Commit | Line | Data |
---|---|---|
130c507e | 1 | ;;; cc-langs.el --- language specific settings for CC Mode |
785eecbb | 2 | |
d9e94c22 | 3 | ;; Copyright (C) 1985,1987,1992-2003 Free Software Foundation, Inc. |
785eecbb | 4 | |
d9e94c22 MS |
5 | ;; Authors: 1998- Martin Stjernholm |
6 | ;; 1992-1999 Barry A. Warsaw | |
785eecbb RS |
7 | ;; 1987 Dave Detlefs and Stewart Clamen |
8 | ;; 1985 Richard M. Stallman | |
0ec8351b | 9 | ;; Maintainer: bug-cc-mode@gnu.org |
785eecbb | 10 | ;; Created: 22-Apr-1997 (split from cc-mode.el) |
81eb2ff9 | 11 | ;; Version: See cc-mode.el |
785eecbb RS |
12 | ;; Keywords: c languages oop |
13 | ||
14 | ;; This file is part of GNU Emacs. | |
15 | ||
16 | ;; GNU Emacs is free software; you can redistribute it and/or modify | |
17 | ;; it under the terms of the GNU General Public License as published by | |
18 | ;; the Free Software Foundation; either version 2, or (at your option) | |
19 | ;; any later version. | |
20 | ||
21 | ;; GNU Emacs is distributed in the hope that it will be useful, | |
22 | ;; but WITHOUT ANY WARRANTY; without even the implied warranty of | |
23 | ;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the | |
24 | ;; GNU General Public License for more details. | |
25 | ||
26 | ;; You should have received a copy of the GNU General Public License | |
a66cd3ee | 27 | ;; along with GNU Emacs; see the file COPYING. If not, write to |
130c507e | 28 | ;; the Free Software Foundation, Inc., 59 Temple Place - Suite 330, |
785eecbb RS |
29 | ;; Boston, MA 02111-1307, USA. |
30 | ||
3afbc435 PJ |
31 | ;;; Commentary: |
32 | ||
d9e94c22 MS |
33 | ;; HACKERS NOTE: There's heavy macro magic here. If you need to make |
34 | ;; changes in this or other files containing `c-lang-defconst' but | |
35 | ;; don't want to read through the longer discussion below then read | |
36 | ;; this: | |
37 | ;; | |
38 | ;; o A change in a `c-lang-defconst' or `c-lang-defvar' will not take | |
39 | ;; effect if the file containing the mode init function (typically | |
40 | ;; cc-mode.el) is byte compiled. | |
41 | ;; o To make changes show in font locking you need to reevaluate the | |
42 | ;; `*-font-lock-keywords-*' constants, which normally is easiest to | |
43 | ;; do with M-x eval-buffer in cc-fonts.el. | |
44 | ;; o In either case it's necessary to reinitialize the mode to make | |
45 | ;; the changes show in an existing buffer. | |
46 | ||
47 | ;;; Introduction to the language dependent variable system: | |
48 | ;; | |
49 | ;; This file contains all the language dependent variables, except | |
50 | ;; those specific for font locking which reside in cc-fonts.el. As | |
51 | ;; far as possible, all the differences between the languages that CC | |
52 | ;; Mode supports are described with these variables only, so that the | |
53 | ;; code can be shared. | |
54 | ;; | |
55 | ;; The language constant system (see cc-defs.el) is used to specify | |
56 | ;; various language dependent info at a high level, such as lists of | |
57 | ;; keywords, and then from them generate - at compile time - the | |
58 | ;; various regexps and other low-level structures actually employed in | |
59 | ;; the code at runtime. | |
60 | ;; | |
61 | ;; This system is also designed to make it easy for developers of | |
62 | ;; derived modes to customize the source constants for new language | |
63 | ;; variants, without having to keep up with the exact regexps etc that | |
64 | ;; are used in each CC Mode version. It's possible from an external | |
65 | ;; package to add a new language by inheriting an existing one, and | |
66 | ;; then change specific constants as necessary for the new language. | |
67 | ;; The old values for those constants (and the values of all the other | |
68 | ;; high-level constants) may be used to build the new ones, and those | |
69 | ;; new values will in turn be used by the low-level definitions here | |
70 | ;; to build the runtime constants appropriately for the new language | |
71 | ;; in the current version of CC Mode. | |
72 | ;; | |
73 | ;; Like elsewhere in CC Mode, the existence of a doc string signifies | |
74 | ;; that a language constant is part of the external API, and that it | |
75 | ;; therefore can be used with a high confidence that it will continue | |
76 | ;; to work with future versions of CC Mode. Even so, it's not | |
77 | ;; unlikely that such constants will change meaning slightly as this | |
78 | ;; system is refined further; a certain degree of dependence on the CC | |
79 | ;; Mode version is unavoidable when hooking in at this level. Also | |
80 | ;; note that there's still work to be done to actually use these | |
81 | ;; constants everywhere inside CC Mode; there are still hardcoded | |
82 | ;; values in many places in the code. | |
83 | ;; | |
84 | ;; Separate packages will also benefit from the compile time | |
85 | ;; evaluation; the byte compiled file(s) for them will contain the | |
86 | ;; compiled runtime constants ready for use by (the byte compiled) CC | |
87 | ;; Mode, and the source definitions in this file don't have to be | |
88 | ;; loaded then. However, if a byte compiled package is loaded that | |
89 | ;; has been compiled with a different version of CC Mode than the one | |
90 | ;; currently loaded, then the compiled-in values will be discarded and | |
91 | ;; new ones will be built when the mode is initialized. That will | |
92 | ;; automatically trig a load of the file(s) containing the source | |
93 | ;; definitions (i.e. this file and/or cc-fonts.el) if necessary. | |
94 | ;; | |
95 | ;; A small example of a derived mode is available at | |
96 | ;; <http://cc-mode.sourceforge.net/derived-mode-ex.el>. It also | |
97 | ;; contains some useful hints for derived mode developers. | |
98 | ||
99 | ;;; Using language variables: | |
100 | ;; | |
101 | ;; The `c-lang-defvar' forms in this file comprise the language | |
102 | ;; variables that CC Mode uses. It does not work to use | |
103 | ;; `c-lang-defvar' anywhere else (which isn't much of a limitation | |
104 | ;; since these variables sole purpose is to interface with the CC Mode | |
105 | ;; core functions). The values in these `c-lang-defvar's are not | |
106 | ;; evaluated right away but instead collected to a single large `setq' | |
107 | ;; that can be inserted for a particular language with the | |
108 | ;; `c-init-language-vars' macro. | |
109 | ||
110 | ;; This file is only required at compile time, or when not running | |
111 | ;; from byte compiled files, or when the source definitions for the | |
112 | ;; language constants are requested. | |
113 | ||
3afbc435 PJ |
114 | ;;; Code: |
115 | ||
51f606de GM |
116 | (eval-when-compile |
117 | (let ((load-path | |
130c507e GM |
118 | (if (and (boundp 'byte-compile-dest-file) |
119 | (stringp byte-compile-dest-file)) | |
120 | (cons (file-name-directory byte-compile-dest-file) load-path) | |
51f606de | 121 | load-path))) |
d9e94c22 | 122 | (load "cc-bytecomp" nil t))) |
51f606de | 123 | |
130c507e GM |
124 | (cc-require 'cc-defs) |
125 | (cc-require 'cc-vars) | |
a6739a05 | 126 | |
785eecbb | 127 | \f |
d9e94c22 | 128 | ;;; Setup for the `c-lang-defvar' system. |
a66cd3ee MS |
129 | |
130 | (eval-and-compile | |
d9e94c22 MS |
131 | ;; These are used to collect the init forms from the subsequent |
132 | ;; `c-lang-defvar'. They are used to build the lambda in | |
133 | ;; `c-make-init-lang-vars-fun' below. | |
2eb455ab MS |
134 | (defvar c-lang-variable-inits nil) |
135 | (defvar c-lang-variable-inits-tail nil) | |
136 | (setq c-lang-variable-inits (list nil) | |
137 | c-lang-variable-inits-tail c-lang-variable-inits)) | |
d9e94c22 MS |
138 | |
139 | (defmacro c-lang-defvar (var val &optional doc) | |
140 | "Declares the buffer local variable VAR to get the value VAL at mode | |
141 | initialization, at which point VAL is evaluated. More accurately, VAL | |
142 | is evaluated and bound to VAR when the result from the macro | |
143 | `c-init-language-vars' is evaluated. | |
144 | ||
145 | `c-lang-const' is typically used in VAL to get the right value for the | |
146 | language being initialized, and such calls will be macro expanded to | |
147 | the evaluated constant value at compile time. | |
148 | ||
149 | This macro does not do any hidden buffer changes." | |
150 | ||
151 | (when (and (not doc) | |
152 | (eq (car-safe val) 'c-lang-const) | |
153 | (eq (nth 1 val) var) | |
154 | (not (nth 2 val))) | |
155 | ;; Special case: If there's no docstring and the value is a | |
156 | ;; simple (c-lang-const foo) where foo is the same name as VAR | |
157 | ;; then take the docstring from the language constant foo. | |
158 | (setq doc (get (intern (symbol-name (nth 1 val)) c-lang-constants) | |
159 | 'variable-documentation))) | |
160 | (or (stringp doc) | |
161 | (setq doc nil)) | |
162 | ||
163 | (let ((elem (assq var (cdr c-lang-variable-inits)))) | |
164 | (if elem | |
165 | (setcdr elem (list val doc)) | |
166 | (setcdr c-lang-variable-inits-tail (list (list var val doc))) | |
167 | (setq c-lang-variable-inits-tail (cdr c-lang-variable-inits-tail)))) | |
168 | ||
169 | ;; Return the symbol, like the other def* forms. | |
170 | `',var) | |
171 | ||
172 | (put 'c-lang-defvar 'lisp-indent-function 'defun) | |
173 | (eval-after-load "edebug" | |
174 | '(def-edebug-spec c-lang-defvar | |
175 | (&define name def-form &optional stringp))) | |
c55676a1 | 176 | |
51f606de | 177 | \f |
d9e94c22 MS |
178 | ;;; Various mode specific values that aren't language related. |
179 | ||
180 | (c-lang-defconst c-mode-menu | |
181 | ;; The definition for the mode menu. The menu title is prepended to | |
182 | ;; this before it's fed to `easy-menu-define'. | |
183 | t `(["Comment Out Region" comment-region | |
184 | (c-fn-region-is-active-p)] | |
185 | ["Uncomment Region" (comment-region (region-beginning) | |
186 | (region-end) '(4)) | |
187 | (c-fn-region-is-active-p)] | |
188 | ["Indent Expression" c-indent-exp | |
189 | (memq (char-after) '(?\( ?\[ ?\{))] | |
190 | ["Indent Line or Region" c-indent-line-or-region t] | |
191 | ["Fill Comment Paragraph" c-fill-paragraph t] | |
192 | "----" | |
193 | ["Backward Statement" c-beginning-of-statement t] | |
194 | ["Forward Statement" c-end-of-statement t] | |
195 | ,@(when (c-lang-const c-opt-cpp-prefix) | |
196 | ;; Only applicable if there's a cpp preprocessor. | |
197 | `(["Up Conditional" c-up-conditional t] | |
198 | ["Backward Conditional" c-backward-conditional t] | |
199 | ["Forward Conditional" c-forward-conditional t] | |
200 | "----" | |
201 | ["Macro Expand Region" c-macro-expand | |
202 | (c-fn-region-is-active-p)] | |
203 | ["Backslashify" c-backslash-region | |
204 | (c-fn-region-is-active-p)])) | |
205 | "----" | |
206 | ("Toggle..." | |
207 | ["Syntactic indentation" c-toggle-syntactic-indentation t] | |
208 | ["Auto newline" c-toggle-auto-state t] | |
209 | ["Hungry delete" c-toggle-hungry-state t]))) | |
a66cd3ee | 210 | |
d9e94c22 MS |
211 | \f |
212 | ;;; Syntax tables. | |
213 | ||
214 | (defun c-populate-syntax-table (table) | |
215 | "Populate the given syntax table as necessary for a C-like language. | |
216 | This includes setting ' and \" as string delimiters, and setting up | |
217 | the comment syntax to handle both line style \"//\" and block style | |
218 | \"/*\" \"*/\" comments." | |
219 | ||
220 | (modify-syntax-entry ?_ "_" table) | |
221 | (modify-syntax-entry ?\\ "\\" table) | |
222 | (modify-syntax-entry ?+ "." table) | |
223 | (modify-syntax-entry ?- "." table) | |
224 | (modify-syntax-entry ?= "." table) | |
225 | (modify-syntax-entry ?% "." table) | |
226 | (modify-syntax-entry ?< "." table) | |
227 | (modify-syntax-entry ?> "." table) | |
228 | (modify-syntax-entry ?& "." table) | |
229 | (modify-syntax-entry ?| "." table) | |
230 | (modify-syntax-entry ?\' "\"" table) | |
231 | (modify-syntax-entry ?\240 "." table) | |
232 | ||
233 | ;; Set up block and line oriented comments. The new C | |
234 | ;; standard mandates both comment styles even in C, so since | |
235 | ;; all languages now require dual comments, we make this the | |
236 | ;; default. | |
237 | (cond | |
238 | ;; XEmacs | |
239 | ((memq '8-bit c-emacs-features) | |
240 | (modify-syntax-entry ?/ ". 1456" table) | |
241 | (modify-syntax-entry ?* ". 23" table)) | |
242 | ;; Emacs | |
243 | ((memq '1-bit c-emacs-features) | |
244 | (modify-syntax-entry ?/ ". 124b" table) | |
245 | (modify-syntax-entry ?* ". 23" table)) | |
246 | ;; incompatible | |
247 | (t (error "CC Mode is incompatible with this version of Emacs"))) | |
248 | ||
249 | (modify-syntax-entry ?\n "> b" table) | |
250 | ;; Give CR the same syntax as newline, for selective-display | |
251 | (modify-syntax-entry ?\^m "> b" table)) | |
252 | ||
253 | (c-lang-defconst c-make-mode-syntax-table | |
254 | "Functions that generates the mode specific syntax tables. | |
255 | The syntax tables aren't stored directly since they're quite large." | |
256 | t `(lambda () | |
257 | (let ((table (make-syntax-table))) | |
258 | (c-populate-syntax-table table) | |
259 | ;; Mode specific syntaxes. | |
260 | ,(cond ((c-major-mode-is 'objc-mode) | |
261 | `(modify-syntax-entry ?@ "_" table)) | |
262 | ((c-major-mode-is 'pike-mode) | |
263 | `(modify-syntax-entry ?@ "." table))) | |
264 | table))) | |
265 | ||
266 | (c-lang-defconst c-mode-syntax-table | |
267 | ;; The syntax tables in evaluated form. Only used temporarily when | |
268 | ;; the constants in this file are evaluated. | |
269 | t (funcall (c-lang-const c-make-mode-syntax-table))) | |
270 | ||
f75ef66d | 271 | (c-lang-defconst c++-make-template-syntax-table |
d9e94c22 MS |
272 | ;; A variant of `c++-mode-syntax-table' that defines `<' and `>' as |
273 | ;; parenthesis characters. Used temporarily when template argument | |
274 | ;; lists are parsed. Note that this encourages incorrect parsing of | |
275 | ;; templates since they might contain normal operators that uses the | |
276 | ;; '<' and '>' characters. Therefore this syntax table might go | |
277 | ;; away when CC Mode handles templates correctly everywhere. | |
278 | t nil | |
279 | c++ `(lambda () | |
280 | (let ((table (funcall ,(c-lang-const c-make-mode-syntax-table)))) | |
281 | (modify-syntax-entry ?< "(>" table) | |
282 | (modify-syntax-entry ?> ")<" table) | |
283 | table))) | |
284 | (c-lang-defvar c++-template-syntax-table | |
f75ef66d MS |
285 | (and (c-lang-const c++-make-template-syntax-table) |
286 | (funcall (c-lang-const c++-make-template-syntax-table)))) | |
d9e94c22 MS |
287 | |
288 | (c-lang-defconst c-identifier-syntax-modifications | |
289 | "A list that describes the modifications that should be done to the | |
290 | mode syntax table to get a syntax table that matches all identifiers | |
291 | and keywords as words. | |
292 | ||
293 | The list is just like the one used in `font-lock-defaults': Each | |
294 | element is a cons where the car is the character to modify and the cdr | |
295 | the new syntax, as accepted by `modify-syntax-entry'." | |
296 | ;; The $ character is not allowed in most languages (one exception | |
297 | ;; is Java which allows it for legacy reasons) but we still classify | |
298 | ;; it as an indentifier character since it's often used in various | |
299 | ;; machine generated identifiers. | |
300 | t '((?_ . "w") (?$ . "w")) | |
301 | objc (append '((?@ . "w")) | |
302 | (c-lang-const c-identifier-syntax-modifications)) | |
303 | awk '((?_ . "w"))) | |
304 | (c-lang-defvar c-identifier-syntax-modifications | |
305 | (c-lang-const c-identifier-syntax-modifications)) | |
306 | ||
307 | (c-lang-defvar c-identifier-syntax-table | |
308 | (let ((table (copy-syntax-table (c-mode-var "mode-syntax-table"))) | |
309 | (mods c-identifier-syntax-modifications) | |
310 | mod) | |
311 | (while mods | |
312 | (setq mod (car mods) | |
313 | mods (cdr mods)) | |
314 | (modify-syntax-entry (car mod) (cdr mod) table)) | |
315 | table) | |
316 | "Syntax table built on the mode syntax table but additionally | |
317 | classifies symbol constituents like '_' and '$' as word constituents, | |
318 | so that all identifiers are recognized as words.") | |
319 | ||
320 | \f | |
321 | ;;; Lexer-level syntax (identifiers, tokens etc). | |
322 | ||
323 | (c-lang-defconst c-symbol-start | |
324 | "Regexp that matches the start of a symbol, i.e. any identifier or | |
325 | keyword. It's unspecified how far it matches. Does not contain a \\| | |
326 | operator at the top level." | |
327 | t (concat "[" c-alpha "_]") | |
328 | pike (concat "[" c-alpha "_`]")) | |
329 | (c-lang-defvar c-symbol-start (c-lang-const c-symbol-start)) | |
330 | ||
331 | (c-lang-defconst c-symbol-chars | |
332 | "Set of characters that can be part of a symbol. | |
333 | This is on the form that fits inside [ ] in a regexp." | |
334 | ;; Pike note: With the backquote identifiers this would include most | |
335 | ;; operator chars too, but they are handled with other means instead. | |
336 | t (concat c-alnum "_$") | |
337 | objc (concat c-alnum "_$@")) | |
a66cd3ee | 338 | |
a66cd3ee | 339 | (c-lang-defconst c-symbol-key |
d9e94c22 MS |
340 | "Regexp matching identifiers and keywords. Assumed to match if |
341 | `c-symbol-start' matches on the same position." | |
342 | t (concat (c-lang-const c-symbol-start) | |
343 | "[" (c-lang-const c-symbol-chars) "]*") | |
344 | pike (concat | |
345 | ;; Use the value from C here since the operator backquote is | |
346 | ;; covered by the other alternative. | |
347 | (c-lang-const c-symbol-key c) | |
348 | "\\|" | |
349 | (c-make-keywords-re nil | |
350 | (c-lang-const c-overloadable-operators)))) | |
351 | (c-lang-defvar c-symbol-key (c-lang-const c-symbol-key)) | |
352 | ||
353 | (c-lang-defconst c-symbol-key-depth | |
354 | ;; Number of regexp grouping parens in `c-symbol-key'. | |
355 | t (c-regexp-opt-depth (c-lang-const c-symbol-key))) | |
356 | ||
357 | (c-lang-defconst c-nonsymbol-chars | |
358 | "This is the set of chars that can't be part of a symbol, i.e. the | |
359 | negation of `c-symbol-chars'." | |
360 | t (concat "^" (c-lang-const c-symbol-chars))) | |
361 | (c-lang-defvar c-nonsymbol-chars (c-lang-const c-nonsymbol-chars)) | |
362 | ||
363 | (c-lang-defconst c-nonsymbol-key | |
364 | "Regexp that matches any character that can't be part of a symbol. | |
365 | It's usually appended to other regexps to avoid matching a prefix. | |
366 | It's assumed to not contain any submatchers." | |
367 | ;; The same thing regarding Unicode identifiers applies here as to | |
368 | ;; `c-symbol-key'. | |
369 | t (concat "[" (c-lang-const c-nonsymbol-chars) "]")) | |
370 | ||
371 | (c-lang-defconst c-opt-identifier-concat-key | |
372 | "Regexp matching the operators that join symbols to fully qualified | |
373 | identifiers, or nil in languages that don't have such things. Does | |
374 | not contain a \\| operator at the top level." | |
375 | t nil | |
376 | c++ "::" | |
377 | java "\\." | |
378 | idl "::" | |
379 | pike "\\(::\\|\\.\\)") | |
380 | (c-lang-defvar c-opt-identifier-concat-key | |
381 | (c-lang-const c-opt-identifier-concat-key) | |
382 | 'dont-doc) | |
383 | ||
384 | (c-lang-defconst c-opt-after-id-concat-key | |
385 | "Regexp that must match the token after `c-opt-identifier-concat-key' | |
386 | for it to be considered an identifier concatenation operator (which | |
387 | e.g. causes the preceding identifier to be fontified as a reference). | |
388 | Assumed to be a string if `c-opt-identifier-concat-key' is." | |
389 | t (if (c-lang-const c-opt-identifier-concat-key) | |
390 | (c-lang-const c-symbol-start)) | |
391 | c++ (concat (c-lang-const c-symbol-start) | |
392 | "\\|[~*]") | |
393 | java (concat (c-lang-const c-symbol-start) | |
394 | "\\|\\*")) | |
395 | ||
396 | (c-lang-defconst c-identifier-start | |
397 | "Regexp that matches the start of an \(optionally qualified) | |
398 | identifier. It should also match all keywords. It's unspecified how | |
399 | far it matches." | |
400 | t (concat (c-lang-const c-symbol-start) | |
401 | (if (c-lang-const c-opt-identifier-concat-key) | |
402 | (concat "\\|" (c-lang-const c-opt-identifier-concat-key)) | |
403 | "")) | |
404 | c++ (concat (c-lang-const c-identifier-start) | |
405 | "\\|" | |
406 | "[~*][ \t\n\r\f\v]*" (c-lang-const c-symbol-start)) | |
407 | ;; Java does not allow a leading qualifier operator. | |
408 | java (c-lang-const c-symbol-start)) | |
409 | (c-lang-defvar c-identifier-start (c-lang-const c-identifier-start)) | |
410 | ||
411 | (c-lang-defconst c-identifier-key | |
412 | "Regexp matching a fully qualified identifier, like \"A::B::c\" in | |
413 | C++. It does not recognize the full range of syntactic whitespace | |
414 | between the tokens; `c-forward-name' has to be used for that." | |
415 | t (c-lang-const c-symbol-key) ; Default to `c-symbol-key'. | |
416 | ;; C++ allows a leading qualifier operator and a `~' before the last | |
417 | ;; symbol. This regexp is more complex than strictly necessary to | |
418 | ;; ensure that it can be matched with a minimum of backtracking. | |
419 | c++ (concat | |
420 | "\\(" (c-lang-const c-opt-identifier-concat-key) "[ \t\n\r\f\v]*\\)?" | |
421 | (concat | |
422 | "\\(" | |
423 | ;; The submatch below is depth of `c-opt-identifier-concat-key' + 3. | |
424 | "\\(" (c-lang-const c-symbol-key) "\\)" | |
425 | (concat "\\(" | |
426 | "[ \t\n\r\f\v]*" | |
427 | (c-lang-const c-opt-identifier-concat-key) | |
428 | "[ \t\n\r\f\v]*" | |
429 | ;; The submatch below is: `c-symbol-key-depth' + | |
430 | ;; 2 * depth of `c-opt-identifier-concat-key' + 5. | |
431 | "\\(" (c-lang-const c-symbol-key) "\\)" | |
432 | "\\)*") | |
433 | (concat "\\(" | |
434 | "[ \t\n\r\f\v]*" | |
435 | (c-lang-const c-opt-identifier-concat-key) | |
436 | "[ \t\n\r\f\v]*" | |
437 | "[~*]" | |
438 | "[ \t\n\r\f\v]*" | |
439 | ;; The submatch below is: 2 * `c-symbol-key-depth' + | |
440 | ;; 3 * depth of `c-opt-identifier-concat-key' + 7. | |
441 | "\\(" (c-lang-const c-symbol-key) "\\)" | |
442 | "\\)?") | |
443 | "\\|" | |
444 | "~[ \t\n\r\f\v]*" | |
445 | ;; The submatch below is: 3 * `c-symbol-key-depth' + | |
446 | ;; 3 * depth of `c-opt-identifier-concat-key' + 8. | |
447 | "\\(" (c-lang-const c-symbol-key) "\\)" | |
448 | "\\)")) | |
449 | ;; IDL and Pike allows a leading qualifier operator. | |
450 | (idl pike) (concat | |
451 | "\\(" | |
452 | (c-lang-const c-opt-identifier-concat-key) | |
453 | "[ \t\n\r\f\v]*" | |
454 | "\\)?" | |
455 | ;; The submatch below is depth of | |
456 | ;; `c-opt-identifier-concat-key' + 2. | |
457 | "\\(" (c-lang-const c-symbol-key) "\\)" | |
458 | (concat "\\(" | |
459 | "[ \t\n\r\f\v]*" | |
460 | (c-lang-const c-opt-identifier-concat-key) | |
461 | "[ \t\n\r\f\v]*" | |
462 | ;; The submatch below is: `c-symbol-key-depth' + | |
463 | ;; 2 * depth of `c-opt-identifier-concat-key' + 4. | |
464 | "\\(" (c-lang-const c-symbol-key) "\\)" | |
465 | "\\)*")) | |
466 | ;; Java does not allow a leading qualifier operator. If it ends | |
467 | ;; with ".*" (used in import declarations) we also consider that as | |
468 | ;; part of the name. ("*" is actually recognized in any position | |
469 | ;; except the first by this regexp, but we don't bother.) | |
470 | java (concat "\\(" (c-lang-const c-symbol-key) "\\)" ; 1 | |
471 | (concat "\\(" | |
472 | "[ \t\n\r\f\v]*" | |
473 | (c-lang-const c-opt-identifier-concat-key) | |
474 | "[ \t\n\r\f\v]*" | |
475 | (concat "\\(" | |
476 | ;; The submatch below is `c-symbol-key-depth' + | |
477 | ;; depth of `c-opt-identifier-concat-key' + 4. | |
478 | "\\(" (c-lang-const c-symbol-key) "\\)" | |
479 | "\\|\\*\\)") | |
480 | "\\)*"))) | |
481 | (c-lang-defvar c-identifier-key (c-lang-const c-identifier-key)) | |
482 | ||
483 | (c-lang-defconst c-identifier-last-sym-match | |
484 | "Used to identify the submatch in `c-identifier-key' that surrounds | |
485 | the last symbol in the qualified identifier. It's a list of submatch | |
486 | numbers, of which the first that has a match is taken. It's assumed | |
487 | that at least one does when the regexp has matched." | |
488 | t '(0) | |
489 | c++ (list (+ (* 3 (c-lang-const c-symbol-key-depth)) | |
490 | (* 3 (c-regexp-opt-depth | |
491 | (c-lang-const c-opt-identifier-concat-key))) | |
492 | 8) | |
493 | (+ (* 2 (c-lang-const c-symbol-key-depth)) | |
494 | (* 3 (c-regexp-opt-depth | |
495 | (c-lang-const c-opt-identifier-concat-key))) | |
496 | 7) | |
497 | (+ (c-lang-const c-symbol-key-depth) | |
498 | (* 2 (c-regexp-opt-depth | |
499 | (c-lang-const c-opt-identifier-concat-key))) | |
500 | 5) | |
501 | (+ (c-regexp-opt-depth | |
502 | (c-lang-const c-opt-identifier-concat-key)) | |
503 | 3)) | |
504 | (idl pike) (list (+ (c-lang-const c-symbol-key-depth) | |
505 | (* 2 (c-regexp-opt-depth | |
506 | (c-lang-const c-opt-identifier-concat-key))) | |
507 | 4) | |
508 | (+ (c-regexp-opt-depth | |
509 | (c-lang-const c-opt-identifier-concat-key)) | |
510 | 2)) | |
511 | java (list (+ (c-lang-const c-symbol-key-depth) | |
512 | (c-regexp-opt-depth | |
513 | (c-lang-const c-opt-identifier-concat-key)) | |
514 | 4) | |
515 | 1)) | |
516 | (c-lang-defvar c-identifier-last-sym-match | |
517 | (c-lang-const c-identifier-last-sym-match) | |
518 | 'dont-doc) | |
519 | ||
520 | (c-lang-defconst c-opt-cpp-prefix | |
521 | "Regexp matching the prefix of a cpp directive in the languages that | |
522 | normally use that macro preprocessor. Tested at bol or at boi. | |
523 | Assumed to not contain any submatches or \\| operators." | |
524 | t "\\s *#\\s *" | |
525 | (java awk) nil) | |
526 | (c-lang-defvar c-opt-cpp-prefix (c-lang-const c-opt-cpp-prefix)) | |
527 | ||
528 | (c-lang-defconst c-opt-cpp-start | |
529 | "Regexp matching the prefix of a cpp directive including the directive | |
530 | name, or nil in languages without preprocessor support. The first | |
531 | submatch surrounds the directive name." | |
532 | t (if (c-lang-const c-opt-cpp-prefix) | |
533 | (concat (c-lang-const c-opt-cpp-prefix) | |
534 | "\\([" c-alnum "]+\\)")) | |
535 | ;; Pike, being a scripting language, recognizes hash-bangs too. | |
536 | pike (concat (c-lang-const c-opt-cpp-prefix) | |
537 | "\\([" c-alnum "]+\\|!\\)")) | |
538 | (c-lang-defvar c-opt-cpp-start (c-lang-const c-opt-cpp-start)) | |
539 | ||
540 | (c-lang-defconst c-cpp-defined-fns | |
541 | ;; Name of functions in cpp expressions that take an identifier as | |
542 | ;; the argument. | |
543 | t (if (c-lang-const c-opt-cpp-prefix) | |
544 | '("defined")) | |
545 | pike '("defined" "efun" "constant")) | |
546 | ||
547 | (c-lang-defconst c-operators | |
548 | "List describing all operators, along with their precedence and | |
549 | associativity. The order in the list corresponds to the precedence of | |
550 | the operators: The operators in each element is a group with the same | |
551 | precedence, and the group has higher precedence than the groups in all | |
552 | following elements. The car of each element describes the type of of | |
553 | the operator group, and the cdr is a list of the operator tokens in | |
554 | it. The operator group types are: | |
555 | ||
556 | 'prefix Unary prefix operators. | |
557 | 'postfix Unary postfix operators. | |
558 | 'left-assoc Binary left associative operators (i.e. a+b+c means (a+b)+c). | |
559 | 'right-assoc Binary right associative operators (i.e. a=b=c means a=(b=c)). | |
560 | 'right-assoc-sequence | |
561 | Right associative operator that constitutes of a | |
562 | sequence of tokens that separate expressions. All the | |
563 | tokens in the group are in this case taken as | |
564 | describing the sequence in one such operator, and the | |
565 | order between them is therefore significant. | |
566 | ||
567 | Operators containing a character with paren syntax are taken to match | |
568 | with a corresponding open/close paren somewhere else. A postfix | |
569 | operator with close paren syntax is taken to end a postfix expression | |
570 | started somewhere earlier, rather than start a new one at point. Vice | |
571 | versa for prefix operators with open paren syntax. | |
572 | ||
573 | Note that operators like \".\" and \"->\" which in language references | |
574 | often are described as postfix operators are considered binary here, | |
575 | since CC Mode treats every identifier as an expression." | |
576 | ||
577 | ;; There's currently no code in CC Mode that exploit all the info | |
578 | ;; in this variable; precedence, associativity etc are present as a | |
579 | ;; preparation for future work. | |
580 | ||
581 | t `(;; Preprocessor. | |
582 | ,@(when (c-lang-const c-opt-cpp-prefix) | |
583 | `((prefix "#" | |
584 | ,@(when (c-major-mode-is '(c-mode c++-mode)) | |
585 | '("%:" "??="))) | |
586 | (left-assoc "##" | |
587 | ,@(when (c-major-mode-is '(c-mode c++-mode)) | |
588 | '("%:%:" "??=??="))))) | |
589 | ||
590 | ;; Primary. Info duplicated in `c-opt-identifier-concat-key' | |
591 | ;; and `c-identifier-key'. | |
592 | ,@(cond ((c-major-mode-is 'c++-mode) | |
593 | `((postfix-if-paren "<" ">") ; Templates. | |
594 | (prefix "~" "??-" "compl") | |
595 | (right-assoc "::") | |
596 | (prefix "::"))) | |
597 | ((c-major-mode-is 'pike-mode) | |
598 | `((left-assoc "::") | |
599 | (prefix "::" "global" "predef"))) | |
600 | ((c-major-mode-is 'java-mode) | |
601 | `(;; Not necessary since it's also in the postfix group below. | |
602 | ;;(left-assoc ".") | |
603 | (prefix "super")))) | |
604 | ||
605 | ;; Postfix. | |
606 | ,@(when (c-major-mode-is 'c++-mode) | |
607 | ;; The following need special treatment. | |
608 | `((prefix "dynamic_cast" "static_cast" | |
609 | "reinterpret_cast" "const_cast" "typeid"))) | |
610 | (left-assoc "." | |
611 | ,@(unless (c-major-mode-is 'java-mode) | |
612 | '("->"))) | |
613 | (postfix "++" "--" "[" "]" "(" ")" | |
614 | ,@(when (c-major-mode-is '(c-mode c++-mode)) | |
615 | '("<:" ":>" "??(" "??)"))) | |
616 | ||
617 | ;; Unary. | |
618 | (prefix "++" "--" "+" "-" "!" "~" | |
619 | ,@(when (c-major-mode-is 'c++-mode) '("not" "compl")) | |
620 | ,@(when (c-major-mode-is '(c-mode c++-mode)) | |
621 | '("*" "&" "sizeof" "??-")) | |
622 | ,@(when (c-major-mode-is 'objc-mode) | |
623 | '("@selector" "@protocol" "@encode")) | |
624 | ;; The following need special treatment. | |
625 | ,@(cond ((c-major-mode-is 'c++-mode) | |
626 | '("new" "delete")) | |
627 | ((c-major-mode-is 'java-mode) | |
628 | '("new")) | |
629 | ((c-major-mode-is 'pike-mode) | |
630 | '("class" "lambda" "catch" "throw" "gauge"))) | |
631 | "(" ")" ; Cast. | |
632 | ,@(when (c-major-mode-is 'pike-mode) | |
633 | '("[" "]"))) ; Type cast. | |
634 | ||
635 | ;; Member selection. | |
636 | ,@(when (c-major-mode-is 'c++-mode) | |
637 | `((left-assoc ".*" "->*"))) | |
638 | ||
639 | ;; Multiplicative. | |
640 | (left-assoc "*" "/" "%") | |
641 | ||
642 | ;; Additive. | |
643 | (left-assoc "+" "-") | |
644 | ||
645 | ;; Shift. | |
646 | (left-assoc "<<" ">>" | |
647 | ,@(when (c-major-mode-is 'java-mode) | |
648 | '(">>>"))) | |
649 | ||
650 | ;; Relational. | |
651 | (left-assoc "<" ">" "<=" ">=" | |
652 | ,@(when (c-major-mode-is 'java-mode) | |
653 | '("instanceof"))) | |
654 | ||
655 | ;; Equality. | |
656 | (left-assoc "==" "!=" | |
657 | ,@(when (c-major-mode-is 'c++-mode) '("not_eq"))) | |
658 | ||
659 | ;; Bitwise and. | |
660 | (left-assoc "&" | |
661 | ,@(when (c-major-mode-is 'c++-mode) '("bitand"))) | |
662 | ||
663 | ;; Bitwise exclusive or. | |
664 | (left-assoc "^" | |
665 | ,@(when (c-major-mode-is '(c-mode c++-mode)) | |
666 | '("??'")) | |
667 | ,@(when (c-major-mode-is 'c++-mode) '("xor"))) | |
668 | ||
669 | ;; Bitwise or. | |
670 | (left-assoc "|" | |
671 | ,@(when (c-major-mode-is '(c-mode c++-mode)) | |
672 | '("??!")) | |
673 | ,@(when (c-major-mode-is 'c++-mode) '("bitor"))) | |
674 | ||
675 | ;; Logical and. | |
676 | (left-assoc "&&" | |
677 | ,@(when (c-major-mode-is 'c++-mode) '("and"))) | |
678 | ||
679 | ;; Logical or. | |
680 | (left-assoc "||" | |
681 | ,@(when (c-major-mode-is '(c-mode c++-mode)) | |
682 | '("??!??!")) | |
683 | ,@(when (c-major-mode-is 'c++-mode) '("or"))) | |
684 | ||
685 | ;; Conditional. | |
686 | (right-assoc-sequence "?" ":") | |
687 | ||
688 | ;; Assignment. | |
463f5630 KH |
689 | (right-assoc "=" "*=" "/=" "%=" "+=" "-=" ">>=" "<<=" "&=" "^=" "|=" |
690 | ,@(when (c-major-mode-is 'java-mode) | |
691 | '(">>>=")) | |
692 | ,@(when (c-major-mode-is 'c++-mode) | |
693 | '("and_eq" "or_eq" "xor_eq"))) | |
d9e94c22 MS |
694 | |
695 | ;; Exception. | |
696 | ,@(when (c-major-mode-is 'c++-mode) | |
697 | '((prefix "throw"))) | |
698 | ||
699 | ;; Sequence. | |
700 | (left-assoc ",")) | |
701 | ||
702 | ;; IDL got its own definition since it has a much smaller operator | |
703 | ;; set than the other languages. | |
704 | idl `(;; Preprocessor. | |
705 | (prefix "#") | |
706 | (left-assoc "##") | |
707 | ;; Primary. Info duplicated in `c-opt-identifier-concat-key' | |
708 | ;; and `c-identifier-key'. | |
709 | (left-assoc "::") | |
710 | (prefix "::") | |
711 | ;; Unary. | |
712 | (prefix "+" "-" "~") | |
713 | ;; Multiplicative. | |
714 | (left-assoc "*" "/" "%") | |
715 | ;; Additive. | |
716 | (left-assoc "+" "-") | |
717 | ;; Shift. | |
718 | (left-assoc "<<" ">>") | |
719 | ;; And. | |
720 | (left-assoc "&") | |
721 | ;; Xor. | |
722 | (left-assoc "^") | |
723 | ;; Or. | |
724 | (left-assoc "|"))) | |
725 | ||
726 | (c-lang-defconst c-operator-list | |
727 | ;; The operators as a flat list (without duplicates). | |
728 | t (delete-duplicates (mapcan (lambda (elem) (append (cdr elem) nil)) | |
729 | (c-lang-const c-operators)) | |
730 | :test 'string-equal)) | |
731 | ||
732 | (c-lang-defconst c-overloadable-operators | |
733 | "List of the operators that are overloadable, in their \"identifier form\"." | |
734 | t nil | |
735 | ;; The preceding "operator" keyword is treated separately in C++. | |
736 | c++ '("new" "delete" ;; Can be followed by "[]" but we ignore that. | |
737 | "+" "-" "*" "/" "%" | |
738 | "^" "??'" "xor" "&" "bitand" "|" "??!" "bitor" "~" "??-" "compl" | |
739 | "!" "=" "<" ">" "+=" "-=" "*=" "/=" "%=" "^=" | |
740 | "??'=" "xor_eq" "&=" "and_eq" "|=" "??!=" "or_eq" | |
741 | "<<" ">>" ">>=" "<<=" "==" "!=" "not_eq" "<=" ">=" | |
742 | "&&" "and" "||" "??!??!" "or" "++" "--" "," "->*" "->" | |
743 | "()" "[]" "<::>" "??(??)") | |
744 | ;; These work like identifiers in Pike. | |
745 | pike '("`+" "`-" "`&" "`|" "`^" "`<<" "`>>" "`*" "`/" "`%" "`~" | |
746 | "`==" "`<" "`>" "`!" "`[]" "`[]=" "`->" "`->=" "`()" "``+" | |
747 | "``-" "``&" "``|" "``^" "``<<" "``>>" "``*" "``/" "``%" | |
748 | "`+=")) | |
749 | ||
750 | (c-lang-defconst c-overloadable-operators-regexp | |
751 | ;; Regexp tested after an "operator" token in C++. | |
752 | t nil | |
753 | c++ (c-make-keywords-re nil (c-lang-const c-overloadable-operators))) | |
754 | (c-lang-defvar c-overloadable-operators-regexp | |
755 | (c-lang-const c-overloadable-operators-regexp)) | |
756 | ||
757 | (c-lang-defconst c-other-op-syntax-tokens | |
758 | "List of the tokens made up of characters in the punctuation or | |
759 | parenthesis syntax classes that have uses other than as expression | |
760 | operators." | |
761 | t '("{" "}" "(" ")" "[" "]" ";" ":" "," "=" "/*" "*/" "//") | |
762 | (c c++ pike) (append '("#" "##" ; Used by cpp. | |
763 | "::" "...") | |
764 | (c-lang-const c-other-op-syntax-tokens)) | |
765 | (c c++) (append '("<%" "%>" "<:" ":>" "%:" "%:%:" "*") | |
766 | (c-lang-const c-other-op-syntax-tokens)) | |
767 | c++ (append '("&") (c-lang-const c-other-op-syntax-tokens)) | |
768 | objc (append '("#" "##" ; Used by cpp. | |
769 | "+" "-") (c-lang-const c-other-op-syntax-tokens)) | |
770 | idl (append '("#" "##") ; Used by cpp. | |
771 | (c-lang-const c-other-op-syntax-tokens)) | |
772 | pike (append '("..") | |
773 | (c-lang-const c-other-op-syntax-tokens) | |
774 | (c-lang-const c-overloadable-operators)) | |
775 | awk '("{" "}" "(" ")" "[" "]" ";" "," "=" "/")) | |
776 | ||
777 | (c-lang-defconst c-nonsymbol-token-regexp | |
778 | ;; Regexp matching all tokens in the punctuation and parenthesis | |
779 | ;; syntax classes. Note that this also matches ".", which can start | |
780 | ;; a float. | |
781 | t (c-make-keywords-re nil | |
782 | (c-with-syntax-table (c-lang-const c-mode-syntax-table) | |
783 | (mapcan (lambda (op) | |
784 | (if (string-match "\\`\\(\\s.\\|\\s\(\\|\\s\)\\)+\\'" op) | |
785 | (list op))) | |
786 | (append (c-lang-const c-other-op-syntax-tokens) | |
787 | (c-lang-const c-operator-list)))))) | |
788 | (c-lang-defvar c-nonsymbol-token-regexp | |
789 | (c-lang-const c-nonsymbol-token-regexp)) | |
790 | ||
791 | (c-lang-defconst c-<-op-cont-regexp | |
792 | ;; Regexp matching the second and subsequent characters of all | |
793 | ;; multicharacter tokens that begin with "<". | |
794 | t (c-make-keywords-re nil | |
795 | (mapcan (lambda (op) | |
796 | (if (string-match "\\`<." op) | |
797 | (list (substring op 1)))) | |
798 | (append (c-lang-const c-other-op-syntax-tokens) | |
799 | (c-lang-const c-operator-list))))) | |
800 | (c-lang-defvar c-<-op-cont-regexp (c-lang-const c-<-op-cont-regexp)) | |
801 | ||
802 | (c-lang-defconst c->-op-cont-regexp | |
803 | ;; Regexp matching the second and subsequent characters of all | |
804 | ;; multicharacter tokens that begin with ">". | |
805 | t (c-make-keywords-re nil | |
806 | (mapcan (lambda (op) | |
807 | (if (string-match "\\`>." op) | |
808 | (list (substring op 1)))) | |
809 | (append (c-lang-const c-other-op-syntax-tokens) | |
810 | (c-lang-const c-operator-list))))) | |
811 | (c-lang-defvar c->-op-cont-regexp (c-lang-const c->-op-cont-regexp)) | |
812 | ||
813 | (c-lang-defconst c-stmt-delim-chars | |
814 | ;; The characters that should be considered to bound statements. To | |
815 | ;; optimize `c-crosses-statement-barrier-p' somewhat, it's assumed to | |
816 | ;; begin with "^" to negate the set. If ? : operators should be | |
817 | ;; detected then the string must end with "?:". | |
818 | t "^;{}?:" | |
819 | awk "^;{}\n\r?:") ; The newline chars gets special treatment. | |
820 | (c-lang-defvar c-stmt-delim-chars (c-lang-const c-stmt-delim-chars)) | |
821 | ||
822 | (c-lang-defconst c-stmt-delim-chars-with-comma | |
823 | ;; Variant of `c-stmt-delim-chars' that additionally contains ','. | |
824 | t "^;,{}?:" | |
825 | awk "^;,{}\n\r?:") ; The newline chars gets special treatment. | |
826 | (c-lang-defvar c-stmt-delim-chars-with-comma | |
827 | (c-lang-const c-stmt-delim-chars-with-comma)) | |
828 | ||
829 | \f | |
830 | ;;; Syntactic whitespace. | |
831 | ||
832 | (c-lang-defconst c-comment-start-regexp | |
833 | ;; Regexp to match the start of any type of comment. | |
834 | ;; | |
835 | ;; TODO: Ought to use `c-comment-prefix-regexp' with some | |
836 | ;; modifications instead of this. | |
837 | t "/[/*]" | |
838 | awk "#") | |
839 | (c-lang-defvar c-comment-start-regexp (c-lang-const c-comment-start-regexp)) | |
840 | ||
841 | (c-lang-defconst c-literal-start-regexp | |
842 | ;; Regexp to match the start of comments and string literals. | |
843 | t (concat (c-lang-const c-comment-start-regexp) | |
844 | "\\|" | |
845 | (if (memq 'gen-string-delim c-emacs-features) | |
846 | "\"|" | |
847 | "\""))) | |
848 | (c-lang-defvar c-literal-start-regexp (c-lang-const c-literal-start-regexp)) | |
849 | ||
850 | (c-lang-defconst c-doc-comment-start-regexp | |
851 | "Regexp to match the start of documentation comments." | |
852 | t "\\<\\>" | |
853 | ;; From font-lock.el: `doxygen' uses /*! while others use /**. | |
854 | (c c++ objc) "/\\*[*!]" | |
855 | java "/\\*\\*" | |
856 | pike "/[/*]!") | |
857 | (c-lang-defvar c-doc-comment-start-regexp | |
858 | (c-lang-const c-doc-comment-start-regexp)) | |
859 | ||
860 | (c-lang-defconst comment-start | |
861 | "String that starts comments inserted with M-; etc. | |
862 | `comment-start' is initialized from this." | |
863 | t "// " | |
864 | c "/* " | |
865 | awk "# ") | |
866 | (c-lang-defvar comment-start (c-lang-const comment-start) | |
867 | 'dont-doc) | |
868 | ||
869 | (c-lang-defconst comment-end | |
870 | "String that ends comments inserted with M-; etc. | |
871 | `comment-end' is initialized from this." | |
872 | t "" | |
873 | c " */") | |
874 | (c-lang-defvar comment-end (c-lang-const comment-end) | |
875 | 'dont-doc) | |
876 | ||
877 | (c-lang-defconst comment-start-skip | |
878 | "Regexp to match the start of a comment plus everything up to its body. | |
879 | `comment-start-skip' is initialized from this." | |
880 | t "/\\*+ *\\|//+ *" | |
881 | awk "#+ *") | |
882 | (c-lang-defvar comment-start-skip (c-lang-const comment-start-skip) | |
883 | 'dont-doc) | |
884 | ||
f75ef66d | 885 | (c-lang-defconst c-syntactic-ws-start |
d9e94c22 MS |
886 | "Regexp matching any sequence that can start syntactic whitespace. |
887 | The only uncertain case is '#' when there are cpp directives." | |
888 | t "[ \n\t\r\v\f#]\\|/[/*]\\|\\\\[\n\r]" | |
889 | awk "[ \n\t\r\v\f#]\\|\\\\[\n\r]") | |
f75ef66d | 890 | (c-lang-defvar c-syntactic-ws-start (c-lang-const c-syntactic-ws-start) |
d9e94c22 MS |
891 | 'dont-doc) |
892 | ||
f75ef66d | 893 | (c-lang-defconst c-syntactic-ws-end |
d9e94c22 MS |
894 | "Regexp matching any single character that might end syntactic whitespace." |
895 | t "[ \n\t\r\v\f/]" | |
896 | awk "[ \n\t\r\v\f]") | |
f75ef66d | 897 | (c-lang-defvar c-syntactic-ws-end (c-lang-const c-syntactic-ws-end) |
d9e94c22 MS |
898 | 'dont-doc) |
899 | ||
900 | (c-lang-defconst c-nonwhite-syntactic-ws | |
901 | ;; Regexp matching a piece of syntactic whitespace that isn't a | |
902 | ;; sequence of simple whitespace characters. As opposed to | |
903 | ;; `c-(forward|backward)-syntactic-ws', this doesn't regard cpp | |
904 | ;; directives as syntactic whitespace. | |
905 | t (concat "/" (concat | |
906 | "\\(" | |
907 | "/[^\n\r]*[\n\r]" ; Line comment. | |
908 | "\\|" | |
909 | ;; Block comment. We intentionally don't allow line | |
910 | ;; breaks in them to avoid going very far and risk | |
911 | ;; running out of regexp stack; this regexp is | |
912 | ;; intended to handle only short comments that | |
913 | ;; might be put in the middle of limited constructs | |
914 | ;; like declarations. | |
915 | "\\*\\([^*\n\r]\\|\\*[^/\n\r]\\)*\\*/" | |
916 | "\\)") | |
917 | "\\|" | |
918 | "\\\\[\n\r]") ; Line continuations. | |
919 | awk ("#.*[\n\r]\\|\\\\[\n\r]")) | |
920 | ||
921 | (c-lang-defconst c-syntactic-ws | |
922 | ;; Regexp matching syntactic whitespace, including possibly the | |
923 | ;; empty string. As opposed to `c-(forward|backward)-syntactic-ws', | |
924 | ;; this doesn't regard cpp directives as syntactic whitespace. Does | |
925 | ;; not contain a \| operator at the top level. | |
926 | t (concat "[ \t\n\r\f\v]*\\(" | |
927 | "\\(" (c-lang-const c-nonwhite-syntactic-ws) "\\)" | |
928 | "[ \t\n\r\f\v]*\\)*")) | |
929 | ||
930 | (c-lang-defconst c-syntactic-ws-depth | |
931 | ;; Number of regexp grouping parens in `c-syntactic-ws'. | |
932 | t (c-regexp-opt-depth (c-lang-const c-syntactic-ws))) | |
933 | ||
934 | (c-lang-defconst c-nonempty-syntactic-ws | |
935 | ;; Regexp matching syntactic whitespace, which is at least one | |
936 | ;; character long. As opposed to `c-(forward|backward)-syntactic-ws', | |
937 | ;; this doesn't regard cpp directives as syntactic whitespace. Does | |
938 | ;; not contain a \| operator at the top level. | |
939 | t (concat "\\([ \t\n\r\f\v]\\|" | |
940 | (c-lang-const c-nonwhite-syntactic-ws) | |
941 | "\\)+")) | |
942 | ||
943 | (c-lang-defconst c-nonempty-syntactic-ws-depth | |
944 | ;; Number of regexp grouping parens in `c-nonempty-syntactic-ws'. | |
945 | t (c-regexp-opt-depth (c-lang-const c-nonempty-syntactic-ws))) | |
946 | ||
947 | (c-lang-defconst c-single-line-syntactic-ws | |
948 | ;; Regexp matching syntactic whitespace without any line breaks. As | |
949 | ;; opposed to `c-(forward|backward)-syntactic-ws', this doesn't | |
950 | ;; regard cpp directives as syntactic whitespace. Does not contain | |
951 | ;; a \| operator at the top level. | |
952 | t (concat "[ \t]*\\(" | |
953 | "/\\*\\([^*\n\r]\\|\\*[^/\n\r]\\)*\\*/" ; Block comment | |
954 | "[ \t]*\\)*") | |
955 | awk ("[ \t]*\\(#.*$\\)?")) | |
956 | ||
957 | (c-lang-defconst c-single-line-syntactic-ws-depth | |
958 | ;; Number of regexp grouping parens in `c-single-line-syntactic-ws'. | |
959 | t (c-regexp-opt-depth (c-lang-const c-single-line-syntactic-ws))) | |
960 | ||
961 | (c-lang-defvar c-syntactic-eol | |
962 | ;; Regexp that matches when there is no syntactically significant | |
963 | ;; text before eol. Macros are regarded as syntactically | |
964 | ;; significant text here. | |
965 | (concat (concat | |
966 | ;; Match horizontal whitespace and block comments that | |
967 | ;; don't contain newlines. | |
968 | "\\(\\s \\|" | |
969 | (concat "/\\*" | |
970 | "\\([^*\n\r]\\|\\*[^/\n\r]\\)*" | |
971 | "\\*/") | |
972 | "\\)*") | |
973 | (concat | |
974 | ;; Match eol (possibly inside a block comment or preceded | |
975 | ;; by a line continuation backslash), or the beginning of a | |
976 | ;; line comment. Note: This has to be modified for awk | |
977 | ;; where line comments start with '#'. | |
978 | "\\(" | |
979 | (concat "\\(" | |
980 | "/\\*\\([^*\n\r]\\|\\*[^/\n\r]\\)*" | |
981 | "\\|" | |
982 | "\\\\" | |
983 | "\\)?" | |
984 | "$") | |
985 | "\\|//\\)"))) | |
986 | ||
987 | \f | |
988 | ;;; In-comment text handling. | |
989 | ||
990 | (c-lang-defconst c-paragraph-start | |
991 | "Regexp to append to `paragraph-start'." | |
992 | t "$" | |
993 | java "\\(@[a-zA-Z]+\\>\\|$\\)" ; For Javadoc. | |
994 | pike "\\(@[a-zA-Z_-]+\\>\\([^{]\\|$\\)\\|$\\)") ; For Pike refdoc. | |
995 | (c-lang-defvar c-paragraph-start (c-lang-const c-paragraph-start)) | |
996 | ||
997 | (c-lang-defconst c-paragraph-separate | |
998 | "Regexp to append to `paragraph-separate'." | |
999 | t "$" | |
1000 | pike (c-lang-const c-paragraph-start)) | |
1001 | (c-lang-defvar c-paragraph-separate (c-lang-const c-paragraph-separate)) | |
1002 | ||
1003 | \f | |
1004 | ;;; Keyword lists. | |
1005 | ||
1006 | ;; Note: All and only all language constants containing keyword lists | |
1007 | ;; should end with "-kwds"; they're automatically collected into the | |
1008 | ;; `c-kwds-lang-consts' list below and used to build `c-keywords' etc. | |
1009 | ||
a66cd3ee | 1010 | (c-lang-defconst c-primitive-type-kwds |
d9e94c22 MS |
1011 | "Primitive type keywords. As opposed to the other keyword lists, the |
1012 | keywords listed here are fontified with the type face instead of the | |
1013 | keyword face. | |
1014 | ||
1015 | If any of these also are on `c-type-list-kwds', `c-ref-list-kwds', | |
1016 | `c-colon-type-list-kwds', `c-paren-nontype-kwds', `c-paren-type-kwds', | |
1017 | `c-<>-type-kwds', or `c-<>-arglist-kwds' then the associated clauses | |
1018 | will be handled. | |
1019 | ||
1020 | Do not try to modify this list for end user customizations; the | |
1021 | `*-font-lock-extra-types' variable, where `*' is the mode prefix, is | |
1022 | the appropriate place for that." | |
1023 | t '("char" "double" "float" "int" "long" "short" "signed" | |
1024 | "unsigned" "void") | |
1025 | c (append | |
1026 | '("_Bool" "_Complex" "_Imaginary") ; Conditionally defined in C99. | |
1027 | (c-lang-const c-primitive-type-kwds)) | |
1028 | c++ (append | |
1029 | '("bool" "wchar_t") | |
1030 | (c-lang-const c-primitive-type-kwds)) | |
1031 | ;; Objective-C extends C, but probably not the new stuff in C99. | |
1032 | objc (append | |
1033 | '("id" "Class" "SEL" "IMP" "BOOL") | |
1034 | (c-lang-const c-primitive-type-kwds)) | |
a66cd3ee | 1035 | java '("boolean" "byte" "char" "double" "float" "int" "long" "short" "void") |
d9e94c22 MS |
1036 | idl '("Object" "ValueBase" "any" "boolean" "char" "double" "fixed" "float" |
1037 | "long" "octet" "sequence" "short" "string" "void" "wchar" "wstring" | |
1038 | ;; In CORBA PSDL: | |
1039 | "ref" | |
1040 | ;; The following can't really end a type, but we have to specify them | |
1041 | ;; here due to the assumption in `c-primitive-type-prefix-kwds'. It | |
1042 | ;; doesn't matter that much. | |
1043 | "unsigned" "strong") | |
1044 | pike '(;; this_program isn't really a keyword, but it's practically | |
1045 | ;; used as a builtin type. | |
1046 | "array" "float" "function" "int" "mapping" "mixed" "multiset" | |
1047 | "object" "program" "string" "this_program" "void")) | |
1048 | ||
1049 | (c-lang-defconst c-primitive-type-key | |
1050 | ;; An adorned regexp that matches `c-primitive-type-kwds'. | |
1051 | t (c-make-keywords-re t (c-lang-const c-primitive-type-kwds))) | |
1052 | (c-lang-defvar c-primitive-type-key (c-lang-const c-primitive-type-key)) | |
1053 | ||
1054 | (c-lang-defconst c-primitive-type-prefix-kwds | |
1055 | "Keywords that might act as prefixes for primitive types. Assumed to | |
1056 | be a subset of `c-primitive-type-kwds'." | |
1057 | t nil | |
1058 | (c c++) '("long" "short" "signed" "unsigned") | |
1059 | idl '("long" "unsigned" | |
1060 | ;; In CORBA PSDL: | |
1061 | "strong")) | |
1062 | ||
1063 | (c-lang-defconst c-type-prefix-kwds | |
1064 | "Keywords where the following name - if any - is a type name, and | |
1065 | where the keyword together with the symbol works as a type in | |
1066 | declarations. | |
1067 | ||
1068 | Note that an alternative if the second part doesn't hold is | |
1069 | `c-type-list-kwds'. Keywords on this list are typically also present | |
1070 | on one of the `*-decl-kwds' lists." | |
1071 | t nil | |
1072 | c '("struct" "union" "enum") | |
1073 | c++ (append '("class" "typename") | |
1074 | (c-lang-const c-type-prefix-kwds c))) | |
1075 | ||
1076 | (c-lang-defconst c-type-prefix-key | |
1077 | ;; Adorned regexp matching `c-type-prefix-kwds'. | |
1078 | t (c-make-keywords-re t (c-lang-const c-type-prefix-kwds))) | |
1079 | (c-lang-defvar c-type-prefix-key (c-lang-const c-type-prefix-key)) | |
1080 | ||
1081 | (c-lang-defconst c-type-modifier-kwds | |
1082 | "Type modifier keywords. These can occur almost anywhere in types | |
1083 | but they don't build a type of themselves. Unlike the keywords on | |
1084 | `c-primitive-type-kwds', they are fontified with the keyword face and | |
1085 | not the type face." | |
1086 | t nil | |
1087 | c '("const" "restrict" "volatile") | |
1088 | c++ '("const" "volatile" "throw") | |
1089 | objc '("const" "volatile")) | |
1090 | ||
1091 | (c-lang-defconst c-opt-type-modifier-key | |
1092 | ;; Adorned regexp matching `c-type-modifier-kwds', or nil in | |
1093 | ;; languages without such keywords. | |
1094 | t (and (c-lang-const c-type-modifier-kwds) | |
1095 | (c-make-keywords-re t (c-lang-const c-type-modifier-kwds)))) | |
1096 | (c-lang-defvar c-opt-type-modifier-key (c-lang-const c-opt-type-modifier-key)) | |
1097 | ||
1098 | (c-lang-defconst c-opt-type-component-key | |
1099 | ;; An adorned regexp that matches `c-primitive-type-prefix-kwds' and | |
1100 | ;; `c-type-modifier-kwds', or nil in languages without any of them. | |
1101 | t (and (or (c-lang-const c-primitive-type-prefix-kwds) | |
1102 | (c-lang-const c-type-modifier-kwds)) | |
1103 | (c-make-keywords-re t | |
1104 | (append (c-lang-const c-primitive-type-prefix-kwds) | |
1105 | (c-lang-const c-type-modifier-kwds))))) | |
1106 | (c-lang-defvar c-opt-type-component-key | |
1107 | (c-lang-const c-opt-type-component-key)) | |
1108 | ||
1109 | (c-lang-defconst c-class-decl-kwds | |
1110 | "Keywords introducing declarations where the following block (if any) | |
1111 | contains another declaration level that should be considered a class. | |
1112 | ||
1113 | If any of these also are on `c-type-list-kwds', `c-ref-list-kwds', | |
1114 | `c-colon-type-list-kwds', `c-paren-nontype-kwds', `c-paren-type-kwds', | |
1115 | `c-<>-type-kwds', or `c-<>-arglist-kwds' then the associated clauses | |
1116 | will be handled. | |
1117 | ||
1118 | Note that presence on this list does not automatically treat the | |
1119 | following identifier as a type; the keyword must also be present on | |
1120 | `c-type-prefix-kwds' or `c-type-list-kwds' to accomplish that." | |
1121 | t nil | |
1122 | c '("struct" "union") | |
1123 | c++ '("class" "struct" "union") | |
1124 | objc '("struct" "union" | |
1125 | "@interface" "@implementation" "@protocol") | |
a66cd3ee | 1126 | java '("class" "interface") |
d9e94c22 MS |
1127 | idl '("component" "eventtype" "exception" "home" "interface" "struct" |
1128 | "union" "valuetype" | |
1129 | ;; In CORBA PSDL: | |
1130 | "storagehome" "storagetype" | |
1131 | ;; In CORBA CIDL: | |
1132 | "catalog" "executor" "manages" "segment") | |
a66cd3ee MS |
1133 | pike '("class")) |
1134 | ||
a66cd3ee | 1135 | (c-lang-defconst c-class-key |
d9e94c22 MS |
1136 | ;; Regexp matching the start of a class. |
1137 | t (c-make-keywords-re t (c-lang-const c-class-decl-kwds))) | |
1138 | (c-lang-defvar c-class-key (c-lang-const c-class-key)) | |
1139 | ||
1140 | (c-lang-defconst c-brace-list-decl-kwds | |
1141 | "Keywords introducing declarations where the following block (if | |
1142 | any) is a brace list. | |
1143 | ||
1144 | If any of these also are on `c-type-list-kwds', `c-ref-list-kwds', | |
1145 | `c-colon-type-list-kwds', `c-paren-nontype-kwds', `c-paren-type-kwds', | |
1146 | `c-<>-type-kwds', or `c-<>-arglist-kwds' then the associated clauses | |
1147 | will be handled." | |
1148 | t '("enum") | |
1149 | (java awk) nil) | |
1150 | ||
1151 | (c-lang-defconst c-brace-list-key | |
1152 | ;; Regexp matching the start of declarations where the following | |
1153 | ;; block is a brace list. | |
1154 | t (c-make-keywords-re t (c-lang-const c-brace-list-decl-kwds))) | |
1155 | (c-lang-defvar c-brace-list-key (c-lang-const c-brace-list-key)) | |
1156 | ||
1157 | (c-lang-defconst c-other-block-decl-kwds | |
1158 | "Keywords where the following block (if any) contain another | |
1159 | declaration level that should not be considered a class. | |
1160 | ||
1161 | If any of these also are on `c-type-list-kwds', `c-ref-list-kwds', | |
1162 | `c-colon-type-list-kwds', `c-paren-nontype-kwds', `c-paren-type-kwds', | |
1163 | `c-<>-type-kwds', or `c-<>-arglist-kwds' then the associated clauses | |
1164 | will be handled." | |
1165 | t nil | |
1166 | c '("extern") | |
a66cd3ee | 1167 | c++ '("namespace" "extern") |
d9e94c22 MS |
1168 | idl '("module" |
1169 | ;; In CORBA CIDL: | |
1170 | "composition")) | |
a66cd3ee | 1171 | |
a66cd3ee | 1172 | (c-lang-defconst c-other-decl-block-key |
d9e94c22 MS |
1173 | ;; Regexp matching the start of blocks besides classes that contain |
1174 | ;; another declaration level. | |
1175 | t (c-make-keywords-re t (c-lang-const c-other-block-decl-kwds))) | |
1176 | (c-lang-defvar c-other-decl-block-key (c-lang-const c-other-decl-block-key)) | |
1177 | ||
1178 | (c-lang-defconst c-typedef-decl-kwds | |
1179 | "Keywords introducing declarations where the identifiers are defined | |
1180 | to be types. | |
1181 | ||
1182 | If any of these also are on `c-type-list-kwds', `c-ref-list-kwds', | |
1183 | `c-colon-type-list-kwds', `c-paren-nontype-kwds', `c-paren-type-kwds', | |
1184 | `c-<>-type-kwds', or `c-<>-arglist-kwds' then the associated clauses | |
1185 | will be handled." | |
1186 | t '("typedef") | |
1187 | (java awk) nil) | |
1188 | ||
1189 | (c-lang-defconst c-typeless-decl-kwds | |
1190 | "Keywords introducing declarations where the identifier (declarator) | |
1191 | list follows directly after the keyword, without any type. | |
1192 | ||
1193 | If any of these also are on `c-type-list-kwds', `c-ref-list-kwds', | |
1194 | `c-colon-type-list-kwds', `c-paren-nontype-kwds', `c-paren-type-kwds', | |
1195 | `c-<>-type-kwds', or `c-<>-arglist-kwds' then the associated clauses | |
1196 | will be handled." | |
1197 | t nil | |
1198 | ;; Unlike most other languages, exception names are not handled as | |
1199 | ;; types in IDL since they only can occur in "raises" specs. | |
1200 | idl '("exception" "factory" "finder" "native" | |
1201 | ;; In CORBA PSDL: | |
1202 | "key" "stores" | |
1203 | ;; In CORBA CIDL: | |
1204 | ;; Note that "manages" here clashes with its presence on | |
1205 | ;; `c-type-list-kwds' for IDL. | |
1206 | "executor" "facet" "manages" "segment") | |
1207 | pike '("constant")) | |
1208 | ||
1209 | (c-lang-defconst c-modifier-kwds | |
1210 | "Keywords that can prefix normal declarations of identifiers | |
1211 | \(and typically acts as flags). Things like argument declarations | |
1212 | inside function headers are also considered declarations in this | |
1213 | sense. | |
1214 | ||
1215 | If any of these also are on `c-type-list-kwds', `c-ref-list-kwds', | |
1216 | `c-colon-type-list-kwds', `c-paren-nontype-kwds', `c-paren-type-kwds', | |
1217 | `c-<>-type-kwds', or `c-<>-arglist-kwds' then the associated clauses | |
1218 | will be handled." | |
1219 | t nil | |
1220 | (c c++) '("auto" "extern" "inline" "register" "static") | |
1221 | c++ (append '("explicit" "friend" "mutable" "template" "using" "virtual") | |
1222 | (c-lang-const c-modifier-kwds)) | |
1223 | objc '("auto" "bycopy" "byref" "extern" "in" "inout" "oneway" "out" "static") | |
1224 | ;; FIXME: Some of those below ought to be on `c-other-decl-kwds' instead. | |
1225 | idl '("abstract" "attribute" "const" "consumes" "custom" "emits" "import" | |
1226 | "in" "inout" "local" "multiple" "oneway" "out" "private" "provides" | |
1227 | "public" "publishes" "readonly" "typeid" "typeprefix" "uses" | |
1228 | ;; In CORBA PSDL: | |
1229 | "primary" "state" | |
1230 | ;; In CORBA CIDL: | |
1231 | "bindsTo" "delegatesTo" "implements" "proxy" "storedOn") | |
1232 | ;; Note: "const" is not used in Java, but it's still a reserved keyword. | |
1233 | java '("abstract" "const" "final" "native" "private" "protected" "public" | |
1234 | "static" "strictfp" "synchronized" "transient" "volatile") | |
1235 | pike '("final" "inline" "local" "nomask" "optional" "private" "protected" | |
1236 | "public" "static" "variant")) | |
a66cd3ee | 1237 | |
d9e94c22 MS |
1238 | (c-lang-defconst c-other-decl-kwds |
1239 | "Keywords that can start or prefix any declaration level construct, | |
1240 | besides those on `c-class-decl-kwds', `c-brace-list-decl-kwds', | |
1241 | `c-other-block-decl-kwds', `c-typedef-decl-kwds', | |
1242 | `c-typeless-decl-kwds' and `c-modifier-kwds'. In a declaration, these | |
1243 | keywords are also recognized inside or after the identifiers that | |
1244 | makes up the type. | |
1245 | ||
1246 | If any of these also are on `c-type-list-kwds', `c-ref-list-kwds', | |
1247 | `c-colon-type-list-kwds', `c-paren-nontype-kwds', `c-paren-type-kwds', | |
1248 | `c-<>-type-kwds', or `c-<>-arglist-kwds' then the associated clauses | |
1249 | will be handled." | |
1250 | t nil | |
1251 | (c c++) '("__declspec") ; MSVC extension. | |
1252 | objc '("@class" "@end" "@defs") | |
1253 | java '("import" "package") | |
1254 | pike '("import" "inherit")) | |
1255 | ||
1256 | (c-lang-defconst c-specifier-key | |
1257 | ;; Adorned regexp matching keywords that can start a declaration but | |
1258 | ;; not a type. | |
1259 | t (c-make-keywords-re t | |
1260 | (set-difference (append (c-lang-const c-class-decl-kwds) | |
1261 | (c-lang-const c-brace-list-decl-kwds) | |
1262 | (c-lang-const c-other-block-decl-kwds) | |
1263 | (c-lang-const c-typedef-decl-kwds) | |
1264 | (c-lang-const c-typeless-decl-kwds) | |
1265 | (c-lang-const c-modifier-kwds) | |
1266 | (c-lang-const c-other-decl-kwds)) | |
1267 | (append (c-lang-const c-primitive-type-kwds) | |
1268 | (c-lang-const c-type-prefix-kwds) | |
1269 | (c-lang-const c-type-modifier-kwds)) | |
1270 | :test 'string-equal))) | |
1271 | (c-lang-defvar c-specifier-key (c-lang-const c-specifier-key)) | |
a66cd3ee | 1272 | |
d9e94c22 MS |
1273 | (c-lang-defconst c-protection-kwds |
1274 | "Protection label keywords in classes." | |
1275 | t nil | |
1276 | c++ '("private" "protected" "public") | |
1277 | objc '("@private" "@protected" "@public")) | |
a66cd3ee | 1278 | |
d9e94c22 MS |
1279 | (c-lang-defconst c-opt-access-key |
1280 | ;; Regexp matching an access protection label in a class, or nil in | |
1281 | ;; languages that don't have such things. | |
1282 | t (if (c-lang-const c-protection-kwds) | |
1283 | (c-make-keywords-re t (c-lang-const c-protection-kwds))) | |
1284 | c++ (concat "\\(" | |
1285 | (c-make-keywords-re nil (c-lang-const c-protection-kwds)) | |
1286 | "\\)[ \t\n\r\f\v]*:")) | |
1287 | (c-lang-defvar c-opt-access-key (c-lang-const c-opt-access-key)) | |
a66cd3ee | 1288 | |
d9e94c22 MS |
1289 | (c-lang-defconst c-block-decls-with-vars |
1290 | "Keywords introducing declarations that can contain a block which | |
1291 | might be followed by variable declarations, e.g. like \"foo\" in | |
1292 | \"class Foo { ... } foo;\". So if there is a block in a declaration | |
1293 | like that, it ends with the following ';' and not right away. | |
130c507e | 1294 | |
d9e94c22 MS |
1295 | The keywords on list are assumed to also be present on one of the |
1296 | `*-decl-kwds' lists." | |
1297 | t nil | |
1298 | (c objc) '("struct" "union" "enum" "typedef") | |
1299 | c++ '("class" "struct" "union" "enum" "typedef")) | |
1300 | ||
1301 | (c-lang-defconst c-opt-block-decls-with-vars-key | |
1302 | ;; Regexp matching the `c-block-decls-with-vars' keywords, or nil in | |
1303 | ;; languages without such constructs. | |
1304 | t (and (c-lang-const c-block-decls-with-vars) | |
1305 | (c-make-keywords-re t (c-lang-const c-block-decls-with-vars)))) | |
1306 | (c-lang-defvar c-opt-block-decls-with-vars-key | |
1307 | (c-lang-const c-opt-block-decls-with-vars-key)) | |
1308 | ||
1309 | (c-lang-defconst c-postfix-decl-spec-kwds | |
1310 | "Keywords introducing extra declaration specifiers in the region | |
1311 | between the header and the body \(i.e. the \"K&R-region\") in | |
1312 | declarations." | |
1313 | t nil | |
1314 | (c c++) '("__attribute__") ; GCC extension. | |
1315 | java '("extends" "implements" "throws") | |
1316 | idl '("context" "getraises" "manages" "primarykey" "raises" "setraises" | |
1317 | "supports" | |
1318 | ;; In CORBA PSDL: | |
1319 | "as" "const" "implements" "of" "ref")) | |
1320 | ||
1321 | (c-lang-defconst c-nonsymbol-sexp-kwds | |
1322 | "Keywords that may be followed by a nonsymbol sexp before whatever | |
1323 | construct it's part of continues." | |
1324 | t nil | |
1325 | (c c++ objc) '("extern")) | |
1326 | ||
1327 | (c-lang-defconst c-type-list-kwds | |
1328 | "Keywords that may be followed by a comma separated list of type | |
1329 | identifiers, where each optionally can be prefixed by keywords. (Can | |
1330 | also be used for the special case when the list can contain only one | |
1331 | element.) | |
1332 | ||
1333 | Assumed to be mutually exclusive with `c-ref-list-kwds'. There's no | |
1334 | reason to put keywords on this list if they are on `c-type-prefix-kwds'. | |
1335 | There's also no reason to add keywords that prefixes a normal | |
1336 | declaration consisting of a type followed by a declarator (list), so | |
1337 | the keywords on `c-modifier-kwds' should normally not be listed here | |
1338 | too. | |
1339 | ||
1340 | Note: Use `c-typeless-decl-kwds' for keywords followed by a function | |
1341 | or variable identifier (that's being defined)." | |
1342 | t '("struct" "union" "enum") | |
1343 | (c c++ awk) nil | |
1344 | objc (append '("@class" "@interface" "@implementation" "@protocol") | |
1345 | (c-lang-const c-type-list-kwds)) | |
1346 | java '("class" "import" "interface" "new" "extends" "implements" "throws") | |
1347 | idl (append '("component" "eventtype" "home" "interface" "manages" "native" | |
1348 | "primarykey" "supports" "valuetype" | |
1349 | ;; In CORBA PSDL: | |
1350 | "as" "implements" "of" "scope" "storagehome" "storagetype") | |
1351 | (c-lang-const c-type-list-kwds)) | |
1352 | pike '("class" "enum" "inherit")) | |
1353 | ||
1354 | (c-lang-defconst c-ref-list-kwds | |
1355 | "Keywords that may be followed by a comma separated list of | |
1356 | reference (i.e. namespace/scope/module) identifiers, where each | |
1357 | optionally can be prefixed by keywords. (Can also be used for the | |
1358 | special case when the list can contain only one element.) Assumed to | |
1359 | be mutually exclusive with `c-type-list-kwds'. | |
1360 | ||
1361 | Note: Use `c-typeless-decl-kwds' for keywords followed by a function | |
1362 | or variable identifier (that's being defined)." | |
1363 | t nil | |
1364 | c++ '("namespace") | |
1365 | java '("package") | |
1366 | idl '("import" "module" | |
1367 | ;; In CORBA CIDL: | |
1368 | "composition") | |
1369 | pike '("import")) | |
1370 | ||
1371 | (c-lang-defconst c-colon-type-list-kwds | |
1372 | "Keywords that may be followed (not necessarily directly) by a colon | |
1373 | and then a comma separated list of type identifiers, where each | |
1374 | optionally can be prefixed by keywords. (Can also be used for the | |
1375 | special case when the list can contain only one element.)" | |
1376 | t nil | |
1377 | c++ '("class" "struct") | |
1378 | idl '("component" "eventtype" "home" "interface" "valuetype" | |
1379 | ;; In CORBA PSDL: | |
1380 | "storagehome" "storagetype")) | |
1381 | ||
1382 | (c-lang-defconst c-colon-type-list-re | |
1383 | "Regexp matched after the keywords in `c-colon-type-list-kwds' to skip | |
1384 | forward to the colon. The end of the match is assumed to be directly | |
1385 | after the colon, so the regexp should end with \":\" although that | |
1386 | isn't necessary. Must be a regexp if `c-colon-type-list-kwds' isn't | |
1387 | nil." | |
1388 | t (if (c-lang-const c-colon-type-list-kwds) | |
1389 | ;; Disallow various common punctuation chars that can't come | |
1390 | ;; before the ":" that starts the inherit list after "class" | |
1391 | ;; or "struct" in C++. (Also used as default for other | |
1392 | ;; languages.) | |
1393 | "[^\]\[{}();,/#=:]*:")) | |
1394 | (c-lang-defvar c-colon-type-list-re (c-lang-const c-colon-type-list-re)) | |
1395 | ||
1396 | (c-lang-defconst c-paren-nontype-kwds | |
1397 | "Keywords that may be followed by a parenthesis expression that doesn't | |
1398 | contain type identifiers." | |
1399 | t nil | |
1400 | (c c++) '("__declspec")) ; MSVC extension. | |
1401 | ||
1402 | (c-lang-defconst c-paren-type-kwds | |
1403 | "Keywords that may be followed by a parenthesis expression containing | |
1404 | type identifiers separated by arbitrary tokens." | |
1405 | t nil | |
1406 | c++ '("throw") | |
1407 | objc '("@defs") | |
1408 | idl '("switch") | |
1409 | pike '("array" "function" "int" "mapping" "multiset" "object" "program")) | |
1410 | ||
1411 | (c-lang-defconst c-paren-any-kwds | |
1412 | t (delete-duplicates (append (c-lang-const c-paren-nontype-kwds) | |
1413 | (c-lang-const c-paren-type-kwds)) | |
1414 | :test 'string-equal)) | |
1415 | ||
1416 | (c-lang-defconst c-<>-type-kwds | |
1417 | "Keywords that may be followed by an angle bracket expression | |
1418 | containing type identifiers separated by \",\". The difference from | |
1419 | `c-<>-arglist-kwds' is that unknown names are taken to be types and | |
1420 | not other identifiers. `c-recognize-<>-arglists' is assumed to be set | |
1421 | if this isn't nil." | |
1422 | t nil | |
1423 | objc '("id") | |
1424 | idl '("sequence" | |
1425 | ;; In CORBA PSDL: | |
1426 | "ref")) | |
1427 | ||
1428 | (c-lang-defconst c-<>-arglist-kwds | |
1429 | "Keywords that can be followed by a C++ style template arglist; see | |
1430 | `c-recognize-<>-arglists' for details. That language constant is | |
1431 | assumed to be set if this isn't nil." | |
1432 | t nil | |
1433 | c++ '("template") | |
1434 | idl '("fixed" "string" "wstring")) | |
1435 | ||
1436 | (c-lang-defconst c-<>-sexp-kwds | |
1437 | ;; All keywords that can be followed by an angle bracket sexp. | |
1438 | t (delete-duplicates (append (c-lang-const c-<>-type-kwds) | |
1439 | (c-lang-const c-<>-arglist-kwds)) | |
1440 | :test 'string-equal)) | |
1441 | ||
1442 | (c-lang-defconst c-opt-<>-sexp-key | |
1443 | ;; Adorned regexp matching keywords that can be followed by an angle | |
463f5630 | 1444 | ;; bracket sexp. |
d9e94c22 MS |
1445 | t (if (c-lang-const c-recognize-<>-arglists) |
1446 | (c-make-keywords-re t (c-lang-const c-<>-sexp-kwds)))) | |
1447 | (c-lang-defvar c-opt-<>-sexp-key (c-lang-const c-opt-<>-sexp-key)) | |
1448 | ||
1449 | (c-lang-defconst c-brace-id-list-kwds | |
1450 | "Keywords that may be followed by a brace block containing a comma | |
1451 | separated list of identifier definitions, i.e. like the list of | |
1452 | identifiers that follows the type in a normal declaration." | |
1453 | t (c-lang-const c-brace-list-decl-kwds)) | |
a66cd3ee | 1454 | |
a66cd3ee | 1455 | (c-lang-defconst c-block-stmt-1-kwds |
d9e94c22 MS |
1456 | "Statement keywords followed directly by a substatement." |
1457 | t '("do" "else") | |
1458 | c++ '("do" "else" "try") | |
1459 | java '("do" "else" "finally" "try") | |
1460 | idl nil) | |
a66cd3ee | 1461 | |
a66cd3ee | 1462 | (c-lang-defconst c-block-stmt-1-key |
d9e94c22 MS |
1463 | ;; Regexp matching the start of any statement followed directly by a |
1464 | ;; substatement (doesn't match a bare block, however). | |
1465 | t (c-make-keywords-re t (c-lang-const c-block-stmt-1-kwds))) | |
1466 | (c-lang-defvar c-block-stmt-1-key (c-lang-const c-block-stmt-1-key)) | |
a66cd3ee | 1467 | |
a66cd3ee | 1468 | (c-lang-defconst c-block-stmt-2-kwds |
d9e94c22 MS |
1469 | "Statement keywords followed by a paren sexp and then by a substatement." |
1470 | t '("for" "if" "switch" "while") | |
1471 | c++ '("for" "if" "switch" "while" "catch") | |
a66cd3ee | 1472 | java '("for" "if" "switch" "while" "catch" "synchronized") |
d9e94c22 MS |
1473 | idl nil |
1474 | pike '("for" "if" "switch" "while" "foreach") | |
1475 | awk '("for" "if" "while")) | |
a66cd3ee | 1476 | |
a66cd3ee | 1477 | (c-lang-defconst c-block-stmt-2-key |
d9e94c22 MS |
1478 | ;; Regexp matching the start of any statement followed by a paren sexp |
1479 | ;; and then by a substatement. | |
1480 | t (c-make-keywords-re t (c-lang-const c-block-stmt-2-kwds))) | |
1481 | (c-lang-defvar c-block-stmt-2-key (c-lang-const c-block-stmt-2-key)) | |
a66cd3ee | 1482 | |
a66cd3ee | 1483 | (c-lang-defconst c-opt-block-stmt-key |
d9e94c22 MS |
1484 | ;; Regexp matching the start of any statement that has a |
1485 | ;; substatement (except a bare block). Nil in languages that | |
1486 | ;; don't have such constructs. | |
1487 | t (if (or (c-lang-const c-block-stmt-1-kwds) | |
1488 | (c-lang-const c-block-stmt-2-kwds)) | |
1489 | (c-make-keywords-re t | |
1490 | (append (c-lang-const c-block-stmt-1-kwds) | |
1491 | (c-lang-const c-block-stmt-2-kwds))))) | |
1492 | (c-lang-defvar c-opt-block-stmt-key (c-lang-const c-opt-block-stmt-key)) | |
1493 | ||
a66cd3ee | 1494 | (c-lang-defconst c-simple-stmt-kwds |
d9e94c22 MS |
1495 | "Statement keywords followed by an expression or nothing." |
1496 | t '("break" "continue" "goto" "return") | |
a66cd3ee MS |
1497 | ;; Note: `goto' is not valid in Java, but the keyword is still reserved. |
1498 | java '("break" "continue" "goto" "return" "throw") | |
d9e94c22 MS |
1499 | idl nil |
1500 | pike '("break" "continue" "return") | |
1501 | awk '(;; Not sure about "delete", "exit", "getline", etc. ; ACM 2002/5/30 | |
1502 | "break" "continue" "return" "delete" "exit" "getline" "next" | |
1503 | "nextfile" "print" "printf")) | |
1504 | ||
1505 | (c-lang-defconst c-simple-stmt-key | |
1506 | ;; Adorned regexp matching `c-simple-stmt-kwds'. | |
1507 | t (c-make-keywords-re t (c-lang-const c-simple-stmt-kwds))) | |
1508 | (c-lang-defvar c-simple-stmt-key (c-lang-const c-simple-stmt-key)) | |
1509 | ||
1510 | (c-lang-defconst c-paren-stmt-kwds | |
1511 | "Statement keywords followed by a parenthesis expression that | |
1512 | nevertheless contains a list separated with ';' and not ','." | |
1513 | t '("for") | |
1514 | idl nil) | |
1515 | ||
1516 | (c-lang-defconst c-paren-stmt-key | |
1517 | ;; Adorned regexp matching `c-paren-stmt-kwds'. | |
1518 | t (c-make-keywords-re t (c-lang-const c-paren-stmt-kwds))) | |
1519 | (c-lang-defvar c-paren-stmt-key (c-lang-const c-paren-stmt-key)) | |
a66cd3ee | 1520 | |
a66cd3ee | 1521 | (c-lang-defconst c-asm-stmt-kwds |
d9e94c22 MS |
1522 | "Statement keywords followed by an assembler expression." |
1523 | t nil | |
1524 | (c c++) '("asm" "__asm__")) ;; Not standard, but common. | |
a66cd3ee | 1525 | |
a66cd3ee | 1526 | (c-lang-defconst c-opt-asm-stmt-key |
d9e94c22 MS |
1527 | ;; Regexp matching the start of an assembler statement. Nil in |
1528 | ;; languages that don't support that. | |
1529 | t (if (c-lang-const c-asm-stmt-kwds) | |
1530 | (c-make-keywords-re t (c-lang-const c-asm-stmt-kwds)))) | |
1531 | (c-lang-defvar c-opt-asm-stmt-key (c-lang-const c-opt-asm-stmt-key)) | |
1532 | ||
1533 | (c-lang-defconst c-label-kwds | |
1534 | "Keywords introducing labels in blocks." | |
1535 | t '("case" "default") | |
1536 | awk nil) | |
1537 | ||
1538 | (c-lang-defconst c-before-label-kwds | |
1539 | "Keywords that might be followed by a label identifier." | |
1540 | t '("goto") | |
1541 | (java pike) (append '("break" "continue") | |
1542 | (c-lang-const c-before-label-kwds)) | |
1543 | idl nil | |
1544 | awk nil) | |
130c507e | 1545 | |
a66cd3ee | 1546 | (c-lang-defconst c-label-kwds-regexp |
d9e94c22 MS |
1547 | ;; Regexp matching any keyword that introduces a label. |
1548 | t (c-make-keywords-re t (c-lang-const c-label-kwds))) | |
1549 | (c-lang-defvar c-label-kwds-regexp (c-lang-const c-label-kwds-regexp)) | |
1550 | ||
1551 | (c-lang-defconst c-constant-kwds | |
1552 | "Keywords for constants." | |
1553 | t nil | |
1554 | (c c++) '("NULL" ;; Not a keyword, but practically works as one. | |
1555 | "false" "true") ; Defined in C99. | |
1556 | objc '("nil" "Nil") | |
1557 | idl '("TRUE" "FALSE") | |
1558 | pike '("UNDEFINED")) ;; Not a keyword, but practically works as one. | |
1559 | ||
1560 | (c-lang-defconst c-primary-expr-kwds | |
1561 | "Keywords besides constants and operators that start primary expressions." | |
1562 | t nil | |
1563 | c++ '("operator" "this") | |
1564 | objc '("super" "self") | |
1565 | java '("this") | |
1566 | pike '("this")) ;; Not really a keyword, but practically works as one. | |
130c507e | 1567 | |
a66cd3ee | 1568 | (c-lang-defconst c-expr-kwds |
d9e94c22 MS |
1569 | ;; Keywords that can occur anywhere in expressions. Built from |
1570 | ;; `c-primary-expr-kwds' and all keyword operators in `c-operators'. | |
1571 | t (delete-duplicates | |
1572 | (append (c-lang-const c-primary-expr-kwds) | |
1573 | (c-with-syntax-table (c-lang-const c-mode-syntax-table) | |
1574 | (mapcan (lambda (op) | |
1575 | (and (string-match "\\`\\(\\w\\|\\s_\\)+\\'" op) | |
1576 | (list op))) | |
1577 | (c-lang-const c-operator-list)))) | |
1578 | :test 'string-equal)) | |
1579 | ||
1580 | (c-lang-defconst c-lambda-kwds | |
1581 | "Keywords that start lambda constructs, i.e. function definitions in | |
1582 | expressions." | |
1583 | t nil | |
1584 | pike '("lambda")) | |
a66cd3ee | 1585 | |
a66cd3ee | 1586 | (c-lang-defconst c-opt-lambda-key |
d9e94c22 MS |
1587 | ;; Adorned regexp matching the start of lambda constructs, or nil in |
1588 | ;; languages that don't have such things. | |
1589 | t (and (c-lang-const c-lambda-kwds) | |
1590 | (c-make-keywords-re t (c-lang-const c-lambda-kwds)))) | |
1591 | (c-lang-defvar c-opt-lambda-key (c-lang-const c-opt-lambda-key)) | |
1592 | ||
1593 | (c-lang-defconst c-inexpr-block-kwds | |
1594 | "Keywords that start constructs followed by statement blocks which can | |
1595 | be used in expressions \(the gcc extension for this in C and C++ is | |
1596 | handled separately)." | |
1597 | t nil | |
1598 | pike '("catch" "gauge")) | |
a66cd3ee | 1599 | |
a66cd3ee | 1600 | (c-lang-defconst c-opt-inexpr-block-key |
d9e94c22 MS |
1601 | ;; Regexp matching the start of in-expression statements, or nil in |
1602 | ;; languages that don't have such things. | |
1603 | t nil | |
1604 | pike (c-make-keywords-re t (c-lang-const c-inexpr-block-kwds))) | |
1605 | (c-lang-defvar c-opt-inexpr-block-key (c-lang-const c-opt-inexpr-block-key)) | |
a66cd3ee | 1606 | |
a66cd3ee | 1607 | (c-lang-defconst c-inexpr-class-kwds |
d9e94c22 MS |
1608 | "Keywords that can start classes inside expressions." |
1609 | t nil | |
a66cd3ee MS |
1610 | java '("new") |
1611 | pike '("class")) | |
1612 | ||
a66cd3ee | 1613 | (c-lang-defconst c-opt-inexpr-class-key |
d9e94c22 MS |
1614 | ;; Regexp matching the start of a class in an expression, or nil in |
1615 | ;; languages that don't have such things. | |
1616 | t (and (c-lang-const c-inexpr-class-kwds) | |
1617 | (c-make-keywords-re t (c-lang-const c-inexpr-class-kwds)))) | |
1618 | (c-lang-defvar c-opt-inexpr-class-key (c-lang-const c-opt-inexpr-class-key)) | |
1619 | ||
1620 | (c-lang-defconst c-inexpr-brace-list-kwds | |
1621 | "Keywords that can start brace list blocks inside expressions. | |
1622 | Note that Java specific rules are currently applied to tell this from | |
1623 | `c-inexpr-class-kwds'." | |
1624 | t nil | |
1625 | java '("new")) | |
1626 | ||
1627 | (c-lang-defconst c-opt-inexpr-brace-list-key | |
1628 | ;; Regexp matching the start of a brace list in an expression, or | |
1629 | ;; nil in languages that don't have such things. This should not | |
1630 | ;; match brace lists recognized through `c-special-brace-lists'. | |
1631 | t (and (c-lang-const c-inexpr-brace-list-kwds) | |
1632 | (c-make-keywords-re t (c-lang-const c-inexpr-brace-list-kwds)))) | |
1633 | (c-lang-defvar c-opt-inexpr-brace-list-key | |
1634 | (c-lang-const c-opt-inexpr-brace-list-key)) | |
a66cd3ee | 1635 | |
a66cd3ee | 1636 | (c-lang-defconst c-any-class-key |
d9e94c22 MS |
1637 | ;; Regexp matching the start of any class, both at top level and in |
1638 | ;; expressions. | |
1639 | t (c-make-keywords-re t | |
1640 | (append (c-lang-const c-class-decl-kwds) | |
1641 | (c-lang-const c-inexpr-class-kwds)))) | |
1642 | (c-lang-defvar c-any-class-key (c-lang-const c-any-class-key)) | |
1643 | ||
a66cd3ee | 1644 | (c-lang-defconst c-decl-block-key |
d9e94c22 MS |
1645 | ;; Regexp matching the start of any declaration-level block that |
1646 | ;; contain another declaration level, i.e. that isn't a function | |
1647 | ;; block or brace list. | |
1648 | t (c-make-keywords-re t | |
1649 | (append (c-lang-const c-class-decl-kwds) | |
1650 | (c-lang-const c-other-block-decl-kwds) | |
1651 | (c-lang-const c-inexpr-class-kwds))) | |
1652 | ;; In Pike modifiers might be followed by a block | |
1653 | ;; to apply to several declarations. | |
1654 | pike (concat (c-lang-const c-decl-block-key) | |
1655 | "\\|" | |
1656 | "\\(" (c-make-keywords-re nil | |
1657 | (c-lang-const c-modifier-kwds)) "\\)" | |
1658 | (c-lang-const c-syntactic-ws) | |
1659 | "{")) | |
1660 | (c-lang-defvar c-decl-block-key (c-lang-const c-decl-block-key)) | |
1661 | ||
a66cd3ee | 1662 | (c-lang-defconst c-bitfield-kwds |
d9e94c22 MS |
1663 | "Keywords that can introduce bitfields." |
1664 | t nil | |
1665 | (c c++ objc) '("char" "int" "long" "signed" "unsigned")) | |
a66cd3ee | 1666 | |
a66cd3ee | 1667 | (c-lang-defconst c-opt-bitfield-key |
d9e94c22 MS |
1668 | ;; Regexp matching the start of a bitfield (not uniquely), or nil in |
1669 | ;; languages without bitfield support. | |
1670 | t nil | |
1671 | (c c++) (c-make-keywords-re t (c-lang-const c-bitfield-kwds))) | |
1672 | (c-lang-defvar c-opt-bitfield-key (c-lang-const c-opt-bitfield-key)) | |
1673 | ||
1674 | (c-lang-defconst c-other-kwds | |
1675 | "Keywords not accounted for by any other `*-kwds' language constant." | |
1676 | t nil | |
1677 | idl '("truncatable" | |
1678 | ;; In CORBA CIDL: (These are declaration keywords that never | |
1679 | ;; can start a declaration.) | |
1680 | "entity" "process" "service" "session" "storage")) | |
1681 | ||
1682 | \f | |
1683 | ;;; Constants built from keywords. | |
1684 | ||
1685 | ;; Note: No `*-kwds' language constants may be defined below this point. | |
1686 | ||
1687 | (eval-and-compile | |
1688 | (defconst c-kwds-lang-consts | |
1689 | ;; List of all the language constants that contain keyword lists. | |
1690 | (let (list) | |
1691 | (mapatoms (lambda (sym) | |
1692 | (when (and (boundp sym) | |
1693 | (string-match "-kwds\\'" (symbol-name sym))) | |
1694 | ;; Make the list of globally interned symbols | |
1695 | ;; instead of ones interned in `c-lang-constants'. | |
1696 | (setq list (cons (intern (symbol-name sym)) list)))) | |
1697 | c-lang-constants) | |
1698 | list))) | |
a66cd3ee | 1699 | |
a66cd3ee | 1700 | (c-lang-defconst c-keywords |
d9e94c22 MS |
1701 | ;; All keywords as a list. |
1702 | t (delete-duplicates | |
1703 | (c-lang-defconst-eval-immediately | |
1704 | `(append ,@(mapcar (lambda (kwds-lang-const) | |
1705 | `(c-lang-const ,kwds-lang-const)) | |
1706 | c-kwds-lang-consts) | |
1707 | nil)) | |
1708 | :test 'string-equal)) | |
1709 | ||
a66cd3ee | 1710 | (c-lang-defconst c-keywords-regexp |
d9e94c22 MS |
1711 | ;; All keywords as an adorned regexp. |
1712 | t (c-make-keywords-re t (c-lang-const c-keywords))) | |
1713 | (c-lang-defvar c-keywords-regexp (c-lang-const c-keywords-regexp)) | |
1714 | ||
1715 | (c-lang-defconst c-keyword-member-alist | |
1716 | ;; An alist with all the keywords in the cars. The cdr for each | |
1717 | ;; keyword is a list of the symbols for the `*-kwds' lists that | |
1718 | ;; contains it. | |
1719 | t (let ((kwd-list-alist | |
1720 | (c-lang-defconst-eval-immediately | |
1721 | `(list ,@(mapcar (lambda (kwds-lang-const) | |
1722 | `(cons ',kwds-lang-const | |
1723 | (c-lang-const ,kwds-lang-const))) | |
1724 | c-kwds-lang-consts)))) | |
1725 | lang-const kwd-list kwd | |
1726 | result-alist elem) | |
1727 | (while kwd-list-alist | |
1728 | (setq lang-const (caar kwd-list-alist) | |
1729 | kwd-list (cdar kwd-list-alist) | |
1730 | kwd-list-alist (cdr kwd-list-alist)) | |
1731 | (while kwd-list | |
1732 | (setq kwd (car kwd-list) | |
1733 | kwd-list (cdr kwd-list)) | |
1734 | (unless (setq elem (assoc kwd result-alist)) | |
1735 | (setq result-alist (cons (setq elem (list kwd)) result-alist))) | |
1736 | (unless (memq lang-const (cdr elem)) | |
1737 | (setcdr elem (cons lang-const (cdr elem)))))) | |
1738 | result-alist)) | |
1739 | ||
1740 | (c-lang-defvar c-keywords-obarray | |
1741 | ;; An obarray containing all keywords as symbols. The property list | |
1742 | ;; of each symbol has a non-nil entry for the specific `*-kwds' | |
1743 | ;; lists it's a member of. | |
1744 | ;; | |
1745 | ;; E.g. to see whether the string str contains a keyword on | |
1746 | ;; `c-class-decl-kwds', one can do like this: | |
1747 | ;; (get (intern-soft str c-keyword-obarray) 'c-class-decl-kwds) | |
1748 | ;; Which preferably is written using the associated functions in | |
1749 | ;; cc-engine: | |
1750 | ;; (c-keyword-member (c-keyword-sym str) 'c-class-decl-kwds) | |
1751 | ||
1752 | ;; The obarray is not stored directly as a language constant since | |
1753 | ;; the printed representation for obarrays used in .elc files isn't | |
1754 | ;; complete. | |
1755 | ||
1756 | (let* ((alist (c-lang-const c-keyword-member-alist)) | |
1757 | kwd lang-const-list | |
1758 | (obarray (make-vector (* (length alist) 2) 0))) | |
1759 | (while alist | |
1760 | (setq kwd (caar alist) | |
1761 | lang-const-list (cdar alist) | |
1762 | alist (cdr alist)) | |
1763 | (setplist (intern kwd obarray) | |
1764 | ;; Emacs has an odd bug that causes `mapcan' to fail | |
1765 | ;; with unintelligible errors. (XEmacs >= 20 works.) | |
1766 | ;;(mapcan (lambda (lang-const) | |
1767 | ;; (list lang-const t)) | |
1768 | ;; lang-const-list) | |
1769 | (apply 'nconc (mapcar (lambda (lang-const) | |
1770 | (list lang-const t)) | |
1771 | lang-const-list)))) | |
1772 | obarray)) | |
1773 | ||
1774 | (c-lang-defconst c-regular-keywords-regexp | |
1775 | ;; Adorned regexp matching all keywords that aren't types or | |
1776 | ;; constants. | |
1777 | t (c-make-keywords-re t | |
1778 | (set-difference (c-lang-const c-keywords) | |
1779 | (append (c-lang-const c-primitive-type-kwds) | |
1780 | (c-lang-const c-constant-kwds)) | |
1781 | :test 'string-equal))) | |
1782 | (c-lang-defvar c-regular-keywords-regexp | |
1783 | (c-lang-const c-regular-keywords-regexp)) | |
1784 | ||
1785 | (c-lang-defconst c-not-decl-init-keywords | |
1786 | ;; Adorned regexp matching all keywords that can't appear at the | |
1787 | ;; start of a declaration. | |
1788 | t (c-make-keywords-re t | |
1789 | (set-difference (c-lang-const c-keywords) | |
1790 | (append (c-lang-const c-primitive-type-kwds) | |
1791 | (c-lang-const c-type-prefix-kwds) | |
1792 | (c-lang-const c-type-modifier-kwds) | |
1793 | (c-lang-const c-class-decl-kwds) | |
1794 | (c-lang-const c-brace-list-decl-kwds) | |
1795 | (c-lang-const c-other-block-decl-kwds) | |
1796 | (c-lang-const c-typedef-decl-kwds) | |
1797 | (c-lang-const c-typeless-decl-kwds) | |
1798 | (c-lang-const c-modifier-kwds) | |
1799 | (c-lang-const c-other-decl-kwds)) | |
1800 | :test 'string-equal))) | |
1801 | (c-lang-defvar c-not-decl-init-keywords | |
1802 | (c-lang-const c-not-decl-init-keywords)) | |
1803 | ||
1804 | (c-lang-defconst c-primary-expr-regexp | |
1805 | ;; Regexp matching the start of any primary expression, i.e. any | |
1806 | ;; literal, symbol, prefix operator, and '('. It doesn't need to | |
1807 | ;; exclude keywords; they are excluded afterwards unless the second | |
1808 | ;; submatch matches. If the first but not the second submatch | |
1809 | ;; matches then it is an ambiguous primary expression; it could also | |
1810 | ;; be a match of e.g. an infix operator. (The case with ambiguous | |
1811 | ;; keyword operators isn't handled.) | |
1812 | ||
1813 | t (c-with-syntax-table (c-lang-const c-mode-syntax-table) | |
1814 | (let* ((prefix-ops | |
1815 | (mapcan (lambda (op) | |
1816 | ;; Filter out the special case prefix | |
1817 | ;; operators that are close parens. | |
1818 | (unless (string-match "\\s\)" op) | |
1819 | (list op))) | |
1820 | (mapcan | |
1821 | (lambda (opclass) | |
1822 | (when (eq (car opclass) 'prefix) | |
1823 | (append (cdr opclass) nil))) | |
1824 | (c-lang-const c-operators)))) | |
1825 | ||
1826 | (nonkeyword-prefix-ops | |
1827 | (mapcan (lambda (op) | |
1828 | (unless (string-match "\\`\\(\\w\\|\\s_\\)+\\'" op) | |
1829 | (list op))) | |
1830 | prefix-ops)) | |
1831 | ||
1832 | (in-or-postfix-ops | |
1833 | (mapcan (lambda (opclass) | |
1834 | (when (memq (car opclass) | |
1835 | '(postfix | |
1836 | left-assoc | |
1837 | right-assoc | |
1838 | right-assoc-sequence)) | |
1839 | (append (cdr opclass) nil))) | |
1840 | (c-lang-const c-operators))) | |
1841 | ||
1842 | (unambiguous-prefix-ops (set-difference nonkeyword-prefix-ops | |
1843 | in-or-postfix-ops | |
1844 | :test 'string-equal)) | |
1845 | (ambiguous-prefix-ops (intersection nonkeyword-prefix-ops | |
1846 | in-or-postfix-ops | |
1847 | :test 'string-equal))) | |
1848 | ||
1849 | (concat | |
1850 | "\\(" | |
1851 | ;; Take out all symbol class operators from `prefix-ops' and make the | |
1852 | ;; first submatch from them together with `c-primary-expr-kwds'. | |
1853 | (c-make-keywords-re t | |
1854 | (append (c-lang-const c-primary-expr-kwds) | |
1855 | (set-difference prefix-ops nonkeyword-prefix-ops | |
1856 | :test 'string-equal))) | |
1857 | ||
1858 | "\\|" | |
1859 | ;; Match all ambiguous operators. | |
1860 | (c-make-keywords-re nil | |
1861 | (intersection nonkeyword-prefix-ops in-or-postfix-ops | |
1862 | :test 'string-equal)) | |
1863 | "\\)" | |
1864 | ||
1865 | "\\|" | |
1866 | ;; Now match all other symbols. | |
1867 | (c-lang-const c-symbol-start) | |
1868 | ||
1869 | "\\|" | |
1870 | ;; The chars that can start integer and floating point | |
1871 | ;; constants. | |
1872 | "\\.?[0-9]" | |
1873 | ||
1874 | "\\|" | |
1875 | ;; The nonambiguous operators from `prefix-ops'. | |
1876 | (c-make-keywords-re nil | |
1877 | (set-difference nonkeyword-prefix-ops in-or-postfix-ops | |
1878 | :test 'string-equal)) | |
1879 | ||
1880 | "\\|" | |
1881 | ;; Match string and character literals. | |
1882 | "\\s\"" | |
1883 | (if (memq 'gen-string-delim c-emacs-features) | |
1884 | "\\|\\s|" | |
1885 | ""))))) | |
1886 | (c-lang-defvar c-primary-expr-regexp (c-lang-const c-primary-expr-regexp)) | |
a66cd3ee | 1887 | |
d9e94c22 MS |
1888 | \f |
1889 | ;;; Additional constants for parser-level constructs. | |
1890 | ||
1891 | (c-lang-defconst c-decl-prefix-re | |
1892 | "Regexp matching something that might precede a declaration or a cast, | |
1893 | such as the last token of a preceding statement or declaration. It | |
1894 | should not match bob, though. It can't require a match longer than | |
1895 | one token. The end of the token is taken to be at the end of the | |
1896 | first submatch. It must not include any following whitespace. It's | |
1897 | undefined whether identifier syntax (see `c-identifier-syntax-table') | |
1898 | is in effect or not." | |
1899 | ;; We match a sequence of characters to skip over things like \"};\" | |
1900 | ;; more quickly. We match ")" in C for K&R region declarations, and | |
1901 | ;; in all languages except Java for when a cpp macro definition | |
1902 | ;; begins with a declaration. | |
1903 | t "\\([\{\}\(\);,]+\\)" | |
1904 | java "\\([\{\}\(;,]+\\)" | |
1905 | ;; Match "<" in C++ to get the first argument in a template arglist. | |
1906 | ;; In that case there's an additional check in `c-find-decl-spots' | |
1907 | ;; that it got open paren syntax. | |
1908 | ;; | |
1909 | ;; Also match a single ":" for protection labels. We cheat a little | |
1910 | ;; and require a symbol immediately before to avoid false matches | |
1911 | ;; when starting directly on a single ":", which can be the start of | |
1912 | ;; the base class initializer list in a constructor. | |
1913 | c++ "\\([\{\}\(\);,<]+\\|\\(\\w\\|\\s_\\):\\)\\([^:]\\|\\'\\)" | |
1914 | ;; Additionally match the protection directives in Objective-C. | |
1915 | ;; Note that this doesn't cope with the longer directives, which we | |
1916 | ;; would have to match from start to end since they don't end with | |
1917 | ;; any easily recognized characters. | |
1918 | objc (concat "\\([\{\}\(\);,]+\\|" | |
1919 | (c-make-keywords-re nil (c-lang-const c-protection-kwds)) | |
1920 | "\\)") | |
1921 | ;; Match ":" for switch labels inside union declarations in IDL. | |
1922 | idl "\\([\{\}\(\);:,]+\\)\\([^:]\\|\\'\\)" | |
1923 | ;; Pike is like C but we also match "[" for multiple value | |
1924 | ;; assignments and type casts. | |
1925 | pike "\\([\{\}\(\)\[;,]+\\)") | |
1926 | (c-lang-defvar c-decl-prefix-re (c-lang-const c-decl-prefix-re) | |
1927 | 'dont-doc) | |
1928 | ||
1929 | (c-lang-defconst c-cast-parens | |
1930 | ;; List containing the paren characters that can open a cast, or nil in | |
1931 | ;; languages without casts. | |
1932 | t (c-with-syntax-table (c-lang-const c-mode-syntax-table) | |
1933 | (mapcan (lambda (opclass) | |
1934 | (when (eq (car opclass) 'prefix) | |
1935 | (mapcan (lambda (op) | |
1936 | (when (string-match "\\`\\s\(\\'" op) | |
1937 | (list (elt op 0)))) | |
1938 | (cdr opclass)))) | |
1939 | (c-lang-const c-operators)))) | |
1940 | (c-lang-defvar c-cast-parens (c-lang-const c-cast-parens)) | |
1941 | ||
1942 | (c-lang-defconst c-type-decl-prefix-key | |
1943 | "Regexp matching the operators that might precede the identifier in a | |
1944 | declaration, e.g. the \"*\" in \"char *argv\". This regexp should | |
1945 | match \"(\" if parentheses are valid in type declarations. The end of | |
1946 | the first submatch is taken as the end of the operator. Identifier | |
1947 | syntax is in effect when this is matched (see `c-identifier-syntax-table')." | |
1948 | t (if (c-lang-const c-type-modifier-kwds) | |
1949 | (concat (c-regexp-opt (c-lang-const c-type-modifier-kwds) t) "\\>") | |
1950 | ;; Default to a regexp that never matches. | |
1951 | "\\<\\>") | |
1952 | (c objc) (concat "\\(" | |
1953 | "[*\(]" | |
1954 | "\\|" | |
1955 | (c-lang-const c-type-decl-prefix-key) | |
1956 | "\\)" | |
1957 | "\\([^=]\\|$\\)") | |
1958 | c++ (concat "\\(" | |
1959 | "[*\(&]" | |
1960 | "\\|" | |
1961 | (concat "\\(" ; 2 | |
1962 | ;; If this matches there's special treatment in | |
1963 | ;; `c-font-lock-declarators' and | |
1964 | ;; `c-font-lock-declarations' that check for a | |
1965 | ;; complete name followed by ":: *". | |
1966 | (c-lang-const c-identifier-start) | |
1967 | "\\)") | |
1968 | "\\|" | |
1969 | (c-lang-const c-type-decl-prefix-key) | |
1970 | "\\)" | |
1971 | "\\([^=]\\|$\\)") | |
1972 | pike "\\([*\(!~]\\)\\([^=]\\|$\\)") | |
1973 | (c-lang-defvar c-type-decl-prefix-key (c-lang-const c-type-decl-prefix-key) | |
1974 | 'dont-doc) | |
1975 | ||
1976 | (c-lang-defconst c-type-decl-suffix-key | |
1977 | "Regexp matching the operators that might follow after the identifier | |
1978 | in a declaration, e.g. the \"[\" in \"char argv[]\". This regexp | |
1979 | should match \")\" if parentheses are valid in type declarations. If | |
1980 | it matches an open paren of some kind, the type declaration check | |
1981 | continues at the corresponding close paren, otherwise the end of the | |
1982 | first submatch is taken as the end of the operator. Identifier syntax | |
1983 | is in effect when this is matched (see `c-identifier-syntax-table')." | |
1984 | ;; Default to a regexp that matches `c-type-modifier-kwds' and a | |
1985 | ;; function argument list parenthesis. | |
1986 | t (if (c-lang-const c-type-modifier-kwds) | |
1987 | (concat "\\(\(\\|" | |
1988 | (c-regexp-opt (c-lang-const c-type-modifier-kwds) t) "\\>" | |
1989 | "\\)") | |
1990 | "\\(\(\\)") | |
1991 | (c c++ objc) (concat | |
1992 | "\\(" | |
1993 | "[\)\[\(]" | |
1994 | "\\|" | |
1995 | ;; "throw" in `c-type-modifier-kwds' is followed by a | |
1996 | ;; parenthesis list, but no extra measures are | |
1997 | ;; necessary to handle that. | |
1998 | (c-regexp-opt (c-lang-const c-type-modifier-kwds) t) "\\>" | |
1999 | "\\)") | |
2000 | (java idl) "\\([\[\(]\\)") | |
2001 | (c-lang-defvar c-type-decl-suffix-key (c-lang-const c-type-decl-suffix-key) | |
2002 | 'dont-doc) | |
2003 | ||
2004 | (c-lang-defconst c-after-suffixed-type-decl-key | |
2005 | "This regexp is matched after a type declaration expression where | |
2006 | `c-type-decl-suffix-key' has matched. If it matches then the | |
2007 | construct is taken as a declaration. It's typically used to match the | |
2008 | beginning of a function body or whatever might occur after the | |
2009 | function header in a function declaration or definition. It's | |
2010 | undefined whether identifier syntax (see `c-identifier-syntax-table') | |
2011 | is in effect or not. | |
2012 | ||
2013 | Note that it's used in cases like after \"foo (bar)\" so it should | |
2014 | only match when it's certain that it's a declaration, e.g \"{\" but | |
2015 | not \",\" or \";\"." | |
2016 | t "{" | |
2017 | ;; If K&R style declarations should be recognized then one could | |
2018 | ;; consider to match the start of any symbol since we want to match | |
2019 | ;; the start of the first declaration in the "K&R region". That | |
2020 | ;; could however produce false matches on code like "FOO(bar) x" | |
2021 | ;; where FOO is a cpp macro, so it's better to leave it out and rely | |
2022 | ;; on the other heuristics in that case. | |
2023 | t (if (c-lang-const c-postfix-decl-spec-kwds) | |
2024 | ;; Add on the keywords in `c-postfix-decl-spec-kwds'. | |
2025 | (concat (c-lang-const c-after-suffixed-type-decl-key) | |
2026 | "\\|" | |
2027 | (c-make-keywords-re t (c-lang-const c-postfix-decl-spec-kwds))) | |
2028 | (c-lang-const c-after-suffixed-type-decl-key)) | |
2029 | ;; Also match the colon that starts a base class initializer list in | |
2030 | ;; C++. That can be confused with a function call before the colon | |
2031 | ;; in a ? : operator, but we count on that `c-decl-prefix-re' won't | |
2032 | ;; match before such a thing (as a declaration-level construct; | |
2033 | ;; matches inside arglist contexts are already excluded). | |
2034 | c++ "[{:]") | |
2035 | (c-lang-defvar c-after-suffixed-type-decl-key | |
2036 | (c-lang-const c-after-suffixed-type-decl-key) | |
2037 | 'dont-doc) | |
2038 | ||
2039 | (c-lang-defconst c-after-suffixed-type-maybe-decl-key | |
2040 | ;; Regexp that in addition to `c-after-suffixed-type-decl-key' | |
2041 | ;; matches ";" and ",". | |
2042 | t (concat "\\(" (c-lang-const c-after-suffixed-type-decl-key) "\\)" | |
2043 | "\\|[;,]")) | |
2044 | (c-lang-defvar c-after-suffixed-type-maybe-decl-key | |
2045 | (c-lang-const c-after-suffixed-type-maybe-decl-key)) | |
2046 | ||
2047 | (c-lang-defconst c-opt-type-concat-key | |
2048 | "Regexp matching operators that concatenate types, e.g. the \"|\" in | |
2049 | \"int|string\" in Pike. The end of the first submatch is taken as the | |
2050 | end of the operator. nil in languages without such operators. It's | |
2051 | undefined whether identifier syntax (see `c-identifier-syntax-table') | |
2052 | is in effect or not." | |
2053 | t nil | |
2054 | pike "\\([|.&]\\)\\($\\|[^|.&]\\)") | |
2055 | (c-lang-defvar c-opt-type-concat-key (c-lang-const c-opt-type-concat-key) | |
2056 | 'dont-doc) | |
2057 | ||
2058 | (c-lang-defconst c-opt-type-suffix-key | |
2059 | "Regexp matching operators that might follow after a type, or nil in | |
2060 | languages that don't have such operators. The end of the first | |
2061 | submatch is taken as the end of the operator. This should not match | |
2062 | things like C++ template arglists if `c-recognize-<>-arglists' is set. | |
2063 | It's undefined whether identifier syntax (see `c-identifier-syntax-table') | |
2064 | is in effect or not." | |
2065 | t nil | |
2066 | (c c++ objc pike) "\\(\\.\\.\\.\\)" | |
2067 | java "\\(\\[[ \t\n\r\f\v]*\\]\\)") | |
2068 | (c-lang-defvar c-opt-type-suffix-key (c-lang-const c-opt-type-suffix-key)) | |
2069 | ||
2070 | (c-lang-defvar c-known-type-key | |
2071 | ;; Regexp matching the known type identifiers. This is initialized | |
2072 | ;; from the type keywords and `*-font-lock-extra-types'. The first | |
2073 | ;; submatch is the one that matches the type. Note that this regexp | |
2074 | ;; assumes that symbol constituents like '_' and '$' have word | |
2075 | ;; syntax. | |
2076 | (let ((extra-types (when (boundp (c-mode-symbol "font-lock-extra-types")) | |
2077 | (c-mode-var "font-lock-extra-types")))) | |
2078 | (concat "\\<\\(" | |
2079 | (c-make-keywords-re nil (c-lang-const c-primitive-type-kwds)) | |
2080 | (if (consp extra-types) | |
2081 | (concat "\\|" (mapconcat 'identity extra-types "\\|")) | |
2082 | "") | |
2083 | "\\)\\>"))) | |
2084 | ||
2085 | (c-lang-defconst c-special-brace-lists | |
2086 | "List of open- and close-chars that makes up a pike-style brace list, | |
2087 |