Commit | Line | Data |
---|---|---|
c38e0c97 | 1 | /* Tags file maker to go with GNU Emacs -*- coding: utf-8 -*- |
3b7ad313 | 2 | |
b7e1b21b | 3 | Copyright (C) 1984 The Regents of the University of California |
c6d46f5f | 4 | |
b7e1b21b GM |
5 | Redistribution and use in source and binary forms, with or without |
6 | modification, are permitted provided that the following conditions are | |
7 | met: | |
8 | 1. Redistributions of source code must retain the above copyright | |
9 | notice, this list of conditions and the following disclaimer. | |
10 | 2. Redistributions in binary form must reproduce the above copyright | |
11 | notice, this list of conditions and the following disclaimer in the | |
12 | documentation and/or other materials provided with the | |
13 | distribution. | |
14 | 3. Neither the name of the University nor the names of its | |
15 | contributors may be used to endorse or promote products derived | |
16 | from this software without specific prior written permission. | |
c6d46f5f | 17 | |
b7e1b21b GM |
18 | THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' |
19 | AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, | |
20 | THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR | |
21 | PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS | |
22 | BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR | |
23 | CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF | |
24 | SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR | |
25 | BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, | |
26 | WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE | |
27 | OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN | |
28 | IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. | |
29 | ||
30 | ||
ba318903 | 31 | Copyright (C) 1984, 1987-1989, 1993-1995, 1998-2014 Free Software |
ab422c4d | 32 | Foundation, Inc. |
b7e1b21b GM |
33 | |
34 | This file is not considered part of GNU Emacs. | |
35 | ||
294981c7 | 36 | This program is free software: you can redistribute it and/or modify |
b7e1b21b | 37 | it under the terms of the GNU General Public License as published by |
294981c7 GM |
38 | the Free Software Foundation, either version 3 of the License, or |
39 | (at your option) any later version. | |
b7e1b21b GM |
40 | |
41 | This program is distributed in the hope that it will be useful, | |
42 | but WITHOUT ANY WARRANTY; without even the implied warranty of | |
43 | MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the | |
44 | GNU General Public License for more details. | |
45 | ||
46 | You should have received a copy of the GNU General Public License | |
294981c7 | 47 | along with this program. If not, see <http://www.gnu.org/licenses/>. */ |
b7e1b21b GM |
48 | |
49 | ||
50 | /* NB To comply with the above BSD license, copyright information is | |
51 | reproduced in etc/ETAGS.README. That file should be updated when the | |
52 | above notices are. | |
53 | ||
54 | To the best of our knowledge, this code was originally based on the | |
55 | ctags.c distributed with BSD4.2, which was copyrighted by the | |
56 | University of California, as described above. */ | |
c6d46f5f | 57 | |
c6d46f5f JB |
58 | |
59 | /* | |
60 | * Authors: | |
9817cf3f FP |
61 | * 1983 Ctags originally by Ken Arnold. |
62 | * 1984 Fortran added by Jim Kleckner. | |
63 | * 1984 Ed Pelegri-Llopart added C typedefs. | |
64 | * 1985 Emacs TAGS format by Richard Stallman. | |
65 | * 1989 Sam Kendall added C++. | |
eaaacacd | 66 | * 1992 Joseph B. Wells improved C and C++ parsing. |
c38e0c97 | 67 | * 1993 Francesco Potortì reorganized C and C++. |
9817cf3f | 68 | * 1994 Line-by-line regexp tags by Tom Tromey. |
c38e0c97 PE |
69 | * 2001 Nested classes by Francesco Potortì (concept by Mykola Dzyuba). |
70 | * 2002 #line directives by Francesco Potortì. | |
31d4b314 | 71 | * |
c38e0c97 | 72 | * Francesco Potortì <pot@gnu.org> has maintained and improved it since 1993. |
c6d46f5f JB |
73 | */ |
74 | ||
05d9a399 FP |
75 | /* |
76 | * If you want to add support for a new language, start by looking at the LUA | |
9817cf3f FP |
77 | * language, which is the simplest. Alternatively, consider distributing etags |
78 | * together with a configuration file containing regexp definitions for etags. | |
05d9a399 FP |
79 | */ |
80 | ||
d3706fa9 | 81 | char pot_etags_version[] = "@(#) pot revision number is 17.38.1.4"; |
75bdbc6a | 82 | |
b2521b0a FP |
83 | #ifdef DEBUG |
84 | # undef DEBUG | |
96c06863 | 85 | # define DEBUG true |
b2521b0a | 86 | #else |
96c06863 | 87 | # define DEBUG false |
b2521b0a | 88 | # define NDEBUG /* disable assert */ |
75bdbc6a | 89 | #endif |
46c145db | 90 | |
cf38a720 | 91 | #include <config.h> |
4ee9629e | 92 | |
944bdd72 FP |
93 | #ifndef _GNU_SOURCE |
94 | # define _GNU_SOURCE 1 /* enables some compiler checks on GNU */ | |
95 | #endif | |
96 | ||
35a4c758 | 97 | /* WIN32_NATIVE is for XEmacs. |
b28e26be FP |
98 | MSDOS, WINDOWSNT, DOS_NT are for Emacs. */ |
99 | #ifdef WIN32_NATIVE | |
100 | # undef MSDOS | |
101 | # undef WINDOWSNT | |
102 | # define WINDOWSNT | |
103 | #endif /* WIN32_NATIVE */ | |
104 | ||
c6880c90 | 105 | #ifdef MSDOS |
71cbb895 | 106 | # undef MSDOS |
96c06863 | 107 | # define MSDOS true |
8dc7496c FP |
108 | # include <fcntl.h> |
109 | # include <sys/param.h> | |
93b7ac65 | 110 | # include <io.h> |
944bdd72 | 111 | #else |
96c06863 | 112 | # define MSDOS false |
c6880c90 RS |
113 | #endif /* MSDOS */ |
114 | ||
c05b6df5 | 115 | #ifdef WINDOWSNT |
8dc7496c | 116 | # include <fcntl.h> |
944bdd72 | 117 | # include <direct.h> |
a65c6351 | 118 | # include <io.h> |
8dc7496c | 119 | # define MAXPATHLEN _MAX_PATH |
b28e26be FP |
120 | # undef HAVE_NTGUI |
121 | # undef DOS_NT | |
122 | # define DOS_NT | |
9239d970 | 123 | #endif /* WINDOWSNT */ |
cec68fcb | 124 | |
4004364e | 125 | #include <unistd.h> |
db5a3003 | 126 | #include <stdarg.h> |
0e926e56 PE |
127 | #include <stdlib.h> |
128 | #include <string.h> | |
918f9ad1 JB |
129 | #include <stdio.h> |
130 | #include <ctype.h> | |
dcc89e63 | 131 | #include <errno.h> |
918f9ad1 JB |
132 | #include <sys/types.h> |
133 | #include <sys/stat.h> | |
fee5959d | 134 | #include <c-strcase.h> |
918f9ad1 | 135 | |
8f79fe72 FP |
136 | #include <assert.h> |
137 | #ifdef NDEBUG | |
138 | # undef assert /* some systems have a buggy assert.h */ | |
139 | # define assert(x) ((void) 0) | |
914d7258 AI |
140 | #endif |
141 | ||
9239d970 | 142 | #include <getopt.h> |
35a4c758 | 143 | #include <regex.h> |
918f9ad1 | 144 | |
32daa216 | 145 | /* Define CTAGS to make the program "ctags" compatible with the usual one. |
85e77a3f | 146 | Leave it undefined to make the program "etags", which makes emacs-style |
32daa216 FP |
147 | tag tables and tags typedefs, #defines and struct/union/enum by default. */ |
148 | #ifdef CTAGS | |
149 | # undef CTAGS | |
96c06863 | 150 | # define CTAGS true |
32daa216 | 151 | #else |
96c06863 | 152 | # define CTAGS false |
c6d46f5f JB |
153 | #endif |
154 | ||
5e617bc2 | 155 | #define streq(s,t) (assert ((s)!=NULL || (t)!=NULL), !strcmp (s, t)) |
fee5959d | 156 | #define strcaseeq(s,t) (assert ((s)!=NULL && (t)!=NULL), !c_strcasecmp (s, t)) |
5e617bc2 | 157 | #define strneq(s,t,n) (assert ((s)!=NULL || (t)!=NULL), !strncmp (s, t, n)) |
fee5959d | 158 | #define strncaseeq(s,t,n) (assert ((s)!=NULL && (t)!=NULL), !c_strncasecmp (s, t, n)) |
c6d46f5f | 159 | |
30903246 | 160 | #define CHARS 256 /* 2^sizeof(char) */ |
89d57763 | 161 | #define CHAR(x) ((unsigned int)(x) & (CHARS - 1)) |
5e617bc2 JB |
162 | #define iswhite(c) (_wht[CHAR (c)]) /* c is white (see white) */ |
163 | #define notinname(c) (_nin[CHAR (c)]) /* c is not in a name (see nonam) */ | |
164 | #define begtoken(c) (_btk[CHAR (c)]) /* c can start token (see begtk) */ | |
165 | #define intoken(c) (_itk[CHAR (c)]) /* c can be in token (see midtk) */ | |
166 | #define endtoken(c) (_etk[CHAR (c)]) /* c ends tokens (see endtk) */ | |
c6d46f5f | 167 | |
5e617bc2 JB |
168 | #define ISALNUM(c) isalnum (CHAR (c)) |
169 | #define ISALPHA(c) isalpha (CHAR (c)) | |
170 | #define ISDIGIT(c) isdigit (CHAR (c)) | |
171 | #define ISLOWER(c) islower (CHAR (c)) | |
89d57763 | 172 | |
5e617bc2 | 173 | #define lowcase(c) tolower (CHAR (c)) |
af03e6ab | 174 | |
b2db879b | 175 | |
55597f90 | 176 | /* |
93b7ac65 | 177 | * xnew, xrnew -- allocate, reallocate storage |
55597f90 FP |
178 | * |
179 | * SYNOPSIS: Type *xnew (int n, Type); | |
b2521b0a | 180 | * void xrnew (OldPointer, int n, Type); |
55597f90 | 181 | */ |
b2521b0a | 182 | #if DEBUG |
2f608d34 | 183 | # include "chkmalloc.h" |
93b7ac65 FP |
184 | # define xnew(n,Type) ((Type *) trace_malloc (__FILE__, __LINE__, \ |
185 | (n) * sizeof (Type))) | |
b2521b0a FP |
186 | # define xrnew(op,n,Type) ((op) = (Type *) trace_realloc (__FILE__, __LINE__, \ |
187 | (char *) (op), (n) * sizeof (Type))) | |
2f608d34 | 188 | #else |
93b7ac65 | 189 | # define xnew(n,Type) ((Type *) xmalloc ((n) * sizeof (Type))) |
b2521b0a FP |
190 | # define xrnew(op,n,Type) ((op) = (Type *) xrealloc ( \ |
191 | (char *) (op), (n) * sizeof (Type))) | |
2f608d34 | 192 | #endif |
c6d46f5f | 193 | |
f57e2426 | 194 | typedef void Lang_function (FILE *); |
c6d46f5f | 195 | |
93b7ac65 FP |
196 | typedef struct |
197 | { | |
988e88ab J |
198 | const char *suffix; /* file name suffix for this compressor */ |
199 | const char *command; /* takes one arg and decompresses to stdout */ | |
93b7ac65 | 200 | } compressor; |
c6d46f5f | 201 | |
93b7ac65 FP |
202 | typedef struct |
203 | { | |
988e88ab J |
204 | const char *name; /* language name */ |
205 | const char *help; /* detailed help for the language */ | |
9e0a3f98 | 206 | Lang_function *function; /* parse function */ |
988e88ab J |
207 | const char **suffixes; /* name suffixes of this language's files */ |
208 | const char **filenames; /* names of this language's files */ | |
209 | const char **interpreters; /* interpreters for this language */ | |
4fdb4553 | 210 | bool metasource; /* source used to generate other sources */ |
93b7ac65 FP |
211 | } language; |
212 | ||
9e0a3f98 FP |
213 | typedef struct fdesc |
214 | { | |
215 | struct fdesc *next; /* for the linked list */ | |
216 | char *infname; /* uncompressed input file name */ | |
217 | char *infabsname; /* absolute uncompressed input file name */ | |
218 | char *infabsdir; /* absolute dir of input file */ | |
219 | char *taggedfname; /* file name to write in tagfile */ | |
220 | language *lang; /* language of file */ | |
221 | char *prop; /* file properties to write in tagfile */ | |
222 | bool usecharno; /* etags tags shall contain char number */ | |
61a1f6fa | 223 | bool written; /* entry written in the tags file */ |
9e0a3f98 FP |
224 | } fdesc; |
225 | ||
944bdd72 | 226 | typedef struct node_st |
9e0a3f98 FP |
227 | { /* sorting structure */ |
228 | struct node_st *left, *right; /* left and right sons */ | |
229 | fdesc *fdp; /* description of file to whom tag belongs */ | |
988e88ab | 230 | char *name; /* tag name */ |
61a1f6fa | 231 | char *regex; /* search regexp */ |
9e0a3f98 | 232 | bool valid; /* write this tag on the tag file */ |
61a1f6fa | 233 | bool is_func; /* function tag: use regexp in CTAGS mode */ |
9e0a3f98 FP |
234 | bool been_warned; /* warning already given for duplicated tag */ |
235 | int lno; /* line number tag is on */ | |
944bdd72 | 236 | long cno; /* character number line starts on */ |
944bdd72 FP |
237 | } node; |
238 | ||
239 | /* | |
240 | * A `linebuffer' is a structure which holds a line of text. | |
241 | * `readline_internal' reads a line from a stream into a linebuffer | |
242 | * and works regardless of the length of the line. | |
243 | * SIZE is the size of BUFFER, LEN is the length of the string in | |
244 | * BUFFER after readline reads it. | |
245 | */ | |
246 | typedef struct | |
247 | { | |
248 | long size; | |
249 | int len; | |
250 | char *buffer; | |
251 | } linebuffer; | |
b9755a12 | 252 | |
9e0a3f98 FP |
253 | /* Used to support mixing of --lang and file names. */ |
254 | typedef struct | |
255 | { | |
256 | enum { | |
257 | at_language, /* a language specification */ | |
258 | at_regexp, /* a regular expression */ | |
97b90b0a | 259 | at_filename, /* a file name */ |
4fdb4553 FP |
260 | at_stdin, /* read from stdin here */ |
261 | at_end /* stop parsing the list */ | |
9e0a3f98 FP |
262 | } arg_type; /* argument type */ |
263 | language *lang; /* language associated with the argument */ | |
264 | char *what; /* the argument itself */ | |
265 | } argument; | |
266 | ||
9e0a3f98 | 267 | /* Structure defining a regular expression. */ |
61a1f6fa | 268 | typedef struct regexp |
9e0a3f98 | 269 | { |
61a1f6fa FP |
270 | struct regexp *p_next; /* pointer to next in list */ |
271 | language *lang; /* if set, use only for this language */ | |
272 | char *pattern; /* the regexp pattern */ | |
273 | char *name; /* tag name */ | |
274 | struct re_pattern_buffer *pat; /* the compiled pattern */ | |
275 | struct re_registers regs; /* re registers */ | |
276 | bool error_signaled; /* already signaled for this regexp */ | |
53964682 | 277 | bool force_explicit_name; /* do not allow implicit tag name */ |
61a1f6fa FP |
278 | bool ignore_case; /* ignore case when matching */ |
279 | bool multi_line; /* do a multi-line match on the whole file */ | |
280 | } regexp; | |
9e0a3f98 FP |
281 | |
282 | ||
30903246 | 283 | /* Many compilers barf on this: |
93b7ac65 | 284 | Lang_function Ada_funcs; |
30903246 | 285 | so let's write it this way */ |
f57e2426 J |
286 | static void Ada_funcs (FILE *); |
287 | static void Asm_labels (FILE *); | |
288 | static void C_entries (int c_ext, FILE *); | |
289 | static void default_C_entries (FILE *); | |
290 | static void plain_C_entries (FILE *); | |
291 | static void Cjava_entries (FILE *); | |
292 | static void Cobol_paragraphs (FILE *); | |
293 | static void Cplusplus_entries (FILE *); | |
294 | static void Cstar_entries (FILE *); | |
295 | static void Erlang_functions (FILE *); | |
296 | static void Forth_words (FILE *); | |
297 | static void Fortran_functions (FILE *); | |
298 | static void HTML_labels (FILE *); | |
299 | static void Lisp_functions (FILE *); | |
300 | static void Lua_functions (FILE *); | |
301 | static void Makefile_targets (FILE *); | |
302 | static void Pascal_functions (FILE *); | |
303 | static void Perl_functions (FILE *); | |
304 | static void PHP_functions (FILE *); | |
305 | static void PS_functions (FILE *); | |
306 | static void Prolog_functions (FILE *); | |
307 | static void Python_functions (FILE *); | |
308 | static void Scheme_functions (FILE *); | |
309 | static void TeX_commands (FILE *); | |
310 | static void Texinfo_nodes (FILE *); | |
311 | static void Yacc_entries (FILE *); | |
312 | static void just_read_file (FILE *); | |
313 | ||
f57e2426 | 314 | static language *get_language_from_langname (const char *); |
f57e2426 J |
315 | static void readline (linebuffer *, FILE *); |
316 | static long readline_internal (linebuffer *, FILE *); | |
988e88ab | 317 | static bool nocase_tail (const char *); |
f57e2426 J |
318 | static void get_tag (char *, char **); |
319 | ||
320 | static void analyse_regex (char *); | |
321 | static void free_regexps (void); | |
322 | static void regex_tag_multiline (void); | |
db5a3003 | 323 | static void error (const char *, ...) ATTRIBUTE_FORMAT_PRINTF (1, 2); |
845ca893 PE |
324 | static _Noreturn void suggest_asking_for_help (void); |
325 | _Noreturn void fatal (const char *, const char *); | |
326 | static _Noreturn void pfatal (const char *); | |
f57e2426 J |
327 | static void add_node (node *, node **); |
328 | ||
329 | static void init (void); | |
330 | static void process_file_name (char *, language *); | |
331 | static void process_file (FILE *, char *, language *); | |
332 | static void find_entries (FILE *); | |
333 | static void free_tree (node *); | |
334 | static void free_fdesc (fdesc *); | |
335 | static void pfnote (char *, bool, char *, int, int, long); | |
f57e2426 J |
336 | static void invalidate_nodes (fdesc *, node **); |
337 | static void put_entries (node *); | |
338 | ||
988e88ab | 339 | static char *concat (const char *, const char *, const char *); |
f57e2426 J |
340 | static char *skip_spaces (char *); |
341 | static char *skip_non_spaces (char *); | |
1cbaa705 | 342 | static char *skip_name (char *); |
988e88ab J |
343 | static char *savenstr (const char *, int); |
344 | static char *savestr (const char *); | |
f57e2426 J |
345 | static char *etags_strchr (const char *, int); |
346 | static char *etags_strrchr (const char *, int); | |
f57e2426 J |
347 | static char *etags_getcwd (void); |
348 | static char *relative_filename (char *, char *); | |
349 | static char *absolute_filename (char *, char *); | |
350 | static char *absolute_dirname (char *, char *); | |
351 | static bool filename_is_absolute (char *f); | |
352 | static void canonicalize_filename (char *); | |
353 | static void linebuffer_init (linebuffer *); | |
354 | static void linebuffer_setlen (linebuffer *, int); | |
261cb4bb PE |
355 | static void *xmalloc (size_t); |
356 | static void *xrealloc (char *, size_t); | |
93b7ac65 | 357 | |
c6d46f5f | 358 | \f |
db590582 | 359 | static char searchar = '/'; /* use /.../ searches */ |
c6d46f5f | 360 | |
db590582 FP |
361 | static char *tagfile; /* output file */ |
362 | static char *progname; /* name this program was invoked with */ | |
363 | static char *cwd; /* current working directory */ | |
364 | static char *tagfiledir; /* directory of tagfile */ | |
365 | static FILE *tagf; /* ioptr for tags file */ | |
9250f758 | 366 | static ptrdiff_t whatlen_max; /* maximum length of any 'what' member */ |
93b7ac65 | 367 | |
9e0a3f98 FP |
368 | static fdesc *fdhead; /* head of file description list */ |
369 | static fdesc *curfdp; /* current file description */ | |
db590582 FP |
370 | static int lineno; /* line number of current line */ |
371 | static long charno; /* current character number */ | |
372 | static long linecharno; /* charno of start of current line */ | |
373 | static char *dbp; /* pointer to start of current tag */ | |
9e0a3f98 | 374 | |
db590582 | 375 | static const int invalidcharno = -1; |
93b7ac65 | 376 | |
9e0a3f98 | 377 | static node *nodehead; /* the head of the binary tree of tags */ |
8378bcd3 | 378 | static node *last_node; /* the last node created */ |
c6d46f5f | 379 | |
db590582 | 380 | static linebuffer lb; /* the current line */ |
24dbe96a | 381 | static linebuffer filebuf; /* a buffer containing the whole file */ |
61a1f6fa | 382 | static linebuffer token_name; /* a buffer containing a tag name */ |
c6d46f5f | 383 | |
55597f90 | 384 | /* boolean "functions" (see init) */ |
db590582 | 385 | static bool _wht[CHARS], _nin[CHARS], _itk[CHARS], _btk[CHARS], _etk[CHARS]; |
988e88ab | 386 | static const char |
d8913c1c | 387 | /* white chars */ |
71cbb895 | 388 | *white = " \f\t\n\r\v", |
30903246 | 389 | /* not in a name */ |
2201e3dc | 390 | *nonam = " \f\t\n\r()=,;", /* look at make_tag before modifying! */ |
d8913c1c | 391 | /* token ending chars */ |
93b7ac65 | 392 | *endtk = " \t\n\r\"'#()[]{}=-+%*/&|^~!<>;,.:?", |
d8913c1c FP |
393 | /* token starting chars */ |
394 | *begtk = "ABCDEFGHIJKLMNOPQRSTUVWXYZ_abcdefghijklmnopqrstuvwxyz$~@", | |
395 | /* valid in-token chars */ | |
30903246 | 396 | *midtk = "ABCDEFGHIJKLMNOPQRSTUVWXYZ_abcdefghijklmnopqrstuvwxyz$0123456789"; |
c6d46f5f | 397 | |
db590582 | 398 | static bool append_to_tagfile; /* -a: append to tags */ |
96c06863 | 399 | /* The next five default to true in C and derived languages. */ |
db590582 FP |
400 | static bool typedefs; /* -t: create tags for C and Ada typedefs */ |
401 | static bool typedefs_or_cplusplus; /* -T: create tags for C typedefs, level */ | |
c6d46f5f | 402 | /* 0 struct/enum/union decls, and C++ */ |
32daa216 | 403 | /* member functions. */ |
db590582 | 404 | static bool constantypedefs; /* -d: create tags for C #define, enum */ |
30903246 | 405 | /* constants and variables. */ |
c6d46f5f | 406 | /* -D: opposite of -d. Default under ctags. */ |
96c06863 PE |
407 | static int globals; /* create tags for global variables */ |
408 | static int members; /* create tags for C member variables */ | |
409 | static int declarations; /* --declarations: tag them and extern in C&Co*/ | |
410 | static int no_line_directive; /* ignore #line directives (undocumented) */ | |
411 | static int no_duplicates; /* no duplicate tags for ctags (undocumented) */ | |
db590582 FP |
412 | static bool update; /* -u: update tags */ |
413 | static bool vgrind_style; /* -v: create vgrind style index output */ | |
ed8bbc0e | 414 | static bool no_warnings; /* -w: suppress warnings (undocumented) */ |
db590582 | 415 | static bool cxref_style; /* -x: create cxref style output */ |
f96732e5 | 416 | static bool cplusplus; /* .[hc] means C++, not C (undocumented) */ |
3c04a71a | 417 | static bool ignoreindent; /* -I: ignore indentation in C */ |
96c06863 | 418 | static int packages_only; /* --packages-only: in Ada, only tag packages*/ |
30903246 | 419 | |
ce0ae3a9 EZ |
420 | /* STDIN is defined in LynxOS system headers */ |
421 | #ifdef STDIN | |
2e0bea68 | 422 | # undef STDIN |
ce0ae3a9 EZ |
423 | #endif |
424 | ||
97b90b0a FP |
425 | #define STDIN 0x1001 /* returned by getopt_long on --parse-stdin */ |
426 | static bool parsing_stdin; /* --parse-stdin used */ | |
427 | ||
61a1f6fa | 428 | static regexp *p_head; /* list of all regexps */ |
24dbe96a | 429 | static bool need_filebuf; /* some regexes are multi-line */ |
9e0a3f98 | 430 | |
db590582 | 431 | static struct option longopts[] = |
7537186d | 432 | { |
ed8bbc0e | 433 | { "append", no_argument, NULL, 'a' }, |
96c06863 | 434 | { "packages-only", no_argument, &packages_only, 1 }, |
ed8bbc0e | 435 | { "c++", no_argument, NULL, 'C' }, |
96c06863 PE |
436 | { "declarations", no_argument, &declarations, 1 }, |
437 | { "no-line-directive", no_argument, &no_line_directive, 1 }, | |
438 | { "no-duplicates", no_argument, &no_duplicates, 1 }, | |
ed8bbc0e FP |
439 | { "help", no_argument, NULL, 'h' }, |
440 | { "help", no_argument, NULL, 'H' }, | |
441 | { "ignore-indentation", no_argument, NULL, 'I' }, | |
442 | { "language", required_argument, NULL, 'l' }, | |
96c06863 PE |
443 | { "members", no_argument, &members, 1 }, |
444 | { "no-members", no_argument, &members, 0 }, | |
ed8bbc0e FP |
445 | { "output", required_argument, NULL, 'o' }, |
446 | { "regex", required_argument, NULL, 'r' }, | |
447 | { "no-regex", no_argument, NULL, 'R' }, | |
448 | { "ignore-case-regex", required_argument, NULL, 'c' }, | |
97b90b0a | 449 | { "parse-stdin", required_argument, NULL, STDIN }, |
ed8bbc0e | 450 | { "version", no_argument, NULL, 'V' }, |
9e0a3f98 | 451 | |
55102b5d | 452 | #if CTAGS /* Ctags options */ |
ed8bbc0e FP |
453 | { "backward-search", no_argument, NULL, 'B' }, |
454 | { "cxref", no_argument, NULL, 'x' }, | |
455 | { "defines", no_argument, NULL, 'd' }, | |
96c06863 | 456 | { "globals", no_argument, &globals, 1 }, |
ed8bbc0e FP |
457 | { "typedefs", no_argument, NULL, 't' }, |
458 | { "typedefs-and-c++", no_argument, NULL, 'T' }, | |
459 | { "update", no_argument, NULL, 'u' }, | |
460 | { "vgrind", no_argument, NULL, 'v' }, | |
461 | { "no-warn", no_argument, NULL, 'w' }, | |
9e0a3f98 | 462 | |
55102b5d | 463 | #else /* Etags options */ |
ed8bbc0e | 464 | { "no-defines", no_argument, NULL, 'D' }, |
96c06863 | 465 | { "no-globals", no_argument, &globals, 0 }, |
ed8bbc0e | 466 | { "include", required_argument, NULL, 'i' }, |
9e0a3f98 | 467 | #endif |
93b7ac65 | 468 | { NULL } |
4746118a JB |
469 | }; |
470 | ||
db590582 | 471 | static compressor compressors[] = |
93b7ac65 FP |
472 | { |
473 | { "z", "gzip -d -c"}, | |
474 | { "Z", "gzip -d -c"}, | |
475 | { "gz", "gzip -d -c"}, | |
476 | { "GZ", "gzip -d -c"}, | |
477 | { "bz2", "bzip2 -d -c" }, | |
4c964351 | 478 | { "xz", "xz -d -c" }, |
93b7ac65 FP |
479 | { NULL } |
480 | }; | |
481 | ||
1f638249 FP |
482 | /* |
483 | * Language stuff. | |
484 | */ | |
b9755a12 | 485 | |
93b7ac65 | 486 | /* Ada code */ |
988e88ab | 487 | static const char *Ada_suffixes [] = |
93b7ac65 | 488 | { "ads", "adb", "ada", NULL }; |
988e88ab | 489 | static const char Ada_help [] = |
4fdb4553 FP |
490 | "In Ada code, functions, procedures, packages, tasks and types are\n\ |
491 | tags. Use the `--packages-only' option to create tags for\n\ | |
492 | packages only.\n\ | |
493 | Ada tag names have suffixes indicating the type of entity:\n\ | |
494 | Entity type: Qualifier:\n\ | |
495 | ------------ ----------\n\ | |
496 | function /f\n\ | |
497 | procedure /p\n\ | |
498 | package spec /s\n\ | |
499 | package body /b\n\ | |
500 | type /t\n\ | |
501 | task /k\n\ | |
502 | Thus, `M-x find-tag <RET> bidule/b <RET>' will go directly to the\n\ | |
503 | body of the package `bidule', while `M-x find-tag <RET> bidule <RET>'\n\ | |
504 | will just search for any tag `bidule'."; | |
1f638249 FP |
505 | |
506 | /* Assembly code */ | |
988e88ab | 507 | static const char *Asm_suffixes [] = |
db590582 FP |
508 | { "a", /* Unix assembler */ |
509 | "asm", /* Microcontroller assembly */ | |
510 | "def", /* BSO/Tasking definition includes */ | |
511 | "inc", /* Microcontroller include files */ | |
512 | "ins", /* Microcontroller include files */ | |
513 | "s", "sa", /* Unix assembler */ | |
514 | "S", /* cpp-processed Unix assembler */ | |
515 | "src", /* BSO/Tasking C compiler output */ | |
516 | NULL | |
517 | }; | |
988e88ab | 518 | static const char Asm_help [] = |
4fdb4553 FP |
519 | "In assembler code, labels appearing at the beginning of a line,\n\ |
520 | followed by a colon, are tags."; | |
521 | ||
1f638249 FP |
522 | |
523 | /* Note that .c and .h can be considered C++, if the --c++ flag was | |
f96732e5 | 524 | given, or if the `class' or `template' keywords are met inside the file. |
b28e26be | 525 | That is why default_C_entries is called for these. */ |
988e88ab | 526 | static const char *default_C_suffixes [] = |
1f638249 | 527 | { "c", "h", NULL }; |
f96732e5 | 528 | #if CTAGS /* C help for Ctags */ |
988e88ab | 529 | static const char default_C_help [] = |
f96732e5 FP |
530 | "In C code, any C function is a tag. Use -t to tag typedefs.\n\ |
531 | Use -T to tag definitions of `struct', `union' and `enum'.\n\ | |
532 | Use -d to tag `#define' macro definitions and `enum' constants.\n\ | |
533 | Use --globals to tag global variables.\n\ | |
534 | You can tag function declarations and external variables by\n\ | |
535 | using `--declarations', and struct members by using `--members'."; | |
536 | #else /* C help for Etags */ | |
988e88ab | 537 | static const char default_C_help [] = |
4fdb4553 FP |
538 | "In C code, any C function or typedef is a tag, and so are\n\ |
539 | definitions of `struct', `union' and `enum'. `#define' macro\n\ | |
540 | definitions and `enum' constants are tags unless you specify\n\ | |
541 | `--no-defines'. Global variables are tags unless you specify\n\ | |
78064948 FP |
542 | `--no-globals' and so are struct members unless you specify\n\ |
543 | `--no-members'. Use of `--no-globals', `--no-defines' and\n\ | |
544 | `--no-members' can make the tags table file much smaller.\n\ | |
4fdb4553 | 545 | You can tag function declarations and external variables by\n\ |
78064948 | 546 | using `--declarations'."; |
f96732e5 | 547 | #endif /* C help for Ctags and Etags */ |
1f638249 | 548 | |
988e88ab | 549 | static const char *Cplusplus_suffixes [] = |
89d57763 | 550 | { "C", "c++", "cc", "cpp", "cxx", "H", "h++", "hh", "hpp", "hxx", |
cec68fcb | 551 | "M", /* Objective C++ */ |
7877f373 | 552 | "pdb", /* PostScript with C syntax */ |
cec68fcb | 553 | NULL }; |
988e88ab | 554 | static const char Cplusplus_help [] = |
373df066 DL |
555 | "In C++ code, all the tag constructs of C code are tagged. (Use\n\ |
556 | --help --lang=c --lang=c++ for full help.)\n\ | |
78064948 FP |
557 | In addition to C tags, member functions are also recognized. Member\n\ |
558 | variables are recognized unless you use the `--no-members' option.\n\ | |
4fdb4553 FP |
559 | Tags for variables and functions in classes are named `CLASS::VARIABLE'\n\ |
560 | and `CLASS::FUNCTION'. `operator' definitions have tag names like\n\ | |
561 | `operator+'."; | |
cec68fcb | 562 | |
988e88ab | 563 | static const char *Cjava_suffixes [] = |
cec68fcb | 564 | { "java", NULL }; |
4fdb4553 FP |
565 | static char Cjava_help [] = |
566 | "In Java code, all the tags constructs of C and C++ code are\n\ | |
373df066 | 567 | tagged. (Use --help --lang=c --lang=c++ --lang=java for full help.)"; |
4fdb4553 | 568 | |
1f638249 | 569 | |
988e88ab | 570 | static const char *Cobol_suffixes [] = |
30903246 | 571 | { "COB", "cob", NULL }; |
4fdb4553 FP |
572 | static char Cobol_help [] = |
573 | "In Cobol code, tags are paragraph names; that is, any word\n\ | |
574 | starting in column 8 and followed by a period."; | |
30903246 | 575 | |
988e88ab | 576 | static const char *Cstar_suffixes [] = |
1f638249 FP |
577 | { "cs", "hs", NULL }; |
578 | ||
988e88ab | 579 | static const char *Erlang_suffixes [] = |
8dc7496c | 580 | { "erl", "hrl", NULL }; |
988e88ab | 581 | static const char Erlang_help [] = |
4fdb4553 FP |
582 | "In Erlang code, the tags are the functions, records and macros\n\ |
583 | defined in the file."; | |
8dc7496c | 584 | |
988e88ab | 585 | const char *Forth_suffixes [] = |
55102b5d | 586 | { "fth", "tok", NULL }; |
988e88ab | 587 | static const char Forth_help [] = |
55102b5d FP |
588 | "In Forth code, tags are words defined by `:',\n\ |
589 | constant, code, create, defer, value, variable, buffer:, field."; | |
590 | ||
988e88ab | 591 | static const char *Fortran_suffixes [] = |
1f638249 | 592 | { "F", "f", "f90", "for", NULL }; |
988e88ab | 593 | static const char Fortran_help [] = |
4fdb4553 | 594 | "In Fortran code, functions, subroutines and block data are tags."; |
1f638249 | 595 | |
988e88ab | 596 | static const char *HTML_suffixes [] = |
61a1f6fa | 597 | { "htm", "html", "shtml", NULL }; |
988e88ab | 598 | static const char HTML_help [] = |
4fdb4553 FP |
599 | "In HTML input files, the tags are the `title' and the `h1', `h2',\n\ |
600 | `h3' headers. Also, tags are `name=' in anchors and all\n\ | |
601 | occurrences of `id='."; | |
61a1f6fa | 602 | |
988e88ab | 603 | static const char *Lisp_suffixes [] = |
89d57763 | 604 | { "cl", "clisp", "el", "l", "lisp", "LSP", "lsp", "ml", NULL }; |
988e88ab | 605 | static const char Lisp_help [] = |
4fdb4553 FP |
606 | "In Lisp code, any function defined with `defun', any variable\n\ |
607 | defined with `defvar' or `defconst', and in general the first\n\ | |
608 | argument of any expression that starts with `(def' in column zero\n\ | |
1cbaa705 KR |
609 | is a tag.\n\ |
610 | The `--declarations' option tags \"(defvar foo)\" constructs too."; | |
89d57763 | 611 | |
988e88ab | 612 | static const char *Lua_suffixes [] = |
05d9a399 | 613 | { "lua", "LUA", NULL }; |
988e88ab | 614 | static const char Lua_help [] = |
05d9a399 FP |
615 | "In Lua scripts, all functions are tags."; |
616 | ||
988e88ab | 617 | static const char *Makefile_filenames [] = |
89d57763 | 618 | { "Makefile", "makefile", "GNUMakefile", "Makefile.in", "Makefile.am", NULL}; |
988e88ab | 619 | static const char Makefile_help [] = |
4fdb4553 FP |
620 | "In makefiles, targets are tags; additionally, variables are tags\n\ |
621 | unless you specify `--no-globals'."; | |
622 | ||
988e88ab | 623 | static const char *Objc_suffixes [] = |
4fdb4553 FP |
624 | { "lm", /* Objective lex file */ |
625 | "m", /* Objective C file */ | |
626 | NULL }; | |
988e88ab | 627 | static const char Objc_help [] = |
4fdb4553 FP |
628 | "In Objective C code, tags include Objective C definitions for classes,\n\ |
629 | class categories, methods and protocols. Tags for variables and\n\ | |
35a4c758 FP |
630 | functions in classes are named `CLASS::VARIABLE' and `CLASS::FUNCTION'.\n\ |
631 | (Use --help --lang=c --lang=objc --lang=java for full help.)"; | |
1f638249 | 632 | |
988e88ab | 633 | static const char *Pascal_suffixes [] = |
1f638249 | 634 | { "p", "pas", NULL }; |
988e88ab | 635 | static const char Pascal_help [] = |
4fdb4553 FP |
636 | "In Pascal code, the tags are the functions and procedures defined\n\ |
637 | in the file."; | |
35a4c758 | 638 | /* " // this is for working around an Emacs highlighting bug... */ |
1f638249 | 639 | |
988e88ab | 640 | static const char *Perl_suffixes [] = |
1f638249 | 641 | { "pl", "pm", NULL }; |
988e88ab | 642 | static const char *Perl_interpreters [] = |
d8913c1c | 643 | { "perl", "@PERL@", NULL }; |
988e88ab | 644 | static const char Perl_help [] = |
4fdb4553 FP |
645 | "In Perl code, the tags are the packages, subroutines and variables\n\ |
646 | defined by the `package', `sub', `my' and `local' keywords. Use\n\ | |
647 | `--globals' if you want to tag global variables. Tags for\n\ | |
648 | subroutines are named `PACKAGE::SUB'. The name for subroutines\n\ | |
649 | defined in the default package is `main::SUB'."; | |
1f638249 | 650 | |
988e88ab | 651 | static const char *PHP_suffixes [] = |
62aec606 | 652 | { "php", "php3", "php4", NULL }; |
988e88ab | 653 | static const char PHP_help [] = |
78064948 FP |
654 | "In PHP code, tags are functions, classes and defines. Unless you use\n\ |
655 | the `--no-members' option, vars are tags too."; | |
62aec606 | 656 | |
988e88ab | 657 | static const char *plain_C_suffixes [] = |
4fdb4553 | 658 | { "pc", /* Pro*C file */ |
d8913c1c | 659 | NULL }; |
1f638249 | 660 | |
988e88ab | 661 | static const char *PS_suffixes [] = |
93b7ac65 | 662 | { "ps", "psw", NULL }; /* .psw is for PSWrap */ |
988e88ab | 663 | static const char PS_help [] = |
4fdb4553 | 664 | "In PostScript code, the tags are the functions."; |
cec68fcb | 665 | |
988e88ab | 666 | static const char *Prolog_suffixes [] = |
1f638249 | 667 | { "prolog", NULL }; |
988e88ab | 668 | static const char Prolog_help [] = |
4fdb4553 FP |
669 | "In Prolog code, tags are predicates and rules at the beginning of\n\ |
670 | line."; | |
1f638249 | 671 | |
988e88ab | 672 | static const char *Python_suffixes [] = |
93b7ac65 | 673 | { "py", NULL }; |
988e88ab | 674 | static const char Python_help [] = |
4fdb4553 FP |
675 | "In Python code, `def' or `class' at the beginning of a line\n\ |
676 | generate a tag."; | |
93b7ac65 | 677 | |
d8913c1c | 678 | /* Can't do the `SCM' or `scm' prefix with a version number. */ |
988e88ab | 679 | static const char *Scheme_suffixes [] = |
89d57763 | 680 | { "oak", "sch", "scheme", "SCM", "scm", "SM", "sm", "ss", "t", NULL }; |
988e88ab | 681 | static const char Scheme_help [] = |
4fdb4553 FP |
682 | "In Scheme code, tags include anything defined with `def' or with a\n\ |
683 | construct whose name starts with `def'. They also include\n\ | |
684 | variables set with `set!' at top level in the file."; | |
1f638249 | 685 | |
988e88ab | 686 | static const char *TeX_suffixes [] = |
89d57763 | 687 | { "bib", "clo", "cls", "ltx", "sty", "TeX", "tex", NULL }; |
988e88ab | 688 | static const char TeX_help [] = |
4fdb4553 FP |
689 | "In LaTeX text, the argument of any of the commands `\\chapter',\n\ |
690 | `\\section', `\\subsection', `\\subsubsection', `\\eqno', `\\label',\n\ | |
691 | `\\ref', `\\cite', `\\bibitem', `\\part', `\\appendix', `\\entry',\n\ | |
692 | `\\index', `\\def', `\\newcommand', `\\renewcommand',\n\ | |
693 | `\\newenvironment' or `\\renewenvironment' is a tag.\n\ | |
694 | \n\ | |
695 | Other commands can be specified by setting the environment variable\n\ | |
696 | `TEXTAGS' to a colon-separated list like, for example,\n\ | |
697 | TEXTAGS=\"mycommand:myothercommand\"."; | |
698 | ||
1f638249 | 699 | |
988e88ab | 700 | static const char *Texinfo_suffixes [] = |
89d57763 | 701 | { "texi", "texinfo", "txi", NULL }; |
988e88ab | 702 | static const char Texinfo_help [] = |
4fdb4553 | 703 | "for texinfo files, lines starting with @node are tagged."; |
97052c63 | 704 | |
988e88ab | 705 | static const char *Yacc_suffixes [] = |
89d57763 | 706 | { "y", "y++", "ym", "yxx", "yy", NULL }; /* .ym is Objective yacc file */ |
988e88ab | 707 | static const char Yacc_help [] = |
4fdb4553 FP |
708 | "In Bison or Yacc input files, each rule defines as a tag the\n\ |
709 | nonterminal it constructs. The portions of the file that contain\n\ | |
710 | C code are parsed as C code (use --help --lang=c --lang=yacc\n\ | |
711 | for full help)."; | |
712 | ||
988e88ab | 713 | static const char auto_help [] = |
4fdb4553 FP |
714 | "`auto' is not a real language, it indicates to use\n\ |
715 | a default language for files base on file name suffix and file contents."; | |
716 | ||
988e88ab | 717 | static const char none_help [] = |
4fdb4553 FP |
718 | "`none' is not a real language, it indicates to only do\n\ |
719 | regexp processing on files."; | |
720 | ||
988e88ab | 721 | static const char no_lang_help [] = |
4fdb4553 FP |
722 | "No detailed help available for this language."; |
723 | ||
1f638249 | 724 | |
93b7ac65 FP |
725 | /* |
726 | * Table of languages. | |
727 | * | |
728 | * It is ok for a given function to be listed under more than one | |
729 | * name. I just didn't. | |
730 | */ | |
b9755a12 | 731 | |
db590582 | 732 | static language lang_names [] = |
b9755a12 | 733 | { |
4fdb4553 FP |
734 | { "ada", Ada_help, Ada_funcs, Ada_suffixes }, |
735 | { "asm", Asm_help, Asm_labels, Asm_suffixes }, | |
736 | { "c", default_C_help, default_C_entries, default_C_suffixes }, | |
737 | { "c++", Cplusplus_help, Cplusplus_entries, Cplusplus_suffixes }, | |
738 | { "c*", no_lang_help, Cstar_entries, Cstar_suffixes }, | |
739 | { "cobol", Cobol_help, Cobol_paragraphs, Cobol_suffixes }, | |
740 | { "erlang", Erlang_help, Erlang_functions, Erlang_suffixes }, | |
55102b5d | 741 | { "forth", Forth_help, Forth_words, Forth_suffixes }, |
4fdb4553 FP |
742 | { "fortran", Fortran_help, Fortran_functions, Fortran_suffixes }, |
743 | { "html", HTML_help, HTML_labels, HTML_suffixes }, | |
744 | { "java", Cjava_help, Cjava_entries, Cjava_suffixes }, | |
745 | { "lisp", Lisp_help, Lisp_functions, Lisp_suffixes }, | |
05d9a399 | 746 | { "lua", Lua_help, Lua_functions, Lua_suffixes }, |
4fdb4553 FP |
747 | { "makefile", Makefile_help,Makefile_targets,NULL,Makefile_filenames}, |
748 | { "objc", Objc_help, plain_C_entries, Objc_suffixes }, | |
749 | { "pascal", Pascal_help, Pascal_functions, Pascal_suffixes }, | |
750 | { "perl",Perl_help,Perl_functions,Perl_suffixes,NULL,Perl_interpreters}, | |
751 | { "php", PHP_help, PHP_functions, PHP_suffixes }, | |
752 | { "postscript",PS_help, PS_functions, PS_suffixes }, | |
753 | { "proc", no_lang_help, plain_C_entries, plain_C_suffixes }, | |
754 | { "prolog", Prolog_help, Prolog_functions, Prolog_suffixes }, | |
755 | { "python", Python_help, Python_functions, Python_suffixes }, | |
756 | { "scheme", Scheme_help, Scheme_functions, Scheme_suffixes }, | |
757 | { "tex", TeX_help, TeX_commands, TeX_suffixes }, | |
758 | { "texinfo", Texinfo_help, Texinfo_nodes, Texinfo_suffixes }, | |
96c06863 | 759 | { "yacc", Yacc_help,Yacc_entries,Yacc_suffixes,NULL,NULL,true}, |
4fdb4553 FP |
760 | { "auto", auto_help }, /* default guessing scheme */ |
761 | { "none", none_help, just_read_file }, /* regexp matching only */ | |
762 | { NULL } /* end of list */ | |
b9755a12 | 763 | }; |
b2521b0a | 764 | |
c6d46f5f | 765 | \f |
944bdd72 | 766 | static void |
873fbd0b | 767 | print_language_names (void) |
b9755a12 | 768 | { |
93b7ac65 | 769 | language *lang; |
988e88ab | 770 | const char **name, **ext; |
b9755a12 FP |
771 | |
772 | puts ("\nThese are the currently supported languages, along with the\n\ | |
83be933c | 773 | default file names and dot suffixes:"); |
1f638249 | 774 | for (lang = lang_names; lang->name != NULL; lang++) |
b9755a12 | 775 | { |
83be933c FP |
776 | printf (" %-*s", 10, lang->name); |
777 | if (lang->filenames != NULL) | |
778 | for (name = lang->filenames; *name != NULL; name++) | |
779 | printf (" %s", *name); | |
1f638249 FP |
780 | if (lang->suffixes != NULL) |
781 | for (ext = lang->suffixes; *ext != NULL; ext++) | |
782 | printf (" .%s", *ext); | |
b9755a12 FP |
783 | puts (""); |
784 | } | |
4fdb4553 | 785 | puts ("where `auto' means use default language for files based on file\n\ |
79263656 FP |
786 | name suffix, and `none' means only do regexp processing on files.\n\ |
787 | If no language is specified and no matching suffix is found,\n\ | |
1f638249 FP |
788 | the first line of the file is read for a sharp-bang (#!) sequence\n\ |
789 | followed by the name of an interpreter. If no such sequence is found,\n\ | |
93b7ac65 | 790 | Fortran is tried first; if no tags are found, C is tried next.\n\ |
4fdb4553 FP |
791 | When parsing any C file, a \"class\" or \"template\" keyword\n\ |
792 | switches to C++."); | |
4c964351 | 793 | puts ("Compressed files are supported using gzip, bzip2, and xz.\n\ |
4fdb4553 FP |
794 | \n\ |
795 | For detailed help on a given language use, for example,\n\ | |
796 | etags --help --lang=ada."); | |
b9755a12 FP |
797 | } |
798 | ||
944bdd72 | 799 | #ifndef EMACS_NAME |
24dbe96a | 800 | # define EMACS_NAME "standalone" |
944bdd72 | 801 | #endif |
c5007f46 | 802 | #ifndef VERSION |
d3706fa9 | 803 | # define VERSION "17.38.1.4" |
c5007f46 | 804 | #endif |
31ff141c | 805 | static _Noreturn void |
873fbd0b | 806 | print_version (void) |
4746118a | 807 | { |
78f83752 | 808 | char emacs_copyright[] = COPYRIGHT; |
d16c1140 | 809 | |
944bdd72 | 810 | printf ("%s (%s %s)\n", (CTAGS) ? "ctags" : "etags", EMACS_NAME, VERSION); |
d16c1140 | 811 | puts (emacs_copyright); |
b7e1b21b | 812 | puts ("This program is distributed under the terms in ETAGS.README"); |
4746118a | 813 | |
3f0656ff | 814 | exit (EXIT_SUCCESS); |
4746118a JB |
815 | } |
816 | ||
172aa4c1 | 817 | #ifndef PRINT_UNDOCUMENTED_OPTIONS_HELP |
96c06863 | 818 | # define PRINT_UNDOCUMENTED_OPTIONS_HELP false |
172aa4c1 FP |
819 | #endif |
820 | ||
31ff141c | 821 | static _Noreturn void |
873fbd0b | 822 | print_help (argument *argbuffer) |
4746118a | 823 | { |
96c06863 | 824 | bool help_for_lang = false; |
4fdb4553 FP |
825 | |
826 | for (; argbuffer->arg_type != at_end; argbuffer++) | |
827 | if (argbuffer->arg_type == at_language) | |
828 | { | |
829 | if (help_for_lang) | |
830 | puts (""); | |
831 | puts (argbuffer->lang->help); | |
96c06863 | 832 | help_for_lang = true; |
4fdb4553 FP |
833 | } |
834 | ||
835 | if (help_for_lang) | |
3f0656ff | 836 | exit (EXIT_SUCCESS); |
4fdb4553 | 837 | |
30903246 FP |
838 | printf ("Usage: %s [options] [[regex-option ...] file-name] ...\n\ |
839 | \n\ | |
840 | These are the options accepted by %s.\n", progname, progname); | |
9239d970 | 841 | puts ("You may use unambiguous abbreviations for the long option names."); |
9e0a3f98 FP |
842 | puts (" A - as file name means read names from stdin (one per line).\n\ |
843 | Absolute names are stored in the output file as they are.\n\ | |
844 | Relative ones are stored relative to the output file's directory.\n"); | |
4746118a | 845 | |
55102b5d | 846 | puts ("-a, --append\n\ |
52cc7c59 | 847 | Append tag entries to existing tags file."); |
1a0d8c80 | 848 | |
93b7ac65 | 849 | puts ("--packages-only\n\ |
9e0a3f98 | 850 | For Ada files, only generate tags for packages."); |
93b7ac65 | 851 | |
32daa216 FP |
852 | if (CTAGS) |
853 | puts ("-B, --backward-search\n\ | |
1a0d8c80 | 854 | Write the search commands for the tag entries using '?', the\n\ |
3ad2882c | 855 | backward-search command instead of '/', the forward-search command."); |
1a0d8c80 | 856 | |
8c463abe FP |
857 | /* This option is mostly obsolete, because etags can now automatically |
858 | detect C++. Retained for backward compatibility and for debugging and | |
859 | experimentation. In principle, we could want to tag as C++ even | |
4fdb4553 | 860 | before any "class" or "template" keyword. |
52cc7c59 | 861 | puts ("-C, --c++\n\ |
79263656 | 862 | Treat files whose name suffix defaults to C language as C++ files."); |
8c463abe | 863 | */ |
4746118a | 864 | |
93b7ac65 FP |
865 | puts ("--declarations\n\ |
866 | In C and derived languages, create tags for function declarations,"); | |
867 | if (CTAGS) | |
868 | puts ("\tand create tags for extern variables if --globals is used."); | |
869 | else | |
870 | puts | |
871 | ("\tand create tags for extern variables unless --no-globals is used."); | |
872 | ||
32daa216 FP |
873 | if (CTAGS) |
874 | puts ("-d, --defines\n\ | |
3ca80e28 | 875 | Create tag entries for C #define constants and enum constants, too."); |
32daa216 FP |
876 | else |
877 | puts ("-D, --no-defines\n\ | |
3ca80e28 FP |
878 | Don't create tag entries for C #define constants and enum constants.\n\ |
879 | This makes the tags file smaller."); | |
4746118a | 880 | |
32daa216 | 881 | if (!CTAGS) |
9e0a3f98 | 882 | puts ("-i FILE, --include=FILE\n\ |
1a0d8c80 FP |
883 | Include a note in tag file indicating that, when searching for\n\ |
884 | a tag, one should also consult the tags file FILE after\n\ | |
885 | checking the current file."); | |
9e0a3f98 FP |
886 | |
887 | puts ("-l LANG, --language=LANG\n\ | |
b9755a12 FP |
888 | Force the following files to be considered as written in the\n\ |
889 | named language up to the next --language=LANG option."); | |
7537186d | 890 | |
30903246 FP |
891 | if (CTAGS) |
892 | puts ("--globals\n\ | |
93b7ac65 | 893 | Create tag entries for global variables in some languages."); |
30903246 FP |
894 | else |
895 | puts ("--no-globals\n\ | |
93b7ac65 FP |
896 | Do not create tag entries for global variables in some\n\ |
897 | languages. This makes the tags file smaller."); | |
172aa4c1 FP |
898 | |
899 | if (PRINT_UNDOCUMENTED_OPTIONS_HELP) | |
900 | puts ("--no-line-directive\n\ | |
901 | Ignore #line preprocessor directives in C and derived languages."); | |
902 | ||
eac50785 FP |
903 | if (CTAGS) |
904 | puts ("--members\n\ | |
905 | Create tag entries for members of structures in some languages."); | |
906 | else | |
907 | puts ("--no-members\n\ | |
78064948 FP |
908 | Do not create tag entries for members of structures\n\ |
909 | in some languages."); | |
30903246 | 910 | |
89fb2be1 | 911 | puts ("-r REGEXP, --regex=REGEXP or --regex=@regexfile\n\ |
3c04a71a | 912 | Make a tag for each line matching a regular expression pattern\n\ |
89fb2be1 FP |
913 | in the following files. {LANGUAGE}REGEXP uses REGEXP for LANGUAGE\n\ |
914 | files only. REGEXFILE is a file containing one REGEXP per line.\n\ | |
915 | REGEXP takes the form /TAGREGEXP/TAGNAME/MODS, where TAGNAME/ is\n\ | |
916 | optional. The TAGREGEXP pattern is anchored (as if preceded by ^)."); | |
917 | puts (" If TAGNAME/ is present, the tags created are named.\n\ | |
93b7ac65 | 918 | For example Tcl named tags can be created with:\n\ |
89fb2be1 FP |
919 | --regex=\"/proc[ \\t]+\\([^ \\t]+\\)/\\1/.\".\n\ |
920 | MODS are optional one-letter modifiers: `i' means to ignore case,\n\ | |
921 | `m' means to allow multi-line matches, `s' implies `m' and\n\ | |
3c04a71a | 922 | causes dot to match any character, including newline."); |
172aa4c1 | 923 | |
7537186d | 924 | puts ("-R, --no-regex\n\ |
b9755a12 | 925 | Don't create tags from regexps for the following files."); |
172aa4c1 | 926 | |
201f9f2b | 927 | puts ("-I, --ignore-indentation\n\ |
3c04a71a FP |
928 | In C and C++ do not assume that a closing brace in the first\n\ |
929 | column is the final brace of a function or structure definition."); | |
172aa4c1 | 930 | |
e7d3b099 FP |
931 | puts ("-o FILE, --output=FILE\n\ |
932 | Write the tags to FILE."); | |
172aa4c1 | 933 | |
e7d3b099 FP |
934 | puts ("--parse-stdin=NAME\n\ |
935 | Read from standard input and record tags as belonging to file NAME."); | |
4746118a | 936 | |
32daa216 FP |
937 | if (CTAGS) |
938 | { | |
939 | puts ("-t, --typedefs\n\ | |
93b7ac65 | 940 | Generate tag entries for C and Ada typedefs."); |
32daa216 FP |
941 | puts ("-T, --typedefs-and-c++\n\ |
942 | Generate tag entries for C typedefs, C struct/enum/union tags,\n\ | |
943 | and C++ member functions."); | |
9e0a3f98 FP |
944 | } |
945 | ||
946 | if (CTAGS) | |
947 | puts ("-u, --update\n\ | |
4746118a JB |
948 | Update the tag entries for the given files, leaving tag\n\ |
949 | entries for other files in place. Currently, this is\n\ | |
950 | implemented by deleting the existing entries for the given\n\ | |
951 | files and then rewriting the new entries at the end of the\n\ | |
952 | tags file. It is often faster to simply rebuild the entire\n\ | |
52cc7c59 | 953 | tag file than to use this."); |
9e0a3f98 FP |
954 | |
955 | if (CTAGS) | |
956 | { | |
32daa216 | 957 | puts ("-v, --vgrind\n\ |
2e0bea68 FP |
958 | Print on the standard output an index of items intended for\n\ |
959 | human consumption, similar to the output of vgrind. The index\n\ | |
960 | is sorted, and gives the page number of each item."); | |
172aa4c1 FP |
961 | |
962 | if (PRINT_UNDOCUMENTED_OPTIONS_HELP) | |
963 | puts ("-w, --no-duplicates\n\ | |
ed8bbc0e FP |
964 | Do not create duplicate tag entries, for compatibility with\n\ |
965 | traditional ctags."); | |
172aa4c1 FP |
966 | |
967 | if (PRINT_UNDOCUMENTED_OPTIONS_HELP) | |
968 | puts ("-w, --no-warn\n\ | |
ed8bbc0e | 969 | Suppress warning messages about duplicate tag entries."); |
172aa4c1 | 970 | |
32daa216 | 971 | puts ("-x, --cxref\n\ |
4746118a JB |
972 | Like --vgrind, but in the style of cxref, rather than vgrind.\n\ |
973 | The output uses line numbers instead of page numbers, but\n\ | |
974 | beyond that the differences are cosmetic; try both to see\n\ | |
52cc7c59 | 975 | which you like."); |
32daa216 | 976 | } |
4746118a JB |
977 | |
978 | puts ("-V, --version\n\ | |
979 | Print the version of the program.\n\ | |
7537186d | 980 | -h, --help\n\ |
4fdb4553 FP |
981 | Print this help message.\n\ |
982 | Followed by one or more `--language' options prints detailed\n\ | |
983 | help about tag generation for the specified languages."); | |
4746118a | 984 | |
b9755a12 FP |
985 | print_language_names (); |
986 | ||
b6b48846 | 987 | puts (""); |
a5c7ef8b | 988 | puts ("Report bugs to bug-gnu-emacs@gnu.org"); |
b6b48846 | 989 | |
3f0656ff | 990 | exit (EXIT_SUCCESS); |
4746118a JB |
991 | } |
992 | ||
993 | \f | |
d0dff6e5 | 994 | int |
873fbd0b | 995 | main (int argc, char **argv) |
c6d46f5f | 996 | { |
c6d46f5f | 997 | int i; |
2f608d34 FP |
998 | unsigned int nincluded_files; |
999 | char **included_files; | |
1f638249 | 1000 | argument *argbuffer; |
2f608d34 | 1001 | int current_arg, file_count; |
93b7ac65 | 1002 | linebuffer filename_lb; |
96c06863 | 1003 | bool help_asked = false; |
9250f758 | 1004 | ptrdiff_t len; |
97b90b0a FP |
1005 | char *optstring; |
1006 | int opt; | |
1007 | ||
c5007f46 | 1008 | |
c05b6df5 | 1009 | #ifdef DOS_NT |
42680d3c | 1010 | _fmode = O_BINARY; /* all of files are treated as binary files */ |
c05b6df5 | 1011 | #endif /* DOS_NT */ |
c6880c90 | 1012 | |
c6d46f5f | 1013 | progname = argv[0]; |
2f608d34 FP |
1014 | nincluded_files = 0; |
1015 | included_files = xnew (argc, char *); | |
1016 | current_arg = 0; | |
1017 | file_count = 0; | |
c6d46f5f | 1018 | |
b9755a12 FP |
1019 | /* Allocate enough no matter what happens. Overkill, but each one |
1020 | is small. */ | |
1f638249 | 1021 | argbuffer = xnew (argc, argument); |
b9755a12 | 1022 | |
c6d46f5f | 1023 | /* |
89d8309f | 1024 | * Always find typedefs and structure tags. |
78064948 | 1025 | * Also default to find macro constants, enum constants, struct |
89d8309f | 1026 | * members and global variables. Do it for both etags and ctags. |
c6d46f5f | 1027 | */ |
96c06863 PE |
1028 | typedefs = typedefs_or_cplusplus = constantypedefs = true; |
1029 | globals = members = true; | |
c6d46f5f | 1030 | |
55102b5d FP |
1031 | /* When the optstring begins with a '-' getopt_long does not rearrange the |
1032 | non-options arguments to be at the end, but leaves them alone. */ | |
9239d970 PE |
1033 | optstring = concat ("-ac:Cf:Il:o:r:RSVhH", |
1034 | (CTAGS) ? "BxdtTuvw" : "Di:", | |
1035 | ""); | |
4746118a | 1036 | |
55102b5d | 1037 | while ((opt = getopt_long (argc, argv, optstring, longopts, NULL)) != EOF) |
97b90b0a FP |
1038 | switch (opt) |
1039 | { | |
1040 | case 0: | |
1041 | /* If getopt returns 0, then it has already processed a | |
1042 | long-named option. We should do nothing. */ | |
4746118a JB |
1043 | break; |
1044 | ||
97b90b0a FP |
1045 | case 1: |
1046 | /* This means that a file name has been seen. Record it. */ | |
1047 | argbuffer[current_arg].arg_type = at_filename; | |
1048 | argbuffer[current_arg].what = optarg; | |
9250f758 PE |
1049 | len = strlen (optarg); |
1050 | if (whatlen_max < len) | |
1051 | whatlen_max = len; | |
97b90b0a FP |
1052 | ++current_arg; |
1053 | ++file_count; | |
1054 | break; | |
4746118a | 1055 | |
97b90b0a FP |
1056 | case STDIN: |
1057 | /* Parse standard input. Idea by Vivek <vivek@etla.org>. */ | |
1058 | argbuffer[current_arg].arg_type = at_stdin; | |
1059 | argbuffer[current_arg].what = optarg; | |
9250f758 PE |
1060 | len = strlen (optarg); |
1061 | if (whatlen_max < len) | |
1062 | whatlen_max = len; | |
97b90b0a FP |
1063 | ++current_arg; |
1064 | ++file_count; | |
e7d3b099 FP |
1065 | if (parsing_stdin) |
1066 | fatal ("cannot parse standard input more than once", (char *)NULL); | |
96c06863 | 1067 | parsing_stdin = true; |
97b90b0a | 1068 | break; |
b9755a12 | 1069 | |
97b90b0a | 1070 | /* Common options. */ |
96c06863 PE |
1071 | case 'a': append_to_tagfile = true; break; |
1072 | case 'C': cplusplus = true; break; | |
97b90b0a FP |
1073 | case 'f': /* for compatibility with old makefiles */ |
1074 | case 'o': | |
1075 | if (tagfile) | |
93b7ac65 | 1076 | { |
db5a3003 | 1077 | error ("-o option may only be given once."); |
97b90b0a | 1078 | suggest_asking_for_help (); |
4fdb4553 | 1079 | /* NOTREACHED */ |
93b7ac65 | 1080 | } |
97b90b0a FP |
1081 | tagfile = optarg; |
1082 | break; | |
1083 | case 'I': | |
1084 | case 'S': /* for backward compatibility */ | |
96c06863 | 1085 | ignoreindent = true; |
97b90b0a FP |
1086 | break; |
1087 | case 'l': | |
1088 | { | |
1089 | language *lang = get_language_from_langname (optarg); | |
1090 | if (lang != NULL) | |
1091 | { | |
1092 | argbuffer[current_arg].lang = lang; | |
1093 | argbuffer[current_arg].arg_type = at_language; | |
1094 | ++current_arg; | |
1095 | } | |
c6d46f5f | 1096 | } |
97b90b0a | 1097 | break; |
24dbe96a FP |
1098 | case 'c': |
1099 | /* Backward compatibility: support obsolete --ignore-case-regexp. */ | |
1100 | optarg = concat (optarg, "i", ""); /* memory leak here */ | |
1101 | /* FALLTHRU */ | |
97b90b0a FP |
1102 | case 'r': |
1103 | argbuffer[current_arg].arg_type = at_regexp; | |
1104 | argbuffer[current_arg].what = optarg; | |
9250f758 PE |
1105 | len = strlen (optarg); |
1106 | if (whatlen_max < len) | |
1107 | whatlen_max = len; | |
97b90b0a FP |
1108 | ++current_arg; |
1109 | break; | |
1110 | case 'R': | |
1111 | argbuffer[current_arg].arg_type = at_regexp; | |
1112 | argbuffer[current_arg].what = NULL; | |
1113 | ++current_arg; | |
1114 | break; | |
97b90b0a FP |
1115 | case 'V': |
1116 | print_version (); | |
1117 | break; | |
1118 | case 'h': | |
1119 | case 'H': | |
96c06863 | 1120 | help_asked = true; |
97b90b0a FP |
1121 | break; |
1122 | ||
1123 | /* Etags options */ | |
96c06863 | 1124 | case 'D': constantypedefs = false; break; |
97b90b0a FP |
1125 | case 'i': included_files[nincluded_files++] = optarg; break; |
1126 | ||
1127 | /* Ctags options. */ | |
1128 | case 'B': searchar = '?'; break; | |
96c06863 PE |
1129 | case 'd': constantypedefs = true; break; |
1130 | case 't': typedefs = true; break; | |
1131 | case 'T': typedefs = typedefs_or_cplusplus = true; break; | |
1132 | case 'u': update = true; break; | |
1133 | case 'v': vgrind_style = true; /*FALLTHRU*/ | |
1134 | case 'x': cxref_style = true; break; | |
1135 | case 'w': no_warnings = true; break; | |
97b90b0a FP |
1136 | default: |
1137 | suggest_asking_for_help (); | |
4fdb4553 | 1138 | /* NOTREACHED */ |
97b90b0a | 1139 | } |
c6d46f5f | 1140 | |
55102b5d | 1141 | /* No more options. Store the rest of arguments. */ |
4fdb4553 | 1142 | for (; optind < argc; optind++) |
b9755a12 FP |
1143 | { |
1144 | argbuffer[current_arg].arg_type = at_filename; | |
1145 | argbuffer[current_arg].what = argv[optind]; | |
9250f758 PE |
1146 | len = strlen (argv[optind]); |
1147 | if (whatlen_max < len) | |
1148 | whatlen_max = len; | |
b9755a12 FP |
1149 | ++current_arg; |
1150 | ++file_count; | |
1151 | } | |
1152 | ||
4fdb4553 FP |
1153 | argbuffer[current_arg].arg_type = at_end; |
1154 | ||
1155 | if (help_asked) | |
1156 | print_help (argbuffer); | |
1157 | /* NOTREACHED */ | |
1158 | ||
b9755a12 | 1159 | if (nincluded_files == 0 && file_count == 0) |
c6d46f5f | 1160 | { |
db5a3003 | 1161 | error ("no input files specified."); |
d8913c1c | 1162 | suggest_asking_for_help (); |
4fdb4553 | 1163 | /* NOTREACHED */ |
c6d46f5f JB |
1164 | } |
1165 | ||
6dd5561c | 1166 | if (tagfile == NULL) |
d3706fa9 | 1167 | tagfile = savestr (CTAGS ? "tags" : "TAGS"); |
b02c5fea | 1168 | cwd = etags_getcwd (); /* the current working directory */ |
f55aa55a | 1169 | if (cwd[strlen (cwd) - 1] != '/') |
cec68fcb FP |
1170 | { |
1171 | char *oldcwd = cwd; | |
1172 | cwd = concat (oldcwd, "/", ""); | |
1173 | free (oldcwd); | |
1174 | } | |
d3706fa9 FP |
1175 | |
1176 | /* Compute base directory for relative file names. */ | |
91702d63 FP |
1177 | if (streq (tagfile, "-") |
1178 | || strneq (tagfile, "/dev/", 5)) | |
d3706fa9 | 1179 | tagfiledir = cwd; /* relative file names are relative to cwd */ |
46c145db | 1180 | else |
9817cf3f FP |
1181 | { |
1182 | canonicalize_filename (tagfile); | |
1183 | tagfiledir = absolute_dirname (tagfile, cwd); | |
1184 | } | |
c6d46f5f | 1185 | |
b9755a12 | 1186 | init (); /* set up boolean "functions" */ |
c6d46f5f | 1187 | |
61a1f6fa FP |
1188 | linebuffer_init (&lb); |
1189 | linebuffer_init (&filename_lb); | |
1190 | linebuffer_init (&filebuf); | |
1191 | linebuffer_init (&token_name); | |
b9755a12 | 1192 | |
32daa216 | 1193 | if (!CTAGS) |
c6d46f5f | 1194 | { |
6dd5561c | 1195 | if (streq (tagfile, "-")) |
ce7f6d62 RS |
1196 | { |
1197 | tagf = stdout; | |
1198 | #ifdef DOS_NT | |
1199 | /* Switch redirected `stdout' to binary mode (setting `_fmode' | |
cf347d3c | 1200 | doesn't take effect until after `stdout' is already open). */ |
ce7f6d62 RS |
1201 | if (!isatty (fileno (stdout))) |
1202 | setmode (fileno (stdout), O_BINARY); | |
1203 | #endif /* DOS_NT */ | |
1204 | } | |
c6d46f5f | 1205 | else |
6dd5561c FP |
1206 | tagf = fopen (tagfile, append_to_tagfile ? "a" : "w"); |
1207 | if (tagf == NULL) | |
cdc1f6a7 | 1208 | pfatal (tagfile); |
c6d46f5f JB |
1209 | } |
1210 | ||
b9755a12 FP |
1211 | /* |
1212 | * Loop through files finding functions. | |
1213 | */ | |
4fdb4553 | 1214 | for (i = 0; i < current_arg; i++) |
c6d46f5f | 1215 | { |
9e0a3f98 FP |
1216 | static language *lang; /* non-NULL if language is forced */ |
1217 | char *this_file; | |
1218 | ||
b9755a12 | 1219 | switch (argbuffer[i].arg_type) |
c6d46f5f | 1220 | { |
b9755a12 | 1221 | case at_language: |
9e0a3f98 | 1222 | lang = argbuffer[i].lang; |
b9755a12 | 1223 | break; |
b9755a12 | 1224 | case at_regexp: |
24dbe96a | 1225 | analyse_regex (argbuffer[i].what); |
b9755a12 | 1226 | break; |
b9755a12 | 1227 | case at_filename: |
b9755a12 | 1228 | this_file = argbuffer[i].what; |
b9755a12 | 1229 | /* Input file named "-" means read file names from stdin |
93b7ac65 | 1230 | (one per line) and use them. */ |
b9755a12 | 1231 | if (streq (this_file, "-")) |
97b90b0a FP |
1232 | { |
1233 | if (parsing_stdin) | |
1234 | fatal ("cannot parse standard input AND read file names from it", | |
1235 | (char *)NULL); | |
1236 | while (readline_internal (&filename_lb, stdin) > 0) | |
1237 | process_file_name (filename_lb.buffer, lang); | |
1238 | } | |
b9755a12 | 1239 | else |
97b90b0a | 1240 | process_file_name (this_file, lang); |
b9755a12 | 1241 | break; |
97b90b0a FP |
1242 | case at_stdin: |
1243 | this_file = argbuffer[i].what; | |
1244 | process_file (stdin, this_file, lang); | |
1245 | break; | |
c6d46f5f | 1246 | } |
46c145db | 1247 | } |
9cb0aa73 | 1248 | |
61a1f6fa | 1249 | free_regexps (); |
61a1f6fa | 1250 | free (lb.buffer); |
24dbe96a | 1251 | free (filebuf.buffer); |
61a1f6fa | 1252 | free (token_name.buffer); |
93b7ac65 | 1253 | |
db590582 | 1254 | if (!CTAGS || cxref_style) |
c6d46f5f | 1255 | { |
2e0bea68 FP |
1256 | /* Write the remaining tags to tagf (ETAGS) or stdout (CXREF). */ |
1257 | put_entries (nodehead); | |
9e0a3f98 FP |
1258 | free_tree (nodehead); |
1259 | nodehead = NULL; | |
db590582 | 1260 | if (!CTAGS) |
61a1f6fa FP |
1261 | { |
1262 | fdesc *fdp; | |
1263 | ||
1264 | /* Output file entries that have no tags. */ | |
1265 | for (fdp = fdhead; fdp != NULL; fdp = fdp->next) | |
1266 | if (!fdp->written) | |
1267 | fprintf (tagf, "\f\n%s,0\n", fdp->taggedfname); | |
1268 | ||
1269 | while (nincluded_files-- > 0) | |
1270 | fprintf (tagf, "\f\n%s,include\n", *included_files++); | |
2e0bea68 FP |
1271 | |
1272 | if (fclose (tagf) == EOF) | |
1273 | pfatal (tagfile); | |
61a1f6fa | 1274 | } |
db590582 | 1275 | |
3f0656ff | 1276 | exit (EXIT_SUCCESS); |
c6d46f5f | 1277 | } |
55597f90 | 1278 | |
89d8309f | 1279 | /* From here on, we are in (CTAGS && !cxref_style) */ |
4746118a | 1280 | if (update) |
c6d46f5f | 1281 | { |
9250f758 PE |
1282 | char *cmd = |
1283 | xmalloc (strlen (tagfile) + whatlen_max + | |
1284 | sizeof "mv..OTAGS;fgrep -v '\t\t' OTAGS >;rm OTAGS"); | |
b9755a12 | 1285 | for (i = 0; i < current_arg; ++i) |
c6d46f5f | 1286 | { |
97b90b0a FP |
1287 | switch (argbuffer[i].arg_type) |
1288 | { | |
1289 | case at_filename: | |
1290 | case at_stdin: | |
1291 | break; | |
1292 | default: | |
1293 | continue; /* the for loop */ | |
1294 | } | |
9250f758 PE |
1295 | strcpy (cmd, "mv "); |
1296 | strcat (cmd, tagfile); | |
1297 | strcat (cmd, " OTAGS;fgrep -v '\t"); | |
1298 | strcat (cmd, argbuffer[i].what); | |
1299 | strcat (cmd, "\t' OTAGS >"); | |
1300 | strcat (cmd, tagfile); | |
1301 | strcat (cmd, ";rm OTAGS"); | |
3f0656ff | 1302 | if (system (cmd) != EXIT_SUCCESS) |
086eac13 | 1303 | fatal ("failed to execute shell command", (char *)NULL); |
c6d46f5f | 1304 | } |
9250f758 | 1305 | free (cmd); |
96c06863 | 1306 | append_to_tagfile = true; |
c6d46f5f | 1307 | } |
55597f90 | 1308 | |
6dd5561c FP |
1309 | tagf = fopen (tagfile, append_to_tagfile ? "a" : "w"); |
1310 | if (tagf == NULL) | |
55597f90 | 1311 | pfatal (tagfile); |
89fb2be1 | 1312 | put_entries (nodehead); /* write all the tags (CTAGS) */ |
9e0a3f98 FP |
1313 | free_tree (nodehead); |
1314 | nodehead = NULL; | |
db590582 FP |
1315 | if (fclose (tagf) == EOF) |
1316 | pfatal (tagfile); | |
55597f90 | 1317 | |
b9509712 FP |
1318 | if (CTAGS) |
1319 | if (append_to_tagfile || update) | |
1320 | { | |
9250f758 | 1321 | char *cmd = xmalloc (2 * strlen (tagfile) + sizeof "sort -u -o.."); |
660222c7 FP |
1322 | /* Maybe these should be used: |
1323 | setenv ("LC_COLLATE", "C", 1); | |
1324 | setenv ("LC_ALL", "C", 1); */ | |
9250f758 PE |
1325 | strcpy (cmd, "sort -u -o "); |
1326 | strcat (cmd, tagfile); | |
1327 | strcat (cmd, " "); | |
1328 | strcat (cmd, tagfile); | |
b9509712 FP |
1329 | exit (system (cmd)); |
1330 | } | |
3f0656ff | 1331 | return EXIT_SUCCESS; |
c6d46f5f JB |
1332 | } |
1333 | ||
1334 | ||
93b7ac65 FP |
1335 | /* |
1336 | * Return a compressor given the file name. If EXTPTR is non-zero, | |
1337 | * return a pointer into FILE where the compressor-specific | |
1338 | * extension begins. If no compressor is found, NULL is returned | |
1339 | * and EXTPTR is not significant. | |
b2521b0a | 1340 | * Idea by Vladimir Alexiev <vladimir@cs.ualberta.ca> (1998) |
93b7ac65 | 1341 | */ |
944bdd72 | 1342 | static compressor * |
873fbd0b | 1343 | get_compressor_from_suffix (char *file, char **extptr) |
93b7ac65 FP |
1344 | { |
1345 | compressor *compr; | |
1346 | char *slash, *suffix; | |
1347 | ||
9817cf3f | 1348 | /* File has been processed by canonicalize_filename, |
93b7ac65 FP |
1349 | so we don't need to consider backslashes on DOS_NT. */ |
1350 | slash = etags_strrchr (file, '/'); | |
1351 | suffix = etags_strrchr (file, '.'); | |
1352 | if (suffix == NULL || suffix < slash) | |
1353 | return NULL; | |
1354 | if (extptr != NULL) | |
1355 | *extptr = suffix; | |
1356 | suffix += 1; | |
1357 | /* Let those poor souls who live with DOS 8+3 file name limits get | |
1358 | some solace by treating foo.cgz as if it were foo.c.gz, etc. | |
1359 | Only the first do loop is run if not MSDOS */ | |
1360 | do | |
1361 | { | |
1362 | for (compr = compressors; compr->suffix != NULL; compr++) | |
1363 | if (streq (compr->suffix, suffix)) | |
1364 | return compr; | |
944bdd72 FP |
1365 | if (!MSDOS) |
1366 | break; /* do it only once: not really a loop */ | |
93b7ac65 FP |
1367 | if (extptr != NULL) |
1368 | *extptr = ++suffix; | |
1369 | } while (*suffix != '\0'); | |
1370 | return NULL; | |
1371 | } | |
1372 | ||
1373 | ||
1374 | ||
b9755a12 | 1375 | /* |
93b7ac65 | 1376 | * Return a language given the name. |
b9755a12 | 1377 | */ |
944bdd72 | 1378 | static language * |
873fbd0b | 1379 | get_language_from_langname (const char *name) |
b9755a12 | 1380 | { |
93b7ac65 | 1381 | language *lang; |
b9755a12 | 1382 | |
93b7ac65 | 1383 | if (name == NULL) |
db5a3003 | 1384 | error ("empty language name"); |
93b7ac65 FP |
1385 | else |
1386 | { | |
1387 | for (lang = lang_names; lang->name != NULL; lang++) | |
d8913c1c | 1388 | if (streq (name, lang->name)) |
93b7ac65 FP |
1389 | return lang; |
1390 | error ("unknown language \"%s\"", name); | |
1391 | } | |
d8913c1c | 1392 | |
93b7ac65 | 1393 | return NULL; |
1f638249 FP |
1394 | } |
1395 | ||
1396 | ||
1397 | /* | |
93b7ac65 | 1398 | * Return a language given the interpreter name. |
1f638249 | 1399 | */ |
944bdd72 | 1400 | static language * |
873fbd0b | 1401 | get_language_from_interpreter (char *interpreter) |
1f638249 | 1402 | { |
93b7ac65 | 1403 | language *lang; |
988e88ab | 1404 | const char **iname; |
1f638249 FP |
1405 | |
1406 | if (interpreter == NULL) | |
1407 | return NULL; | |
1408 | for (lang = lang_names; lang->name != NULL; lang++) | |
1409 | if (lang->interpreters != NULL) | |
1410 | for (iname = lang->interpreters; *iname != NULL; iname++) | |
1411 | if (streq (*iname, interpreter)) | |
93b7ac65 | 1412 | return lang; |
1f638249 FP |
1413 | |
1414 | return NULL; | |
1415 | } | |
1416 | ||
1417 | ||
1418 | ||
1419 | /* | |
93b7ac65 | 1420 | * Return a language given the file name. |
1f638249 | 1421 | */ |
944bdd72 | 1422 | static language * |
873fbd0b | 1423 | get_language_from_filename (char *file, int case_sensitive) |
1f638249 | 1424 | { |
93b7ac65 | 1425 | language *lang; |
988e88ab | 1426 | const char **name, **ext, *suffix; |
89d57763 FP |
1427 | |
1428 | /* Try whole file name first. */ | |
1429 | for (lang = lang_names; lang->name != NULL; lang++) | |
1430 | if (lang->filenames != NULL) | |
1431 | for (name = lang->filenames; *name != NULL; name++) | |
a5363890 FP |
1432 | if ((case_sensitive) |
1433 | ? streq (*name, file) | |
1434 | : strcaseeq (*name, file)) | |
89d57763 | 1435 | return lang; |
1f638249 | 1436 | |
89d57763 | 1437 | /* If not found, try suffix after last dot. */ |
93b7ac65 | 1438 | suffix = etags_strrchr (file, '.'); |
1f638249 FP |
1439 | if (suffix == NULL) |
1440 | return NULL; | |
93b7ac65 | 1441 | suffix += 1; |
1f638249 FP |
1442 | for (lang = lang_names; lang->name != NULL; lang++) |
1443 | if (lang->suffixes != NULL) | |
1444 | for (ext = lang->suffixes; *ext != NULL; ext++) | |
a5363890 FP |
1445 | if ((case_sensitive) |
1446 | ? streq (*ext, suffix) | |
1447 | : strcaseeq (*ext, suffix)) | |
93b7ac65 | 1448 | return lang; |
1f638249 | 1449 | return NULL; |
b9755a12 FP |
1450 | } |
1451 | ||
9e0a3f98 | 1452 | \f |
c6d46f5f JB |
1453 | /* |
1454 | * This routine is called on each file argument. | |
1455 | */ | |
944bdd72 | 1456 | static void |
873fbd0b | 1457 | process_file_name (char *file, language *lang) |
c6d46f5f JB |
1458 | { |
1459 | struct stat stat_buf; | |
55597f90 | 1460 | FILE *inf; |
9e0a3f98 | 1461 | fdesc *fdp; |
93b7ac65 FP |
1462 | compressor *compr; |
1463 | char *compressed_name, *uncompressed_name; | |
1464 | char *ext, *real_name; | |
db590582 | 1465 | int retval; |
ce7f6d62 | 1466 | |
93b7ac65 | 1467 | canonicalize_filename (file); |
6dd5561c | 1468 | if (streq (file, tagfile) && !streq (tagfile, "-")) |
c6d46f5f | 1469 | { |
30903246 | 1470 | error ("skipping inclusion of %s in self.", file); |
c6d46f5f JB |
1471 | return; |
1472 | } | |
93b7ac65 FP |
1473 | if ((compr = get_compressor_from_suffix (file, &ext)) == NULL) |
1474 | { | |
1475 | compressed_name = NULL; | |
1476 | real_name = uncompressed_name = savestr (file); | |
1477 | } | |
1478 | else | |
1479 | { | |
1480 | real_name = compressed_name = savestr (file); | |
1481 | uncompressed_name = savenstr (file, ext - file); | |
1482 | } | |
1483 | ||
9e0a3f98 FP |
1484 | /* If the canonicalized uncompressed name |
1485 | has already been dealt with, skip it silently. */ | |
1486 | for (fdp = fdhead; fdp != NULL; fdp = fdp->next) | |
93b7ac65 | 1487 | { |
9e0a3f98 FP |
1488 | assert (fdp->infname != NULL); |
1489 | if (streq (uncompressed_name, fdp->infname)) | |
1490 | goto cleanup; | |
1491 | } | |
1492 | ||
93b7ac65 FP |
1493 | if (stat (real_name, &stat_buf) != 0) |
1494 | { | |
1495 | /* Reset real_name and try with a different name. */ | |
1496 | real_name = NULL; | |
1497 | if (compressed_name != NULL) /* try with the given suffix */ | |
1498 | { | |
1499 | if (stat (uncompressed_name, &stat_buf) == 0) | |
1500 | real_name = uncompressed_name; | |
1501 | } | |
1502 | else /* try all possible suffixes */ | |
1503 | { | |
1504 | for (compr = compressors; compr->suffix != NULL; compr++) | |
1505 | { | |
1506 | compressed_name = concat (file, ".", compr->suffix); | |
1507 | if (stat (compressed_name, &stat_buf) != 0) | |
1508 | { | |
944bdd72 | 1509 | if (MSDOS) |
93b7ac65 | 1510 | { |
944bdd72 FP |
1511 | char *suf = compressed_name + strlen (file); |
1512 | size_t suflen = strlen (compr->suffix) + 1; | |
1513 | for ( ; suf[1]; suf++, suflen--) | |
93b7ac65 | 1514 | { |
944bdd72 FP |
1515 | memmove (suf, suf + 1, suflen); |
1516 | if (stat (compressed_name, &stat_buf) == 0) | |
1517 | { | |
1518 | real_name = compressed_name; | |
1519 | break; | |
1520 | } | |
93b7ac65 | 1521 | } |
944bdd72 FP |
1522 | if (real_name != NULL) |
1523 | break; | |
1524 | } /* MSDOS */ | |
93b7ac65 FP |
1525 | free (compressed_name); |
1526 | compressed_name = NULL; | |
1527 | } | |
1528 | else | |
1529 | { | |
1530 | real_name = compressed_name; | |
1531 | break; | |
1532 | } | |
1533 | } | |
1534 | } | |
1535 | if (real_name == NULL) | |
1536 | { | |
1537 | perror (file); | |
9e0a3f98 | 1538 | goto cleanup; |
93b7ac65 FP |
1539 | } |
1540 | } /* try with a different name */ | |
1541 | ||
1542 | if (!S_ISREG (stat_buf.st_mode)) | |
1543 | { | |
1544 | error ("skipping %s: it is not a regular file.", real_name); | |
9e0a3f98 | 1545 | goto cleanup; |
93b7ac65 FP |
1546 | } |
1547 | if (real_name == compressed_name) | |
1548 | { | |
1549 | char *cmd = concat (compr->command, " ", real_name); | |
8c463abe | 1550 | inf = (FILE *) popen (cmd, "r"); |
93b7ac65 FP |
1551 | free (cmd); |
1552 | } | |
1553 | else | |
1554 | inf = fopen (real_name, "r"); | |
55597f90 | 1555 | if (inf == NULL) |
42680d3c | 1556 | { |
93b7ac65 | 1557 | perror (real_name); |
9e0a3f98 | 1558 | goto cleanup; |
42680d3c | 1559 | } |
55597f90 | 1560 | |
97b90b0a FP |
1561 | process_file (inf, uncompressed_name, lang); |
1562 | ||
1563 | if (real_name == compressed_name) | |
1564 | retval = pclose (inf); | |
1565 | else | |
1566 | retval = fclose (inf); | |
1567 | if (retval < 0) | |
1568 | pfatal (file); | |
1569 | ||
1570 | cleanup: | |
c2cd06e6 JM |
1571 | free (compressed_name); |
1572 | free (uncompressed_name); | |
97b90b0a FP |
1573 | last_node = NULL; |
1574 | curfdp = NULL; | |
1575 | return; | |
1576 | } | |
1577 | ||
1578 | static void | |
873fbd0b | 1579 | process_file (FILE *fh, char *fn, language *lang) |
97b90b0a FP |
1580 | { |
1581 | static const fdesc emptyfdesc; | |
1582 | fdesc *fdp; | |
1583 | ||
a5363890 FP |
1584 | /* Create a new input file description entry. */ |
1585 | fdp = xnew (1, fdesc); | |
1586 | *fdp = emptyfdesc; | |
1587 | fdp->next = fdhead; | |
97b90b0a | 1588 | fdp->infname = savestr (fn); |
a5363890 | 1589 | fdp->lang = lang; |
97b90b0a FP |
1590 | fdp->infabsname = absolute_filename (fn, cwd); |
1591 | fdp->infabsdir = absolute_dirname (fn, cwd); | |
1592 | if (filename_is_absolute (fn)) | |
db590582 | 1593 | { |
97b90b0a FP |
1594 | /* An absolute file name. Canonicalize it. */ |
1595 | fdp->taggedfname = absolute_filename (fn, NULL); | |
db590582 FP |
1596 | } |
1597 | else | |
1598 | { | |
97b90b0a | 1599 | /* A file name relative to cwd. Make it relative |
db590582 | 1600 | to the directory of the tags file. */ |
97b90b0a | 1601 | fdp->taggedfname = relative_filename (fn, tagfiledir); |
db590582 | 1602 | } |
96c06863 | 1603 | fdp->usecharno = true; /* use char position when making tags */ |
a5363890 | 1604 | fdp->prop = NULL; |
96c06863 | 1605 | fdp->written = false; /* not written on tags file yet */ |
9e0a3f98 | 1606 | |
a5363890 | 1607 | fdhead = fdp; |
9e0a3f98 FP |
1608 | curfdp = fdhead; /* the current file description */ |
1609 | ||
97b90b0a | 1610 | find_entries (fh); |
93b7ac65 | 1611 | |
2431364f | 1612 | /* If not Ctags, and if this is not metasource and if it contained no #line |
8378bcd3 FP |
1613 | directives, we can write the tags and free all nodes pointing to |
1614 | curfdp. */ | |
2431364f | 1615 | if (!CTAGS |
8378bcd3 | 1616 | && curfdp->usecharno /* no #line directives in this file */ |
2431364f FP |
1617 | && !curfdp->lang->metasource) |
1618 | { | |
8378bcd3 FP |
1619 | node *np, *prev; |
1620 | ||
1621 | /* Look for the head of the sublist relative to this file. See add_node | |
1622 | for the structure of the node tree. */ | |
1623 | prev = NULL; | |
1624 | for (np = nodehead; np != NULL; prev = np, np = np->left) | |
1625 | if (np->fdp == curfdp) | |
1626 | break; | |
1627 | ||
1628 | /* If we generated tags for this file, write and delete them. */ | |
1629 | if (np != NULL) | |
1630 | { | |
1631 | /* This is the head of the last sublist, if any. The following | |
1632 | instructions depend on this being true. */ | |
1633 | assert (np->left == NULL); | |
1634 | ||
1635 | assert (fdhead == curfdp); | |
1636 | assert (last_node->fdp == curfdp); | |
1637 | put_entries (np); /* write tags for file curfdp->taggedfname */ | |
1638 | free_tree (np); /* remove the written nodes */ | |
1639 | if (prev == NULL) | |
1640 | nodehead = NULL; /* no nodes left */ | |
1641 | else | |
1642 | prev->left = NULL; /* delete the pointer to the sublist */ | |
1643 | } | |
2431364f | 1644 | } |
c6d46f5f JB |
1645 | } |
1646 | ||
1647 | /* | |
eb8c3be9 | 1648 | * This routine sets up the boolean pseudo-functions which work |
93b7ac65 | 1649 | * by setting boolean flags dependent upon the corresponding character. |
c6d46f5f | 1650 | * Every char which is NOT in that string is not a white char. Therefore, |
96c06863 PE |
1651 | * all of the array "_wht" is set to false, and then the elements |
1652 | * subscripted by the chars in "white" are set to true. Thus "_wht" | |
1653 | * of a char is true if it is the string "white", else false. | |
c6d46f5f | 1654 | */ |
944bdd72 | 1655 | static void |
873fbd0b | 1656 | init (void) |
c6d46f5f | 1657 | { |
96c06863 PE |
1658 | const char *sp; |
1659 | int i; | |
c6d46f5f | 1660 | |
30903246 | 1661 | for (i = 0; i < CHARS; i++) |
96c06863 PE |
1662 | iswhite (i) = notinname (i) = begtoken (i) = intoken (i) = endtoken (i) |
1663 | = false; | |
1664 | for (sp = white; *sp != '\0'; sp++) iswhite (*sp) = true; | |
1665 | for (sp = nonam; *sp != '\0'; sp++) notinname (*sp) = true; | |
5e617bc2 | 1666 | notinname ('\0') = notinname ('\n'); |
96c06863 | 1667 | for (sp = begtk; *sp != '\0'; sp++) begtoken (*sp) = true; |
5e617bc2 | 1668 | begtoken ('\0') = begtoken ('\n'); |
96c06863 | 1669 | for (sp = midtk; *sp != '\0'; sp++) intoken (*sp) = true; |
5e617bc2 | 1670 | intoken ('\0') = intoken ('\n'); |
96c06863 | 1671 | for (sp = endtk; *sp != '\0'; sp++) endtoken (*sp) = true; |
5e617bc2 | 1672 | endtoken ('\0') = endtoken ('\n'); |
c6d46f5f JB |
1673 | } |
1674 | ||
1675 | /* | |
1676 | * This routine opens the specified file and calls the function | |
1677 | * which finds the function and type definitions. | |
1678 | */ | |
944bdd72 | 1679 | static void |
873fbd0b | 1680 | find_entries (FILE *inf) |
c6d46f5f | 1681 | { |
b9755a12 | 1682 | char *cp; |
9e0a3f98 FP |
1683 | language *lang = curfdp->lang; |
1684 | Lang_function *parser = NULL; | |
93b7ac65 | 1685 | |
b9755a12 | 1686 | /* If user specified a language, use it. */ |
93b7ac65 | 1687 | if (lang != NULL && lang->function != NULL) |
13fde0cd | 1688 | { |
9e0a3f98 | 1689 | parser = lang->function; |
13fde0cd | 1690 | } |
b9755a12 | 1691 | |
9e0a3f98 FP |
1692 | /* Else try to guess the language given the file name. */ |
1693 | if (parser == NULL) | |
1f638249 | 1694 | { |
96c06863 | 1695 | lang = get_language_from_filename (curfdp->infname, true); |
9e0a3f98 FP |
1696 | if (lang != NULL && lang->function != NULL) |
1697 | { | |
1698 | curfdp->lang = lang; | |
1699 | parser = lang->function; | |
1700 | } | |
1f638249 FP |
1701 | } |
1702 | ||
9e0a3f98 FP |
1703 | /* Else look for sharp-bang as the first two characters. */ |
1704 | if (parser == NULL | |
1705 | && readline_internal (&lb, inf) > 0 | |
e0903a92 | 1706 | && lb.len >= 2 |
1f638249 FP |
1707 | && lb.buffer[0] == '#' |
1708 | && lb.buffer[1] == '!') | |
c6d46f5f | 1709 | { |
1f638249 FP |
1710 | char *lp; |
1711 | ||
1712 | /* Set lp to point at the first char after the last slash in the | |
1713 | line or, if no slashes, at the first nonblank. Then set cp to | |
1714 | the first successive blank and terminate the string. */ | |
1715 | lp = etags_strrchr (lb.buffer+2, '/'); | |
1716 | if (lp != NULL) | |
1717 | lp += 1; | |
1718 | else | |
93b7ac65 FP |
1719 | lp = skip_spaces (lb.buffer + 2); |
1720 | cp = skip_non_spaces (lp); | |
1f638249 FP |
1721 | *cp = '\0'; |
1722 | ||
1723 | if (strlen (lp) > 0) | |
b9755a12 | 1724 | { |
93b7ac65 FP |
1725 | lang = get_language_from_interpreter (lp); |
1726 | if (lang != NULL && lang->function != NULL) | |
b9755a12 | 1727 | { |
9e0a3f98 FP |
1728 | curfdp->lang = lang; |
1729 | parser = lang->function; | |
b9755a12 FP |
1730 | } |
1731 | } | |
c6d46f5f | 1732 | } |
9e0a3f98 | 1733 | |
a5363890 FP |
1734 | /* We rewind here, even if inf may be a pipe. We fail if the |
1735 | length of the first line is longer than the pipe block size, | |
1736 | which is unlikely. */ | |
24dbe96a | 1737 | rewind (inf); |
54ef70a2 | 1738 | |
a5363890 FP |
1739 | /* Else try to guess the language given the case insensitive file name. */ |
1740 | if (parser == NULL) | |
1741 | { | |
96c06863 | 1742 | lang = get_language_from_filename (curfdp->infname, false); |
a5363890 FP |
1743 | if (lang != NULL && lang->function != NULL) |
1744 | { | |
1745 | curfdp->lang = lang; | |
1746 | parser = lang->function; | |
1747 | } | |
1748 | } | |
54ef70a2 | 1749 | |
24dbe96a FP |
1750 | /* Else try Fortran or C. */ |
1751 | if (parser == NULL) | |
1752 | { | |
1753 | node *old_last_node = last_node; | |
1754 | ||
1755 | curfdp->lang = get_language_from_langname ("fortran"); | |
1756 | find_entries (inf); | |
1757 | ||
1758 | if (old_last_node == last_node) | |
1759 | /* No Fortran entries found. Try C. */ | |
1760 | { | |
1761 | /* We do not tag if rewind fails. | |
1762 | Only the file name will be recorded in the tags file. */ | |
1763 | rewind (inf); | |
1764 | curfdp->lang = get_language_from_langname (cplusplus ? "c++" : "c"); | |
1765 | find_entries (inf); | |
1766 | } | |
1767 | return; | |
1768 | } | |
1769 | ||
9e0a3f98 FP |
1770 | if (!no_line_directive |
1771 | && curfdp->lang != NULL && curfdp->lang->metasource) | |
a5363890 | 1772 | /* It may be that this is a bingo.y file, and we already parsed a bingo.c |
9e0a3f98 | 1773 | file, or anyway we parsed a file that is automatically generated from |
a5363890 | 1774 | this one. If this is the case, the bingo.c file contained #line |
9e0a3f98 FP |
1775 | directives that generated tags pointing to this file. Let's delete |
1776 | them all before parsing this file, which is the real source. */ | |
1777 | { | |
1778 | fdesc **fdpp = &fdhead; | |
1779 | while (*fdpp != NULL) | |
1780 | if (*fdpp != curfdp | |
1781 | && streq ((*fdpp)->taggedfname, curfdp->taggedfname)) | |
1782 | /* We found one of those! We must delete both the file description | |
1783 | and all tags referring to it. */ | |
1784 | { | |
1785 | fdesc *badfdp = *fdpp; | |
1786 | ||
9c485bbe FP |
1787 | /* Delete the tags referring to badfdp->taggedfname |
1788 | that were obtained from badfdp->infname. */ | |
8378bcd3 | 1789 | invalidate_nodes (badfdp, &nodehead); |
9e0a3f98 | 1790 | |
2431364f | 1791 | *fdpp = badfdp->next; /* remove the bad description from the list */ |
8378bcd3 | 1792 | free_fdesc (badfdp); |
9e0a3f98 FP |
1793 | } |
1794 | else | |
1795 | fdpp = &(*fdpp)->next; /* advance the list pointer */ | |
1796 | } | |
1797 | ||
24dbe96a | 1798 | assert (parser != NULL); |
e7d3b099 | 1799 | |
e1dbe924 | 1800 | /* Generic initializations before reading from file. */ |
61a1f6fa | 1801 | linebuffer_setlen (&filebuf, 0); /* reset the file buffer */ |
9e0a3f98 | 1802 | |
e1dbe924 | 1803 | /* Generic initializations before parsing file with readline. */ |
24dbe96a FP |
1804 | lineno = 0; /* reset global line number */ |
1805 | charno = 0; /* reset global char number */ | |
1806 | linecharno = 0; /* reset global char number of line start */ | |
c6d46f5f | 1807 | |
24dbe96a FP |
1808 | parser (inf); |
1809 | ||
24dbe96a | 1810 | regex_tag_multiline (); |
c6d46f5f | 1811 | } |
b2521b0a | 1812 | |
c6d46f5f | 1813 | \f |
b2521b0a | 1814 | /* |
2201e3dc FP |
1815 | * Check whether an implicitly named tag should be created, |
1816 | * then call `pfnote'. | |
1817 | * NAME is a string that is internally copied by this function. | |
1818 | * | |
b2521b0a FP |
1819 | * TAGS format specification |
1820 | * Idea by Sam Kendall <kendall@mv.mv.com> (1997) | |
081b7c70 | 1821 | * The following is explained in some more detail in etc/ETAGS.EBNF. |
30903246 | 1822 | * |
2201e3dc FP |
1823 | * make_tag creates tags with "implicit tag names" (unnamed tags) |
1824 | * if the following are all true, assuming NONAM=" \f\t\n\r()=,;": | |
1825 | * 1. NAME does not contain any of the characters in NONAM; | |
1826 | * 2. LINESTART contains name as either a rightmost, or rightmost but | |
30903246 | 1827 | * one character, substring; |
2201e3dc FP |
1828 | * 3. the character, if any, immediately before NAME in LINESTART must |
1829 | * be a character in NONAM; | |
1830 | * 4. the character, if any, immediately after NAME in LINESTART must | |
1831 | * also be a character in NONAM. | |
93b7ac65 | 1832 | * |
e1dbe924 | 1833 | * The implementation uses the notinname() macro, which recognizes the |
2201e3dc FP |
1834 | * characters stored in the string `nonam'. |
1835 | * etags.el needs to use the same characters that are in NONAM. | |
30903246 | 1836 | */ |
0dacfc4b | 1837 | static void |
988e88ab J |
1838 | make_tag (const char *name, /* tag name, or NULL if unnamed */ |
1839 | int namelen, /* tag length */ | |
96c06863 | 1840 | bool is_func, /* tag is a function */ |
988e88ab J |
1841 | char *linestart, /* start of the line where tag is */ |
1842 | int linelen, /* length of the line where tag is */ | |
1843 | int lno, /* line number */ | |
1844 | long int cno) /* character number */ | |
30903246 | 1845 | { |
61a1f6fa | 1846 | bool named = (name != NULL && namelen > 0); |
988e88ab | 1847 | char *nname = NULL; |
30903246 | 1848 | |
61a1f6fa FP |
1849 | if (!CTAGS && named) /* maybe set named to false */ |
1850 | /* Let's try to make an implicit tag name, that is, create an unnamed tag | |
1851 | such that etags.el can guess a name from it. */ | |
30903246 | 1852 | { |
74032cc3 | 1853 | int i; |
988e88ab | 1854 | register const char *cp = name; |
74032cc3 FP |
1855 | |
1856 | for (i = 0; i < namelen; i++) | |
1857 | if (notinname (*cp++)) | |
1858 | break; | |
1859 | if (i == namelen) /* rule #1 */ | |
30903246 FP |
1860 | { |
1861 | cp = linestart + linelen - namelen; | |
1862 | if (notinname (linestart[linelen-1])) | |
1863 | cp -= 1; /* rule #4 */ | |
1864 | if (cp >= linestart /* rule #2 */ | |
1865 | && (cp == linestart | |
1866 | || notinname (cp[-1])) /* rule #3 */ | |
1867 | && strneq (name, cp, namelen)) /* rule #2 */ | |
96c06863 | 1868 | named = false; /* use implicit tag name */ |
30903246 FP |
1869 | } |
1870 | } | |
93b7ac65 | 1871 | |
30903246 | 1872 | if (named) |
988e88ab J |
1873 | nname = savenstr (name, namelen); |
1874 | ||
1875 | pfnote (nname, is_func, linestart, linelen, lno, cno); | |
30903246 FP |
1876 | } |
1877 | ||
081b7c70 FP |
1878 | /* Record a tag. */ |
1879 | static void | |
96c06863 PE |
1880 | pfnote (char *name, bool is_func, char *linestart, int linelen, int lno, |
1881 | long int cno) | |
873fbd0b DN |
1882 | /* tag name, or NULL if unnamed */ |
1883 | /* tag is a function */ | |
1884 | /* start of the line where tag is */ | |
1885 | /* length of the line where tag is */ | |
1886 | /* line number */ | |
1887 | /* character number */ | |
081b7c70 FP |
1888 | { |
1889 | register node *np; | |
1890 | ||
a127d423 | 1891 | assert (name == NULL || name[0] != '\0'); |
081b7c70 FP |
1892 | if (CTAGS && name == NULL) |
1893 | return; | |
1894 | ||
1895 | np = xnew (1, node); | |
1896 | ||
1897 | /* If ctags mode, change name "main" to M<thisfilename>. */ | |
1898 | if (CTAGS && !cxref_style && streq (name, "main")) | |
1899 | { | |
1900 | register char *fp = etags_strrchr (curfdp->taggedfname, '/'); | |
1901 | np->name = concat ("M", fp == NULL ? curfdp->taggedfname : fp + 1, ""); | |
1902 | fp = etags_strrchr (np->name, '.'); | |
1903 | if (fp != NULL && fp[1] != '\0' && fp[2] == '\0') | |
1904 | fp[0] = '\0'; | |
1905 | } | |
1906 | else | |
1907 | np->name = name; | |
96c06863 PE |
1908 | np->valid = true; |
1909 | np->been_warned = false; | |
081b7c70 FP |
1910 | np->fdp = curfdp; |
1911 | np->is_func = is_func; | |
1912 | np->lno = lno; | |
1913 | if (np->fdp->usecharno) | |
1914 | /* Our char numbers are 0-base, because of C language tradition? | |
1915 | ctags compatibility? old versions compatibility? I don't know. | |
1916 | Anyway, since emacs's are 1-base we expect etags.el to take care | |
1917 | of the difference. If we wanted to have 1-based numbers, we would | |
1918 | uncomment the +1 below. */ | |
1919 | np->cno = cno /* + 1 */ ; | |
1920 | else | |
1921 | np->cno = invalidcharno; | |
1922 | np->left = np->right = NULL; | |
1923 | if (CTAGS && !cxref_style) | |
1924 | { | |
1925 | if (strlen (linestart) < 50) | |
61a1f6fa | 1926 | np->regex = concat (linestart, "$", ""); |
081b7c70 | 1927 | else |
61a1f6fa | 1928 | np->regex = savenstr (linestart, 50); |
081b7c70 FP |
1929 | } |
1930 | else | |
61a1f6fa | 1931 | np->regex = savenstr (linestart, linelen); |
081b7c70 FP |
1932 | |
1933 | add_node (np, &nodehead); | |
1934 | } | |
1935 | ||
c6d46f5f JB |
1936 | /* |
1937 | * free_tree () | |
1938 | * recurse on left children, iterate on right children. | |
1939 | */ | |
944bdd72 | 1940 | static void |
873fbd0b | 1941 | free_tree (register node *np) |
c6d46f5f | 1942 | { |
93b7ac65 | 1943 | while (np) |
c6d46f5f | 1944 | { |
93b7ac65 FP |
1945 | register node *node_right = np->right; |
1946 | free_tree (np->left); | |
c2cd06e6 | 1947 | free (np->name); |
61a1f6fa | 1948 | free (np->regex); |
93b7ac65 FP |
1949 | free (np); |
1950 | np = node_right; | |
c6d46f5f JB |
1951 | } |
1952 | } | |
1953 | ||
8378bcd3 FP |
1954 | /* |
1955 | * free_fdesc () | |
1956 | * delete a file description | |
1957 | */ | |
1958 | static void | |
873fbd0b | 1959 | free_fdesc (register fdesc *fdp) |
8378bcd3 | 1960 | { |
c2cd06e6 JM |
1961 | free (fdp->infname); |
1962 | free (fdp->infabsname); | |
1963 | free (fdp->infabsdir); | |
1964 | free (fdp->taggedfname); | |
1965 | free (fdp->prop); | |
8378bcd3 FP |
1966 | free (fdp); |
1967 | } | |
1968 | ||
c6d46f5f JB |
1969 | /* |
1970 | * add_node () | |
db590582 FP |
1971 | * Adds a node to the tree of nodes. In etags mode, sort by file |
1972 | * name. In ctags mode, sort by tag name. Make no attempt at | |
1973 | * balancing. | |
c6d46f5f JB |
1974 | * |
1975 | * add_node is the only function allowed to add nodes, so it can | |
1976 | * maintain state. | |
1977 | */ | |
944bdd72 | 1978 | static void |
873fbd0b | 1979 | add_node (node *np, node **cur_node_p) |
c6d46f5f JB |
1980 | { |
1981 | register int dif; | |
93b7ac65 | 1982 | register node *cur_node = *cur_node_p; |
c6d46f5f JB |
1983 | |
1984 | if (cur_node == NULL) | |
1985 | { | |
93b7ac65 FP |
1986 | *cur_node_p = np; |
1987 | last_node = np; | |
c6d46f5f JB |
1988 | return; |
1989 | } | |
1990 | ||
32daa216 | 1991 | if (!CTAGS) |
8378bcd3 | 1992 | /* Etags Mode */ |
c6d46f5f | 1993 | { |
db590582 FP |
1994 | /* For each file name, tags are in a linked sublist on the right |
1995 | pointer. The first tags of different files are a linked list | |
1996 | on the left pointer. last_node points to the end of the last | |
1997 | used sublist. */ | |
8378bcd3 | 1998 | if (last_node != NULL && last_node->fdp == np->fdp) |
db590582 FP |
1999 | { |
2000 | /* Let's use the same sublist as the last added node. */ | |
8378bcd3 | 2001 | assert (last_node->right == NULL); |
db590582 FP |
2002 | last_node->right = np; |
2003 | last_node = np; | |
2004 | } | |
9e0a3f98 | 2005 | else if (cur_node->fdp == np->fdp) |
db590582 FP |
2006 | { |
2007 | /* Scanning the list we found the head of a sublist which is | |
2008 | good for us. Let's scan this sublist. */ | |
2009 | add_node (np, &cur_node->right); | |
2010 | } | |
2011 | else | |
2012 | /* The head of this sublist is not good for us. Let's try the | |
2013 | next one. */ | |
2014 | add_node (np, &cur_node->left); | |
8378bcd3 FP |
2015 | } /* if ETAGS mode */ |
2016 | ||
c6d46f5f JB |
2017 | else |
2018 | { | |
2019 | /* Ctags Mode */ | |
93b7ac65 | 2020 | dif = strcmp (np->name, cur_node->name); |
c6d46f5f JB |
2021 | |
2022 | /* | |
2023 | * If this tag name matches an existing one, then | |
2024 | * do not add the node, but maybe print a warning. | |
2025 | */ | |
ed8bbc0e | 2026 | if (no_duplicates && !dif) |
c6d46f5f | 2027 | { |
9e0a3f98 | 2028 | if (np->fdp == cur_node->fdp) |
c6d46f5f JB |
2029 | { |
2030 | if (!no_warnings) | |
2031 | { | |
2032 | fprintf (stderr, "Duplicate entry in file %s, line %d: %s\n", | |
9e0a3f98 | 2033 | np->fdp->infname, lineno, np->name); |
c6d46f5f JB |
2034 | fprintf (stderr, "Second entry ignored\n"); |
2035 | } | |
c6d46f5f | 2036 | } |
c5007f46 | 2037 | else if (!cur_node->been_warned && !no_warnings) |
c6d46f5f | 2038 | { |
c5007f46 FP |
2039 | fprintf |
2040 | (stderr, | |
2041 | "Duplicate entry in files %s and %s: %s (Warning only)\n", | |
9e0a3f98 | 2042 | np->fdp->infname, cur_node->fdp->infname, np->name); |
96c06863 | 2043 | cur_node->been_warned = true; |
c6d46f5f | 2044 | } |
c6d46f5f JB |
2045 | return; |
2046 | } | |
2047 | ||
c6d46f5f | 2048 | /* Actually add the node */ |
93b7ac65 | 2049 | add_node (np, dif < 0 ? &cur_node->left : &cur_node->right); |
8378bcd3 | 2050 | } /* if CTAGS mode */ |
c6d46f5f | 2051 | } |
b2521b0a | 2052 | |
9e0a3f98 FP |
2053 | /* |
2054 | * invalidate_nodes () | |
2055 | * Scan the node tree and invalidate all nodes pointing to the | |
8378bcd3 | 2056 | * given file description (CTAGS case) or free them (ETAGS case). |
9e0a3f98 FP |
2057 | */ |
2058 | static void | |
873fbd0b | 2059 | invalidate_nodes (fdesc *badfdp, node **npp) |
9e0a3f98 | 2060 | { |
8378bcd3 FP |
2061 | node *np = *npp; |
2062 | ||
2063 | if (np == NULL) | |
2064 | return; | |
2065 | ||
2066 | if (CTAGS) | |
2067 | { | |
2068 | if (np->left != NULL) | |
2069 | invalidate_nodes (badfdp, &np->left); | |
2070 | if (np->fdp == badfdp) | |
96c06863 | 2071 | np->valid = false; |
8378bcd3 FP |
2072 | if (np->right != NULL) |
2073 | invalidate_nodes (badfdp, &np->right); | |
2074 | } | |
2075 | else | |
2076 | { | |
89fb2be1 | 2077 | assert (np->fdp != NULL); |
8378bcd3 FP |
2078 | if (np->fdp == badfdp) |
2079 | { | |
89fb2be1 | 2080 | *npp = np->left; /* detach the sublist from the list */ |
8378bcd3 FP |
2081 | np->left = NULL; /* isolate it */ |
2082 | free_tree (np); /* free it */ | |
89fb2be1 | 2083 | invalidate_nodes (badfdp, npp); |
8378bcd3 | 2084 | } |
89fb2be1 FP |
2085 | else |
2086 | invalidate_nodes (badfdp, &np->left); | |
8378bcd3 | 2087 | } |
9e0a3f98 FP |
2088 | } |
2089 | ||
c6d46f5f | 2090 | \f |
f57e2426 | 2091 | static int total_size_of_entries (node *); |
5994c183 | 2092 | static int number_len (long) ATTRIBUTE_CONST; |
db590582 | 2093 | |
9e0a3f98 | 2094 | /* Length of a non-negative number's decimal representation. */ |
db590582 | 2095 | static int |
873fbd0b | 2096 | number_len (long int num) |
db590582 FP |
2097 | { |
2098 | int len = 1; | |
2099 | while ((num /= 10) > 0) | |
2100 | len += 1; | |
2101 | return len; | |
2102 | } | |
2103 | ||
2104 | /* | |
2105 | * Return total number of characters that put_entries will output for | |
2106 | * the nodes in the linked list at the right of the specified node. | |
2107 | * This count is irrelevant with etags.el since emacs 19.34 at least, | |
2108 | * but is still supplied for backward compatibility. | |
2109 | */ | |
2110 | static int | |
873fbd0b | 2111 | total_size_of_entries (register node *np) |
db590582 FP |
2112 | { |
2113 | register int total = 0; | |
2114 | ||
2115 | for (; np != NULL; np = np->right) | |
61a1f6fa FP |
2116 | if (np->valid) |
2117 | { | |
2118 | total += strlen (np->regex) + 1; /* pat\177 */ | |
2119 | if (np->name != NULL) | |
2120 | total += strlen (np->name) + 1; /* name\001 */ | |
2121 | total += number_len ((long) np->lno) + 1; /* lno, */ | |
2122 | if (np->cno != invalidcharno) /* cno */ | |
2123 | total += number_len (np->cno); | |
2124 | total += 1; /* newline */ | |
2125 | } | |
db590582 FP |
2126 | |
2127 | return total; | |
2128 | } | |
db590582 | 2129 | |
944bdd72 | 2130 | static void |
873fbd0b | 2131 | put_entries (register node *np) |
c6d46f5f | 2132 | { |
13fde0cd | 2133 | register char *sp; |
9e0a3f98 | 2134 | static fdesc *fdp = NULL; |
c6d46f5f | 2135 | |
93b7ac65 | 2136 | if (np == NULL) |
c6d46f5f JB |
2137 | return; |
2138 | ||
2139 | /* Output subentries that precede this one */ | |
db590582 FP |
2140 | if (CTAGS) |
2141 | put_entries (np->left); | |
c6d46f5f JB |
2142 | |
2143 | /* Output this entry */ | |
9e0a3f98 | 2144 | if (np->valid) |
c6d46f5f | 2145 | { |
9e0a3f98 | 2146 | if (!CTAGS) |
db590582 | 2147 | { |
9e0a3f98 FP |
2148 | /* Etags mode */ |
2149 | if (fdp != np->fdp) | |
2150 | { | |
2151 | fdp = np->fdp; | |
2152 | fprintf (tagf, "\f\n%s,%d\n", | |
2153 | fdp->taggedfname, total_size_of_entries (np)); | |
96c06863 | 2154 | fdp->written = true; |
9e0a3f98 | 2155 | } |
61a1f6fa | 2156 | fputs (np->regex, tagf); |
9e0a3f98 FP |
2157 | fputc ('\177', tagf); |
2158 | if (np->name != NULL) | |
2159 | { | |
2160 | fputs (np->name, tagf); | |
2161 | fputc ('\001', tagf); | |
2162 | } | |
2163 | fprintf (tagf, "%d,", np->lno); | |
2164 | if (np->cno != invalidcharno) | |
2165 | fprintf (tagf, "%ld", np->cno); | |
2166 | fputs ("\n", tagf); | |
db590582 | 2167 | } |
c6d46f5f | 2168 | else |
d8913c1c | 2169 | { |
9e0a3f98 FP |
2170 | /* Ctags mode */ |
2171 | if (np->name == NULL) | |
db5a3003 | 2172 | error ("internal error: NULL name in ctags mode."); |
9e0a3f98 FP |
2173 | |
2174 | if (cxref_style) | |
2175 | { | |
2176 | if (vgrind_style) | |
2177 | fprintf (stdout, "%s %s %d\n", | |
2178 | np->name, np->fdp->taggedfname, (np->lno + 63) / 64); | |
2179 | else | |
2180 | fprintf (stdout, "%-16s %3d %-16s %s\n", | |
61a1f6fa | 2181 | np->name, np->lno, np->fdp->taggedfname, np->regex); |
9e0a3f98 | 2182 | } |
d8913c1c | 2183 | else |
9e0a3f98 FP |
2184 | { |
2185 | fprintf (tagf, "%s\t%s\t", np->name, np->fdp->taggedfname); | |
d8913c1c | 2186 | |
9e0a3f98 FP |
2187 | if (np->is_func) |
2188 | { /* function or #define macro with args */ | |
2189 | putc (searchar, tagf); | |
2190 | putc ('^', tagf); | |
d8913c1c | 2191 | |
61a1f6fa | 2192 | for (sp = np->regex; *sp; sp++) |
9e0a3f98 FP |
2193 | { |
2194 | if (*sp == '\\' || *sp == searchar) | |
2195 | putc ('\\', tagf); | |
2196 | putc (*sp, tagf); | |
2197 | } | |
2198 | putc (searchar, tagf); | |
d8913c1c | 2199 | } |
9e0a3f98 FP |
2200 | else |
2201 | { /* anything else; text pattern inadequate */ | |
2202 | fprintf (tagf, "%d", np->lno); | |
2203 | } | |
2204 | putc ('\n', tagf); | |
d8913c1c | 2205 | } |
c6d46f5f | 2206 | } |
9e0a3f98 | 2207 | } /* if this node contains a valid tag */ |
db590582 | 2208 | |
c6d46f5f | 2209 | /* Output subentries that follow this one */ |
93b7ac65 | 2210 | put_entries (np->right); |
db590582 FP |
2211 | if (!CTAGS) |
2212 | put_entries (np->left); | |
c6d46f5f | 2213 | } |
b2521b0a | 2214 | |
c6d46f5f | 2215 | \f |
b28e26be FP |
2216 | /* C extensions. */ |
2217 | #define C_EXT 0x00fff /* C extensions */ | |
2218 | #define C_PLAIN 0x00000 /* C */ | |
2219 | #define C_PLPL 0x00001 /* C++ */ | |
2220 | #define C_STAR 0x00003 /* C* */ | |
2221 | #define C_JAVA 0x00005 /* JAVA */ | |
2222 | #define C_AUTO 0x01000 /* C, but switch to C++ if `class' is met */ | |
2223 | #define YACC 0x10000 /* yacc file */ | |
2224 | ||
c6d46f5f JB |
2225 | /* |
2226 | * The C symbol tables. | |
2227 | */ | |
55597f90 FP |
2228 | enum sym_type |
2229 | { | |
30903246 FP |
2230 | st_none, |
2231 | st_C_objprot, st_C_objimpl, st_C_objend, | |
2232 | st_C_gnumacro, | |
05d9a399 | 2233 | st_C_ignore, st_C_attribute, |
30903246 | 2234 | st_C_javastruct, |
93b7ac65 | 2235 | st_C_operator, |
f55ae599 | 2236 | st_C_class, st_C_template, |
05d9a399 | 2237 | st_C_struct, st_C_extern, st_C_enum, st_C_define, st_C_typedef |
55597f90 | 2238 | }; |
c6d46f5f | 2239 | |
42680d3c | 2240 | /* Feed stuff between (but not including) %[ and %] lines to: |
05d9a399 | 2241 | gperf -m 5 |
42680d3c | 2242 | %[ |
05d9a399 FP |
2243 | %compare-strncmp |
2244 | %enum | |
2245 | %struct-type | |
42680d3c FP |
2246 | struct C_stab_entry { char *name; int c_ext; enum sym_type type; } |
2247 | %% | |
05d9a399 FP |
2248 | if, 0, st_C_ignore |
2249 | for, 0, st_C_ignore | |
2250 | while, 0, st_C_ignore | |
2251 | switch, 0, st_C_ignore | |
2252 | return, 0, st_C_ignore | |
2253 | __attribute__, 0, st_C_attribute | |
36c51859 | 2254 | GTY, 0, st_C_attribute |
05d9a399 FP |
2255 | @interface, 0, st_C_objprot |
2256 | @protocol, 0, st_C_objprot | |
2257 | @implementation,0, st_C_objimpl | |
2258 | @end, 0, st_C_objend | |
47fd18c6 FP |
2259 | import, (C_JAVA & ~C_PLPL), st_C_ignore |
2260 | package, (C_JAVA & ~C_PLPL), st_C_ignore | |
05d9a399 | 2261 | friend, C_PLPL, st_C_ignore |
47fd18c6 FP |
2262 | extends, (C_JAVA & ~C_PLPL), st_C_javastruct |
2263 | implements, (C_JAVA & ~C_PLPL), st_C_javastruct | |
2264 | interface, (C_JAVA & ~C_PLPL), st_C_struct | |
05d9a399 FP |
2265 | class, 0, st_C_class |
2266 | namespace, C_PLPL, st_C_struct | |
2267 | domain, C_STAR, st_C_struct | |
2268 | union, 0, st_C_struct | |
2269 | struct, 0, st_C_struct | |
2270 | extern, 0, st_C_extern | |
2271 | enum, 0, st_C_enum | |
2272 | typedef, 0, st_C_typedef | |
2273 | define, 0, st_C_define | |
327891eb | 2274 | undef, 0, st_C_define |
05d9a399 FP |
2275 | operator, C_PLPL, st_C_operator |
2276 | template, 0, st_C_template | |
d8913c1c | 2277 | # DEFUN used in emacs, the next three used in glibc (SYSCALL only for mach). |
05d9a399 FP |
2278 | DEFUN, 0, st_C_gnumacro |
2279 | SYSCALL, 0, st_C_gnumacro | |
2280 | ENTRY, 0, st_C_gnumacro | |
2281 | PSEUDO, 0, st_C_gnumacro | |
d8913c1c FP |
2282 | # These are defined inside C functions, so currently they are not met. |
2283 | # EXFUN used in glibc, DEFVAR_* in emacs. | |
05d9a399 FP |
2284 | #EXFUN, 0, st_C_gnumacro |
2285 | #DEFVAR_, 0, st_C_gnumacro | |
42680d3c | 2286 | %] |
05d9a399 FP |
2287 | and replace lines between %< and %> with its output, then: |
2288 | - remove the #if characterset check | |
2289 | - make in_word_set static and not inline. */ | |
42680d3c | 2290 | /*%<*/ |
05d9a399 FP |
2291 | /* C code produced by gperf version 3.0.1 */ |
2292 | /* Command-line: gperf -m 5 */ | |
327891eb | 2293 | /* Computed positions: -k'2-3' */ |
42680d3c | 2294 | |
988e88ab | 2295 | struct C_stab_entry { const char *name; int c_ext; enum sym_type type; }; |
327891eb | 2296 | /* maximum key range = 33, duplicates = 0 */ |
42680d3c | 2297 | |
164b1ba3 PE |
2298 | static int |
2299 | hash (const char *str, int len) | |
42680d3c | 2300 | { |
164b1ba3 | 2301 | static char const asso_values[] = |
42680d3c | 2302 | { |
327891eb FP |
2303 | 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, |
2304 | 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, | |
2305 | 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, | |
2306 | 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, | |
2307 | 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, | |
2308 | 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, | |
36c51859 CY |
2309 | 35, 35, 35, 35, 35, 35, 35, 35, 35, 3, |
2310 | 26, 35, 35, 35, 35, 35, 35, 35, 27, 35, | |
2311 | 35, 35, 35, 24, 0, 35, 35, 35, 35, 0, | |
327891eb FP |
2312 | 35, 35, 35, 35, 35, 1, 35, 16, 35, 6, |
2313 | 23, 0, 0, 35, 22, 0, 35, 35, 5, 0, | |
2314 | 0, 15, 1, 35, 6, 35, 8, 19, 35, 16, | |
2315 | 4, 5, 35, 35, 35, 35, 35, 35, 35, 35, | |
2316 | 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, | |
2317 | 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, | |
2318 | 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, | |
2319 | 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, | |
2320 | 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, | |
2321 | 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, | |
2322 | 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, | |
2323 | 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, | |
2324 | 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, | |
2325 | 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, | |
2326 | 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, | |
2327 | 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, | |
2328 | 35, 35, 35, 35, 35, 35 | |
93b7ac65 | 2329 | }; |
164b1ba3 | 2330 | int hval = len; |
327891eb FP |
2331 | |
2332 | switch (hval) | |
2333 | { | |
2334 | default: | |
164b1ba3 | 2335 | hval += asso_values[(unsigned char) str[2]]; |
327891eb FP |
2336 | /*FALLTHROUGH*/ |
2337 | case 2: | |
164b1ba3 | 2338 | hval += asso_values[(unsigned char) str[1]]; |
327891eb FP |
2339 | break; |
2340 | } | |
2341 | return hval; | |
42680d3c | 2342 | } |
c6d46f5f | 2343 | |
30526cc6 | 2344 | static struct C_stab_entry * |
873fbd0b | 2345 | in_word_set (register const char *str, register unsigned int len) |
42680d3c | 2346 | { |
05d9a399 FP |
2347 | enum |
2348 | { | |
36c51859 | 2349 | TOTAL_KEYWORDS = 33, |
05d9a399 FP |
2350 | MIN_WORD_LENGTH = 2, |
2351 | MAX_WORD_LENGTH = 15, | |
327891eb FP |
2352 | MIN_HASH_VALUE = 2, |
2353 | MAX_HASH_VALUE = 34 | |
05d9a399 FP |
2354 | }; |
2355 | ||
93b7ac65 | 2356 | static struct C_stab_entry wordlist[] = |
42680d3c | 2357 | { |
327891eb | 2358 | {""}, {""}, |
05d9a399 | 2359 | {"if", 0, st_C_ignore}, |
36c51859 | 2360 | {"GTY", 0, st_C_attribute}, |
05d9a399 | 2361 | {"@end", 0, st_C_objend}, |
327891eb | 2362 | {"union", 0, st_C_struct}, |
05d9a399 | 2363 | {"define", 0, st_C_define}, |
47fd18c6 | 2364 | {"import", (C_JAVA & ~C_PLPL), st_C_ignore}, |
05d9a399 | 2365 | {"template", 0, st_C_template}, |
327891eb FP |
2366 | {"operator", C_PLPL, st_C_operator}, |
2367 | {"@interface", 0, st_C_objprot}, | |
47fd18c6 | 2368 | {"implements", (C_JAVA & ~C_PLPL), st_C_javastruct}, |
327891eb | 2369 | {"friend", C_PLPL, st_C_ignore}, |
05d9a399 | 2370 | {"typedef", 0, st_C_typedef}, |
327891eb FP |
2371 | {"return", 0, st_C_ignore}, |
2372 | {"@implementation",0, st_C_objimpl}, | |
2373 | {"@protocol", 0, st_C_objprot}, | |
47fd18c6 | 2374 | {"interface", (C_JAVA & ~C_PLPL), st_C_struct}, |
327891eb | 2375 | {"extern", 0, st_C_extern}, |
47fd18c6 | 2376 | {"extends", (C_JAVA & ~C_PLPL), st_C_javastruct}, |
05d9a399 | 2377 | {"struct", 0, st_C_struct}, |
327891eb | 2378 | {"domain", C_STAR, st_C_struct}, |
05d9a399 | 2379 | {"switch", 0, st_C_ignore}, |
327891eb FP |
2380 | {"enum", 0, st_C_enum}, |
2381 | {"for", 0, st_C_ignore}, | |
2382 | {"namespace", C_PLPL, st_C_struct}, | |
05d9a399 | 2383 | {"class", 0, st_C_class}, |
327891eb FP |
2384 | {"while", 0, st_C_ignore}, |
2385 | {"undef", 0, st_C_define}, | |
47fd18c6 | 2386 | {"package", (C_JAVA & ~C_PLPL), st_C_ignore}, |
05d9a399 FP |
2387 | {"__attribute__", 0, st_C_attribute}, |
2388 | {"SYSCALL", 0, st_C_gnumacro}, | |
05d9a399 | 2389 | {"ENTRY", 0, st_C_gnumacro}, |
327891eb | 2390 | {"PSEUDO", 0, st_C_gnumacro}, |
05d9a399 | 2391 | {"DEFUN", 0, st_C_gnumacro} |
42680d3c FP |
2392 | }; |
2393 | ||
2394 | if (len <= MAX_WORD_LENGTH && len >= MIN_WORD_LENGTH) | |
2395 | { | |
164b1ba3 | 2396 | int key = hash (str, len); |
42680d3c | 2397 | |
93b7ac65 | 2398 | if (key <= MAX_HASH_VALUE && key >= 0) |
42680d3c | 2399 | { |
164b1ba3 | 2400 | const char *s = wordlist[key].name; |
42680d3c | 2401 | |
05d9a399 | 2402 | if (*str == *s && !strncmp (str + 1, s + 1, len - 1) && s[len] == '\0') |
42680d3c FP |
2403 | return &wordlist[key]; |
2404 | } | |
2405 | } | |
2406 | return 0; | |
c6d46f5f | 2407 | } |
42680d3c | 2408 | /*%>*/ |
c6d46f5f | 2409 | |
944bdd72 | 2410 | static enum sym_type |
873fbd0b | 2411 | C_symtype (char *str, int len, int c_ext) |
c6d46f5f | 2412 | { |
3f1c8fcd | 2413 | register struct C_stab_entry *se = in_word_set (str, len); |
c6d46f5f | 2414 | |
42680d3c FP |
2415 | if (se == NULL || (se->c_ext && !(c_ext & se->c_ext))) |
2416 | return st_none; | |
2417 | return se->type; | |
c6d46f5f | 2418 | } |
b2521b0a | 2419 | |
c6d46f5f | 2420 | \f |
05d9a399 FP |
2421 | /* |
2422 | * Ignoring __attribute__ ((list)) | |
2423 | */ | |
2424 | static bool inattribute; /* looking at an __attribute__ construct */ | |
2425 | ||
b2521b0a FP |
2426 | /* |
2427 | * C functions and variables are recognized using a simple | |
2428 | * finite automaton. fvdef is its state variable. | |
2429 | */ | |
db590582 | 2430 | static enum |
13fde0cd | 2431 | { |
30903246 | 2432 | fvnone, /* nothing seen */ |
b2521b0a FP |
2433 | fdefunkey, /* Emacs DEFUN keyword seen */ |
2434 | fdefunname, /* Emacs DEFUN name seen */ | |
93b7ac65 | 2435 | foperator, /* func: operator keyword seen (cplpl) */ |
30903246 FP |
2436 | fvnameseen, /* function or variable name seen */ |
2437 | fstartlist, /* func: just after open parenthesis */ | |
2438 | finlist, /* func: in parameter list */ | |
2439 | flistseen, /* func: after parameter list */ | |
2440 | fignore, /* func: before open brace */ | |
2441 | vignore /* var-like: ignore until ';' */ | |
2442 | } fvdef; | |
13fde0cd | 2443 | |
db590582 | 2444 | static bool fvextern; /* func or var: extern keyword seen; */ |
13fde0cd | 2445 | |
b2521b0a FP |
2446 | /* |
2447 | * typedefs are recognized using a simple finite automaton. | |
2448 | * typdef is its state variable. | |
2449 | */ | |
db590582 | 2450 | static enum |
13fde0cd | 2451 | { |
31d4b314 | 2452 | tnone, /* nothing seen */ |
93b7ac65 FP |
2453 | tkeyseen, /* typedef keyword seen */ |
2454 | ttypeseen, /* defined type seen */ | |
31d4b314 | 2455 | tinbody, /* inside typedef body */ |
46c145db FP |
2456 | tend, /* just before typedef tag */ |
2457 | tignore /* junk after typedef tag */ | |
d8913c1c | 2458 | } typdef; |
13fde0cd | 2459 | |
b2521b0a FP |
2460 | /* |
2461 | * struct-like structures (enum, struct and union) are recognized | |
2462 | * using another simple finite automaton. `structdef' is its state | |
2463 | * variable. | |
2464 | */ | |
db590582 | 2465 | static enum |
13fde0cd | 2466 | { |
8c463abe | 2467 | snone, /* nothing seen yet, |
05d9a399 | 2468 | or in struct body if bracelev > 0 */ |
13fde0cd RS |
2469 | skeyseen, /* struct-like keyword seen */ |
2470 | stagseen, /* struct-like tag seen */ | |
8c463abe | 2471 | scolonseen /* colon seen after struct-like tag */ |
d8913c1c | 2472 | } structdef; |
46c145db | 2473 | |
d8913c1c FP |
2474 | /* |
2475 | * When objdef is different from onone, objtag is the name of the class. | |
2476 | */ | |
988e88ab | 2477 | static const char *objtag = "<uninited>"; |
d8913c1c | 2478 | |
13fde0cd RS |
2479 | /* |
2480 | * Yet another little state machine to deal with preprocessor lines. | |
2481 | */ | |
db590582 | 2482 | static enum |
13fde0cd RS |
2483 | { |
2484 | dnone, /* nothing seen */ | |
2485 | dsharpseen, /* '#' seen as first char on line */ | |
2486 | ddefineseen, /* '#' and 'define' seen */ | |
46e4cb76 | 2487 | dignorerest /* ignore rest of line */ |
d8913c1c FP |
2488 | } definedef; |
2489 | ||
2490 | /* | |
2491 | * State machine for Objective C protocols and implementations. | |
b2521b0a | 2492 | * Idea by Tom R.Hageman <tom@basil.icce.rug.nl> (1995) |
d8913c1c | 2493 | */ |
db590582 | 2494 | static enum |
d8913c1c FP |
2495 | { |
2496 | onone, /* nothing seen */ | |
2497 | oprotocol, /* @interface or @protocol seen */ | |
2498 | oimplementation, /* @implementations seen */ | |
2499 | otagseen, /* class name seen */ | |
2500 | oparenseen, /* parenthesis before category seen */ | |
2501 | ocatseen, /* category name seen */ | |
2502 | oinbody, /* in @implementation body */ | |
2503 | omethodsign, /* in @implementation body, after +/- */ | |
2504 | omethodtag, /* after method name */ | |
2505 | omethodcolon, /* after method colon */ | |
2506 | omethodparm, /* after method parameter */ | |
ba1fe3a8 | 2507 | oignore /* wait for @end */ |
d8913c1c | 2508 | } objdef; |
13fde0cd | 2509 | |
30903246 FP |
2510 | |
2511 | /* | |
2512 | * Use this structure to keep info about the token read, and how it | |
2513 | * should be tagged. Used by the make_C_tag function to build a tag. | |
2514 | */ | |
db590582 | 2515 | static struct tok |
30903246 | 2516 | { |
9c485bbe FP |
2517 | char *line; /* string containing the token */ |
2518 | int offset; /* where the token starts in LINE */ | |
2519 | int length; /* token length */ | |
2520 | /* | |
2521 | The previous members can be used to pass strings around for generic | |
2522 | purposes. The following ones specifically refer to creating tags. In this | |
2523 | case the token contained here is the pattern that will be used to create a | |
2524 | tag. | |
2525 | */ | |
2526 | bool valid; /* do not create a tag; the token should be | |
2527 | invalidated whenever a state machine is | |
2528 | reset prematurely */ | |
2529 | bool named; /* create a named tag */ | |
2530 | int lineno; /* source line number of tag */ | |
2531 | long linepos; /* source char number of tag */ | |
b2521b0a | 2532 | } token; /* latest token read */ |
d8913c1c | 2533 | |
8c463abe FP |
2534 | /* |
2535 | * Variables and functions for dealing with nested structures. | |
2536 | * Idea by Mykola Dzyuba <mdzyuba@yahoo.com> (2001) | |
2537 | */ | |
f57e2426 J |
2538 | static void pushclass_above (int, char *, int); |
2539 | static void popclass_above (int); | |
988e88ab | 2540 | static void write_classname (linebuffer *, const char *qualifier); |
8c463abe | 2541 | |
db590582 | 2542 | static struct { |
8c463abe | 2543 | char **cname; /* nested class names */ |
05d9a399 | 2544 | int *bracelev; /* nested class brace level */ |
8c463abe FP |
2545 | int nl; /* class nesting level (elements used) */ |
2546 | int size; /* length of the array */ | |
2547 | } cstack; /* stack for nested declaration tags */ | |
2548 | /* Current struct nesting depth (namespace, class, struct, union, enum). */ | |
2549 | #define nestlev (cstack.nl) | |
3c04a71a | 2550 | /* After struct keyword or in struct body, not inside a nested function. */ |
8c463abe | 2551 | #define instruct (structdef == snone && nestlev > 0 \ |
05d9a399 | 2552 | && bracelev == cstack.bracelev[nestlev-1] + 1) |
8c463abe FP |
2553 | |
2554 | static void | |
873fbd0b | 2555 | pushclass_above (int bracelev, char *str, int len) |
8c463abe FP |
2556 | { |
2557 | int nl; | |
2558 | ||
05d9a399 | 2559 | popclass_above (bracelev); |
8c463abe FP |
2560 | nl = cstack.nl; |
2561 | if (nl >= cstack.size) | |
2562 | { | |
2563 | int size = cstack.size *= 2; | |
2564 | xrnew (cstack.cname, size, char *); | |
05d9a399 | 2565 | xrnew (cstack.bracelev, size, int); |
8c463abe | 2566 | } |
05d9a399 | 2567 | assert (nl == 0 || cstack.bracelev[nl-1] < bracelev); |
8c463abe | 2568 | cstack.cname[nl] = (str == NULL) ? NULL : savenstr (str, len); |
05d9a399 | 2569 | cstack.bracelev[nl] = bracelev; |
8c463abe FP |
2570 | cstack.nl = nl + 1; |
2571 | } | |
2572 | ||
2573 | static void | |
873fbd0b | 2574 | popclass_above (int bracelev) |
8c463abe FP |
2575 | { |
2576 | int nl; | |
2577 | ||
2578 | for (nl = cstack.nl - 1; | |
05d9a399 | 2579 | nl >= 0 && cstack.bracelev[nl] >= bracelev; |
8c463abe FP |
2580 | nl--) |
2581 | { | |
c2cd06e6 | 2582 | free (cstack.cname[nl]); |
8c463abe FP |
2583 | cstack.nl = nl; |
2584 | } | |
2585 | } | |
2586 | ||
2587 | static void | |
988e88ab | 2588 | write_classname (linebuffer *cn, const char *qualifier) |
8c463abe FP |
2589 | { |
2590 | int i, len; | |
2591 | int qlen = strlen (qualifier); | |
2592 | ||
2593 | if (cstack.nl == 0 || cstack.cname[0] == NULL) | |
2594 | { | |
2595 | len = 0; | |
2596 | cn->len = 0; | |
2597 | cn->buffer[0] = '\0'; | |
2598 | } | |
2599 | else | |
2600 | { | |
2601 | len = strlen (cstack.cname[0]); | |
2602 | linebuffer_setlen (cn, len); | |
2603 | strcpy (cn->buffer, cstack.cname[0]); | |
2604 | } | |
2605 | for (i = 1; i < cstack.nl; i++) | |
2606 | { | |
e99a530f | 2607 | char *s = cstack.cname[i]; |
8c463abe FP |
2608 | if (s == NULL) |
2609 | continue; | |
e99a530f PE |
2610 | linebuffer_setlen (cn, len + qlen + strlen (s)); |
2611 | len += sprintf (cn->buffer + len, "%s%s", qualifier, s); | |
8c463abe FP |
2612 | } |
2613 | } | |
2614 | ||
2615 | \f | |
f57e2426 J |
2616 | static bool consider_token (char *, int, int, int *, int, int, bool *); |
2617 | static void make_C_tag (bool); | |
71cbb895 | 2618 | |
6dd5561c FP |
2619 | /* |
2620 | * consider_token () | |
2621 | * checks to see if the current token is at the start of a | |
30903246 FP |
2622 | * function or variable, or corresponds to a typedef, or |
2623 | * is a struct/union/enum tag, or #define, or an enum constant. | |
6dd5561c | 2624 | * |
96c06863 | 2625 | * *IS_FUNC_OR_VAR gets true if the token is a function or #define macro |
8c463abe | 2626 | * with args. C_EXTP points to which language we are looking at. |
6dd5561c | 2627 | * |
6dd5561c | 2628 | * Globals |
30903246 | 2629 | * fvdef IN OUT |
6dd5561c FP |
2630 | * structdef IN OUT |
2631 | * definedef IN OUT | |
2632 | * typdef IN OUT | |
d8913c1c | 2633 | * objdef IN OUT |
6dd5561c FP |
2634 | */ |
2635 | ||
944bdd72 | 2636 | static bool |
96c06863 PE |
2637 | consider_token (char *str, int len, int c, int *c_extp, |
2638 | int bracelev, int parlev, bool *is_func_or_var) | |
873fbd0b DN |
2639 | /* IN: token pointer */ |
2640 | /* IN: token length */ | |
2641 | /* IN: first char after the token */ | |
2642 | /* IN, OUT: C extensions mask */ | |
2643 | /* IN: brace level */ | |
2644 | /* IN: parenthesis level */ | |
2645 | /* OUT: function or variable found */ | |
6dd5561c | 2646 | { |
05d9a399 | 2647 | /* When structdef is stagseen, scolonseen, or snone with bracelev > 0, |
8c463abe | 2648 | structtype is the type of the preceding struct-like keyword, and |
05d9a399 | 2649 | structbracelev is the brace level where it has been seen. */ |
b2521b0a | 2650 | static enum sym_type structtype; |
05d9a399 | 2651 | static int structbracelev; |
b2521b0a FP |
2652 | static enum sym_type toktype; |
2653 | ||
2654 | ||
8c463abe | 2655 | toktype = C_symtype (str, len, *c_extp); |
6dd5561c FP |
2656 | |
2657 | /* | |
05d9a399 | 2658 | * Skip __attribute__ |
6dd5561c | 2659 | */ |
05d9a399 | 2660 | if (toktype == st_C_attribute) |
6dd5561c | 2661 | { |
96c06863 PE |
2662 | inattribute = true; |
2663 | return false; | |
05d9a399 FP |
2664 | } |
2665 | ||
2666 | /* | |
2667 | * Advance the definedef state machine. | |
2668 | */ | |
2669 | switch (definedef) | |
2670 | { | |
2671 | case dnone: | |
2672 | /* We're not on a preprocessor line. */ | |
2673 | if (toktype == st_C_gnumacro) | |
2674 | { | |
2675 | fvdef = fdefunkey; | |
96c06863 | 2676 | return false; |
05d9a399 FP |
2677 | } |
2678 | break; | |
2679 | case dsharpseen: | |
2680 | if (toktype == st_C_define) | |
2681 | { | |
2682 | definedef = ddefineseen; | |
2683 | } | |
2684 | else | |
2685 | { | |
2686 | definedef = dignorerest; | |
2687 | } | |
96c06863 | 2688 | return false; |
05d9a399 FP |
2689 | case ddefineseen: |
2690 | /* | |
2691 | * Make a tag for any macro, unless it is a constant | |
96c06863 | 2692 | * and constantypedefs is false. |
05d9a399 FP |
2693 | */ |
2694 | definedef = dignorerest; | |
2695 | *is_func_or_var = (c == '('); | |
2696 | if (!*is_func_or_var && !constantypedefs) | |
96c06863 | 2697 | return false; |
05d9a399 | 2698 | else |
96c06863 | 2699 | return true; |
05d9a399 | 2700 | case dignorerest: |
96c06863 | 2701 | return false; |
05d9a399 | 2702 | default: |
db5a3003 | 2703 | error ("internal error: definedef value."); |
05d9a399 FP |
2704 | } |
2705 | ||
2706 | /* | |
2707 | * Now typedefs | |
2708 | */ | |
2709 | switch (typdef) | |
2710 | { | |
2711 | case tnone: | |
2712 | if (toktype == st_C_typedef) | |
2713 | { | |
2714 | if (typedefs) | |
2715 | typdef = tkeyseen; | |
96c06863 | 2716 | fvextern = false; |
05d9a399 | 2717 | fvdef = fvnone; |
96c06863 | 2718 | return false; |
05d9a399 FP |
2719 | } |
2720 | break; | |
2721 | case tkeyseen: | |
2722 | switch (toktype) | |
2723 | { | |
2724 | case st_none: | |
2725 | case st_C_class: | |
2726 | case st_C_struct: | |
2727 | case st_C_enum: | |
2728 | typdef = ttypeseen; | |
2729 | } | |
2730 | break; | |
2731 | case ttypeseen: | |
2732 | if (structdef == snone && fvdef == fvnone) | |
2733 | { | |
2734 | fvdef = fvnameseen; | |
96c06863 | 2735 | return true; |
05d9a399 FP |
2736 | } |
2737 | break; | |
2738 | case tend: | |
2739 | switch (toktype) | |
2740 | { | |
2741 | case st_C_class: | |
2742 | case st_C_struct: | |
2743 | case st_C_enum: | |
96c06863 | 2744 | return false; |
05d9a399 | 2745 | } |
96c06863 | 2746 | return true; |
05d9a399 FP |
2747 | } |
2748 | ||
05d9a399 FP |
2749 | switch (toktype) |
2750 | { | |
2751 | case st_C_javastruct: | |
2752 | if (structdef == stagseen) | |
2753 | structdef = scolonseen; | |
96c06863 | 2754 | return false; |
05d9a399 FP |
2755 | case st_C_template: |
2756 | case st_C_class: | |
2757 | if ((*c_extp & C_AUTO) /* automatic detection of C++ language */ | |
2758 | && bracelev == 0 | |
2759 | && definedef == dnone && structdef == snone | |
2760 | && typdef == tnone && fvdef == fvnone) | |
2761 | *c_extp = (*c_extp | C_PLPL) & ~C_AUTO; | |
2762 | if (toktype == st_C_template) | |
2763 | break; | |
2764 | /* FALLTHRU */ | |
2765 | case st_C_struct: | |
2766 | case st_C_enum: | |
2767 | if (parlev == 0 | |
2768 | && fvdef != vignore | |
2769 | && (typdef == tkeyseen | |
2770 | || (typedefs_or_cplusplus && structdef == snone))) | |
2771 | { | |
2772 | structdef = skeyseen; | |
2773 | structtype = toktype; | |
2774 | structbracelev = bracelev; | |
2775 | if (fvdef == fvnameseen) | |
2776 | fvdef = fvnone; | |
2777 | } | |
96c06863 | 2778 | return false; |
05d9a399 FP |
2779 | } |
2780 | ||
2781 | if (structdef == skeyseen) | |
2782 | { | |
2783 | structdef = stagseen; | |
96c06863 | 2784 | return true; |
05d9a399 FP |
2785 | } |
2786 | ||
2787 | if (typdef != tnone) | |
2788 | definedef = dnone; | |
2789 | ||
2790 | /* Detect Objective C constructs. */ | |
2791 | switch (objdef) | |
2792 | { | |
2793 | case onone: | |
2794 | switch (toktype) | |
2795 | { | |
2796 | case st_C_objprot: | |
2797 | objdef = oprotocol; | |
96c06863 | 2798 | return false; |
05d9a399 FP |
2799 | case st_C_objimpl: |
2800 | objdef = oimplementation; | |
96c06863 | 2801 | return false; |
05d9a399 FP |
2802 | } |
2803 | break; | |
2804 | case oimplementation: | |
2805 | /* Save the class tag for functions or variables defined inside. */ | |
2806 | objtag = savenstr (str, len); | |
2807 | objdef = oinbody; | |
96c06863 | 2808 | return false; |
05d9a399 FP |
2809 | case oprotocol: |
2810 | /* Save the class tag for categories. */ | |
2811 | objtag = savenstr (str, len); | |
2812 | objdef = otagseen; | |
96c06863 PE |
2813 | *is_func_or_var = true; |
2814 | return true; | |
05d9a399 FP |
2815 | case oparenseen: |
2816 | objdef = ocatseen; | |
96c06863 PE |
2817 | *is_func_or_var = true; |
2818 | return true; | |
05d9a399 FP |
2819 | case oinbody: |
2820 | break; | |
2821 | case omethodsign: | |
2822 | if (parlev == 0) | |
2823 | { | |
2824 | fvdef = fvnone; | |
2825 | objdef = omethodtag; | |
2826 | linebuffer_setlen (&token_name, len); | |
e99a530f | 2827 | memcpy (token_name.buffer, str, len); |
05d9a399 | 2828 | token_name.buffer[len] = '\0'; |
96c06863 | 2829 | return true; |
05d9a399 | 2830 | } |
96c06863 | 2831 | return false; |
05d9a399 FP |
2832 | case omethodcolon: |
2833 | if (parlev == 0) | |
2834 | objdef = omethodparm; | |
96c06863 | 2835 | return false; |
05d9a399 FP |
2836 | case omethodparm: |
2837 | if (parlev == 0) | |
2838 | { | |
e99a530f | 2839 | int oldlen = token_name.len; |
05d9a399 FP |
2840 | fvdef = fvnone; |
2841 | objdef = omethodtag; | |
e99a530f PE |
2842 | linebuffer_setlen (&token_name, oldlen + len); |
2843 | memcpy (token_name.buffer + oldlen, str, len); | |
5bf64749 | 2844 | token_name.buffer[oldlen + len] = '\0'; |
96c06863 | 2845 | return true; |
05d9a399 | 2846 | } |
96c06863 | 2847 | return false; |
05d9a399 FP |
2848 | case oignore: |
2849 | if (toktype == st_C_objend) | |
2850 | { | |
2851 | /* Memory leakage here: the string pointed by objtag is | |
2852 | never released, because many tests would be needed to | |
2853 | avoid breaking on incorrect input code. The amount of | |
2854 | memory leaked here is the sum of the lengths of the | |
2855 | class tags. | |
2856 | free (objtag); */ | |
2857 | objdef = onone; | |
2858 | } | |
96c06863 | 2859 | return false; |
05d9a399 FP |
2860 | } |
2861 | ||
2862 | /* A function, variable or enum constant? */ | |
2863 | switch (toktype) | |
2864 | { | |
2865 | case st_C_extern: | |
96c06863 | 2866 | fvextern = true; |
05d9a399 FP |
2867 | switch (fvdef) |
2868 | { | |
2869 | case finlist: | |
2870 | case flistseen: | |
2871 | case fignore: | |
2872 | case vignore: | |
2873 | break; | |
2874 | default: | |
2875 | fvdef = fvnone; | |
2876 | } | |
96c06863 | 2877 | return false; |
05d9a399 | 2878 | case st_C_ignore: |
96c06863 | 2879 | fvextern = false; |
05d9a399 | 2880 | fvdef = vignore; |
96c06863 | 2881 | return false; |
05d9a399 FP |
2882 | case st_C_operator: |
2883 | fvdef = foperator; | |
96c06863 PE |
2884 | *is_func_or_var = true; |
2885 | return true; | |
05d9a399 FP |
2886 | case st_none: |
2887 | if (constantypedefs | |
2888 | && structdef == snone | |
2889 | && structtype == st_C_enum && bracelev > structbracelev) | |
96c06863 | 2890 | return true; /* enum constant */ |
05d9a399 FP |
2891 | switch (fvdef) |
2892 | { | |
2893 | case fdefunkey: | |
2894 | if (bracelev > 0) | |
2895 | break; | |
2896 | fvdef = fdefunname; /* GNU macro */ | |
96c06863 PE |
2897 | *is_func_or_var = true; |
2898 | return true; | |
05d9a399 FP |
2899 | case fvnone: |
2900 | switch (typdef) | |
2901 | { | |
2902 | case ttypeseen: | |
96c06863 | 2903 | return false; |
05d9a399 FP |
2904 | case tnone: |
2905 | if ((strneq (str, "asm", 3) && endtoken (str[3])) | |
2906 | || (strneq (str, "__asm__", 7) && endtoken (str[7]))) | |
2907 | { | |
2908 | fvdef = vignore; | |
96c06863 | 2909 | return false; |
05d9a399 FP |
2910 | } |
2911 | break; | |
2912 | } | |
2913 | /* FALLTHRU */ | |
2914 | case fvnameseen: | |
2bc2e3fc | 2915 | if (len >= 10 && strneq (str+len-10, "::operator", 10)) |
8c463abe | 2916 | { |
9c485bbe FP |
2917 | if (*c_extp & C_AUTO) /* automatic detection of C++ */ |
2918 | *c_extp = (*c_extp | C_PLPL) & ~C_AUTO; | |
8c463abe | 2919 | fvdef = foperator; |
96c06863 PE |
2920 | *is_func_or_var = true; |
2921 | return true; | |
8c463abe | 2922 | } |
05d9a399 | 2923 | if (bracelev > 0 && !instruct) |
8c463abe | 2924 | break; |
30903246 | 2925 | fvdef = fvnameseen; /* function or variable */ |
96c06863 PE |
2926 | *is_func_or_var = true; |
2927 | return true; | |
6dd5561c | 2928 | } |
93b7ac65 | 2929 | break; |
6dd5561c FP |
2930 | } |
2931 | ||
96c06863 | 2932 | return false; |
6dd5561c FP |
2933 | } |
2934 | ||
b2521b0a | 2935 | \f |
c6d46f5f | 2936 | /* |
b2521b0a FP |
2937 | * C_entries often keeps pointers to tokens or lines which are older than |
2938 | * the line currently read. By keeping two line buffers, and switching | |
2939 | * them at end of line, it is possible to use those pointers. | |
c6d46f5f | 2940 | */ |
db590582 | 2941 | static struct |
b2521b0a FP |
2942 | { |
2943 | long linepos; | |
2944 | linebuffer lb; | |
2945 | } lbs[2]; | |
2946 | ||
55597f90 FP |
2947 | #define current_lb_is_new (newndx == curndx) |
2948 | #define switch_line_buffers() (curndx = 1 - curndx) | |
c6d46f5f | 2949 | |
13fde0cd | 2950 | #define curlb (lbs[curndx].lb) |
13fde0cd RS |
2951 | #define newlb (lbs[newndx].lb) |
2952 | #define curlinepos (lbs[curndx].linepos) | |
13fde0cd RS |
2953 | #define newlinepos (lbs[newndx].linepos) |
2954 | ||
9c485bbe FP |
2955 | #define plainc ((c_ext & C_EXT) == C_PLAIN) |
2956 | #define cplpl (c_ext & C_PLPL) | |
3c04a71a FP |
2957 | #define cjava ((c_ext & C_JAVA) == C_JAVA) |
2958 | ||
93b7ac65 | 2959 | #define CNL_SAVE_DEFINEDEF() \ |
13fde0cd | 2960 | do { \ |
55597f90 | 2961 | curlinepos = charno; \ |
e7d3b099 | 2962 | readline (&curlb, inf); \ |
13fde0cd | 2963 | lp = curlb.buffer; \ |
96c06863 | 2964 | quotednl = false; \ |
13fde0cd | 2965 | newndx = curndx; \ |
b9755a12 | 2966 | } while (0) |
c6d46f5f | 2967 | |
93b7ac65 | 2968 | #define CNL() \ |
13fde0cd | 2969 | do { \ |
93b7ac65 | 2970 | CNL_SAVE_DEFINEDEF(); \ |
b2521b0a | 2971 | if (savetoken.valid) \ |
55597f90 | 2972 | { \ |
b2521b0a | 2973 | token = savetoken; \ |
96c06863 | 2974 | savetoken.valid = false; \ |
55597f90 | 2975 | } \ |
c6d46f5f | 2976 | definedef = dnone; \ |
b9755a12 | 2977 | } while (0) |
13fde0cd | 2978 | |
086eac13 | 2979 | |
944bdd72 | 2980 | static void |
96c06863 | 2981 | make_C_tag (bool isfun) |
086eac13 | 2982 | { |
96c06863 | 2983 | /* This function is never called when token.valid is false, but |
086eac13 | 2984 | we must protect against invalid input or internal errors. */ |
9c485bbe | 2985 | if (token.valid) |
2201e3dc FP |
2986 | make_tag (token_name.buffer, token_name.len, isfun, token.line, |
2987 | token.offset+token.length+1, token.lineno, token.linepos); | |
89d8309f | 2988 | else if (DEBUG) |
e1dbe924 | 2989 | { /* this branch is optimized away if !DEBUG */ |
89d8309f FP |
2990 | make_tag (concat ("INVALID TOKEN:-->", token_name.buffer, ""), |
2991 | token_name.len + 17, isfun, token.line, | |
2992 | token.offset+token.length+1, token.lineno, token.linepos); | |
db5a3003 | 2993 | error ("INVALID TOKEN"); |
89d8309f | 2994 | } |
2201e3dc | 2995 | |
96c06863 | 2996 | token.valid = false; |
086eac13 FP |
2997 | } |
2998 | ||
c6d46f5f | 2999 | |
b2521b0a FP |
3000 | /* |
3001 | * C_entries () | |
3002 | * This routine finds functions, variables, typedefs, | |
3003 | * #define's, enum constants and struct/union/enum definitions in | |
3004 | * C syntax and adds them to the list. | |
3005 | */ | |
0dacfc4b | 3006 | static void |
873fbd0b DN |
3007 | C_entries (int c_ext, FILE *inf) |
3008 | /* extension of C */ | |
3009 | /* input file */ | |
c6d46f5f | 3010 | { |
13fde0cd | 3011 | register char c; /* latest char read; '\0' for end of line */ |
c6d46f5f | 3012 | register char *lp; /* pointer one beyond the character `c' */ |
13fde0cd | 3013 | int curndx, newndx; /* indices for current and new lb */ |
55597f90 FP |
3014 | register int tokoff; /* offset in line of start of current token */ |
3015 | register int toklen; /* length of current token */ | |
988e88ab | 3016 | const char *qualifier; /* string used to qualify names */ |
30903246 | 3017 | int qlen; /* length of qualifier */ |
05d9a399 FP |
3018 | int bracelev; /* current brace level */ |
3019 | int bracketlev; /* current bracket level */ | |
b12756c8 | 3020 | int parlev; /* current parenthesis level */ |
05d9a399 FP |
3021 | int attrparlev; /* __attribute__ parenthesis level */ |
3022 | int templatelev; /* current template level */ | |
3023 | int typdefbracelev; /* bracelev where a typedef struct body begun */ | |
30903246 | 3024 | bool incomm, inquote, inchar, quotednl, midtoken; |
b2521b0a | 3025 | bool yacc_rules; /* in the rules part of a yacc file */ |
35a4c758 | 3026 | struct tok savetoken = {0}; /* token saved during preprocessor handling */ |
c6d46f5f | 3027 | |
75bdbc6a | 3028 | |
61a1f6fa FP |
3029 | linebuffer_init (&lbs[0].lb); |
3030 | linebuffer_init (&lbs[1].lb); | |
8c463abe FP |
3031 | if (cstack.size == 0) |
3032 | { | |
3033 | cstack.size = (DEBUG) ? 1 : 4; | |
3034 | cstack.nl = 0; | |
3035 | cstack.cname = xnew (cstack.size, char *); | |
05d9a399 | 3036 | cstack.bracelev = xnew (cstack.size, int); |
8c463abe | 3037 | } |
b2521b0a | 3038 | |
05d9a399 | 3039 | tokoff = toklen = typdefbracelev = 0; /* keep compiler quiet */ |
13fde0cd | 3040 | curndx = newndx = 0; |
13fde0cd | 3041 | lp = curlb.buffer; |
c6d46f5f JB |
3042 | *lp = 0; |
3043 | ||
96c06863 | 3044 | fvdef = fvnone; fvextern = false; typdef = tnone; |
93b7ac65 | 3045 | structdef = snone; definedef = dnone; objdef = onone; |
96c06863 PE |
3046 | yacc_rules = false; |
3047 | midtoken = inquote = inchar = incomm = quotednl = false; | |
3048 | token.valid = savetoken.valid = false; | |
05d9a399 | 3049 | bracelev = bracketlev = parlev = attrparlev = templatelev = 0; |
30903246 FP |
3050 | if (cjava) |
3051 | { qualifier = "."; qlen = 1; } | |
3052 | else | |
3053 | { qualifier = "::"; qlen = 2; } | |
c6d46f5f | 3054 | |
b2521b0a | 3055 | |
c6d46f5f JB |
3056 | while (!feof (inf)) |
3057 | { | |
3058 | c = *lp++; | |
c6d46f5f JB |
3059 | if (c == '\\') |
3060 | { | |
05d9a399 FP |
3061 | /* If we are at the end of the line, the next character is a |
3062 | '\0'; do not skip it, because it is what tells us | |
4746118a | 3063 | to read the next line. */ |
13fde0cd | 3064 | if (*lp == '\0') |
99e0a2e0 | 3065 | { |
96c06863 | 3066 | quotednl = true; |
99e0a2e0 RS |
3067 | continue; |
3068 | } | |
1e134a5f | 3069 | lp++; |
c6d46f5f JB |
3070 | c = ' '; |
3071 | } | |
3072 | else if (incomm) | |
3073 | { | |
13fde0cd | 3074 | switch (c) |
c6d46f5f | 3075 | { |
13fde0cd RS |
3076 | case '*': |
3077 | if (*lp == '/') | |
3078 | { | |
3079 | c = *lp++; | |
96c06863 | 3080 | incomm = false; |
13fde0cd RS |
3081 | } |
3082 | break; | |
3083 | case '\0': | |
3084 | /* Newlines inside comments do not end macro definitions in | |
3085 | traditional cpp. */ | |
93b7ac65 | 3086 | CNL_SAVE_DEFINEDEF (); |
13fde0cd | 3087 | break; |
c6d46f5f | 3088 | } |
13fde0cd | 3089 | continue; |
c6d46f5f JB |
3090 | } |
3091 | else if (inquote) | |
3092 | { | |
13fde0cd RS |
3093 | switch (c) |
3094 | { | |
3095 | case '"': | |
96c06863 | 3096 | inquote = false; |
13fde0cd RS |
3097 | break; |
3098 | case '\0': | |
42680d3c | 3099 | /* Newlines inside strings do not end macro definitions |
13fde0cd RS |
3100 | in traditional cpp, even though compilers don't |
3101 | usually accept them. */ | |
93b7ac65 | 3102 | CNL_SAVE_DEFINEDEF (); |
13fde0cd RS |
3103 | break; |
3104 | } | |
3105 | continue; | |
c6d46f5f JB |
3106 | } |
3107 | else if (inchar) | |
3108 | { | |
42680d3c FP |
3109 | switch (c) |
3110 | { | |
3111 | case '\0': | |
3112 | /* Hmmm, something went wrong. */ | |
93b7ac65 | 3113 | CNL (); |
42680d3c FP |
3114 | /* FALLTHRU */ |
3115 | case '\'': | |
96c06863 | 3116 | inchar = false; |
42680d3c FP |
3117 | break; |
3118 | } | |
c6d46f5f JB |
3119 | continue; |
3120 | } | |
05d9a399 FP |
3121 | else switch (c) |
3122 | { | |
3123 | case '"': | |
96c06863 | 3124 | inquote = true; |
28796b3a AS |
3125 | if (bracketlev > 0) |
3126 | continue; | |
05d9a399 FP |
3127 | if (inattribute) |
3128 | break; | |
3129 | switch (fvdef) | |
3130 | { | |
3131 | case fdefunkey: | |
3132 | case fstartlist: | |
3133 | case finlist: | |
3134 | case fignore: | |
3135 | case vignore: | |
3136 | break; | |
3137 | default: | |
96c06863 | 3138 | fvextern = false; |
05d9a399 FP |
3139 | fvdef = fvnone; |
3140 | } | |
3141 | continue; | |
3142 | case '\'': | |
96c06863 | 3143 | inchar = true; |
28796b3a AS |
3144 | if (bracketlev > 0) |
3145 | continue; | |
05d9a399 FP |
3146 | if (inattribute) |
3147 | break; | |
28796b3a | 3148 | if (fvdef != finlist && fvdef != fignore && fvdef != vignore) |
05d9a399 | 3149 | { |
96c06863 | 3150 | fvextern = false; |
05d9a399 FP |
3151 | fvdef = fvnone; |
3152 | } | |
3153 | continue; | |
3154 | case '/': | |
3155 | if (*lp == '*') | |
3156 | { | |
96c06863 | 3157 | incomm = true; |
eac50785 FP |
3158 | lp++; |
3159 | c = ' '; | |
28796b3a AS |
3160 | if (bracketlev > 0) |
3161 | continue; | |
05d9a399 FP |
3162 | } |
3163 | else if (/* cplpl && */ *lp == '/') | |
3164 | { | |
3165 | c = '\0'; | |
05d9a399 | 3166 | } |
eac50785 | 3167 | break; |
05d9a399 FP |
3168 | case '%': |
3169 | if ((c_ext & YACC) && *lp == '%') | |
3170 | { | |
3171 | /* Entering or exiting rules section in yacc file. */ | |
3172 | lp++; | |
96c06863 | 3173 | definedef = dnone; fvdef = fvnone; fvextern = false; |
05d9a399 | 3174 | typdef = tnone; structdef = snone; |
96c06863 | 3175 | midtoken = inquote = inchar = incomm = quotednl = false; |
05d9a399 FP |
3176 | bracelev = 0; |
3177 | yacc_rules = !yacc_rules; | |
3178 | continue; | |
3179 | } | |
3180 | else | |
3181 | break; | |
3182 | case '#': | |
3183 | if (definedef == dnone) | |
3184 | { | |
3185 | char *cp; | |
96c06863 | 3186 | bool cpptoken = true; |
05d9a399 FP |
3187 | |
3188 | /* Look back on this line. If all blanks, or nonblanks | |
3189 | followed by an end of comment, this is a preprocessor | |
3190 | token. */ | |
3191 | for (cp = newlb.buffer; cp < lp-1; cp++) | |
3192 | if (!iswhite (*cp)) | |
3193 | { | |
28796b3a | 3194 | if (*cp == '*' && cp[1] == '/') |
05d9a399 FP |
3195 | { |
3196 | cp++; | |
96c06863 | 3197 | cpptoken = true; |
05d9a399 FP |
3198 | } |
3199 | else | |
96c06863 | 3200 | cpptoken = false; |
05d9a399 FP |
3201 | } |
3202 | if (cpptoken) | |
3203 | definedef = dsharpseen; | |
3204 | } /* if (definedef == dnone) */ | |
3205 | continue; | |
3206 | case '[': | |
3207 | bracketlev++; | |
28796b3a AS |
3208 | continue; |
3209 | default: | |
3210 | if (bracketlev > 0) | |
3211 | { | |
3212 | if (c == ']') | |
3213 | --bracketlev; | |
3214 | else if (c == '\0') | |
3215 | CNL_SAVE_DEFINEDEF (); | |
3216 | continue; | |
3217 | } | |
3218 | break; | |
05d9a399 | 3219 | } /* switch (c) */ |
c6d46f5f | 3220 | |
c6d46f5f | 3221 | |
8c463abe | 3222 | /* Consider token only if some involved conditions are satisfied. */ |
b2521b0a | 3223 | if (typdef != tignore |
13fde0cd | 3224 | && definedef != dignorerest |
b2521b0a | 3225 | && fvdef != finlist |
05d9a399 | 3226 | && templatelev == 0 |
b2521b0a | 3227 | && (definedef != dnone |
05d9a399 FP |
3228 | || structdef != scolonseen) |
3229 | && !inattribute) | |
c6d46f5f JB |
3230 | { |
3231 | if (midtoken) | |
3232 | { | |
3233 | if (endtoken (c)) | |
3234 | { | |
9c485bbe FP |
3235 | if (c == ':' && *lp == ':' && begtoken (lp[1])) |
3236 | /* This handles :: in the middle, | |
3237 | but not at the beginning of an identifier. | |
e1dbe924 | 3238 | Also, space-separated :: is not recognized. */ |
c6d46f5f | 3239 | { |
9c485bbe FP |
3240 | if (c_ext & C_AUTO) /* automatic detection of C++ */ |
3241 | c_ext = (c_ext | C_PLPL) & ~C_AUTO; | |
c6d46f5f | 3242 | lp += 2; |
93b7ac65 FP |
3243 | toklen += 2; |
3244 | c = lp[-1]; | |
8f79fe72 | 3245 | goto still_in_token; |
c6d46f5f JB |
3246 | } |
3247 | else | |
3248 | { | |
96c06863 | 3249 | bool funorvar = false; |
b2521b0a | 3250 | |
13fde0cd | 3251 | if (yacc_rules |
d8913c1c | 3252 | || consider_token (newlb.buffer + tokoff, toklen, c, |
05d9a399 FP |
3253 | &c_ext, bracelev, parlev, |
3254 | &funorvar)) | |
c6d46f5f | 3255 | { |
93b7ac65 FP |
3256 | if (fvdef == foperator) |
3257 | { | |
3258 | char *oldlp = lp; | |
3259 | lp = skip_spaces (lp-1); | |
3260 | if (*lp != '\0') | |
3261 | lp += 1; | |
3262 | while (*lp != '\0' | |
71cbb895 | 3263 | && !iswhite (*lp) && *lp != '(') |
93b7ac65 FP |
3264 | lp += 1; |
3265 | c = *lp++; | |
3266 | toklen += lp - oldlp; | |
3267 | } | |
96c06863 | 3268 | token.named = false; |
9c485bbe | 3269 | if (!plainc |
8c463abe FP |
3270 | && nestlev > 0 && definedef == dnone) |
3271 | /* in struct body */ | |
fe0b3356 | 3272 | { |
e99a530f | 3273 | int len; |
8c463abe | 3274 | write_classname (&token_name, qualifier); |
e99a530f PE |
3275 | len = token_name.len; |
3276 | linebuffer_setlen (&token_name, len+qlen+toklen); | |
3277 | sprintf (token_name.buffer + len, "%s%.*s", | |
3278 | qualifier, toklen, newlb.buffer + tokoff); | |
96c06863 | 3279 | token.named = true; |
c6d46f5f | 3280 | } |
d8913c1c FP |
3281 | else if (objdef == ocatseen) |
3282 | /* Objective C category */ | |
3283 | { | |
30903246 | 3284 | int len = strlen (objtag) + 2 + toklen; |
b2521b0a | 3285 | linebuffer_setlen (&token_name, len); |
e99a530f PE |
3286 | sprintf (token_name.buffer, "%s(%.*s)", |
3287 | objtag, toklen, newlb.buffer + tokoff); | |
96c06863 | 3288 | token.named = true; |
d8913c1c FP |
3289 | } |
3290 | else if (objdef == omethodtag | |
3291 | || objdef == omethodparm) | |
3292 | /* Objective C method */ | |
3293 | { | |
96c06863 | 3294 | token.named = true; |
b2521b0a FP |
3295 | } |
3296 | else if (fvdef == fdefunname) | |
8c463abe | 3297 | /* GNU DEFUN and similar macros */ |
b2521b0a FP |
3298 | { |
3299 | bool defun = (newlb.buffer[tokoff] == 'F'); | |
3300 | int off = tokoff; | |
3301 | int len = toklen; | |
3302 | ||
3303 | /* Rewrite the tag so that emacs lisp DEFUNs | |
3304 | can be found by their elisp name */ | |
3305 | if (defun) | |
3306 | { | |
3307 | off += 1; | |
3308 | len -= 1; | |
3309 | } | |
b2521b0a | 3310 | linebuffer_setlen (&token_name, len); |
e99a530f PE |
3311 | memcpy (token_name.buffer, |
3312 | newlb.buffer + off, len); | |
b2521b0a FP |
3313 | token_name.buffer[len] = '\0'; |
3314 | if (defun) | |
3315 | while (--len >= 0) | |
3316 | if (token_name.buffer[len] == '_') | |
3317 | token_name.buffer[len] = '-'; | |
3318 | token.named = defun; | |
d8913c1c | 3319 | } |
c6d46f5f JB |
3320 | else |
3321 | { | |
b2521b0a | 3322 | linebuffer_setlen (&token_name, toklen); |
e99a530f PE |
3323 | memcpy (token_name.buffer, |
3324 | newlb.buffer + tokoff, toklen); | |
75bdbc6a | 3325 | token_name.buffer[toklen] = '\0'; |
93b7ac65 | 3326 | /* Name macros and members. */ |
b2521b0a FP |
3327 | token.named = (structdef == stagseen |
3328 | || typdef == ttypeseen | |
3329 | || typdef == tend | |
3330 | || (funorvar | |
3331 | && definedef == dignorerest) | |
3332 | || (funorvar | |
3333 | && definedef == dnone | |
8c463abe | 3334 | && structdef == snone |
05d9a399 | 3335 | && bracelev > 0)); |
c6d46f5f | 3336 | } |
b2521b0a | 3337 | token.lineno = lineno; |
8c463abe FP |
3338 | token.offset = tokoff; |
3339 | token.length = toklen; | |
b2521b0a FP |
3340 | token.line = newlb.buffer; |
3341 | token.linepos = newlinepos; | |
96c06863 | 3342 | token.valid = true; |
fe0b3356 | 3343 | |
b12756c8 | 3344 | if (definedef == dnone |
30903246 | 3345 | && (fvdef == fvnameseen |
93b7ac65 | 3346 | || fvdef == foperator |
b12756c8 | 3347 | || structdef == stagseen |
d8913c1c | 3348 | || typdef == tend |
8c463abe | 3349 | || typdef == ttypeseen |
d8913c1c | 3350 | || objdef != onone)) |
13fde0cd | 3351 | { |
55597f90 FP |
3352 | if (current_lb_is_new) |
3353 | switch_line_buffers (); | |
13fde0cd | 3354 | } |
8c463abe FP |
3355 | else if (definedef != dnone |
3356 | || fvdef == fdefunname | |
3357 | || instruct) | |
30903246 | 3358 | make_C_tag (funorvar); |
c6d46f5f | 3359 | } |
05d9a399 FP |
3360 | else /* not yacc and consider_token failed */ |
3361 | { | |
3362 | if (inattribute && fvdef == fignore) | |
3363 | { | |
3364 | /* We have just met __attribute__ after a | |
3365 | function parameter list: do not tag the | |
3366 | function again. */ | |
3367 | fvdef = fvnone; | |
3368 | } | |
3369 | } | |
96c06863 | 3370 | midtoken = false; |
c6d46f5f | 3371 | } |
13fde0cd | 3372 | } /* if (endtoken (c)) */ |
c6d46f5f | 3373 | else if (intoken (c)) |
8f79fe72 | 3374 | still_in_token: |
13fde0cd RS |
3375 | { |
3376 | toklen++; | |
3377 | continue; | |
3378 | } | |
3379 | } /* if (midtoken) */ | |
c6d46f5f JB |
3380 | else if (begtoken (c)) |
3381 | { | |
b12756c8 | 3382 | switch (definedef) |
13fde0cd | 3383 | { |
b12756c8 | 3384 | case dnone: |
30903246 | 3385 | switch (fvdef) |
b12756c8 FP |
3386 | { |
3387 | case fstartlist: | |
05d9a399 FP |
3388 | /* This prevents tagging fb in |
3389 | void (__attribute__((noreturn)) *fb) (void); | |
3390 | Fixing this is not easy and not very important. */ | |
30903246 | 3391 | fvdef = finlist; |
b12756c8 FP |
3392 | continue; |
3393 | case flistseen: | |
9c485bbe FP |
3394 | if (plainc || declarations) |
3395 | { | |
96c06863 | 3396 | make_C_tag (true); /* a function */ |
9c485bbe FP |
3397 | fvdef = fignore; |
3398 | } | |
b12756c8 | 3399 | break; |
b12756c8 | 3400 | } |
cec68fcb | 3401 | if (structdef == stagseen && !cjava) |
8c463abe | 3402 | { |
05d9a399 | 3403 | popclass_above (bracelev); |
8c463abe FP |
3404 | structdef = snone; |
3405 | } | |
13fde0cd | 3406 | break; |
b12756c8 | 3407 | case dsharpseen: |
b2521b0a | 3408 | savetoken = token; |
f6566f90 | 3409 | break; |
13fde0cd | 3410 | } |
13fde0cd RS |
3411 | if (!yacc_rules || lp == newlb.buffer + 1) |
3412 | { | |
3413 | tokoff = lp - 1 - newlb.buffer; | |
3414 | toklen = 1; | |
96c06863 | 3415 | midtoken = true; |
13fde0cd RS |
3416 | } |
3417 | continue; | |
4b533b5b | 3418 | } /* if (begtoken) */ |
13fde0cd RS |
3419 | } /* if must look at token */ |
3420 | ||
3421 | ||
3422 | /* Detect end of line, colon, comma, semicolon and various braces | |
b12756c8 | 3423 | after having handled a token.*/ |
13fde0cd | 3424 | switch (c) |
1e134a5f | 3425 | { |
13fde0cd | 3426 | case ':': |
05d9a399 FP |
3427 | if (inattribute) |
3428 | break; | |
8c463abe FP |
3429 | if (yacc_rules && token.offset == 0 && token.valid) |
3430 | { | |
96c06863 | 3431 | make_C_tag (false); /* a yacc function */ |
8c463abe FP |
3432 | break; |
3433 | } | |
b12756c8 FP |
3434 | if (definedef != dnone) |
3435 | break; | |
d8913c1c FP |
3436 | switch (objdef) |
3437 | { | |
3438 | case otagseen: | |
3439 | objdef = oignore; | |
96c06863 | 3440 | make_C_tag (true); /* an Objective C class */ |
d8913c1c FP |
3441 | break; |
3442 | case omethodtag: | |
3443 | case omethodparm: | |
3444 | objdef = omethodcolon; | |
b2521b0a | 3445 | linebuffer_setlen (&token_name, token_name.len + 1); |
d8913c1c FP |
3446 | strcat (token_name.buffer, ":"); |
3447 | break; | |
3448 | } | |
13fde0cd | 3449 | if (structdef == stagseen) |
3c04a71a FP |
3450 | { |
3451 | structdef = scolonseen; | |
3452 | break; | |
3453 | } | |
9c485bbe | 3454 | /* Should be useless, but may be work as a safety net. */ |
3c04a71a FP |
3455 | if (cplpl && fvdef == flistseen) |
3456 | { | |
96c06863 | 3457 | make_C_tag (true); /* a function */ |
3c04a71a FP |
3458 | fvdef = fignore; |
3459 | break; | |
3460 | } | |
13fde0cd RS |
3461 | break; |
3462 | case ';': | |
05d9a399 | 3463 | if (definedef != dnone || inattribute) |
b12756c8 | 3464 | break; |
8c463abe | 3465 | switch (typdef) |
1f638249 | 3466 | { |
8c463abe FP |
3467 | case tend: |
3468 | case ttypeseen: | |
96c06863 | 3469 | make_C_tag (false); /* a typedef */ |
8c463abe | 3470 | typdef = tnone; |
93b7ac65 | 3471 | fvdef = fvnone; |
93b7ac65 | 3472 | break; |
8c463abe FP |
3473 | case tnone: |
3474 | case tinbody: | |
3475 | case tignore: | |
3476 | switch (fvdef) | |
3477 | { | |
3478 | case fignore: | |
3c04a71a | 3479 | if (typdef == tignore || cplpl) |
8c463abe FP |
3480 | fvdef = fvnone; |
3481 | break; | |
3482 | case fvnameseen: | |
05d9a399 | 3483 | if ((globals && bracelev == 0 && (!fvextern || declarations)) |
8c463abe | 3484 | || (members && instruct)) |
96c06863 PE |
3485 | make_C_tag (false); /* a variable */ |
3486 | fvextern = false; | |
8c463abe | 3487 | fvdef = fvnone; |
96c06863 | 3488 | token.valid = false; |
8c463abe FP |
3489 | break; |
3490 | case flistseen: | |
05d9a399 FP |
3491 | if ((declarations |
3492 | && (cplpl || !instruct) | |
3493 | && (typdef == tnone || (typdef != tignore && instruct))) | |
3494 | || (members | |
3495 | && plainc && instruct)) | |
96c06863 | 3496 | make_C_tag (true); /* a function */ |
8c463abe FP |
3497 | /* FALLTHRU */ |
3498 | default: | |
96c06863 | 3499 | fvextern = false; |
8c463abe FP |
3500 | fvdef = fvnone; |
3501 | if (declarations | |
9c485bbe | 3502 | && cplpl && structdef == stagseen) |
96c06863 | 3503 | make_C_tag (false); /* forward declaration */ |
8c463abe | 3504 | else |
96c06863 | 3505 | token.valid = false; |
8c463abe | 3506 | } /* switch (fvdef) */ |
30903246 FP |
3507 | /* FALLTHRU */ |
3508 | default: | |
8c463abe | 3509 | if (!instruct) |
b2521b0a | 3510 | typdef = tnone; |
8c463abe | 3511 | } |
46c145db FP |
3512 | if (structdef == stagseen) |
3513 | structdef = snone; | |
3514 | break; | |
13fde0cd | 3515 | case ',': |
05d9a399 | 3516 | if (definedef != dnone || inattribute) |
46c145db | 3517 | break; |
d8913c1c FP |
3518 | switch (objdef) |
3519 | { | |
3520 | case omethodtag: | |
3521 | case omethodparm: | |
96c06863 | 3522 | make_C_tag (true); /* an Objective C method */ |
d8913c1c FP |
3523 | objdef = oinbody; |
3524 | break; | |
3525 | } | |
30903246 FP |
3526 | switch (fvdef) |
3527 | { | |
b2521b0a | 3528 | case fdefunkey: |
93b7ac65 | 3529 | case foperator: |
8c463abe | 3530 | case fstartlist: |
30903246 FP |
3531 | case finlist: |
3532 | case fignore: | |
3533 | case vignore: | |
3534 | break; | |
b2521b0a FP |
3535 | case fdefunname: |
3536 | fvdef = fignore; | |
30903246 | 3537 | break; |
05d9a399 FP |
3538 | case fvnameseen: |
3539 | if (parlev == 0 | |
3540 | && ((globals | |
3541 | && bracelev == 0 | |
3542 | && templatelev == 0 | |
3543 | && (!fvextern || declarations)) | |
3544 | || (members && instruct))) | |
96c06863 | 3545 | make_C_tag (false); /* a variable */ |
8c463abe | 3546 | break; |
05d9a399 | 3547 | case flistseen: |
8c463abe FP |
3548 | if ((declarations && typdef == tnone && !instruct) |
3549 | || (members && typdef != tignore && instruct)) | |
b2521b0a | 3550 | { |
96c06863 | 3551 | make_C_tag (true); /* a function */ |
8c463abe | 3552 | fvdef = fvnameseen; |
b2521b0a | 3553 | } |
8c463abe FP |
3554 | else if (!declarations) |
3555 | fvdef = fvnone; | |
96c06863 | 3556 | token.valid = false; |
8c463abe | 3557 | break; |
30903246 FP |
3558 | default: |
3559 | fvdef = fvnone; | |
3560 | } | |
46c145db FP |
3561 | if (structdef == stagseen) |
3562 | structdef = snone; | |
3563 | break; | |
05d9a399 FP |
3564 | case ']': |
3565 | if (definedef != dnone || inattribute) | |
b12756c8 | 3566 | break; |
8c463abe FP |
3567 | if (structdef == stagseen) |
3568 | structdef = snone; | |
3569 | switch (typdef) | |
46c145db | 3570 | { |
8c463abe FP |
3571 | case ttypeseen: |
3572 | case tend: | |
46c145db | 3573 | typdef = tignore; |
96c06863 | 3574 | make_C_tag (false); /* a typedef */ |
46c145db | 3575 | break; |
8c463abe FP |
3576 | case tnone: |
3577 | case tinbody: | |
3578 | switch (fvdef) | |
3579 | { | |
3580 | case foperator: | |
3581 | case finlist: | |
3582 | case fignore: | |
3583 | case vignore: | |
3584 | break; | |
3585 | case fvnameseen: | |
05d9a399 FP |
3586 | if ((members && bracelev == 1) |
3587 | || (globals && bracelev == 0 | |
8c463abe | 3588 | && (!fvextern || declarations))) |
96c06863 | 3589 | make_C_tag (false); /* a variable */ |
8c463abe FP |
3590 | /* FALLTHRU */ |
3591 | default: | |
3592 | fvdef = fvnone; | |
3593 | } | |
30903246 | 3594 | break; |
30903246 | 3595 | } |
13fde0cd RS |
3596 | break; |
3597 | case '(': | |
05d9a399 FP |
3598 | if (inattribute) |
3599 | { | |
3600 | attrparlev++; | |
3601 | break; | |
3602 | } | |
b12756c8 FP |
3603 | if (definedef != dnone) |
3604 | break; | |
d8913c1c FP |
3605 | if (objdef == otagseen && parlev == 0) |
3606 | objdef = oparenseen; | |
30903246 | 3607 | switch (fvdef) |
57e83cfe | 3608 | { |
30903246 | 3609 | case fvnameseen: |
93b7ac65 | 3610 | if (typdef == ttypeseen |
93b7ac65 | 3611 | && *lp != '*' |
8c463abe | 3612 | && !instruct) |
93b7ac65 FP |
3613 | { |
3614 | /* This handles constructs like: | |
3615 | typedef void OperatorFun (int fun); */ | |
96c06863 | 3616 | make_C_tag (false); |
93b7ac65 | 3617 | typdef = tignore; |
8c463abe FP |
3618 | fvdef = fignore; |
3619 | break; | |
93b7ac65 FP |
3620 | } |
3621 | /* FALLTHRU */ | |
3622 | case foperator: | |
3623 | fvdef = fstartlist; | |
13fde0cd | 3624 | break; |
13fde0cd | 3625 | case flistseen: |
30903246 | 3626 | fvdef = finlist; |
13fde0cd | 3627 | break; |
57e83cfe | 3628 | } |
b12756c8 | 3629 | parlev++; |
13fde0cd RS |
3630 | break; |
3631 | case ')': | |
05d9a399 FP |
3632 | if (inattribute) |
3633 | { | |
3634 | if (--attrparlev == 0) | |
96c06863 | 3635 | inattribute = false; |
05d9a399 FP |
3636 | break; |
3637 | } | |
b12756c8 FP |
3638 | if (definedef != dnone) |
3639 | break; | |
d8913c1c FP |
3640 | if (objdef == ocatseen && parlev == 1) |
3641 | { | |
96c06863 | 3642 | make_C_tag (true); /* an Objective C category */ |
d8913c1c FP |
3643 | objdef = oignore; |
3644 | } | |
b12756c8 FP |
3645 | if (--parlev == 0) |
3646 | { | |
30903246 | 3647 | switch (fvdef) |
b12756c8 FP |
3648 | { |
3649 | case fstartlist: | |
3650 | case finlist: | |
30903246 | 3651 | fvdef = flistseen; |
b12756c8 FP |
3652 | break; |
3653 | } | |
8c463abe FP |
3654 | if (!instruct |
3655 | && (typdef == tend | |
3656 | || typdef == ttypeseen)) | |
46c145db FP |
3657 | { |
3658 | typdef = tignore; | |
96c06863 | 3659 | make_C_tag (false); /* a typedef */ |
46c145db | 3660 | } |
b12756c8 FP |
3661 | } |
3662 | else if (parlev < 0) /* can happen due to ill-conceived #if's. */ | |
3663 | parlev = 0; | |
13fde0cd RS |
3664 | break; |
3665 | case '{': | |
b12756c8 FP |
3666 | if (definedef != dnone) |
3667 | break; | |
93b7ac65 | 3668 | if (typdef == ttypeseen) |
8c463abe | 3669 | { |
f55ae599 | 3670 | /* Whenever typdef is set to tinbody (currently only |
05d9a399 | 3671 | here), typdefbracelev should be set to bracelev. */ |
8c463abe | 3672 | typdef = tinbody; |
05d9a399 | 3673 | typdefbracelev = bracelev; |
8c463abe | 3674 | } |
30903246 | 3675 | switch (fvdef) |
31d4b314 FP |
3676 | { |
3677 | case flistseen: | |
96c06863 | 3678 | make_C_tag (true); /* a function */ |
31d4b314 FP |
3679 | /* FALLTHRU */ |
3680 | case fignore: | |
30903246 | 3681 | fvdef = fvnone; |
46c145db | 3682 | break; |
30903246 | 3683 | case fvnone: |
d8913c1c FP |
3684 | switch (objdef) |
3685 | { | |
3686 | case otagseen: | |
96c06863 | 3687 | make_C_tag (true); /* an Objective C class */ |
d8913c1c FP |
3688 | objdef = oignore; |
3689 | break; | |
3690 | case omethodtag: | |
3691 | case omethodparm: | |
96c06863 | 3692 | make_C_tag (true); /* an Objective C method */ |
d8913c1c FP |
3693 | objdef = oinbody; |
3694 | break; | |
3695 | default: | |
3ca80e28 | 3696 | /* Neutralize `extern "C" {' grot. */ |
05d9a399 | 3697 | if (bracelev == 0 && structdef == snone && nestlev == 0 |
8c463abe | 3698 | && typdef == tnone) |
05d9a399 | 3699 | bracelev = -1; |
d8913c1c | 3700 | } |
f6566f90 | 3701 | break; |
31d4b314 | 3702 | } |
b2521b0a FP |
3703 | switch (structdef) |
3704 | { | |
3705 | case skeyseen: /* unnamed struct */ | |
05d9a399 | 3706 | pushclass_above (bracelev, NULL, 0); |
8c463abe | 3707 | structdef = snone; |
b2521b0a | 3708 | break; |
8c463abe FP |
3709 | case stagseen: /* named struct or enum */ |
3710 | case scolonseen: /* a class */ | |
05d9a399 | 3711 | pushclass_above (bracelev,token.line+token.offset, token.length); |
8c463abe | 3712 | structdef = snone; |
96c06863 | 3713 | make_C_tag (false); /* a struct or enum */ |
b2521b0a FP |
3714 | break; |
3715 | } | |
89d8309f | 3716 | bracelev += 1; |
31d4b314 | 3717 | break; |
13fde0cd | 3718 | case '*': |
b12756c8 FP |
3719 | if (definedef != dnone) |
3720 | break; | |
30903246 | 3721 | if (fvdef == fstartlist) |
9c485bbe FP |
3722 | { |
3723 | fvdef = fvnone; /* avoid tagging `foo' in `foo (*bar()) ()' */ | |
96c06863 | 3724 | token.valid = false; |
9c485bbe | 3725 | } |
13fde0cd RS |
3726 | break; |
3727 | case '}': | |
b12756c8 FP |
3728 | if (definedef != dnone) |
3729 | break; | |
89d8309f | 3730 | bracelev -= 1; |
3c04a71a | 3731 | if (!ignoreindent && lp == newlb.buffer + 1) |
b12756c8 | 3732 | { |
05d9a399 | 3733 | if (bracelev != 0) |
96c06863 | 3734 | token.valid = false; /* unexpected value, token unreliable */ |
05d9a399 | 3735 | bracelev = 0; /* reset brace level if first column */ |
b12756c8 FP |
3736 | parlev = 0; /* also reset paren level, just in case... */ |
3737 | } | |
89d8309f | 3738 | else if (bracelev < 0) |
172aa4c1 | 3739 | { |
96c06863 | 3740 | token.valid = false; /* something gone amiss, token unreliable */ |
89d8309f | 3741 | bracelev = 0; |
172aa4c1 | 3742 | } |
f14a3ad4 FP |
3743 | if (bracelev == 0 && fvdef == vignore) |
3744 | fvdef = fvnone; /* end of function */ | |
05d9a399 | 3745 | popclass_above (bracelev); |
8c463abe | 3746 | structdef = snone; |
05d9a399 FP |
3747 | /* Only if typdef == tinbody is typdefbracelev significant. */ |
3748 | if (typdef == tinbody && bracelev <= typdefbracelev) | |
13fde0cd | 3749 | { |
05d9a399 | 3750 | assert (bracelev == typdefbracelev); |
8c463abe | 3751 | typdef = tend; |
13fde0cd RS |
3752 | } |
3753 | break; | |
30903246 FP |
3754 | case '=': |
3755 | if (definedef != dnone) | |
3756 | break; | |
3757 | switch (fvdef) | |
3758 | { | |
93b7ac65 | 3759 | case foperator: |
30903246 FP |
3760 | case finlist: |
3761 | case fignore: | |
3762 | case vignore: | |
3763 | break; | |
3764 | case fvnameseen: | |
05d9a399 FP |
3765 | if ((members && bracelev == 1) |
3766 | || (globals && bracelev == 0 && (!fvextern || declarations))) | |
96c06863 | 3767 | make_C_tag (false); /* a variable */ |
30903246 FP |
3768 | /* FALLTHRU */ |
3769 | default: | |
3770 | fvdef = vignore; | |
3771 | } | |
3772 | break; | |
8c463abe | 3773 | case '<': |
05d9a399 FP |
3774 | if (cplpl |
3775 | && (structdef == stagseen || fvdef == fvnameseen)) | |
8c463abe | 3776 | { |
05d9a399 | 3777 | templatelev++; |
8c463abe FP |
3778 | break; |
3779 | } | |
3780 | goto resetfvdef; | |
3781 | case '>': | |
05d9a399 | 3782 | if (templatelev > 0) |
8c463abe | 3783 | { |
05d9a399 | 3784 | templatelev--; |
8c463abe FP |
3785 | break; |
3786 | } | |
3787 | goto resetfvdef; | |
d8913c1c FP |
3788 | case '+': |
3789 | case '-': | |
05d9a399 | 3790 | if (objdef == oinbody && bracelev == 0) |
d8913c1c FP |
3791 | { |
3792 | objdef = omethodsign; | |
3793 | break; | |
3794 | } | |
3795 | /* FALLTHRU */ | |
8c463abe | 3796 | resetfvdef: |
05d9a399 FP |
3797 | case '#': case '~': case '&': case '%': case '/': |
3798 | case '|': case '^': case '!': case '.': case '?': | |
b12756c8 FP |
3799 | if (definedef != dnone) |
3800 | break; | |
93b7ac65 FP |
3801 | /* These surely cannot follow a function tag in C. */ |
3802 | switch (fvdef) | |
3803 | { | |
3804 | case foperator: | |
3805 | case finlist: | |
3806 | case fignore: | |
3807 | case vignore: | |
3808 | break; | |
3809 | default: | |
3810 | fvdef = fvnone; | |
3811 | } | |
b12756c8 | 3812 | break; |
13fde0cd | 3813 | case '\0': |
d8913c1c FP |
3814 | if (objdef == otagseen) |
3815 | { | |
96c06863 | 3816 | make_C_tag (true); /* an Objective C class */ |
d8913c1c FP |
3817 | objdef = oignore; |
3818 | } | |
13fde0cd RS |
3819 | /* If a macro spans multiple lines don't reset its state. */ |
3820 | if (quotednl) | |
93b7ac65 | 3821 | CNL_SAVE_DEFINEDEF (); |
13fde0cd | 3822 | else |
93b7ac65 | 3823 | CNL (); |
13fde0cd RS |
3824 | break; |
3825 | } /* switch (c) */ | |
3826 | ||
3827 | } /* while not eof */ | |
b2521b0a | 3828 | |
b2521b0a FP |
3829 | free (lbs[0].lb.buffer); |
3830 | free (lbs[1].lb.buffer); | |
c6d46f5f | 3831 | } |
b9755a12 FP |
3832 | |
3833 | /* | |
3834 | * Process either a C++ file or a C file depending on the setting | |
3835 | * of a global flag. | |
3836 | */ | |
944bdd72 | 3837 | static void |
873fbd0b | 3838 | default_C_entries (FILE *inf) |
b9755a12 | 3839 | { |
8c463abe | 3840 | C_entries (cplusplus ? C_PLPL : C_AUTO, inf); |
b9755a12 FP |
3841 | } |
3842 | ||
8c463abe | 3843 | /* Always do plain C. */ |
944bdd72 | 3844 | static void |
873fbd0b | 3845 | plain_C_entries (FILE *inf) |
79263656 FP |
3846 | { |
3847 | C_entries (0, inf); | |
3848 | } | |
3849 | ||
b9755a12 | 3850 | /* Always do C++. */ |
944bdd72 | 3851 | static void |
873fbd0b | 3852 | Cplusplus_entries (FILE *inf) |
b9755a12 FP |
3853 | { |
3854 | C_entries (C_PLPL, inf); | |
3855 | } | |
3856 | ||
cec68fcb | 3857 | /* Always do Java. */ |
944bdd72 | 3858 | static void |
873fbd0b | 3859 | Cjava_entries (FILE *inf) |
cec68fcb FP |
3860 | { |
3861 | C_entries (C_JAVA, inf); | |
3862 | } | |
3863 | ||
b9755a12 | 3864 | /* Always do C*. */ |
944bdd72 | 3865 | static void |
873fbd0b | 3866 | Cstar_entries (FILE *inf) |
b9755a12 FP |
3867 | { |
3868 | C_entries (C_STAR, inf); | |
3869 | } | |
3870 | ||
3871 | /* Always do Yacc. */ | |
944bdd72 | 3872 | static void |
873fbd0b | 3873 | Yacc_entries (FILE *inf) |
b9755a12 FP |
3874 | { |
3875 | C_entries (YACC, inf); | |
3876 | } | |
b2521b0a | 3877 | |
6dd5561c | 3878 | \f |
62aec606 | 3879 | /* Useful macros. */ |
93b7ac65 | 3880 | #define LOOP_ON_INPUT_LINES(file_pointer, line_buffer, char_pointer) \ |
e7d3b099 | 3881 | for (; /* loop initialization */ \ |
93b7ac65 | 3882 | !feof (file_pointer) /* loop test */ \ |
2b749964 FP |
3883 | && /* instructions at start of loop */ \ |
3884 | (readline (&line_buffer, file_pointer), \ | |
3885 | char_pointer = line_buffer.buffer, \ | |
96c06863 | 3886 | true); \ |
93b7ac65 | 3887 | ) |
55102b5d FP |
3888 | |
3889 | #define LOOKING_AT(cp, kw) /* kw is the keyword, a literal string */ \ | |
96c06863 | 3890 | ((assert ("" kw), true) /* syntax error if not a literal string */ \ |
5e617bc2 JB |
3891 | && strneq ((cp), kw, sizeof (kw)-1) /* cp points at kw */ \ |
3892 | && notinname ((cp)[sizeof (kw)-1]) /* end of kw */ \ | |
3893 | && ((cp) = skip_spaces ((cp)+sizeof (kw)-1))) /* skip spaces */ | |
55102b5d FP |
3894 | |
3895 | /* Similar to LOOKING_AT but does not use notinname, does not skip */ | |
3896 | #define LOOKING_AT_NOCASE(cp, kw) /* the keyword is a literal string */ \ | |
96c06863 | 3897 | ((assert ("" kw), true) /* syntax error if not a literal string */ \ |
5e617bc2 JB |
3898 | && strncaseeq ((cp), kw, sizeof (kw)-1) /* cp points at kw */ \ |
3899 | && ((cp) += sizeof (kw)-1)) /* skip spaces */ | |
93b7ac65 FP |
3900 | |
3901 | /* | |
3902 | * Read a file, but do no processing. This is used to do regexp | |
3903 | * matching on files that have no language defined. | |
3904 | */ | |
944bdd72 | 3905 | static void |
873fbd0b | 3906 | just_read_file (FILE *inf) |
93b7ac65 | 3907 | { |
8c422c30 PE |
3908 | while (!feof (inf)) |
3909 | readline (&lb, inf); | |
93b7ac65 | 3910 | } |
b2521b0a | 3911 | |
93b7ac65 FP |
3912 | \f |
3913 | /* Fortran parsing */ | |
c6d46f5f | 3914 | |
f57e2426 J |
3915 | static void F_takeprec (void); |
3916 | static void F_getit (FILE *); | |
13fde0cd | 3917 | |
944bdd72 | 3918 | static void |
873fbd0b | 3919 | F_takeprec (void) |
6dd5561c | 3920 | { |
93b7ac65 | 3921 | dbp = skip_spaces (dbp); |
6dd5561c FP |
3922 | if (*dbp != '*') |
3923 | return; | |
3924 | dbp++; | |
93b7ac65 | 3925 | dbp = skip_spaces (dbp); |
79263656 FP |
3926 | if (strneq (dbp, "(*)", 3)) |
3927 | { | |
3928 | dbp += 3; | |
3929 | return; | |
3930 | } | |
af03e6ab | 3931 | if (!ISDIGIT (*dbp)) |
c6d46f5f | 3932 | { |
6dd5561c FP |
3933 | --dbp; /* force failure */ |
3934 | return; | |
c6d46f5f | 3935 | } |
6dd5561c FP |
3936 | do |
3937 | dbp++; | |
af03e6ab | 3938 | while (ISDIGIT (*dbp)); |
6dd5561c | 3939 | } |
13fde0cd | 3940 | |
944bdd72 | 3941 | static void |
873fbd0b | 3942 | F_getit (FILE *inf) |
6dd5561c FP |
3943 | { |
3944 | register char *cp; | |
13fde0cd | 3945 | |
93b7ac65 | 3946 | dbp = skip_spaces (dbp); |
6dd5561c | 3947 | if (*dbp == '\0') |
c6d46f5f | 3948 | { |
e7d3b099 | 3949 | readline (&lb, inf); |
6dd5561c FP |
3950 | dbp = lb.buffer; |
3951 | if (dbp[5] != '&') | |
3952 | return; | |
3953 | dbp += 6; | |
93b7ac65 | 3954 | dbp = skip_spaces (dbp); |
c6d46f5f | 3955 | } |
af03e6ab | 3956 | if (!ISALPHA (*dbp) && *dbp != '_' && *dbp != '$') |
6dd5561c | 3957 | return; |
93b7ac65 | 3958 | for (cp = dbp + 1; *cp != '\0' && intoken (*cp); cp++) |
6dd5561c | 3959 | continue; |
96c06863 | 3960 | make_tag (dbp, cp-dbp, true, |
081b7c70 | 3961 | lb.buffer, cp - lb.buffer + 1, lineno, linecharno); |
c6d46f5f | 3962 | } |
c6d46f5f | 3963 | |
93b7ac65 | 3964 | |
944bdd72 | 3965 | static void |
873fbd0b | 3966 | Fortran_functions (FILE *inf) |
c6d46f5f | 3967 | { |
93b7ac65 | 3968 | LOOP_ON_INPUT_LINES (inf, lb, dbp) |
c6d46f5f | 3969 | { |
c6d46f5f JB |
3970 | if (*dbp == '%') |
3971 | dbp++; /* Ratfor escape to fortran */ | |
93b7ac65 | 3972 | dbp = skip_spaces (dbp); |
108c932a | 3973 | if (*dbp == '\0') |
c6d46f5f | 3974 | continue; |
9d77a91f CY |
3975 | |
3976 | if (LOOKING_AT_NOCASE (dbp, "recursive")) | |
3977 | dbp = skip_spaces (dbp); | |
3978 | ||
d0417b4c DS |
3979 | if (LOOKING_AT_NOCASE (dbp, "pure")) |
3980 | dbp = skip_spaces (dbp); | |
3981 | ||
0f054abc GM |
3982 | if (LOOKING_AT_NOCASE (dbp, "elemental")) |
3983 | dbp = skip_spaces (dbp); | |
3984 | ||
79263656 | 3985 | switch (lowcase (*dbp)) |
c6d46f5f JB |
3986 | { |
3987 | case 'i': | |
eaaacacd FP |
3988 | if (nocase_tail ("integer")) |
3989 | F_takeprec (); | |
c6d46f5f JB |
3990 | break; |
3991 | case 'r': | |
eaaacacd FP |
3992 | if (nocase_tail ("real")) |
3993 | F_takeprec (); | |
c6d46f5f JB |
3994 | break; |
3995 | case 'l': | |
eaaacacd FP |
3996 | if (nocase_tail ("logical")) |
3997 | F_takeprec (); | |
c6d46f5f JB |
3998 | break; |
3999 | case 'c': | |
eaaacacd FP |
4000 | if (nocase_tail ("complex") || nocase_tail ("character")) |
4001 | F_takeprec (); | |
c6d46f5f JB |
4002 | break; |
4003 | case 'd': | |
eaaacacd | 4004 | if (nocase_tail ("double")) |
c6d46f5f | 4005 | { |
93b7ac65 | 4006 | dbp = skip_spaces (dbp); |
108c932a | 4007 | if (*dbp == '\0') |
c6d46f5f | 4008 | continue; |
eaaacacd | 4009 | if (nocase_tail ("precision")) |
c6d46f5f JB |
4010 | break; |
4011 | continue; | |
4012 | } | |
4013 | break; | |
4014 | } | |
93b7ac65 | 4015 | dbp = skip_spaces (dbp); |
108c932a | 4016 | if (*dbp == '\0') |
c6d46f5f | 4017 | continue; |
79263656 | 4018 | switch (lowcase (*dbp)) |
c6d46f5f JB |
4019 | { |
4020 | case 'f': | |
eaaacacd FP |
4021 | if (nocase_tail ("function")) |
4022 | F_getit (inf); | |
c6d46f5f JB |
4023 | continue; |
4024 | case 's': | |
eaaacacd FP |
4025 | if (nocase_tail ("subroutine")) |
4026 | F_getit (inf); | |
c6d46f5f | 4027 | continue; |
8a6c8bcf | 4028 | case 'e': |
eaaacacd FP |
4029 | if (nocase_tail ("entry")) |
4030 | F_getit (inf); | |
8a6c8bcf | 4031 | continue; |
15d5b889 | 4032 | case 'b': |
eaaacacd | 4033 | if (nocase_tail ("blockdata") || nocase_tail ("block data")) |
15d5b889 | 4034 | { |
93b7ac65 | 4035 | dbp = skip_spaces (dbp); |
15d5b889 | 4036 | if (*dbp == '\0') /* assume un-named */ |
96c06863 | 4037 | make_tag ("blockdata", 9, true, |
081b7c70 | 4038 | lb.buffer, dbp - lb.buffer, lineno, linecharno); |
15d5b889 | 4039 | else |
eaaacacd | 4040 | F_getit (inf); /* look for name */ |
15d5b889 RS |
4041 | } |
4042 | continue; | |
c6d46f5f JB |
4043 | } |
4044 | } | |
c6d46f5f | 4045 | } |
b2521b0a | 4046 | |
6dd5561c | 4047 | \f |
93b7ac65 | 4048 | /* |
93b7ac65 | 4049 | * Ada parsing |
eaaacacd | 4050 | * Original code by |
73d3586d | 4051 | * Philippe Waroquiers (1998) |
93b7ac65 | 4052 | */ |
71cbb895 | 4053 | |
93b7ac65 FP |
4054 | /* Once we are positioned after an "interesting" keyword, let's get |
4055 | the real tag value necessary. */ | |
944bdd72 | 4056 | static void |
988e88ab | 4057 | Ada_getit (FILE *inf, const char *name_qualifier) |
93b7ac65 FP |
4058 | { |
4059 | register char *cp; | |
4060 | char *name; | |
4061 | char c; | |
4062 | ||
4063 | while (!feof (inf)) | |
4064 | { | |
4065 | dbp = skip_spaces (dbp); | |
4066 | if (*dbp == '\0' | |
4067 | || (dbp[0] == '-' && dbp[1] == '-')) | |
4068 | { | |
e7d3b099 | 4069 | readline (&lb, inf); |
93b7ac65 FP |
4070 | dbp = lb.buffer; |
4071 | } | |
5e617bc2 | 4072 | switch (lowcase (*dbp)) |
93b7ac65 FP |
4073 | { |
4074 | case 'b': | |
eaaacacd | 4075 | if (nocase_tail ("body")) |
93b7ac65 FP |
4076 | { |
4077 | /* Skipping body of procedure body or package body or .... | |
4078 | resetting qualifier to body instead of spec. */ | |
4079 | name_qualifier = "/b"; | |
4080 | continue; | |
4081 | } | |
4082 | break; | |
4083 | case 't': | |
93b7ac65 | 4084 | /* Skipping type of task type or protected type ... */ |
eaaacacd | 4085 | if (nocase_tail ("type")) |
93b7ac65 FP |
4086 | continue; |
4087 | break; | |
4088 | } | |
4089 | if (*dbp == '"') | |
4090 | { | |
4091 | dbp += 1; | |
4092 | for (cp = dbp; *cp != '\0' && *cp != '"'; cp++) | |
4093 | continue; | |
4094 | } | |
4095 | else | |
4096 | { | |
4097 | dbp = skip_spaces (dbp); | |
4098 | for (cp = dbp; | |
4099 | (*cp != '\0' | |
af03e6ab | 4100 | && (ISALPHA (*cp) || ISDIGIT (*cp) || *cp == '_' || *cp == '.')); |
93b7ac65 FP |
4101 | cp++) |
4102 | continue; | |
4103 | if (cp == dbp) | |
4104 | return; | |
4105 | } | |
4106 | c = *cp; | |
4107 | *cp = '\0'; | |
4108 | name = concat (dbp, name_qualifier, ""); | |
4109 | *cp = c; | |
96c06863 | 4110 | make_tag (name, strlen (name), true, |
081b7c70 FP |
4111 | lb.buffer, cp - lb.buffer + 1, lineno, linecharno); |
4112 | free (name); | |
93b7ac65 FP |
4113 | if (c == '"') |
4114 | dbp = cp + 1; | |
4115 | return; | |
4116 | } | |
4117 | } | |
4118 | ||
944bdd72 | 4119 | static void |
873fbd0b | 4120 | Ada_funcs (FILE *inf) |
93b7ac65 | 4121 | { |
96c06863 PE |
4122 | bool inquote = false; |
4123 | bool skip_till_semicolumn = false; | |
93b7ac65 FP |
4124 | |
4125 | LOOP_ON_INPUT_LINES (inf, lb, dbp) | |
4126 | { | |
4127 | while (*dbp != '\0') | |
4128 | { | |
4129 | /* Skip a string i.e. "abcd". */ | |
4130 | if (inquote || (*dbp == '"')) | |
4131 | { | |
58cb46fb | 4132 | dbp = etags_strchr (dbp + !inquote, '"'); |
93b7ac65 FP |
4133 | if (dbp != NULL) |
4134 | { | |
96c06863 | 4135 | inquote = false; |
93b7ac65 FP |
4136 | dbp += 1; |
4137 | continue; /* advance char */ | |
4138 | } | |
4139 | else | |
4140 | { | |
96c06863 | 4141 | inquote = true; |
93b7ac65 FP |
4142 | break; /* advance line */ |
4143 | } | |
4144 | } | |
4145 | ||
4146 | /* Skip comments. */ | |
4147 | if (dbp[0] == '-' && dbp[1] == '-') | |
4148 | break; /* advance line */ | |
4149 | ||
4150 | /* Skip character enclosed in single quote i.e. 'a' | |
4151 | and skip single quote starting an attribute i.e. 'Image. */ | |
4152 | if (*dbp == '\'') | |
4153 | { | |
4154 | dbp++ ; | |
4155 | if (*dbp != '\0') | |
4156 | dbp++; | |
4157 | continue; | |
4158 | } | |
4159 | ||
61a1f6fa FP |
4160 | if (skip_till_semicolumn) |
4161 | { | |
4162 | if (*dbp == ';') | |
96c06863 | 4163 | skip_till_semicolumn = false; |
61a1f6fa FP |
4164 | dbp++; |
4165 | continue; /* advance char */ | |
4166 | } | |
4167 | ||
93b7ac65 FP |
4168 | /* Search for beginning of a token. */ |
4169 | if (!begtoken (*dbp)) | |
4170 | { | |
4171 | dbp++; | |
4172 | continue; /* advance char */ | |
4173 | } | |
4174 | ||
4175 | /* We are at the beginning of a token. */ | |
5e617bc2 | 4176 | switch (lowcase (*dbp)) |
93b7ac65 FP |
4177 | { |
4178 | case 'f': | |
eaaacacd FP |
4179 | if (!packages_only && nocase_tail ("function")) |
4180 | Ada_getit (inf, "/f"); | |
93b7ac65 FP |
4181 | else |
4182 | break; /* from switch */ | |
4183 | continue; /* advance char */ | |
4184 | case 'p': | |
eaaacacd FP |
4185 | if (!packages_only && nocase_tail ("procedure")) |
4186 | Ada_getit (inf, "/p"); | |
4187 | else if (nocase_tail ("package")) | |
4188 | Ada_getit (inf, "/s"); | |
4189 | else if (nocase_tail ("protected")) /* protected type */ | |
4190 | Ada_getit (inf, "/t"); | |
93b7ac65 FP |
4191 | else |
4192 | break; /* from switch */ | |
4193 | continue; /* advance char */ | |
61a1f6fa FP |
4194 | |
4195 | case 'u': | |
4196 | if (typedefs && !packages_only && nocase_tail ("use")) | |
4197 | { | |
4198 | /* when tagging types, avoid tagging use type Pack.Typename; | |
4199 | for this, we will skip everything till a ; */ | |
96c06863 | 4200 | skip_till_semicolumn = true; |
61a1f6fa FP |
4201 | continue; /* advance char */ |
4202 | } | |
4203 | ||
93b7ac65 | 4204 | case 't': |
eaaacacd FP |
4205 | if (!packages_only && nocase_tail ("task")) |
4206 | Ada_getit (inf, "/k"); | |
4207 | else if (typedefs && !packages_only && nocase_tail ("type")) | |
93b7ac65 | 4208 | { |
eaaacacd | 4209 | Ada_getit (inf, "/t"); |
93b7ac65 FP |
4210 | while (*dbp != '\0') |
4211 | dbp += 1; | |
4212 | } | |
4213 | else | |
4214 | break; /* from switch */ | |
4215 | continue; /* advance char */ | |
4216 | } | |
4217 | ||
4218 | /* Look for the end of the token. */ | |
4219 | while (!endtoken (*dbp)) | |
4220 | dbp++; | |
4221 | ||
4222 | } /* advance char */ | |
4223 | } /* advance line */ | |
4224 | } | |
b2521b0a | 4225 | |
93b7ac65 | 4226 | \f |
6dd5561c | 4227 | /* |
6dd5561c | 4228 | * Unix and microcontroller assembly tag handling |
eaaacacd FP |
4229 | * Labels: /^[a-zA-Z_.$][a-zA_Z0-9_.$]*[: ^I^J]/ |
4230 | * Idea by Bob Weiner, Motorola Inc. (1994) | |
6dd5561c | 4231 | */ |
944bdd72 | 4232 | static void |
873fbd0b | 4233 | Asm_labels (FILE *inf) |
c6d46f5f JB |
4234 | { |
4235 | register char *cp; | |
c6d46f5f | 4236 | |
93b7ac65 | 4237 | LOOP_ON_INPUT_LINES (inf, lb, cp) |
c6d46f5f | 4238 | { |
6dd5561c FP |
4239 | /* If first char is alphabetic or one of [_.$], test for colon |
4240 | following identifier. */ | |
af03e6ab | 4241 | if (ISALPHA (*cp) || *cp == '_' || *cp == '.' || *cp == '$') |
6dd5561c FP |
4242 | { |
4243 | /* Read past label. */ | |
4244 | cp++; | |
af03e6ab | 4245 | while (ISALNUM (*cp) || *cp == '_' || *cp == '.' || *cp == '$') |
6dd5561c | 4246 | cp++; |
71cbb895 | 4247 | if (*cp == ':' || iswhite (*cp)) |
081b7c70 | 4248 | /* Found end of label, so copy it and add it to the table. */ |
96c06863 | 4249 | make_tag (lb.buffer, cp - lb.buffer, true, |
55597f90 | 4250 | lb.buffer, cp - lb.buffer + 1, lineno, linecharno); |
6dd5561c | 4251 | } |
c6d46f5f JB |
4252 | } |
4253 | } | |
b2521b0a | 4254 | |
c6d46f5f | 4255 | \f |
1f638249 | 4256 | /* |
b2521b0a | 4257 | * Perl support |
eaaacacd | 4258 | * Perl sub names: /^sub[ \t\n]+[^ \t\n{]+/ |
00054d21 | 4259 | * /^use constant[ \t\n]+[^ \t\n{=,;]+/ |
93b7ac65 | 4260 | * Perl variable names: /^(my|local).../ |
eaaacacd FP |
4261 | * Original code by Bart Robinson <lomew@cs.utah.edu> (1995) |
4262 | * Additions by Michael Ernst <mernst@alum.mit.edu> (1997) | |
c38e0c97 | 4263 | * Ideas by Kai Großjohann <Kai.Grossjohann@CS.Uni-Dortmund.DE> (2001) |
1f638249 | 4264 | */ |
944bdd72 | 4265 | static void |
873fbd0b | 4266 | Perl_functions (FILE *inf) |
1f638249 | 4267 | { |
eaaacacd | 4268 | char *package = savestr ("main"); /* current package name */ |
1f638249 FP |
4269 | register char *cp; |
4270 | ||
93b7ac65 | 4271 | LOOP_ON_INPUT_LINES (inf, lb, cp) |
1f638249 | 4272 | { |
1797886f | 4273 | cp = skip_spaces (cp); |
eaaacacd FP |
4274 | |
4275 | if (LOOKING_AT (cp, "package")) | |
93b7ac65 | 4276 | { |
eaaacacd | 4277 | free (package); |
081b7c70 | 4278 | get_tag (cp, &package); |
eaaacacd FP |
4279 | } |
4280 | else if (LOOKING_AT (cp, "sub")) | |
4281 | { | |
00054d21 | 4282 | char *pos, *sp; |
eaaacacd | 4283 | |
00054d21 KR |
4284 | subr: |
4285 | sp = cp; | |
eaaacacd FP |
4286 | while (!notinname (*cp)) |
4287 | cp++; | |
4288 | if (cp == sp) | |
081b7c70 FP |
4289 | continue; /* nothing found */ |
4290 | if ((pos = etags_strchr (sp, ':')) != NULL | |
4291 | && pos < cp && pos[1] == ':') | |
4292 | /* The name is already qualified. */ | |
96c06863 | 4293 | make_tag (sp, cp - sp, true, |
081b7c70 | 4294 | lb.buffer, cp - lb.buffer + 1, lineno, linecharno); |
eaaacacd | 4295 | else |
081b7c70 FP |
4296 | /* Qualify it. */ |
4297 | { | |
4298 | char savechar, *name; | |
4299 | ||
4300 | savechar = *cp; | |
4301 | *cp = '\0'; | |
4302 | name = concat (package, "::", sp); | |
4303 | *cp = savechar; | |
96c06863 | 4304 | make_tag (name, strlen (name), true, |
081b7c70 FP |
4305 | lb.buffer, cp - lb.buffer + 1, lineno, linecharno); |
4306 | free (name); | |
4307 | } | |
00054d21 KR |
4308 | } |
4309 | else if (LOOKING_AT (cp, "use constant") | |
4310 | || LOOKING_AT (cp, "use constant::defer")) | |
4311 | { | |
4312 | /* For hash style multi-constant like | |
4313 | use constant { FOO => 123, | |
4314 | BAR => 456 }; | |
4315 | only the first FOO is picked up. Parsing across the value | |
4316 | expressions would be difficult in general, due to possible nested | |
4317 | hashes, here-documents, etc. */ | |
4318 | if (*cp == '{') | |
4319 | cp = skip_spaces (cp+1); | |
4320 | goto subr; | |
93b7ac65 | 4321 | } |
00054d21 | 4322 | else if (globals) /* only if we are tagging global vars */ |
93b7ac65 | 4323 | { |
081b7c70 FP |
4324 | /* Skip a qualifier, if any. */ |
4325 | bool qual = LOOKING_AT (cp, "my") || LOOKING_AT (cp, "local"); | |
93b7ac65 | 4326 | /* After "my" or "local", but before any following paren or space. */ |
081b7c70 | 4327 | char *varstart = cp; |
1f638249 | 4328 | |
081b7c70 FP |
4329 | if (qual /* should this be removed? If yes, how? */ |
4330 | && (*cp == '$' || *cp == '@' || *cp == '%')) | |
93b7ac65 | 4331 | { |
081b7c70 FP |
4332 | varstart += 1; |
4333 | do | |
93b7ac65 | 4334 | cp++; |
081b7c70 | 4335 | while (ISALNUM (*cp) || *cp == '_'); |
93b7ac65 | 4336 | } |
081b7c70 | 4337 | else if (qual) |
93b7ac65 FP |
4338 | { |
4339 | /* Should be examining a variable list at this point; | |
4340 | could insist on seeing an open parenthesis. */ | |
4341 | while (*cp != '\0' && *cp != ';' && *cp != '=' && *cp != ')') | |
4342 | cp++; | |
4343 | } | |
081b7c70 FP |
4344 | else |
4345 | continue; | |
93b7ac65 | 4346 | |
96c06863 | 4347 | make_tag (varstart, cp - varstart, false, |
081b7c70 | 4348 | lb.buffer, cp - lb.buffer + 1, lineno, linecharno); |
93b7ac65 FP |
4349 | } |
4350 | } | |
7e237d24 | 4351 | free (package); |
93b7ac65 | 4352 | } |
b2521b0a | 4353 | |
62aec606 | 4354 | |
93b7ac65 | 4355 | /* |
b2521b0a | 4356 | * Python support |
49adb67a | 4357 | * Look for /^[\t]*def[ \t\n]+[^ \t\n(:]+/ or /^class[ \t\n]+[^ \t\n(:]+/ |
eaaacacd | 4358 | * Idea by Eric S. Raymond <esr@thyrsus.com> (1997) |
49adb67a | 4359 | * More ideas by seb bacon <seb@jamkit.com> (2002) |
93b7ac65 | 4360 | */ |
944bdd72 | 4361 | static void |
873fbd0b | 4362 | Python_functions (FILE *inf) |
93b7ac65 FP |
4363 | { |
4364 | register char *cp; | |
4365 | ||
62aec606 | 4366 | LOOP_ON_INPUT_LINES (inf, lb, cp) |
49adb67a FP |
4367 | { |
4368 | cp = skip_spaces (cp); | |
4369 | if (LOOKING_AT (cp, "def") || LOOKING_AT (cp, "class")) | |
4370 | { | |
4371 | char *name = cp; | |
4372 | while (!notinname (*cp) && *cp != ':') | |
4373 | cp++; | |
96c06863 | 4374 | make_tag (name, cp - name, true, |
081b7c70 | 4375 | lb.buffer, cp - lb.buffer + 1, lineno, linecharno); |
49adb67a FP |
4376 | } |
4377 | } | |
62aec606 FP |
4378 | } |
4379 | ||
4380 | \f | |
4381 | /* | |
4382 | * PHP support | |
4383 | * Look for: | |
4384 | * - /^[ \t]*function[ \t\n]+[^ \t\n(]+/ | |
4385 | * - /^[ \t]*class[ \t\n]+[^ \t\n]+/ | |
4386 | * - /^[ \t]*define\(\"[^\"]+/ | |
4387 | * Only with --members: | |
4388 | * - /^[ \t]*var[ \t\n]+\$[^ \t\n=;]/ | |
eaaacacd | 4389 | * Idea by Diez B. Roggisch (2001) |
62aec606 FP |
4390 | */ |
4391 | static void | |
873fbd0b | 4392 | PHP_functions (FILE *inf) |
62aec606 | 4393 | { |
96c06863 PE |
4394 | char *cp, *name; |
4395 | bool search_identifier = false; | |
62aec606 | 4396 | |
93b7ac65 FP |
4397 | LOOP_ON_INPUT_LINES (inf, lb, cp) |
4398 | { | |
62aec606 | 4399 | cp = skip_spaces (cp); |
49adb67a | 4400 | name = cp; |
62aec606 FP |
4401 | if (search_identifier |
4402 | && *cp != '\0') | |
1f638249 | 4403 | { |
eaaacacd | 4404 | while (!notinname (*cp)) |
1f638249 | 4405 | cp++; |
96c06863 | 4406 | make_tag (name, cp - name, true, |
081b7c70 | 4407 | lb.buffer, cp - lb.buffer + 1, lineno, linecharno); |
96c06863 | 4408 | search_identifier = false; |
93b7ac65 | 4409 | } |
62aec606 FP |
4410 | else if (LOOKING_AT (cp, "function")) |
4411 | { | |
5e617bc2 | 4412 | if (*cp == '&') |
62aec606 | 4413 | cp = skip_spaces (cp+1); |
5e617bc2 | 4414 | if (*cp != '\0') |
62aec606 | 4415 | { |
49adb67a | 4416 | name = cp; |
eaaacacd | 4417 | while (!notinname (*cp)) |
62aec606 | 4418 | cp++; |
96c06863 | 4419 | make_tag (name, cp - name, true, |
081b7c70 | 4420 | lb.buffer, cp - lb.buffer + 1, lineno, linecharno); |
62aec606 FP |
4421 | } |
4422 | else | |
96c06863 | 4423 | search_identifier = true; |
62aec606 FP |
4424 | } |
4425 | else if (LOOKING_AT (cp, "class")) | |
4426 | { | |
4427 | if (*cp != '\0') | |
4428 | { | |
49adb67a | 4429 | name = cp; |
62aec606 FP |
4430 | while (*cp != '\0' && !iswhite (*cp)) |
4431 | cp++; | |
96c06863 | 4432 | make_tag (name, cp - name, false, |
081b7c70 | 4433 | lb.buffer, cp - lb.buffer + 1, lineno, linecharno); |
62aec606 FP |
4434 | } |
4435 | else | |
96c06863 | 4436 | search_identifier = true; |
62aec606 FP |
4437 | } |
4438 | else if (strneq (cp, "define", 6) | |
4439 | && (cp = skip_spaces (cp+6)) | |
4440 | && *cp++ == '(' | |
4441 | && (*cp == '"' || *cp == '\'')) | |
93b7ac65 | 4442 | { |
62aec606 | 4443 | char quote = *cp++; |
49adb67a | 4444 | name = cp; |
62aec606 | 4445 | while (*cp != quote && *cp != '\0') |
1f638249 | 4446 | cp++; |
96c06863 | 4447 | make_tag (name, cp - name, false, |
081b7c70 | 4448 | lb.buffer, cp - lb.buffer + 1, lineno, linecharno); |
62aec606 FP |
4449 | } |
4450 | else if (members | |
4451 | && LOOKING_AT (cp, "var") | |
4452 | && *cp == '$') | |
4453 | { | |
49adb67a | 4454 | name = cp; |
5e617bc2 | 4455 | while (!notinname (*cp)) |
62aec606 | 4456 | cp++; |
96c06863 | 4457 | make_tag (name, cp - name, false, |
081b7c70 | 4458 | lb.buffer, cp - lb.buffer + 1, lineno, linecharno); |
1f638249 FP |
4459 | } |
4460 | } | |
4461 | } | |
b2521b0a | 4462 | |
1f638249 | 4463 | \f |
eaaacacd | 4464 | /* |
30903246 FP |
4465 | * Cobol tag functions |
4466 | * We could look for anything that could be a paragraph name. | |
4467 | * i.e. anything that starts in column 8 is one word and ends in a full stop. | |
eaaacacd | 4468 | * Idea by Corny de Souza (1993) |
30903246 | 4469 | */ |
944bdd72 | 4470 | static void |
873fbd0b | 4471 | Cobol_paragraphs (FILE *inf) |
30903246 | 4472 | { |
93b7ac65 | 4473 | register char *bp, *ep; |
30903246 | 4474 | |
93b7ac65 | 4475 | LOOP_ON_INPUT_LINES (inf, lb, bp) |
30903246 | 4476 | { |
30903246 FP |
4477 | if (lb.len < 9) |
4478 | continue; | |
93b7ac65 | 4479 | bp += 8; |
30903246 FP |
4480 | |
4481 | /* If eoln, compiler option or comment ignore whole line. */ | |
af03e6ab | 4482 | if (bp[-1] != ' ' || !ISALNUM (bp[0])) |
30903246 FP |
4483 | continue; |
4484 | ||
af03e6ab | 4485 | for (ep = bp; ISALNUM (*ep) || *ep == '-'; ep++) |
30903246 | 4486 | continue; |
93b7ac65 | 4487 | if (*ep++ == '.') |
96c06863 | 4488 | make_tag (bp, ep - bp, true, |
081b7c70 | 4489 | lb.buffer, ep - lb.buffer + 1, lineno, linecharno); |
30903246 FP |
4490 | } |
4491 | } | |
b2521b0a | 4492 | |
30903246 | 4493 | \f |
89d57763 FP |
4494 | /* |
4495 | * Makefile support | |
2b749964 | 4496 | * Ideas by Assar Westerlund <assar@sics.se> (2001) |
89d57763 FP |
4497 | */ |
4498 | static void | |
873fbd0b | 4499 | Makefile_targets (FILE *inf) |
89d57763 FP |
4500 | { |
4501 | register char *bp; | |
4502 | ||
4503 | LOOP_ON_INPUT_LINES (inf, lb, bp) | |
4504 | { | |
4505 | if (*bp == '\t' || *bp == '#') | |
4506 | continue; | |
4507 | while (*bp != '\0' && *bp != '=' && *bp != ':') | |
4508 | bp++; | |
2b749964 | 4509 | if (*bp == ':' || (globals && *bp == '=')) |
6598a3d4 FP |
4510 | { |
4511 | /* We should detect if there is more than one tag, but we do not. | |
4512 | We just skip initial and final spaces. */ | |
4513 | char * namestart = skip_spaces (lb.buffer); | |
4514 | while (--bp > namestart) | |
4515 | if (!notinname (*bp)) | |
4516 | break; | |
96c06863 | 4517 | make_tag (namestart, bp - namestart + 1, true, |
6598a3d4 FP |
4518 | lb.buffer, bp - lb.buffer + 2, lineno, linecharno); |
4519 | } | |
89d57763 FP |
4520 | } |
4521 | } | |
b2521b0a | 4522 | |
89d57763 | 4523 | \f |
aab1fdae | 4524 | /* |
eaaacacd FP |
4525 | * Pascal parsing |
4526 | * Original code by Mosur K. Mohan (1989) | |
4527 | * | |
aab1fdae FP |
4528 | * Locates tags for procedures & functions. Doesn't do any type- or |
4529 | * var-definitions. It does look for the keyword "extern" or | |
4530 | * "forward" immediately following the procedure statement; if found, | |
4531 | * the tag is skipped. | |
c6d46f5f | 4532 | */ |
944bdd72 | 4533 | static void |
873fbd0b | 4534 | Pascal_functions (FILE *inf) |
c6d46f5f | 4535 | { |
93b7ac65 | 4536 | linebuffer tline; /* mostly copied from C_entries */ |
c6d46f5f | 4537 | long save_lcno; |
081b7c70 FP |
4538 | int save_lineno, namelen, taglen; |
4539 | char c, *name; | |
c6d46f5f | 4540 | |
96c06863 | 4541 | bool /* each of these flags is true if: */ |
b9755a12 | 4542 | incomment, /* point is inside a comment */ |
c6d46f5f | 4543 | inquote, /* point is inside '..' string */ |
108c932a FP |
4544 | get_tagname, /* point is after PROCEDURE/FUNCTION |
4545 | keyword, so next item = potential tag */ | |
c6d46f5f JB |
4546 | found_tag, /* point is after a potential tag */ |
4547 | inparms, /* point is within parameter-list */ | |
108c932a FP |
4548 | verify_tag; /* point has passed the parm-list, so the |
4549 | next token will determine whether this | |
4550 | is a FORWARD/EXTERN to be ignored, or | |
4551 | whether it is a real tag */ | |
c6d46f5f | 4552 | |
081b7c70 FP |
4553 | save_lcno = save_lineno = namelen = taglen = 0; /* keep compiler quiet */ |
4554 | name = NULL; /* keep compiler quiet */ | |
c6d46f5f | 4555 | dbp = lb.buffer; |
108c932a | 4556 | *dbp = '\0'; |
61a1f6fa | 4557 | linebuffer_init (&tline); |
c6d46f5f | 4558 | |
96c06863 PE |
4559 | incomment = inquote = false; |
4560 | found_tag = false; /* have a proc name; check if extern */ | |
4561 | get_tagname = false; /* found "procedure" keyword */ | |
4562 | inparms = false; /* found '(' after "proc" */ | |
4563 | verify_tag = false; /* check if "extern" is ahead */ | |
c6d46f5f | 4564 | |
93b7ac65 FP |
4565 | |
4566 | while (!feof (inf)) /* long main loop to get next char */ | |
c6d46f5f JB |
4567 | { |
4568 | c = *dbp++; | |
55597f90 | 4569 | if (c == '\0') /* if end of line */ |
c6d46f5f | 4570 | { |
e7d3b099 | 4571 | readline (&lb, inf); |
cf347d3c | 4572 | dbp = lb.buffer; |
55597f90 | 4573 | if (*dbp == '\0') |
c6d46f5f | 4574 | continue; |
93b7ac65 FP |
4575 | if (!((found_tag && verify_tag) |
4576 | || get_tagname)) | |
108c932a FP |
4577 | c = *dbp++; /* only if don't need *dbp pointing |
4578 | to the beginning of the name of | |
4579 | the procedure or function */ | |
c6d46f5f | 4580 | } |
b9755a12 | 4581 | if (incomment) |
c6d46f5f | 4582 | { |
108c932a | 4583 | if (c == '}') /* within { } comments */ |
96c06863 | 4584 | incomment = false; |
108c932a | 4585 | else if (c == '*' && *dbp == ')') /* within (* *) comments */ |
c6d46f5f | 4586 | { |
b9755a12 | 4587 | dbp++; |
96c06863 | 4588 | incomment = false; |
c6d46f5f JB |
4589 | } |
4590 | continue; | |
4591 | } | |
4592 | else if (inquote) | |
4593 | { | |
4594 | if (c == '\'') | |
96c06863 | 4595 | inquote = false; |
c6d46f5f JB |
4596 | continue; |
4597 | } | |
55597f90 | 4598 | else |
c6d46f5f JB |
4599 | switch (c) |
4600 | { | |
4601 | case '\'': | |
96c06863 | 4602 | inquote = true; /* found first quote */ |
c6d46f5f | 4603 | continue; |
108c932a | 4604 | case '{': /* found open { comment */ |
96c06863 | 4605 | incomment = true; |
c6d46f5f JB |
4606 | continue; |
4607 | case '(': | |
108c932a | 4608 | if (*dbp == '*') /* found open (* comment */ |
c6d46f5f | 4609 | { |
96c06863 | 4610 | incomment = true; |
c6d46f5f JB |
4611 | dbp++; |
4612 | } | |
4613 | else if (found_tag) /* found '(' after tag, i.e., parm-list */ | |
96c06863 | 4614 | inparms = true; |
c6d46f5f JB |
4615 | continue; |
4616 | case ')': /* end of parms list */ | |
4617 | if (inparms) | |
96c06863 | 4618 | inparms = false; |
c6d46f5f JB |
4619 | continue; |
4620 | case ';': | |
108c932a | 4621 | if (found_tag && !inparms) /* end of proc or fn stmt */ |
c6d46f5f | 4622 | { |
96c06863 | 4623 | verify_tag = true; |
c6d46f5f JB |
4624 | break; |
4625 | } | |
4626 | continue; | |
4627 | } | |
108c932a | 4628 | if (found_tag && verify_tag && (*dbp != ' ')) |
c6d46f5f | 4629 | { |
081b7c70 | 4630 | /* Check if this is an "extern" declaration. */ |
108c932a | 4631 | if (*dbp == '\0') |
c6d46f5f | 4632 | continue; |
8453bb55 | 4633 | if (lowcase (*dbp) == 'e') |
c6d46f5f | 4634 | { |
eaaacacd | 4635 | if (nocase_tail ("extern")) /* superfluous, really! */ |
c6d46f5f | 4636 | { |
96c06863 PE |
4637 | found_tag = false; |
4638 | verify_tag = false; | |
c6d46f5f JB |
4639 | } |
4640 | } | |
108c932a | 4641 | else if (lowcase (*dbp) == 'f') |
c6d46f5f | 4642 | { |
61a1f6fa | 4643 | if (nocase_tail ("forward")) /* check for forward reference */ |
c6d46f5f | 4644 | { |
96c06863 PE |
4645 | found_tag = false; |
4646 | verify_tag = false; | |
c6d46f5f JB |
4647 | } |
4648 | } | |
108c932a | 4649 | if (found_tag && verify_tag) /* not external proc, so make tag */ |
c6d46f5f | 4650 | { |
96c06863 PE |
4651 | found_tag = false; |
4652 | verify_tag = false; | |
4653 | make_tag (name, namelen, true, | |
081b7c70 | 4654 | tline.buffer, taglen, save_lineno, save_lcno); |
c6d46f5f JB |
4655 | continue; |
4656 | } | |
4657 | } | |
4658 | if (get_tagname) /* grab name of proc or fn */ | |
4659 | { | |
081b7c70 FP |
4660 | char *cp; |
4661 | ||
108c932a | 4662 | if (*dbp == '\0') |
c6d46f5f JB |
4663 | continue; |
4664 | ||
081b7c70 FP |
4665 | /* Find block name. */ |
4666 | for (cp = dbp + 1; *cp != '\0' && !endtoken (*cp); cp++) | |
4667 | continue; | |
4668 | ||
4669 | /* Save all values for later tagging. */ | |
b2521b0a | 4670 | linebuffer_setlen (&tline, lb.len); |
c6d46f5f JB |
4671 | strcpy (tline.buffer, lb.buffer); |
4672 | save_lineno = lineno; | |
4673 | save_lcno = linecharno; | |
081b7c70 FP |
4674 | name = tline.buffer + (dbp - lb.buffer); |
4675 | namelen = cp - dbp; | |
4676 | taglen = cp - lb.buffer + 1; | |
c6d46f5f | 4677 | |
d8913c1c | 4678 | dbp = cp; /* set dbp to e-o-token */ |
96c06863 PE |
4679 | get_tagname = false; |
4680 | found_tag = true; | |
c6d46f5f JB |
4681 | continue; |
4682 | ||
081b7c70 | 4683 | /* And proceed to check for "extern". */ |
c6d46f5f | 4684 | } |
55597f90 | 4685 | else if (!incomment && !inquote && !found_tag) |
c6d46f5f | 4686 | { |
081b7c70 | 4687 | /* Check for proc/fn keywords. */ |
79263656 | 4688 | switch (lowcase (c)) |
c6d46f5f JB |
4689 | { |
4690 | case 'p': | |
eaaacacd | 4691 | if (nocase_tail ("rocedure")) /* c = 'p', dbp has advanced */ |
96c06863 | 4692 | get_tagname = true; |
c6d46f5f JB |
4693 | continue; |
4694 | case 'f': | |
eaaacacd | 4695 | if (nocase_tail ("unction")) |
96c06863 | 4696 | get_tagname = true; |
c6d46f5f JB |
4697 | continue; |
4698 | } | |
4699 | } | |
081b7c70 | 4700 | } /* while not eof */ |
c5007f46 | 4701 | |
108c932a | 4702 | free (tline.buffer); |
c6d46f5f | 4703 | } |
b2521b0a | 4704 | |
c6d46f5f JB |
4705 | \f |
4706 | /* | |
b2521b0a | 4707 | * Lisp tag functions |
55597f90 | 4708 | * look for (def or (DEF, quote or QUOTE |
c6d46f5f | 4709 | */ |
71cbb895 | 4710 | |
f57e2426 | 4711 | static void L_getit (void); |
c6d46f5f | 4712 | |
944bdd72 | 4713 | static void |
873fbd0b | 4714 | L_getit (void) |
c6d46f5f | 4715 | { |
31d4b314 FP |
4716 | if (*dbp == '\'') /* Skip prefix quote */ |
4717 | dbp++; | |
93b7ac65 | 4718 | else if (*dbp == '(') |
31d4b314 | 4719 | { |
eaaacacd FP |
4720 | dbp++; |
4721 | /* Try to skip "(quote " */ | |
4722 | if (!LOOKING_AT (dbp, "quote") && !LOOKING_AT (dbp, "QUOTE")) | |
4723 | /* Ok, then skip "(" before name in (defstruct (foo)) */ | |
4724 | dbp = skip_spaces (dbp); | |
31d4b314 | 4725 | } |
081b7c70 | 4726 | get_tag (dbp, NULL); |
c6d46f5f | 4727 | } |
6dd5561c | 4728 | |
944bdd72 | 4729 | static void |
873fbd0b | 4730 | Lisp_functions (FILE *inf) |
6dd5561c | 4731 | { |
93b7ac65 | 4732 | LOOP_ON_INPUT_LINES (inf, lb, dbp) |
6dd5561c | 4733 | { |
eaaacacd FP |
4734 | if (dbp[0] != '(') |
4735 | continue; | |
4736 | ||
1cbaa705 KR |
4737 | /* "(defvar foo)" is a declaration rather than a definition. */ |
4738 | if (! declarations) | |
4739 | { | |
4740 | char *p = dbp + 1; | |
4741 | if (LOOKING_AT (p, "defvar")) | |
4742 | { | |
4743 | p = skip_name (p); /* past var name */ | |
4744 | p = skip_spaces (p); | |
4745 | if (*p == ')') | |
4746 | continue; | |
4747 | } | |
4748 | } | |
4749 | ||
eaaacacd | 4750 | if (strneq (dbp+1, "def", 3) || strneq (dbp+1, "DEF", 3)) |
6dd5561c | 4751 | { |
eaaacacd FP |
4752 | dbp = skip_non_spaces (dbp); |
4753 | dbp = skip_spaces (dbp); | |
4754 | L_getit (); | |
4755 | } | |
4756 | else | |
4757 | { | |
4758 | /* Check for (foo::defmumble name-defined ... */ | |
4759 | do | |
4760 | dbp++; | |
4761 | while (!notinname (*dbp) && *dbp != ':'); | |
4762 | if (*dbp == ':') | |
6dd5561c | 4763 | { |
6dd5561c FP |
4764 | do |
4765 | dbp++; | |
eaaacacd | 4766 | while (*dbp == ':'); |
6dd5561c | 4767 | |
eaaacacd FP |
4768 | if (strneq (dbp, "def", 3) || strneq (dbp, "DEF", 3)) |
4769 | { | |
4770 | dbp = skip_non_spaces (dbp); | |
4771 | dbp = skip_spaces (dbp); | |
4772 | L_getit (); | |
6dd5561c FP |
4773 | } |
4774 | } | |
4775 | } | |
4776 | } | |
4777 | } | |
b2521b0a | 4778 | |
c6d46f5f | 4779 | \f |
05d9a399 FP |
4780 | /* |
4781 | * Lua script language parsing | |
4782 | * Original code by David A. Capello <dacap@users.sourceforge.net> (2004) | |
4783 | * | |
4784 | * "function" and "local function" are tags if they start at column 1. | |
4785 | */ | |
4786 | static void | |
873fbd0b | 4787 | Lua_functions (FILE *inf) |
05d9a399 FP |
4788 | { |
4789 | register char *bp; | |
4790 | ||
4791 | LOOP_ON_INPUT_LINES (inf, lb, bp) | |
4792 | { | |
4793 | if (bp[0] != 'f' && bp[0] != 'l') | |
4794 | continue; | |
4795 | ||
21cb180b | 4796 | (void)LOOKING_AT (bp, "local"); /* skip possible "local" */ |
05d9a399 FP |
4797 | |
4798 | if (LOOKING_AT (bp, "function")) | |
4799 | get_tag (bp, NULL); | |
4800 | } | |
4801 | } | |
4802 | ||
4803 | \f | |
cec68fcb | 4804 | /* |
7877f373 | 4805 | * PostScript tags |
cec68fcb | 4806 | * Just look for lines where the first character is '/' |
93b7ac65 | 4807 | * Also look at "defineps" for PSWrap |
eaaacacd FP |
4808 | * Ideas by: |
4809 | * Richard Mlynarik <mly@adoc.xerox.com> (1997) | |
4810 | * Masatake Yamato <masata-y@is.aist-nara.ac.jp> (1999) | |
cec68fcb | 4811 | */ |
944bdd72 | 4812 | static void |
873fbd0b | 4813 | PS_functions (FILE *inf) |
cec68fcb | 4814 | { |
93b7ac65 | 4815 | register char *bp, *ep; |
30903246 | 4816 | |
93b7ac65 | 4817 | LOOP_ON_INPUT_LINES (inf, lb, bp) |
cec68fcb | 4818 | { |
93b7ac65 | 4819 | if (bp[0] == '/') |
cec68fcb | 4820 | { |
93b7ac65 FP |
4821 | for (ep = bp+1; |
4822 | *ep != '\0' && *ep != ' ' && *ep != '{'; | |
4823 | ep++) | |
cec68fcb | 4824 | continue; |
96c06863 | 4825 | make_tag (bp, ep - bp, true, |
081b7c70 | 4826 | lb.buffer, ep - lb.buffer + 1, lineno, linecharno); |
93b7ac65 | 4827 | } |
eaaacacd | 4828 | else if (LOOKING_AT (bp, "defineps")) |
081b7c70 | 4829 | get_tag (bp, NULL); |
cec68fcb FP |
4830 | } |
4831 | } | |
4832 | ||
4833 | \f | |
55102b5d FP |
4834 | /* |
4835 | * Forth tags | |
4836 | * Ignore anything after \ followed by space or in ( ) | |
4837 | * Look for words defined by : | |
4838 | * Look for constant, code, create, defer, value, and variable | |
4839 | * OBP extensions: Look for buffer:, field, | |
4840 | * Ideas by Eduardo Horvath <eeh@netbsd.org> (2004) | |
4841 | */ | |
4842 | static void | |
873fbd0b | 4843 | Forth_words (FILE *inf) |
55102b5d FP |
4844 | { |
4845 | register char *bp; | |
4846 | ||
4847 | LOOP_ON_INPUT_LINES (inf, lb, bp) | |
4848 | while ((bp = skip_spaces (bp))[0] != '\0') | |
5e617bc2 | 4849 | if (bp[0] == '\\' && iswhite (bp[1])) |
55102b5d | 4850 | break; /* read next line */ |
5e617bc2 | 4851 | else if (bp[0] == '(' && iswhite (bp[1])) |
55102b5d FP |
4852 | do /* skip to ) or eol */ |
4853 | bp++; | |
4854 | while (*bp != ')' && *bp != '\0'); | |
5e617bc2 | 4855 | else if ((bp[0] == ':' && iswhite (bp[1]) && bp++) |
55102b5d FP |
4856 | || LOOKING_AT_NOCASE (bp, "constant") |
4857 | || LOOKING_AT_NOCASE (bp, "code") | |
4858 | || LOOKING_AT_NOCASE (bp, "create") | |
4859 | || LOOKING_AT_NOCASE (bp, "defer") | |
4860 | || LOOKING_AT_NOCASE (bp, "value") | |
4861 | || LOOKING_AT_NOCASE (bp, "variable") | |
4862 | || LOOKING_AT_NOCASE (bp, "buffer:") | |
4863 | || LOOKING_AT_NOCASE (bp, "field")) | |
4864 | get_tag (skip_spaces (bp), NULL); /* Yay! A definition! */ | |
4865 | else | |
4866 | bp = skip_non_spaces (bp); | |
4867 | } | |
4868 | ||
4869 | \f | |
c6d46f5f JB |
4870 | /* |
4871 | * Scheme tag functions | |
4872 | * look for (def... xyzzy | |
eaaacacd FP |
4873 | * (def... (xyzzy |
4874 | * (def ... ((...(xyzzy .... | |
4875 | * (set! xyzzy | |
4876 | * Original code by Ken Haase (1985?) | |
c6d46f5f | 4877 | */ |
944bdd72 | 4878 | static void |
873fbd0b | 4879 | Scheme_functions (FILE *inf) |
c6d46f5f | 4880 | { |
93b7ac65 | 4881 | register char *bp; |
c6d46f5f | 4882 | |
93b7ac65 | 4883 | LOOP_ON_INPUT_LINES (inf, lb, bp) |
c6d46f5f | 4884 | { |
eaaacacd | 4885 | if (strneq (bp, "(def", 4) || strneq (bp, "(DEF", 4)) |
c6d46f5f | 4886 | { |
eaaacacd | 4887 | bp = skip_non_spaces (bp+4); |
f9b84f9f CY |
4888 | /* Skip over open parens and white space. Don't continue past |
4889 | '\0'. */ | |
4890 | while (*bp && notinname (*bp)) | |
93b7ac65 | 4891 | bp++; |
081b7c70 | 4892 | get_tag (bp, NULL); |
c6d46f5f | 4893 | } |
62aec606 | 4894 | if (LOOKING_AT (bp, "(SET!") || LOOKING_AT (bp, "(set!")) |
081b7c70 | 4895 | get_tag (bp, NULL); |
c6d46f5f JB |
4896 | } |
4897 | } | |
b2521b0a | 4898 | |
c6d46f5f JB |
4899 | \f |
4900 | /* Find tags in TeX and LaTeX input files. */ | |
4901 | ||
4902 | /* TEX_toktab is a table of TeX control sequences that define tags. | |
f6566f90 FP |
4903 | * Each entry records one such control sequence. |
4904 | * | |
4905 | * Original code from who knows whom. | |
4906 | * Ideas by: | |
4907 | * Stefan Monnier (2002) | |
4908 | */ | |
c6d46f5f | 4909 | |
f6566f90 | 4910 | static linebuffer *TEX_toktab = NULL; /* Table with tag tokens */ |
c6d46f5f JB |
4911 | |
4912 | /* Default set of control sequences to put into TEX_toktab. | |
4913 | The value of environment var TEXTAGS is prepended to this. */ | |
988e88ab | 4914 | static const char *TEX_defenv = "\ |
c5007f46 | 4915 | :chapter:section:subsection:subsubsection:eqno:label:ref:cite:bibitem\ |
f6566f90 FP |
4916 | :part:appendix:entry:index:def\ |
4917 | :newcommand:renewcommand:newenvironment:renewenvironment"; | |
c6d46f5f | 4918 | |
f57e2426 | 4919 | static void TEX_mode (FILE *); |
988e88ab | 4920 | static void TEX_decode_env (const char *, const char *); |
c6d46f5f | 4921 | |
db590582 FP |
4922 | static char TEX_esc = '\\'; |
4923 | static char TEX_opgrp = '{'; | |
4924 | static char TEX_clgrp = '}'; | |
c6d46f5f JB |
4925 | |
4926 | /* | |
4927 | * TeX/LaTeX scanning loop. | |
4928 | */ | |
944bdd72 | 4929 | static void |
873fbd0b | 4930 | TeX_commands (FILE *inf) |
c6d46f5f | 4931 | { |
f6566f90 FP |
4932 | char *cp; |
4933 | linebuffer *key; | |
c6d46f5f | 4934 | |
c6d46f5f | 4935 | /* Select either \ or ! as escape character. */ |
6dd5561c | 4936 | TEX_mode (inf); |
c6d46f5f JB |
4937 | |
4938 | /* Initialize token table once from environment. */ | |
f6566f90 FP |
4939 | if (TEX_toktab == NULL) |
4940 | TEX_decode_env ("TEXTAGS", TEX_defenv); | |
c6d46f5f | 4941 | |
93b7ac65 FP |
4942 | LOOP_ON_INPUT_LINES (inf, lb, cp) |
4943 | { | |
f6566f90 FP |
4944 | /* Look at each TEX keyword in line. */ |
4945 | for (;;) | |
8a6c8bcf | 4946 | { |
f6566f90 FP |
4947 | /* Look for a TEX escape. */ |
4948 | while (*cp++ != TEX_esc) | |
4949 | if (cp[-1] == '\0' || cp[-1] == '%') | |
4950 | goto tex_next_line; | |
4951 | ||
4952 | for (key = TEX_toktab; key->buffer != NULL; key++) | |
4953 | if (strneq (cp, key->buffer, key->len)) | |
4954 | { | |
96c06863 | 4955 | char *p; |
081b7c70 | 4956 | int namelen, linelen; |
96c06863 | 4957 | bool opgrp = false; |
f6566f90 FP |
4958 | |
4959 | cp = skip_spaces (cp + key->len); | |
4960 | if (*cp == TEX_opgrp) | |
4961 | { | |
96c06863 | 4962 | opgrp = true; |
f6566f90 FP |
4963 | cp++; |
4964 | } | |
4965 | for (p = cp; | |
4966 | (!iswhite (*p) && *p != '#' && | |
4967 | *p != TEX_opgrp && *p != TEX_clgrp); | |
4968 | p++) | |
4969 | continue; | |
081b7c70 | 4970 | namelen = p - cp; |
f6566f90 FP |
4971 | linelen = lb.len; |
4972 | if (!opgrp || *p == TEX_clgrp) | |
4973 | { | |
4974 | while (*p != '\0' && *p != TEX_opgrp && *p != TEX_clgrp) | |
21cb180b | 4975 | p++; |
f6566f90 FP |
4976 | linelen = p - lb.buffer + 1; |
4977 | } | |
96c06863 | 4978 | make_tag (cp, namelen, true, |
081b7c70 | 4979 | lb.buffer, linelen, lineno, linecharno); |
f6566f90 FP |
4980 | goto tex_next_line; /* We only tag a line once */ |
4981 | } | |
c6d46f5f | 4982 | } |
f6566f90 FP |
4983 | tex_next_line: |
4984 | ; | |
c6d46f5f JB |
4985 | } |
4986 | } | |
4987 | ||
4988 | #define TEX_LESC '\\' | |
4989 | #define TEX_SESC '!' | |
c6d46f5f | 4990 | |
aab1fdae FP |
4991 | /* Figure out whether TeX's escapechar is '\\' or '!' and set grouping |
4992 | chars accordingly. */ | |
944bdd72 | 4993 | static void |
873fbd0b | 4994 | TEX_mode (FILE *inf) |
c6d46f5f JB |
4995 | { |
4996 | int c; | |
4997 | ||
6dd5561c | 4998 | while ((c = getc (inf)) != EOF) |
c6d46f5f JB |
4999 | { |
5000 | /* Skip to next line if we hit the TeX comment char. */ | |
f6566f90 | 5001 | if (c == '%') |
867cc23e | 5002 | while (c != '\n' && c != EOF) |
6dd5561c | 5003 | c = getc (inf); |
c6d46f5f JB |
5004 | else if (c == TEX_LESC || c == TEX_SESC ) |
5005 | break; | |
5006 | } | |
5007 | ||
5008 | if (c == TEX_LESC) | |
5009 | { | |
5010 | TEX_esc = TEX_LESC; | |
5011 | TEX_opgrp = '{'; | |
5012 | TEX_clgrp = '}'; | |
5013 | } | |
5014 | else | |
5015 | { | |
5016 | TEX_esc = TEX_SESC; | |
5017 | TEX_opgrp = '<'; | |
5018 | TEX_clgrp = '>'; | |
5019 | } | |
93b7ac65 FP |
5020 | /* If the input file is compressed, inf is a pipe, and rewind may fail. |
5021 | No attempt is made to correct the situation. */ | |
6dd5561c | 5022 | rewind (inf); |
c6d46f5f JB |
5023 | } |
5024 | ||
aab1fdae FP |
5025 | /* Read environment and prepend it to the default string. |
5026 | Build token table. */ | |
f6566f90 | 5027 | static void |
988e88ab | 5028 | TEX_decode_env (const char *evarname, const char *defenv) |
c6d46f5f | 5029 | { |
988e88ab | 5030 | register const char *env, *p; |
f6566f90 | 5031 | int i, len; |
c6d46f5f JB |
5032 | |
5033 | /* Append default string to environment. */ | |
5034 | env = getenv (evarname); | |
5035 | if (!env) | |
5036 | env = defenv; | |
5037 | else | |
988e88ab | 5038 | env = concat (env, defenv, ""); |
c6d46f5f JB |
5039 | |
5040 | /* Allocate a token table */ | |
f6566f90 | 5041 | for (len = 1, p = env; p;) |
93b7ac65 | 5042 | if ((p = etags_strchr (p, ':')) && *++p != '\0') |
f6566f90 FP |
5043 | len++; |
5044 | TEX_toktab = xnew (len, linebuffer); | |
c6d46f5f JB |
5045 | |
5046 | /* Unpack environment string into token table. Be careful about */ | |
5047 | /* zero-length strings (leading ':', "::" and trailing ':') */ | |
f6566f90 | 5048 | for (i = 0; *env != '\0';) |
c6d46f5f | 5049 | { |
b02c5fea | 5050 | p = etags_strchr (env, ':'); |
c6d46f5f JB |
5051 | if (!p) /* End of environment string. */ |
5052 | p = env + strlen (env); | |
5053 | if (p - env > 0) | |
5054 | { /* Only non-zero strings. */ | |
f6566f90 FP |
5055 | TEX_toktab[i].buffer = savenstr (env, p - env); |
5056 | TEX_toktab[i].len = p - env; | |
c6d46f5f JB |
5057 | i++; |
5058 | } | |
5059 | if (*p) | |
5060 | env = p + 1; | |
5061 | else | |
5062 | { | |
f6566f90 FP |
5063 | TEX_toktab[i].buffer = NULL; /* Mark end of table. */ |
5064 | TEX_toktab[i].len = 0; | |
c6d46f5f JB |
5065 | break; |
5066 | } | |
5067 | } | |
c6d46f5f | 5068 | } |
b2521b0a | 5069 | |
c6d46f5f | 5070 | \f |
97052c63 DL |
5071 | /* Texinfo support. Dave Love, Mar. 2000. */ |
5072 | static void | |
873fbd0b | 5073 | Texinfo_nodes (FILE *inf) |
97052c63 DL |
5074 | { |
5075 | char *cp, *start; | |
5076 | LOOP_ON_INPUT_LINES (inf, lb, cp) | |
62aec606 FP |
5077 | if (LOOKING_AT (cp, "@node")) |
5078 | { | |
5079 | start = cp; | |
5080 | while (*cp != '\0' && *cp != ',') | |
5081 | cp++; | |
96c06863 | 5082 | make_tag (start, cp - start, true, |
081b7c70 | 5083 | lb.buffer, cp - lb.buffer + 1, lineno, linecharno); |
62aec606 | 5084 | } |
97052c63 | 5085 | } |
b2521b0a | 5086 | |
97052c63 | 5087 | \f |
61a1f6fa FP |
5088 | /* |
5089 | * HTML support. | |
5090 | * Contents of <title>, <h1>, <h2>, <h3> are tags. | |
5091 | * Contents of <a name=xxx> are tags with name xxx. | |
5092 | * | |
c38e0c97 | 5093 | * Francesco Potortì, 2002. |
61a1f6fa FP |
5094 | */ |
5095 | static void | |
873fbd0b | 5096 | HTML_labels (FILE *inf) |
61a1f6fa | 5097 | { |
96c06863 PE |
5098 | bool getnext = false; /* next text outside of HTML tags is a tag */ |
5099 | bool skiptag = false; /* skip to the end of the current HTML tag */ | |
5100 | bool intag = false; /* inside an html tag, looking for ID= */ | |
5101 | bool inanchor = false; /* when INTAG, is an anchor, look for NAME= */ | |
61a1f6fa FP |
5102 | char *end; |
5103 | ||
5104 | ||
5105 | linebuffer_setlen (&token_name, 0); /* no name in buffer */ | |
5106 | ||
5107 | LOOP_ON_INPUT_LINES (inf, lb, dbp) | |
3c04a71a FP |
5108 | for (;;) /* loop on the same line */ |
5109 | { | |
5110 | if (skiptag) /* skip HTML tag */ | |
61a1f6fa FP |
5111 | { |
5112 | while (*dbp != '\0' && *dbp != '>') | |
5113 | dbp++; | |
5114 | if (*dbp == '>') | |
5115 | { | |
5116 | dbp += 1; | |
96c06863 | 5117 | skiptag = false; |
61a1f6fa FP |
5118 | continue; /* look on the same line */ |
5119 | } | |
5120 | break; /* go to next line */ | |
5121 | } | |
5122 | ||
3c04a71a | 5123 | else if (intag) /* look for "name=" or "id=" */ |
61a1f6fa | 5124 | { |
3c04a71a FP |
5125 | while (*dbp != '\0' && *dbp != '>' |
5126 | && lowcase (*dbp) != 'n' && lowcase (*dbp) != 'i') | |
61a1f6fa FP |
5127 | dbp++; |
5128 | if (*dbp == '\0') | |
5129 | break; /* go to next line */ | |
5130 | if (*dbp == '>') | |
5131 | { | |
5132 | dbp += 1; | |
96c06863 | 5133 | intag = false; |
61a1f6fa FP |
5134 | continue; /* look on the same line */ |
5135 | } | |
3c04a71a FP |
5136 | if ((inanchor && LOOKING_AT_NOCASE (dbp, "name=")) |
5137 | || LOOKING_AT_NOCASE (dbp, "id=")) | |
61a1f6fa FP |
5138 | { |
5139 | bool quoted = (dbp[0] == '"'); | |
5140 | ||
5141 | if (quoted) | |
5142 | for (end = ++dbp; *end != '\0' && *end != '"'; end++) | |
5143 | continue; | |
5144 | else | |
5145 | for (end = dbp; *end != '\0' && intoken (*end); end++) | |
5146 | continue; | |
5147 | linebuffer_setlen (&token_name, end - dbp); | |
e99a530f | 5148 | memcpy (token_name.buffer, dbp, end - dbp); |
61a1f6fa FP |
5149 | token_name.buffer[end - dbp] = '\0'; |
5150 | ||
5151 | dbp = end; | |
96c06863 PE |
5152 | intag = false; /* we found what we looked for */ |
5153 | skiptag = true; /* skip to the end of the tag */ | |
5154 | getnext = true; /* then grab the text */ | |
61a1f6fa FP |
5155 | continue; /* look on the same line */ |
5156 | } | |
3c04a71a | 5157 | dbp += 1; |
61a1f6fa FP |
5158 | } |
5159 | ||
5160 | else if (getnext) /* grab next tokens and tag them */ | |
5161 | { | |
5162 | dbp = skip_spaces (dbp); | |
5163 | if (*dbp == '\0') | |
5164 | break; /* go to next line */ | |
5165 | if (*dbp == '<') | |
5166 | { | |
96c06863 | 5167 | intag = true; |
3c04a71a | 5168 | inanchor = (lowcase (dbp[1]) == 'a' && !intoken (dbp[2])); |
61a1f6fa FP |
5169 | continue; /* look on the same line */ |
5170 | } | |
5171 | ||
5172 | for (end = dbp + 1; *end != '\0' && *end != '<'; end++) | |
5173 | continue; | |
96c06863 | 5174 | make_tag (token_name.buffer, token_name.len, true, |
61a1f6fa FP |
5175 | dbp, end - dbp, lineno, linecharno); |
5176 | linebuffer_setlen (&token_name, 0); /* no name in buffer */ | |
96c06863 | 5177 | getnext = false; |
61a1f6fa FP |
5178 | break; /* go to next line */ |
5179 | } | |
5180 | ||
5181 | else /* look for an interesting HTML tag */ | |
5182 | { | |
5183 | while (*dbp != '\0' && *dbp != '<') | |
5184 | dbp++; | |
5185 | if (*dbp == '\0') | |
5186 | break; /* go to next line */ | |
96c06863 | 5187 | intag = true; |
3c04a71a | 5188 | if (lowcase (dbp[1]) == 'a' && !intoken (dbp[2])) |
61a1f6fa | 5189 | { |
96c06863 | 5190 | inanchor = true; |
3c04a71a FP |
5191 | continue; /* look on the same line */ |
5192 | } | |
5193 | else if (LOOKING_AT_NOCASE (dbp, "<title>") | |
5194 | || LOOKING_AT_NOCASE (dbp, "<h1>") | |
5195 | || LOOKING_AT_NOCASE (dbp, "<h2>") | |
5196 | || LOOKING_AT_NOCASE (dbp, "<h3>")) | |
5197 | { | |
96c06863 PE |
5198 | intag = false; |
5199 | getnext = true; | |
61a1f6fa FP |
5200 | continue; /* look on the same line */ |
5201 | } | |
3c04a71a | 5202 | dbp += 1; |
61a1f6fa | 5203 | } |
3c04a71a | 5204 | } |
61a1f6fa FP |
5205 | } |
5206 | ||
5207 | \f | |
8dc7496c | 5208 | /* |
eaaacacd | 5209 | * Prolog support |
8dc7496c | 5210 | * |
eaaacacd FP |
5211 | * Assumes that the predicate or rule starts at column 0. |
5212 | * Only the first clause of a predicate or rule is added. | |
5213 | * Original code by Sunichirou Sugou (1989) | |
5214 | * Rewritten by Anders Lindgren (1996) | |
8dc7496c | 5215 | */ |
3ef271f2 | 5216 | static size_t prolog_pr (char *, char *); |
f57e2426 | 5217 | static void prolog_skip_comment (linebuffer *, FILE *); |
3ef271f2 | 5218 | static size_t prolog_atom (char *, size_t); |
30903246 | 5219 | |
944bdd72 | 5220 | static void |
873fbd0b | 5221 | Prolog_functions (FILE *inf) |
8dc7496c | 5222 | { |
93b7ac65 | 5223 | char *cp, *last; |
3ef271f2 PE |
5224 | size_t len; |
5225 | size_t allocated; | |
8dc7496c FP |
5226 | |
5227 | allocated = 0; | |
5228 | len = 0; | |
5229 | last = NULL; | |
5230 | ||
93b7ac65 | 5231 | LOOP_ON_INPUT_LINES (inf, lb, cp) |
8dc7496c | 5232 | { |
93b7ac65 | 5233 | if (cp[0] == '\0') /* Empty line */ |
8dc7496c | 5234 | continue; |
71cbb895 | 5235 | else if (iswhite (cp[0])) /* Not a predicate */ |
8dc7496c | 5236 | continue; |
93b7ac65 | 5237 | else if (cp[0] == '/' && cp[1] == '*') /* comment. */ |
3f1c8fcd | 5238 | prolog_skip_comment (&lb, inf); |
62aec606 | 5239 | else if ((len = prolog_pr (cp, last)) > 0) |
8dc7496c | 5240 | { |
eaaacacd FP |
5241 | /* Predicate or rule. Store the function name so that we |
5242 | only generate a tag for the first clause. */ | |
8dc7496c | 5243 | if (last == NULL) |
5e617bc2 | 5244 | last = xnew (len + 1, char); |
8dc7496c | 5245 | else if (len + 1 > allocated) |
b2521b0a | 5246 | xrnew (last, len + 1, char); |
8dc7496c | 5247 | allocated = len + 1; |
e99a530f | 5248 | memcpy (last, cp, len); |
8dc7496c FP |
5249 | last[len] = '\0'; |
5250 | } | |
5251 | } | |
c2cd06e6 | 5252 | free (last); |
8dc7496c FP |
5253 | } |
5254 | ||
c6d46f5f | 5255 | |
944bdd72 | 5256 | static void |
873fbd0b | 5257 | prolog_skip_comment (linebuffer *plb, FILE *inf) |
8dc7496c FP |
5258 | { |
5259 | char *cp; | |
5260 | ||
5261 | do | |
5262 | { | |
5263 | for (cp = plb->buffer; *cp != '\0'; cp++) | |
5264 | if (cp[0] == '*' && cp[1] == '/') | |
5265 | return; | |
e7d3b099 | 5266 | readline (plb, inf); |
8dc7496c | 5267 | } |
5e617bc2 | 5268 | while (!feof (inf)); |
8dc7496c FP |
5269 | } |
5270 | ||
5271 | /* | |
62aec606 | 5272 | * A predicate or rule definition is added if it matches: |
8dc7496c | 5273 | * <beginning of line><Prolog Atom><whitespace>( |
62aec606 | 5274 | * or <beginning of line><Prolog Atom><whitespace>:- |
8dc7496c FP |
5275 | * |
5276 | * It is added to the tags database if it doesn't match the | |
5277 | * name of the previous clause header. | |
5278 | * | |
62aec606 FP |
5279 | * Return the size of the name of the predicate or rule, or 0 if no |
5280 | * header was found. | |
8dc7496c | 5281 | */ |
3ef271f2 | 5282 | static size_t |
873fbd0b | 5283 | prolog_pr (char *s, char *last) |
f68c809d | 5284 | |
873fbd0b | 5285 | /* Name of last clause. */ |
c6d46f5f | 5286 | { |
3ef271f2 PE |
5287 | size_t pos; |
5288 | size_t len; | |
8dc7496c | 5289 | |
30903246 | 5290 | pos = prolog_atom (s, 0); |
3ef271f2 | 5291 | if (! pos) |
8dc7496c FP |
5292 | return 0; |
5293 | ||
5294 | len = pos; | |
93b7ac65 | 5295 | pos = skip_spaces (s + pos) - s; |
8dc7496c | 5296 | |
62aec606 FP |
5297 | if ((s[pos] == '.' |
5298 | || (s[pos] == '(' && (pos += 1)) | |
5299 | || (s[pos] == ':' && s[pos + 1] == '-' && (pos += 2))) | |
5300 | && (last == NULL /* save only the first clause */ | |
3ef271f2 | 5301 | || len != strlen (last) |
62aec606 | 5302 | || !strneq (s, last, len))) |
c6d46f5f | 5303 | { |
96c06863 | 5304 | make_tag (s, len, true, s, pos, lineno, linecharno); |
8dc7496c | 5305 | return len; |
c6d46f5f | 5306 | } |
62aec606 FP |
5307 | else |
5308 | return 0; | |
8dc7496c FP |
5309 | } |
5310 | ||
5311 | /* | |
5312 | * Consume a Prolog atom. | |
3ef271f2 | 5313 | * Return the number of bytes consumed, or 0 if there was an error. |
8dc7496c FP |
5314 | * |
5315 | * A prolog atom, in this context, could be one of: | |
5316 | * - An alphanumeric sequence, starting with a lower case letter. | |
5317 | * - A quoted arbitrary string. Single quotes can escape themselves. | |
5318 | * Backslash quotes everything. | |
5319 | */ | |
3ef271f2 PE |
5320 | static size_t |
5321 | prolog_atom (char *s, size_t pos) | |
8dc7496c | 5322 | { |
3ef271f2 | 5323 | size_t origpos; |
8dc7496c FP |
5324 | |
5325 | origpos = pos; | |
5326 | ||
5e617bc2 | 5327 | if (ISLOWER (s[pos]) || (s[pos] == '_')) |
8dc7496c FP |
5328 | { |
5329 | /* The atom is unquoted. */ | |
5330 | pos++; | |
5e617bc2 | 5331 | while (ISALNUM (s[pos]) || (s[pos] == '_')) |
c6d46f5f | 5332 | { |
8dc7496c | 5333 | pos++; |
c6d46f5f | 5334 | } |
8dc7496c FP |
5335 | return pos - origpos; |
5336 | } | |
5337 | else if (s[pos] == '\'') | |
5338 | { | |
5339 | pos++; | |
5340 | ||
97b90b0a | 5341 | for (;;) |
c6d46f5f | 5342 | { |
8dc7496c FP |
5343 | if (s[pos] == '\'') |
5344 | { | |
5345 | pos++; | |
5346 | if (s[pos] != '\'') | |
5347 | break; | |
5348 | pos++; /* A double quote */ | |
5349 | } | |
5350 | else if (s[pos] == '\0') | |
5351 | /* Multiline quoted atoms are ignored. */ | |
3ef271f2 | 5352 | return 0; |
8dc7496c FP |
5353 | else if (s[pos] == '\\') |
5354 | { | |
5355 | if (s[pos+1] == '\0') | |
3ef271f2 | 5356 | return 0; |
8dc7496c FP |
5357 | pos += 2; |
5358 | } | |
5359 | else | |
5360 | pos++; | |
c6d46f5f | 5361 | } |
8dc7496c | 5362 | return pos - origpos; |
c6d46f5f | 5363 | } |
8dc7496c | 5364 | else |
3ef271f2 | 5365 | return 0; |
c6d46f5f | 5366 | } |
b2521b0a | 5367 | |
8dc7496c | 5368 | \f |
93b7ac65 | 5369 | /* |
eaaacacd | 5370 | * Support for Erlang |
8dc7496c FP |
5371 | * |
5372 | * Generates tags for functions, defines, and records. | |
8dc7496c | 5373 | * Assumes that Erlang functions start at column 0. |
eaaacacd | 5374 | * Original code by Anders Lindgren (1996) |
8dc7496c | 5375 | */ |
f57e2426 J |
5376 | static int erlang_func (char *, char *); |
5377 | static void erlang_attribute (char *); | |
5378 | static int erlang_atom (char *); | |
30903246 | 5379 | |
944bdd72 | 5380 | static void |
873fbd0b | 5381 | Erlang_functions (FILE *inf) |
c6d46f5f | 5382 | { |
93b7ac65 | 5383 | char *cp, *last; |
8dc7496c FP |
5384 | int len; |
5385 | int allocated; | |
5386 | ||
5387 | allocated = 0; | |
5388 | len = 0; | |
5389 | last = NULL; | |
5390 | ||
93b7ac65 | 5391 | LOOP_ON_INPUT_LINES (inf, lb, cp) |
c6d46f5f | 5392 | { |
93b7ac65 | 5393 | if (cp[0] == '\0') /* Empty line */ |
c6d46f5f | 5394 | continue; |
71cbb895 | 5395 | else if (iswhite (cp[0])) /* Not function nor attribute */ |
c6d46f5f | 5396 | continue; |
93b7ac65 | 5397 | else if (cp[0] == '%') /* comment */ |
8dc7496c | 5398 | continue; |
93b7ac65 | 5399 | else if (cp[0] == '"') /* Sometimes, strings start in column one */ |
8dc7496c | 5400 | continue; |
93b7ac65 | 5401 | else if (cp[0] == '-') /* attribute, e.g. "-define" */ |
8dc7496c | 5402 | { |
93b7ac65 | 5403 | erlang_attribute (cp); |
7e237d24 FP |
5404 | if (last != NULL) |
5405 | { | |
5406 | free (last); | |
5407 | last = NULL; | |
5408 | } | |
8dc7496c | 5409 | } |
93b7ac65 | 5410 | else if ((len = erlang_func (cp, last)) > 0) |
8dc7496c | 5411 | { |
93b7ac65 | 5412 | /* |
8dc7496c FP |
5413 | * Function. Store the function name so that we only |
5414 | * generates a tag for the first clause. | |
5415 | */ | |
5416 | if (last == NULL) | |
30903246 | 5417 | last = xnew (len + 1, char); |
8dc7496c | 5418 | else if (len + 1 > allocated) |
b2521b0a | 5419 | xrnew (last, len + 1, char); |
8dc7496c | 5420 | allocated = len + 1; |
e99a530f | 5421 | memcpy (last, cp, len); |
8dc7496c FP |
5422 | last[len] = '\0'; |
5423 | } | |
5424 | } | |
c2cd06e6 | 5425 | free (last); |
8dc7496c FP |
5426 | } |
5427 | ||
5428 | ||
5429 | /* | |
5430 | * A function definition is added if it matches: | |
5431 | * <beginning of line><Erlang Atom><whitespace>( | |
5432 | * | |
5433 | * It is added to the tags database if it doesn't match the | |
5434 | * name of the previous clause header. | |
5435 | * | |
5436 | * Return the size of the name of the function, or 0 if no function | |
5437 | * was found. | |
5438 | */ | |
0dacfc4b | 5439 | static int |
873fbd0b | 5440 | erlang_func (char *s, char *last) |
f68c809d | 5441 | |
873fbd0b | 5442 | /* Name of last clause. */ |
8dc7496c | 5443 | { |
8dc7496c FP |
5444 | int pos; |
5445 | int len; | |
5446 | ||
89fb2be1 | 5447 | pos = erlang_atom (s); |
8dc7496c FP |
5448 | if (pos < 1) |
5449 | return 0; | |
5450 | ||
5451 | len = pos; | |
93b7ac65 | 5452 | pos = skip_spaces (s + pos) - s; |
8dc7496c | 5453 | |
30903246 FP |
5454 | /* Save only the first clause. */ |
5455 | if (s[pos++] == '(' | |
5456 | && (last == NULL | |
93b7ac65 | 5457 | || len != (int)strlen (last) |
30903246 | 5458 | || !strneq (s, last, len))) |
8dc7496c | 5459 | { |
96c06863 | 5460 | make_tag (s, len, true, s, pos, lineno, linecharno); |
8dc7496c FP |
5461 | return len; |
5462 | } | |
30903246 | 5463 | |
8dc7496c | 5464 | return 0; |
c6d46f5f JB |
5465 | } |
5466 | ||
8dc7496c FP |
5467 | |
5468 | /* | |
93b7ac65 | 5469 | * Handle attributes. Currently, tags are generated for defines |
8dc7496c FP |
5470 | * and records. |
5471 | * | |
5472 | * They are on the form: | |
5473 | * -define(foo, bar). | |
5474 | * -define(Foo(M, N), M+N). | |
5475 | * -record(graph, {vtab = notable, cyclic = true}). | |
5476 | */ | |
944bdd72 | 5477 | static void |
873fbd0b | 5478 | erlang_attribute (char *s) |
c6d46f5f | 5479 | { |
89fb2be1 | 5480 | char *cp = s; |
8dc7496c | 5481 | |
89fb2be1 FP |
5482 | if ((LOOKING_AT (cp, "-define") || LOOKING_AT (cp, "-record")) |
5483 | && *cp++ == '(') | |
c6d46f5f | 5484 | { |
89fb2be1 FP |
5485 | int len = erlang_atom (skip_spaces (cp)); |
5486 | if (len > 0) | |
96c06863 | 5487 | make_tag (cp, len, true, s, cp + len - s, lineno, linecharno); |
b9755a12 | 5488 | } |
8dc7496c FP |
5489 | return; |
5490 | } | |
5491 | ||
5492 | ||
5493 | /* | |
5494 | * Consume an Erlang atom (or variable). | |
5495 | * Return the number of bytes consumed, or -1 if there was an error. | |
5496 | */ | |
944bdd72 | 5497 | static int |
873fbd0b | 5498 | erlang_atom (char *s) |
8dc7496c | 5499 | { |
89fb2be1 | 5500 | int pos = 0; |
8dc7496c | 5501 | |
af03e6ab | 5502 | if (ISALPHA (s[pos]) || s[pos] == '_') |
8dc7496c FP |
5503 | { |
5504 | /* The atom is unquoted. */ | |
89fb2be1 | 5505 | do |
8dc7496c | 5506 | pos++; |
89fb2be1 | 5507 | while (ISALNUM (s[pos]) || s[pos] == '_'); |
8dc7496c FP |
5508 | } |
5509 | else if (s[pos] == '\'') | |
5510 | { | |
89fb2be1 FP |
5511 | for (pos++; s[pos] != '\''; pos++) |
5512 | if (s[pos] == '\0' /* multiline quoted atoms are ignored */ | |
5513 | || (s[pos] == '\\' && s[++pos] == '\0')) | |
5514 | return 0; | |
8dc7496c | 5515 | pos++; |
8dc7496c | 5516 | } |
89fb2be1 FP |
5517 | |
5518 | return pos; | |
8dc7496c | 5519 | } |
b2521b0a | 5520 | |
b9755a12 | 5521 | \f |
f57e2426 J |
5522 | static char *scan_separators (char *); |
5523 | static void add_regex (char *, language *); | |
5524 | static char *substitute (char *, char *, struct re_registers *); | |
71cbb895 | 5525 | |
24dbe96a FP |
5526 | /* |
5527 | * Take a string like "/blah/" and turn it into "blah", verifying | |
5528 | * that the first and last characters are the same, and handling | |
5529 | * quoted separator characters. Actually, stops on the occurrence of | |
5530 | * an unquoted separator. Also process \t, \n, etc. and turn into | |
5531 | * appropriate characters. Works in place. Null terminates name string. | |
5532 | * Returns pointer to terminating separator, or NULL for | |
5533 | * unterminated regexps. | |
5534 | */ | |
944bdd72 | 5535 | static char * |
873fbd0b | 5536 | scan_separators (char *name) |
b9755a12 FP |
5537 | { |
5538 | char sep = name[0]; | |
5539 | char *copyto = name; | |
96c06863 | 5540 | bool quoted = false; |
b9755a12 FP |
5541 | |
5542 | for (++name; *name != '\0'; ++name) | |
5543 | { | |
5544 | if (quoted) | |
5545 | { | |
e7d3b099 | 5546 | switch (*name) |
b9755a12 | 5547 | { |
24dbe96a FP |
5548 | case 'a': *copyto++ = '\007'; break; /* BEL (bell) */ |
5549 | case 'b': *copyto++ = '\b'; break; /* BS (back space) */ | |
5550 | case 'd': *copyto++ = 0177; break; /* DEL (delete) */ | |
5551 | case 'e': *copyto++ = 033; break; /* ESC (delete) */ | |
5552 | case 'f': *copyto++ = '\f'; break; /* FF (form feed) */ | |
5553 | case 'n': *copyto++ = '\n'; break; /* NL (new line) */ | |
5554 | case 'r': *copyto++ = '\r'; break; /* CR (carriage return) */ | |
5555 | case 't': *copyto++ = '\t'; break; /* TAB (horizontal tab) */ | |
5556 | case 'v': *copyto++ = '\v'; break; /* VT (vertical tab) */ | |
e7d3b099 FP |
5557 | default: |
5558 | if (*name == sep) | |
5559 | *copyto++ = sep; | |
5560 | else | |
5561 | { | |
5562 | /* Something else is quoted, so preserve the quote. */ | |
5563 | *copyto++ = '\\'; | |
5564 | *copyto++ = *name; | |
5565 | } | |
5566 | break; | |
b9755a12 | 5567 | } |
96c06863 | 5568 | quoted = false; |
b9755a12 FP |
5569 | } |
5570 | else if (*name == '\\') | |
96c06863 | 5571 | quoted = true; |
b9755a12 FP |
5572 | else if (*name == sep) |
5573 | break; | |
5574 | else | |
5575 | *copyto++ = *name; | |
5576 | } | |
24dbe96a FP |
5577 | if (*name != sep) |
5578 | name = NULL; /* signal unterminated regexp */ | |
c6d46f5f | 5579 | |
b9755a12 FP |
5580 | /* Terminate copied string. */ |
5581 | *copyto = '\0'; | |
5582 | return name; | |
5583 | } | |
c6d46f5f | 5584 | |
cec68fcb | 5585 | /* Look at the argument of --regex or --no-regex and do the right |
93b7ac65 | 5586 | thing. Same for each line of a regexp file. */ |
944bdd72 | 5587 | static void |
873fbd0b | 5588 | analyse_regex (char *regex_arg) |
c6d46f5f | 5589 | { |
cec68fcb | 5590 | if (regex_arg == NULL) |
76fecdff | 5591 | { |
61a1f6fa | 5592 | free_regexps (); /* --no-regex: remove existing regexps */ |
76fecdff FP |
5593 | return; |
5594 | } | |
93b7ac65 FP |
5595 | |
5596 | /* A real --regexp option or a line in a regexp file. */ | |
5597 | switch (regex_arg[0]) | |
b9755a12 | 5598 | { |
93b7ac65 FP |
5599 | /* Comments in regexp file or null arg to --regex. */ |
5600 | case '\0': | |
5601 | case ' ': | |
5602 | case '\t': | |
5603 | break; | |
cec68fcb | 5604 | |
93b7ac65 FP |
5605 | /* Read a regex file. This is recursive and may result in a |
5606 | loop, which will stop when the file descriptors are exhausted. */ | |
5607 | case '@': | |
5608 | { | |
5609 | FILE *regexfp; | |
5610 | linebuffer regexbuf; | |
5611 | char *regexfile = regex_arg + 1; | |
5612 | ||
5613 | /* regexfile is a file containing regexps, one per line. */ | |
5614 | regexfp = fopen (regexfile, "r"); | |
5615 | if (regexfp == NULL) | |
d9df6f40 | 5616 | pfatal (regexfile); |
61a1f6fa | 5617 | linebuffer_init (®exbuf); |
93b7ac65 | 5618 | while (readline_internal (®exbuf, regexfp) > 0) |
24dbe96a | 5619 | analyse_regex (regexbuf.buffer); |
93b7ac65 FP |
5620 | free (regexbuf.buffer); |
5621 | fclose (regexfp); | |
5622 | } | |
5623 | break; | |
5624 | ||
5625 | /* Regexp to be used for a specific language only. */ | |
5626 | case '{': | |
5627 | { | |
5628 | language *lang; | |
5629 | char *lang_name = regex_arg + 1; | |
5630 | char *cp; | |
5631 | ||
5632 | for (cp = lang_name; *cp != '}'; cp++) | |
5633 | if (*cp == '\0') | |
5634 | { | |
5635 | error ("unterminated language name in regex: %s", regex_arg); | |
5636 | return; | |
5637 | } | |
24dbe96a | 5638 | *cp++ = '\0'; |
89d57763 | 5639 | lang = get_language_from_langname (lang_name); |
93b7ac65 | 5640 | if (lang == NULL) |
cec68fcb | 5641 | return; |
24dbe96a | 5642 | add_regex (cp, lang); |
93b7ac65 FP |
5643 | } |
5644 | break; | |
5645 | ||
5646 | /* Regexp to be used for any language. */ | |
5647 | default: | |
24dbe96a | 5648 | add_regex (regex_arg, NULL); |
93b7ac65 | 5649 | break; |
cec68fcb FP |
5650 | } |
5651 | } | |
5652 | ||
61a1f6fa FP |
5653 | /* Separate the regexp pattern, compile it, |
5654 | and care for optional name and modifiers. */ | |
944bdd72 | 5655 | static void |
873fbd0b | 5656 | add_regex (char *regexp_pattern, language *lang) |
cec68fcb | 5657 | { |
7e88eb25 | 5658 | static struct re_pattern_buffer zeropattern; |
24dbe96a | 5659 | char sep, *pat, *name, *modifiers; |
db5a3003 | 5660 | char empty = '\0'; |
cec68fcb FP |
5661 | const char *err; |
5662 | struct re_pattern_buffer *patbuf; | |
61a1f6fa FP |
5663 | regexp *rp; |
5664 | bool | |
96c06863 PE |
5665 | force_explicit_name = true, /* do not use implicit tag names */ |
5666 | ignore_case = false, /* case is significant */ | |
5667 | multi_line = false, /* matches are done one line at a time */ | |
5668 | single_line = false; /* dot does not match newline */ | |
cec68fcb FP |
5669 | |
5670 | ||
5e617bc2 | 5671 | if (strlen (regexp_pattern) < 3) |
b9755a12 | 5672 | { |
db5a3003 | 5673 | error ("null regexp"); |
b9755a12 FP |
5674 | return; |
5675 | } | |
24dbe96a | 5676 | sep = regexp_pattern[0]; |
b9755a12 | 5677 | name = scan_separators (regexp_pattern); |
24dbe96a | 5678 | if (name == NULL) |
b9755a12 | 5679 | { |
24dbe96a FP |
5680 | error ("%s: unterminated regexp", regexp_pattern); |
5681 | return; | |
5682 | } | |
5683 | if (name[1] == sep) | |
5684 | { | |
5685 | error ("null name for regexp \"%s\"", regexp_pattern); | |
b9755a12 FP |
5686 | return; |
5687 | } | |
24dbe96a FP |
5688 | modifiers = scan_separators (name); |
5689 | if (modifiers == NULL) /* no terminating separator --> no name */ | |
5690 | { | |
5691 | modifiers = name; | |
db5a3003 | 5692 | name = ∅ |
24dbe96a FP |
5693 | } |
5694 | else | |
5695 | modifiers += 1; /* skip separator */ | |
5696 | ||
5697 | /* Parse regex modifiers. */ | |
24dbe96a FP |
5698 | for (; modifiers[0] != '\0'; modifiers++) |
5699 | switch (modifiers[0]) | |
5700 | { | |
61a1f6fa FP |
5701 | case 'N': |
5702 | if (modifiers == name) | |
db5a3003 | 5703 | error ("forcing explicit tag name but no name, ignoring"); |
96c06863 | 5704 | force_explicit_name = true; |
61a1f6fa | 5705 | break; |
24dbe96a | 5706 | case 'i': |
96c06863 | 5707 | ignore_case = true; |
24dbe96a FP |
5708 | break; |
5709 | case 's': | |
96c06863 | 5710 | single_line = true; |
24dbe96a FP |
5711 | /* FALLTHRU */ |
5712 | case 'm': | |
96c06863 PE |
5713 | multi_line = true; |
5714 | need_filebuf = true; | |
24dbe96a FP |
5715 | break; |
5716 | default: | |
db5a3003 | 5717 | error ("invalid regexp modifier `%c', ignoring", modifiers[0]); |
2b749964 | 5718 | break; |
24dbe96a | 5719 | } |
b9755a12 FP |
5720 | |
5721 | patbuf = xnew (1, struct re_pattern_buffer); | |
7e88eb25 FP |
5722 | *patbuf = zeropattern; |
5723 | if (ignore_case) | |
24dbe96a FP |
5724 | { |
5725 | static char lc_trans[CHARS]; | |
5726 | int i; | |
5727 | for (i = 0; i < CHARS; i++) | |
5728 | lc_trans[i] = lowcase (i); | |
5729 | patbuf->translate = lc_trans; /* translation table to fold case */ | |
5730 | } | |
5731 | ||
5732 | if (multi_line) | |
5733 | pat = concat ("^", regexp_pattern, ""); /* anchor to beginning of line */ | |
5734 | else | |
5735 | pat = regexp_pattern; | |
b9755a12 | 5736 | |
24dbe96a FP |
5737 | if (single_line) |
5738 | re_set_syntax (RE_SYNTAX_EMACS | RE_DOT_NEWLINE); | |
5739 | else | |
5740 | re_set_syntax (RE_SYNTAX_EMACS); | |
5741 | ||
3d864325 | 5742 | err = re_compile_pattern (pat, strlen (pat), patbuf); |
24dbe96a FP |
5743 | if (multi_line) |
5744 | free (pat); | |
b9755a12 FP |
5745 | if (err != NULL) |
5746 | { | |
5747 | error ("%s while compiling pattern", err); | |
5748 | return; | |
5749 | } | |
5750 | ||
61a1f6fa FP |
5751 | rp = p_head; |
5752 | p_head = xnew (1, regexp); | |
5753 | p_head->pattern = savestr (regexp_pattern); | |
5754 | p_head->p_next = rp; | |
eaaacacd FP |
5755 | p_head->lang = lang; |
5756 | p_head->pat = patbuf; | |
61a1f6fa | 5757 | p_head->name = savestr (name); |
96c06863 | 5758 | p_head->error_signaled = false; |
61a1f6fa | 5759 | p_head->force_explicit_name = force_explicit_name; |
9e0a3f98 | 5760 | p_head->ignore_case = ignore_case; |
24dbe96a | 5761 | p_head->multi_line = multi_line; |
b9755a12 FP |
5762 | } |
5763 | ||
5764 | /* | |
c5007f46 | 5765 | * Do the substitutions indicated by the regular expression and |
b9755a12 FP |
5766 | * arguments. |
5767 | */ | |
944bdd72 | 5768 | static char * |
873fbd0b | 5769 | substitute (char *in, char *out, struct re_registers *regs) |
b9755a12 | 5770 | { |
2f608d34 | 5771 | char *result, *t; |
e0903a92 | 5772 | int size, dig, diglen; |
2f608d34 FP |
5773 | |
5774 | result = NULL; | |
5775 | size = strlen (out); | |
b9755a12 | 5776 | |
e0903a92 FP |
5777 | /* Pass 1: figure out how much to allocate by finding all \N strings. */ |
5778 | if (out[size - 1] == '\\') | |
5779 | fatal ("pattern error in \"%s\"", out); | |
5780 | for (t = etags_strchr (out, '\\'); | |
5781 | t != NULL; | |
5782 | t = etags_strchr (t + 2, '\\')) | |
af03e6ab | 5783 | if (ISDIGIT (t[1])) |
2f608d34 | 5784 | { |
e0903a92 FP |
5785 | dig = t[1] - '0'; |
5786 | diglen = regs->end[dig] - regs->start[dig]; | |
5787 | size += diglen - 2; | |
2f608d34 | 5788 | } |
e0903a92 FP |
5789 | else |
5790 | size -= 1; | |
b9755a12 FP |
5791 | |
5792 | /* Allocate space and do the substitutions. */ | |
a127d423 | 5793 | assert (size >= 0); |
b9755a12 | 5794 | result = xnew (size + 1, char); |
e0903a92 FP |
5795 | |
5796 | for (t = result; *out != '\0'; out++) | |
af03e6ab | 5797 | if (*out == '\\' && ISDIGIT (*++out)) |
e0903a92 | 5798 | { |
e0903a92 FP |
5799 | dig = *out - '0'; |
5800 | diglen = regs->end[dig] - regs->start[dig]; | |
e99a530f | 5801 | memcpy (t, in + regs->start[dig], diglen); |
e0903a92 FP |
5802 | t += diglen; |
5803 | } | |
5804 | else | |
5805 | *t++ = *out; | |
5806 | *t = '\0'; | |
5807 | ||
a127d423 FP |
5808 | assert (t <= result + size); |
5809 | assert (t - result == (int)strlen (result)); | |
b9755a12 FP |
5810 | |
5811 | return result; | |
c6d46f5f | 5812 | } |
93b7ac65 | 5813 | |
61a1f6fa | 5814 | /* Deallocate all regexps. */ |
944bdd72 | 5815 | static void |
873fbd0b | 5816 | free_regexps (void) |
93b7ac65 | 5817 | { |
61a1f6fa | 5818 | regexp *rp; |
93b7ac65 FP |
5819 | while (p_head != NULL) |
5820 | { | |
61a1f6fa FP |
5821 | rp = p_head->p_next; |
5822 | free (p_head->pattern); | |
5823 | free (p_head->name); | |
93b7ac65 | 5824 | free (p_head); |
61a1f6fa | 5825 | p_head = rp; |
93b7ac65 FP |
5826 | } |
5827 | return; | |
5828 | } | |
24dbe96a FP |
5829 | |
5830 | /* | |
5831 | * Reads the whole file as a single string from `filebuf' and looks for | |
5832 | * multi-line regular expressions, creating tags on matches. | |
5833 | * readline already dealt with normal regexps. | |
5834 | * | |
5835 | * Idea by Ben Wing <ben@666.com> (2002). | |
5836 | */ | |
5837 | static void | |
873fbd0b | 5838 | regex_tag_multiline (void) |
24dbe96a FP |
5839 | { |
5840 | char *buffer = filebuf.buffer; | |
61a1f6fa FP |
5841 | regexp *rp; |
5842 | char *name; | |
24dbe96a | 5843 | |
61a1f6fa | 5844 | for (rp = p_head; rp != NULL; rp = rp->p_next) |
24dbe96a FP |
5845 | { |
5846 | int match = 0; | |
5847 | ||
61a1f6fa | 5848 | if (!rp->multi_line) |
24dbe96a FP |
5849 | continue; /* skip normal regexps */ |
5850 | ||
e1dbe924 | 5851 | /* Generic initializations before parsing file from memory. */ |
24dbe96a FP |
5852 | lineno = 1; /* reset global line number */ |
5853 | charno = 0; /* reset global char number */ | |
5854 | linecharno = 0; /* reset global char number of line start */ | |
5855 | ||
5856 | /* Only use generic regexps or those for the current language. */ | |
61a1f6fa | 5857 | if (rp->lang != NULL && rp->lang != curfdp->lang) |
24dbe96a FP |
5858 | continue; |
5859 | ||
5860 | while (match >= 0 && match < filebuf.len) | |
5861 | { | |
61a1f6fa FP |
5862 | match = re_search (rp->pat, buffer, filebuf.len, charno, |
5863 | filebuf.len - match, &rp->regs); | |
24dbe96a FP |
5864 | switch (match) |
5865 | { | |
5866 | case -2: | |
5867 | /* Some error. */ | |
61a1f6fa | 5868 | if (!rp->error_signaled) |
24dbe96a FP |
5869 | { |
5870 | error ("regexp stack overflow while matching \"%s\"", | |
61a1f6fa | 5871 | rp->pattern); |
96c06863 | 5872 | rp->error_signaled = true; |
24dbe96a FP |
5873 | } |
5874 | break; | |
5875 | case -1: | |
5876 | /* No match. */ | |
5877 | break; | |
5878 | default: | |
61a1f6fa | 5879 | if (match == rp->regs.end[0]) |
24dbe96a | 5880 | { |
61a1f6fa | 5881 | if (!rp->error_signaled) |
24dbe96a FP |
5882 | { |
5883 | error ("regexp matches the empty string: \"%s\"", | |
61a1f6fa | 5884 | rp->pattern); |
96c06863 | 5885 | rp->error_signaled = true; |
24dbe96a FP |
5886 | } |
5887 | match = -3; /* exit from while loop */ | |
5888 | break; | |
5889 | } | |
5890 | ||
5891 | /* Match occurred. Construct a tag. */ | |
61a1f6fa | 5892 | while (charno < rp->regs.end[0]) |
24dbe96a FP |
5893 | if (buffer[charno++] == '\n') |
5894 | lineno++, linecharno = charno; | |
61a1f6fa | 5895 | name = rp->name; |
a127d423 FP |
5896 | if (name[0] == '\0') |
5897 | name = NULL; | |
5898 | else /* make a named tag */ | |
61a1f6fa FP |
5899 | name = substitute (buffer, rp->name, &rp->regs); |
5900 | if (rp->force_explicit_name) | |
5901 | /* Force explicit tag name, if a name is there. */ | |
96c06863 | 5902 | pfnote (name, true, buffer + linecharno, |
61a1f6fa | 5903 | charno - linecharno + 1, lineno, linecharno); |
24dbe96a | 5904 | else |
96c06863 | 5905 | make_tag (name, strlen (name), true, buffer + linecharno, |
24dbe96a | 5906 | charno - linecharno + 1, lineno, linecharno); |
24dbe96a FP |
5907 | break; |
5908 | } | |
5909 | } | |
5910 | } | |
5911 | } | |
5912 | ||
c6d46f5f | 5913 | \f |
eaaacacd | 5914 | static bool |
988e88ab | 5915 | nocase_tail (const char *cp) |
eaaacacd FP |
5916 | { |
5917 | register int len = 0; | |
5918 | ||
5919 | while (*cp != '\0' && lowcase (*cp) == lowcase (dbp[len])) | |
5920 | cp++, len++; | |
5921 | if (*cp == '\0' && !intoken (dbp[len])) | |
5922 | { | |
5923 | dbp += len; | |
96c06863 | 5924 | return true; |
eaaacacd | 5925 | } |
96c06863 | 5926 | return false; |
eaaacacd FP |
5927 | } |
5928 | ||
081b7c70 | 5929 | static void |
873fbd0b | 5930 | get_tag (register char *bp, char **namepp) |
93b7ac65 | 5931 | { |
081b7c70 | 5932 | register char *cp = bp; |
93b7ac65 | 5933 | |
081b7c70 FP |
5934 | if (*bp != '\0') |
5935 | { | |
5936 | /* Go till you get to white space or a syntactic break */ | |
5937 | for (cp = bp + 1; !notinname (*cp); cp++) | |
5938 | continue; | |
96c06863 | 5939 | make_tag (bp, cp - bp, true, |
081b7c70 FP |
5940 | lb.buffer, cp - lb.buffer + 1, lineno, linecharno); |
5941 | } | |
5942 | ||
5943 | if (namepp != NULL) | |
5944 | *namepp = savenstr (bp, cp - bp); | |
93b7ac65 FP |
5945 | } |
5946 | ||
c6d46f5f | 5947 | /* |
93b7ac65 FP |
5948 | * Read a line of text from `stream' into `lbp', excluding the |
5949 | * newline or CR-NL, if any. Return the number of characters read from | |
5950 | * `stream', which is the length of the line including the newline. | |
5951 | * | |
24dbe96a FP |
5952 | * On DOS or Windows we do not count the CR character, if any before the |
5953 | * NL, in the returned length; this mirrors the behavior of Emacs on those | |
93b7ac65 FP |
5954 | * platforms (for text files, it translates CR-NL to NL as it reads in the |
5955 | * file). | |
24dbe96a FP |
5956 | * |
5957 | * If multi-line regular expressions are requested, each line read is | |
5958 | * appended to `filebuf'. | |
c6d46f5f | 5959 | */ |
944bdd72 | 5960 | static long |
873fbd0b | 5961 | readline_internal (linebuffer *lbp, register FILE *stream) |
c6d46f5f | 5962 | { |
93b7ac65 FP |
5963 | char *buffer = lbp->buffer; |
5964 | register char *p = lbp->buffer; | |
c6d46f5f | 5965 | register char *pend; |
aab1fdae | 5966 | int chars_deleted; |
c6d46f5f | 5967 | |
93b7ac65 | 5968 | pend = p + lbp->size; /* Separate to avoid 386/IX compiler bug. */ |
c6d46f5f | 5969 | |
97b90b0a | 5970 | for (;;) |
c6d46f5f JB |
5971 | { |
5972 | register int c = getc (stream); | |
5973 | if (p == pend) | |
5974 | { | |
93b7ac65 FP |
5975 | /* We're at the end of linebuffer: expand it. */ |
5976 | lbp->size *= 2; | |
b2521b0a | 5977 | xrnew (buffer, lbp->size, char); |
93b7ac65 FP |
5978 | p += buffer - lbp->buffer; |
5979 | pend = buffer + lbp->size; | |
5980 | lbp->buffer = buffer; | |
c6d46f5f | 5981 | } |
aab1fdae | 5982 | if (c == EOF) |
c6d46f5f | 5983 | { |
8dc7496c | 5984 | *p = '\0'; |
aab1fdae FP |
5985 | chars_deleted = 0; |
5986 | break; | |
5987 | } | |
5988 | if (c == '\n') | |
5989 | { | |
a8d9bd4b | 5990 | if (p > buffer && p[-1] == '\r') |
aab1fdae | 5991 | { |
e0903a92 | 5992 | p -= 1; |
70611080 | 5993 | #ifdef DOS_NT |
b6b48846 FP |
5994 | /* Assume CRLF->LF translation will be performed by Emacs |
5995 | when loading this file, so CRs won't appear in the buffer. | |
5996 | It would be cleaner to compensate within Emacs; | |
5997 | however, Emacs does not know how many CRs were deleted | |
5998 | before any given point in the file. */ | |
70611080 RS |
5999 | chars_deleted = 1; |
6000 | #else | |
aab1fdae | 6001 | chars_deleted = 2; |
70611080 | 6002 | #endif |
aab1fdae FP |
6003 | } |
6004 | else | |
6005 | { | |
aab1fdae FP |
6006 | chars_deleted = 1; |
6007 | } | |
e0903a92 | 6008 | *p = '\0'; |
c6d46f5f JB |
6009 | break; |
6010 | } | |
6011 | *p++ = c; | |
6012 | } | |
93b7ac65 | 6013 | lbp->len = p - buffer; |
c6d46f5f | 6014 | |
24dbe96a FP |
6015 | if (need_filebuf /* we need filebuf for multi-line regexps */ |
6016 | && chars_deleted > 0) /* not at EOF */ | |
6017 | { | |
6018 | while (filebuf.size <= filebuf.len + lbp->len + 1) /* +1 for \n */ | |
6019 | { | |
6020 | /* Expand filebuf. */ | |
6021 | filebuf.size *= 2; | |
6022 | xrnew (filebuf.buffer, filebuf.size, char); | |
6023 | } | |
e99a530f | 6024 | memcpy (filebuf.buffer + filebuf.len, lbp->buffer, lbp->len); |
24dbe96a FP |
6025 | filebuf.len += lbp->len; |
6026 | filebuf.buffer[filebuf.len++] = '\n'; | |
6027 | filebuf.buffer[filebuf.len] = '\0'; | |
6028 | } | |
6029 | ||
93b7ac65 | 6030 | return lbp->len + chars_deleted; |
c6d46f5f | 6031 | } |
b9755a12 FP |
6032 | |
6033 | /* | |
e0903a92 | 6034 | * Like readline_internal, above, but in addition try to match the |
24dbe96a FP |
6035 | * input line against relevant regular expressions and manage #line |
6036 | * directives. | |
b9755a12 | 6037 | */ |
e7d3b099 | 6038 | static void |
873fbd0b | 6039 | readline (linebuffer *lbp, FILE *stream) |
b9755a12 | 6040 | { |
e7d3b099 FP |
6041 | long result; |
6042 | ||
6043 | linecharno = charno; /* update global char number of line start */ | |
6044 | result = readline_internal (lbp, stream); /* read line */ | |
6045 | lineno += 1; /* increment global line number */ | |
6046 | charno += result; /* increment global char number */ | |
db590582 | 6047 | |
f6b1b0a8 | 6048 | /* Honor #line directives. */ |
9e0a3f98 | 6049 | if (!no_line_directive) |
db590582 | 6050 | { |
9e0a3f98 | 6051 | static bool discard_until_line_directive; |
db590582 | 6052 | |
9e0a3f98 FP |
6053 | /* Check whether this is a #line directive. */ |
6054 | if (result > 12 && strneq (lbp->buffer, "#line ", 6)) | |
db590582 | 6055 | { |
21cb180b FP |
6056 | unsigned int lno; |
6057 | int start = 0; | |
db590582 | 6058 | |
21cb180b FP |
6059 | if (sscanf (lbp->buffer, "#line %u \"%n", &lno, &start) >= 1 |
6060 | && start > 0) /* double quote character found */ | |
db590582 | 6061 | { |
21cb180b | 6062 | char *endp = lbp->buffer + start; |
9e0a3f98 | 6063 | |
9e0a3f98 FP |
6064 | while ((endp = etags_strchr (endp, '"')) != NULL |
6065 | && endp[-1] == '\\') | |
6066 | endp++; | |
6067 | if (endp != NULL) | |
6068 | /* Ok, this is a real #line directive. Let's deal with it. */ | |
49adb67a | 6069 | { |
9e0a3f98 FP |
6070 | char *taggedabsname; /* absolute name of original file */ |
6071 | char *taggedfname; /* name of original file as given */ | |
6072 | char *name; /* temp var */ | |
6073 | ||
96c06863 | 6074 | discard_until_line_directive = false; /* found it */ |
9e0a3f98 FP |
6075 | name = lbp->buffer + start; |
6076 | *endp = '\0'; | |
9817cf3f | 6077 | canonicalize_filename (name); |
6598a3d4 | 6078 | taggedabsname = absolute_filename (name, tagfiledir); |
9e0a3f98 FP |
6079 | if (filename_is_absolute (name) |
6080 | || filename_is_absolute (curfdp->infname)) | |
6081 | taggedfname = savestr (taggedabsname); | |
6082 | else | |
6083 | taggedfname = relative_filename (taggedabsname,tagfiledir); | |
db590582 | 6084 | |
9e0a3f98 FP |
6085 | if (streq (curfdp->taggedfname, taggedfname)) |
6086 | /* The #line directive is only a line number change. We | |
6087 | deal with this afterwards. */ | |
6088 | free (taggedfname); | |
6089 | else | |
6090 | /* The tags following this #line directive should be | |
6091 | attributed to taggedfname. In order to do this, set | |
6092 | curfdp accordingly. */ | |
6093 | { | |
6094 | fdesc *fdp; /* file description pointer */ | |
6095 | ||
6096 | /* Go look for a file description already set up for the | |
6097 | file indicated in the #line directive. If there is | |
6098 | one, use it from now until the next #line | |
6099 | directive. */ | |
6100 | for (fdp = fdhead; fdp != NULL; fdp = fdp->next) | |
6101 | if (streq (fdp->infname, curfdp->infname) | |
6102 | && streq (fdp->taggedfname, taggedfname)) | |
6103 | /* If we remove the second test above (after the &&) | |
6104 | then all entries pertaining to the same file are | |
6105 | coalesced in the tags file. If we use it, then | |
6106 | entries pertaining to the same file but generated | |
6107 | from different files (via #line directives) will | |
6108 | go into separate sections in the tags file. These | |
6109 | alternatives look equivalent. The first one | |
6110 | destroys some apparently useless information. */ | |
6111 | { | |
6112 | curfdp = fdp; | |
6113 | free (taggedfname); | |
6114 | break; | |
6115 | } | |
6116 | /* Else, if we already tagged the real file, skip all | |
6117 | input lines until the next #line directive. */ | |
6118 | if (fdp == NULL) /* not found */ | |
6119 | for (fdp = fdhead; fdp != NULL; fdp = fdp->next) | |
6120 | if (streq (fdp->infabsname, taggedabsname)) | |
6121 | { | |
96c06863 | 6122 | discard_until_line_directive = true; |
9e0a3f98 FP |
6123 | free (taggedfname); |
6124 | break; | |
6125 | } | |
6126 | /* Else create a new file description and use that from | |
6127 | now on, until the next #line directive. */ | |
6128 | if (fdp == NULL) /* not found */ | |
6129 | { | |
6130 | fdp = fdhead; | |
6131 | fdhead = xnew (1, fdesc); | |
6132 | *fdhead = *curfdp; /* copy curr. file description */ | |
6133 | fdhead->next = fdp; | |
6134 | fdhead->infname = savestr (curfdp->infname); | |
6135 | fdhead->infabsname = savestr (curfdp->infabsname); | |
6136 | fdhead->infabsdir = savestr (curfdp->infabsdir); | |
6137 | fdhead->taggedfname = taggedfname; | |
96c06863 | 6138 | fdhead->usecharno = false; |
61a1f6fa | 6139 | fdhead->prop = NULL; |
96c06863 | 6140 | fdhead->written = false; |
9e0a3f98 FP |
6141 | curfdp = fdhead; |
6142 | } | |
6143 | } | |
6144 | free (taggedabsname); | |
e7d3b099 FP |
6145 | lineno = lno - 1; |
6146 | readline (lbp, stream); | |
6147 | return; | |
9e0a3f98 | 6148 | } /* if a real #line directive */ |
045b9da7 | 6149 | } /* if #line is followed by a number */ |
9e0a3f98 FP |
6150 | } /* if line begins with "#line " */ |
6151 | ||
6152 | /* If we are here, no #line directive was found. */ | |
6153 | if (discard_until_line_directive) | |
6154 | { | |
6155 | if (result > 0) | |
e7d3b099 | 6156 | { |
24dbe96a FP |
6157 | /* Do a tail recursion on ourselves, thus discarding the contents |
6158 | of the line buffer. */ | |
e7d3b099 FP |
6159 | readline (lbp, stream); |
6160 | return; | |
6161 | } | |
9e0a3f98 | 6162 | /* End of file. */ |
96c06863 | 6163 | discard_until_line_directive = false; |
e7d3b099 | 6164 | return; |
db590582 | 6165 | } |
9e0a3f98 | 6166 | } /* if #line directives should be considered */ |
49adb67a | 6167 | |
db590582 FP |
6168 | { |
6169 | int match; | |
61a1f6fa FP |
6170 | regexp *rp; |
6171 | char *name; | |
d8913c1c | 6172 | |
61a1f6fa | 6173 | /* Match against relevant regexps. */ |
db590582 | 6174 | if (lbp->len > 0) |
61a1f6fa | 6175 | for (rp = p_head; rp != NULL; rp = rp->p_next) |
db590582 | 6176 | { |
24dbe96a FP |
6177 | /* Only use generic regexps or those for the current language. |
6178 | Also do not use multiline regexps, which is the job of | |
6179 | regex_tag_multiline. */ | |
61a1f6fa FP |
6180 | if ((rp->lang != NULL && rp->lang != fdhead->lang) |
6181 | || rp->multi_line) | |
db590582 | 6182 | continue; |
93b7ac65 | 6183 | |
61a1f6fa | 6184 | match = re_match (rp->pat, lbp->buffer, lbp->len, 0, &rp->regs); |
db590582 FP |
6185 | switch (match) |
6186 | { | |
6187 | case -2: | |
6188 | /* Some error. */ | |
61a1f6fa | 6189 | if (!rp->error_signaled) |
db590582 | 6190 | { |
24dbe96a | 6191 | error ("regexp stack overflow while matching \"%s\"", |
61a1f6fa | 6192 | rp->pattern); |
96c06863 | 6193 | rp->error_signaled = true; |
db590582 FP |
6194 | } |
6195 | break; | |
6196 | case -1: | |
6197 | /* No match. */ | |
6198 | break; | |
24dbe96a FP |
6199 | case 0: |
6200 | /* Empty string matched. */ | |
61a1f6fa | 6201 | if (!rp->error_signaled) |
24dbe96a | 6202 | { |
61a1f6fa | 6203 | error ("regexp matches the empty string: \"%s\"", rp->pattern); |
96c06863 | 6204 | rp->error_signaled = true; |
24dbe96a FP |
6205 | } |
6206 | break; | |
db590582 FP |
6207 | default: |
6208 | /* Match occurred. Construct a tag. */ | |
61a1f6fa | 6209 | name = rp->name; |
a127d423 FP |
6210 | if (name[0] == '\0') |
6211 | name = NULL; | |
6212 | else /* make a named tag */ | |
61a1f6fa FP |
6213 | name = substitute (lbp->buffer, rp->name, &rp->regs); |
6214 | if (rp->force_explicit_name) | |
6215 | /* Force explicit tag name, if a name is there. */ | |
96c06863 | 6216 | pfnote (name, true, lbp->buffer, match, lineno, linecharno); |
db590582 | 6217 | else |
96c06863 | 6218 | make_tag (name, strlen (name), true, |
db590582 | 6219 | lbp->buffer, match, lineno, linecharno); |
db590582 FP |
6220 | break; |
6221 | } | |
6222 | } | |
6223 | } | |
b9755a12 | 6224 | } |
b2521b0a | 6225 | |
c6d46f5f | 6226 | \f |
55597f90 FP |
6227 | /* |
6228 | * Return a pointer to a space of size strlen(cp)+1 allocated | |
6229 | * with xnew where the string CP has been copied. | |
6230 | */ | |
944bdd72 | 6231 | static char * |
988e88ab | 6232 | savestr (const char *cp) |
c6d46f5f JB |
6233 | { |
6234 | return savenstr (cp, strlen (cp)); | |
6235 | } | |
6236 | ||
55597f90 FP |
6237 | /* |
6238 | * Return a pointer to a space of size LEN+1 allocated with xnew where | |
6239 | * the string CP has been copied for at most the first LEN characters. | |
6240 | */ | |
944bdd72 | 6241 | static char * |
988e88ab | 6242 | savenstr (const char *cp, int len) |
c6d46f5f | 6243 | { |
d672ac3c | 6244 | char *dp = xnew (len + 1, char); |
c6d46f5f | 6245 | dp[len] = '\0'; |
d672ac3c | 6246 | return memcpy (dp, cp, len); |
c6d46f5f JB |
6247 | } |
6248 | ||
c6d46f5f JB |
6249 | /* |
6250 | * Return the ptr in sp at which the character c last | |
6251 | * appears; NULL if not found | |
6252 | * | |
944bdd72 | 6253 | * Identical to POSIX strrchr, included for portability. |
c6d46f5f | 6254 | */ |
944bdd72 | 6255 | static char * |
873fbd0b | 6256 | etags_strrchr (register const char *sp, register int c) |
c6d46f5f | 6257 | { |
944bdd72 | 6258 | register const char *r; |
c6d46f5f JB |
6259 | |
6260 | r = NULL; | |
6261 | do | |
6262 | { | |
6263 | if (*sp == c) | |
6264 | r = sp; | |
6265 | } while (*sp++); | |
944bdd72 | 6266 | return (char *)r; |
c6d46f5f JB |
6267 | } |
6268 | ||
6269 | /* | |
6270 | * Return the ptr in sp at which the character c first | |
6271 | * appears; NULL if not found | |
6272 | * | |
944bdd72 | 6273 | * Identical to POSIX strchr, included for portability. |
c6d46f5f | 6274 | */ |
944bdd72 | 6275 | static char * |
873fbd0b | 6276 | etags_strchr (register const char *sp, register int c) |
c6d46f5f JB |
6277 | { |
6278 | do | |
6279 | { | |
6280 | if (*sp == c) | |
944bdd72 | 6281 | return (char *)sp; |
b9755a12 FP |
6282 | } while (*sp++); |
6283 | return NULL; | |
c6d46f5f JB |
6284 | } |
6285 | ||
55102b5d | 6286 | /* Skip spaces (end of string is not space), return new pointer. */ |
944bdd72 | 6287 | static char * |
873fbd0b | 6288 | skip_spaces (char *cp) |
93b7ac65 | 6289 | { |
71cbb895 | 6290 | while (iswhite (*cp)) |
93b7ac65 FP |
6291 | cp++; |
6292 | return cp; | |
6293 | } | |
6294 | ||
55102b5d | 6295 | /* Skip non spaces, except end of string, return new pointer. */ |
944bdd72 | 6296 | static char * |
873fbd0b | 6297 | skip_non_spaces (char *cp) |
93b7ac65 | 6298 | { |
71cbb895 | 6299 | while (*cp != '\0' && !iswhite (*cp)) |
93b7ac65 FP |
6300 | cp++; |
6301 | return cp; | |
6302 | } | |
6303 | ||
1cbaa705 KR |
6304 | /* Skip any chars in the "name" class.*/ |
6305 | static char * | |
6306 | skip_name (char *cp) | |
6307 | { | |
6308 | /* '\0' is a notinname() so loop stops there too */ | |
6309 | while (! notinname (*cp)) | |
6310 | cp++; | |
6311 | return cp; | |
6312 | } | |
6313 | ||
c6d46f5f | 6314 | /* Print error message and exit. */ |
b2521b0a | 6315 | void |
988e88ab | 6316 | fatal (const char *s1, const char *s2) |
c6d46f5f JB |
6317 | { |
6318 | error (s1, s2); | |
3f0656ff | 6319 | exit (EXIT_FAILURE); |
c6d46f5f JB |
6320 | } |
6321 | ||
944bdd72 | 6322 | static void |
988e88ab | 6323 | pfatal (const char *s1) |
cdc1f6a7 FP |
6324 | { |
6325 | perror (s1); | |
3f0656ff | 6326 | exit (EXIT_FAILURE); |
cdc1f6a7 FP |
6327 | } |
6328 | ||
944bdd72 | 6329 | static void |
873fbd0b | 6330 | suggest_asking_for_help (void) |
d8913c1c | 6331 | { |
9239d970 PE |
6332 | fprintf (stderr, "\tTry `%s --help' for a complete list of options.\n", |
6333 | progname); | |
3f0656ff | 6334 | exit (EXIT_FAILURE); |
d8913c1c FP |
6335 | } |
6336 | ||
db5a3003 | 6337 | /* Output a diagnostic with printf-style FORMAT and args. */ |
944bdd72 | 6338 | static void |
db5a3003 | 6339 | error (const char *format, ...) |
c6d46f5f | 6340 | { |
db5a3003 PE |
6341 | va_list ap; |
6342 | va_start (ap, format); | |
c6d46f5f | 6343 | fprintf (stderr, "%s: ", progname); |
db5a3003 | 6344 | vfprintf (stderr, format, ap); |
c6d46f5f | 6345 | fprintf (stderr, "\n"); |
db5a3003 | 6346 | va_end (ap); |
c6d46f5f JB |
6347 | } |
6348 | ||
46c145db FP |
6349 | /* Return a newly-allocated string whose contents |
6350 | concatenate those of s1, s2, s3. */ | |
944bdd72 | 6351 | static char * |
988e88ab | 6352 | concat (const char *s1, const char *s2, const char *s3) |
c6d46f5f JB |
6353 | { |
6354 | int len1 = strlen (s1), len2 = strlen (s2), len3 = strlen (s3); | |
6355 | char *result = xnew (len1 + len2 + len3 + 1, char); | |
6356 | ||
1a0d8c80 FP |
6357 | strcpy (result, s1); |
6358 | strcpy (result + len1, s2); | |
6359 | strcpy (result + len1 + len2, s3); | |
c6d46f5f JB |
6360 | |
6361 | return result; | |
6362 | } | |
b2521b0a | 6363 | |
b02c5fea | 6364 | \f |
cdc1f6a7 | 6365 | /* Does the same work as the system V getcwd, but does not need to |
c5007f46 | 6366 | guess the buffer size in advance. */ |
944bdd72 | 6367 | static char * |
873fbd0b | 6368 | etags_getcwd (void) |
915e30c8 | 6369 | { |
cdc1f6a7 FP |
6370 | int bufsize = 200; |
6371 | char *path = xnew (bufsize, char); | |
c6d46f5f | 6372 | |
cdc1f6a7 | 6373 | while (getcwd (path, bufsize) == NULL) |
b02c5fea | 6374 | { |
dcc89e63 | 6375 | if (errno != ERANGE) |
0f394065 | 6376 | pfatal ("getcwd"); |
5e9c8296 | 6377 | bufsize *= 2; |
cec68fcb | 6378 | free (path); |
cdc1f6a7 | 6379 | path = xnew (bufsize, char); |
5e9c8296 | 6380 | } |
b02c5fea | 6381 | |
93b7ac65 | 6382 | canonicalize_filename (path); |
cdc1f6a7 | 6383 | return path; |
b02c5fea FP |
6384 | } |
6385 | ||
93b7ac65 FP |
6386 | /* Return a newly allocated string containing the file name of FILE |
6387 | relative to the absolute directory DIR (which should end with a slash). */ | |
944bdd72 | 6388 | static char * |
873fbd0b | 6389 | relative_filename (char *file, char *dir) |
46c145db | 6390 | { |
93b7ac65 | 6391 | char *fp, *dp, *afn, *res; |
2f608d34 | 6392 | int i; |
46c145db | 6393 | |
3ca80e28 | 6394 | /* Find the common root of file and dir (with a trailing slash). */ |
93b7ac65 FP |
6395 | afn = absolute_filename (file, cwd); |
6396 | fp = afn; | |
46c145db FP |
6397 | dp = dir; |
6398 | while (*fp++ == *dp++) | |
6399 | continue; | |
3f1c8fcd | 6400 | fp--, dp--; /* back to the first differing char */ |
93b7ac65 FP |
6401 | #ifdef DOS_NT |
6402 | if (fp == afn && afn[0] != '/') /* cannot build a relative name */ | |
6403 | return afn; | |
6404 | #endif | |
6405 | do /* look at the equal chars until '/' */ | |
3ca80e28 | 6406 | fp--, dp--; |
93b7ac65 | 6407 | while (*fp != '/'); |
46c145db | 6408 | |
2f608d34 FP |
6409 | /* Build a sequence of "../" strings for the resulting relative file name. */ |
6410 | i = 0; | |
6411 | while ((dp = etags_strchr (dp + 1, '/')) != NULL) | |
9817cf3f | 6412 | i += 1; |
2f608d34 FP |
6413 | res = xnew (3*i + strlen (fp + 1) + 1, char); |
6414 | res[0] = '\0'; | |
6415 | while (i-- > 0) | |
6416 | strcat (res, "../"); | |
6417 | ||
6418 | /* Add the file name relative to the common root of file and dir. */ | |
6419 | strcat (res, fp + 1); | |
93b7ac65 | 6420 | free (afn); |
46c145db | 6421 | |
108c932a | 6422 | return res; |
46c145db FP |
6423 | } |
6424 | ||
93b7ac65 FP |
6425 | /* Return a newly allocated string containing the absolute file name |
6426 | of FILE given DIR (which should end with a slash). */ | |
944bdd72 | 6427 | static char * |
873fbd0b | 6428 | absolute_filename (char *file, char *dir) |
46c145db FP |
6429 | { |
6430 | char *slashp, *cp, *res; | |
6431 | ||
93b7ac65 | 6432 | if (filename_is_absolute (file)) |
2f608d34 | 6433 | res = savestr (file); |
ce7f6d62 | 6434 | #ifdef DOS_NT |
2f608d34 | 6435 | /* We don't support non-absolute file names with a drive |
ce7f6d62 RS |
6436 | letter, like `d:NAME' (it's too much hassle). */ |
6437 | else if (file[1] == ':') | |
2f608d34 | 6438 | fatal ("%s: relative file names with drive letters not supported", file); |
ce7f6d62 | 6439 | #endif |
46c145db | 6440 | else |
93b7ac65 | 6441 | res = concat (dir, file, ""); |
46c145db FP |
6442 | |
6443 | /* Delete the "/dirname/.." and "/." substrings. */ | |
b02c5fea | 6444 | slashp = etags_strchr (res, '/'); |
46c145db FP |
6445 | while (slashp != NULL && slashp[0] != '\0') |
6446 | { | |
6447 | if (slashp[1] == '.') | |
6448 | { | |
6449 | if (slashp[2] == '.' | |
6450 | && (slashp[3] == '/' || slashp[3] == '\0')) | |
6451 | { | |
6452 | cp = slashp; | |
6453 | do | |
6454 | cp--; | |
93b7ac65 | 6455 | while (cp >= res && !filename_is_absolute (cp)); |
2f608d34 FP |
6456 | if (cp < res) |
6457 | cp = slashp; /* the absolute name begins with "/.." */ | |
ce7f6d62 RS |
6458 | #ifdef DOS_NT |
6459 | /* Under MSDOS and NT we get `d:/NAME' as absolute | |
2f608d34 | 6460 | file name, so the luser could say `d:/../NAME'. |
ce7f6d62 | 6461 | We silently treat this as `d:/NAME'. */ |
2f608d34 FP |
6462 | else if (cp[0] != '/') |
6463 | cp = slashp; | |
ce7f6d62 | 6464 | #endif |
46f3381a | 6465 | memmove (cp, slashp + 3, strlen (slashp + 2)); |
46c145db | 6466 | slashp = cp; |
e9b2b94c | 6467 | continue; |
46c145db FP |
6468 | } |
6469 | else if (slashp[2] == '/' || slashp[2] == '\0') | |
6470 | { | |
46f3381a | 6471 | memmove (slashp, slashp + 2, strlen (slashp + 1)); |
e9b2b94c | 6472 | continue; |
46c145db FP |
6473 | } |
6474 | } | |
e9b2b94c FP |
6475 | |
6476 | slashp = etags_strchr (slashp + 1, '/'); | |
46c145db | 6477 | } |
7ebca6a6 | 6478 | |
327891eb FP |
6479 | if (res[0] == '\0') /* just a safety net: should never happen */ |
6480 | { | |
6481 | free (res); | |
6482 | return savestr ("/"); | |
6483 | } | |
2f608d34 FP |
6484 | else |
6485 | return res; | |
46c145db FP |
6486 | } |
6487 | ||
b02c5fea | 6488 | /* Return a newly allocated string containing the absolute |
93b7ac65 | 6489 | file name of dir where FILE resides given DIR (which should |
b02c5fea | 6490 | end with a slash). */ |
944bdd72 | 6491 | static char * |
873fbd0b | 6492 | absolute_dirname (char *file, char *dir) |
46c145db FP |
6493 | { |
6494 | char *slashp, *res; | |
6495 | char save; | |
6496 | ||
b02c5fea | 6497 | slashp = etags_strrchr (file, '/'); |
46c145db | 6498 | if (slashp == NULL) |
93b7ac65 | 6499 | return savestr (dir); |
46c145db FP |
6500 | save = slashp[1]; |
6501 | slashp[1] = '\0'; | |
93b7ac65 | 6502 | res = absolute_filename (file, dir); |
46c145db FP |
6503 | slashp[1] = save; |
6504 | ||
6505 | return res; | |
6506 | } | |
6507 | ||
93b7ac65 FP |
6508 | /* Whether the argument string is an absolute file name. The argument |
6509 | string must have been canonicalized with canonicalize_filename. */ | |
944bdd72 | 6510 | static bool |
873fbd0b | 6511 | filename_is_absolute (char *fn) |
93b7ac65 FP |
6512 | { |
6513 | return (fn[0] == '/' | |
6514 | #ifdef DOS_NT | |
5e617bc2 | 6515 | || (ISALPHA (fn[0]) && fn[1] == ':' && fn[2] == '/') |
93b7ac65 FP |
6516 | #endif |
6517 | ); | |
6518 | } | |
6519 | ||
e5075711 | 6520 | /* Downcase DOS drive letter and collapse separators into single slashes. |
9817cf3f | 6521 | Works in place. */ |
944bdd72 | 6522 | static void |
873fbd0b | 6523 | canonicalize_filename (register char *fn) |
93b7ac65 | 6524 | { |
9817cf3f FP |
6525 | register char* cp; |
6526 | char sep = '/'; | |
6527 | ||
93b7ac65 | 6528 | #ifdef DOS_NT |
c8b04710 | 6529 | /* Canonicalize drive letter case. */ |
5e617bc2 | 6530 | # define ISUPPER(c) isupper (CHAR (c)) |
e5075711 | 6531 | if (fn[0] != '\0' && fn[1] == ':' && ISUPPER (fn[0])) |
32e793fa | 6532 | fn[0] = lowcase (fn[0]); |
9817cf3f FP |
6533 | |
6534 | sep = '\\'; | |
93b7ac65 | 6535 | #endif |
9817cf3f FP |
6536 | |
6537 | /* Collapse multiple separators into a single slash. */ | |
6538 | for (cp = fn; *cp != '\0'; cp++, fn++) | |
6539 | if (*cp == sep) | |
6540 | { | |
6541 | *fn = '/'; | |
6542 | while (cp[1] == sep) | |
6543 | cp++; | |
6544 | } | |
6545 | else | |
6546 | *fn = *cp; | |
6547 | *fn = '\0'; | |
93b7ac65 FP |
6548 | } |
6549 | ||
61a1f6fa | 6550 | \f |
9817cf3f | 6551 | /* Initialize a linebuffer for use. */ |
61a1f6fa | 6552 | static void |
873fbd0b | 6553 | linebuffer_init (linebuffer *lbp) |
61a1f6fa FP |
6554 | { |
6555 | lbp->size = (DEBUG) ? 3 : 200; | |
6556 | lbp->buffer = xnew (lbp->size, char); | |
6557 | lbp->buffer[0] = '\0'; | |
6558 | lbp->len = 0; | |
6559 | } | |
6560 | ||
b2521b0a | 6561 | /* Set the minimum size of a string contained in a linebuffer. */ |
944bdd72 | 6562 | static void |
873fbd0b | 6563 | linebuffer_setlen (linebuffer *lbp, int toksize) |
3f1c8fcd | 6564 | { |
b2521b0a FP |
6565 | while (lbp->size <= toksize) |
6566 | { | |
6567 | lbp->size *= 2; | |
6568 | xrnew (lbp->buffer, lbp->size, char); | |
6569 | } | |
6570 | lbp->len = toksize; | |
3f1c8fcd FP |
6571 | } |
6572 | ||
61a1f6fa | 6573 | /* Like malloc but get fatal error if memory is exhausted. */ |
261cb4bb | 6574 | static void * |
9250f758 | 6575 | xmalloc (size_t size) |
c6d46f5f | 6576 | { |
261cb4bb | 6577 | void *result = malloc (size); |
1a0d8c80 | 6578 | if (result == NULL) |
086eac13 | 6579 | fatal ("virtual memory exhausted", (char *)NULL); |
c6d46f5f JB |
6580 | return result; |
6581 | } | |
6582 | ||
261cb4bb | 6583 | static void * |
9250f758 | 6584 | xrealloc (char *ptr, size_t size) |
c6d46f5f | 6585 | { |
261cb4bb | 6586 | void *result = realloc (ptr, size); |
1a0d8c80 | 6587 | if (result == NULL) |
086eac13 | 6588 | fatal ("virtual memory exhausted", (char *)NULL); |
c6d46f5f JB |
6589 | return result; |
6590 | } | |
62aec606 FP |
6591 | |
6592 | /* | |
6593 | * Local Variables: | |
62aec606 FP |
6594 | * indent-tabs-mode: t |
6595 | * tab-width: 8 | |
9e0a3f98 | 6596 | * fill-column: 79 |
61a1f6fa | 6597 | * c-font-lock-extra-types: ("FILE" "bool" "language" "linebuffer" "fdesc" "node" "regexp") |
a423dfde | 6598 | * c-file-style: "gnu" |
62aec606 FP |
6599 | * End: |
6600 | */ | |
ab5796a9 | 6601 | |
3f0656ff | 6602 | /* etags.c ends here */ |