0ec1d2c5 2858 (save-excursion
267eacb5 2859 (goto-char start-point)
0ec1d2c5
AM
2860 (and (c-beginning-of-macro)
2861 (point)))))
267eacb5 2862 (start-point-actual-macro-end ; End of this macro, (maybe
d23b0804 2863 ; HERE), or nil.
267eacb5 2864 (and start-point-actual-macro-start
0ec1d2c5 2865 (save-excursion
267eacb5 2866 (goto-char start-point-actual-macro-start)
0ec1d2c5
AM
2867 (c-end-of-macro)
2868 (point))))
267eacb5 2869 pps-state ; Will be 9 or 10 elements long.
0ec1d2c5
AM
2870 pos
2871 upper-lim ; ,beyond which `c-state-cache' entries are removed
2872 scan-back-pos
2873 pair-beg pps-point-state target-depth)
2874
d23b0804
AM
2875 ;; Remove entries beyond HERE. Also remove any entries inside
2876 ;; a macro, unless HERE is in the same macro.
0ec1d2c5
AM
2877 (setq upper-lim
2878 (if (or (null c-state-old-cpp-beg)
d23b0804
AM
2879 (and (> here c-state-old-cpp-beg)
2880 (< here c-state-old-cpp-end)))
2881 here
2882 (min here c-state-old-cpp-beg)))
657071fc 2883 (while (and c-state-cache (>= (c-state-cache-top-lparen) upper-lim))
8cff5feb 2884 (setq scan-back-pos (car-safe (car c-state-cache)))
0ec1d2c5 2885 (setq c-state-cache (cdr c-state-cache)))
8cff5feb 2886
0ec1d2c5
AM
2887 ;; If `upper-lim' is inside the last recorded brace pair, remove its
2888 ;; RBrace and indicate we'll need to search backwards for a previous
2889 ;; brace pair.
2890 (when (and c-state-cache
2891 (consp (car c-state-cache))
2892 (> (cdar c-state-cache) upper-lim))
2893 (setcar c-state-cache (caar c-state-cache))
2894 (setq scan-back-pos (car c-state-cache)))
2895
2896 ;; The next loop jumps forward out of a nested level of parens each
2897 ;; time round; the corresponding elements in `c-state-cache' are
2898 ;; removed. `pos' is just after the brace-pair or the open paren at
267eacb5 2899 ;; (car c-state-cache). There can be no open parens/braces/brackets
d23b0804 2900 ;; between `start-point'/`start-point-actual-macro-start' and HERE,
0ec1d2c5 2901 ;; due to the interface spec to this function.
267eacb5
AM
2902 (setq pos (if (and start-point-actual-macro-end
2903 (not (eq start-point-actual-macro-start
eb2f6eeb 2904 in-macro-start)))
267eacb5 2905 (1+ start-point-actual-macro-end) ; get outside the macro as
0ec1d2c5 2906 ; marked by a `category' text property.
267eacb5 2907 start-point))
0ec1d2c5
AM
2908 (goto-char pos)
2909 (while (and c-state-cache
d23b0804
AM
2910 (or (numberp (car c-state-cache)) ; Have we a { at all?
2911 (cdr c-state-cache))
2912 (< (point) here))
0ec1d2c5
AM
2913 (cond
2914 ((null pps-state) ; first time through
2915 (setq target-depth -1))
2916 ((eq (car pps-state) target-depth) ; found closing ),},]
2917 (setq target-depth (1- (car pps-state))))
2918 ;; Do nothing when we've merely reached pps-point.
2919 )
2920
2921 ;; Scan!
2922 (setq pps-state
2923 (parse-partial-sexp
d23b0804 2924 (point) (if (< (point) pps-point) pps-point here)
0ec1d2c5
AM
2925 target-depth
2926 nil pps-state))
2927
2928 (if (= (point) pps-point)
2929 (setq pps-point-state pps-state))
2930
2931 (when (eq (car pps-state) target-depth)
2932 (setq pos (point)) ; POS is now just after an R-paren/brace.
2933 (cond
2934 ((and (consp (car c-state-cache))
2935 (eq (point) (cdar c-state-cache)))
2936 ;; We've just moved out of the paren pair containing the brace-pair
2937 ;; at (car c-state-cache). `pair-beg' is where the open paren is,
2938 ;; and is potentially where the open brace of a cons in
2939 ;; c-state-cache will be.
2940 (setq pair-beg (car-safe (cdr c-state-cache))
2941 c-state-cache (cdr-safe (cdr c-state-cache)))) ; remove {}pair + containing Lparen.
2942 ((numberp (car c-state-cache))
2943 (setq pair-beg (car c-state-cache)
2944 c-state-cache (cdr c-state-cache))) ; remove this
2945 ; containing Lparen
2946 ((numberp (cadr c-state-cache))
2947 (setq pair-beg (cadr c-state-cache)
2948 c-state-cache (cddr c-state-cache))) ; Remove a paren pair
2949 ; together with enclosed brace pair.
2950 ;; (t nil) ; Ignore an unmated Rparen.
2951 )))
2952
2953 (if (< (point) pps-point)
2954 (setq pps-state (parse-partial-sexp (point) pps-point
2955 nil nil ; TARGETDEPTH, STOPBEFORE
2956 pps-state)))
2957
2958 ;; If the last paren pair we moved out of was actually a brace pair,
2959 ;; insert it into `c-state-cache'.
2960 (when (and pair-beg (eq (char-after pair-beg) ?{))
2961 (if (consp (car-safe c-state-cache))
2962 (setq c-state-cache (cdr c-state-cache)))
2963 (setq c-state-cache (cons (cons pair-beg pos)
2964 c-state-cache)))
2965
2966 (list pos scan-back-pos pps-state)))))
2967
267eacb5 2968(defun c-remove-stale-state-cache-backwards (here)
0ec1d2c5
AM
2969 ;; Strip stale elements of `c-state-cache' by moving backwards through the
2970 ;; buffer, and inform the caller of the scenario detected.
2971 ;;
2972 ;; HERE is the position we're setting `c-state-cache' for.
267eacb5
AM
2973 ;; CACHE-POS (a locally bound variable) is just after the latest recorded
2974 ;; position in `c-state-cache' before HERE, or a position at or near
2975 ;; point-min which isn't in a literal.
0ec1d2c5
AM
2976 ;;
2977 ;; This function must only be called only when (> `c-state-cache-good-pos'
2978 ;; HERE). Usually the gap between CACHE-POS and HERE is large. It is thus
e1dbe924 2979 ;; optimized to eliminate (or minimize) scanning between these two
0ec1d2c5
AM
2980 ;; positions.
2981 ;;
2982 ;; Return a three element list (GOOD-POS SCAN-BACK-POS FWD-FLAG), where:
2983 ;; o - GOOD-POS is a "good position", where `c-state-cache' is valid, or
2984 ;; could become so after missing elements are inserted into
2985 ;; `c-state-cache'. This is JUST AFTER an opening or closing
2986 ;; brace/paren/bracket which is already in `c-state-cache' or just before
2987 ;; one otherwise. exceptionally (when there's no such b/p/b handy) the BOL
2988 ;; before `here''s line, or the start of the literal containing it.
2989 ;; o - SCAN-BACK-POS, if non-nil, indicates there may be a brace pair
2990 ;; preceding POS which isn't recorded in `c-state-cache'. It is a position
2991 ;; to scan backwards from.
2992 ;; o - FWD-FLAG, if non-nil, indicates there may be parens/braces between
2993 ;; POS and HERE which aren't recorded in `c-state-cache'.
2994 ;;
2995 ;; The comments in this defun use "paren" to mean parenthesis or square
2996 ;; bracket (as contrasted with a brace), and "(" and ")" likewise.
2997 ;;
267eacb5
AM
2998 ;; . {..} (..) (..) ( .. { } ) (...) ( .... . ..)
2999 ;; | | | | | |
3000 ;; CP E here D C good
3001 (let ((cache-pos (c-get-cache-scan-pos here)) ; highest position below HERE in cache (or 1)
3002 (pos c-state-cache-good-pos)
0ec1d2c5
AM
3003 pa ren ; positions of "(" and ")"
3004 dropped-cons ; whether the last element dropped from `c-state-cache'
3005 ; was a cons (representing a brace-pair)
3006 good-pos ; see above.
3007 lit ; (START . END) of a literal containing some point.
3008 here-lit-start here-lit-end ; bounds of literal containing `here'
3009 ; or `here' itself.
c0209c2c 3010 here- here+ ; start/end of macro around HERE, or HERE
0ec1d2c5 3011 (here-bol (c-point 'bol here))
c400c4d7 3012 (too-far-back (max (- here c-state-cache-too-far) (point-min))))
0ec1d2c5
AM
3013
3014 ;; Remove completely irrelevant entries from `c-state-cache'.
3015 (while (and c-state-cache
3016 (>= (setq pa (c-state-cache-top-lparen)) here))
3017 (setq dropped-cons (consp (car c-state-cache)))
3018 (setq c-state-cache (cdr c-state-cache))
3019 (setq pos pa))
3020 ;; At this stage, (> pos here);
3021 ;; (< (c-state-cache-top-lparen) here) (or is nil).
3022
c0209c2c
AM
3023 (cond
3024 ((and (consp (car c-state-cache))
3025 (> (cdar c-state-cache) here))
3026 ;; CASE 1: The top of the cache is a brace pair which now encloses
3027 ;; `here'. As good-pos, return the address. of the "{". Since we've no
3028 ;; knowledge of what's inside these braces, we have no alternative but
3029 ;; to direct the caller to scan the buffer from the opening brace.
3030 (setq pos (caar c-state-cache))
3031 (setcar c-state-cache pos)
3032 (list (1+ pos) pos t)) ; return value. We've just converted a brace pair
3033 ; entry into a { entry, so the caller needs to
3034 ; search for a brace pair before the {.
3035
3036 ;; `here' might be inside a literal. Check for this.
3037 ((progn
3038 (setq lit (c-state-literal-at here)
3039 here-lit-start (or (car lit) here)
3040 here-lit-end (or (cdr lit) here))
3041 ;; Has `here' just "newly entered" a macro?
3042 (save-excursion
3043 (goto-char here-lit-start)
3044 (if (and (c-beginning-of-macro)
3045 (or (null c-state-old-cpp-beg)
3046 (not (= (point) c-state-old-cpp-beg))))
3047 (progn
3048 (setq here- (point))
3049 (c-end-of-macro)
3050 (setq here+ (point)))
3051 (setq here- here-lit-start
3052 here+ here-lit-end)))
3053
3054 ;; `here' might be nested inside any depth of parens (or brackets but
3055 ;; not braces). Scan backwards to find the outermost such opening
3056 ;; paren, if there is one. This will be the scan position to return.
3057 (save-restriction
3058 (narrow-to-region cache-pos (point-max))
3059 (setq pos (c-state-balance-parens-backwards here- here+ pos)))
3060 nil)) ; for the cond
3061
3062 ((< pos here-lit-start)
3063 ;; CASE 2: Address of outermost ( or [ which now encloses `here', but
3064 ;; didn't enclose the (previous) `c-state-cache-good-pos'. If there is
3065 ;; a brace pair preceding this, it will already be in `c-state-cache',
3066 ;; unless there was a brace pair after it, i.e. there'll only be one to
3067 ;; scan for if we've just deleted one.
3068 (list pos (and dropped-cons pos) t)) ; Return value.
3069
3070 ;; `here' isn't enclosed in a (previously unrecorded) bracket/paren.
3071 ;; Further forward scanning isn't needed, but we still need to find a
3072 ;; GOOD-POS. Step out of all enclosing "("s on HERE's line.
3073 ((progn
0ec1d2c5
AM
3074 (save-restriction
3075 (narrow-to-region here-bol (point-max))
3076 (setq pos here-lit-start)
3077 (c-safe (while (setq pa (scan-lists pos -1 1))
3078 (setq pos pa)))) ; might signal
c0209c2c
AM
3079 nil)) ; for the cond
3080
3081 ((setq ren (c-safe-scan-lists pos -1 -1 too-far-back))
3082 ;; CASE 3: After a }/)/] before `here''s BOL.
3083 (list (1+ ren) (and dropped-cons pos) nil)) ; Return value
3084
3085 (t
3086 ;; CASE 4; Best of a bad job: BOL before `here-bol', or beginning of
3087 ;; literal containing it.
3088 (setq good-pos (c-state-lit-beg (c-point 'bopl here-bol)))
3089 (list good-pos (and dropped-cons good-pos) nil)))))
0ec1d2c5
AM
3090
3091
3092;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
3093;; Externally visible routines.
3094
3095(defun c-state-cache-init ()
3096 (setq c-state-cache nil
3097 c-state-cache-good-pos 1
3098 c-state-nonlit-pos-cache nil
3099 c-state-nonlit-pos-cache-limit 1
8ed43f15
AM
3100 c-state-semi-nonlit-pos-cache nil
3101 c-state-semi-nonlit-pos-cache-limit 1
0ec1d2c5
AM
3102 c-state-brace-pair-desert nil
3103 c-state-point-min 1
3104 c-state-point-min-lit-type nil
3105 c-state-point-min-lit-start nil
3106 c-state-min-scan-pos 1
3107 c-state-old-cpp-beg nil
3108 c-state-old-cpp-end nil)
3109 (c-state-mark-point-min-literal))
3110
e84efb70
AM
3111;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
3112;; Debugging routines to dump `c-state-cache' in a "replayable" form.
3113;; (defmacro c-sc-de (elt) ; "c-state-cache-dump-element"
3114;; `(format ,(concat "(setq " (symbol-name elt) " %s) ") ,elt))
3115;; (defmacro c-sc-qde (elt) ; "c-state-cache-quote-dump-element"
3116;; `(format ,(concat "(setq " (symbol-name elt) " '%s) ") ,elt))
3117;; (defun c-state-dump ()
3118;; ;; For debugging.
3119;; ;(message
3120;; (concat
3121;; (c-sc-qde c-state-cache)
3122;; (c-sc-de c-state-cache-good-pos)
3123;; (c-sc-qde c-state-nonlit-pos-cache)
3124;; (c-sc-de c-state-nonlit-pos-cache-limit)
3125;; (c-sc-qde c-state-brace-pair-desert)
3126;; (c-sc-de c-state-point-min)
3127;; (c-sc-de c-state-point-min-lit-type)
3128;; (c-sc-de c-state-point-min-lit-start)
3129;; (c-sc-de c-state-min-scan-pos)
3130;; (c-sc-de c-state-old-cpp-beg)
3131;; (c-sc-de c-state-old-cpp-end)))
3132;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
3133
0ec1d2c5
AM
3134(defun c-invalidate-state-cache-1 (here)
3135 ;; Invalidate all info on `c-state-cache' that applies to the buffer at HERE
3136 ;; or higher and set `c-state-cache-good-pos' accordingly. The cache is
3137 ;; left in a consistent state.
3138 ;;
3139 ;; This is much like `c-whack-state-after', but it never changes a paren
3140 ;; pair element into an open paren element. Doing that would mean that the
3141 ;; new open paren wouldn't have the required preceding paren pair element.
785eecbb 3142 ;;
0ec1d2c5
AM
3143 ;; This function is called from c-after-change.
3144
56d093a9 3145 ;; The caches of non-literals:
0ec1d2c5
AM
3146 (if (< here c-state-nonlit-pos-cache-limit)
3147 (setq c-state-nonlit-pos-cache-limit here))
56d093a9
AM
3148 (if (< here c-state-semi-nonlit-pos-cache-limit)
3149 (setq c-state-semi-nonlit-pos-cache-limit here))
0ec1d2c5
AM
3150
3151 ;; `c-state-cache':
3152 ;; Case 1: if `here' is in a literal containing point-min, everything
3153 ;; becomes (or is already) nil.
3154 (if (or (null c-state-cache-good-pos)
3155 (< here (c-state-get-min-scan-pos)))
3156 (setq c-state-cache nil
3157 c-state-cache-good-pos nil
3158 c-state-min-scan-pos nil)
3159
c400c4d7
SM
3160 ;; Truncate `c-state-cache' and set `c-state-cache-good-pos' to a value
3161 ;; below `here'. To maintain its consistency, we may need to insert a new
3162 ;; brace pair.
0ec1d2c5
AM
3163 (let ((here-bol (c-point 'bol here))
3164 too-high-pa ; recorded {/(/[ next above here, or nil.
3165 dropped-cons ; was the last removed element a brace pair?
3166 pa)
3167 ;; The easy bit - knock over-the-top bits off `c-state-cache'.
3168 (while (and c-state-cache
3169 (>= (setq pa (c-state-cache-top-paren)) here))
3170 (setq dropped-cons (consp (car c-state-cache))
3171 too-high-pa (c-state-cache-top-lparen)
3172 c-state-cache (cdr c-state-cache)))
3173
3174 ;; Do we need to add in an earlier brace pair, having lopped one off?
3175 (if (and dropped-cons
3176 (< too-high-pa (+ here c-state-cache-too-far)))
d23b0804 3177 (c-append-lower-brace-pair-to-state-cache too-high-pa here here-bol))
0ec1d2c5
AM
3178 (setq c-state-cache-good-pos (or (c-state-cache-after-top-paren)
3179 (c-state-get-min-scan-pos)))))
3180
3181 ;; The brace-pair desert marker:
3182 (when (car c-state-brace-pair-desert)
3183 (if (< here (car c-state-brace-pair-desert))
3184 (setq c-state-brace-pair-desert nil)
3185 (if (< here (cdr c-state-brace-pair-desert))
3186 (setcdr c-state-brace-pair-desert here)))))
3187
3188(defun c-parse-state-1 ()
3189 ;; Find and record all noteworthy parens between some good point earlier in
3190 ;; the file and point. That good point is at least the beginning of the
3191 ;; top-level construct we are in, or the beginning of the preceding
3192 ;; top-level construct if we aren't in one.
3193 ;;
3194 ;; The returned value is a list of the noteworthy parens with the last one
3195 ;; first. If an element in the list is an integer, it's the position of an
3196 ;; open paren (of any type) which has not been closed before the point. If
3197 ;; an element is a cons, it gives the position of a closed BRACE paren
3198 ;; pair[*]; the car is the start brace position and the cdr is the position
3199 ;; following the closing brace. Only the last closed brace paren pair
3200 ;; before each open paren and before the point is recorded, and thus the
3201 ;; state never contains two cons elements in succession. When a close brace
3202 ;; has no matching open brace (e.g., the matching brace is outside the
3203 ;; visible region), it is not represented in the returned value.
3204 ;;
3205 ;; [*] N.B. The close "brace" might be a mismatching close bracket or paren.
3206 ;; This defun explicitly treats mismatching parens/braces/brackets as
3207 ;; matching. It is the open brace which makes it a "brace" pair.
3208 ;;
3209 ;; If POINT is within a macro, open parens and brace pairs within
3210 ;; THIS macro MIGHT be recorded. This depends on whether their
3211 ;; syntactic properties have been suppressed by
3212 ;; `c-neutralize-syntax-in-CPP'. This might need fixing (2008-12-11).
d9e94c22
MS
3213 ;;
3214 ;; Currently no characters which are given paren syntax with the
3215 ;; syntax-table property are recorded, i.e. angle bracket arglist
3216 ;; parens are never present here. Note that this might change.
3217 ;;
0386b551 3218 ;; BUG: This function doesn't cope entirely well with unbalanced
0ec1d2c5
AM
3219 ;; parens in macros. (2008-12-11: this has probably been resolved
3220 ;; by the function `c-neutralize-syntax-in-CPP'.) E.g. in the
3221 ;; following case the brace before the macro isn't balanced with the
3222 ;; one after it:
0386b551
AM
3223 ;;
3224 ;; {
3225 ;; #define X {
3226 ;; }
3227 ;;
0ec1d2c5
AM
3228 ;; Note to maintainers: this function DOES get called with point
3229 ;; within comments and strings, so don't assume it doesn't!
3230 ;;
0386b551 3231 ;; This function might do hidden buffer changes.
0ec1d2c5
AM
3232 (let* ((here (point))
3233 (here-bopl (c-point 'bopl))
3234 strategy ; 'forward, 'backward etc..
3235 ;; Candidate positions to start scanning from:
3236 cache-pos ; highest position below HERE already existing in
3237 ; cache (or 1).
3238 good-pos
267eacb5
AM
3239 start-point ; (when scanning forward) a place below HERE where there
3240 ; are no open parens/braces between it and HERE.
0ec1d2c5
AM
3241 bopl-state
3242 res
3243 scan-backward-pos scan-forward-p) ; used for 'backward.
3244 ;; If POINT-MIN has changed, adjust the cache
3245 (unless (= (point-min) c-state-point-min)
3246 (c-renarrow-state-cache))
3247
3248 ;; Strategy?
3249 (setq res (c-parse-state-get-strategy here c-state-cache-good-pos)
3250 strategy (car res)
267eacb5 3251 start-point (cadr res))
0ec1d2c5 3252
0ec1d2c5 3253 ;; SCAN!
d23b0804
AM
3254 (cond
3255 ((eq strategy 'forward)
3256 (setq res (c-remove-stale-state-cache start-point here here-bopl))
3257 (setq cache-pos (car res)
3258 scan-backward-pos (cadr res)
3259 bopl-state (car (cddr res))) ; will be nil if (< here-bopl
0ec1d2c5 3260 ; start-point)
d23b0804
AM
3261 (if scan-backward-pos
3262 (c-append-lower-brace-pair-to-state-cache scan-backward-pos here))
3263 (setq good-pos
3264 (c-append-to-state-cache cache-pos here))
3265 (setq c-state-cache-good-pos
3266 (if (and bopl-state
3267 (< good-pos (- here c-state-cache-too-far)))
3268 (c-state-cache-non-literal-place here-bopl bopl-state)
3269 good-pos)))
3270
3271 ((eq strategy 'backward)
3272 (setq res (c-remove-stale-state-cache-backwards here)
3273 good-pos (car res)
3274 scan-backward-pos (cadr res)
3275 scan-forward-p (car (cddr res)))
3276 (if scan-backward-pos
3277 (c-append-lower-brace-pair-to-state-cache scan-backward-pos here))
3278 (setq c-state-cache-good-pos
3279 (if scan-forward-p
3280 (c-append-to-state-cache good-pos here)
3281 good-pos)))
3282
3283 (t ; (eq strategy 'IN-LIT)
3284 (setq c-state-cache nil
3285 c-state-cache-good-pos nil))))
0ec1d2c5
AM
3286
3287 c-state-cache)
3288
3289(defun c-invalidate-state-cache (here)
3290 ;; This is a wrapper over `c-invalidate-state-cache-1'.
3291 ;;
3292 ;; It suppresses the syntactic effect of the < and > (template) brackets and
3293 ;; of all parens in preprocessor constructs, except for any such construct
3294 ;; containing point. We can then call `c-invalidate-state-cache-1' without
3295 ;; worrying further about macros and template delimiters.
3296 (c-with-<->-as-parens-suppressed
3e8f7d91
AM
3297 (if (and c-state-old-cpp-beg
3298 (< c-state-old-cpp-beg here))
0ec1d2c5 3299 (c-with-all-but-one-cpps-commented-out
3e8f7d91
AM
3300 c-state-old-cpp-beg
3301 (min c-state-old-cpp-end here)
0ec1d2c5
AM
3302 (c-invalidate-state-cache-1 here))
3303 (c-with-cpps-commented-out
3304 (c-invalidate-state-cache-1 here)))))
3305
8cff5feb
AM
3306(defmacro c-state-maybe-marker (place marker)
3307 ;; If PLACE is non-nil, return a marker marking it, otherwise nil.
3308 ;; We (re)use MARKER.
3309 `(and ,place
3310 (or ,marker (setq ,marker (make-marker)))
3311 (set-marker ,marker ,place)))
3312
0ec1d2c5
AM
3313(defun c-parse-state ()
3314 ;; This is a wrapper over `c-parse-state-1'. See that function for a
3315 ;; description of the functionality and return value.
3316 ;;
3317 ;; It suppresses the syntactic effect of the < and > (template) brackets and
3318 ;; of all parens in preprocessor constructs, except for any such construct
3319 ;; containing point. We can then call `c-parse-state-1' without worrying
3320 ;; further about macros and template delimiters.
3321 (let (here-cpp-beg here-cpp-end)
3322 (save-excursion
3323 (when (c-beginning-of-macro)
3324 (setq here-cpp-beg (point))
3325 (unless
3326 (> (setq here-cpp-end (c-syntactic-end-of-macro))
3327 here-cpp-beg)
3328 (setq here-cpp-beg nil here-cpp-end nil))))
3329 ;; FIXME!!! Put in a `condition-case' here to protect the integrity of the
3330 ;; subsystem.
3331 (prog1
3332 (c-with-<->-as-parens-suppressed
3333 (if (and here-cpp-beg (> here-cpp-end here-cpp-beg))
3334 (c-with-all-but-one-cpps-commented-out
3335 here-cpp-beg here-cpp-end
3336 (c-parse-state-1))
3337 (c-with-cpps-commented-out
3338 (c-parse-state-1))))
8cff5feb
AM
3339 (setq c-state-old-cpp-beg
3340 (c-state-maybe-marker here-cpp-beg c-state-old-cpp-beg-marker)
3341 c-state-old-cpp-end
3342 (c-state-maybe-marker here-cpp-end c-state-old-cpp-end-marker)))))
0ec1d2c5
AM
3343
3344;; Debug tool to catch cache inconsistencies. This is called from
3345;; 000tests.el.
a66cd3ee
MS
3346(defvar c-debug-parse-state nil)
3347(unless (fboundp 'c-real-parse-state)
3348 (fset 'c-real-parse-state (symbol-function 'c-parse-state)))
3349(cc-bytecomp-defun c-real-parse-state)
cd5495ff 3350
b03aabda 3351(defvar c-parse-state-point nil)
cd5495ff
AM
3352(defvar c-parse-state-state nil)
3353(defun c-record-parse-state-state ()
b03aabda 3354 (setq c-parse-state-point (point))
cd5495ff
AM
3355 (setq c-parse-state-state
3356 (mapcar
3357 (lambda (arg)
b03aabda
AM
3358 (let ((val (symbol-value arg)))
3359 (cons arg
3360 (if (consp val)
3361 (copy-tree val)
3362 val))))
cd5495ff
AM
3363 '(c-state-cache
3364 c-state-cache-good-pos
3365 c-state-nonlit-pos-cache
3366 c-state-nonlit-pos-cache-limit
8ed43f15
AM
3367 c-state-semi-nonlit-pos-cache
3368 c-state-semi-nonlit-pos-cache-limit
cd5495ff
AM
3369 c-state-brace-pair-desert
3370 c-state-point-min
3371 c-state-point-min-lit-type
3372 c-state-point-min-lit-start
3373 c-state-min-scan-pos
3374 c-state-old-cpp-beg
b03aabda
AM
3375 c-state-old-cpp-end
3376 c-parse-state-point))))
cd5495ff
AM
3377(defun c-replay-parse-state-state ()
3378 (message
3379 (concat "(setq "
3380 (mapconcat
3381 (lambda (arg)
3382 (format "%s %s%s" (car arg) (if (atom (cdr arg)) "" "'") (cdr arg)))
3383 c-parse-state-state " ")
3384 ")")))
3385
b03aabda
AM
3386(defun c-debug-parse-state-double-cons (state)
3387 (let (state-car conses-not-ok)
3388 (while state
3389 (setq state-car (car state)
3390 state (cdr state))
3391 (if (and (consp state-car)
3392 (consp (car state)))
3393 (setq conses-not-ok t)))
3394 conses-not-ok))
3395
a66cd3ee 3396(defun c-debug-parse-state ()
0ec1d2c5 3397 (let ((here (point)) (res1 (c-real-parse-state)) res2)
0386b551 3398 (let ((c-state-cache nil)
0ec1d2c5
AM
3399 (c-state-cache-good-pos 1)
3400 (c-state-nonlit-pos-cache nil)
3401 (c-state-nonlit-pos-cache-limit 1)
3402 (c-state-brace-pair-desert nil)
3403 (c-state-point-min 1)
3404 (c-state-point-min-lit-type nil)
3405 (c-state-point-min-lit-start nil)
3406 (c-state-min-scan-pos 1)
3407 (c-state-old-cpp-beg nil)
3408 (c-state-old-cpp-end nil))
a66cd3ee
MS
3409 (setq res2 (c-real-parse-state)))
3410 (unless (equal res1 res2)
0386b551
AM
3411 ;; The cache can actually go further back due to the ad-hoc way
3412 ;; the first paren is found, so try to whack off a bit of its
3413 ;; start before complaining.
cd5495ff
AM
3414 ;; (save-excursion
3415 ;; (goto-char (or (c-least-enclosing-brace res2) (point)))
3416 ;; (c-beginning-of-defun-1)
3417 ;; (while (not (or (bobp) (eq (char-after) ?{)))
3418 ;; (c-beginning-of-defun-1))
3419 ;; (unless (equal (c-whack-state-before (point) res1) res2)
3420 ;; (message (concat "c-parse-state inconsistency at %s: "
3421 ;; "using cache: %s, from scratch: %s")
3422 ;; here res1 res2)))
3423 (message (concat "c-parse-state inconsistency at %s: "
3424 "using cache: %s, from scratch: %s")
3425 here res1 res2)
3426 (message "Old state:")
3427 (c-replay-parse-state-state))
b03aabda
AM
3428
3429 (when (c-debug-parse-state-double-cons res1)
3430 (message "c-parse-state INVALIDITY at %s: %s"
3431 here res1)
3432 (message "Old state:")
3433 (c-replay-parse-state-state))
3434
cd5495ff 3435 (c-record-parse-state-state)
b03aabda
AM
3436 res2 ; res1 correct a cascading series of errors ASAP
3437 ))
0ec1d2c5 3438
a66cd3ee
MS
3439(defun c-toggle-parse-state-debug (&optional arg)
3440 (interactive "P")
3441 (setq c-debug-parse-state (c-calculate-state arg c-debug-parse-state))
3442 (fset 'c-parse-state (symbol-function (if c-debug-parse-state
3443 'c-debug-parse-state
3444 'c-real-parse-state)))
b03aabda
AM
3445 (c-keep-region-active)
3446 (message "c-debug-parse-state %sabled"
3447 (if c-debug-parse-state "en" "dis")))
0386b551
AM
3448(when c-debug-parse-state
3449 (c-toggle-parse-state-debug 1))
a66cd3ee 3450
0ec1d2c5 3451\f
d9e94c22
MS
3452(defun c-whack-state-before (bufpos paren-state)
3453 ;; Whack off any state information from PAREN-STATE which lies
3454 ;; before BUFPOS. Not destructive on PAREN-STATE.
d9e94c22
MS
3455 (let* ((newstate (list nil))
3456 (ptr newstate)
3457 car)
3458 (while paren-state
3459 (setq car (car paren-state)
3460 paren-state (cdr paren-state))
3461 (if (< (if (consp car) (car car) car) bufpos)
3462 (setq paren-state nil)
3463 (setcdr ptr (list car))
3464 (setq ptr (cdr ptr))))
3465 (cdr newstate)))
3466
3467(defun c-whack-state-after (bufpos paren-state)
3468 ;; Whack off any state information from PAREN-STATE which lies at or
3469 ;; after BUFPOS. Not destructive on PAREN-STATE.
d9e94c22
MS
3470 (catch 'done
3471 (while paren-state
3472 (let ((car (car paren-state)))
3473 (if (consp car)
3474 ;; just check the car, because in a balanced brace
3475 ;; expression, it must be impossible for the corresponding
3476 ;; close brace to be before point, but the open brace to
3477 ;; be after.
3478 (if (<= bufpos (car car))
3479 nil ; whack it off
3480 (if (< bufpos (cdr car))
3481 ;; its possible that the open brace is before
3482 ;; bufpos, but the close brace is after. In that
3483 ;; case, convert this to a non-cons element. The
3484 ;; rest of the state is before bufpos, so we're
3485 ;; done.
3486 (throw 'done (cons (car car) (cdr paren-state)))
3487 ;; we know that both the open and close braces are
3488 ;; before bufpos, so we also know that everything else
3489 ;; on state is before bufpos.
3490 (throw 'done paren-state)))
3491 (if (<= bufpos car)
3492 nil ; whack it off
3493 ;; it's before bufpos, so everything else should too.
3494 (throw 'done paren-state)))
3495 (setq paren-state (cdr paren-state)))
3496 nil)))
3497
3498(defun c-most-enclosing-brace (paren-state &optional bufpos)
3499 ;; Return the bufpos of the innermost enclosing open paren before
0386b551 3500 ;; bufpos, or nil if none was found.
d9e94c22
MS
3501 (let (enclosingp)
3502 (or bufpos (setq bufpos 134217727))
3503 (while paren-state
3504 (setq enclosingp (car paren-state)
3505 paren-state (cdr paren-state))
3506 (if (or (consp enclosingp)
3507 (>= enclosingp bufpos))
3508 (setq enclosingp nil)
d9e94c22
MS
3509 (setq paren-state nil)))
3510 enclosingp))
3511
0386b551
AM
3512(defun c-least-enclosing-brace (paren-state)
3513 ;; Return the bufpos of the outermost enclosing open paren, or nil
3514 ;; if none was found.
d9e94c22 3515 (let (pos elem)
d9e94c22
MS
3516 (while paren-state
3517 (setq elem (car paren-state)
3518 paren-state (cdr paren-state))
0386b551
AM
3519 (if (integerp elem)
3520 (setq pos elem)))
d9e94c22
MS
3521 pos))
3522
3523(defun c-safe-position (bufpos paren-state)
0386b551
AM
3524 ;; Return the closest "safe" position recorded on PAREN-STATE that
3525 ;; is higher up than BUFPOS. Return nil if PAREN-STATE doesn't
3526 ;; contain any. Return nil if BUFPOS is nil, which is useful to
3527 ;; find the closest limit before a given limit that might be nil.
d9e94c22 3528 ;;
0386b551
AM
3529 ;; A "safe" position is a position at or after a recorded open
3530 ;; paren, or after a recorded close paren. The returned position is
3531 ;; thus either the first position after a close brace, or the first
3532 ;; position after an enclosing paren, or at the enclosing paren in
3533 ;; case BUFPOS is immediately after it.
d9e94c22
MS
3534 (when bufpos
3535 (let (elem)
3536 (catch 'done
3537 (while paren-state
3538 (setq elem (car paren-state))
3539 (if (consp elem)
3540 (cond ((< (cdr elem) bufpos)
3541 (throw 'done (cdr elem)))
3542 ((< (car elem) bufpos)
3543 ;; See below.
3544 (throw 'done (min (1+ (car elem)) bufpos))))
3545 (if (< elem bufpos)
3546 ;; elem is the position at and not after the opening paren, so
3547 ;; we can go forward one more step unless it's equal to
3548 ;; bufpos. This is useful in some cases avoid an extra paren
3549 ;; level between the safe position and bufpos.
3550 (throw 'done (min (1+ elem) bufpos))))
3551 (setq paren-state (cdr paren-state)))))))
3552
3553(defun c-beginning-of-syntax ()
3554 ;; This is used for `font-lock-beginning-of-syntax-function'. It
3555 ;; goes to the closest previous point that is known to be outside
3556 ;; any string literal or comment. `c-state-cache' is used if it has
3557 ;; a position in the vicinity.
3558 (let* ((paren-state c-state-cache)
3559 elem
3560
3561 (pos (catch 'done
3562 ;; Note: Similar code in `c-safe-position'. The
3563 ;; difference is that we accept a safe position at
3564 ;; the point and don't bother to go forward past open
3565 ;; parens.
3566 (while paren-state
3567 (setq elem (car paren-state))
3568 (if (consp elem)
3569 (cond ((<= (cdr elem) (point))
3570 (throw 'done (cdr elem)))
3571 ((<= (car elem) (point))
3572 (throw 'done (car elem))))
3573 (if (<= elem (point))
3574 (throw 'done elem)))
3575 (setq paren-state (cdr paren-state)))
3576 (point-min))))
3577
3578 (if (> pos (- (point) 4000))
3579 (goto-char pos)
3580 ;; The position is far back. Try `c-beginning-of-defun-1'
3581 ;; (although we can't be entirely sure it will go to a position
3582 ;; outside a comment or string in current emacsen). FIXME:
3583 ;; Consult `syntax-ppss' here.
3584 (c-beginning-of-defun-1)
3585 (if (< (point) pos)
3586 (goto-char pos)))))
3587
3588\f
3589;; Tools for scanning identifiers and other tokens.
3590
3591(defun c-on-identifier ()
3592 "Return non-nil if the point is on or directly after an identifier.
3593Keywords are recognized and not considered identifiers. If an
3594identifier is detected, the returned value is its starting position.
0386b551
AM
3595If an identifier ends at the point and another begins at it \(can only
3596happen in Pike) then the point for the preceding one is returned.
d9e94c22 3597
0386b551
AM
3598Note that this function might do hidden buffer changes. See the
3599comment at the start of cc-engine.el for more info."
3600
3601 ;; FIXME: Shouldn't this function handle "operator" in C++?
d9e94c22
MS
3602
3603 (save-excursion
0386b551
AM
3604 (skip-syntax-backward "w_")
3605
3606 (or
3607
3608 ;; Check for a normal (non-keyword) identifier.
3609 (and (looking-at c-symbol-start)
3610 (not (looking-at c-keywords-regexp))
3611 (point))
3612
3613 (when (c-major-mode-is 'pike-mode)
3614 ;; Handle the `<operator> syntax in Pike.
3615 (let ((pos (point)))
3616 (skip-chars-backward "-!%&*+/<=>^|~[]()")
3617 (and (if (< (skip-chars-backward "`") 0)
3618 t
3619 (goto-char pos)
3620 (eq (char-after) ?\`))
3621 (looking-at c-symbol-key)
3622 (>= (match-end 0) pos)
3623 (point))))
3624
3625 ;; Handle the "operator +" syntax in C++.
3626 (when (and c-overloadable-operators-regexp
3627 (= (c-backward-token-2 0) 0))
3628
3629 (cond ((and (looking-at c-overloadable-operators-regexp)
51c9af45 3630 (or (not c-opt-op-identifier-prefix)
0386b551 3631 (and (= (c-backward-token-2 1) 0)
51c9af45 3632 (looking-at c-opt-op-identifier-prefix))))
0386b551
AM
3633 (point))
3634
3635 ((save-excursion
51c9af45
AM
3636 (and c-opt-op-identifier-prefix
3637 (looking-at c-opt-op-identifier-prefix)
0386b551
AM
3638 (= (c-forward-token-2 1) 0)
3639 (looking-at c-overloadable-operators-regexp)))
3640 (point))))
3641
3642 )))
d9e94c22
MS
3643
3644(defsubst c-simple-skip-symbol-backward ()
3645 ;; If the point is at the end of a symbol then skip backward to the
3646 ;; beginning of it. Don't move otherwise. Return non-nil if point
3647 ;; moved.
0386b551
AM
3648 ;;
3649 ;; This function might do hidden buffer changes.
d9e94c22
MS
3650 (or (< (skip-syntax-backward "w_") 0)
3651 (and (c-major-mode-is 'pike-mode)
3652 ;; Handle the `<operator> syntax in Pike.
3653 (let ((pos (point)))
2a15eb73 3654 (if (and (< (skip-chars-backward "-!%&*+/<=>^|~[]()") 0)
d9e94c22
MS
3655 (< (skip-chars-backward "`") 0)
3656 (looking-at c-symbol-key)
3657 (>= (match-end 0) pos))
3658 t
3659 (goto-char pos)
3660 nil)))))
3661
0386b551 3662(defun c-beginning-of-current-token (&optional back-limit)
d9e94c22
MS
3663 ;; Move to the beginning of the current token. Do not move if not
3664 ;; in the middle of one. BACK-LIMIT may be used to bound the
3665 ;; backward search; if given it's assumed to be at the boundary
a85fd6da 3666 ;; between two tokens. Return non-nil if the point is moved, nil
580fba94 3667 ;; otherwise.
0386b551
AM
3668 ;;
3669 ;; This function might do hidden buffer changes.
d9e94c22 3670 (let ((start (point)))
580fba94
AM
3671 (if (looking-at "\\w\\|\\s_")
3672 (skip-syntax-backward "w_" back-limit)
3673 (when (< (skip-syntax-backward ".()" back-limit) 0)
3674 (while (let ((pos (or (and (looking-at c-nonsymbol-token-regexp)
3675 (match-end 0))
3676 ;; `c-nonsymbol-token-regexp' should always match
3677 ;; since we've skipped backward over punctuator
3678 ;; or paren syntax, but consume one char in case
3679 ;; it doesn't so that we don't leave point before
3680 ;; some earlier incorrect token.
3681 (1+ (point)))))
3682 (if (<= pos start)
3683 (goto-char pos))))))
3684 (< (point) start)))
d9e94c22 3685
ff959bab 3686(defun c-end-of-current-token (&optional back-limit)
d9e94c22
MS
3687 ;; Move to the end of the current token. Do not move if not in the
3688 ;; middle of one. BACK-LIMIT may be used to bound the backward
3689 ;; search; if given it's assumed to be at the boundary between two
ff959bab 3690 ;; tokens. Return non-nil if the point is moved, nil otherwise.
0386b551
AM
3691 ;;
3692 ;; This function might do hidden buffer changes.
d9e94c22
MS
3693 (let ((start (point)))
3694 (cond ((< (skip-syntax-backward "w_" (1- start)) 0)
3695 (skip-syntax-forward "w_"))
3696 ((< (skip-syntax-backward ".()" back-limit) 0)
3697 (while (progn
3698 (if (looking-at c-nonsymbol-token-regexp)
3699 (goto-char (match-end 0))
3700 ;; `c-nonsymbol-token-regexp' should always match since
3701 ;; we've skipped backward over punctuator or paren
3702 ;; syntax, but move forward in case it doesn't so that
3703 ;; we don't leave point earlier than we started with.
3704 (forward-char))
ff959bab
MS
3705 (< (point) start)))))
3706 (> (point) start)))
d9e94c22
MS
3707
3708(defconst c-jump-syntax-balanced
3709 (if (memq 'gen-string-delim c-emacs-features)
3710 "\\w\\|\\s_\\|\\s\(\\|\\s\)\\|\\s\"\\|\\s|"
3711 "\\w\\|\\s_\\|\\s\(\\|\\s\)\\|\\s\""))
3712
3713(defconst c-jump-syntax-unbalanced
3714 (if (memq 'gen-string-delim c-emacs-features)
3715 "\\w\\|\\s_\\|\\s\"\\|\\s|"
3716 "\\w\\|\\s_\\|\\s\""))
3717
3718(defun c-forward-token-2 (&optional count balanced limit)
3719 "Move forward by tokens.
3720A token is defined as all symbols and identifiers which aren't
3721syntactic whitespace \(note that multicharacter tokens like \"==\" are
3722treated properly). Point is always either left at the beginning of a
3723token or not moved at all. COUNT specifies the number of tokens to
3724move; a negative COUNT moves in the opposite direction. A COUNT of 0
3725moves to the next token beginning only if not already at one. If
3726BALANCED is true, move over balanced parens, otherwise move into them.
3727Also, if BALANCED is true, never move out of an enclosing paren.
3728
3729LIMIT sets the limit for the movement and defaults to the point limit.
3730The case when LIMIT is set in the middle of a token, comment or macro
3731is handled correctly, i.e. the point won't be left there.
3732
3733Return the number of tokens left to move \(positive or negative). If
3734BALANCED is true, a move over a balanced paren counts as one. Note
3735that if COUNT is 0 and no appropriate token beginning is found, 1 will
3736be returned. Thus, a return value of 0 guarantees that point is at
3737the requested position and a return value less \(without signs) than
0386b551
AM
3738COUNT guarantees that point is at the beginning of some token.
3739
3740Note that this function might do hidden buffer changes. See the
3741comment at the start of cc-engine.el for more info."
d9e94c22
MS
3742
3743 (or count (setq count 1))
3744 (if (< count 0)
3745 (- (c-backward-token-2 (- count) balanced limit))
3746
3747 (let ((jump-syntax (if balanced
3748 c-jump-syntax-balanced
3749 c-jump-syntax-unbalanced))
3750 (last (point))
3751 (prev (point)))
3752
3753 (if (zerop count)
3754 ;; If count is zero we should jump if in the middle of a token.
3755 (c-end-of-current-token))
3756
3757 (save-restriction
3758 (if limit (narrow-to-region (point-min) limit))
3759 (if (/= (point)
3760 (progn (c-forward-syntactic-ws) (point)))
3761 ;; Skip whitespace. Count this as a move if we did in
3762 ;; fact move.
3763 (setq count (max (1- count) 0)))
3764
3765 (if (eobp)
3766 ;; Moved out of bounds. Make sure the returned count isn't zero.
3767 (progn
3768 (if (zerop count) (setq count 1))
3769 (goto-char last))
3770
3771 ;; Use `condition-case' to avoid having the limit tests
3772 ;; inside the loop.
3773 (condition-case nil
3774 (while (and
3775 (> count 0)
3776 (progn
3777 (setq last (point))
3778 (cond ((looking-at jump-syntax)
3779 (goto-char (scan-sexps (point) 1))
3780 t)
3781 ((looking-at c-nonsymbol-token-regexp)
3782 (goto-char (match-end 0))
3783 t)
3784 ;; `c-nonsymbol-token-regexp' above should always
3785 ;; match if there are correct tokens. Try to
3786 ;; widen to see if the limit was set in the
3787 ;; middle of one, else fall back to treating
3788 ;; the offending thing as a one character token.
3789 ((and limit
3790 (save-restriction
3791 (widen)
3792 (looking-at c-nonsymbol-token-regexp)))
3793 nil)
3794 (t
3795 (forward-char)
3796 t))))
3797 (c-forward-syntactic-ws)
3798 (setq prev last
3799 count (1- count)))
3800 (error (goto-char last)))
3801
3802 (when (eobp)
3803 (goto-char prev)
3804 (setq count (1+ count)))))
3805
3806 count)))
3807
3808(defun c-backward-token-2 (&optional count balanced limit)
3809 "Move backward by tokens.
3810See `c-forward-token-2' for details."
3811
3812 (or count (setq count 1))
3813 (if (< count 0)
3814 (- (c-forward-token-2 (- count) balanced limit))
3815
3816 (or limit (setq limit (point-min)))
3817 (let ((jump-syntax (if balanced
3818 c-jump-syntax-balanced
3819 c-jump-syntax-unbalanced))
3820 (last (point)))
3821
3822 (if (zerop count)
3823 ;; The count is zero so try to skip to the beginning of the
3824 ;; current token.
3825 (if (> (point)
3826 (progn (c-beginning-of-current-token) (point)))
3827 (if (< (point) limit)
3828 ;; The limit is inside the same token, so return 1.
3829 (setq count 1))
3830
3831 ;; We're not in the middle of a token. If there's
3832 ;; whitespace after the point then we must move backward,
3833 ;; so set count to 1 in that case.
3834 (and (looking-at c-syntactic-ws-start)
3835 ;; If we're looking at a '#' that might start a cpp
3836 ;; directive then we have to do a more elaborate check.
3837 (or (/= (char-after) ?#)
3838 (not c-opt-cpp-prefix)
3839 (save-excursion
3840 (and (= (point)
3841 (progn (beginning-of-line)
3842 (looking-at "[ \t]*")
3843 (match-end 0)))
3844 (or (bobp)
3845 (progn (backward-char)
3846 (not (eq (char-before) ?\\)))))))
3847 (setq count 1))))
3848
3849 ;; Use `condition-case' to avoid having to check for buffer
3850 ;; limits in `backward-char', `scan-sexps' and `goto-char' below.
3851 (condition-case nil
3852 (while (and
3853 (> count 0)
3854 (progn
3855 (c-backward-syntactic-ws)
3856 (backward-char)
3857 (if (looking-at jump-syntax)
3858 (goto-char (scan-sexps (1+ (point)) -1))
3859 ;; This can be very inefficient if there's a long
3860 ;; sequence of operator tokens without any separation.
3861 ;; That doesn't happen in practice, anyway.
3862 (c-beginning-of-current-token))
3863 (>= (point) limit)))
3864 (setq last (point)
3865 count (1- count)))
3866 (error (goto-char last)))
3867
3868 (if (< (point) limit)
3869 (goto-char last))
3870
3871 count)))
3872
3873(defun c-forward-token-1 (&optional count balanced limit)
3874 "Like `c-forward-token-2' but doesn't treat multicharacter operator
3875tokens like \"==\" as single tokens, i.e. all sequences of symbol
3876characters are jumped over character by character. This function is
3877for compatibility only; it's only a wrapper over `c-forward-token-2'."
3878 (let ((c-nonsymbol-token-regexp "\\s.\\|\\s\(\\|\\s\)"))
3879 (c-forward-token-2 count balanced limit)))
3880
3881(defun c-backward-token-1 (&optional count balanced limit)
3882 "Like `c-backward-token-2' but doesn't treat multicharacter operator
3883tokens like \"==\" as single tokens, i.e. all sequences of symbol
3884characters are jumped over character by character. This function is
3885for compatibility only; it's only a wrapper over `c-backward-token-2'."
3886 (let ((c-nonsymbol-token-regexp "\\s.\\|\\s\(\\|\\s\)"))
3887 (c-backward-token-2 count balanced limit)))
3888
3889\f
3890;; Tools for doing searches restricted to syntactically relevant text.
3891
3892(defun c-syntactic-re-search-forward (regexp &optional bound noerror
3893 paren-level not-inside-token
3894 lookbehind-submatch)
3895 "Like `re-search-forward', but only report matches that are found
3896in syntactically significant text. I.e. matches in comments, macros
3897or string literals are ignored. The start point is assumed to be
3898outside any comment, macro or string literal, or else the content of
3899that region is taken as syntactically significant text.
3900
3901If PAREN-LEVEL is non-nil, an additional restriction is added to
2a15eb73
MS
3902ignore matches in nested paren sexps. The search will also not go
3903outside the current list sexp, which has the effect that if the point
3904should be moved to BOUND when no match is found \(i.e. NOERROR is
3905neither nil nor t), then it will be at the closing paren if the end of
3906the current list sexp is encountered first.
d9e94c22
MS
3907
3908If NOT-INSIDE-TOKEN is non-nil, matches in the middle of tokens are
3909ignored. Things like multicharacter operators and special symbols
3910\(e.g. \"`()\" in Pike) are handled but currently not floating point
3911constants.
3912
3913If LOOKBEHIND-SUBMATCH is non-nil, it's taken as a number of a
3914subexpression in REGEXP. The end of that submatch is used as the
3915position to check for syntactic significance. If LOOKBEHIND-SUBMATCH
3916isn't used or if that subexpression didn't match then the start
3917position of the whole match is used instead. The \"look behind\"
3918subexpression is never tested before the starting position, so it
3919might be a good idea to include \\=\\= as a match alternative in it.
3920
3921Optimization note: Matches might be missed if the \"look behind\"
2a15eb73 3922subexpression can match the end of nonwhite syntactic whitespace,
d9e94c22 3923i.e. the end of comments or cpp directives. This since the function
2a15eb73
MS
3924skips over such things before resuming the search. It's on the other
3925hand not safe to assume that the \"look behind\" subexpression never
3926matches syntactic whitespace.
3927
3928Bug: Unbalanced parens inside cpp directives are currently not handled
3929correctly \(i.e. they don't get ignored as they should) when
0386b551
AM
3930PAREN-LEVEL is set.
3931
3932Note that this function might do hidden buffer changes. See the
3933comment at the start of cc-engine.el for more info."
d9e94c22
MS
3934
3935 (or bound (setq bound (point-max)))
3936 (if paren-level (setq paren-level -1))
3937
3938 ;;(message "c-syntactic-re-search-forward %s %s %S" (point) bound regexp)
3939
3940 (let ((start (point))
2a15eb73
MS
3941 tmp
3942 ;; Start position for the last search.
3943 search-pos
3944 ;; The `parse-partial-sexp' state between the start position
3945 ;; and the point.
3946 state
3947 ;; The current position after the last state update. The next
3948 ;; `parse-partial-sexp' continues from here.
3949 (state-pos (point))
3950 ;; The position at which to check the state and the state
3951 ;; there. This is separate from `state-pos' since we might
3952 ;; need to back up before doing the next search round.
3953 check-pos check-state
3954 ;; Last position known to end a token.
d9e94c22 3955 (last-token-end-pos (point-min))
2a15eb73
MS
3956 ;; Set when a valid match is found.
3957 found)
d9e94c22
MS
3958
3959 (condition-case err
3960 (while
3961 (and
2a15eb73
MS
3962 (progn
3963 (setq search-pos (point))
3964 (re-search-forward regexp bound noerror))
d9e94c22
MS
3965
3966 (progn
2a15eb73
MS
3967 (setq state (parse-partial-sexp
3968 state-pos (match-beginning 0) paren-level nil state)
3969 state-pos (point))
d9e94c22 3970 (if (setq check-pos (and lookbehind-submatch
2a15eb73
MS
3971 (or (not paren-level)
3972 (>= (car state) 0))
d9e94c22
MS
3973 (match-end lookbehind-submatch)))
3974 (setq check-state (parse-partial-sexp
2a15eb73
MS
3975 state-pos check-pos paren-level nil state))
3976 (setq check-pos state-pos
d9e94c22
MS
3977 check-state state))
3978
2a15eb73
MS
3979 ;; NOTE: If we got a look behind subexpression and get
3980 ;; an insignificant match in something that isn't
d9e94c22
MS
3981 ;; syntactic whitespace (i.e. strings or in nested
3982 ;; parentheses), then we can never skip more than a
2a15eb73
MS
3983 ;; single character from the match start position
3984 ;; (i.e. `state-pos' here) before continuing the
3985 ;; search. That since the look behind subexpression
3986 ;; might match the end of the insignificant region in
3987 ;; the next search.
d9e94c22
MS
3988
3989 (cond
d9e94c22
MS
3990 ((elt check-state 7)
3991 ;; Match inside a line comment. Skip to eol. Use
3992 ;; `re-search-forward' instead of `skip-chars-forward' to get
3993 ;; the right bound behavior.
3994 (re-search-forward "[\n\r]" bound noerror))
3995
3996 ((elt check-state 4)
3997 ;; Match inside a block comment. Skip to the '*/'.
3998 (search-forward "*/" bound noerror))
3999
4000 ((and (not (elt check-state 5))
4001 (eq (char-before check-pos) ?/)
2a15eb73 4002 (not (c-get-char-property (1- check-pos) 'syntax-table))
d9e94c22
MS
4003 (memq (char-after check-pos) '(?/ ?*)))
4004 ;; Match in the middle of the opener of a block or line
4005 ;; comment.
4006 (if (= (char-after check-pos) ?/)
4007 (re-search-forward "[\n\r]" bound noerror)
4008 (search-forward "*/" bound noerror)))
4009
2a15eb73
MS
4010 ;; The last `parse-partial-sexp' above might have
4011 ;; stopped short of the real check position if the end
4012 ;; of the current sexp was encountered in paren-level
4013 ;; mode. The checks above are always false in that
4014 ;; case, and since they can do better skipping in
4015 ;; lookbehind-submatch mode, we do them before
4016 ;; checking the paren level.
4017
4018 ((and paren-level
4019 (/= (setq tmp (car check-state)) 0))
4020 ;; Check the paren level first since we're short of the
4021 ;; syntactic checking position if the end of the
4022 ;; current sexp was encountered by `parse-partial-sexp'.
4023 (if (> tmp 0)
4024
4025 ;; Inside a nested paren sexp.
4026 (if lookbehind-submatch
4027 ;; See the NOTE above.
4028 (progn (goto-char state-pos) t)
4029 ;; Skip out of the paren quickly.
4030 (setq state (parse-partial-sexp state-pos bound 0 nil state)
4031 state-pos (point)))
4032
4033 ;; Have exited the current paren sexp.
4034 (if noerror
4035 (progn
4036 ;; The last `parse-partial-sexp' call above
4037 ;; has left us just after the closing paren
4038 ;; in this case, so we can modify the bound
4039 ;; to leave the point at the right position
4040 ;; upon return.
4041 (setq bound (1- (point)))
4042 nil)
4043 (signal 'search-failed (list regexp)))))
4044
4045 ((setq tmp (elt check-state 3))
4046 ;; Match inside a string.
4047 (if (or lookbehind-submatch
4048 (not (integerp tmp)))
4049 ;; See the NOTE above.
4050 (progn (goto-char state-pos) t)
4051 ;; Skip to the end of the string before continuing.
4052 (let ((ender (make-string 1 tmp)) (continue t))
4053 (while (if (search-forward ender bound noerror)
4054 (progn
4055 (setq state (parse-partial-sexp
4056 state-pos (point) nil nil state)
4057 state-pos (point))
4058 (elt state 3))
4059 (setq continue nil)))
4060 continue)))
d9e94c22
MS
4061
4062 ((save-excursion
4063 (save-match-data
4064 (c-beginning-of-macro start)))
4065 ;; Match inside a macro. Skip to the end of it.
4066 (c-end-of-macro)
4067 (cond ((<= (point) bound) t)
4068 (noerror nil)
2a15eb73 4069 (t (signal 'search-failed (list regexp)))))
d9e94c22 4070
2a15eb73
MS
4071 ((and not-inside-token
4072 (or (< check-pos last-token-end-pos)
4073 (< check-pos
4074 (save-excursion
4075 (goto-char check-pos)
4076 (save-match-data
4077 (c-end-of-current-token last-token-end-pos))
4078 (setq last-token-end-pos (point))))))
4079 ;; Inside a token.
4080 (if lookbehind-submatch
4081 ;; See the NOTE above.
4082 (goto-char state-pos)
4083 (goto-char (min last-token-end-pos bound))))
d9e94c22
MS
4084
4085 (t
4086 ;; A real match.
4087 (setq found t)
2a15eb73
MS
4088 nil)))
4089
4090 ;; Should loop to search again, but take care to avoid
4091 ;; looping on the same spot.
4092 (or (/= search-pos (point))
4093 (if (= (point) bound)
4094 (if noerror
4095 nil
4096 (signal 'search-failed (list regexp)))
4097 (forward-char)
4098 t))))
d9e94c22
MS
4099
4100 (error
4101 (goto-char start)
4102 (signal (car err) (cdr err))))
4103
2a15eb73 4104 ;;(message "c-syntactic-re-search-forward done %s" (or (match-end 0) (point)))
d9e94c22
MS
4105
4106 (if found
4107 (progn
2a15eb73
MS
4108 (goto-char (match-end 0))
4109 (match-end 0))
d9e94c22
MS
4110
4111 ;; Search failed. Set point as appropriate.
2a15eb73
MS
4112 (if (eq noerror t)
4113 (goto-char start)
4114 (goto-char bound))
d9e94c22
MS
4115 nil)))
4116
47641aac
GM
4117(defvar safe-pos-list) ; bound in c-syntactic-skip-backward
4118
d0fcee66
AM
4119(defsubst c-ssb-lit-begin ()
4120 ;; Return the start of the literal point is in, or nil.
4121 ;; We read and write the variables `safe-pos', `safe-pos-list', `state'
4122 ;; bound in the caller.
4123
4124 ;; Use `parse-partial-sexp' from a safe position down to the point to check
4125 ;; if it's outside comments and strings.
4126 (save-excursion
4127 (let ((pos (point)) safe-pos state pps-end-pos)
4128 ;; Pick a safe position as close to the point as possible.
4129 ;;
4130 ;; FIXME: Consult `syntax-ppss' here if our cache doesn't give a good
4131 ;; position.
47641aac 4132
d0fcee66
AM
4133 (while (and safe-pos-list
4134 (> (car safe-pos-list) (point)))
4135 (setq safe-pos-list (cdr safe-pos-list)))
4136 (unless (setq safe-pos (car-safe safe-pos-list))
4137 (setq safe-pos (max (or (c-safe-position
4138 (point) (or c-state-cache
4139 (c-parse-state)))
4140 0)
4141 (point-min))
4142 safe-pos-list (list safe-pos)))
4143
4144 ;; Cache positions along the way to use if we have to back up more. We
4145 ;; cache every closing paren on the same level. If the paren cache is
4146 ;; relevant in this region then we're typically already on the same
4147 ;; level as the target position. Note that we might cache positions
4148 ;; after opening parens in case safe-pos is in a nested list. That's
4149 ;; both uncommon and harmless.
4150 (while (progn
4151 (setq state (parse-partial-sexp
4152 safe-pos pos 0))
4153 (< (point) pos))
4154 (setq safe-pos (point)
4155 safe-pos-list (cons safe-pos safe-pos-list)))
4156
4157 ;; If the state contains the start of the containing sexp we cache that
4158 ;; position too, so that parse-partial-sexp in the next run has a bigger
4159 ;; chance of starting at the same level as the target position and thus
4160 ;; will get more good safe positions into the list.
4161 (if (elt state 1)
4162 (setq safe-pos (1+ (elt state 1))
4163 safe-pos-list (cons safe-pos safe-pos-list)))
4164
4165 (if (or (elt state 3) (elt state 4))
4166 ;; Inside string or comment. Continue search at the
4167 ;; beginning of it.
4168 (elt state 8)))))
4169
0386b551 4170(defun c-syntactic-skip-backward (skip-chars &optional limit paren-level)
d9e94c22
MS
4171 "Like `skip-chars-backward' but only look at syntactically relevant chars,
4172i.e. don't stop at positions inside syntactic whitespace or string
4173literals. Preprocessor directives are also ignored, with the exception
4174of the one that the point starts within, if any. If LIMIT is given,
0386b551
AM
4175it's assumed to be at a syntactically relevant position.
4176
4177If PAREN-LEVEL is non-nil, the function won't stop in nested paren
4178sexps, and the search will also not go outside the current paren sexp.
4179However, if LIMIT or the buffer limit is reached inside a nested paren
4180then the point will be left at the limit.
4181
4182Non-nil is returned if the point moved, nil otherwise.
4183
4184Note that this function might do hidden buffer changes. See the
4185comment at the start of cc-engine.el for more info."
d9e94c22
MS
4186
4187 (let ((start (point))
d0fcee66 4188 state-2
d9e94c22
MS
4189 ;; A list of syntactically relevant positions in descending
4190 ;; order. It's used to avoid scanning repeatedly over
4191 ;; potentially large regions with `parse-partial-sexp' to verify
d0fcee66 4192 ;; each position. Used in `c-ssb-lit-begin'
d9e94c22
MS
4193 safe-pos-list
4194 ;; The result from `c-beginning-of-macro' at the start position or the
4195 ;; start position itself if it isn't within a macro. Evaluated on
4196 ;; demand.
0386b551
AM
4197 start-macro-beg
4198 ;; The earliest position after the current one with the same paren
4199 ;; level. Used only when `paren-level' is set.
d0fcee66 4200 lit-beg
0386b551 4201 (paren-level-pos (point)))
d9e94c22 4202
d0fcee66
AM
4203 (while
4204 (progn
4205 ;; The next loop "tries" to find the end point each time round,
4206 ;; loops when it hasn't succeeded.
4207 (while
4208 (and
4209 (< (skip-chars-backward skip-chars limit) 0)
4210
4211 (let ((pos (point)) state-2 pps-end-pos)
4212
4213 (cond
4214 ;; Don't stop inside a literal
4215 ((setq lit-beg (c-ssb-lit-begin))
4216 (goto-char lit-beg)
4217 t)
4218
4219 ((and paren-level
4220 (save-excursion
4221 (setq state-2 (parse-partial-sexp
4222 pos paren-level-pos -1)
4223 pps-end-pos (point))
4224 (/= (car state-2) 0)))
4225 ;; Not at the right level.
4226
4227 (if (and (< (car state-2) 0)
4228 ;; We stop above if we go out of a paren.
4229 ;; Now check whether it precedes or is
4230 ;; nested in the starting sexp.
4231 (save-excursion
4232 (setq state-2
4233 (parse-partial-sexp
4234 pps-end-pos paren-level-pos
4235 nil nil state-2))
4236 (< (car state-2) 0)))
4237
4238 ;; We've stopped short of the starting position
4239 ;; so the hit was inside a nested list. Go up
4240 ;; until we are at the right level.
4241 (condition-case nil
4242 (progn
4243 (goto-char (scan-lists pos -1
4244 (- (car state-2))))
4245 (setq paren-level-pos (point))
4246 (if (and limit (>= limit paren-level-pos))
0386b551 4247 (progn
d0fcee66
AM
4248 (goto-char limit)
4249 nil)
4250 t))
4251 (error
4252 (goto-char (or limit (point-min)))
4253 nil))
4254
4255 ;; The hit was outside the list at the start
4256 ;; position. Go to the start of the list and exit.
4257 (goto-char (1+ (elt state-2 1)))
4258 nil))
4259
4260 ((c-beginning-of-macro limit)
4261 ;; Inside a macro.
4262 (if (< (point)
4263 (or start-macro-beg
4264 (setq start-macro-beg
4265 (save-excursion
4266 (goto-char start)
4267 (c-beginning-of-macro limit)
4268 (point)))))
4269 t
4270
4271 ;; It's inside the same macro we started in so it's
4272 ;; a relevant match.
4273 (goto-char pos)
4274 nil))))))
91af3942 4275
d0fcee66
AM
4276 (> (point)
4277 (progn
4278 ;; Skip syntactic ws afterwards so that we don't stop at the
4279 ;; end of a comment if `skip-chars' is something like "^/".
4280 (c-backward-syntactic-ws)
4281 (point)))))
d9e94c22 4282
0386b551
AM
4283 ;; We might want to extend this with more useful return values in
4284 ;; the future.
4285 (/= (point) start)))
4286
4287;; The following is an alternative implementation of
4288;; `c-syntactic-skip-backward' that uses backward movement to keep
4289;; track of the syntactic context. It turned out to be generally
4290;; slower than the one above which uses forward checks from earlier
4291;; safe positions.
4292;;
4293;;(defconst c-ssb-stop-re
4294;; ;; The regexp matching chars `c-syntactic-skip-backward' needs to
4295;; ;; stop at to avoid going into comments and literals.
4296;; (concat
4297;; ;; Match comment end syntax and string literal syntax. Also match
4298;; ;; '/' for block comment endings (not covered by comment end
4299;; ;; syntax).
4300;; "\\s>\\|/\\|\\s\""
4301;; (if (memq 'gen-string-delim c-emacs-features)
4302;; "\\|\\s|"
4303;; "")
4304;; (if (memq 'gen-comment-delim c-emacs-features)
4305;; "\\|\\s!"
4306;; "")))
4307;;
4308;;(defconst c-ssb-stop-paren-re
4309;; ;; Like `c-ssb-stop-re' but also stops at paren chars.
4310;; (concat c-ssb-stop-re "\\|\\s(\\|\\s)"))
4311;;
4312;;(defconst c-ssb-sexp-end-re
4313;; ;; Regexp matching the ending syntax of a complex sexp.
4314;; (concat c-string-limit-regexp "\\|\\s)"))
4315;;
4316;;(defun c-syntactic-skip-backward (skip-chars &optional limit paren-level)
4317;; "Like `skip-chars-backward' but only look at syntactically relevant chars,
4318;;i.e. don't stop at positions inside syntactic whitespace or string
4319;;literals. Preprocessor directives are also ignored. However, if the
4320;;point is within a comment, string literal or preprocessor directory to
4321;;begin with, its contents is treated as syntactically relevant chars.
4322;;If LIMIT is given, it limits the backward search and the point will be
4323;;left there if no earlier position is found.
4324;;
4325;;If PAREN-LEVEL is non-nil, the function won't stop in nested paren
4326;;sexps, and the search will also not go outside the current paren sexp.
4327;;However, if LIMIT or the buffer limit is reached inside a nested paren
4328;;then the point will be left at the limit.
4329;;
4330;;Non-nil is returned if the point moved, nil otherwise.
4331;;
4332;;Note that this function might do hidden buffer changes. See the
4333;;comment at the start of cc-engine.el for more info."
4334;;
4335;; (save-restriction
4336;; (when limit
4337;; (narrow-to-region limit (point-max)))
4338;;
4339;; (let ((start (point)))
4340;; (catch 'done
4341;; (while (let ((last-pos (point))
4342;; (stop-pos (progn
4343;; (skip-chars-backward skip-chars)
4344;; (point))))
4345;;
4346;; ;; Skip back over the same region as
4347;; ;; `skip-chars-backward' above, but keep to
4348;; ;; syntactically relevant positions.
4349;; (goto-char last-pos)
4350;; (while (and
4351;; ;; `re-search-backward' with a single char regexp
4352;; ;; should be fast.
4353;; (re-search-backward
4354;; (if paren-level c-ssb-stop-paren-re c-ssb-stop-re)
4355;; stop-pos 'move)
4356;;
4357;; (progn
4358;; (cond
4359;; ((looking-at "\\s(")
4360;; ;; `paren-level' is set and we've found the
4361;; ;; start of the containing paren.
4362;; (forward-char)
4363;; (throw 'done t))
4364;;
4365;; ((looking-at c-ssb-sexp-end-re)
4366;; ;; We're at the end of a string literal or paren
4367;; ;; sexp (if `paren-level' is set).
4368;; (forward-char)
4369;; (condition-case nil
4370;; (c-backward-sexp)
4371;; (error
4372;; (goto-char limit)
4373;; (throw 'done t))))
4374;;
4375;; (t
4376;; (forward-char)
4377;; ;; At the end of some syntactic ws or possibly
4378;; ;; after a plain '/' operator.
4379;; (let ((pos (point)))
4380;; (c-backward-syntactic-ws)
4381;; (if (= pos (point))
4382;; ;; Was a plain '/' operator. Go past it.
4383;; (backward-char)))))
4384;;
4385;; (> (point) stop-pos))))
4386;;
4387;; ;; Now the point is either at `stop-pos' or at some
4388;; ;; position further back if `stop-pos' was at a
4389;; ;; syntactically irrelevant place.
4390;;
4391;; ;; Skip additional syntactic ws so that we don't stop
4392;; ;; at the end of a comment if `skip-chars' is
4393;; ;; something like "^/".
4394;; (c-backward-syntactic-ws)
4395;;
4396;; (< (point) stop-pos))))
4397;;
4398;; ;; We might want to extend this with more useful return values
4399;; ;; in the future.
4400;; (/= (point) start))))
d9e94c22
MS
4401
4402\f
4403;; Tools for handling comments and string literals.
4404
13d49cbb 4405(defun c-in-literal (&optional lim detect-cpp)
d9e94c22
MS
4406 "Return the type of literal point is in, if any.
4407The return value is `c' if in a C-style comment, `c++' if in a C++
4408style comment, `string' if in a string literal, `pound' if DETECT-CPP
4409is non-nil and in a preprocessor line, or nil if somewhere else.
4410Optional LIM is used as the backward limit of the search. If omitted,
4411or nil, `c-beginning-of-defun' is used.
4412
4413The last point calculated is cached if the cache is enabled, i.e. if
4414`c-in-literal-cache' is bound to a two element vector.
4415
0386b551
AM
4416Note that this function might do hidden buffer changes. See the
4417comment at the start of cc-engine.el for more info."
2cc769a8
AM
4418 (save-restriction
4419 (widen)
56d093a9 4420 (let* ((safe-place (c-state-semi-safe-place (point)))
2cc769a8
AM
4421 (lit (c-state-pp-to-literal safe-place (point))))
4422 (or (cadr lit)
4423 (and detect-cpp
4424 (save-excursion (c-beginning-of-macro))
4425 'pound)))))
d9e94c22
MS
4426
4427(defun c-literal-limits (&optional lim near not-in-delimiter)
4428 "Return a cons of the beginning and end positions of the comment or
4429string surrounding point (including both delimiters), or nil if point
4430isn't in one. If LIM is non-nil, it's used as the \"safe\" position
4431to start parsing from. If NEAR is non-nil, then the limits of any
4432literal next to point is returned. \"Next to\" means there's only
4433spaces and tabs between point and the literal. The search for such a
4434literal is done first in forward direction. If NOT-IN-DELIMITER is
4435non-nil, the case when point is inside a starting delimiter won't be
a85fd6da 4436recognized. This only has effect for comments which have starting
d9e94c22
MS
4437delimiters with more than one character.
4438
0386b551
AM
4439Note that this function might do hidden buffer changes. See the
4440comment at the start of cc-engine.el for more info."
d9e94c22
MS
4441
4442 (save-excursion
4443 (let* ((pos (point))
56d093a9 4444 (lim (or lim (c-state-semi-safe-place pos)))
2cc769a8
AM
4445 (pp-to-lit (save-restriction
4446 (widen)
4447 (c-state-pp-to-literal lim pos)))
13d49cbb 4448 (state (car pp-to-lit))
13d49cbb 4449 (lit-limits (car (cddr pp-to-lit))))
d9e94c22 4450
13d49cbb
AM
4451 (cond
4452 (lit-limits)
4453 ((and (not not-in-delimiter)
4454 (not (elt state 5))
4455 (eq (char-before) ?/)
4456 (looking-at "[/*]")) ; FIXME!!! use c-line/block-comment-starter. 2008-09-28.
4457 ;; We're standing in a comment starter.
4458 (backward-char 1)
4459 (cons (point) (progn (c-forward-single-comment) (point))))
4460
4461 (near
4462 (goto-char pos)
4463 ;; Search forward for a literal.
4464 (skip-chars-forward " \t")
4465 (cond
4466 ((looking-at c-string-limit-regexp) ; String.
4467 (cons (point) (or (c-safe (c-forward-sexp 1) (point))
4468 (point-max))))
d9e94c22 4469
13d49cbb
AM
4470 ((looking-at c-comment-start-regexp) ; Line or block comment.
4471 (cons (point) (progn (c-forward-single-comment) (point))))
4472
4473 (t
4474 ;; Search backward.
4475 (skip-chars-backward " \t")
4476
4477 (let ((end (point)) beg)
4478 (cond
4479 ((save-excursion
4480 (< (skip-syntax-backward c-string-syntax) 0)) ; String.
4481 (setq beg (c-safe (c-backward-sexp 1) (point))))
4482
4483 ((and (c-safe (forward-char -2) t)
4484 (looking-at "*/"))
4485 ;; Block comment. Due to the nature of line
4486 ;; comments, they will always be covered by the
4487 ;; normal case above.
4488 (goto-char end)
4489 (c-backward-single-comment)
4490 ;; If LIM is bogus, beg will be bogus.
4491 (setq beg (point))))
4492
4493 (if beg (cons beg end))))))
4494 ))))
d9e94c22 4495
0386b551
AM
4496;; In case external callers use this; it did have a docstring.
4497(defalias 'c-literal-limits-fast 'c-literal-limits)
d9e94c22 4498
0386b551
AM
4499(defun c-collect-line-comments (range)
4500 "If the argument is a cons of two buffer positions (such as returned by
4501`c-literal-limits'), and that range contains a C++ style line comment,
4502then an extended range is returned that contains all adjacent line
4503comments (i.e. all comments that starts in the same column with no
4504empty lines or non-whitespace characters between them). Otherwise the
4505argument is returned.
d9e94c22 4506
0386b551
AM
4507Note that this function might do hidden buffer changes. See the
4508comment at the start of cc-engine.el for more info."
d9e94c22
MS
4509
4510 (save-excursion
0386b551
AM
4511 (condition-case nil
4512 (if (and (consp range) (progn
4513 (goto-char (car range))
4514 (looking-at c-line-comment-starter)))
b414f371 4515 (let ((col (current-column))
0386b551
AM
4516 (beg (point))
4517 (bopl (c-point 'bopl))
4518 (end (cdr range)))
4519 ;; Got to take care in the backward direction to handle
4520 ;; comments which are preceded by code.
4521 (while (and (c-backward-single-comment)
4522 (>= (point) bopl)
4523 (looking-at c-line-comment-starter)
4524 (= col (current-column)))
4525 (setq beg (point)
4526 bopl (c-point 'bopl)))
4527 (goto-char end)
4528 (while (and (progn (skip-chars-forward " \t")
4529 (looking-at c-line-comment-starter))
4530 (= col (current-column))
4531 (prog1 (zerop (forward-line 1))
4532 (setq end (point)))))
4533 (cons beg end))
4534 range)
4535 (error range))))
d9e94c22
MS
4536
4537(defun c-literal-type (range)
4538 "Convenience function that given the result of `c-literal-limits',
a85fd6da
AM
4539returns nil or the type of literal that the range surrounds, one
4540of the symbols 'c, 'c++ or 'string. It's much faster than using
4541`c-in-literal' and is intended to be used when you need both the
4542type of a literal and its limits.
d9e94c22 4543
0386b551
AM
4544Note that this function might do hidden buffer changes. See the
4545comment at the start of cc-engine.el for more info."
4546
d9e94c22
MS
4547 (if (consp range)
4548 (save-excursion
4549 (goto-char (car range))
4550 (cond ((looking-at c-string-limit-regexp) 'string)
4551 ((or (looking-at "//") ; c++ line comment
4552 (and (looking-at "\\s<") ; comment starter
4553 (looking-at "#"))) ; awk comment.
4554 'c++)
4555 (t 'c))) ; Assuming the range is valid.
4556 range))
4557
9657183b
AM
4558(defsubst c-determine-limit-get-base (start try-size)
4559 ;; Get a "safe place" approximately TRY-SIZE characters before START.
4560 ;; This doesn't preserve point.
4561 (let* ((pos (max (- start try-size) (point-min)))
56d093a9 4562 (base (c-state-semi-safe-place pos))
9657183b
AM
4563 (s (parse-partial-sexp base pos)))
4564 (if (or (nth 4 s) (nth 3 s)) ; comment or string
4565 (nth 8 s)
4566 (point))))
4567
4568(defun c-determine-limit (how-far-back &optional start try-size)
4569 ;; Return a buffer position HOW-FAR-BACK non-literal characters from START
4570 ;; (default point). This is done by going back further in the buffer then
4571 ;; searching forward for literals. The position found won't be in a
4572 ;; literal. We start searching for the sought position TRY-SIZE (default
4573 ;; twice HOW-FAR-BACK) bytes back from START. This function must be fast.
4574 ;; :-)
4575 (save-excursion
4576 (let* ((start (or start (point)))
4577 (try-size (or try-size (* 2 how-far-back)))
4578 (base (c-determine-limit-get-base start try-size))
4579 (pos base)
4580
4581 (s (parse-partial-sexp pos pos)) ; null state.
4582 stack elt size
4583 (count 0))
4584 (while (< pos start)
4585 ;; Move forward one literal each time round this loop.
4586 ;; Move forward to the start of a comment or string.
4587 (setq s (parse-partial-sexp
4588 pos
4589 start
4590 nil ; target-depth
4591 nil ; stop-before
4592 s ; state
4593 'syntax-table)) ; stop-comment
4594
4595 ;; Gather details of the non-literal-bit - starting pos and size.
4596 (setq size (- (if (or (nth 4 s) (nth 3 s))
4597 (nth 8 s)
4598 (point))
4599 pos))
4600 (if (> size 0)
4601 (setq stack (cons (cons pos size) stack)))
4602
4603 ;; Move forward to the end of the comment/string.
4604 (if (or (nth 4 s) (nth 3 s))
4605 (setq s (parse-partial-sexp
4606 (point)
4607 start
4608 nil ; target-depth
4609 nil ; stop-before
4610 s ; state
4611 'syntax-table))) ; stop-comment
4612 (setq pos (point)))
4613
4614 ;; Now try and find enough non-literal characters recorded on the stack.
4615 ;; Go back one recorded literal each time round this loop.
4616 (while (and (< count how-far-back)
4617 stack)
4618 (setq elt (car stack)
4619 stack (cdr stack))
4620 (setq count (+ count (cdr elt))))
4621
4622 ;; Have we found enough yet?
4623 (cond
4624 ((>= count how-far-back)
4625 (+ (car elt) (- count how-far-back)))
4626 ((eq base (point-min))
4627 (point-min))
4628 (t
4629 (c-determine-limit (- how-far-back count) base try-size))))))
2e492df3
AM
4630
4631(defun c-determine-+ve-limit (how-far &optional start-pos)
4632 ;; Return a buffer position about HOW-FAR non-literal characters forward
4633 ;; from START-POS (default point), which must not be inside a literal.
4634 (save-excursion
4635 (let ((pos (or start-pos (point)))
4636 (count how-far)
4637 (s (parse-partial-sexp (point) (point)))) ; null state
4638 (while (and (not (eobp))
4639 (> count 0))
4640 ;; Scan over counted characters.
4641 (setq s (parse-partial-sexp
4642 pos
4643 (min (+ pos count) (point-max))
4644 nil ; target-depth
4645 nil ; stop-before
4646 s ; state
4647 'syntax-table)) ; stop-comment
4648 (setq count (- count (- (point) pos) 1)
4649 pos (point))
4650 ;; Scan over literal characters.
4651 (if (nth 8 s)
4652 (setq s (parse-partial-sexp
4653 pos
4654 (point-max)
4655 nil ; target-depth
4656 nil ; stop-before
4657 s ; state
4658 'syntax-table) ; stop-comment
4659 pos (point))))
4660 (point))))
4661
d9e94c22
MS
4662\f
4663;; `c-find-decl-spots' and accompanying stuff.
4664
4665;; Variables used in `c-find-decl-spots' to cache the search done for
4666;; the first declaration in the last call. When that function starts,
4667;; it needs to back up over syntactic whitespace to look at the last
4668;; token before the region being searched. That can sometimes cause
4669;; moves back and forth over a quite large region of comments and
4670;; macros, which would be repeated for each changed character when
4671;; we're called during fontification, since font-lock refontifies the
4672;; current line for each change. Thus it's worthwhile to cache the
4673;; first match.
4674;;
4675;; `c-find-decl-syntactic-pos' is a syntactically relevant position in
4676;; the syntactic whitespace less or equal to some start position.
4677;; There's no cached value if it's nil.
4678;;
4679;; `c-find-decl-match-pos' is the match position if
4680;; `c-find-decl-prefix-search' matched before the syntactic whitespace
4681;; at `c-find-decl-syntactic-pos', or nil if there's no such match.
4682(defvar c-find-decl-syntactic-pos nil)
4683(make-variable-buffer-local 'c-find-decl-syntactic-pos)
4684(defvar c-find-decl-match-pos nil)
4685(make-variable-buffer-local 'c-find-decl-match-pos)
4686
4687(defsubst c-invalidate-find-decl-cache (change-min-pos)
4688 (and c-find-decl-syntactic-pos
4689 (< change-min-pos c-find-decl-syntactic-pos)
4690 (setq c-find-decl-syntactic-pos nil)))
4691
4692; (defface c-debug-decl-spot-face
4693; '((t (:background "Turquoise")))
4694; "Debug face to mark the spots where `c-find-decl-spots' stopped.")
4695; (defface c-debug-decl-sws-face
4696; '((t (:background "Khaki")))
4697; "Debug face to mark the syntactic whitespace between the declaration
4698; spots and the preceding token end.")
4699
4700(defmacro c-debug-put-decl-spot-faces (match-pos decl-pos)
4701 (when (facep 'c-debug-decl-spot-face)
0386b551 4702 `(c-save-buffer-state ((match-pos ,match-pos) (decl-pos ,decl-pos))
d9e94c22
MS
4703 (c-debug-add-face (max match-pos (point-min)) decl-pos
4704 'c-debug-decl-sws-face)
4705 (c-debug-add-face decl-pos (min (1+ decl-pos) (point-max))
4706 'c-debug-decl-spot-face))))
4707(defmacro c-debug-remove-decl-spot-faces (beg end)
4708 (when (facep 'c-debug-decl-spot-face)
0386b551 4709 `(c-save-buffer-state ()
d9e94c22
MS
4710 (c-debug-remove-face ,beg ,end 'c-debug-decl-spot-face)
4711 (c-debug-remove-face ,beg ,end 'c-debug-decl-sws-face))))
4712
4713(defmacro c-find-decl-prefix-search ()
4714 ;; Macro used inside `c-find-decl-spots'. It ought to be a defun,
4715 ;; but it contains lots of free variables that refer to things
4716 ;; inside `c-find-decl-spots'. The point is left at `cfd-match-pos'
4717 ;; if there is a match, otherwise at `cfd-limit'.
0386b551
AM
4718 ;;
4719 ;; This macro might do hidden buffer changes.
d9e94c22
MS
4720
4721 '(progn
4722 ;; Find the next property match position if we haven't got one already.
4723 (unless cfd-prop-match
4724 (save-excursion
4725 (while (progn
4726 (goto-char (next-single-property-change
4727 (point) 'c-type nil cfd-limit))
4728 (and (< (point) cfd-limit)
4729 (not (eq (c-get-char-property (1- (point)) 'c-type)
4730 'c-decl-end)))))
4731 (setq cfd-prop-match (point))))
4732
0386b551
AM
4733 ;; Find the next `c-decl-prefix-or-start-re' match if we haven't
4734 ;; got one already.
d9e94c22 4735 (unless cfd-re-match
0386b551
AM
4736
4737 (if (> cfd-re-match-end (point))
4738 (goto-char cfd-re-match-end))
4739
4740 (while (if (setq cfd-re-match-end
4741 (re-search-forward c-decl-prefix-or-start-re
4742 cfd-limit 'move))
4743
4744 ;; Match. Check if it's inside a comment or string literal.
4745 (c-got-face-at
4746 (if (setq cfd-re-match (match-end 1))
4747 ;; Matched the end of a token preceding a decl spot.
4748 (progn
4749 (goto-char cfd-re-match)
4750 (1- cfd-re-match))
4751 ;; Matched a token that start a decl spot.
4752 (goto-char (match-beginning 0))
4753 (point))
4754 c-literal-faces)
4755
4756 ;; No match. Finish up and exit the loop.
4757 (setq cfd-re-match cfd-limit)
4758 nil)
4759
4760 ;; Skip out of comments and string literals.
d9e94c22
MS
4761 (while (progn
4762 (goto-char (next-single-property-change
0386b551 4763 (point) 'face nil cfd-limit))
d9e94c22 4764 (and (< (point) cfd-limit)
0386b551
AM
4765 (c-got-face-at (point) c-literal-faces)))))
4766
4767 ;; If we matched at the decl start, we have to back up over the
4768 ;; preceding syntactic ws to set `cfd-match-pos' and to catch
4769 ;; any decl spots in the syntactic ws.
d9e94c22 4770 (unless cfd-re-match
0386b551
AM
4771 (c-backward-syntactic-ws)
4772 (setq cfd-re-match (point))))
d9e94c22
MS
4773
4774 ;; Choose whichever match is closer to the start.
4775 (if (< cfd-re-match cfd-prop-match)
4776 (setq cfd-match-pos cfd-re-match
4777 cfd-re-match nil)
4778 (setq cfd-match-pos cfd-prop-match
4779 cfd-prop-match nil))
4780
4781 (goto-char cfd-match-pos)
4782
4783 (when (< cfd-match-pos cfd-limit)
4784 ;; Skip forward past comments only so we don't skip macros.
4785 (c-forward-comments)
4786 ;; Set the position to continue at. We can avoid going over
4787 ;; the comments skipped above a second time, but it's possible
4788 ;; that the comment skipping has taken us past `cfd-prop-match'
4789 ;; since the property might be used inside comments.
4790 (setq cfd-continue-pos (if cfd-prop-match
4791 (min cfd-prop-match (point))
4792 (point))))))
4793
4794(defun c-find-decl-spots (cfd-limit cfd-decl-re cfd-face-checklist cfd-fun)
0386b551 4795 ;; Call CFD-FUN for each possible spot for a declaration, cast or
a85fd6da
AM
4796 ;; label from the point to CFD-LIMIT.
4797 ;;
9657183b
AM
4798 ;; CFD-FUN is called with point at the start of the spot. It's passed two
4799 ;; arguments: The first is the end position of the token preceding the spot,
4800 ;; or 0 for the implicit match at bob. The second is a flag that is t when
4801 ;; the match is inside a macro. Point should be moved forward by at least
4802 ;; one token.
4803 ;;
4804 ;; If CFD-FUN adds `c-decl-end' properties somewhere below the current spot,
4805 ;; it should return non-nil to ensure that the next search will find them.
0386b551 4806 ;;
a85fd6da 4807 ;; Such a spot is:
0386b551
AM
4808 ;; o The first token after bob.
4809 ;; o The first token after the end of submatch 1 in
4810 ;; `c-decl-prefix-or-start-re' when that submatch matches.
4811 ;; o The start of each `c-decl-prefix-or-start-re' match when
4812 ;; submatch 1 doesn't match.
5a89f0a7 4813 ;; o The first token after the end of each occurrence of the
0386b551
AM
4814 ;; `c-type' text property with the value `c-decl-end', provided
4815 ;; `c-type-decl-end-used' is set.
4816 ;;
4817 ;; Only a spot that match CFD-DECL-RE and whose face is in the
4818 ;; CFD-FACE-CHECKLIST list causes CFD-FUN to be called. The face
4819 ;; check is disabled if CFD-FACE-CHECKLIST is nil.
d9e94c22
MS
4820 ;;
4821 ;; If the match is inside a macro then the buffer is narrowed to the
4822 ;; end of it, so that CFD-FUN can investigate the following tokens
4823 ;; without matching something that begins inside a macro and ends
4824 ;; outside it. It's to avoid this work that the CFD-DECL-RE and
4825 ;; CFD-FACE-CHECKLIST checks exist.
4826 ;;
0386b551
AM
4827 ;; The spots are visited approximately in order from top to bottom.
4828 ;; It's however the positions where `c-decl-prefix-or-start-re'
4829 ;; matches and where `c-decl-end' properties are found that are in
4830 ;; order. Since the spots often are at the following token, they
4831 ;; might be visited out of order insofar as more spots are reported
4832 ;; later on within the syntactic whitespace between the match
4833 ;; positions and their spots.
4834 ;;
4835 ;; It's assumed that comments and strings are fontified in the
d9e94c22
MS
4836 ;; searched range.
4837 ;;
4838 ;; This is mainly used in fontification, and so has an elaborate
4839 ;; cache to handle repeated calls from the same start position; see
4840 ;; the variables above.
4841 ;;
4842 ;; All variables in this function begin with `cfd-' to avoid name
4843 ;; collision with the (dynamically bound) variables used in CFD-FUN.
0386b551
AM
4844 ;;
4845 ;; This function might do hidden buffer changes.
d9e94c22 4846
0386b551
AM
4847 (let ((cfd-start-pos (point))
4848 (cfd-buffer-end (point-max))
4849 ;; The end of the token preceding the decl spot last found
4850 ;; with `c-decl-prefix-or-start-re'. `cfd-limit' if there's
4851 ;; no match.
d9e94c22 4852 cfd-re-match
0386b551
AM
4853 ;; The end position of the last `c-decl-prefix-or-start-re'
4854 ;; match. If this is greater than `cfd-continue-pos', the
4855 ;; next regexp search is started here instead.
4856 (cfd-re-match-end (point-min))
4857 ;; The end of the last `c-decl-end' found by
4858 ;; `c-find-decl-prefix-search'. `cfd-limit' if there's no
4859 ;; match. If searching for the property isn't needed then we
4860 ;; disable it by setting it to `cfd-limit' directly.
d9e94c22 4861 (cfd-prop-match (unless c-type-decl-end-used cfd-limit))
0386b551
AM
4862 ;; The end of the token preceding the decl spot last found by
4863 ;; `c-find-decl-prefix-search'. 0 for the implicit match at
4864 ;; bob. `cfd-limit' if there's no match. In other words,
4865 ;; this is the minimum of `cfd-re-match' and `cfd-prop-match'.
d9e94c22
MS
4866 (cfd-match-pos cfd-limit)
4867 ;; The position to continue searching at.
4868 cfd-continue-pos
4869 ;; The position of the last "real" token we've stopped at.
4870 ;; This can be greater than `cfd-continue-pos' when we get
4871 ;; hits inside macros or at `c-decl-end' positions inside
4872 ;; comments.
4873 (cfd-token-pos 0)
4874 ;; The end position of the last entered macro.
4875 (cfd-macro-end 0))
4876
4877 ;; Initialize by finding a syntactically relevant start position
0386b551
AM
4878 ;; before the point, and do the first `c-decl-prefix-or-start-re'
4879 ;; search unless we're at bob.
d9e94c22 4880
0386b551 4881 (let (start-in-literal start-in-macro syntactic-pos)
d9e94c22
MS
4882 ;; Must back up a bit since we look for the end of the previous
4883 ;; statement or declaration, which is earlier than the first
4884 ;; returned match.
4885
0386b551
AM
4886 (cond
4887 ;; First we need to move to a syntactically relevant position.
4888 ;; Begin by backing out of comment or string literals.
4889 ((and
4890 (when (c-got-face-at (point) c-literal-faces)
4891 ;; Try to use the faces to back up to the start of the
4892 ;; literal. FIXME: What if the point is on a declaration
4893 ;; inside a comment?
4894 (while (and (not (bobp))
4895 (c-got-face-at (1- (point)) c-literal-faces))
4896 (goto-char (previous-single-property-change
4897 (point) 'face nil (point-min))))
4898
4899 ;; XEmacs doesn't fontify the quotes surrounding string
4900 ;; literals.
4901 (and (featurep 'xemacs)
4902 (eq (get-text-property (point) 'face)
4903 'font-lock-string-face)
4904 (not (bobp))
4905 (progn (backward-char)
4906 (not (looking-at c-string-limit-regexp)))
4907 (forward-char))
4908
4909 ;; Don't trust the literal to contain only literal faces
4910 ;; (the font lock package might not have fontified the
4911 ;; start of it at all, for instance) so check that we have
4912 ;; arrived at something that looks like a start or else
4913 ;; resort to `c-literal-limits'.
4914 (unless (looking-at c-literal-start-regexp)
4915 (let ((range (c-literal-limits)))
4916 (if range (goto-char (car range)))))
4917
4918 (setq start-in-literal (point)))
4919
4920 ;; The start is in a literal. If the limit is in the same
4921 ;; one we don't have to find a syntactic position etc. We
4922 ;; only check that if the limit is at or before bonl to save
4923 ;; time; it covers the by far most common case when font-lock
4924 ;; refontifies the current line only.
4925 (<= cfd-limit (c-point 'bonl cfd-start-pos))
4926 (save-excursion
4927 (goto-char cfd-start-pos)
4928 (while (progn
4929 (goto-char (next-single-property-change
4930 (point) 'face nil cfd-limit))
4931 (and (< (point) cfd-limit)
4932 (c-got-face-at (point) c-literal-faces))))
4933 (= (point) cfd-limit)))
4934
4935 ;; Completely inside a literal. Set up variables to trig the
4936 ;; (< cfd-continue-pos cfd-start-pos) case below and it'll
4937 ;; find a suitable start position.
4938 (setq cfd-continue-pos start-in-literal))
4939
4940 ;; Check if the region might be completely inside a macro, to
4941 ;; optimize that like the completely-inside-literal above.
4942 ((save-excursion
4943 (and (= (forward-line 1) 0)
4944 (bolp) ; forward-line has funny behavior at eob.
4945 (>= (point) cfd-limit)
4946 (progn (backward-char)
4947 (eq (char-before) ?\\))))
4948 ;; (Maybe) completely inside a macro. Only need to trig the
4949 ;; (< cfd-continue-pos cfd-start-pos) case below to make it
4950 ;; set things up.
4951 (setq cfd-continue-pos (1- cfd-start-pos)
4952 start-in-macro t))
d9e94c22 4953
0386b551
AM
4954 (t
4955 ;; Back out of any macro so we don't miss any declaration
4956 ;; that could follow after it.
4957 (when (c-beginning-of-macro)
4958 (setq start-in-macro t))
4959
4960 ;; Now we're at a proper syntactically relevant position so we
4961 ;; can use the cache. But first clear it if it applied
4962 ;; further down.
4963 (c-invalidate-find-decl-cache cfd-start-pos)
4964
4965 (setq syntactic-pos (point))
4966 (unless (eq syntactic-pos c-find-decl-syntactic-pos)
4967 ;; Don't have to do this if the cache is relevant here,
4968 ;; typically if the same line is refontified again. If
4969 ;; we're just some syntactic whitespace further down we can
4970 ;; still use the cache to limit the skipping.
4971 (c-backward-syntactic-ws c-find-decl-syntactic-pos))
4972
4973 ;; If we hit `c-find-decl-syntactic-pos' and
4974 ;; `c-find-decl-match-pos' is set then we install the cached
4975 ;; values. If we hit `c-find-decl-syntactic-pos' and
4976 ;; `c-find-decl-match-pos' is nil then we know there's no decl
4977 ;; prefix in the whitespace before `c-find-decl-syntactic-pos'
4978 ;; and so we can continue the search from this point. If we
4979 ;; didn't hit `c-find-decl-syntactic-pos' then we're now in
4980 ;; the right spot to begin searching anyway.
4981 (if (and (eq (point) c-find-decl-syntactic-pos)
4982 c-find-decl-match-pos)
d9e94c22
MS
4983 (setq cfd-match-pos c-find-decl-match-pos
4984 cfd-continue-pos syntactic-pos)
0386b551
AM
4985
4986 (setq c-find-decl-syntactic-pos syntactic-pos)
4987
4988 (when (if (bobp)
4989 ;; Always consider bob a match to get the first
4990 ;; declaration in the file. Do this separately instead of
4991 ;; letting `c-decl-prefix-or-start-re' match bob, so that
4992 ;; regexp always can consume at least one character to
4993 ;; ensure that we won't get stuck in an infinite loop.
4994 (setq cfd-re-match 0)
4995 (backward-char)
4996 (c-beginning-of-current-token)
4997 (< (point) cfd-limit))
4998 ;; Do an initial search now. In the bob case above it's
4999 ;; only done to search for a `c-decl-end' spot.
5000 (c-find-decl-prefix-search))
5001
5002 (setq c-find-decl-match-pos (and (< cfd-match-pos cfd-start-pos)
5003 cfd-match-pos)))))
5004
5005 ;; Advance `cfd-continue-pos' if it's before the start position.
5006 ;; The closest continue position that might have effect at or
5007 ;; after the start depends on what we started in. This also
5008 ;; finds a suitable start position in the special cases when the
5009 ;; region is completely within a literal or macro.
5010 (when (and cfd-continue-pos (< cfd-continue-pos cfd-start-pos))
5011
5012 (cond
5013 (start-in-macro
5014 ;; If we're in a macro then it's the closest preceding token
5015 ;; in the macro. Check this before `start-in-literal',
5016 ;; since if we're inside a literal in a macro, the preceding
5017 ;; token is earlier than any `c-decl-end' spot inside the
5018 ;; literal (comment).
5019 (goto-char (or start-in-literal cfd-start-pos))
5020 ;; The only syntactic ws in macros are comments.
d9e94c22 5021 (c-backward-comments)
0386b551
AM
5022 (backward-char)
5023 (c-beginning-of-current-token))
5024
5025 (start-in-literal
5026 ;; If we're in a comment it can only be the closest
5027 ;; preceding `c-decl-end' position within that comment, if
5028 ;; any. Go back to the beginning of such a property so that
5029 ;; `c-find-decl-prefix-search' will find the end of it.
5030 ;; (Can't stop at the end and install it directly on
5031 ;; `cfd-prop-match' since that variable might be cleared
5032 ;; after `cfd-fun' below.)
5033 ;;
5034 ;; Note that if the literal is a string then the property
5035 ;; search will simply skip to the beginning of it right
5036 ;; away.
5037 (if (not c-type-decl-end-used)
5038 (goto-char start-in-literal)
5039 (goto-char cfd-start-pos)
5040 (while (progn
5041 (goto-char (previous-single-property-change
5042 (point) 'c-type nil start-in-literal))
5043 (and (> (point) start-in-literal)
5044 (not (eq (c-get-char-property (point) 'c-type)
5045 'c-decl-end))))))
5046
5047 (when (= (point) start-in-literal)
5048 ;; Didn't find any property inside the comment, so we can
5049 ;; skip it entirely. (This won't skip past a string, but
5050 ;; that'll be handled quickly by the next
5051 ;; `c-find-decl-prefix-search' anyway.)
5052 (c-forward-single-comment)
5053 (if (> (point) cfd-limit)
5054 (goto-char cfd-limit))))
d9e94c22 5055
0386b551
AM
5056 (t
5057 ;; If we started in normal code, the only match that might
5058 ;; apply before the start is what we already got in
5059 ;; `cfd-match-pos' so we can continue at the start position.
5060 ;; (Note that we don't get here if the first match is below
5061 ;; it.)
5062 (goto-char cfd-start-pos)))
5063
5064 ;; Delete found matches if they are before our new continue
5065 ;; position, so that `c-find-decl-prefix-search' won't back up
5066 ;; to them later on.
5067 (setq cfd-continue-pos (point))
5068 (when (and cfd-re-match (< cfd-re-match cfd-continue-pos))
5069 (setq cfd-re-match nil))
5070 (when (and cfd-prop-match (< cfd-prop-match cfd-continue-pos))
5071 (setq cfd-prop-match nil)))
5072
5073 (if syntactic-pos
5074 ;; This is the normal case and we got a proper syntactic
5075 ;; position. If there's a match then it's always outside
5076 ;; macros and comments, so advance to the next token and set
5077 ;; `cfd-token-pos'. The loop below will later go back using
5078 ;; `cfd-continue-pos' to fix declarations inside the
5079 ;; syntactic ws.
5080 (when (and cfd-match-pos (< cfd-match-pos syntactic-pos))
5081 (goto-char syntactic-pos)
5082 (c-forward-syntactic-ws)
5083 (and cfd-continue-pos
5084 (< cfd-continue-pos (point))
5085 (setq cfd-token-pos (point))))
5086
5087 ;; Have one of the special cases when the region is completely
5088 ;; within a literal or macro. `cfd-continue-pos' is set to a
5089 ;; good start position for the search, so do it.
5090 (c-find-decl-prefix-search)))
d9e94c22 5091
51c9af45 5092 ;; Now loop. Round what? (ACM, 2006/7/5). We already got the first match.
d9e94c22
MS
5093
5094 (while (progn
5095 (while (and
5096 (< cfd-match-pos cfd-limit)
5097
5098 (or
5099 ;; Kludge to filter out matches on the "<" that
5100 ;; aren't open parens, for the sake of languages
5101 ;; that got `c-recognize-<>-arglists' set.
5102 (and (eq (char-before cfd-match-pos) ?<)
5103 (not (c-get-char-property (1- cfd-match-pos)
5104 'syntax-table)))
5105
5106 ;; If `cfd-continue-pos' is less or equal to
5107 ;; `cfd-token-pos', we've got a hit inside a macro
5108 ;; that's in the syntactic whitespace before the last
5109 ;; "real" declaration we've checked. If they're equal
5110 ;; we've arrived at the declaration a second time, so
5111 ;; there's nothing to do.
5112 (= cfd-continue-pos cfd-token-pos)
5113
5114 (progn
5115 ;; If `cfd-continue-pos' is less than `cfd-token-pos'
5116 ;; we're still searching for declarations embedded in
5117 ;; the syntactic whitespace. In that case we need
5118 ;; only to skip comments and not macros, since they
5119 ;; can't be nested, and that's already been done in
5120 ;; `c-find-decl-prefix-search'.
5121 (when (> cfd-continue-pos cfd-token-pos)
5122 (c-forward-syntactic-ws)
5123 (setq cfd-token-pos (point)))
5124
5125 ;; Continue if the following token fails the
5126 ;; CFD-DECL-RE and CFD-FACE-CHECKLIST checks.
5127 (when (or (>= (point) cfd-limit)
5128 (not (looking-at cfd-decl-re))
5129 (and cfd-face-checklist
5130 (not (c-got-face-at
5131 (point) cfd-face-checklist))))
5132 (goto-char cfd-continue-pos)
5133 t)))
5134
5135 (< (point) cfd-limit))
5136 (c-find-decl-prefix-search))
5137
5138 (< (point) cfd-limit))
5139
0386b551
AM
5140 (when (and
5141 (>= (point) cfd-start-pos)
d9e94c22 5142
0386b551
AM
5143 (progn
5144 ;; Narrow to the end of the macro if we got a hit inside
5145 ;; one, to avoid recognizing things that start inside the
5146 ;; macro and end outside it.
5147 (when (> cfd-match-pos cfd-macro-end)
5148 ;; Not in the same macro as in the previous round.
5149 (save-excursion
5150 (goto-char cfd-match-pos)
5151 (setq cfd-macro-end
5152 (if (save-excursion (and (c-beginning-of-macro)
5153 (< (point) cfd-match-pos)))
5154 (progn (c-end-of-macro)
5155 (point))
5156 0))))
5157
5158 (if (zerop cfd-macro-end)
5159 t
5160 (if (> cfd-macro-end (point))
5161 (progn (narrow-to-region (point-min) cfd-macro-end)
5162 t)
5163 ;; The matched token was the last thing in the macro,
5164 ;; so the whole match is bogus.
5165 (setq cfd-macro-end 0)
5166 nil))))
d9e94c22
MS
5167
5168 (c-debug-put-decl-spot-faces cfd-match-pos (point))
0386b551
AM
5169 (if (funcall cfd-fun cfd-match-pos (/= cfd-macro-end 0))
5170 (setq cfd-prop-match nil))
d9e94c22
MS
5171
5172 (when (/= cfd-macro-end 0)
e1dbe924 5173 ;; Restore limits if we did macro narrowing above.
d9e94c22
MS
5174 (narrow-to-region (point-min) cfd-buffer-end)))
5175
5176 (goto-char cfd-continue-pos)
5177 (if (= cfd-continue-pos cfd-limit)
5178 (setq cfd-match-pos cfd-limit)
9657183b
AM
5179 (c-find-decl-prefix-search))))) ; Moves point, sets cfd-continue-pos,
5180 ; cfd-match-pos, etc.
d9e94c22
MS
5181
5182\f
5183;; A cache for found types.
5184
5185;; Buffer local variable that contains an obarray with the types we've
5186;; found. If a declaration is recognized somewhere we record the
5187;; fully qualified identifier in it to recognize it as a type
5188;; elsewhere in the file too. This is not accurate since we do not
5189;; bother with the scoping rules of the languages, but in practice the
5190;; same name is seldom used as both a type and something else in a
5191;; file, and we only use this as a last resort in ambiguous cases (see
0386b551
AM
5192;; `c-forward-decl-or-cast-1').
5193;;
580fba94
AM
5194;; Not every type need be in this cache. However, things which have
5195;; ceased to be types must be removed from it.
5196;;
0386b551
AM
5197;; Template types in C++ are added here too but with the template
5198;; arglist replaced with "<>" in references or "<" for the one in the
5199;; primary type. E.g. the type "Foo<A,B>::Bar<C>" is stored as
5200;; "Foo<>::Bar<". This avoids storing very long strings (since C++
5201;; template specs can be fairly sized programs in themselves) and
5202;; improves the hit ratio (it's a type regardless of the template
5203;; args; it's just not the same type, but we're only interested in
5204;; recognizing types, not telling distinct types apart). Note that
5205;; template types in references are added here too; from the example
5206;; above there will also be an entry "Foo<".
d9e94c22
MS
5207(defvar c-found-types nil)
5208(make-variable-buffer-local 'c-found-types)
5209
5210(defsubst c-clear-found-types ()
5211 ;; Clears `c-found-types'.
d9e94c22
MS
5212 (setq c-found-types (make-vector 53 0)))
5213
5214(defun c-add-type (from to)
5215 ;; Add the given region as a type in `c-found-types'. If the region
5216 ;; doesn't match an existing type but there is a type which is equal
5217 ;; to the given one except that the last character is missing, then
5218 ;; the shorter type is removed. That's done to avoid adding all
5219 ;; prefixes of a type as it's being entered and font locked. This
5220 ;; doesn't cover cases like when characters are removed from a type
5221 ;; or added in the middle. We'd need the position of point when the
5222 ;; font locking is invoked to solve this well.
0386b551
AM
5223 ;;
5224 ;; This function might do hidden buffer changes.
5225 (let ((type (c-syntactic-content from to c-recognize-<>-arglists)))
5226 (unless (intern-soft type c-found-types)
5227 (unintern (substring type 0 -1) c-found-types)
5228 (intern type c-found-types))))
d9e94c22 5229
580fba94
AM
5230(defun c-unfind-type (name)
5231 ;; Remove the "NAME" from c-found-types, if present.
5232 (unintern name c-found-types))
5233
d9e94c22
MS
5234(defsubst c-check-type (from to)
5235 ;; Return non-nil if the given region contains a type in
5236 ;; `c-found-types'.
0386b551
AM
5237 ;;
5238 ;; This function might do hidden buffer changes.
5239 (intern-soft (c-syntactic-content from to c-recognize-<>-arglists)
5240 c-found-types))
d9e94c22
MS
5241
5242(defun c-list-found-types ()
5243 ;; Return all the types in `c-found-types' as a sorted list of
5244 ;; strings.
5245 (let (type-list)
5246 (mapatoms (lambda (type)
5247 (setq type-list (cons (symbol-name type)
5248 type-list)))
5249 c-found-types)
5250 (sort type-list 'string-lessp)))
a66cd3ee 5251
2f42c75f
DN
5252;; Shut up the byte compiler.
5253(defvar c-maybe-stale-found-type)
5254
580fba94
AM
5255(defun c-trim-found-types (beg end old-len)
5256 ;; An after change function which, in conjunction with the info in
5257 ;; c-maybe-stale-found-type (set in c-before-change), removes a type
5258 ;; from `c-found-types', should this type have become stale. For
5259 ;; example, this happens to "foo" when "foo \n bar();" becomes
5260 ;; "foo(); \n bar();". Such stale types, if not removed, foul up
5261 ;; the fontification.
b414f371 5262 ;;
580fba94
AM
5263 ;; Have we, perhaps, added non-ws characters to the front/back of a found
5264 ;; type?
5265 (when (> end beg)
5266 (save-excursion
5267 (when (< end (point-max))
5268 (goto-char end)
5269 (if (and (c-beginning-of-current-token) ; only moves when we started in the middle
5270 (progn (goto-char end)
5271 (c-end-of-current-token)))
5272 (c-unfind-type (buffer-substring-no-properties
5273 end (point)))))
5274 (when (> beg (point-min))
5275 (goto-char beg)
5276 (if (and (c-end-of-current-token) ; only moves when we started in the middle
5277 (progn (goto-char beg)
5278 (c-beginning-of-current-token)))
5279 (c-unfind-type (buffer-substring-no-properties
5280 (point) beg))))))
b414f371 5281
580fba94
AM
5282 (if c-maybe-stale-found-type ; e.g. (c-decl-id-start "foo" 97 107 " (* ooka) " "o")
5283 (cond
5284 ;; Changing the amount of (already existing) whitespace - don't do anything.
5285 ((and (c-partial-ws-p beg end)
5286 (or (= beg end) ; removal of WS
580fba94
AM
5287 (string-match "^[ \t\n\r\f\v]*$" (nth 5 c-maybe-stale-found-type)))))
5288
5289 ;; The syntactic relationship which defined a "found type" has been
5290 ;; destroyed.
5291 ((eq (car c-maybe-stale-found-type) 'c-decl-id-start)
5292 (c-unfind-type (cadr c-maybe-stale-found-type)))
5293;; ((eq (car c-maybe-stale-found-type) 'c-decl-type-start) FIXME!!!
5294 )))
5295
d9e94c22 5296\f
dd969a56
AM
5297;; Setting and removing syntax properties on < and > in languages (C++
5298;; and Java) where they can be template/generic delimiters as well as
5299;; their normal meaning of "less/greater than".
5300
5301;; Normally, < and > have syntax 'punctuation'. When they are found to
5302;; be delimiters, they are marked as such with the category properties
5303;; c-<-as-paren-syntax, c->-as-paren-syntax respectively.
5304
5305;; STRATEGY:
5306;;
5307;; It is impossible to determine with certainty whether a <..> pair in
5308;; C++ is two comparison operators or is template delimiters, unless
5309;; one duplicates a lot of a C++ compiler. For example, the following
5310;; code fragment:
5311;;
5312;; foo (a < b, c > d) ;
5313;;
5314;; could be a function call with two integer parameters (each a
5315;; relational expression), or it could be a constructor for class foo
5316;; taking one parameter d of templated type "a < b, c >". They are
5317;; somewhat easier to distinguish in Java.
5318;;
5319;; The strategy now (2010-01) adopted is to mark and unmark < and
5320;; > IN MATCHING PAIRS ONLY. [Previously, they were marked
5321;; individually when their context so indicated. This gave rise to
53964682 5322;; intractable problems when one of a matching pair was deleted, or
dd969a56
AM
5323;; pulled into a literal.]
5324;;
5325;; At each buffer change, the syntax-table properties are removed in a
5326;; before-change function and reapplied, when needed, in an
5327;; after-change function. It is far more important that the
5328;; properties get removed when they they are spurious than that they
5329;; be present when wanted.
5330;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
5331(defun c-clear-<-pair-props (&optional pos)
5332 ;; POS (default point) is at a < character. If it is marked with
5333 ;; open paren syntax-table text property, remove the property,
5334 ;; together with the close paren property on the matching > (if
5335 ;; any).
5336 (save-excursion
5337 (if pos
5338 (goto-char pos)
5339 (setq pos (point)))
5340 (when (equal (c-get-char-property (point) 'syntax-table)
5341 c-<-as-paren-syntax)
5342 (with-syntax-table c-no-parens-syntax-table ; ignore unbalanced [,{,(,..
5343 (c-go-list-forward))
5344 (when (equal (c-get-char-property (1- (point)) 'syntax-table)
5345 c->-as-paren-syntax) ; should always be true.
8a249abc
AM
5346 (c-clear-char-property (1- (point)) 'category))
5347 (c-clear-char-property pos 'category))))
dd969a56
AM
5348
5349(defun c-clear->-pair-props (&optional pos)
5350 ;; POS (default point) is at a > character. If it is marked with
5351 ;; close paren syntax-table property, remove the property, together
5352 ;; with the open paren property on the matching < (if any).
5353 (save-excursion
5354 (if pos
5355 (goto-char pos)
5356 (setq pos (point)))
5357 (when (equal (c-get-char-property (point) 'syntax-table)
5358 c->-as-paren-syntax)
5359 (with-syntax-table c-no-parens-syntax-table ; ignore unbalanced [,{,(,..
5360 (c-go-up-list-backward))
5361 (when (equal (c-get-char-property (point) 'syntax-table)
5362 c-<-as-paren-syntax) ; should always be true.
8a249abc
AM
5363 (c-clear-char-property (point) 'category))
5364 (c-clear-char-property pos 'category))))
dd969a56
AM
5365
5366(defun c-clear-<>-pair-props (&optional pos)
5367 ;; POS (default point) is at a < or > character. If it has an
5368 ;; open/close paren syntax-table property, remove this property both
5369 ;; from the current character and its partner (which will also be
5370 ;; thusly marked).
5371 (cond
5372 ((eq (char-after) ?\<)
5373 (c-clear-<-pair-props pos))
5374 ((eq (char-after) ?\>)
5375 (c-clear->-pair-props pos))
5376 (t (c-benign-error
5377 "c-clear-<>-pair-props called from wrong position"))))
5378
5379(defun c-clear-<-pair-props-if-match-after (lim &optional pos)
5380 ;; POS (default point) is at a < character. If it is both marked
5381 ;; with open/close paren syntax-table property, and has a matching >
5382 ;; (also marked) which is after LIM, remove the property both from
43a91810
AM
5383 ;; the current > and its partner. Return t when this happens, nil
5384 ;; when it doesn't.
dd969a56
AM
5385 (save-excursion
5386 (if pos
5387 (goto-char pos)
5388 (setq pos (point)))
5389 (when (equal (c-get-char-property (point) 'syntax-table)
5390 c-<-as-paren-syntax)
5391 (with-syntax-table c-no-parens-syntax-table ; ignore unbalanced [,{,(,..
5392 (c-go-list-forward))
5393 (when (and (>= (point) lim)
5394 (equal (c-get-char-property (1- (point)) 'syntax-table)
5395 c->-as-paren-syntax)) ; should always be true.
5396 (c-unmark-<->-as-paren (1- (point)))
43a91810
AM
5397 (c-unmark-<->-as-paren pos))
5398 t)))
dd969a56
AM
5399
5400(defun c-clear->-pair-props-if-match-before (lim &optional pos)
5401 ;; POS (default point) is at a > character. If it is both marked
5402 ;; with open/close paren syntax-table property, and has a matching <
5403 ;; (also marked) which is before LIM, remove the property both from
43a91810
AM
5404 ;; the current < and its partner. Return t when this happens, nil
5405 ;; when it doesn't.
dd969a56
AM
5406 (save-excursion
5407 (if pos
5408 (goto-char pos)
5409 (setq pos (point)))
5410 (when (equal (c-get-char-property (point) 'syntax-table)
5411 c->-as-paren-syntax)
5412 (with-syntax-table c-no-parens-syntax-table ; ignore unbalanced [,{,(,..
5413 (c-go-up-list-backward))
5414 (when (and (<= (point) lim)
5415 (equal (c-get-char-property (point) 'syntax-table)
5416 c-<-as-paren-syntax)) ; should always be true.
5417 (c-unmark-<->-as-paren (point))
43a91810
AM
5418 (c-unmark-<->-as-paren pos))
5419 t)))
dd969a56 5420
f3b554af
GM
5421;; Set by c-common-init in cc-mode.el.
5422(defvar c-new-BEG)
5423(defvar c-new-END)
5424
dd969a56
AM
5425(defun c-before-change-check-<>-operators (beg end)
5426 ;; Unmark certain pairs of "< .... >" which are currently marked as
5427 ;; template/generic delimiters. (This marking is via syntax-table
5428 ;; text properties).
5429 ;;
5430 ;; These pairs are those which are in the current "statement" (i.e.,
5431 ;; the region between the {, }, or ; before BEG and the one after
5432 ;; END), and which enclose any part of the interval (BEG END).
5433 ;;
5434 ;; Note that in C++ (?and Java), template/generic parens cannot
5435 ;; enclose a brace or semicolon, so we use these as bounds on the
5436 ;; region we must work on.
5437 ;;
5438 ;; This function is called from before-change-functions (via
5439 ;; c-get-state-before-change-functions). Thus the buffer is widened,
5440 ;; and point is undefined, both at entry and exit.
5441 ;;
5442 ;; FIXME!!! This routine ignores the possibility of macros entirely.
5443 ;; 2010-01-29.
5444 (save-excursion
5445 (let ((beg-lit-limits (progn (goto-char beg) (c-literal-limits)))
43a91810
AM
5446 (end-lit-limits (progn (goto-char end) (c-literal-limits)))
5447 new-beg new-end need-new-beg need-new-end)
dd969a56
AM
5448 ;; Locate the barrier before the changed region
5449 (goto-char (if beg-lit-limits (car beg-lit-limits) beg))
0de3da9f 5450 (c-syntactic-skip-backward "^;{}" (c-determine-limit 512))
43a91810 5451 (setq new-beg (point))
dd969a56
AM
5452
5453 ;; Remove the syntax-table properties from each pertinent <...> pair.
5454 ;; Firsly, the ones with the < before beg and > after beg.
5455 (while (c-search-forward-char-property 'category 'c-<-as-paren-syntax beg)
43a91810
AM
5456 (if (c-clear-<-pair-props-if-match-after beg (1- (point)))
5457 (setq need-new-beg t)))
dd969a56
AM
5458
5459 ;; Locate the barrier after END.
5460 (goto-char (if end-lit-limits (cdr end-lit-limits) end))
0de3da9f 5461 (c-syntactic-re-search-forward "[;{}]" (c-determine-+ve-limit 512) 'end)
43a91810 5462 (setq new-end (point))
dd969a56
AM
5463
5464 ;; Remove syntax-table properties from the remaining pertinent <...>
5465 ;; pairs, those with a > after end and < before end.
5466 (while (c-search-backward-char-property 'category 'c->-as-paren-syntax end)
43a91810
AM
5467 (if (c-clear->-pair-props-if-match-before end)
5468 (setq need-new-end t)))
5469
5470 ;; Extend the fontification region, if needed.
5471 (when need-new-beg
5472 (goto-char new-beg)
5473 (c-forward-syntactic-ws)
5474 (and (< (point) c-new-BEG) (setq c-new-BEG (point))))
5475
5476 (when need-new-end
5477 (and (> new-end c-new-END) (setq c-new-END new-end))))))
dd969a56
AM
5478
5479
d9e94c22 5480
0386b551
AM
5481(defun c-after-change-check-<>-operators (beg end)
5482 ;; This is called from `after-change-functions' when
5483 ;; c-recognize-<>-arglists' is set. It ensures that no "<" or ">"
5484 ;; chars with paren syntax become part of another operator like "<<"
5485 ;; or ">=".
5486 ;;
5487 ;; This function might do hidden buffer changes.
5488
cb694ab7
AM
5489 (save-excursion
5490 (goto-char beg)
5491 (when (or (looking-at "[<>]")
5492 (< (skip-chars-backward "<>") 0))
5493
0386b551 5494 (goto-char beg)
cb694ab7
AM
5495 (c-beginning-of-current-token)
5496 (when (and (< (point) beg)
5497 (looking-at c-<>-multichar-token-regexp)
5498 (< beg (setq beg (match-end 0))))
5499 (while (progn (skip-chars-forward "^<>" beg)
5500 (< (point) beg))
dd969a56 5501 (c-clear-<>-pair-props)
cb694ab7
AM
5502 (forward-char))))
5503
5504 (when (< beg end)
5505 (goto-char end)
0386b551
AM
5506 (when (or (looking-at "[<>]")
5507 (< (skip-chars-backward "<>") 0))
5508
cb694ab7 5509 (goto-char end)
0386b551 5510 (c-beginning-of-current-token)
cb694ab7 5511 (when (and (< (point) end)
0386b551 5512 (looking-at c-<>-multichar-token-regexp)
cb694ab7
AM
5513 (< end (setq end (match-end 0))))
5514 (while (progn (skip-chars-forward "^<>" end)
5515 (< (point) end))
dd969a56 5516 (c-clear-<>-pair-props)
cb694ab7 5517 (forward-char)))))))
d9e94c22 5518
dd969a56
AM
5519
5520\f
5521;; Handling of small scale constructs like types and names.
5522
d9e94c22
MS
5523;; Dynamically bound variable that instructs `c-forward-type' to also
5524;; treat possible types (i.e. those that it normally returns 'maybe or
5525;; 'found for) as actual types (and always return 'found for them).
5526;; This means that it records them in `c-record-type-identifiers' if
5527;; that is set, and that it adds them to `c-found-types'.
5528(defvar c-promote-possible-types nil)
5529
0386b551
AM
5530;; Dynamically bound variable that instructs `c-forward-<>-arglist' to
5531;; mark up successfully parsed arglists with paren syntax properties on
5532;; the surrounding angle brackets and with `c-<>-arg-sep' in the
5533;; `c-type' property of each argument separating comma.
5534;;
5535;; Setting this variable also makes `c-forward-<>-arglist' recurse into
5536;; all arglists for side effects (i.e. recording types), otherwise it
5537;; exploits any existing paren syntax properties to quickly jump to the
5538;; end of already parsed arglists.
5539;;
5540;; Marking up the arglists is not the default since doing that correctly
5541;; depends on a proper value for `c-restricted-<>-arglists'.
5542(defvar c-parse-and-markup-<>-arglists nil)
5543
d9e94c22 5544;; Dynamically bound variable that instructs `c-forward-<>-arglist' to
037558bf
MS
5545;; not accept arglists that contain binary operators.
5546;;
5547;; This is primarily used to handle C++ template arglists. C++
5548;; disambiguates them by checking whether the preceding name is a
5549;; template or not. We can't do that, so we assume it is a template
5550;; if it can be parsed as one. That usually works well since
5551;; comparison expressions on the forms "a < b > c" or "a < b, c > d"
5552;; in almost all cases would be pointless.
5553;;
5554;; However, in function arglists, e.g. in "foo (a < b, c > d)", we
5555;; should let the comma separate the function arguments instead. And
5556;; in a context where the value of the expression is taken, e.g. in
5557;; "if (a < b || c > d)", it's probably not a template.
5558(defvar c-restricted-<>-arglists nil)
d9e94c22 5559
0386b551
AM
5560;; Dynamically bound variables that instructs
5561;; `c-forward-keyword-clause', `c-forward-<>-arglist',
5562;; `c-forward-name', `c-forward-type', `c-forward-decl-or-cast-1', and
5563;; `c-forward-label' to record the ranges of all the type and
5564;; reference identifiers they encounter. They will build lists on
5565;; these variables where each element is a cons of the buffer
5566;; positions surrounding each identifier. This recording is only
5567;; activated when `c-record-type-identifiers' is non-nil.
d9e94c22
MS
5568;;
5569;; All known types that can't be identifiers are recorded, and also
5570;; other possible types if `c-promote-possible-types' is set.
5571;; Recording is however disabled inside angle bracket arglists that
5572;; are encountered inside names and other angle bracket arglists.
0386b551 5573;; Such occurrences are taken care of by `c-font-lock-<>-arglists'
d9e94c22
MS
5574;; instead.
5575;;
5576;; Only the names in C++ template style references (e.g. "tmpl" in
5577;; "tmpl<a,b>::foo") are recorded as references, other references
5578;; aren't handled here.
0386b551
AM
5579;;
5580;; `c-forward-label' records the label identifier(s) on
5581;; `c-record-ref-identifiers'.
d9e94c22
MS
5582(defvar c-record-type-identifiers nil)
5583(defvar c-record-ref-identifiers nil)
5584
0386b551
AM
5585;; This variable will receive a cons cell of the range of the last
5586;; single identifier symbol stepped over by `c-forward-name' if it's
5587;; successful. This is the range that should be put on one of the
5588;; record lists above by the caller. It's assigned nil if there's no
5589;; such symbol in the name.
d9e94c22
MS
5590(defvar c-last-identifier-range nil)
5591
5592(defmacro c-record-type-id (range)
5593 (if (eq (car-safe range) 'cons)
5594 ;; Always true.
5595 `(setq c-record-type-identifiers
5596 (cons ,range c-record-type-identifiers))
5597 `(let ((range ,range))
5598 (if range
5599 (setq c-record-type-identifiers
5600 (cons range c-record-type-identifiers))))))
5601
5602(defmacro c-record-ref-id (range)
5603 (if (eq (car-safe range) 'cons)
5604 ;; Always true.
5605 `(setq c-record-ref-identifiers
5606 (cons ,range c-record-ref-identifiers))
5607 `(let ((range ,range))
5608 (if range
5609 (setq c-record-ref-identifiers
5610 (cons range c-record-ref-identifiers))))))
5611
5612;; Dynamically bound variable that instructs `c-forward-type' to
5613;; record the ranges of types that only are found. Behaves otherwise
5614;; like `c-record-type-identifiers'.
5615(defvar c-record-found-types nil)
5616
5617(defmacro c-forward-keyword-prefixed-id (type)
5618 ;; Used internally in `c-forward-keyword-clause' to move forward
5619 ;; over a type (if TYPE is 'type) or a name (otherwise) which
5620 ;; possibly is prefixed by keywords and their associated clauses.
5621 ;; Try with a type/name first to not trip up on those that begin
5622 ;; with a keyword. Return t if a known or found type is moved
5623 ;; over. The point is clobbered if nil is returned. If range
5624 ;; recording is enabled, the identifier is recorded on as a type
5625 ;; if TYPE is 'type or as a reference if TYPE is 'ref.
0386b551
AM
5626 ;;
5627 ;; This macro might do hidden buffer changes.
d9e94c22
MS
5628 `(let (res)
5629 (while (if (setq res ,(if (eq type 'type)
5630 `(c-forward-type)
5631 `(c-forward-name)))
5632 nil
5633 (and (looking-at c-keywords-regexp)
0386b551 5634 (c-forward-keyword-clause 1))))
d9e94c22
MS
5635 (when (memq res '(t known found prefix))
5636 ,(when (eq type 'ref)
5637 `(when c-record-type-identifiers
5638 (c-record-ref-id c-last-identifier-range)))
5639 t)))
5640
0386b551 5641(defmacro c-forward-id-comma-list (type update-safe-pos)
d9e94c22
MS
5642 ;; Used internally in `c-forward-keyword-clause' to move forward
5643 ;; over a comma separated list of types or names using
5644 ;; `c-forward-keyword-prefixed-id'.
0386b551
AM
5645 ;;
5646 ;; This macro might do hidden buffer changes.
d9e94c22 5647 `(while (and (progn
0386b551
AM
5648 ,(when update-safe-pos
5649 `(setq safe-pos (point)))
d9e94c22
MS
5650 (eq (char-after) ?,))
5651 (progn
5652 (forward-char)
5653 (c-forward-syntactic-ws)
5654 (c-forward-keyword-prefixed-id ,type)))))
5655
0386b551
AM
5656(defun c-forward-keyword-clause (match)
5657 ;; Submatch MATCH in the current match data is assumed to surround a
5658 ;; token. If it's a keyword, move over it and any immediately
5659 ;; following clauses associated with it, stopping at the start of
5660 ;; the next token. t is returned in that case, otherwise the point
d9e94c22
MS
5661 ;; stays and nil is returned. The kind of clauses that are
5662 ;; recognized are those specified by `c-type-list-kwds',
5663 ;; `c-ref-list-kwds', `c-colon-type-list-kwds',
5664 ;; `c-paren-nontype-kwds', `c-paren-type-kwds', `c-<>-type-kwds',
5665 ;; and `c-<>-arglist-kwds'.
0386b551
AM
5666 ;;
5667 ;; This function records identifier ranges on
5668 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
5669 ;; `c-record-type-identifiers' is non-nil.
5670 ;;
5671 ;; Note that for `c-colon-type-list-kwds', which doesn't necessary
5672 ;; apply directly after the keyword, the type list is moved over
5673 ;; only when there is no unaccounted token before it (i.e. a token
5674 ;; that isn't moved over due to some other keyword list). The
5675 ;; identifier ranges in the list are still recorded if that should
5676 ;; be done, though.
5677 ;;
5678 ;; This function might do hidden buffer changes.
5679
5680 (let ((kwd-sym (c-keyword-sym (match-string match))) safe-pos pos
5681 ;; The call to `c-forward-<>-arglist' below is made after
5682 ;; `c-<>-sexp-kwds' keywords, so we're certain they actually
5683 ;; are angle bracket arglists and `c-restricted-<>-arglists'
5684 ;; should therefore be nil.
5685 (c-parse-and-markup-<>-arglists t)
5686 c-restricted-<>-arglists)
d9e94c22 5687
d9e94c22 5688 (when kwd-sym
0386b551 5689 (goto-char (match-end match))
d9e94c22
MS
5690 (c-forward-syntactic-ws)
5691 (setq safe-pos (point))
a66cd3ee 5692
d9e94c22
MS
5693 (cond
5694 ((and (c-keyword-member kwd-sym 'c-type-list-kwds)
5695 (c-forward-keyword-prefixed-id type))
5696 ;; There's a type directly after a keyword in `c-type-list-kwds'.
0386b551 5697 (c-forward-id-comma-list type t))
d9e94c22
MS
5698
5699 ((and (c-keyword-member kwd-sym 'c-ref-list-kwds)
5700 (c-forward-keyword-prefixed-id ref))
5701 ;; There's a name directly after a keyword in `c-ref-list-kwds'.
0386b551 5702 (c-forward-id-comma-list ref t))
d9e94c22
MS
5703
5704 ((and (c-keyword-member kwd-sym 'c-paren-any-kwds)
5705 (eq (char-after) ?\())
5706 ;; There's an open paren after a keyword in `c-paren-any-kwds'.
5707
5708 (forward-char)
5709 (when (and (setq pos (c-up-list-forward))
5710 (eq (char-before pos) ?\)))
5711 (when (and c-record-type-identifiers
5712 (c-keyword-member kwd-sym 'c-paren-type-kwds))
5713 ;; Use `c-forward-type' on every identifier we can find
5714 ;; inside the paren, to record the types.
5715 (while (c-syntactic-re-search-forward c-symbol-start pos t)
5716 (goto-char (match-beginning 0))
5717 (unless (c-forward-type)
5718 (looking-at c-symbol-key) ; Always matches.
5719 (goto-char (match-end 0)))))
5720
5721 (goto-char pos)
5722 (c-forward-syntactic-ws)
5723 (setq safe-pos (point))))
5724
5725 ((and (c-keyword-member kwd-sym 'c-<>-sexp-kwds)
5726 (eq (char-after) ?<)
0386b551 5727 (c-forward-<>-arglist (c-keyword-member kwd-sym 'c-<>-type-kwds)))
d9e94c22
MS
5728 (c-forward-syntactic-ws)
5729 (setq safe-pos (point)))
5730
5731 ((and (c-keyword-member kwd-sym 'c-nonsymbol-sexp-kwds)
449a2b0d
MS
5732 (not (looking-at c-symbol-start))
5733 (c-safe (c-forward-sexp) t))
d9e94c22
MS
5734 (c-forward-syntactic-ws)
5735 (setq safe-pos (point))))
5736
0386b551
AM
5737 (when (c-keyword-member kwd-sym 'c-colon-type-list-kwds)
5738 (if (eq (char-after) ?:)
5739 ;; If we are at the colon already, we move over the type
5740 ;; list after it.
5741 (progn
5742 (forward-char)
5743 (c-forward-syntactic-ws)
5744 (when (c-forward-keyword-prefixed-id type)
5745 (c-forward-id-comma-list type t)))
5746 ;; Not at the colon, so stop here. But the identifier
5747 ;; ranges in the type list later on should still be
5748 ;; recorded.
5749 (and c-record-type-identifiers
5750 (progn
5751 ;; If a keyword matched both one of the types above and
5752 ;; this one, we match `c-colon-type-list-re' after the
5753 ;; clause matched above.
5754 (goto-char safe-pos)
5755 (looking-at c-colon-type-list-re))
5756 (progn
5757 (goto-char (match-end 0))
5758 (c-forward-syntactic-ws)
5759 (c-forward-keyword-prefixed-id type))
5760 ;; There's a type after the `c-colon-type-list-re' match
5761 ;; after a keyword in `c-colon-type-list-kwds'.
5762 (c-forward-id-comma-list type nil))))
d9e94c22
MS
5763
5764 (goto-char safe-pos)
5765 t)))
5766
f3b554af
GM
5767;; cc-mode requires cc-fonts.
5768(declare-function c-fontify-recorded-types-and-refs "cc-fonts" ())
5769
0386b551
AM
5770(defun c-forward-<>-arglist (all-types)
5771 ;; The point is assumed to be at a "<". Try to treat it as the open
b4dc7d98 5772 ;; paren of an angle bracket arglist and move forward to the
0386b551
AM
5773 ;; corresponding ">". If successful, the point is left after the
5774 ;; ">" and t is returned, otherwise the point isn't moved and nil is
d9e94c22
MS
5775 ;; returned. If ALL-TYPES is t then all encountered arguments in
5776 ;; the arglist that might be types are treated as found types.
5777 ;;
0386b551
AM
5778 ;; The variable `c-parse-and-markup-<>-arglists' controls how this
5779 ;; function handles text properties on the angle brackets and argument
5780 ;; separating commas.
d9e94c22 5781 ;;
0386b551
AM
5782 ;; `c-restricted-<>-arglists' controls how lenient the template
5783 ;; arglist recognition should be.
5784 ;;
5785 ;; This function records identifier ranges on
5786 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
5787 ;; `c-record-type-identifiers' is non-nil.
5788 ;;
5789 ;; This function might do hidden buffer changes.
d9e94c22
MS
5790
5791 (let ((start (point))
5792 ;; If `c-record-type-identifiers' is set then activate
5793 ;; recording of any found types that constitute an argument in
5794 ;; the arglist.
abfc152b 5795 (c-record-found-types (if c-record-type-identifiers t)))
d9e94c22
MS
5796 (if (catch 'angle-bracket-arglist-escape
5797 (setq c-record-found-types
0386b551 5798 (c-forward-<>-arglist-recur all-types)))
d9e94c22
MS
5799 (progn
5800 (when (consp c-record-found-types)
5801 (setq c-record-type-identifiers
5802 ;; `nconc' doesn't mind that the tail of
5803 ;; `c-record-found-types' is t.
5804 (nconc c-record-found-types c-record-type-identifiers)))
452ea855 5805 (if (c-major-mode-is 'java-mode) (c-fontify-recorded-types-and-refs))
d9e94c22
MS
5806 t)
5807
5808 (goto-char start)
a66cd3ee 5809 nil)))
785eecbb 5810
0386b551 5811(defun c-forward-<>-arglist-recur (all-types)
d9e94c22 5812 ;; Recursive part of `c-forward-<>-arglist'.
0386b551
AM
5813 ;;
5814 ;; This function might do hidden buffer changes.
d9e94c22
MS
5815
5816 (let ((start (point)) res pos tmp
5817 ;; Cover this so that any recorded found type ranges are
5818 ;; automatically lost if it turns out to not be an angle
5819 ;; bracket arglist. It's propagated through the return value
5820 ;; on successful completion.
5821 (c-record-found-types c-record-found-types)
5822 ;; List that collects the positions after the argument
5823 ;; separating ',' in the arglist.
5824 arg-start-pos)
0386b551
AM
5825 ;; If the '<' has paren open syntax then we've marked it as an angle
5826 ;; bracket arglist before, so skip to the end.
5827 (if (and (not c-parse-and-markup-<>-arglists)
5828 (c-get-char-property (point) 'syntax-table))
5829
5830 (progn
5831 (forward-char)
5832 (if (and (c-go-up-list-forward)
5833 (eq (char-before) ?>))
5834 t
0386b551
AM
5835 ;; Got unmatched paren angle brackets. We don't clear the paren
5836 ;; syntax properties and retry, on the basis that it's very
5837 ;; unlikely that paren angle brackets become operators by code
5838 ;; manipulation. It's far more likely that it doesn't match due
5839 ;; to narrowing or some temporary change.
5840 (goto-char start)
5841 nil))
d9e94c22 5842
abfc152b 5843 (forward-char) ; Forward over the opening '<'.
452ea855 5844
d9e94c22 5845 (unless (looking-at c-<-op-cont-regexp)
abfc152b
AM
5846 ;; go forward one non-alphanumeric character (group) per iteration of
5847 ;; this loop.
22c3ce97 5848 (while (and
d9e94c22 5849 (progn
22c3ce97
AM
5850 (c-forward-syntactic-ws)
5851 (let ((orig-record-found-types c-record-found-types))
5852 (when (or (and c-record-type-identifiers all-types)
5853 (c-major-mode-is 'java-mode))
5854 ;; All encountered identifiers are types, so set the
5855 ;; promote flag and parse the type.
5856 (progn
5857 (c-forward-syntactic-ws)
5858 (if (looking-at "\\?")
5859 (forward-char)
5860 (when (looking-at c-identifier-start)
5861 (let ((c-promote-possible-types t)
5862 (c-record-found-types t))
5863 (c-forward-type))))
5864
5865 (c-forward-syntactic-ws)
5866
5867 (when (or (looking-at "extends")
5868 (looking-at "super"))
5869 (forward-word)
5870 (c-forward-syntactic-ws)
5871 (let ((c-promote-possible-types t)
5872 (c-record-found-types t))
5873 (c-forward-type)
5874 (c-forward-syntactic-ws))))))
5875
abfc152b 5876 (setq pos (point)) ; e.g. first token inside the '<'
22c3ce97 5877
a4ee83cc
AM
5878 ;; Note: These regexps exploit the match order in \| so
5879 ;; that "<>" is matched by "<" rather than "[^>:-]>".
5880 (c-syntactic-re-search-forward
5881 ;; Stop on ',', '|', '&', '+' and '-' to catch
5882 ;; common binary operators that could be between
5883 ;; two comparison expressions "a<b" and "c>d".
5884 "[<;{},|+&-]\\|[>)]"
5885 nil t t))
22c3ce97
AM
5886
5887 (cond
5888 ((eq (char-before) ?>)
d9e94c22
MS
5889 ;; Either an operator starting with '>' or the end of
5890 ;; the angle bracket arglist.
5891
0386b551 5892 (if (looking-at c->-op-cont-regexp)
d9e94c22 5893 (progn
0386b551
AM
5894 (goto-char (match-end 0))
5895 t) ; Continue the loop.
d9e94c22 5896
0386b551
AM
5897 ;; The angle bracket arglist is finished.
5898 (when c-parse-and-markup-<>-arglists
d9e94c22 5899 (while arg-start-pos
0386b551
AM
5900 (c-put-c-type-property (1- (car arg-start-pos))
5901 'c-<>-arg-sep)
d9e94c22
MS
5902 (setq arg-start-pos (cdr arg-start-pos)))
5903 (c-mark-<-as-paren start)
0386b551
AM
5904 (c-mark->-as-paren (1- (point))))
5905 (setq res t)
5906 nil)) ; Exit the loop.
d9e94c22
MS
5907
5908 ((eq (char-before) ?<)
5909 ;; Either an operator starting with '<' or a nested arglist.
d9e94c22
MS
5910 (setq pos (point))
5911 (let (id-start id-end subres keyword-match)
abfc152b
AM
5912 (cond
5913 ;; The '<' begins a multi-char operator.
5914 ((looking-at c-<-op-cont-regexp)
5915 (setq tmp (match-end 0))
5916 (goto-char (match-end 0)))
5917 ;; We're at a nested <.....>
5918 ((progn
5919 (setq tmp pos)
5920 (backward-char) ; to the '<'
5921 (and
5922 (save-excursion
5923 ;; There's always an identifier before an angle
5924 ;; bracket arglist, or a keyword in `c-<>-type-kwds'
5925 ;; or `c-<>-arglist-kwds'.
5926 (c-backward-syntactic-ws)
5927 (setq id-end (point))
5928 (c-simple-skip-symbol-backward)
5929 (when (or (setq keyword-match
5930 (looking-at c-opt-<>-sexp-key))
5931 (not (looking-at c-keywords-regexp)))
5932 (setq id-start (point))))
5933 (setq subres
5934 (let ((c-promote-possible-types t)
5935 (c-record-found-types t))
5936 (c-forward-<>-arglist-recur
5937 (and keyword-match
5938 (c-keyword-member
5939 (c-keyword-sym (match-string 1))
5940 'c-<>-type-kwds)))))))
d9e94c22
MS
5941
5942 ;; It was an angle bracket arglist.
5943 (setq c-record-found-types subres)
5944
5945 ;; Record the identifier before the template as a type
5946 ;; or reference depending on whether the arglist is last
5947 ;; in a qualified identifier.
5948 (when (and c-record-type-identifiers
5949 (not keyword-match))
5950 (if (and c-opt-identifier-concat-key
5951 (progn
5952 (c-forward-syntactic-ws)
5953 (looking-at c-opt-identifier-concat-key)))
5954 (c-record-ref-id (cons id-start id-end))
abfc152b
AM
5955 (c-record-type-id (cons id-start id-end)))))
5956
5957 ;; At a "less than" operator.
5958 (t
5959 (forward-char)
5960 )))
5961 t) ; carry on looping.
22c3ce97
AM
5962
5963 ((and (not c-restricted-<>-arglists)
5964 (or (and (eq (char-before) ?&)
5965 (not (eq (char-after) ?&)))
5966 (eq (char-before) ?,)))
5967 ;; Just another argument. Record the position. The
5968 ;; type check stuff that made us stop at it is at
5969 ;; the top of the loop.
d9e94c22
MS
5970 (setq arg-start-pos (cons (point) arg-start-pos)))
5971
5972 (t
5973 ;; Got a character that can't be in an angle bracket
5974 ;; arglist argument. Abort using `throw', since
5975 ;; it's useless to try to find a surrounding arglist
5976 ;; if we're nested.
5977 (throw 'angle-bracket-arglist-escape nil))))))
d9e94c22
MS
5978 (if res
5979 (or c-record-found-types t)))))
5980
0386b551
AM
5981(defun c-backward-<>-arglist (all-types &optional limit)
5982 ;; The point is assumed to be directly after a ">". Try to treat it
5983 ;; as the close paren of an angle bracket arglist and move back to
5984 ;; the corresponding "<". If successful, the point is left at
5985 ;; the "<" and t is returned, otherwise the point isn't moved and
5986 ;; nil is returned. ALL-TYPES is passed on to
5987 ;; `c-forward-<>-arglist'.
5988 ;;
5989 ;; If the optional LIMIT is given, it bounds the backward search.
5990 ;; It's then assumed to be at a syntactically relevant position.
5991 ;;
5992 ;; This is a wrapper around `c-forward-<>-arglist'. See that
5993 ;; function for more details.
5994
5995 (let ((start (point)))
5996 (backward-char)
5997 (if (and (not c-parse-and-markup-<>-arglists)
5998 (c-get-char-property (point) 'syntax-table))
5999
6000 (if (and (c-go-up-list-backward)
6001 (eq (char-after) ?<))
6002 t
6003 ;; See corresponding note in `c-forward-<>-arglist'.
6004 (goto-char start)
6005 nil)
6006
51c9af45 6007 (while (progn
0386b551
AM
6008 (c-syntactic-skip-backward "^<;{}" limit t)
6009
51c9af45
AM
6010 (and
6011 (if (eq (char-before) ?<)
6012 t
6013 ;; Stopped at bob or a char that isn't allowed in an
6014 ;; arglist, so we've failed.
6015 (goto-char start)
6016 nil)
0386b551 6017
51c9af45
AM
6018 (if (> (point)
6019 (progn (c-beginning-of-current-token)
6020 (point)))
6021 ;; If we moved then the "<" was part of some
6022 ;; multicharacter token.
6023 t
0386b551 6024
51c9af45
AM
6025 (backward-char)
6026 (let ((beg-pos (point)))
6027 (if (c-forward-<>-arglist all-types)
6028 (cond ((= (point) start)
6029 ;; Matched the arglist. Break the while.
6030 (goto-char beg-pos)
6031 nil)
6032 ((> (point) start)
6033 ;; We started from a non-paren ">" inside an
6034 ;; arglist.
6035 (goto-char start)
6036 nil)
6037 (t
6038 ;; Matched a shorter arglist. Can be a nested
6039 ;; one so continue looking.
6040 (goto-char beg-pos)
6041 t))
6042 t))))))
0386b551
AM
6043
6044 (/= (point) start))))
6045
d9e94c22
MS
6046(defun c-forward-name ()
6047 ;; Move forward over a complete name if at the beginning of one,
e15f8aaa
AM
6048 ;; stopping at the next following token. A keyword, as such,
6049 ;; doesn't count as a name. If the point is not at something that
6050 ;; is recognized as a name then it stays put.
6051 ;;
6052 ;; A name could be something as simple as "foo" in C or something as
d9e94c22
MS
6053 ;; complex as "X<Y<class A<int>::B, BIT_MAX >> b>, ::operator<> ::
6054 ;; Z<(a>b)> :: operator const X<&foo>::T Q::G<unsigned short
6055 ;; int>::*volatile const" in C++ (this function is actually little
6056 ;; more than a `looking-at' call in all modes except those that,
e15f8aaa
AM
6057 ;; like C++, have `c-recognize-<>-arglists' set).
6058 ;;
6059 ;; Return
6060 ;; o - nil if no name is found;
6061 ;; o - 'template if it's an identifier ending with an angle bracket
6062 ;; arglist;
6063 ;; o - 'operator of it's an operator identifier;
6064 ;; o - t if it's some other kind of name.
0386b551
AM
6065 ;;
6066 ;; This function records identifier ranges on
6067 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
6068 ;; `c-record-type-identifiers' is non-nil.
6069 ;;
6070 ;; This function might do hidden buffer changes.
d9e94c22 6071
0386b551 6072 (let ((pos (point)) (start (point)) res id-start id-end
d9e94c22
MS
6073 ;; Turn off `c-promote-possible-types' here since we might
6074 ;; call `c-forward-<>-arglist' and we don't want it to promote
6075 ;; every suspect thing in the arglist to a type. We're
6076 ;; typically called from `c-forward-type' in this case, and
6077 ;; the caller only wants the top level type that it finds to
6078 ;; be promoted.
6079 c-promote-possible-types)
6080 (while
6081 (and
6082 (looking-at c-identifier-key)
6083
6084 (progn
6085 ;; Check for keyword. We go to the last symbol in
6086 ;; `c-identifier-key' first.
0386b551
AM
6087 (goto-char (setq id-end (match-end 0)))
6088 (c-simple-skip-symbol-backward)
6089 (setq id-start (point))
d9e94c22
MS
6090
6091 (if (looking-at c-keywords-regexp)
6092 (when (and (c-major-mode-is 'c++-mode)
6093 (looking-at
6094 (cc-eval-when-compile
6095 (concat "\\(operator\\|\\(template\\)\\)"
6096 "\\(" (c-lang-const c-nonsymbol-key c++)
6097 "\\|$\\)")))
6098 (if (match-beginning 2)
6099 ;; "template" is only valid inside an
6100 ;; identifier if preceded by "::".
6101 (save-excursion
6102 (c-backward-syntactic-ws)
6103 (and (c-safe (backward-char 2) t)
6104 (looking-at "::")))
6105 t))
6106
6107 ;; Handle a C++ operator or template identifier.
6108 (goto-char id-end)
6109 (c-forward-syntactic-ws)
6110 (cond ((eq (char-before id-end) ?e)
6111 ;; Got "... ::template".
6112 (let ((subres (c-forward-name)))
6113 (when subres
6114 (setq pos (point)
6115 res subres))))
6116
6117 ((looking-at c-identifier-start)
6118 ;; Got a cast operator.
6119 (when (c-forward-type)
6120 (setq pos (point)
6121 res 'operator)
6122 ;; Now we should match a sequence of either
6123 ;; '*', '&' or a name followed by ":: *",
6124 ;; where each can be followed by a sequence
6125 ;; of `c-opt-type-modifier-key'.
6126 (while (cond ((looking-at "[*&]")
6127 (goto-char (match-end 0))
6128 t)
6129 ((looking-at c-identifier-start)
6130 (and (c-forward-name)
6131 (looking-at "::")
6132 (progn
6133 (goto-char (match-end 0))
6134 (c-forward-syntactic-ws)
6135 (eq (char-after) ?*))
6136 (progn
6137 (forward-char)
6138 t))))
6139 (while (progn
6140 (c-forward-syntactic-ws)
6141 (setq pos (point))
6142 (looking-at c-opt-type-modifier-key))
6143 (goto-char (match-end 1))))))
6144
6145 ((looking-at c-overloadable-operators-regexp)
6146 ;; Got some other operator.
0386b551
AM
6147 (setq c-last-identifier-range
6148 (cons (point) (match-end 0)))
d9e94c22
MS
6149 (goto-char (match-end 0))
6150 (c-forward-syntactic-ws)
6151 (setq pos (point)
6152 res 'operator)))
6153
6154 nil)
6155
0386b551
AM
6156 ;; `id-start' is equal to `id-end' if we've jumped over
6157 ;; an identifier that doesn't end with a symbol token.
6158 ;; That can occur e.g. for Java import directives on the
6159 ;; form "foo.bar.*".
6160 (when (and id-start (/= id-start id-end))
d9e94c22
MS
6161 (setq c-last-identifier-range
6162 (cons id-start id-end)))
6163 (goto-char id-end)
6164 (c-forward-syntactic-ws)
6165 (setq pos (point)
6166 res t)))
6167
6168 (progn
6169 (goto-char pos)
6170 (when (or c-opt-identifier-concat-key
6171 c-recognize-<>-arglists)
6172
6173 (cond
6174 ((and c-opt-identifier-concat-key
6175 (looking-at c-opt-identifier-concat-key))
6176 ;; Got a concatenated identifier. This handles the
6177 ;; cases with tricky syntactic whitespace that aren't
6178 ;; covered in `c-identifier-key'.
6179 (goto-char (match-end 0))
6180 (c-forward-syntactic-ws)
6181 t)
6182
6183 ((and c-recognize-<>-arglists
6184 (eq (char-after) ?<))
6185 ;; Maybe an angle bracket arglist.
452ea855
AM
6186 (when (let ((c-record-type-identifiers t)
6187 (c-record-found-types t))
0386b551
AM
6188 (c-forward-<>-arglist nil))
6189
6190 (c-add-type start (1+ pos))
d9e94c22 6191 (c-forward-syntactic-ws)
0386b551
AM
6192 (setq pos (point)
6193 c-last-identifier-range nil)
6194
d9e94c22
MS
6195 (if (and c-opt-identifier-concat-key
6196 (looking-at c-opt-identifier-concat-key))
0386b551 6197
d9e94c22
MS
6198 ;; Continue if there's an identifier concatenation
6199 ;; operator after the template argument.
6200 (progn
0386b551
AM
6201 (when (and c-record-type-identifiers id-start)
6202 (c-record-ref-id (cons id-start id-end)))
d9e94c22
MS
6203 (forward-char 2)
6204 (c-forward-syntactic-ws)
6205 t)
0386b551
AM
6206
6207 (when (and c-record-type-identifiers id-start)
6208 (c-record-type-id (cons id-start id-end)))
d9e94c22
MS
6209 (setq res 'template)
6210 nil)))
6211 )))))
6212
6213 (goto-char pos)
6214 res))
6215
e15f8aaa 6216(defun c-forward-type (&optional brace-block-too)
d9e94c22 6217 ;; Move forward over a type spec if at the beginning of one,
e15f8aaa
AM
6218 ;; stopping at the next following token. The keyword "typedef"
6219 ;; isn't part of a type spec here.
6220 ;;
6221 ;; BRACE-BLOCK-TOO, when non-nil, means move over the brace block in
6222 ;; constructs like "struct foo {...} bar ;" or "struct {...} bar;".
6223 ;; The current (2009-03-10) intention is to convert all uses of
6224 ;; `c-forward-type' to call with this parameter set, then to
6225 ;; eliminate it.
6226 ;;
6227 ;; Return
6228 ;; o - t if it's a known type that can't be a name or other
6229 ;; expression;
6230 ;; o - 'known if it's an otherwise known type (according to
6231 ;; `*-font-lock-extra-types');
6232 ;; o - 'prefix if it's a known prefix of a type;
6233 ;; o - 'found if it's a type that matches one in `c-found-types';
53964682 6234 ;; o - 'maybe if it's an identifier that might be a type; or
e15f8aaa
AM
6235 ;; o - nil if it can't be a type (the point isn't moved then).
6236 ;;
6237 ;; The point is assumed to be at the beginning of a token.
d9e94c22
MS
6238 ;;
6239 ;; Note that this function doesn't skip past the brace definition
6240 ;; that might be considered part of the type, e.g.
6241 ;; "enum {a, b, c} foo".
0386b551
AM
6242 ;;
6243 ;; This function records identifier ranges on
6244 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
6245 ;; `c-record-type-identifiers' is non-nil.
6246 ;;
6247 ;; This function might do hidden buffer changes.
0e5cf2b8
AM
6248 (when (and c-recognize-<>-arglists
6249 (looking-at "<"))
452ea855
AM
6250 (c-forward-<>-arglist t)
6251 (c-forward-syntactic-ws))
0386b551
AM
6252
6253 (let ((start (point)) pos res name-res id-start id-end id-range)
d9e94c22
MS
6254
6255 ;; Skip leading type modifiers. If any are found we know it's a
6256 ;; prefix of a type.
e15f8aaa 6257 (when c-opt-type-modifier-key ; e.g. "const" "volatile", but NOT "typedef"
d9e94c22
MS
6258 (while (looking-at c-opt-type-modifier-key)
6259 (goto-char (match-end 1))
6260 (c-forward-syntactic-ws)
6261 (setq res 'prefix)))
6262
6263 (cond
e15f8aaa
AM
6264 ((looking-at c-type-prefix-key) ; e.g. "struct", "class", but NOT
6265 ; "typedef".
d9e94c22
MS
6266 (goto-char (match-end 1))
6267 (c-forward-syntactic-ws)
6268 (setq pos (point))
e15f8aaa
AM
6269
6270 (setq name-res (c-forward-name))
6271 (setq res (not (null name-res)))
6272 (when (eq name-res t)
6273 ;; In many languages the name can be used without the
6274 ;; prefix, so we add it to `c-found-types'.
6275 (c-add-type pos (point))
6276 (when (and c-record-type-identifiers
6277 c-last-identifier-range)
6278 (c-record-type-id c-last-identifier-range)))
6279 (when (and brace-block-too
6280 (memq res '(t nil))
6281 (eq (char-after) ?\{)
6282 (save-excursion
6283 (c-safe
6284 (progn (c-forward-sexp)
6285 (c-forward-syntactic-ws)
6286 (setq pos (point))))))
6287 (goto-char pos)
6288 (setq res t))
6289 (unless res (goto-char start))) ; invalid syntax
d9e94c22
MS
6290
6291 ((progn
6292 (setq pos nil)
6293 (if (looking-at c-identifier-start)
6294 (save-excursion
6295 (setq id-start (point)
0386b551
AM
6296 name-res (c-forward-name))
6297 (when name-res
d9e94c22
MS
6298 (setq id-end (point)
6299 id-range c-last-identifier-range))))
6300 (and (cond ((looking-at c-primitive-type-key)
6301 (setq res t))
6302 ((c-with-syntax-table c-identifier-syntax-table
6303 (looking-at c-known-type-key))
6304 (setq res 'known)))
6305 (or (not id-end)
6306 (>= (save-excursion
6307 (save-match-data
6308 (goto-char (match-end 1))
6309 (c-forward-syntactic-ws)
6310 (setq pos (point))))
6311 id-end)
6312 (setq res nil))))
6313 ;; Looking at a primitive or known type identifier. We've
6314 ;; checked for a name first so that we don't go here if the
6315 ;; known type match only is a prefix of another name.
6316
6317 (setq id-end (match-end 1))
6318
6319 (when (and c-record-type-identifiers
6320 (or c-promote-possible-types (eq res t)))
6321 (c-record-type-id (cons (match-beginning 1) (match-end 1))))
6322
6323 (if (and c-opt-type-component-key
6324 (save-match-data
6325 (looking-at c-opt-type-component-key)))
6326 ;; There might be more keywords for the type.
6327 (let (safe-pos)
0386b551 6328 (c-forward-keyword-clause 1)
d9e94c22
MS
6329 (while (progn
6330 (setq safe-pos (point))
6331 (looking-at c-opt-type-component-key))
6332 (when (and c-record-type-identifiers
6333 (looking-at c-primitive-type-key))
6334 (c-record-type-id (cons (match-beginning 1)
6335 (match-end 1))))
0386b551 6336 (c-forward-keyword-clause 1))
d9e94c22
MS
6337 (if (looking-at c-primitive-type-key)
6338 (progn
6339 (when c-record-type-identifiers
6340 (c-record-type-id (cons (match-beginning 1)
6341 (match-end 1))))
0386b551 6342 (c-forward-keyword-clause 1)
d9e94c22
MS
6343 (setq res t))
6344 (goto-char safe-pos)
6345 (setq res 'prefix)))
0386b551 6346 (unless (save-match-data (c-forward-keyword-clause 1))
d9e94c22
MS
6347 (if pos
6348 (goto-char pos)
6349 (goto-char (match-end 1))
6350 (c-forward-syntactic-ws)))))
6351
0386b551
AM
6352 (name-res
6353 (cond ((eq name-res t)
d9e94c22
MS
6354 ;; A normal identifier.
6355 (goto-char id-end)
6356 (if (or res c-promote-possible-types)
6357 (progn
6358 (c-add-type id-start id-end)
0386b551 6359 (when (and c-record-type-identifiers id-range)
d9e94c22
MS
6360 (c-record-type-id id-range))
6361 (unless res
6362 (setq res 'found)))
6363 (setq res (if (c-check-type id-start id-end)
6364 ;; It's an identifier that has been used as
6365 ;; a type somewhere else.
6366 'found
6367 ;; It's an identifier that might be a type.
6368 'maybe))))
0386b551 6369 ((eq name-res 'template)
d9e94c22
MS
6370 ;; A template is a type.
6371 (goto-char id-end)
6372 (setq res t))
6373 (t
6374 ;; Otherwise it's an operator identifier, which is not a type.
6375 (goto-char start)
6376 (setq res nil)))))
6377
6378 (when res
e15f8aaa 6379 ;; Skip trailing type modifiers. If any are found we know it's
d9e94c22
MS
6380 ;; a type.
6381 (when c-opt-type-modifier-key
e15f8aaa 6382 (while (looking-at c-opt-type-modifier-key) ; e.g. "const", "volatile"
d9e94c22
MS
6383 (goto-char (match-end 1))
6384 (c-forward-syntactic-ws)
6385 (setq res t)))
d9e94c22
MS
6386 ;; Step over any type suffix operator. Do not let the existence
6387 ;; of these alter the classification of the found type, since
6388 ;; these operators typically are allowed in normal expressions
6389 ;; too.
6390 (when c-opt-type-suffix-key
6391 (while (looking-at c-opt-type-suffix-key)
6392 (goto-char (match-end 1))
6393 (c-forward-syntactic-ws)))
6394
e15f8aaa 6395 (when c-opt-type-concat-key ; Only/mainly for pike.
0386b551
AM
6396 ;; Look for a trailing operator that concatenates the type
6397 ;; with a following one, and if so step past that one through
6398 ;; a recursive call. Note that we don't record concatenated
6399 ;; types in `c-found-types' - it's the component types that
6400 ;; are recorded when appropriate.
d9e94c22
MS
6401 (setq pos (point))
6402 (let* ((c-promote-possible-types (or (memq res '(t known))
6403 c-promote-possible-types))
6404 ;; If we can't promote then set `c-record-found-types' so that
6405 ;; we can merge in the types from the second part afterwards if
6406 ;; it turns out to be a known type there.
6407 (c-record-found-types (and c-record-type-identifiers
0386b551
AM
6408 (not c-promote-possible-types)))
6409 subres)
d9e94c22
MS
6410 (if (and (looking-at c-opt-type-concat-key)
6411
6412 (progn
6413 (goto-char (match-end 1))
6414 (c-forward-syntactic-ws)
0386b551 6415 (setq subres (c-forward-type))))
d9e94c22
MS
6416
6417 (progn
6418 ;; If either operand certainly is a type then both are, but we
6419 ;; don't let the existence of the operator itself promote two
6420 ;; uncertain types to a certain one.
6421 (cond ((eq res t))
0386b551
AM
6422 ((eq subres t)
6423 (unless (eq name-res 'template)
6424 (c-add-type id-start id-end))
6425 (when (and c-record-type-identifiers id-range)
d9e94c22
MS
6426 (c-record-type-id id-range))
6427 (setq res t))
2a15eb73 6428 ((eq res 'known))
0386b551 6429 ((eq subres 'known)
2a15eb73 6430 (setq res 'known))
d9e94c22 6431 ((eq res 'found))
0386b551 6432 ((eq subres 'found)
d9e94c22
MS
6433 (setq res 'found))
6434 (t
6435 (setq res 'maybe)))
6436
6437 (when (and (eq res t)
6438 (consp c-record-found-types))
6439 ;; Merge in the ranges of any types found by the second
6440 ;; `c-forward-type'.
6441 (setq c-record-type-identifiers
6442 ;; `nconc' doesn't mind that the tail of
6443 ;; `c-record-found-types' is t.
6444 (nconc c-record-found-types
6445 c-record-type-identifiers))))
6446
6447 (goto-char pos))))
6448
6449 (when (and c-record-found-types (memq res '(known found)) id-range)
6450 (setq c-record-found-types
6451 (cons id-range c-record-found-types))))
6452
6453 ;;(message "c-forward-type %s -> %s: %s" start (point) res)
6454
6455 res))
6456
452ea855
AM
6457(defun c-forward-annotation ()
6458 ;; Used for Java code only at the moment. Assumes point is on the
6459 ;; @, moves forward an annotation. returns nil if there is no
6460 ;; annotation at point.
6461 (and (looking-at "@")
6462 (progn (forward-char) t)
6463 (c-forward-type)
6464 (progn (c-forward-syntactic-ws) t)
6465 (if (looking-at "(")
6466 (c-go-list-forward)
6467 t)))
6468
785eecbb 6469\f
d9e94c22
MS
6470;; Handling of large scale constructs like statements and declarations.
6471
0386b551
AM
6472;; Macro used inside `c-forward-decl-or-cast-1'. It ought to be a
6473;; defsubst or perhaps even a defun, but it contains lots of free
6474;; variables that refer to things inside `c-forward-decl-or-cast-1'.
6475(defmacro c-fdoc-shift-type-backward (&optional short)
6476 ;; `c-forward-decl-or-cast-1' can consume an arbitrary length list
6477 ;; of types when parsing a declaration, which means that it
6478 ;; sometimes consumes the identifier in the declaration as a type.
6479 ;; This is used to "backtrack" and make the last type be treated as
6480 ;; an identifier instead.
6481 `(progn
6482 ,(unless short
6483 ;; These identifiers are bound only in the inner let.
6484 '(setq identifier-type at-type
6485 identifier-start type-start
6486 got-parens nil
6487 got-identifier t
6488 got-suffix t
6489 got-suffix-after-parens id-start
6490 paren-depth 0))
6491
6492 (if (setq at-type (if (eq backup-at-type 'prefix)
6493 t
6494 backup-at-type))
6495 (setq type-start backup-type-start
6496 id-start backup-id-start)
6497 (setq type-start start-pos
6498 id-start start-pos))
6499
6500 ;; When these flags already are set we've found specifiers that
6501 ;; unconditionally signal these attributes - backtracking doesn't
6502 ;; change that. So keep them set in that case.
6503 (or at-type-decl
6504 (setq at-type-decl backup-at-type-decl))
6505 (or maybe-typeless
6506 (setq maybe-typeless backup-maybe-typeless))
6507
6508 ,(unless short
6509 ;; This identifier is bound only in the inner let.
6510 '(setq start id-start))))
6511
6512(defun c-forward-decl-or-cast-1 (preceding-token-end context last-cast-end)
6513 ;; Move forward over a declaration or a cast if at the start of one.
6514 ;; The point is assumed to be at the start of some token. Nil is
6515 ;; returned if no declaration or cast is recognized, and the point
6516 ;; is clobbered in that case.
6517 ;;
6518 ;; If a declaration is parsed:
6519 ;;
6520 ;; The point is left at the first token after the first complete
6521 ;; declarator, if there is one. The return value is a cons where
022d0cf4
AM
6522 ;; the car is the position of the first token in the declarator. (See
6523 ;; below for the cdr.)
0386b551
AM
6524 ;; Some examples:
6525 ;;
6526 ;; void foo (int a, char *b) stuff ...
6527 ;; car ^ ^ point
6528 ;; float (*a)[], b;
6529 ;; car ^ ^ point
6530 ;; unsigned int a = c_style_initializer, b;
6531 ;; car ^ ^ point
6532 ;; unsigned int a (cplusplus_style_initializer), b;
6533 ;; car ^ ^ point (might change)
6534 ;; class Foo : public Bar {}
6535 ;; car ^ ^ point
6536 ;; class PikeClass (int a, string b) stuff ...
6537 ;; car ^ ^ point
6538 ;; enum bool;
6539 ;; car ^ ^ point
6540 ;; enum bool flag;
6541 ;; car ^ ^ point
6542 ;; void cplusplus_function (int x) throw (Bad);
6543 ;; car ^ ^ point
6544 ;; Foo::Foo (int b) : Base (b) {}
6545 ;; car ^ ^ point
91af3942 6546 ;;
e15f8aaa
AM
6547 ;; The cdr of the return value is non-nil when a
6548 ;; `c-typedef-decl-kwds' specifier is found in the declaration.
6549 ;; Specifically it is a dotted pair (A . B) where B is t when a
6550 ;; `c-typedef-kwds' ("typedef") is present, and A is t when some
6551 ;; other `c-typedef-decl-kwds' (e.g. class, struct, enum)
6552 ;; specifier is present. I.e., (some of) the declared
6553 ;; identifier(s) are types.
91af3942 6554 ;;
0386b551
AM
6555 ;; If a cast is parsed:
6556 ;;
6557 ;; The point is left at the first token after the closing paren of
6558 ;; the cast. The return value is `cast'. Note that the start
6559 ;; position must be at the first token inside the cast parenthesis
6560 ;; to recognize it.
6561 ;;
6562 ;; PRECEDING-TOKEN-END is the first position after the preceding
6563 ;; token, i.e. on the other side of the syntactic ws from the point.
6564 ;; Use a value less than or equal to (point-min) if the point is at
6565 ;; the first token in (the visible part of) the buffer.
6566 ;;
6567 ;; CONTEXT is a symbol that describes the context at the point:
022d0cf4 6568 ;; 'decl In a comma-separated declaration context (typically
0386b551
AM
6569 ;; inside a function declaration arglist).
6570 ;; '<> In an angle bracket arglist.
6571 ;; 'arglist Some other type of arglist.
a85fd6da
AM
6572 ;; nil Some other context or unknown context. Includes
6573 ;; within the parens of an if, for, ... construct.
0386b551
AM
6574 ;;
6575 ;; LAST-CAST-END is the first token after the closing paren of a
6576 ;; preceding cast, or nil if none is known. If
6577 ;; `c-forward-decl-or-cast-1' is used in succession, it should be
6578 ;; the position after the closest preceding call where a cast was
6579 ;; matched. In that case it's used to discover chains of casts like
6580 ;; "(a) (b) c".
6581 ;;
6582 ;; This function records identifier ranges on
6583 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
6584 ;; `c-record-type-identifiers' is non-nil.
6585 ;;
6586 ;; This function might do hidden buffer changes.
6587
6588 (let (;; `start-pos' is used below to point to the start of the
6589 ;; first type, i.e. after any leading specifiers. It might
6590 ;; also point at the beginning of the preceding syntactic
6591 ;; whitespace.
6592 (start-pos (point))
6593 ;; Set to the result of `c-forward-type'.
6594 at-type
6595 ;; The position of the first token in what we currently
6596 ;; believe is the type in the declaration or cast, after any
6597 ;; specifiers and their associated clauses.
6598 type-start
6599 ;; The position of the first token in what we currently
6600 ;; believe is the declarator for the first identifier. Set
6601 ;; when the type is found, and moved forward over any
6602 ;; `c-decl-hangon-kwds' and their associated clauses that
6603 ;; occurs after the type.
6604 id-start
6605 ;; These store `at-type', `type-start' and `id-start' of the
6606 ;; identifier before the one in those variables. The previous
6607 ;; identifier might turn out to be the real type in a
6608 ;; declaration if the last one has to be the declarator in it.
6609 ;; If `backup-at-type' is nil then the other variables have
6610 ;; undefined values.
6611 backup-at-type backup-type-start backup-id-start
e15f8aaa
AM
6612 ;; Set if we've found a specifier (apart from "typedef") that makes
6613 ;; the defined identifier(s) types.
0386b551 6614 at-type-decl
e15f8aaa
AM
6615 ;; Set if we've a "typedef" keyword.
6616 at-typedef
0386b551
AM
6617 ;; Set if we've found a specifier that can start a declaration
6618 ;; where there's no type.
6619 maybe-typeless
6620 ;; If a specifier is found that also can be a type prefix,
6621 ;; these flags are set instead of those above. If we need to
6622 ;; back up an identifier, they are copied to the real flag
6623 ;; variables. Thus they only take effect if we fail to
6624 ;; interpret it as a type.
6625 backup-at-type-decl backup-maybe-typeless
6626 ;; Whether we've found a declaration or a cast. We might know
6627 ;; this before we've found the type in it. It's 'ids if we've
6628 ;; found two consecutive identifiers (usually a sure sign, but
6629 ;; we should allow that in labels too), and t if we've found a
6630 ;; specifier keyword (a 100% sure sign).
6631 at-decl-or-cast
6632 ;; Set when we need to back up to parse this as a declaration
6633 ;; but not as a cast.
6634 backup-if-not-cast
6635 ;; For casts, the return position.
6636 cast-end
6637 ;; Save `c-record-type-identifiers' and
6638 ;; `c-record-ref-identifiers' since ranges are recorded
6639 ;; speculatively and should be thrown away if it turns out
6640 ;; that it isn't a declaration or cast.
6641 (save-rec-type-ids c-record-type-identifiers)
6642 (save-rec-ref-ids c-record-ref-identifiers))
6643
452ea855
AM
6644 (while (c-forward-annotation)
6645 (c-forward-syntactic-ws))
6646
0386b551
AM
6647 ;; Check for a type. Unknown symbols are treated as possible
6648 ;; types, but they could also be specifiers disguised through
6649 ;; macros like __INLINE__, so we recognize both types and known
6650 ;; specifiers after them too.
6651 (while
6652 (let* ((start (point)) kwd-sym kwd-clause-end found-type)
6653
6654 ;; Look for a specifier keyword clause.
e3715033
AM
6655 (when (or (looking-at c-prefix-spec-kwds-re)
6656 (and (c-major-mode-is 'java-mode)
6657 (looking-at "@[A-Za-z0-9]+")))
e15f8aaa
AM
6658 (if (looking-at c-typedef-key)
6659 (setq at-typedef t))
0386b551
AM
6660 (setq kwd-sym (c-keyword-sym (match-string 1)))
6661 (save-excursion
6662 (c-forward-keyword-clause 1)
6663 (setq kwd-clause-end (point))))
6664
e15f8aaa 6665 (when (setq found-type (c-forward-type t)) ; brace-block-too
0386b551
AM
6666 ;; Found a known or possible type or a prefix of a known type.
6667
6668 (when at-type
6669 ;; Got two identifiers with nothing but whitespace
6670 ;; between them. That can only happen in declarations.
6671 (setq at-decl-or-cast 'ids)
6672
6673 (when (eq at-type 'found)
6674 ;; If the previous identifier is a found type we
6675 ;; record it as a real one; it might be some sort of
6676 ;; alias for a prefix like "unsigned".
6677 (save-excursion
6678 (goto-char type-start)
6679 (let ((c-promote-possible-types t))
6680 (c-forward-type)))))
6681
6682 (setq backup-at-type at-type
6683 backup-type-start type-start
6684 backup-id-start id-start
6685 at-type found-type
6686 type-start start
6687 id-start (point)
6688 ;; The previous ambiguous specifier/type turned out
6689 ;; to be a type since we've parsed another one after
6690 ;; it, so clear these backup flags.
6691 backup-at-type-decl nil
6692 backup-maybe-typeless nil))
6693
6694 (if kwd-sym
6695 (progn
6696 ;; Handle known specifier keywords and
6697 ;; `c-decl-hangon-kwds' which can occur after known
6698 ;; types.
6699
6700 (if (c-keyword-member kwd-sym 'c-decl-hangon-kwds)
6701 ;; It's a hang-on keyword that can occur anywhere.
6702 (progn
6703 (setq at-decl-or-cast t)
6704 (if at-type
6705 ;; Move the identifier start position if
6706 ;; we've passed a type.
6707 (setq id-start kwd-clause-end)
6708 ;; Otherwise treat this as a specifier and
6709 ;; move the fallback position.
6710 (setq start-pos kwd-clause-end))
6711 (goto-char kwd-clause-end))
6712
6713 ;; It's an ordinary specifier so we know that
6714 ;; anything before this can't be the type.
6715 (setq backup-at-type nil
6716 start-pos kwd-clause-end)
6717
6718 (if found-type
6719 ;; It's ambiguous whether this keyword is a
6720 ;; specifier or a type prefix, so set the backup
6721 ;; flags. (It's assumed that `c-forward-type'
6722 ;; moved further than `c-forward-keyword-clause'.)
6723 (progn
6724 (when (c-keyword-member kwd-sym 'c-typedef-decl-kwds)
6725 (setq backup-at-type-decl t))
6726 (when (c-keyword-member kwd-sym 'c-typeless-decl-kwds)
6727 (setq backup-maybe-typeless t)))
6728
6729 (when (c-keyword-member kwd-sym 'c-typedef-decl-kwds)
e15f8aaa
AM
6730 ;; This test only happens after we've scanned a type.
6731 ;; So, with valid syntax, kwd-sym can't be 'typedef.
0386b551
AM
6732 (setq at-type-decl t))
6733 (when (c-keyword-member kwd-sym 'c-typeless-decl-kwds)
6734 (setq maybe-typeless t))
6735
ee7683eb 6736 ;; Haven't matched a type so it's an unambiguous
0386b551
AM
6737 ;; specifier keyword and we know we're in a
6738 ;; declaration.
6739 (setq at-decl-or-cast t)
6740
6741 (goto-char kwd-clause-end))))
6742
6743 ;; If the type isn't known we continue so that we'll jump
6744 ;; over all specifiers and type identifiers. The reason
6745 ;; to do this for a known type prefix is to make things
6746 ;; like "unsigned INT16" work.
6747 (and found-type (not (eq found-type t))))))
6748
6749 (cond
6750 ((eq at-type t)
6751 ;; If a known type was found, we still need to skip over any
6752 ;; hangon keyword clauses after it. Otherwise it has already
6753 ;; been done in the loop above.
6754 (while (looking-at c-decl-hangon-key)
6755 (c-forward-keyword-clause 1))
6756 (setq id-start (point)))
6757
6758 ((eq at-type 'prefix)
6759 ;; A prefix type is itself a primitive type when it's not
6760 ;; followed by another type.
6761 (setq at-type t))
6762
6763 ((not at-type)
6764 ;; Got no type but set things up to continue anyway to handle
6765 ;; the various cases when a declaration doesn't start with a
6766 ;; type.
6767 (setq id-start start-pos))
6768
6769 ((and (eq at-type 'maybe)
6770 (c-major-mode-is 'c++-mode))
6771 ;; If it's C++ then check if the last "type" ends on the form
6772 ;; "foo::foo" or "foo::~foo", i.e. if it's the name of a
6773 ;; (con|de)structor.
6774 (save-excursion
6775 (let (name end-2 end-1)
6776 (goto-char id-start)
6777 (c-backward-syntactic-ws)
6778 (setq end-2 (point))
6779 (when (and
6780 (c-simple-skip-symbol-backward)
6781 (progn
6782 (setq name
6783 (buffer-substring-no-properties (point) end-2))
6784 ;; Cheating in the handling of syntactic ws below.
6785 (< (skip-chars-backward ":~ \t\n\r\v\f") 0))
6786 (progn
6787 (setq end-1 (point))
6788 (c-simple-skip-symbol-backward))
6789 (>= (point) type-start)
6790 (equal (buffer-substring-no-properties (point) end-1)
6791 name))
6792 ;; It is a (con|de)structor name. In that case the
6793 ;; declaration is typeless so zap out any preceding
6794 ;; identifier(s) that we might have taken as types.
6795 (goto-char type-start)
6796 (setq at-type nil
6797 backup-at-type nil
6798 id-start type-start))))))
6799
6800 ;; Check for and step over a type decl expression after the thing
6801 ;; that is or might be a type. This can't be skipped since we
6802 ;; need the correct end position of the declarator for
6803 ;; `max-type-decl-end-*'.
6804 (let ((start (point)) (paren-depth 0) pos
6805 ;; True if there's a non-open-paren match of
6806 ;; `c-type-decl-prefix-key'.
6807 got-prefix
6808 ;; True if the declarator is surrounded by a parenthesis pair.
6809 got-parens
6810 ;; True if there is an identifier in the declarator.
6811 got-identifier
6812 ;; True if there's a non-close-paren match of
6813 ;; `c-type-decl-suffix-key'.
6814 got-suffix
6815 ;; True if there's a prefix match outside the outermost
6816 ;; paren pair that surrounds the declarator.
6817 got-prefix-before-parens
1379f2c5 6818 ;; True if there's a suffix match outside the outermost
0386b551
AM
6819 ;; paren pair that surrounds the declarator. The value is
6820 ;; the position of the first suffix match.
6821 got-suffix-after-parens
6822 ;; True if we've parsed the type decl to a token that is
6823 ;; known to end declarations in this context.
6824 at-decl-end
6825 ;; The earlier values of `at-type' and `type-start' if we've
6826 ;; shifted the type backwards.
6827 identifier-type identifier-start
6828 ;; If `c-parse-and-markup-<>-arglists' is set we need to
6829 ;; turn it off during the name skipping below to avoid
6830 ;; getting `c-type' properties that might be bogus. That
6831 ;; can happen since we don't know if
6832 ;; `c-restricted-<>-arglists' will be correct inside the
6833 ;; arglist paren that gets entered.
6834 c-parse-and-markup-<>-arglists)
6835
6836 (goto-char id-start)
6837
6838 ;; Skip over type decl prefix operators. (Note similar code in
6839 ;; `c-font-lock-declarators'.)
6840 (while (and (looking-at c-type-decl-prefix-key)
6841 (if (and (c-major-mode-is 'c++-mode)
05842630 6842 (match-beginning 3))
0386b551
AM
6843 ;; If the second submatch matches in C++ then
6844 ;; we're looking at an identifier that's a
6845 ;; prefix only if it specifies a member pointer.
6846 (when (setq got-identifier (c-forward-name))
6847 (if (looking-at "\\(::\\)")
6848 ;; We only check for a trailing "::" and
6849 ;; let the "*" that should follow be
6850 ;; matched in the next round.
6851 (progn (setq got-identifier nil) t)
6852 ;; It turned out to be the real identifier,
6853 ;; so stop.
6854 nil))
6855 t))
6856
6857 (if (eq (char-after) ?\()
6858 (progn
6859 (setq paren-depth (1+ paren-depth))
6860 (forward-char))
6861 (unless got-prefix-before-parens
6862 (setq got-prefix-before-parens (= paren-depth 0)))
6863 (setq got-prefix t)
6864 (goto-char (match-end 1)))
6865 (c-forward-syntactic-ws))
6866
6867 (setq got-parens (> paren-depth 0))
6868
6869 ;; Skip over an identifier.
6870 (or got-identifier
6871 (and (looking-at c-identifier-start)
6872 (setq got-identifier (c-forward-name))))
6873
6874 ;; Skip over type decl suffix operators.
6875 (while (if (looking-at c-type-decl-suffix-key)
6876
6877 (if (eq (char-after) ?\))
6878 (when (> paren-depth 0)
6879 (setq paren-depth (1- paren-depth))
6880 (forward-char)
6881 t)
6882 (when (if (save-match-data (looking-at "\\s\("))
6883 (c-safe (c-forward-sexp 1) t)
6884 (goto-char (match-end 1))
6885 t)
6886 (when (and (not got-suffix-after-parens)
6887 (= paren-depth 0))
6888 (setq got-suffix-after-parens (match-beginning 0)))
6889 (setq got-suffix t)))
6890
6891 ;; No suffix matched. We might have matched the
6892 ;; identifier as a type and the open paren of a
6893 ;; function arglist as a type decl prefix. In that
6894 ;; case we should "backtrack": Reinterpret the last
6895 ;; type as the identifier, move out of the arglist and
6896 ;; continue searching for suffix operators.
6897 ;;
6898 ;; Do this even if there's no preceding type, to cope
6899 ;; with old style function declarations in K&R C,
6900 ;; (con|de)structors in C++ and `c-typeless-decl-kwds'
6901 ;; style declarations. That isn't applicable in an
6902 ;; arglist context, though.
6903 (when (and (= paren-depth 1)
6904 (not got-prefix-before-parens)
6905 (not (eq at-type t))
6906 (or backup-at-type
6907 maybe-typeless
6908 backup-maybe-typeless
6909 (when c-recognize-typeless-decls
6910 (not context)))
6911 (setq pos (c-up-list-forward (point)))
6912 (eq (char-before pos) ?\)))
6913 (c-fdoc-shift-type-backward)
6914 (goto-char pos)
6915 t))
6916
6917 (c-forward-syntactic-ws))
6918
6919 (when (and (or maybe-typeless backup-maybe-typeless)
6920 (not got-identifier)
6921 (not got-prefix)
6922 at-type)
6923 ;; Have found no identifier but `c-typeless-decl-kwds' has
6924 ;; matched so we know we're inside a declaration. The
6925 ;; preceding type must be the identifier instead.
6926 (c-fdoc-shift-type-backward))
6927
6928 (setq
6929 at-decl-or-cast
6930 (catch 'at-decl-or-cast
6931
a85fd6da 6932 ;; CASE 1
0386b551
AM
6933 (when (> paren-depth 0)
6934 ;; Encountered something inside parens that isn't matched by
6935 ;; the `c-type-decl-*' regexps, so it's not a type decl
6936 ;; expression. Try to skip out to the same paren depth to
6937 ;; not confuse the cast check below.
6938 (c-safe (goto-char (scan-lists (point) 1 paren-depth)))
6939 ;; If we've found a specifier keyword then it's a
6940 ;; declaration regardless.
6941 (throw 'at-decl-or-cast (eq at-decl-or-cast t)))
6942
6943 (setq at-decl-end
6944 (looking-at (cond ((eq context '<>) "[,>]")
6945 (context "[,\)]")
6946 (t "[,;]"))))
6947
6948 ;; Now we've collected info about various characteristics of
6949 ;; the construct we're looking at. Below follows a decision
6950 ;; tree based on that. It's ordered to check more certain
6951 ;; signs before less certain ones.
6952
6953 (if got-identifier
6954 (progn
6955
a85fd6da 6956 ;; CASE 2
0386b551
AM
6957 (when (and (or at-type maybe-typeless)
6958 (not (or got-prefix got-parens)))
6959 ;; Got another identifier directly after the type, so it's a
6960 ;; declaration.
6961 (throw 'at-decl-or-cast t))
6962
6963 (when (and got-parens
6964 (not got-prefix)
6965 (not got-suffix-after-parens)
6966 (or backup-at-type
6967 maybe-typeless
6968 backup-maybe-typeless))
6969 ;; Got a declaration of the form "foo bar (gnu);" where we've
6970 ;; recognized "bar" as the type and "gnu" as the declarator.
6971 ;; In this case it's however more likely that "bar" is the
6972 ;; declarator and "gnu" a function argument or initializer (if
6973 ;; `c-recognize-paren-inits' is set), since the parens around
6974 ;; "gnu" would be superfluous if it's a declarator. Shift the
6975 ;; type one step backward.
6976 (c-fdoc-shift-type-backward)))
6977
6978 ;; Found no identifier.
6979
6980 (if backup-at-type
6981 (progn
6982
b248a85d
AM
6983
6984 ;; CASE 3
6985 (when (= (point) start)
6986 ;; Got a plain list of identifiers. If a colon follows it's
6987 ;; a valid label, or maybe a bitfield. Otherwise the last
6988 ;; one probably is the declared identifier and we should
6989 ;; back up to the previous type, providing it isn't a cast.
452ea855
AM
6990 (if (and (eq (char-after) ?:)
6991 (not (c-major-mode-is 'java-mode)))
b248a85d
AM
6992 (cond
6993 ;; If we've found a specifier keyword then it's a
6994 ;; declaration regardless.
6995 ((eq at-decl-or-cast t)
6996 (throw 'at-decl-or-cast t))
6997 ((and c-has-bitfields
6998 (eq at-decl-or-cast 'ids)) ; bitfield.
6999 (setq backup-if-not-cast t)
7000 (throw 'at-decl-or-cast t)))
7001
7002 (setq backup-if-not-cast t)
7003 (throw 'at-decl-or-cast t)))
0386b551 7004
a85fd6da 7005 ;; CASE 4
0386b551
AM
7006 (when (and got-suffix
7007 (not got-prefix)
7008 (not got-parens))
7009 ;; Got a plain list of identifiers followed by some suffix.
7010 ;; If this isn't a cast then the last identifier probably is
7011 ;; the declared one and we should back up to the previous
7012 ;; type.
7013 (setq backup-if-not-cast t)
7014 (throw 'at-decl-or-cast t)))
7015
a85fd6da 7016 ;; CASE 5
0386b551
AM
7017 (when (eq at-type t)
7018 ;; If the type is known we know that there can't be any
7019 ;; identifier somewhere else, and it's only in declarations in
7020 ;; e.g. function prototypes and in casts that the identifier may
7021 ;; be left out.
7022 (throw 'at-decl-or-cast t))
7023
7024 (when (= (point) start)
7025 ;; Only got a single identifier (parsed as a type so far).
a85fd6da 7026 ;; CASE 6
0386b551
AM
7027 (if (and
7028 ;; Check that the identifier isn't at the start of an
7029 ;; expression.
7030 at-decl-end
7031 (cond
7032 ((eq context 'decl)
7033 ;; Inside an arglist that contains declarations. If K&R
7034 ;; style declarations and parenthesis style initializers
7035 ;; aren't allowed then the single identifier must be a
7036 ;; type, else we require that it's known or found
7037 ;; (primitive types are handled above).
7038 (or (and (not c-recognize-knr-p)
7039 (not c-recognize-paren-inits))
7040 (memq at-type '(known found))))
7041 ((eq context '<>)
7042 ;; Inside a template arglist. Accept known and found
7043 ;; types; other identifiers could just as well be
7044 ;; constants in C++.
7045 (memq at-type '(known found)))))
7046 (throw 'at-decl-or-cast t)
a85fd6da 7047 ;; CASE 7
0386b551
AM
7048 ;; Can't be a valid declaration or cast, but if we've found a
7049 ;; specifier it can't be anything else either, so treat it as
7050 ;; an invalid/unfinished declaration or cast.
7051 (throw 'at-decl-or-cast at-decl-or-cast))))
7052
7053 (if (and got-parens
7054 (not got-prefix)
7055 (not context)
7056 (not (eq at-type t))
7057 (or backup-at-type
7058 maybe-typeless
7059 backup-maybe-typeless
7060 (when c-recognize-typeless-decls
7061 (or (not got-suffix)
7062 (not (looking-at
7063 c-after-suffixed-type-maybe-decl-key))))))
7064 ;; Got an empty paren pair and a preceding type that probably
7065 ;; really is the identifier. Shift the type backwards to make
7066 ;; the last one the identifier. This is analogous to the
7067 ;; "backtracking" done inside the `c-type-decl-suffix-key' loop
7068 ;; above.
7069 ;;
7070 ;; Exception: In addition to the conditions in that
7071 ;; "backtracking" code, do not shift backward if we're not
7072 ;; looking at either `c-after-suffixed-type-decl-key' or "[;,]".
7073 ;; Since there's no preceding type, the shift would mean that
7074 ;; the declaration is typeless. But if the regexp doesn't match
7075 ;; then we will simply fall through in the tests below and not
7076 ;; recognize it at all, so it's better to try it as an abstract
7077 ;; declarator instead.
7078 (c-fdoc-shift-type-backward)
7079
7080 ;; Still no identifier.
a85fd6da 7081 ;; CASE 8
0386b551
AM
7082 (when (and got-prefix (or got-parens got-suffix))
7083 ;; Require `got-prefix' together with either `got-parens' or
7084 ;; `got-suffix' to recognize it as an abstract declarator:
7085 ;; `got-parens' only is probably an empty function call.
7086 ;; `got-suffix' only can build an ordinary expression together
7087 ;; with the preceding identifier which we've taken as a type.
7088 ;; We could actually accept on `got-prefix' only, but that can
7089 ;; easily occur temporarily while writing an expression so we
7090 ;; avoid that case anyway. We could do a better job if we knew
7091 ;; the point when the fontification was invoked.
7092 (throw 'at-decl-or-cast t))
7093
a85fd6da 7094 ;; CASE 9
0386b551
AM
7095 (when (and at-type
7096 (not got-prefix)
7097 (not got-parens)
7098 got-suffix-after-parens
7099 (eq (char-after got-suffix-after-parens) ?\())
7100 ;; Got a type, no declarator but a paren suffix. I.e. it's a
fa463103 7101 ;; normal function call after all (or perhaps a C++ style object
0386b551
AM
7102 ;; instantiation expression).
7103 (throw 'at-decl-or-cast nil))))
7104
a85fd6da 7105 ;; CASE 10
0386b551
AM
7106 (when at-decl-or-cast
7107 ;; By now we've located the type in the declaration that we know
7108 ;; we're in.
7109 (throw 'at-decl-or-cast t))
7110
a85fd6da 7111 ;; CASE 11
0386b551
AM
7112 (when (and got-identifier
7113 (not context)
7114 (looking-at c-after-suffixed-type-decl-key)
7115 (if (and got-parens
7116 (not got-prefix)
7117 (not got-suffix)
7118 (not (eq at-type t)))
7119 ;; Shift the type backward in the case that there's a
7120 ;; single identifier inside parens. That can only
7121 ;; occur in K&R style function declarations so it's
7122 ;; more likely that it really is a function call.
7123 ;; Therefore we only do this after
7124 ;; `c-after-suffixed-type-decl-key' has matched.
7125 (progn (c-fdoc-shift-type-backward) t)
7126 got-suffix-after-parens))
7127 ;; A declaration according to `c-after-suffixed-type-decl-key'.
7128 (throw 'at-decl-or-cast t))
7129
a85fd6da 7130 ;; CASE 12
0386b551
AM
7131 (when (and (or got-prefix (not got-parens))
7132 (memq at-type '(t known)))
7133 ;; It's a declaration if a known type precedes it and it can't be a
7134 ;; function call.
7135 (throw 'at-decl-or-cast t))
7136
7137 ;; If we get here we can't tell if this is a type decl or a normal
7138 ;; expression by looking at it alone. (That's under the assumption
7139 ;; that normal expressions always can look like type decl expressions,
7140 ;; which isn't really true but the cases where it doesn't hold are so
7141 ;; uncommon (e.g. some placements of "const" in C++) it's not worth
7142 ;; the effort to look for them.)
7143
7144 (unless (or at-decl-end (looking-at "=[^=]"))
7145 ;; If this is a declaration it should end here or its initializer(*)
7146 ;; should start here, so check for allowed separation tokens. Note
7147 ;; that this rule doesn't work e.g. with a K&R arglist after a
7148 ;; function header.
7149 ;;
7150 ;; *) Don't check for C++ style initializers using parens
7151 ;; since those already have been matched as suffixes.
7152 ;;
7153 ;; If `at-decl-or-cast' is then we've found some other sign that
7154 ;; it's a declaration or cast, so then it's probably an
7155 ;; invalid/unfinished one.
7156 (throw 'at-decl-or-cast at-decl-or-cast))
7157
7158 ;; Below are tests that only should be applied when we're certain to
7159 ;; not have parsed halfway through an expression.
7160
a85fd6da 7161 ;; CASE 14
0386b551
AM
7162 (when (memq at-type '(t known))
7163 ;; The expression starts with a known type so treat it as a
7164 ;; declaration.
7165 (throw 'at-decl-or-cast t))
7166
a85fd6da 7167 ;; CASE 15
0386b551
AM
7168 (when (and (c-major-mode-is 'c++-mode)
7169 ;; In C++ we check if the identifier is a known type, since
7170 ;; (con|de)structors use the class name as identifier.
7171 ;; We've always shifted over the identifier as a type and
7172 ;; then backed up again in this case.
7173 identifier-type
7174 (or (memq identifier-type '(found known))
7175 (and (eq (char-after identifier-start) ?~)
7176 ;; `at-type' probably won't be 'found for
7177 ;; destructors since the "~" is then part of the
7178 ;; type name being checked against the list of
7179 ;; known types, so do a check without that
7180 ;; operator.
7181 (or (save-excursion
7182 (goto-char (1+ identifier-start))
7183 (c-forward-syntactic-ws)
7184 (c-with-syntax-table
7185 c-identifier-syntax-table
7186 (looking-at c-known-type-key)))
7187 (save-excursion
7188 (goto-char (1+ identifier-start))
7189 ;; We have already parsed the type earlier,
7190 ;; so it'd be possible to cache the end
7191 ;; position instead of redoing it here, but
7192 ;; then we'd need to keep track of another
7193 ;; position everywhere.
7194 (c-check-type (point)
7195 (progn (c-forward-type)
7196 (point))))))))
7197 (throw 'at-decl-or-cast t))
7198
7199 (if got-identifier
7200 (progn
a85fd6da 7201 ;; CASE 16
0386b551
AM
7202 (when (and got-prefix-before-parens
7203 at-type
7204 (or at-decl-end (looking-at "=[^=]"))
7205 (not context)
7206 (not got-suffix))
7207 ;; Got something like "foo * bar;". Since we're not inside an
7208 ;; arglist it would be a meaningless expression because the
7209 ;; result isn't used. We therefore choose to recognize it as
7210 ;; a declaration. Do not allow a suffix since it could then
7211 ;; be a function call.
7212 (throw 'at-decl-or-cast t))
7213
a85fd6da 7214 ;; CASE 17
0386b551
AM
7215 (when (and (or got-suffix-after-parens
7216 (looking-at "=[^=]"))
7217 (eq at-type 'found)
7218 (not (eq context 'arglist)))
7219 ;; Got something like "a (*b) (c);" or "a (b) = c;". It could
7220 ;; be an odd expression or it could be a declaration. Treat
7221 ;; it as a declaration if "a" has been used as a type
7222 ;; somewhere else (if it's a known type we won't get here).
7223 (throw 'at-decl-or-cast t)))
7224
a85fd6da 7225 ;; CASE 18
0386b551
AM
7226 (when (and context
7227 (or got-prefix
7228 (and (eq context 'decl)
7229 (not c-recognize-paren-inits)
7230 (or got-parens got-suffix))))
7231 ;; Got a type followed by an abstract declarator. If `got-prefix'
7232 ;; is set it's something like "a *" without anything after it. If
7233 ;; `got-parens' or `got-suffix' is set it's "a()", "a[]", "a()[]",
7234 ;; or similar, which we accept only if the context rules out
7235 ;; expressions.
7236 (throw 'at-decl-or-cast t)))
7237
7238 ;; If we had a complete symbol table here (which rules out
7239 ;; `c-found-types') we should return t due to the disambiguation rule
7240 ;; (in at least C++) that anything that can be parsed as a declaration
7241 ;; is a declaration. Now we're being more defensive and prefer to
7242 ;; highlight things like "foo (bar);" as a declaration only if we're
7243 ;; inside an arglist that contains declarations.
7244 (eq context 'decl))))
7245
7246 ;; The point is now after the type decl expression.
7247
7248 (cond
7249 ;; Check for a cast.
7250 ((save-excursion
7251 (and
7252 c-cast-parens
7253
7254 ;; Should be the first type/identifier in a cast paren.
7255 (> preceding-token-end (point-min))
7256 (memq (char-before preceding-token-end) c-cast-parens)
7257
7258 ;; The closing paren should follow.
7259 (progn
7260 (c-forward-syntactic-ws)
7261 (looking-at "\\s\)"))
7262
7263 ;; There should be a primary expression after it.
7264 (let (pos)
7265 (forward-char)
7266 (c-forward-syntactic-ws)
7267 (setq cast-end (point))
7268 (and (looking-at c-primary-expr-regexp)
7269 (progn
7270 (setq pos (match-end 0))
7271 (or
7272 ;; Check if the expression begins with a prefix keyword.
7273 (match-beginning 2)
7274 (if (match-beginning 1)
7275 ;; Expression begins with an ambiguous operator. Treat
7276 ;; it as a cast if it's a type decl or if we've
7277 ;; recognized the type somewhere else.
7278 (or at-decl-or-cast
7279 (memq at-type '(t known found)))
7280 ;; Unless it's a keyword, it's the beginning of a primary
7281 ;; expression.
7282 (not (looking-at c-keywords-regexp)))))
7283 ;; If `c-primary-expr-regexp' matched a nonsymbol token, check
7284 ;; that it matched a whole one so that we don't e.g. confuse
7285 ;; the operator '-' with '->'. It's ok if it matches further,
7286 ;; though, since it e.g. can match the float '.5' while the
7287 ;; operator regexp only matches '.'.
7288 (or (not (looking-at c-nonsymbol-token-regexp))
7289 (<= (match-end 0) pos))))
7290
7291 ;; There should either be a cast before it or something that isn't an
7292 ;; identifier or close paren.
7293 (> preceding-token-end (point-min))
7294 (progn
7295 (goto-char (1- preceding-token-end))
7296 (or (eq (point) last-cast-end)
7297 (progn
7298 (c-backward-syntactic-ws)
7299 (if (< (skip-syntax-backward "w_") 0)
7300 ;; It's a symbol. Accept it only if it's one of the
7301 ;; keywords that can precede an expression (without
7302 ;; surrounding parens).
7303 (looking-at c-simple-stmt-key)
7304 (and
7305 ;; Check that it isn't a close paren (block close is ok,
7306 ;; though).
7307 (not (memq (char-before) '(?\) ?\])))
7308 ;; Check that it isn't a nonsymbol identifier.
7309 (not (c-on-identifier)))))))))
7310
7311 ;; Handle the cast.
7312 (when (and c-record-type-identifiers at-type (not (eq at-type t)))
7313 (let ((c-promote-possible-types t))
7314 (goto-char type-start)
7315 (c-forward-type)))
7316
7317 (goto-char cast-end)
7318 'cast)
7319
7320 (at-decl-or-cast
7321 ;; We're at a declaration. Highlight the type and the following
7322 ;; declarators.
7323
7324 (when backup-if-not-cast
7325 (c-fdoc-shift-type-backward t))
7326
7327 (when (and (eq context 'decl) (looking-at ","))
7328 ;; Make sure to propagate the `c-decl-arg-start' property to
7329 ;; the next argument if it's set in this one, to cope with
7330 ;; interactive refontification.
7331 (c-put-c-type-property (point) 'c-decl-arg-start))
7332
7333 (when (and c-record-type-identifiers at-type (not (eq at-type t)))
7334 (let ((c-promote-possible-types t))
7335 (save-excursion
7336 (goto-char type-start)
7337 (c-forward-type))))
7338
e15f8aaa
AM
7339 (cons id-start
7340 (and (or at-type-decl at-typedef)
7341 (cons at-type-decl at-typedef))))
0386b551
AM
7342
7343 (t
7344 ;; False alarm. Restore the recorded ranges.
7345 (setq c-record-type-identifiers save-rec-type-ids
7346 c-record-ref-identifiers save-rec-ref-ids)
7347 nil))))
7348
7349(defun c-forward-label (&optional assume-markup preceding-token-end limit)
51c9af45 7350 ;; Assuming that point is at the beginning of a token, check if it starts a
1379f2c5
AM
7351 ;; label and if so move over it and return non-nil (t in default situations,
7352 ;; specific symbols (see below) for interesting situations), otherwise don't
7353 ;; move and return nil. "Label" here means "most things with a colon".
51c9af45
AM
7354 ;;
7355 ;; More precisely, a "label" is regarded as one of:
1379f2c5
AM
7356 ;; (i) a goto target like "foo:" - returns the symbol `goto-target';
7357 ;; (ii) A case label - either the entire construct "case FOO:", or just the
7358 ;; bare "case", should the colon be missing. We return t;
7359 ;; (iii) a keyword which needs a colon, like "default:" or "private:"; We
7360 ;; return t;
51c9af45 7361 ;; (iv) One of QT's "extended" C++ variants of
1379f2c5
AM
7362 ;; "private:"/"protected:"/"public:"/"more:" looking like "public slots:".
7363 ;; Returns the symbol `qt-2kwds-colon'.
7364 ;; (v) QT's construct "signals:". Returns the symbol `qt-1kwd-colon'.
82ba65cf 7365 ;; (vi) One of the keywords matched by `c-opt-extra-label-key' (without any
51c9af45 7366 ;; colon). Currently (2006-03), this applies only to Objective C's
1379f2c5 7367 ;; keywords "@private", "@protected", and "@public". Returns t.
51c9af45 7368 ;;
e1dbe924 7369 ;; One of the things which will NOT be recognized as a label is a bit-field
51c9af45
AM
7370 ;; element of a struct, something like "int foo:5".
7371 ;;
7372 ;; The end of the label is taken to be just after the colon, or the end of
7373 ;; the first submatch in `c-opt-extra-label-key'. The point is directly
7374 ;; after the end on return. The terminating char gets marked with
7375 ;; `c-decl-end' to improve recognition of the following declaration or
7376 ;; statement.
0386b551
AM
7377 ;;
7378 ;; If ASSUME-MARKUP is non-nil, it's assumed that the preceding
51c9af45 7379 ;; label, if any, has already been marked up like that.
0386b551
AM
7380 ;;
7381 ;; If PRECEDING-TOKEN-END is given, it should be the first position
7382 ;; after the preceding token, i.e. on the other side of the
7383 ;; syntactic ws from the point. Use a value less than or equal to
7384 ;; (point-min) if the point is at the first token in (the visible
7385 ;; part of) the buffer.
7386 ;;
7387 ;; The optional LIMIT limits the forward scan for the colon.
7388 ;;
7389 ;; This function records the ranges of the label symbols on
7390 ;; `c-record-ref-identifiers' if `c-record-type-identifiers' (!) is
7391 ;; non-nil.
7392 ;;
7393 ;; This function might do hidden buffer changes.
7394
51c9af45 7395 (let ((start (point))
1379f2c5 7396 label-end
51c9af45 7397 qt-symbol-idx
1379f2c5 7398 macro-start ; if we're in one.
f412a567
AM
7399 label-type
7400 kwd)
0386b551 7401 (cond
b414f371 7402 ;; "case" or "default" (Doesn't apply to AWK).
0386b551
AM
7403 ((looking-at c-label-kwds-regexp)
7404 (let ((kwd-end (match-end 1)))
7405 ;; Record only the keyword itself for fontification, since in
7406 ;; case labels the following is a constant expression and not
7407 ;; a label.
7408 (when c-record-type-identifiers
7409 (c-record-ref-id (cons (match-beginning 1) kwd-end)))
7410
7411 ;; Find the label end.
7412 (goto-char kwd-end)
1379f2c5
AM
7413 (setq label-type
7414 (if (and (c-syntactic-re-search-forward
7415 ;; Stop on chars that aren't allowed in expressions,
7416 ;; and on operator chars that would be meaningless
7417 ;; there. FIXME: This doesn't cope with ?: operators.
7418 "[;{=,@]\\|\\(\\=\\|[^:]\\):\\([^:]\\|\\'\\)"
7419 limit t t nil 1)
7420 (match-beginning 2))
7421
7422 (progn ; there's a proper :
7423 (goto-char (match-beginning 2)) ; just after the :
7424 (c-put-c-type-property (1- (point)) 'c-decl-end)
7425 t)
7426
7427 ;; It's an unfinished label. We consider the keyword enough
7428 ;; to recognize it as a label, so that it gets fontified.
7429 ;; Leave the point at the end of it, but don't put any
7430 ;; `c-decl-end' marker.
7431 (goto-char kwd-end)
7432 t))))
0386b551 7433
51c9af45 7434 ;; @private, @protected, @public, in Objective C, or similar.
0386b551
AM
7435 ((and c-opt-extra-label-key
7436 (looking-at c-opt-extra-label-key))
7437 ;; For a `c-opt-extra-label-key' match, we record the whole
7438 ;; thing for fontification. That's to get the leading '@' in
7439 ;; Objective-C protection labels fontified.
7440 (goto-char (match-end 1))
7441 (when c-record-type-identifiers
7442 (c-record-ref-id (cons (match-beginning 1) (point))))
7443 (c-put-c-type-property (1- (point)) 'c-decl-end)
1379f2c5 7444 (setq label-type t))
0386b551 7445
51c9af45
AM
7446 ;; All other cases of labels.
7447 ((and c-recognize-colon-labels ; nil for AWK and IDL, otherwise t.
0386b551
AM
7448
7449 ;; A colon label must have something before the colon.
7450 (not (eq (char-after) ?:))
7451
7452 ;; Check that we're not after a token that can't precede a label.
7453 (or
7454 ;; Trivially succeeds when there's no preceding token.
536610a4 7455 ;; Succeeds when we're at a virtual semicolon.
0386b551
AM
7456 (if preceding-token-end
7457 (<= preceding-token-end (point-min))
7458 (save-excursion
7459 (c-backward-syntactic-ws)
7460 (setq preceding-token-end (point))
536610a4
AM
7461 (or (bobp)
7462 (c-at-vsemi-p))))
0386b551
AM
7463
7464 ;; Check if we're after a label, if we're after a closing
7465 ;; paren that belong to statement, and with
7466 ;; `c-label-prefix-re'. It's done in different order
7467 ;; depending on `assume-markup' since the checks have
7468 ;; different expensiveness.
7469 (if assume-markup
7470 (or
7471 (eq (c-get-char-property (1- preceding-token-end) 'c-type)
7472 'c-decl-end)
7473
7474 (save-excursion
7475 (goto-char (1- preceding-token-end))
7476 (c-beginning-of-current-token)
51c9af45
AM
7477 (or (looking-at c-label-prefix-re)
7478 (looking-at c-block-stmt-1-key)))
0386b551
AM
7479
7480 (and (eq (char-before preceding-token-end) ?\))
7481 (c-after-conditional)))
7482
7483 (or
7484 (save-excursion
7485 (goto-char (1- preceding-token-end))
7486 (c-beginning-of-current-token)
51c9af45
AM
7487 (or (looking-at c-label-prefix-re)
7488 (looking-at c-block-stmt-1-key)))
0386b551
AM
7489
7490 (cond
7491 ((eq (char-before preceding-token-end) ?\))
7492 (c-after-conditional))
7493
7494 ((eq (char-before preceding-token-end) ?:)
7495 ;; Might be after another label, so check it recursively.
51c9af45
AM
7496 (save-restriction
7497 (save-excursion
7498 (goto-char (1- preceding-token-end))
7499 ;; Essentially the same as the
7500 ;; `c-syntactic-re-search-forward' regexp below.
7501 (setq macro-start
7502 (save-excursion (and (c-beginning-of-macro)
7503 (point))))
7504 (if macro-start (narrow-to-region macro-start (point-max)))
7505 (c-syntactic-skip-backward "^-]:?;}=*/%&|,<>!@+" nil t)
7506 ;; Note: the following should work instead of the
7507 ;; narrow-to-region above. Investigate why not,
7508 ;; sometime. ACM, 2006-03-31.
7509 ;; (c-syntactic-skip-backward "^-]:?;}=*/%&|,<>!@+"
7510 ;; macro-start t)
7511 (let ((pte (point))
7512 ;; If the caller turned on recording for us,
7513 ;; it shouldn't apply when we check the
7514 ;; preceding label.
7515 c-record-type-identifiers)
7516 ;; A label can't start at a cpp directive. Check for
7517 ;; this, since c-forward-syntactic-ws would foul up on it.
7518 (unless (and c-opt-cpp-prefix (looking-at c-opt-cpp-prefix))
7519 (c-forward-syntactic-ws)
7520 (c-forward-label nil pte start))))))))))
7521
1379f2c5 7522 ;; Point is still at the beginning of the possible label construct.
b414f371 7523 ;;
51c9af45
AM
7524 ;; Check that the next nonsymbol token is ":", or that we're in one
7525 ;; of QT's "slots" declarations. Allow '(' for the sake of macro
7526 ;; arguments. FIXME: Should build this regexp from the language
7527 ;; constants.
1379f2c5
AM
7528 (cond
7529 ;; public: protected: private:
7530 ((and
7531 (c-major-mode-is 'c++-mode)
7532 (search-forward-regexp
7533 "\\=p\\(r\\(ivate\\|otected\\)\\|ublic\\)\\>[^_]" nil t)
7534 (progn (backward-char)
7535 (c-forward-syntactic-ws limit)
7536 (looking-at ":\\([^:]\\|\\'\\)"))) ; A single colon.
7537 (forward-char)
7538 (setq label-type t))
7539 ;; QT double keyword like "protected slots:" or goto target.
7540 ((progn (goto-char start) nil))
7541 ((when (c-syntactic-re-search-forward
7542 "[ \t\n[:?;{=*/%&|,<>!@+-]" limit t t) ; not at EOB
7543 (backward-char)
7544 (setq label-end (point))
7545 (setq qt-symbol-idx
7546 (and (c-major-mode-is 'c++-mode)
7547 (string-match
7548 "\\(p\\(r\\(ivate\\|otected\\)\\|ublic\\)\\|more\\)\\>"
b414f371 7549 (buffer-substring start (point)))))
1379f2c5
AM
7550 (c-forward-syntactic-ws limit)
7551 (cond
7552 ((looking-at ":\\([^:]\\|\\'\\)") ; A single colon.
7553 (forward-char)
7554 (setq label-type
f412a567
AM
7555 (if (or (string= "signals" ; Special QT macro
7556 (setq kwd (buffer-substring-no-properties start label-end)))
7557 (string= "Q_SIGNALS" kwd))
1379f2c5
AM
7558 'qt-1kwd-colon
7559 'goto-target)))
7560 ((and qt-symbol-idx
f412a567 7561 (search-forward-regexp "\\=\\(slots\\|Q_SLOTS\\)\\>" limit t)
1379f2c5
AM
7562 (progn (c-forward-syntactic-ws limit)
7563 (looking-at ":\\([^:]\\|\\'\\)"))) ; A single colon
7564 (forward-char)
7565 (setq label-type 'qt-2kwds-colon)))))))
0386b551
AM
7566
7567 (save-restriction
7568 (narrow-to-region start (point))
7569
7570 ;; Check that `c-nonlabel-token-key' doesn't match anywhere.
7571 (catch 'check-label
7572 (goto-char start)
7573 (while (progn
7574 (when (looking-at c-nonlabel-token-key)
7575 (goto-char start)
1379f2c5 7576 (setq label-type nil)
0386b551
AM
7577 (throw 'check-label nil))
7578 (and (c-safe (c-forward-sexp)
7579 (c-forward-syntactic-ws)
7580 t)
7581 (not (eobp)))))
7582
7583 ;; Record the identifiers in the label for fontification, unless
7584 ;; it begins with `c-label-kwds' in which case the following
7585 ;; identifiers are part of a (constant) expression that
7586 ;; shouldn't be fontified.
7587 (when (and c-record-type-identifiers
7588 (progn (goto-char start)
7589 (not (looking-at c-label-kwds-regexp))))
7590 (while (c-syntactic-re-search-forward c-symbol-key nil t)
7591 (c-record-ref-id (cons (match-beginning 0)
7592 (match-end 0)))))
7593
7594 (c-put-c-type-property (1- (point-max)) 'c-decl-end)
82ba65cf 7595 (goto-char (point-max)))))
0386b551
AM
7596
7597 (t
7598 ;; Not a label.
1379f2c5
AM
7599 (goto-char start)))
7600 label-type))
0386b551
AM
7601
7602(defun c-forward-objc-directive ()
7603 ;; Assuming the point is at the beginning of a token, try to move
7604 ;; forward to the end of the Objective-C directive that starts
7605 ;; there. Return t if a directive was fully recognized, otherwise
7606 ;; the point is moved as far as one could be successfully parsed and
7607 ;; nil is returned.
7608 ;;
7609 ;; This function records identifier ranges on
7610 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
7611 ;; `c-record-type-identifiers' is non-nil.
7612 ;;
7613 ;; This function might do hidden buffer changes.
7614
7615 (let ((start (point))
7616 start-char
7617 (c-promote-possible-types t)
667ced3a 7618 lim
0386b551
AM
7619 ;; Turn off recognition of angle bracket arglists while parsing
7620 ;; types here since the protocol reference list might then be
7621 ;; considered part of the preceding name or superclass-name.
7622 c-recognize-<>-arglists)
7623
7624 (if (or
7625 (when (looking-at
7626 (eval-when-compile
7627 (c-make-keywords-re t
7628 (append (c-lang-const c-protection-kwds objc)
7629 '("@end"))
7630 'objc-mode)))
7631 (goto-char (match-end 1))
7632 t)
7633
7634 (and
7635 (looking-at
7636 (eval-when-compile
7637 (c-make-keywords-re t
7638 '("@interface" "@implementation" "@protocol")
7639 'objc-mode)))
7640
7641 ;; Handle the name of the class itself.
7642 (progn
cb694ab7
AM
7643; (c-forward-token-2) ; 2006/1/13 This doesn't move if the token's
7644; at EOB.
7645 (goto-char (match-end 0))
667ced3a 7646 (setq lim (point))
cb694ab7 7647 (c-skip-ws-forward)
0386b551
AM
7648 (c-forward-type))
7649
7650 (catch 'break
7651 ;; Look for ": superclass-name" or "( category-name )".
7652 (when (looking-at "[:\(]")
7653 (setq start-char (char-after))
7654 (forward-char)
7655 (c-forward-syntactic-ws)
7656 (unless (c-forward-type) (throw 'break nil))
7657 (when (eq start-char ?\()
7658 (unless (eq (char-after) ?\)) (throw 'break nil))
7659 (forward-char)
7660 (c-forward-syntactic-ws)))
7661
7662 ;; Look for a protocol reference list.
7663 (if (eq (char-after) ?<)
7664 (let ((c-recognize-<>-arglists t)
7665 (c-parse-and-markup-<>-arglists t)
7666 c-restricted-<>-arglists)
7667 (c-forward-<>-arglist t))
7668 t))))
7669
7670 (progn
667ced3a 7671 (c-backward-syntactic-ws lim)
0386b551
AM
7672 (c-clear-c-type-property start (1- (point)) 'c-decl-end)
7673 (c-put-c-type-property (1- (point)) 'c-decl-end)
7674 t)
7675
7676 (c-clear-c-type-property start (point) 'c-decl-end)
7677 nil)))
7678
785eecbb
RS
7679(defun c-beginning-of-inheritance-list (&optional lim)
7680 ;; Go to the first non-whitespace after the colon that starts a
7681 ;; multiple inheritance introduction. Optional LIM is the farthest
7682 ;; back we should search.
0386b551
AM
7683 ;;
7684 ;; This function might do hidden buffer changes.
7685 (c-with-syntax-table c++-template-syntax-table
7686 (c-backward-token-2 0 t lim)
7687 (while (and (or (looking-at c-symbol-start)
7688 (looking-at "[<,]\\|::"))
7689 (zerop (c-backward-token-2 1 t lim))))))
785eecbb 7690
785eecbb
RS
7691(defun c-in-method-def-p ()
7692 ;; Return nil if we aren't in a method definition, otherwise the
7693 ;; position of the initial [+-].
0386b551
AM
7694 ;;
7695 ;; This function might do hidden buffer changes.
785eecbb
RS
7696 (save-excursion
7697 (beginning-of-line)
a66cd3ee
MS
7698 (and c-opt-method-key
7699 (looking-at c-opt-method-key)
785eecbb
RS
7700 (point))
7701 ))
7702
a66cd3ee
MS
7703;; Contributed by Kevin Ryde <user42@zip.com.au>.
7704(defun c-in-gcc-asm-p ()
7705 ;; Return non-nil if point is within a gcc \"asm\" block.
7706 ;;
7707 ;; This should be called with point inside an argument list.
7708 ;;
7709 ;; Only one level of enclosing parentheses is considered, so for
7710 ;; instance `nil' is returned when in a function call within an asm
7711 ;; operand.
0386b551
AM
7712 ;;
7713 ;; This function might do hidden buffer changes.
a66cd3ee
MS
7714
7715 (and c-opt-asm-stmt-key
7716 (save-excursion
7717 (beginning-of-line)
7718 (backward-up-list 1)
7719 (c-beginning-of-statement-1 (point-min) nil t)
7720 (looking-at c-opt-asm-stmt-key))))
7721
abb7e5cf 7722(defun c-at-toplevel-p ()
a85fd6da
AM
7723 "Return a determination as to whether point is \"at the top level\".
7724Informally, \"at the top level\" is anywhere where you can write
7725a function.
7726
7727More precisely, being at the top-level means that point is either
7728outside any enclosing block (such as a function definition), or
7729directly inside a class, namespace or other block that contains
7730another declaration level.
abb7e5cf
SM
7731
7732If point is not at the top-level (e.g. it is inside a method
7733definition), then nil is returned. Otherwise, if point is at a
7734top-level not enclosed within a class definition, t is returned.
7735Otherwise, a 2-vector is returned where the zeroth element is the
7736buffer position of the start of the class declaration, and the first
7737element is the buffer position of the enclosing class's opening
0386b551
AM
7738brace.
7739
7740Note that this function might do hidden buffer changes. See the
7741comment at the start of cc-engine.el for more info."
a66cd3ee
MS
7742 (let ((paren-state (c-parse-state)))
7743 (or (not (c-most-enclosing-brace paren-state))
7744 (c-search-uplist-for-classkey paren-state))))
7745
d9e94c22 7746(defun c-just-after-func-arglist-p (&optional lim)
0386b551
AM
7747 ;; Return non-nil if the point is in the region after the argument
7748 ;; list of a function and its opening brace (or semicolon in case it
7749 ;; got no body). If there are K&R style argument declarations in
7750 ;; that region, the point has to be inside the first one for this
7751 ;; function to recognize it.
a66cd3ee 7752 ;;
0386b551
AM
7753 ;; If successful, the point is moved to the first token after the
7754 ;; function header (see `c-forward-decl-or-cast-1' for details) and
7755 ;; the position of the opening paren of the function arglist is
7756 ;; returned.
7757 ;;
7758 ;; The point is clobbered if not successful.
7759 ;;
7760 ;; LIM is used as bound for backward buffer searches.
7761 ;;
7762 ;; This function might do hidden buffer changes.
7763
7764 (let ((beg (point)) end id-start)
7765 (and
7766 (eq (c-beginning-of-statement-1 lim) 'same)
7767
2e492df3
AM
7768 (not (and (c-major-mode-is 'objc-mode)
7769 (c-forward-objc-directive)))
0386b551
AM
7770
7771 (setq id-start
7772 (car-safe (c-forward-decl-or-cast-1 (c-point 'bosws) nil nil)))
7773 (< id-start beg)
7774
7775 ;; There should not be a '=' or ',' between beg and the
7776 ;; start of the declaration since that means we were in the
7777 ;; "expression part" of the declaration.
7778 (or (> (point) beg)
7779 (not (looking-at "[=,]")))
7780
7781 (save-excursion
7782 ;; Check that there's an arglist paren in the
7783 ;; declaration.
7784 (goto-char id-start)
7785 (cond ((eq (char-after) ?\()
7786 ;; The declarator is a paren expression, so skip past it
7787 ;; so that we don't get stuck on that instead of the
7788 ;; function arglist.
7789 (c-forward-sexp))
51c9af45
AM
7790 ((and c-opt-op-identifier-prefix
7791 (looking-at c-opt-op-identifier-prefix))
0386b551
AM
7792 ;; Don't trip up on "operator ()".
7793 (c-forward-token-2 2 t)))
7794 (and (< (point) beg)
7795 (c-syntactic-re-search-forward "(" beg t t)
7796 (1- (point)))))))
785eecbb 7797
a66cd3ee
MS
7798(defun c-in-knr-argdecl (&optional lim)
7799 ;; Return the position of the first argument declaration if point is
7800 ;; inside a K&R style argument declaration list, nil otherwise.
7801 ;; `c-recognize-knr-p' is not checked. If LIM is non-nil, it's a
7802 ;; position that bounds the backward search for the argument list.
7803 ;;
9cf17ef1
AM
7804 ;; Point must be within a possible K&R region, e.g. just before a top-level
7805 ;; "{". It must be outside of parens and brackets. The test can return
7806 ;; false positives otherwise.
0386b551
AM
7807 ;;
7808 ;; This function might do hidden buffer changes.
d9e94c22 7809
a66cd3ee
MS
7810 (save-excursion
7811 (save-restriction
9cf17ef1
AM
7812 ;; If we're in a macro, our search range is restricted to it. Narrow to
7813 ;; the searchable range.
98d7371e
AM
7814 (let* ((macro-start (save-excursion (and (c-beginning-of-macro) (point))))
7815 (macro-end (save-excursion (and macro-start (c-end-of-macro) (point))))
7816 (low-lim (max (or lim (point-min)) (or macro-start (point-min))))
6b6481ed 7817 before-lparen after-rparen
98d7371e
AM
7818 (pp-count-out 20)) ; Max number of paren/brace constructs before
7819 ; we give up
7820 (narrow-to-region low-lim (or macro-end (point-max)))
9cf17ef1
AM
7821
7822 ;; Search backwards for the defun's argument list. We give up if we
98d7371e
AM
7823 ;; encounter a "}" (end of a previous defun) an "=" (which can't be in
7824 ;; a knr region) or BOB.
9cf17ef1
AM
7825 ;;
7826 ;; The criterion for a paren structure being the arg list is:
7827 ;; o - there is non-WS stuff after it but before any "{"; AND
7828 ;; o - the token after it isn't a ";" AND
7829 ;; o - it is preceded by either an identifier (the function name) or
7830 ;; a macro expansion like "DEFUN (...)"; AND
7831 ;; o - its content is a non-empty comma-separated list of identifiers
7832 ;; (an empty arg list won't have a knr region).
7833 ;;
7834 ;; The following snippet illustrates these rules:
7835 ;; int foo (bar, baz, yuk)
7836 ;; int bar [] ;
7837 ;; int (*baz) (my_type) ;
7838 ;; int (*) (void) (*yuk) (void) ;
7839 ;; {
7840
7841 (catch 'knr
6b6481ed
AM
7842 (while (> pp-count-out 0) ; go back one paren/bracket pair each time.
7843 (setq pp-count-out (1- pp-count-out))
98d7371e 7844 (c-syntactic-skip-backward "^)]}=")
9cf17ef1
AM
7845 (cond ((eq (char-before) ?\))
7846 (setq after-rparen (point)))
10489fcb
AM
7847 ((eq (char-before) ?\])
7848 (setq after-rparen nil))
98d7371e
AM
7849 (t ; either } (hit previous defun) or = or no more
7850 ; parens/brackets.
10489fcb 7851 (throw 'knr nil)))
9cf17ef1
AM
7852
7853 (if after-rparen
7854 ;; We're inside a paren. Could it be our argument list....?
7855 (if
7856 (and
7857 (progn
7858 (goto-char after-rparen)
7859 (unless (c-go-list-backward) (throw 'knr nil)) ;
7860 ;; FIXME!!! What about macros between the parens? 2007/01/20
7861 (setq before-lparen (point)))
d9e94c22 7862
9cf17ef1
AM
7863 ;; It can't be the arg list if next token is ; or {
7864 (progn (goto-char after-rparen)
7865 (c-forward-syntactic-ws)
98d7371e 7866 (not (memq (char-after) '(?\; ?\{ ?\=))))
d9e94c22 7867
9cf17ef1
AM
7868 ;; Is the thing preceding the list an identifier (the
7869 ;; function name), or a macro expansion?
7870 (progn
7871 (goto-char before-lparen)
7872 (eq (c-backward-token-2) 0)
98d7371e 7873 (or (eq (c-on-identifier) (point))
9cf17ef1
AM
7874 (and (eq (char-after) ?\))
7875 (c-go-up-list-backward)
7876 (eq (c-backward-token-2) 0)
98d7371e 7877 (eq (c-on-identifier) (point)))))
9cf17ef1
AM
7878
7879 ;; Have we got a non-empty list of comma-separated
7880 ;; identifiers?
7881 (progn
7882 (goto-char before-lparen)
7883 (c-forward-token-2) ; to first token inside parens
7884 (and
7885 (c-on-identifier)
7886 (c-forward-token-2)
7887 (catch 'id-list
7888 (while (eq (char-after) ?\,)
7889 (c-forward-token-2)
7890 (unless (c-on-identifier) (throw 'id-list nil))
7891 (c-forward-token-2))
7892 (eq (char-after) ?\))))))
7893
7894 ;; ...Yes. We've identified the function's argument list.
7895 (throw 'knr
7896 (progn (goto-char after-rparen)
7897 (c-forward-syntactic-ws)
7898 (point)))
7899
7900 ;; ...No. The current parens aren't the function's arg list.
7901 (goto-char before-lparen))
7902
7903 (or (c-go-list-backward) ; backwards over [ .... ]
7904 (throw 'knr nil)))))))))
785eecbb
RS
7905
7906(defun c-skip-conditional ()
7907 ;; skip forward over conditional at point, including any predicate
7908 ;; statements in parentheses. No error checking is performed.
0386b551
AM
7909 ;;
7910 ;; This function might do hidden buffer changes.
0ec8351b
BW
7911 (c-forward-sexp (cond
7912 ;; else if()
a66cd3ee
MS
7913 ((looking-at (concat "\\<else"
7914 "\\([ \t\n]\\|\\\\\n\\)+"
7915 "if\\>\\([^_]\\|$\\)"))
7916 3)
0ec8351b 7917 ;; do, else, try, finally
a66cd3ee
MS
7918 ((looking-at (concat "\\<\\("
7919 "do\\|else\\|try\\|finally"
7920 "\\)\\>\\([^_]\\|$\\)"))
130c507e 7921 1)
ce8c7486 7922 ;; for, if, while, switch, catch, synchronized, foreach
0ec8351b 7923 (t 2))))
785eecbb 7924
a66cd3ee
MS
7925(defun c-after-conditional (&optional lim)
7926 ;; If looking at the token after a conditional then return the
7927 ;; position of its start, otherwise return nil.
0386b551
AM
7928 ;;
7929 ;; This function might do hidden buffer changes.
a66cd3ee 7930 (save-excursion
d9e94c22 7931 (and (zerop (c-backward-token-2 1 t lim))
a66cd3ee
MS
7932 (or (looking-at c-block-stmt-1-key)
7933 (and (eq (char-after) ?\()
d9e94c22 7934 (zerop (c-backward-token-2 1 t lim))
a66cd3ee
MS
7935 (looking-at c-block-stmt-2-key)))
7936 (point))))
7937
0386b551
AM
7938(defun c-after-special-operator-id (&optional lim)
7939 ;; If the point is after an operator identifier that isn't handled
7940 ;; like an ordinary symbol (i.e. like "operator =" in C++) then the
7941 ;; position of the start of that identifier is returned. nil is
7942 ;; returned otherwise. The point may be anywhere in the syntactic
7943 ;; whitespace after the last token of the operator identifier.
7944 ;;
7945 ;; This function might do hidden buffer changes.
7946 (save-excursion
7947 (and c-overloadable-operators-regexp
7948 (zerop (c-backward-token-2 1 nil lim))
7949 (looking-at c-overloadable-operators-regexp)
51c9af45 7950 (or (not c-opt-op-identifier-prefix)
0386b551
AM
7951 (and
7952 (zerop (c-backward-token-2 1 nil lim))
51c9af45 7953 (looking-at c-opt-op-identifier-prefix)))
0386b551
AM
7954 (point))))
7955
a66cd3ee
MS
7956(defsubst c-backward-to-block-anchor (&optional lim)
7957 ;; Assuming point is at a brace that opens a statement block of some
7958 ;; kind, move to the proper anchor point for that block. It might
7959 ;; need to be adjusted further by c-add-stmt-syntax, but the
7960 ;; position at return is suitable as start position for that
7961 ;; function.
0386b551
AM
7962 ;;
7963 ;; This function might do hidden buffer changes.
a66cd3ee
MS
7964 (unless (= (point) (c-point 'boi))
7965 (let ((start (c-after-conditional lim)))
7966 (if start
7967 (goto-char start)))))
7968
037558bf 7969(defsubst c-backward-to-decl-anchor (&optional lim)
a66cd3ee
MS
7970 ;; Assuming point is at a brace that opens the block of a top level
7971 ;; declaration of some kind, move to the proper anchor point for
7972 ;; that block.
0386b551
AM
7973 ;;
7974 ;; This function might do hidden buffer changes.
a66cd3ee 7975 (unless (= (point) (c-point 'boi))
037558bf 7976 (c-beginning-of-statement-1 lim)))
a66cd3ee 7977
ff959bab 7978(defun c-search-decl-header-end ()
a66cd3ee
MS
7979 ;; Search forward for the end of the "header" of the current
7980 ;; declaration. That's the position where the definition body
7981 ;; starts, or the first variable initializer, or the ending
7982 ;; semicolon. I.e. search forward for the closest following
7983 ;; (syntactically relevant) '{', '=' or ';' token. Point is left
7984 ;; _after_ the first found token, or at point-max if none is found.
0386b551
AM
7985 ;;
7986 ;; This function might do hidden buffer changes.
ff959bab
MS
7987
7988 (let ((base (point)))
7989 (if (c-major-mode-is 'c++-mode)
7990
7991 ;; In C++ we need to take special care to handle operator
7992 ;; tokens and those pesky template brackets.
7993 (while (and
7994 (c-syntactic-re-search-forward "[;{<=]" nil 'move t t)
7995 (or
7996 (c-end-of-current-token base)
7997 ;; Handle operator identifiers, i.e. ignore any
7998 ;; operator token preceded by "operator".
7999 (save-excursion
8000 (and (c-safe (c-backward-sexp) t)
51c9af45 8001 (looking-at c-opt-op-identifier-prefix)))
ff959bab
MS
8002 (and (eq (char-before) ?<)
8003 (c-with-syntax-table c++-template-syntax-table
8004 (if (c-safe (goto-char (c-up-list-forward (point))))
8005 t
8006 (goto-char (point-max))
8007 nil)))))
8008 (setq base (point)))
8009
8010 (while (and
8011 (c-syntactic-re-search-forward "[;{=]" nil 'move t t)
8012 (c-end-of-current-token base))
8013 (setq base (point))))))
a66cd3ee
MS
8014
8015(defun c-beginning-of-decl-1 (&optional lim)
8016 ;; Go to the beginning of the current declaration, or the beginning
8017 ;; of the previous one if already at the start of it. Point won't
0386b551 8018 ;; be moved out of any surrounding paren. Return a cons cell of the
a66cd3ee
MS
8019 ;; form (MOVE . KNR-POS). MOVE is like the return value from
8020 ;; `c-beginning-of-statement-1'. If point skipped over some K&R
8021 ;; style argument declarations (and they are to be recognized) then
8022 ;; KNR-POS is set to the start of the first such argument
8023 ;; declaration, otherwise KNR-POS is nil. If LIM is non-nil, it's a
8024 ;; position that bounds the backward search.
8025 ;;
8026 ;; NB: Cases where the declaration continues after the block, as in
8027 ;; "struct foo { ... } bar;", are currently recognized as two
8028 ;; declarations, e.g. "struct foo { ... }" and "bar;" in this case.
0386b551
AM
8029 ;;
8030 ;; This function might do hidden buffer changes.
a66cd3ee
MS
8031 (catch 'return
8032 (let* ((start (point))
d9e94c22 8033 (last-stmt-start (point))
0386b551 8034 (move (c-beginning-of-statement-1 lim nil t)))
a66cd3ee 8035
a66cd3ee
MS
8036 ;; `c-beginning-of-statement-1' stops at a block start, but we
8037 ;; want to continue if the block doesn't begin a top level
2a15eb73
MS
8038 ;; construct, i.e. if it isn't preceded by ';', '}', ':', bob,
8039 ;; or an open paren.
d9e94c22 8040 (let ((beg (point)) tentative-move)
51c9af45
AM
8041 ;; Go back one "statement" each time round the loop until we're just
8042 ;; after a ;, }, or :, or at BOB or the start of a macro or start of
8043 ;; an ObjC method. This will move over a multiple declaration whose
8044 ;; components are comma separated.
d9e94c22
MS
8045 (while (and
8046 ;; Must check with c-opt-method-key in ObjC mode.
8047 (not (and c-opt-method-key
8048 (looking-at c-opt-method-key)))
8049 (/= last-stmt-start (point))
8050 (progn
8051 (c-backward-syntactic-ws lim)
8052 (not (memq (char-before) '(?\; ?} ?: nil))))
2a15eb73
MS
8053 (save-excursion
8054 (backward-char)
8055 (not (looking-at "\\s(")))
d9e94c22
MS
8056 ;; Check that we don't move from the first thing in a
8057 ;; macro to its header.
8058 (not (eq (setq tentative-move
0386b551 8059 (c-beginning-of-statement-1 lim nil t))
d9e94c22
MS
8060 'macro)))
8061 (setq last-stmt-start beg
8062 beg (point)
8063 move tentative-move))
8064 (goto-char beg))
8065
8066 (when c-recognize-knr-p
8067 (let ((fallback-pos (point)) knr-argdecl-start)
8068 ;; Handle K&R argdecls. Back up after the "statement" jumped
8069 ;; over by `c-beginning-of-statement-1', unless it was the
8070 ;; function body, in which case we're sitting on the opening
8071 ;; brace now. Then test if we're in a K&R argdecl region and
8072 ;; that we started at the other side of the first argdecl in
8073 ;; it.
8074 (unless (eq (char-after) ?{)
8075 (goto-char last-stmt-start))
8076 (if (and (setq knr-argdecl-start (c-in-knr-argdecl lim))
8077 (< knr-argdecl-start start)
8078 (progn
8079 (goto-char knr-argdecl-start)
0386b551 8080 (not (eq (c-beginning-of-statement-1 lim nil t) 'macro))))
d9e94c22
MS
8081 (throw 'return
8082 (cons (if (eq (char-after fallback-pos) ?{)
8083 'previous
8084 'same)
8085 knr-argdecl-start))
8086 (goto-char fallback-pos))))
8087
51c9af45
AM
8088 ;; `c-beginning-of-statement-1' counts each brace block as a separate
8089 ;; statement, so the result will be 'previous if we've moved over any.
8090 ;; So change our result back to 'same if necessary.
8091 ;;
8092 ;; If they were brace list initializers we might not have moved over a
8093 ;; declaration boundary though, so change it to 'same if we've moved
8094 ;; past a '=' before '{', but not ';'. (This ought to be integrated
8095 ;; into `c-beginning-of-statement-1', so we avoid this extra pass which
8096 ;; potentially can search over a large amount of text.). Take special
8097 ;; pains not to get mislead by C++'s "operator=", and the like.
d9e94c22
MS
8098 (if (and (eq move 'previous)
8099 (c-with-syntax-table (if (c-major-mode-is 'c++-mode)
8100 c++-template-syntax-table
8101 (syntax-table))
8102 (save-excursion
51c9af45
AM
8103 (and
8104 (progn
8105 (while ; keep going back to "[;={"s until we either find
8106 ; no more, or get to one which isn't an "operator ="
8107 (and (c-syntactic-re-search-forward "[;={]" start t t t)
8108 (eq (char-before) ?=)
8109 c-overloadable-operators-regexp
8110 c-opt-op-identifier-prefix
8111 (save-excursion
8112 (eq (c-backward-token-2) 0)
8113 (looking-at c-overloadable-operators-regexp)
8114 (eq (c-backward-token-2) 0)
8115 (looking-at c-opt-op-identifier-prefix))))
8116 (eq (char-before) ?=))
8117 (c-syntactic-re-search-forward "[;{]" start t t)
8118 (eq (char-before) ?{)
8119 (c-safe (goto-char (c-up-list-forward (point))) t)
8120 (not (c-syntactic-re-search-forward ";" start t t))))))
d9e94c22
MS
8121 (cons 'same nil)
8122 (cons move nil)))))
a66cd3ee
MS
8123
8124(defun c-end-of-decl-1 ()
8125 ;; Assuming point is at the start of a declaration (as detected by
8126 ;; e.g. `c-beginning-of-decl-1'), go to the end of it. Unlike
8127 ;; `c-beginning-of-decl-1', this function handles the case when a
8128 ;; block is followed by identifiers in e.g. struct declarations in C
8129 ;; or C++. If a proper end was found then t is returned, otherwise
8130 ;; point is moved as far as possible within the current sexp and nil
8131 ;; is returned. This function doesn't handle macros; use
8132 ;; `c-end-of-macro' instead in those cases.
0386b551
AM
8133 ;;
8134 ;; This function might do hidden buffer changes.
ce8c7486 8135 (let ((start (point))
a66cd3ee
MS
8136 (decl-syntax-table (if (c-major-mode-is 'c++-mode)
8137 c++-template-syntax-table
8138 (syntax-table))))
8139 (catch 'return
8140 (c-search-decl-header-end)
8141
8142 (when (and c-recognize-knr-p
8143 (eq (char-before) ?\;)
8144 (c-in-knr-argdecl start))
8145 ;; Stopped at the ';' in a K&R argdecl section which is
8146 ;; detected using the same criteria as in
8147 ;; `c-beginning-of-decl-1'. Move to the following block
8148 ;; start.
d9e94c22 8149 (c-syntactic-re-search-forward "{" nil 'move t))
a66cd3ee
MS
8150
8151 (when (eq (char-before) ?{)
8152 ;; Encountered a block in the declaration. Jump over it.
8153 (condition-case nil
8154 (goto-char (c-up-list-forward (point)))
d9e94c22
MS
8155 (error (goto-char (point-max))
8156 (throw 'return nil)))
a66cd3ee
MS
8157 (if (or (not c-opt-block-decls-with-vars-key)
8158 (save-excursion
8159 (c-with-syntax-table decl-syntax-table
8160 (let ((lim (point)))
8161 (goto-char start)
b3cf7e18
MS
8162 (not (and
8163 ;; Check for `c-opt-block-decls-with-vars-key'
8164 ;; before the first paren.
8165 (c-syntactic-re-search-forward
d9e94c22 8166 (concat "[;=\(\[{]\\|\\("
b3cf7e18
MS
8167 c-opt-block-decls-with-vars-key
8168 "\\)")
d9e94c22 8169 lim t t t)
b3cf7e18
MS
8170 (match-beginning 1)
8171 (not (eq (char-before) ?_))
d9e94c22
MS
8172 ;; Check that the first following paren is
8173 ;; the block.
8174 (c-syntactic-re-search-forward "[;=\(\[{]"
8175 lim t t t)
b3cf7e18 8176 (eq (char-before) ?{)))))))
a66cd3ee
MS
8177 ;; The declaration doesn't have any of the
8178 ;; `c-opt-block-decls-with-vars' keywords in the
8179 ;; beginning, so it ends here at the end of the block.
8180 (throw 'return t)))
8181
8182 (c-with-syntax-table decl-syntax-table
8183 (while (progn
8184 (if (eq (char-before) ?\;)
8185 (throw 'return t))
d9e94c22 8186 (c-syntactic-re-search-forward ";" nil 'move t))))
a66cd3ee 8187 nil)))
ce8c7486 8188
0386b551
AM
8189(defun c-looking-at-decl-block (containing-sexp goto-start &optional limit)
8190 ;; Assuming the point is at an open brace, check if it starts a
8191 ;; block that contains another declaration level, i.e. that isn't a
8192 ;; statement block or a brace list, and if so return non-nil.
8193 ;;
8194 ;; If the check is successful, the return value is the start of the
8195 ;; keyword that tells what kind of construct it is, i.e. typically
8196 ;; what `c-decl-block-key' matched. Also, if GOTO-START is set then
8197 ;; the point will be at the start of the construct, before any
8198 ;; leading specifiers, otherwise it's at the returned position.
8199 ;;
8200 ;; The point is clobbered if the check is unsuccessful.
8201 ;;
8202 ;; CONTAINING-SEXP is the position of the open of the surrounding
8203 ;; paren, or nil if none.
8204 ;;
8205 ;; The optional LIMIT limits the backward search for the start of
8206 ;; the construct. It's assumed to be at a syntactically relevant
8207 ;; position.
8208 ;;
8209 ;; If any template arglists are found in the searched region before
8210 ;; the open brace, they get marked with paren syntax.
8211 ;;
8212 ;; This function might do hidden buffer changes.
8213
8214 (let ((open-brace (point)) kwd-start first-specifier-pos)
8215 (c-syntactic-skip-backward c-block-prefix-charset limit t)
8216
8217 (when (and c-recognize-<>-arglists
8218 (eq (char-before) ?>))
8219 ;; Could be at the end of a template arglist.
8220 (let ((c-parse-and-markup-<>-arglists t)
8221 (c-disallow-comma-in-<>-arglists
8222 (and containing-sexp
8223 (not (eq (char-after containing-sexp) ?{)))))
8224 (while (and
8225 (c-backward-<>-arglist nil limit)
8226 (progn
8227 (c-syntactic-skip-backward c-block-prefix-charset limit t)
8228 (eq (char-before) ?>))))))
8229
8230 ;; Note: Can't get bogus hits inside template arglists below since they
8231 ;; have gotten paren syntax above.
8232 (when (and
8233 ;; If `goto-start' is set we begin by searching for the
8234 ;; first possible position of a leading specifier list.
8235 ;; The `c-decl-block-key' search continues from there since
8236 ;; we know it can't match earlier.
8237 (if goto-start
8238 (when (c-syntactic-re-search-forward c-symbol-start
8239 open-brace t t)
8240 (goto-char (setq first-specifier-pos (match-beginning 0)))
8241 t)
8242 t)
8243
8244 (cond
8245 ((c-syntactic-re-search-forward c-decl-block-key open-brace t t t)
8246 (goto-char (setq kwd-start (match-beginning 0)))
8247 (or
8248
8249 ;; Found a keyword that can't be a type?
8250 (match-beginning 1)
8251
8252 ;; Can be a type too, in which case it's the return type of a
8253 ;; function (under the assumption that no declaration level
8254 ;; block construct starts with a type).
8255 (not (c-forward-type))
8256
8257 ;; Jumped over a type, but it could be a declaration keyword
8258 ;; followed by the declared identifier that we've jumped over
8259 ;; instead (e.g. in "class Foo {"). If it indeed is a type
8260 ;; then we should be at the declarator now, so check for a
8261 ;; valid declarator start.
8262 ;;
8263 ;; Note: This doesn't cope with the case when a declared
8264 ;; identifier is followed by e.g. '(' in a language where '('
8265 ;; also might be part of a declarator expression. Currently
8266 ;; there's no such language.
8267 (not (or (looking-at c-symbol-start)
8268 (looking-at c-type-decl-prefix-key)))))
8269
8270 ;; In Pike a list of modifiers may be followed by a brace
8271 ;; to make them apply to many identifiers. Note that the
8272 ;; match data will be empty on return in this case.
8273 ((and (c-major-mode-is 'pike-mode)
8274 (progn
8275 (goto-char open-brace)
8276 (= (c-backward-token-2) 0))
8277 (looking-at c-specifier-key)
8278 ;; Use this variant to avoid yet another special regexp.
8279 (c-keyword-member (c-keyword-sym (match-string 1))
8280 'c-modifier-kwds))
8281 (setq kwd-start (point))
8282 t)))
8283
8284 ;; Got a match.
8285
8286 (if goto-start
8287 ;; Back up over any preceding specifiers and their clauses
8288 ;; by going forward from `first-specifier-pos', which is the
8289 ;; earliest possible position where the specifier list can
8290 ;; start.
8291 (progn
8292 (goto-char first-specifier-pos)
8293
8294 (while (< (point) kwd-start)
8295 (if (looking-at c-symbol-key)
8296 ;; Accept any plain symbol token on the ground that
8297 ;; it's a specifier masked through a macro (just
8298 ;; like `c-forward-decl-or-cast-1' skip forward over
8299 ;; such tokens).
8300 ;;
8301 ;; Could be more restrictive wrt invalid keywords,
8302 ;; but that'd only occur in invalid code so there's
8303 ;; no use spending effort on it.
8304 (let ((end (match-end 0)))
8305 (unless (c-forward-keyword-clause 0)
8306 (goto-char end)
8307 (c-forward-syntactic-ws)))
8308
8309 ;; Can't parse a declaration preamble and is still
8310 ;; before `kwd-start'. That means `first-specifier-pos'
8311 ;; was in some earlier construct. Search again.
8312 (if (c-syntactic-re-search-forward c-symbol-start
8313 kwd-start 'move t)
8314 (goto-char (setq first-specifier-pos (match-beginning 0)))
8315 ;; Got no preamble before the block declaration keyword.
8316 (setq first-specifier-pos kwd-start))))
8317
8318 (goto-char first-specifier-pos))
8319 (goto-char kwd-start))
8320
8321 kwd-start)))
ce8c7486 8322
a66cd3ee 8323(defun c-search-uplist-for-classkey (paren-state)
0386b551
AM
8324 ;; Check if the closest containing paren sexp is a declaration
8325 ;; block, returning a 2 element vector in that case. Aref 0
8326 ;; contains the bufpos at boi of the class key line, and aref 1
8327 ;; contains the bufpos of the open brace. This function is an
8328 ;; obsolete wrapper for `c-looking-at-decl-block'.
8329 ;;
8330 ;; This function might do hidden buffer changes.
8331 (let ((open-paren-pos (c-most-enclosing-brace paren-state)))
8332 (when open-paren-pos
8333 (save-excursion
8334 (goto-char open-paren-pos)
8335 (when (and (eq (char-after) ?{)
8336 (c-looking-at-decl-block
8337 (c-safe-position open-paren-pos paren-state)
8338 nil))
8339 (back-to-indentation)
8340 (vector (point) open-paren-pos))))))
785eecbb 8341
020716e1
AM
8342(defmacro c-pull-open-brace (ps)
8343 ;; Pull the next open brace from PS (which has the form of paren-state),
8344 ;; skipping over any brace pairs. Returns NIL when PS is exhausted.
8345 `(progn
8346 (while (consp (car ,ps))
8347 (setq ,ps (cdr ,ps)))
8348 (prog1 (car ,ps)
8349 (setq ,ps (cdr ,ps)))))
8350
8351(defun c-most-enclosing-decl-block (paren-state)
8352 ;; Return the buffer position of the most enclosing decl-block brace (in the
8353 ;; sense of c-looking-at-decl-block) in the PAREN-STATE structure, or nil if
8354 ;; none was found.
8355 (let* ((open-brace (c-pull-open-brace paren-state))
8356 (next-open-brace (c-pull-open-brace paren-state)))
8357 (while (and open-brace
8358 (save-excursion
8359 (goto-char open-brace)
8360 (not (c-looking-at-decl-block next-open-brace nil))))
8361 (setq open-brace next-open-brace
8362 next-open-brace (c-pull-open-brace paren-state)))
8363 open-brace))
8364
9657183b
AM
8365(defun c-cheap-inside-bracelist-p (paren-state)
8366 ;; Return the position of the L-brace if point is inside a brace list
8367 ;; initialization of an array, etc. This is an approximate function,
8368 ;; designed for speed over accuracy. It will not find every bracelist, but
8369 ;; a non-nil result is reliable. We simply search for "= {" (naturally with
8370 ;; syntactic whitespace allowed). PAREN-STATE is the normal thing that it
8371 ;; is everywhere else.
8372 (let (b-pos)
8373 (save-excursion
8374 (while
8375 (and (setq b-pos (c-pull-open-brace paren-state))
8376 (progn (goto-char b-pos)
8377 (c-backward-sws)
8378 (c-backward-token-2)
8379 (not (looking-at "=")))))
8380 b-pos)))
8381
a66cd3ee 8382(defun c-inside-bracelist-p (containing-sexp paren-state)
785eecbb
RS
8383 ;; return the buffer position of the beginning of the brace list
8384 ;; statement if we're inside a brace list, otherwise return nil.
8385 ;; CONTAINING-SEXP is the buffer pos of the innermost containing
0386b551 8386 ;; paren. PAREN-STATE is the remainder of the state of enclosing
130c507e 8387 ;; braces
785eecbb
RS
8388 ;;
8389 ;; N.B.: This algorithm can potentially get confused by cpp macros
0386b551 8390 ;; placed in inconvenient locations. It's a trade-off we make for
785eecbb 8391 ;; speed.
0386b551
AM
8392 ;;
8393 ;; This function might do hidden buffer changes.
785eecbb 8394 (or
d9e94c22 8395 ;; This will pick up brace list declarations.
b2acd789
RS
8396 (c-safe
8397 (save-excursion
8398 (goto-char containing-sexp)
0ec8351b 8399 (c-forward-sexp -1)
b2acd789 8400 (let (bracepos)
d9e94c22 8401 (if (and (or (looking-at c-brace-list-key)
0ec8351b 8402 (progn (c-forward-sexp -1)
d9e94c22 8403 (looking-at c-brace-list-key)))
a66cd3ee 8404 (setq bracepos (c-down-list-forward (point)))
b2acd789
RS
8405 (not (c-crosses-statement-barrier-p (point)
8406 (- bracepos 2))))
8407 (point)))))
785eecbb
RS
8408 ;; this will pick up array/aggregate init lists, even if they are nested.
8409 (save-excursion
0ec8351b
BW
8410 (let ((class-key
8411 ;; Pike can have class definitions anywhere, so we must
8412 ;; check for the class key here.
8413 (and (c-major-mode-is 'pike-mode)
a66cd3ee
MS
8414 c-decl-block-key))
8415 bufpos braceassignp lim next-containing)
785eecbb
RS
8416 (while (and (not bufpos)
8417 containing-sexp)
a66cd3ee
MS
8418 (when paren-state
8419 (if (consp (car paren-state))
8420 (setq lim (cdr (car paren-state))
8421 paren-state (cdr paren-state))
8422 (setq lim (car paren-state)))
8423 (when paren-state
8424 (setq next-containing (car paren-state)
8425 paren-state (cdr paren-state))))
785eecbb 8426 (goto-char containing-sexp)
a66cd3ee
MS
8427 (if (c-looking-at-inexpr-block next-containing next-containing)
8428 ;; We're in an in-expression block of some kind. Do not
8429 ;; check nesting. We deliberately set the limit to the
8430 ;; containing sexp, so that c-looking-at-inexpr-block
8431 ;; doesn't check for an identifier before it.
0ec8351b
BW
8432 (setq containing-sexp nil)
8433 ;; see if the open brace is preceded by = or [...] in
8434 ;; this statement, but watch out for operator=
a66cd3ee 8435 (setq braceassignp 'dontknow)
d9e94c22 8436 (c-backward-token-2 1 t lim)
6393fef2 8437 ;; Checks to do only on the first sexp before the brace.
d9e94c22 8438 (when (and c-opt-inexpr-brace-list-key
6393fef2
RS
8439 (eq (char-after) ?\[))
8440 ;; In Java, an initialization brace list may follow
8441 ;; directly after "new Foo[]", so check for a "new"
8442 ;; earlier.
8443 (while (eq braceassignp 'dontknow)
8444 (setq braceassignp
d9e94c22
MS
8445 (cond ((/= (c-backward-token-2 1 t lim) 0) nil)
8446 ((looking-at c-opt-inexpr-brace-list-key) t)
6393fef2
RS
8447 ((looking-at "\\sw\\|\\s_\\|[.[]")
8448 ;; Carry on looking if this is an
8449 ;; identifier (may contain "." in Java)
8450 ;; or another "[]" sexp.
8451 'dontknow)
8452 (t nil)))))
8453 ;; Checks to do on all sexps before the brace, up to the
8454 ;; beginning of the statement.
8455 (while (eq braceassignp 'dontknow)
0ec8351b
BW
8456 (cond ((eq (char-after) ?\;)
8457 (setq braceassignp nil))
8458 ((and class-key
8459 (looking-at class-key))
8460 (setq braceassignp nil))
8461 ((eq (char-after) ?=)
8462 ;; We've seen a =, but must check earlier tokens so
8463 ;; that it isn't something that should be ignored.
8464 (setq braceassignp 'maybe)
8465 (while (and (eq braceassignp 'maybe)
d9e94c22 8466 (zerop (c-backward-token-2 1 t lim)))
0ec8351b
BW
8467 (setq braceassignp
8468 (cond
8469 ;; Check for operator =
51c9af45
AM
8470 ((and c-opt-op-identifier-prefix
8471 (looking-at c-opt-op-identifier-prefix))
0386b551 8472 nil)
130c507e
GM
8473 ;; Check for `<opchar>= in Pike.
8474 ((and (c-major-mode-is 'pike-mode)
8475 (or (eq (char-after) ?`)
8476 ;; Special case for Pikes
8477 ;; `[]=, since '[' is not in
8478 ;; the punctuation class.
8479 (and (eq (char-after) ?\[)
8480 (eq (char-before) ?`))))
8481 nil)
0ec8351b
BW
8482 ((looking-at "\\s.") 'maybe)
8483 ;; make sure we're not in a C++ template
8484 ;; argument assignment
a66cd3ee
MS
8485 ((and
8486 (c-major-mode-is 'c++-mode)
8487 (save-excursion
8488 (let ((here (point))
8489 (pos< (progn
8490 (skip-chars-backward "^<>")
8491 (point))))
8492 (and (eq (char-before) ?<)
8493 (not (c-crosses-statement-barrier-p
8494 pos< here))
8495 (not (c-in-literal))
8496 ))))
0ec8351b 8497 nil)
6393fef2
RS
8498 (t t))))))
8499 (if (and (eq braceassignp 'dontknow)
d9e94c22 8500 (/= (c-backward-token-2 1 t lim) 0))
6393fef2
RS
8501 (setq braceassignp nil)))
8502 (if (not braceassignp)
0ec8351b
BW
8503 (if (eq (char-after) ?\;)
8504 ;; Brace lists can't contain a semicolon, so we're done.
8505 (setq containing-sexp nil)
a66cd3ee
MS
8506 ;; Go up one level.
8507 (setq containing-sexp next-containing
8508 lim nil
8509 next-containing nil))
0ec8351b
BW
8510 ;; we've hit the beginning of the aggregate list
8511 (c-beginning-of-statement-1
a66cd3ee 8512 (c-most-enclosing-brace paren-state))
0ec8351b 8513 (setq bufpos (point))))
a66cd3ee 8514 )
785eecbb
RS
8515 bufpos))
8516 ))
8517
0ec8351b
BW
8518(defun c-looking-at-special-brace-list (&optional lim)
8519