Commit | Line | Data |
---|---|---|
9889c728 | 1 | /* Declarations having to do with GNU Emacs syntax tables. |
95df8112 GM |
2 | |
3 | Copyright (C) 1985, 1993-1994, 1997-1998, 2001-2011 | |
4 | Free Software Foundation, Inc. | |
9889c728 JB |
5 | |
6 | This file is part of GNU Emacs. | |
7 | ||
b9b1cc14 | 8 | GNU Emacs is free software: you can redistribute it and/or modify |
9889c728 | 9 | it under the terms of the GNU General Public License as published by |
b9b1cc14 GM |
10 | the Free Software Foundation, either version 3 of the License, or |
11 | (at your option) any later version. | |
9889c728 JB |
12 | |
13 | GNU Emacs is distributed in the hope that it will be useful, | |
14 | but WITHOUT ANY WARRANTY; without even the implied warranty of | |
15 | MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the | |
16 | GNU General Public License for more details. | |
17 | ||
18 | You should have received a copy of the GNU General Public License | |
b9b1cc14 | 19 | along with GNU Emacs. If not, see <http://www.gnu.org/licenses/>. */ |
9889c728 JB |
20 | |
21 | ||
d1dfb56c | 22 | extern void update_syntax_table (EMACS_INT, EMACS_INT, int, Lisp_Object); |
9889c728 JB |
23 | |
24 | /* The standard syntax table is stored where it will automatically | |
25 | be used in all new buffers. */ | |
4b4deea2 | 26 | #define Vstandard_syntax_table BVAR (&buffer_defaults, syntax_table) |
9889c728 | 27 | |
e46c910e RS |
28 | /* A syntax table is a chartable whose elements are cons cells |
29 | (CODE+FLAGS . MATCHING-CHAR). MATCHING-CHAR can be nil if the char | |
30 | is not a kind of parenthesis. | |
9889c728 | 31 | |
e46c910e | 32 | The low 8 bits of CODE+FLAGS is a code, as follows: */ |
9889c728 JB |
33 | |
34 | enum syntaxcode | |
35 | { | |
36 | Swhitespace, /* for a whitespace character */ | |
37 | Spunct, /* for random punctuation characters */ | |
38 | Sword, /* for a word constituent */ | |
39 | Ssymbol, /* symbol constituent but not word constituent */ | |
40 | Sopen, /* for a beginning delimiter */ | |
41 | Sclose, /* for an ending delimiter */ | |
42 | Squote, /* for a prefix character like Lisp ' */ | |
43 | Sstring, /* for a string-grouping character like Lisp " */ | |
5eea1c5a | 44 | Smath, /* for delimiters like $ in Tex. */ |
9889c728 JB |
45 | Sescape, /* for a character that begins a C-style escape */ |
46 | Scharquote, /* for a character that quotes the following character */ | |
47 | Scomment, /* for a comment-starting character */ | |
48 | Sendcomment, /* for a comment-ending character */ | |
c8cdcb16 | 49 | Sinherit, /* use the standard syntax table for this character */ |
5eea1c5a | 50 | Scomment_fence, /* Starts/ends comment which is delimited on the |
47ab3db5 | 51 | other side by any char with the same syntaxcode. */ |
5eea1c5a | 52 | Sstring_fence, /* Starts/ends string which is delimited on the |
47ab3db5 | 53 | other side by any char with the same syntaxcode. */ |
9889c728 JB |
54 | Smax /* Upper bound on codes that are meaningful */ |
55 | }; | |
56 | ||
e0b8ff93 | 57 | /* Set the syntax entry VAL for char C in table TABLE. */ |
e46c910e | 58 | |
f4926ee8 | 59 | #define SET_RAW_SYNTAX_ENTRY(table, c, val) \ |
dcb82a5e | 60 | CHAR_TABLE_SET ((table), c, (val)) |
e0b8ff93 | 61 | |
f4926ee8 KH |
62 | /* Set the syntax entry VAL for char-range RANGE in table TABLE. |
63 | RANGE is a cons (FROM . TO) specifying the range of characters. */ | |
e0b8ff93 | 64 | |
f4926ee8 KH |
65 | #define SET_RAW_SYNTAX_ENTRY_RANGE(table, range, val) \ |
66 | Fset_char_table_range ((table), (range), (val)) | |
9889c728 | 67 | |
5eea1c5a | 68 | /* SYNTAX_ENTRY fetches the information from the entry for character C |
177c0ea7 | 69 | in syntax table TABLE, or from globally kept data (gl_state). |
5eea1c5a RS |
70 | Does inheritance. */ |
71 | /* CURRENT_SYNTAX_TABLE gives the syntax table valid for current | |
72 | position, it is either the buffer's syntax table, or syntax table | |
73 | found in text properties. */ | |
74 | ||
75 | #ifdef SYNTAX_ENTRY_VIA_PROPERTY | |
76 | # define SYNTAX_ENTRY(c) \ | |
77 | (gl_state.use_global ? gl_state.global_code : SYNTAX_ENTRY_INT (c)) | |
78 | # define CURRENT_SYNTAX_TABLE gl_state.current_syntax_table | |
79 | #else | |
80 | # define SYNTAX_ENTRY SYNTAX_ENTRY_INT | |
4b4deea2 | 81 | # define CURRENT_SYNTAX_TABLE BVAR (current_buffer, syntax_table) |
5eea1c5a | 82 | #endif |
e0b8ff93 | 83 | |
501d7ac6 | 84 | #define SYNTAX_ENTRY_INT(c) CHAR_TABLE_REF (CURRENT_SYNTAX_TABLE, (c)) |
e0b8ff93 | 85 | |
e46c910e | 86 | /* Extract the information from the entry for character C |
e0b8ff93 | 87 | in the current syntax table. */ |
c8cdcb16 RS |
88 | |
89 | #ifdef __GNUC__ | |
e46c910e | 90 | #define SYNTAX(c) \ |
874757e8 AS |
91 | ({ Lisp_Object _syntax_temp; \ |
92 | _syntax_temp = SYNTAX_ENTRY (c); \ | |
93 | (CONSP (_syntax_temp) \ | |
94 | ? (enum syntaxcode) (XINT (XCAR (_syntax_temp)) & 0xff) \ | |
e0b8ff93 | 95 | : Swhitespace); }) |
e46c910e RS |
96 | |
97 | #define SYNTAX_WITH_FLAGS(c) \ | |
874757e8 AS |
98 | ({ Lisp_Object _syntax_temp; \ |
99 | _syntax_temp = SYNTAX_ENTRY (c); \ | |
100 | (CONSP (_syntax_temp) \ | |
101 | ? XINT (XCAR (_syntax_temp)) \ | |
e0b8ff93 | 102 | : (int) Swhitespace); }) |
e46c910e RS |
103 | |
104 | #define SYNTAX_MATCH(c) \ | |
874757e8 AS |
105 | ({ Lisp_Object _syntax_temp; \ |
106 | _syntax_temp = SYNTAX_ENTRY (c); \ | |
107 | (CONSP (_syntax_temp) \ | |
108 | ? XCDR (_syntax_temp) \ | |
e0b8ff93 | 109 | : Qnil); }) |
c8cdcb16 | 110 | #else |
3d7db6f1 | 111 | extern Lisp_Object syntax_temp; |
e46c910e | 112 | #define SYNTAX(c) \ |
9d40ebd2 | 113 | (syntax_temp = SYNTAX_ENTRY ((c)), \ |
e46c910e | 114 | (CONSP (syntax_temp) \ |
3331fb06 | 115 | ? (enum syntaxcode) (XINT (XCAR (syntax_temp)) & 0xff) \ |
e0b8ff93 | 116 | : Swhitespace)) |
e46c910e RS |
117 | |
118 | #define SYNTAX_WITH_FLAGS(c) \ | |
9d40ebd2 | 119 | (syntax_temp = SYNTAX_ENTRY ((c)), \ |
e46c910e | 120 | (CONSP (syntax_temp) \ |
3331fb06 | 121 | ? XINT (XCAR (syntax_temp)) \ |
e0b8ff93 | 122 | : (int) Swhitespace)) |
e46c910e RS |
123 | |
124 | #define SYNTAX_MATCH(c) \ | |
9d40ebd2 | 125 | (syntax_temp = SYNTAX_ENTRY ((c)), \ |
e46c910e | 126 | (CONSP (syntax_temp) \ |
3331fb06 | 127 | ? XCDR (syntax_temp) \ |
e0b8ff93 | 128 | : Qnil)) |
c8cdcb16 | 129 | #endif |
9889c728 | 130 | |
a306d6f1 | 131 | |
c5683ceb SM |
132 | /* Whether the syntax of the character C has the prefix flag set. */ |
133 | extern int syntax_prefix_flag_p (int c); | |
c0364919 | 134 | |
9889c728 JB |
135 | /* This array, indexed by a character, contains the syntax code which that |
136 | character signifies (as a char). For example, | |
5eea1c5a | 137 | (enum syntaxcode) syntax_spec_code['w'] is Sword. */ |
9889c728 JB |
138 | |
139 | extern unsigned char syntax_spec_code[0400]; | |
140 | ||
5eea1c5a RS |
141 | /* Indexed by syntax code, give the letter that describes it. */ |
142 | ||
143 | extern char syntax_code_spec[16]; | |
144 | ||
c292db29 | 145 | /* Convert the byte offset BYTEPOS into a character position, |
2f16e7fd RS |
146 | for the object recorded in gl_state with SETUP_SYNTAX_TABLE_FOR_OBJECT. |
147 | ||
148 | The value is meant for use in the UPDATE_SYNTAX_TABLE... macros. | |
149 | These macros do nothing when parse_sexp_lookup_properties is 0, | |
150 | so we return 0 in that case, for speed. */ | |
c292db29 RS |
151 | |
152 | #define SYNTAX_TABLE_BYTE_TO_CHAR(bytepos) \ | |
2f16e7fd RS |
153 | (! parse_sexp_lookup_properties \ |
154 | ? 0 \ | |
155 | : STRINGP (gl_state.object) \ | |
c292db29 RS |
156 | ? string_byte_to_char (gl_state.object, (bytepos)) \ |
157 | : BUFFERP (gl_state.object) \ | |
f79b4b7e KH |
158 | ? buf_bytepos_to_charpos (XBUFFER (gl_state.object), \ |
159 | (bytepos) + BUF_BEGV_BYTE (XBUFFER (gl_state.object)) - 1) - BUF_BEGV (XBUFFER (gl_state.object)) + 1 \ | |
c292db29 | 160 | : NILP (gl_state.object) \ |
f79b4b7e | 161 | ? BYTE_TO_CHAR ((bytepos) + BEGV_BYTE - 1) - BEGV + 1 \ |
c292db29 RS |
162 | : (bytepos)) |
163 | ||
f79b4b7e KH |
164 | /* Make syntax table state (gl_state) good for CHARPOS, assuming it is |
165 | currently good for a position before CHARPOS. */ | |
5eea1c5a | 166 | |
f79b4b7e | 167 | #define UPDATE_SYNTAX_TABLE_FORWARD(charpos) \ |
2f16e7fd | 168 | (parse_sexp_lookup_properties \ |
f79b4b7e KH |
169 | && (charpos) >= gl_state.e_property \ |
170 | ? (update_syntax_table ((charpos) + gl_state.offset, 1, 0, \ | |
c292db29 RS |
171 | gl_state.object), \ |
172 | 1) \ | |
173 | : 0) | |
5eea1c5a | 174 | |
f79b4b7e KH |
175 | /* Make syntax table state (gl_state) good for CHARPOS, assuming it is |
176 | currently good for a position after CHARPOS. */ | |
5eea1c5a | 177 | |
f79b4b7e | 178 | #define UPDATE_SYNTAX_TABLE_BACKWARD(charpos) \ |
2f16e7fd | 179 | (parse_sexp_lookup_properties \ |
4948e1f2 | 180 | && (charpos) < gl_state.b_property \ |
f79b4b7e | 181 | ? (update_syntax_table ((charpos) + gl_state.offset, -1, 0, \ |
c292db29 RS |
182 | gl_state.object), \ |
183 | 1) \ | |
184 | : 0) | |
e2d8d746 | 185 | |
f79b4b7e | 186 | /* Make syntax table good for CHARPOS. */ |
e2d8d746 | 187 | |
f79b4b7e | 188 | #define UPDATE_SYNTAX_TABLE(charpos) \ |
2f16e7fd | 189 | (parse_sexp_lookup_properties \ |
4948e1f2 | 190 | && (charpos) < gl_state.b_property \ |
f79b4b7e | 191 | ? (update_syntax_table ((charpos) + gl_state.offset, -1, 0, \ |
c292db29 RS |
192 | gl_state.object), \ |
193 | 1) \ | |
2f16e7fd | 194 | : (parse_sexp_lookup_properties \ |
f79b4b7e KH |
195 | && (charpos) >= gl_state.e_property \ |
196 | ? (update_syntax_table ((charpos) + gl_state.offset, 1, 0,\ | |
c292db29 RS |
197 | gl_state.object), \ |
198 | 1) \ | |
199 | : 0)) | |
5eea1c5a | 200 | |
d48cd3f4 SM |
201 | /* This macro sets up the buffer-global syntax table. */ |
202 | #define SETUP_BUFFER_SYNTAX_TABLE() \ | |
203 | do \ | |
204 | { \ | |
205 | gl_state.use_global = 0; \ | |
04c56954 | 206 | gl_state.current_syntax_table = BVAR (current_buffer, syntax_table);\ |
d48cd3f4 SM |
207 | } while (0) |
208 | ||
5eea1c5a RS |
209 | /* This macro should be called with FROM at the start of forward |
210 | search, or after the last position of the backward search. It | |
211 | makes sure that the first char is picked up with correct table, so | |
212 | one does not need to call UPDATE_SYNTAX_TABLE immediately after the | |
177c0ea7 | 213 | call. |
5eea1c5a RS |
214 | Sign of COUNT gives the direction of the search. |
215 | */ | |
216 | ||
c292db29 | 217 | #define SETUP_SYNTAX_TABLE(FROM, COUNT) \ |
92413ef3 | 218 | do \ |
2f16e7fd | 219 | { \ |
d48cd3f4 | 220 | SETUP_BUFFER_SYNTAX_TABLE (); \ |
4948e1f2 | 221 | gl_state.b_property = BEGV; \ |
2f16e7fd RS |
222 | gl_state.e_property = ZV + 1; \ |
223 | gl_state.object = Qnil; \ | |
2f16e7fd | 224 | gl_state.offset = 0; \ |
2f16e7fd | 225 | if (parse_sexp_lookup_properties) \ |
9b9794f0 RS |
226 | if ((COUNT) > 0 || (FROM) > BEGV) \ |
227 | update_syntax_table ((COUNT) > 0 ? (FROM) : (FROM) - 1, (COUNT),\ | |
228 | 1, Qnil); \ | |
2f16e7fd | 229 | } \ |
92413ef3 | 230 | while (0) |
5eea1c5a RS |
231 | |
232 | /* Same as above, but in OBJECT. If OBJECT is nil, use current buffer. | |
04c56954 EZ |
233 | If it is t (which is only used in fast_c_string_match_ignore_case), |
234 | ignore properties altogether. | |
e2d8d746 RS |
235 | |
236 | This is meant for regex.c to use. For buffers, regex.c passes arguments | |
237 | to the UPDATE_SYNTAX_TABLE macros which are relative to BEGV. | |
4bbd5bc3 | 238 | So if it is a buffer, we set the offset field to BEGV. */ |
5eea1c5a | 239 | |
c292db29 | 240 | #define SETUP_SYNTAX_TABLE_FOR_OBJECT(OBJECT, FROM, COUNT) \ |
92413ef3 | 241 | do \ |
4bbd5bc3 | 242 | { \ |
d48cd3f4 | 243 | SETUP_BUFFER_SYNTAX_TABLE (); \ |
c292db29 RS |
244 | gl_state.object = (OBJECT); \ |
245 | if (BUFFERP (gl_state.object)) \ | |
246 | { \ | |
247 | struct buffer *buf = XBUFFER (gl_state.object); \ | |
4948e1f2 | 248 | gl_state.b_property = 1; \ |
1d1293dd | 249 | gl_state.e_property = BUF_ZV (buf) - BUF_BEGV (buf) + 1; \ |
c292db29 RS |
250 | gl_state.offset = BUF_BEGV (buf) - 1; \ |
251 | } \ | |
252 | else if (NILP (gl_state.object)) \ | |
4bbd5bc3 | 253 | { \ |
4948e1f2 | 254 | gl_state.b_property = 1; \ |
1d1293dd | 255 | gl_state.e_property = ZV - BEGV + 1; \ |
4bbd5bc3 RS |
256 | gl_state.offset = BEGV - 1; \ |
257 | } \ | |
c292db29 | 258 | else if (EQ (gl_state.object, Qt)) \ |
4bbd5bc3 | 259 | { \ |
4948e1f2 | 260 | gl_state.b_property = 0; \ |
04c56954 | 261 | gl_state.e_property = MOST_POSITIVE_FIXNUM; \ |
4bbd5bc3 RS |
262 | gl_state.offset = 0; \ |
263 | } \ | |
264 | else \ | |
265 | { \ | |
4948e1f2 | 266 | gl_state.b_property = 0; \ |
d5db4077 | 267 | gl_state.e_property = 1 + SCHARS (gl_state.object); \ |
4bbd5bc3 RS |
268 | gl_state.offset = 0; \ |
269 | } \ | |
4bbd5bc3 | 270 | if (parse_sexp_lookup_properties) \ |
f79b4b7e | 271 | update_syntax_table (((FROM) + gl_state.offset \ |
c292db29 RS |
272 | + (COUNT > 0 ? 0 : -1)), \ |
273 | COUNT, 1, gl_state.object); \ | |
4bbd5bc3 | 274 | } \ |
92413ef3 | 275 | while (0) |
5eea1c5a RS |
276 | |
277 | struct gl_state_s | |
278 | { | |
c292db29 | 279 | Lisp_Object object; /* The object we are scanning. */ |
d1dfb56c EZ |
280 | EMACS_INT start; /* Where to stop. */ |
281 | EMACS_INT stop; /* Where to stop. */ | |
5eea1c5a RS |
282 | int use_global; /* Whether to use global_code |
283 | or c_s_t. */ | |
284 | Lisp_Object global_code; /* Syntax code of current char. */ | |
285 | Lisp_Object current_syntax_table; /* Syntax table for current pos. */ | |
286 | Lisp_Object old_prop; /* Syntax-table prop at prev pos. */ | |
d1dfb56c EZ |
287 | EMACS_INT b_property; /* First index where c_s_t is valid. */ |
288 | EMACS_INT e_property; /* First index where c_s_t is | |
5eea1c5a RS |
289 | not valid. */ |
290 | INTERVAL forward_i; /* Where to start lookup on forward */ | |
291 | INTERVAL backward_i; /* or backward movement. The | |
292 | data in c_s_t is valid | |
293 | between these intervals, | |
294 | and possibly at the | |
295 | intervals too, depending | |
296 | on: */ | |
e2d8d746 | 297 | /* Offset for positions specified to UPDATE_SYNTAX_TABLE. */ |
d1dfb56c | 298 | EMACS_INT offset; |
5eea1c5a RS |
299 | }; |
300 | ||
301 | extern struct gl_state_s gl_state; | |
4f3a2f8d | 302 | extern EMACS_INT scan_words (EMACS_INT, EMACS_INT); |