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 | ||
ab422c4d PE |
31 | Copyright (C) 1984, 1987-1989, 1993-1995, 1998-2013 Free Software |
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 FP |
82 | |
83 | #define TRUE 1 | |
84 | #define FALSE 0 | |
d8913c1c | 85 | |
b2521b0a FP |
86 | #ifdef DEBUG |
87 | # undef DEBUG | |
88 | # define DEBUG TRUE | |
89 | #else | |
90 | # define DEBUG FALSE | |
91 | # define NDEBUG /* disable assert */ | |
75bdbc6a | 92 | #endif |
46c145db | 93 | |
cf38a720 | 94 | #include <config.h> |
4ee9629e | 95 | |
944bdd72 FP |
96 | #ifndef _GNU_SOURCE |
97 | # define _GNU_SOURCE 1 /* enables some compiler checks on GNU */ | |
98 | #endif | |
99 | ||
35a4c758 | 100 | /* WIN32_NATIVE is for XEmacs. |
b28e26be FP |
101 | MSDOS, WINDOWSNT, DOS_NT are for Emacs. */ |
102 | #ifdef WIN32_NATIVE | |
103 | # undef MSDOS | |
104 | # undef WINDOWSNT | |
105 | # define WINDOWSNT | |
106 | #endif /* WIN32_NATIVE */ | |
107 | ||
c6880c90 | 108 | #ifdef MSDOS |
71cbb895 | 109 | # undef MSDOS |
944bdd72 | 110 | # define MSDOS TRUE |
8dc7496c FP |
111 | # include <fcntl.h> |
112 | # include <sys/param.h> | |
93b7ac65 | 113 | # include <io.h> |
944bdd72 FP |
114 | #else |
115 | # define MSDOS FALSE | |
c6880c90 RS |
116 | #endif /* MSDOS */ |
117 | ||
c05b6df5 | 118 | #ifdef WINDOWSNT |
8dc7496c | 119 | # include <fcntl.h> |
944bdd72 | 120 | # include <direct.h> |
a65c6351 | 121 | # include <io.h> |
8dc7496c | 122 | # define MAXPATHLEN _MAX_PATH |
b28e26be FP |
123 | # undef HAVE_NTGUI |
124 | # undef DOS_NT | |
125 | # define DOS_NT | |
9239d970 | 126 | #endif /* WINDOWSNT */ |
cec68fcb | 127 | |
4004364e | 128 | #include <unistd.h> |
db5a3003 | 129 | #include <stdarg.h> |
0e926e56 PE |
130 | #include <stdlib.h> |
131 | #include <string.h> | |
918f9ad1 JB |
132 | #include <stdio.h> |
133 | #include <ctype.h> | |
dcc89e63 | 134 | #include <errno.h> |
918f9ad1 JB |
135 | #include <sys/types.h> |
136 | #include <sys/stat.h> | |
fee5959d | 137 | #include <c-strcase.h> |
918f9ad1 | 138 | |
8f79fe72 FP |
139 | #include <assert.h> |
140 | #ifdef NDEBUG | |
141 | # undef assert /* some systems have a buggy assert.h */ | |
142 | # define assert(x) ((void) 0) | |
914d7258 AI |
143 | #endif |
144 | ||
9239d970 | 145 | #include <getopt.h> |
35a4c758 | 146 | #include <regex.h> |
918f9ad1 | 147 | |
32daa216 | 148 | /* Define CTAGS to make the program "ctags" compatible with the usual one. |
85e77a3f | 149 | Leave it undefined to make the program "etags", which makes emacs-style |
32daa216 FP |
150 | tag tables and tags typedefs, #defines and struct/union/enum by default. */ |
151 | #ifdef CTAGS | |
152 | # undef CTAGS | |
153 | # define CTAGS TRUE | |
154 | #else | |
155 | # define CTAGS FALSE | |
c6d46f5f JB |
156 | #endif |
157 | ||
5e617bc2 | 158 | #define streq(s,t) (assert ((s)!=NULL || (t)!=NULL), !strcmp (s, t)) |
fee5959d | 159 | #define strcaseeq(s,t) (assert ((s)!=NULL && (t)!=NULL), !c_strcasecmp (s, t)) |
5e617bc2 | 160 | #define strneq(s,t,n) (assert ((s)!=NULL || (t)!=NULL), !strncmp (s, t, n)) |
fee5959d | 161 | #define strncaseeq(s,t,n) (assert ((s)!=NULL && (t)!=NULL), !c_strncasecmp (s, t, n)) |
c6d46f5f | 162 | |
30903246 | 163 | #define CHARS 256 /* 2^sizeof(char) */ |
89d57763 | 164 | #define CHAR(x) ((unsigned int)(x) & (CHARS - 1)) |
5e617bc2 JB |
165 | #define iswhite(c) (_wht[CHAR (c)]) /* c is white (see white) */ |
166 | #define notinname(c) (_nin[CHAR (c)]) /* c is not in a name (see nonam) */ | |
167 | #define begtoken(c) (_btk[CHAR (c)]) /* c can start token (see begtk) */ | |
168 | #define intoken(c) (_itk[CHAR (c)]) /* c can be in token (see midtk) */ | |
169 | #define endtoken(c) (_etk[CHAR (c)]) /* c ends tokens (see endtk) */ | |
c6d46f5f | 170 | |
5e617bc2 JB |
171 | #define ISALNUM(c) isalnum (CHAR (c)) |
172 | #define ISALPHA(c) isalpha (CHAR (c)) | |
173 | #define ISDIGIT(c) isdigit (CHAR (c)) | |
174 | #define ISLOWER(c) islower (CHAR (c)) | |
89d57763 | 175 | |
5e617bc2 | 176 | #define lowcase(c) tolower (CHAR (c)) |
af03e6ab | 177 | |
b2db879b | 178 | |
55597f90 | 179 | /* |
93b7ac65 | 180 | * xnew, xrnew -- allocate, reallocate storage |
55597f90 FP |
181 | * |
182 | * SYNOPSIS: Type *xnew (int n, Type); | |
b2521b0a | 183 | * void xrnew (OldPointer, int n, Type); |
55597f90 | 184 | */ |
b2521b0a | 185 | #if DEBUG |
2f608d34 | 186 | # include "chkmalloc.h" |
93b7ac65 FP |
187 | # define xnew(n,Type) ((Type *) trace_malloc (__FILE__, __LINE__, \ |
188 | (n) * sizeof (Type))) | |
b2521b0a FP |
189 | # define xrnew(op,n,Type) ((op) = (Type *) trace_realloc (__FILE__, __LINE__, \ |
190 | (char *) (op), (n) * sizeof (Type))) | |
2f608d34 | 191 | #else |
93b7ac65 | 192 | # define xnew(n,Type) ((Type *) xmalloc ((n) * sizeof (Type))) |
b2521b0a FP |
193 | # define xrnew(op,n,Type) ((op) = (Type *) xrealloc ( \ |
194 | (char *) (op), (n) * sizeof (Type))) | |
2f608d34 | 195 | #endif |
c6d46f5f | 196 | |
eaaacacd | 197 | #define bool int |
aab1fdae | 198 | |
f57e2426 | 199 | typedef void Lang_function (FILE *); |
c6d46f5f | 200 | |
93b7ac65 FP |
201 | typedef struct |
202 | { | |
988e88ab J |
203 | const char *suffix; /* file name suffix for this compressor */ |
204 | const char *command; /* takes one arg and decompresses to stdout */ | |
93b7ac65 | 205 | } compressor; |
c6d46f5f | 206 | |
93b7ac65 FP |
207 | typedef struct |
208 | { | |
988e88ab J |
209 | const char *name; /* language name */ |
210 | const char *help; /* detailed help for the language */ | |
9e0a3f98 | 211 | Lang_function *function; /* parse function */ |
988e88ab J |
212 | const char **suffixes; /* name suffixes of this language's files */ |
213 | const char **filenames; /* names of this language's files */ | |
214 | const char **interpreters; /* interpreters for this language */ | |
4fdb4553 | 215 | bool metasource; /* source used to generate other sources */ |
93b7ac65 FP |
216 | } language; |
217 | ||
9e0a3f98 FP |
218 | typedef struct fdesc |
219 | { | |
220 | struct fdesc *next; /* for the linked list */ | |
221 | char *infname; /* uncompressed input file name */ | |
222 | char *infabsname; /* absolute uncompressed input file name */ | |
223 | char *infabsdir; /* absolute dir of input file */ | |
224 | char *taggedfname; /* file name to write in tagfile */ | |
225 | language *lang; /* language of file */ | |
226 | char *prop; /* file properties to write in tagfile */ | |
227 | bool usecharno; /* etags tags shall contain char number */ | |
61a1f6fa | 228 | bool written; /* entry written in the tags file */ |
9e0a3f98 FP |
229 | } fdesc; |
230 | ||
944bdd72 | 231 | typedef struct node_st |
9e0a3f98 FP |
232 | { /* sorting structure */ |
233 | struct node_st *left, *right; /* left and right sons */ | |
234 | fdesc *fdp; /* description of file to whom tag belongs */ | |
988e88ab | 235 | char *name; /* tag name */ |
61a1f6fa | 236 | char *regex; /* search regexp */ |
9e0a3f98 | 237 | bool valid; /* write this tag on the tag file */ |
61a1f6fa | 238 | bool is_func; /* function tag: use regexp in CTAGS mode */ |
9e0a3f98 FP |
239 | bool been_warned; /* warning already given for duplicated tag */ |
240 | int lno; /* line number tag is on */ | |
944bdd72 | 241 | long cno; /* character number line starts on */ |
944bdd72 FP |
242 | } node; |
243 | ||
244 | /* | |
245 | * A `linebuffer' is a structure which holds a line of text. | |
246 | * `readline_internal' reads a line from a stream into a linebuffer | |
247 | * and works regardless of the length of the line. | |
248 | * SIZE is the size of BUFFER, LEN is the length of the string in | |
249 | * BUFFER after readline reads it. | |
250 | */ | |
251 | typedef struct | |
252 | { | |
253 | long size; | |
254 | int len; | |
255 | char *buffer; | |
256 | } linebuffer; | |
b9755a12 | 257 | |
9e0a3f98 FP |
258 | /* Used to support mixing of --lang and file names. */ |
259 | typedef struct | |
260 | { | |
261 | enum { | |
262 | at_language, /* a language specification */ | |
263 | at_regexp, /* a regular expression */ | |
97b90b0a | 264 | at_filename, /* a file name */ |
4fdb4553 FP |
265 | at_stdin, /* read from stdin here */ |
266 | at_end /* stop parsing the list */ | |
9e0a3f98 FP |
267 | } arg_type; /* argument type */ |
268 | language *lang; /* language associated with the argument */ | |
269 | char *what; /* the argument itself */ | |
270 | } argument; | |
271 | ||
9e0a3f98 | 272 | /* Structure defining a regular expression. */ |
61a1f6fa | 273 | typedef struct regexp |
9e0a3f98 | 274 | { |
61a1f6fa FP |
275 | struct regexp *p_next; /* pointer to next in list */ |
276 | language *lang; /* if set, use only for this language */ | |
277 | char *pattern; /* the regexp pattern */ | |
278 | char *name; /* tag name */ | |
279 | struct re_pattern_buffer *pat; /* the compiled pattern */ | |
280 | struct re_registers regs; /* re registers */ | |
281 | bool error_signaled; /* already signaled for this regexp */ | |
53964682 | 282 | bool force_explicit_name; /* do not allow implicit tag name */ |
61a1f6fa FP |
283 | bool ignore_case; /* ignore case when matching */ |
284 | bool multi_line; /* do a multi-line match on the whole file */ | |
285 | } regexp; | |
9e0a3f98 FP |
286 | |
287 | ||
30903246 | 288 | /* Many compilers barf on this: |
93b7ac65 | 289 | Lang_function Ada_funcs; |
30903246 | 290 | so let's write it this way */ |
f57e2426 J |
291 | static void Ada_funcs (FILE *); |
292 | static void Asm_labels (FILE *); | |
293 | static void C_entries (int c_ext, FILE *); | |
294 | static void default_C_entries (FILE *); | |
295 | static void plain_C_entries (FILE *); | |
296 | static void Cjava_entries (FILE *); | |
297 | static void Cobol_paragraphs (FILE *); | |
298 | static void Cplusplus_entries (FILE *); | |
299 | static void Cstar_entries (FILE *); | |
300 | static void Erlang_functions (FILE *); | |
301 | static void Forth_words (FILE *); | |
302 | static void Fortran_functions (FILE *); | |
303 | static void HTML_labels (FILE *); | |
304 | static void Lisp_functions (FILE *); | |
305 | static void Lua_functions (FILE *); | |
306 | static void Makefile_targets (FILE *); | |
307 | static void Pascal_functions (FILE *); | |
308 | static void Perl_functions (FILE *); | |
309 | static void PHP_functions (FILE *); | |
310 | static void PS_functions (FILE *); | |
311 | static void Prolog_functions (FILE *); | |
312 | static void Python_functions (FILE *); | |
313 | static void Scheme_functions (FILE *); | |
314 | static void TeX_commands (FILE *); | |
315 | static void Texinfo_nodes (FILE *); | |
316 | static void Yacc_entries (FILE *); | |
317 | static void just_read_file (FILE *); | |
318 | ||
319 | static void print_language_names (void); | |
320 | static void print_version (void); | |
321 | static void print_help (argument *); | |
322 | int main (int, char **); | |
323 | ||
324 | static compressor *get_compressor_from_suffix (char *, char **); | |
325 | static language *get_language_from_langname (const char *); | |
326 | static language *get_language_from_interpreter (char *); | |
327 | static language *get_language_from_filename (char *, bool); | |
328 | static void readline (linebuffer *, FILE *); | |
329 | static long readline_internal (linebuffer *, FILE *); | |
988e88ab | 330 | static bool nocase_tail (const char *); |
f57e2426 J |
331 | static void get_tag (char *, char **); |
332 | ||
333 | static void analyse_regex (char *); | |
334 | static void free_regexps (void); | |
335 | static void regex_tag_multiline (void); | |
db5a3003 | 336 | static void error (const char *, ...) ATTRIBUTE_FORMAT_PRINTF (1, 2); |
845ca893 PE |
337 | static _Noreturn void suggest_asking_for_help (void); |
338 | _Noreturn void fatal (const char *, const char *); | |
339 | static _Noreturn void pfatal (const char *); | |
f57e2426 J |
340 | static void add_node (node *, node **); |
341 | ||
342 | static void init (void); | |
343 | static void process_file_name (char *, language *); | |
344 | static void process_file (FILE *, char *, language *); | |
345 | static void find_entries (FILE *); | |
346 | static void free_tree (node *); | |
347 | static void free_fdesc (fdesc *); | |
348 | static void pfnote (char *, bool, char *, int, int, long); | |
988e88ab | 349 | static void make_tag (const char *, int, bool, char *, int, int, long); |
f57e2426 J |
350 | static void invalidate_nodes (fdesc *, node **); |
351 | static void put_entries (node *); | |
352 | ||
988e88ab | 353 | static char *concat (const char *, const char *, const char *); |
f57e2426 J |
354 | static char *skip_spaces (char *); |
355 | static char *skip_non_spaces (char *); | |
1cbaa705 | 356 | static char *skip_name (char *); |
988e88ab J |
357 | static char *savenstr (const char *, int); |
358 | static char *savestr (const char *); | |
f57e2426 J |
359 | static char *etags_strchr (const char *, int); |
360 | static char *etags_strrchr (const char *, int); | |
f57e2426 J |
361 | static char *etags_getcwd (void); |
362 | static char *relative_filename (char *, char *); | |
363 | static char *absolute_filename (char *, char *); | |
364 | static char *absolute_dirname (char *, char *); | |
365 | static bool filename_is_absolute (char *f); | |
366 | static void canonicalize_filename (char *); | |
367 | static void linebuffer_init (linebuffer *); | |
368 | static void linebuffer_setlen (linebuffer *, int); | |
261cb4bb PE |
369 | static void *xmalloc (size_t); |
370 | static void *xrealloc (char *, size_t); | |
93b7ac65 | 371 | |
c6d46f5f | 372 | \f |
db590582 | 373 | static char searchar = '/'; /* use /.../ searches */ |
c6d46f5f | 374 | |
db590582 FP |
375 | static char *tagfile; /* output file */ |
376 | static char *progname; /* name this program was invoked with */ | |
377 | static char *cwd; /* current working directory */ | |
378 | static char *tagfiledir; /* directory of tagfile */ | |
379 | static FILE *tagf; /* ioptr for tags file */ | |
9250f758 | 380 | static ptrdiff_t whatlen_max; /* maximum length of any 'what' member */ |
93b7ac65 | 381 | |
9e0a3f98 FP |
382 | static fdesc *fdhead; /* head of file description list */ |
383 | static fdesc *curfdp; /* current file description */ | |
db590582 FP |
384 | static int lineno; /* line number of current line */ |
385 | static long charno; /* current character number */ | |
386 | static long linecharno; /* charno of start of current line */ | |
387 | static char *dbp; /* pointer to start of current tag */ | |
9e0a3f98 | 388 | |
db590582 | 389 | static const int invalidcharno = -1; |
93b7ac65 | 390 | |
9e0a3f98 | 391 | static node *nodehead; /* the head of the binary tree of tags */ |
8378bcd3 | 392 | static node *last_node; /* the last node created */ |
c6d46f5f | 393 | |
db590582 | 394 | static linebuffer lb; /* the current line */ |
24dbe96a | 395 | static linebuffer filebuf; /* a buffer containing the whole file */ |
61a1f6fa | 396 | static linebuffer token_name; /* a buffer containing a tag name */ |
c6d46f5f | 397 | |
55597f90 | 398 | /* boolean "functions" (see init) */ |
db590582 | 399 | static bool _wht[CHARS], _nin[CHARS], _itk[CHARS], _btk[CHARS], _etk[CHARS]; |
988e88ab | 400 | static const char |
d8913c1c | 401 | /* white chars */ |
71cbb895 | 402 | *white = " \f\t\n\r\v", |
30903246 | 403 | /* not in a name */ |
2201e3dc | 404 | *nonam = " \f\t\n\r()=,;", /* look at make_tag before modifying! */ |
d8913c1c | 405 | /* token ending chars */ |
93b7ac65 | 406 | *endtk = " \t\n\r\"'#()[]{}=-+%*/&|^~!<>;,.:?", |
d8913c1c FP |
407 | /* token starting chars */ |
408 | *begtk = "ABCDEFGHIJKLMNOPQRSTUVWXYZ_abcdefghijklmnopqrstuvwxyz$~@", | |
409 | /* valid in-token chars */ | |
30903246 | 410 | *midtk = "ABCDEFGHIJKLMNOPQRSTUVWXYZ_abcdefghijklmnopqrstuvwxyz$0123456789"; |
c6d46f5f | 411 | |
db590582 | 412 | static bool append_to_tagfile; /* -a: append to tags */ |
89d8309f | 413 | /* The next five default to TRUE in C and derived languages. */ |
db590582 FP |
414 | static bool typedefs; /* -t: create tags for C and Ada typedefs */ |
415 | static bool typedefs_or_cplusplus; /* -T: create tags for C typedefs, level */ | |
c6d46f5f | 416 | /* 0 struct/enum/union decls, and C++ */ |
32daa216 | 417 | /* member functions. */ |
db590582 | 418 | static bool constantypedefs; /* -d: create tags for C #define, enum */ |
30903246 | 419 | /* constants and variables. */ |
c6d46f5f | 420 | /* -D: opposite of -d. Default under ctags. */ |
db590582 | 421 | static bool globals; /* create tags for global variables */ |
db590582 | 422 | static bool members; /* create tags for C member variables */ |
78064948 | 423 | static bool declarations; /* --declarations: tag them and extern in C&Co*/ |
50496bd9 | 424 | static bool no_line_directive; /* ignore #line directives (undocumented) */ |
ed8bbc0e | 425 | static bool no_duplicates; /* no duplicate tags for ctags (undocumented) */ |
db590582 FP |
426 | static bool update; /* -u: update tags */ |
427 | static bool vgrind_style; /* -v: create vgrind style index output */ | |
ed8bbc0e | 428 | static bool no_warnings; /* -w: suppress warnings (undocumented) */ |
db590582 | 429 | static bool cxref_style; /* -x: create cxref style output */ |
f96732e5 | 430 | static bool cplusplus; /* .[hc] means C++, not C (undocumented) */ |
3c04a71a | 431 | static bool ignoreindent; /* -I: ignore indentation in C */ |
db590582 | 432 | static bool packages_only; /* --packages-only: in Ada, only tag packages*/ |
30903246 | 433 | |
ce0ae3a9 EZ |
434 | /* STDIN is defined in LynxOS system headers */ |
435 | #ifdef STDIN | |
2e0bea68 | 436 | # undef STDIN |
ce0ae3a9 EZ |
437 | #endif |
438 | ||
97b90b0a FP |
439 | #define STDIN 0x1001 /* returned by getopt_long on --parse-stdin */ |
440 | static bool parsing_stdin; /* --parse-stdin used */ | |
441 | ||
61a1f6fa | 442 | static regexp *p_head; /* list of all regexps */ |
24dbe96a | 443 | static bool need_filebuf; /* some regexes are multi-line */ |
9e0a3f98 | 444 | |
db590582 | 445 | static struct option longopts[] = |
7537186d | 446 | { |
ed8bbc0e FP |
447 | { "append", no_argument, NULL, 'a' }, |
448 | { "packages-only", no_argument, &packages_only, TRUE }, | |
449 | { "c++", no_argument, NULL, 'C' }, | |
450 | { "declarations", no_argument, &declarations, TRUE }, | |
451 | { "no-line-directive", no_argument, &no_line_directive, TRUE }, | |
452 | { "no-duplicates", no_argument, &no_duplicates, TRUE }, | |
453 | { "help", no_argument, NULL, 'h' }, | |
454 | { "help", no_argument, NULL, 'H' }, | |
455 | { "ignore-indentation", no_argument, NULL, 'I' }, | |
456 | { "language", required_argument, NULL, 'l' }, | |
457 | { "members", no_argument, &members, TRUE }, | |
458 | { "no-members", no_argument, &members, FALSE }, | |
459 | { "output", required_argument, NULL, 'o' }, | |
460 | { "regex", required_argument, NULL, 'r' }, | |
461 | { "no-regex", no_argument, NULL, 'R' }, | |
462 | { "ignore-case-regex", required_argument, NULL, 'c' }, | |
97b90b0a | 463 | { "parse-stdin", required_argument, NULL, STDIN }, |
ed8bbc0e | 464 | { "version", no_argument, NULL, 'V' }, |
9e0a3f98 | 465 | |
55102b5d | 466 | #if CTAGS /* Ctags options */ |
ed8bbc0e FP |
467 | { "backward-search", no_argument, NULL, 'B' }, |
468 | { "cxref", no_argument, NULL, 'x' }, | |
469 | { "defines", no_argument, NULL, 'd' }, | |
470 | { "globals", no_argument, &globals, TRUE }, | |
471 | { "typedefs", no_argument, NULL, 't' }, | |
472 | { "typedefs-and-c++", no_argument, NULL, 'T' }, | |
473 | { "update", no_argument, NULL, 'u' }, | |
474 | { "vgrind", no_argument, NULL, 'v' }, | |
475 | { "no-warn", no_argument, NULL, 'w' }, | |
9e0a3f98 | 476 | |
55102b5d | 477 | #else /* Etags options */ |
ed8bbc0e FP |
478 | { "no-defines", no_argument, NULL, 'D' }, |
479 | { "no-globals", no_argument, &globals, FALSE }, | |
480 | { "include", required_argument, NULL, 'i' }, | |
9e0a3f98 | 481 | #endif |
93b7ac65 | 482 | { NULL } |
4746118a JB |
483 | }; |
484 | ||
db590582 | 485 | static compressor compressors[] = |
93b7ac65 FP |
486 | { |
487 | { "z", "gzip -d -c"}, | |
488 | { "Z", "gzip -d -c"}, | |
489 | { "gz", "gzip -d -c"}, | |
490 | { "GZ", "gzip -d -c"}, | |
491 | { "bz2", "bzip2 -d -c" }, | |
4c964351 | 492 | { "xz", "xz -d -c" }, |
93b7ac65 FP |
493 | { NULL } |
494 | }; | |
495 | ||
1f638249 FP |
496 | /* |
497 | * Language stuff. | |
498 | */ | |
b9755a12 | 499 | |
93b7ac65 | 500 | /* Ada code */ |
988e88ab | 501 | static const char *Ada_suffixes [] = |
93b7ac65 | 502 | { "ads", "adb", "ada", NULL }; |
988e88ab | 503 | static const char Ada_help [] = |
4fdb4553 FP |
504 | "In Ada code, functions, procedures, packages, tasks and types are\n\ |
505 | tags. Use the `--packages-only' option to create tags for\n\ | |
506 | packages only.\n\ | |
507 | Ada tag names have suffixes indicating the type of entity:\n\ | |
508 | Entity type: Qualifier:\n\ | |
509 | ------------ ----------\n\ | |
510 | function /f\n\ | |
511 | procedure /p\n\ | |
512 | package spec /s\n\ | |
513 | package body /b\n\ | |
514 | type /t\n\ | |
515 | task /k\n\ | |
516 | Thus, `M-x find-tag <RET> bidule/b <RET>' will go directly to the\n\ | |
517 | body of the package `bidule', while `M-x find-tag <RET> bidule <RET>'\n\ | |
518 | will just search for any tag `bidule'."; | |
1f638249 FP |
519 | |
520 | /* Assembly code */ | |
988e88ab | 521 | static const char *Asm_suffixes [] = |
db590582 FP |
522 | { "a", /* Unix assembler */ |
523 | "asm", /* Microcontroller assembly */ | |
524 | "def", /* BSO/Tasking definition includes */ | |
525 | "inc", /* Microcontroller include files */ | |
526 | "ins", /* Microcontroller include files */ | |
527 | "s", "sa", /* Unix assembler */ | |
528 | "S", /* cpp-processed Unix assembler */ | |
529 | "src", /* BSO/Tasking C compiler output */ | |
530 | NULL | |
531 | }; | |
988e88ab | 532 | static const char Asm_help [] = |
4fdb4553 FP |
533 | "In assembler code, labels appearing at the beginning of a line,\n\ |
534 | followed by a colon, are tags."; | |
535 | ||
1f638249 FP |
536 | |
537 | /* Note that .c and .h can be considered C++, if the --c++ flag was | |
f96732e5 | 538 | given, or if the `class' or `template' keywords are met inside the file. |
b28e26be | 539 | That is why default_C_entries is called for these. */ |
988e88ab | 540 | static const char *default_C_suffixes [] = |
1f638249 | 541 | { "c", "h", NULL }; |
f96732e5 | 542 | #if CTAGS /* C help for Ctags */ |
988e88ab | 543 | static const char default_C_help [] = |
f96732e5 FP |
544 | "In C code, any C function is a tag. Use -t to tag typedefs.\n\ |
545 | Use -T to tag definitions of `struct', `union' and `enum'.\n\ | |
546 | Use -d to tag `#define' macro definitions and `enum' constants.\n\ | |
547 | Use --globals to tag global variables.\n\ | |
548 | You can tag function declarations and external variables by\n\ | |
549 | using `--declarations', and struct members by using `--members'."; | |
550 | #else /* C help for Etags */ | |
988e88ab | 551 | static const char default_C_help [] = |
4fdb4553 FP |
552 | "In C code, any C function or typedef is a tag, and so are\n\ |
553 | definitions of `struct', `union' and `enum'. `#define' macro\n\ | |
554 | definitions and `enum' constants are tags unless you specify\n\ | |
555 | `--no-defines'. Global variables are tags unless you specify\n\ | |
78064948 FP |
556 | `--no-globals' and so are struct members unless you specify\n\ |
557 | `--no-members'. Use of `--no-globals', `--no-defines' and\n\ | |
558 | `--no-members' can make the tags table file much smaller.\n\ | |
4fdb4553 | 559 | You can tag function declarations and external variables by\n\ |
78064948 | 560 | using `--declarations'."; |
f96732e5 | 561 | #endif /* C help for Ctags and Etags */ |
1f638249 | 562 | |
988e88ab | 563 | static const char *Cplusplus_suffixes [] = |
89d57763 | 564 | { "C", "c++", "cc", "cpp", "cxx", "H", "h++", "hh", "hpp", "hxx", |
cec68fcb | 565 | "M", /* Objective C++ */ |
7877f373 | 566 | "pdb", /* PostScript with C syntax */ |
cec68fcb | 567 | NULL }; |
988e88ab | 568 | static const char Cplusplus_help [] = |
373df066 DL |
569 | "In C++ code, all the tag constructs of C code are tagged. (Use\n\ |
570 | --help --lang=c --lang=c++ for full help.)\n\ | |
78064948 FP |
571 | In addition to C tags, member functions are also recognized. Member\n\ |
572 | variables are recognized unless you use the `--no-members' option.\n\ | |
4fdb4553 FP |
573 | Tags for variables and functions in classes are named `CLASS::VARIABLE'\n\ |
574 | and `CLASS::FUNCTION'. `operator' definitions have tag names like\n\ | |
575 | `operator+'."; | |
cec68fcb | 576 | |
988e88ab | 577 | static const char *Cjava_suffixes [] = |
cec68fcb | 578 | { "java", NULL }; |
4fdb4553 FP |
579 | static char Cjava_help [] = |
580 | "In Java code, all the tags constructs of C and C++ code are\n\ | |
373df066 | 581 | tagged. (Use --help --lang=c --lang=c++ --lang=java for full help.)"; |
4fdb4553 | 582 | |
1f638249 | 583 | |
988e88ab | 584 | static const char *Cobol_suffixes [] = |
30903246 | 585 | { "COB", "cob", NULL }; |
4fdb4553 FP |
586 | static char Cobol_help [] = |
587 | "In Cobol code, tags are paragraph names; that is, any word\n\ | |
588 | starting in column 8 and followed by a period."; | |
30903246 | 589 | |
988e88ab | 590 | static const char *Cstar_suffixes [] = |
1f638249 FP |
591 | { "cs", "hs", NULL }; |
592 | ||
988e88ab | 593 | static const char *Erlang_suffixes [] = |
8dc7496c | 594 | { "erl", "hrl", NULL }; |
988e88ab | 595 | static const char Erlang_help [] = |
4fdb4553 FP |
596 | "In Erlang code, the tags are the functions, records and macros\n\ |
597 | defined in the file."; | |
8dc7496c | 598 | |
988e88ab | 599 | const char *Forth_suffixes [] = |
55102b5d | 600 | { "fth", "tok", NULL }; |
988e88ab | 601 | static const char Forth_help [] = |
55102b5d FP |
602 | "In Forth code, tags are words defined by `:',\n\ |
603 | constant, code, create, defer, value, variable, buffer:, field."; | |
604 | ||
988e88ab | 605 | static const char *Fortran_suffixes [] = |
1f638249 | 606 | { "F", "f", "f90", "for", NULL }; |
988e88ab | 607 | static const char Fortran_help [] = |
4fdb4553 | 608 | "In Fortran code, functions, subroutines and block data are tags."; |
1f638249 | 609 | |
988e88ab | 610 | static const char *HTML_suffixes [] = |
61a1f6fa | 611 | { "htm", "html", "shtml", NULL }; |
988e88ab | 612 | static const char HTML_help [] = |
4fdb4553 FP |
613 | "In HTML input files, the tags are the `title' and the `h1', `h2',\n\ |
614 | `h3' headers. Also, tags are `name=' in anchors and all\n\ | |
615 | occurrences of `id='."; | |
61a1f6fa | 616 | |
988e88ab | 617 | static const char *Lisp_suffixes [] = |
89d57763 | 618 | { "cl", "clisp", "el", "l", "lisp", "LSP", "lsp", "ml", NULL }; |
988e88ab | 619 | static const char Lisp_help [] = |
4fdb4553 FP |
620 | "In Lisp code, any function defined with `defun', any variable\n\ |
621 | defined with `defvar' or `defconst', and in general the first\n\ | |
622 | argument of any expression that starts with `(def' in column zero\n\ | |
1cbaa705 KR |
623 | is a tag.\n\ |
624 | The `--declarations' option tags \"(defvar foo)\" constructs too."; | |
89d57763 | 625 | |
988e88ab | 626 | static const char *Lua_suffixes [] = |
05d9a399 | 627 | { "lua", "LUA", NULL }; |
988e88ab | 628 | static const char Lua_help [] = |
05d9a399 FP |
629 | "In Lua scripts, all functions are tags."; |
630 | ||
988e88ab | 631 | static const char *Makefile_filenames [] = |
89d57763 | 632 | { "Makefile", "makefile", "GNUMakefile", "Makefile.in", "Makefile.am", NULL}; |
988e88ab | 633 | static const char Makefile_help [] = |
4fdb4553 FP |
634 | "In makefiles, targets are tags; additionally, variables are tags\n\ |
635 | unless you specify `--no-globals'."; | |
636 | ||
988e88ab | 637 | static const char *Objc_suffixes [] = |
4fdb4553 FP |
638 | { "lm", /* Objective lex file */ |
639 | "m", /* Objective C file */ | |
640 | NULL }; | |
988e88ab | 641 | static const char Objc_help [] = |
4fdb4553 FP |
642 | "In Objective C code, tags include Objective C definitions for classes,\n\ |
643 | class categories, methods and protocols. Tags for variables and\n\ | |
35a4c758 FP |
644 | functions in classes are named `CLASS::VARIABLE' and `CLASS::FUNCTION'.\n\ |
645 | (Use --help --lang=c --lang=objc --lang=java for full help.)"; | |
1f638249 | 646 | |
988e88ab | 647 | static const char *Pascal_suffixes [] = |
1f638249 | 648 | { "p", "pas", NULL }; |
988e88ab | 649 | static const char Pascal_help [] = |
4fdb4553 FP |
650 | "In Pascal code, the tags are the functions and procedures defined\n\ |
651 | in the file."; | |
35a4c758 | 652 | /* " // this is for working around an Emacs highlighting bug... */ |
1f638249 | 653 | |
988e88ab | 654 | static const char *Perl_suffixes [] = |
1f638249 | 655 | { "pl", "pm", NULL }; |
988e88ab | 656 | static const char *Perl_interpreters [] = |
d8913c1c | 657 | { "perl", "@PERL@", NULL }; |
988e88ab | 658 | static const char Perl_help [] = |
4fdb4553 FP |
659 | "In Perl code, the tags are the packages, subroutines and variables\n\ |
660 | defined by the `package', `sub', `my' and `local' keywords. Use\n\ | |
661 | `--globals' if you want to tag global variables. Tags for\n\ | |
662 | subroutines are named `PACKAGE::SUB'. The name for subroutines\n\ | |
663 | defined in the default package is `main::SUB'."; | |
1f638249 | 664 | |
988e88ab | 665 | static const char *PHP_suffixes [] = |
62aec606 | 666 | { "php", "php3", "php4", NULL }; |
988e88ab | 667 | static const char PHP_help [] = |
78064948 FP |
668 | "In PHP code, tags are functions, classes and defines. Unless you use\n\ |
669 | the `--no-members' option, vars are tags too."; | |
62aec606 | 670 | |
988e88ab | 671 | static const char *plain_C_suffixes [] = |
4fdb4553 | 672 | { "pc", /* Pro*C file */ |
d8913c1c | 673 | NULL }; |
1f638249 | 674 | |
988e88ab | 675 | static const char *PS_suffixes [] = |
93b7ac65 | 676 | { "ps", "psw", NULL }; /* .psw is for PSWrap */ |
988e88ab | 677 | static const char PS_help [] = |
4fdb4553 | 678 | "In PostScript code, the tags are the functions."; |
cec68fcb | 679 | |
988e88ab | 680 | static const char *Prolog_suffixes [] = |
1f638249 | 681 | { "prolog", NULL }; |
988e88ab | 682 | static const char Prolog_help [] = |
4fdb4553 FP |
683 | "In Prolog code, tags are predicates and rules at the beginning of\n\ |
684 | line."; | |
1f638249 | 685 | |
988e88ab | 686 | static const char *Python_suffixes [] = |
93b7ac65 | 687 | { "py", NULL }; |
988e88ab | 688 | static const char Python_help [] = |
4fdb4553 FP |
689 | "In Python code, `def' or `class' at the beginning of a line\n\ |
690 | generate a tag."; | |
93b7ac65 | 691 | |
d8913c1c | 692 | /* Can't do the `SCM' or `scm' prefix with a version number. */ |
988e88ab | 693 | static const char *Scheme_suffixes [] = |
89d57763 | 694 | { "oak", "sch", "scheme", "SCM", "scm", "SM", "sm", "ss", "t", NULL }; |
988e88ab | 695 | static const char Scheme_help [] = |
4fdb4553 FP |
696 | "In Scheme code, tags include anything defined with `def' or with a\n\ |
697 | construct whose name starts with `def'. They also include\n\ | |
698 | variables set with `set!' at top level in the file."; | |
1f638249 | 699 | |
988e88ab | 700 | static const char *TeX_suffixes [] = |
89d57763 | 701 | { "bib", "clo", "cls", "ltx", "sty", "TeX", "tex", NULL }; |
988e88ab | 702 | static const char TeX_help [] = |
4fdb4553 FP |
703 | "In LaTeX text, the argument of any of the commands `\\chapter',\n\ |
704 | `\\section', `\\subsection', `\\subsubsection', `\\eqno', `\\label',\n\ | |
705 | `\\ref', `\\cite', `\\bibitem', `\\part', `\\appendix', `\\entry',\n\ | |
706 | `\\index', `\\def', `\\newcommand', `\\renewcommand',\n\ | |
707 | `\\newenvironment' or `\\renewenvironment' is a tag.\n\ | |
708 | \n\ | |
709 | Other commands can be specified by setting the environment variable\n\ | |
710 | `TEXTAGS' to a colon-separated list like, for example,\n\ | |
711 | TEXTAGS=\"mycommand:myothercommand\"."; | |
712 | ||
1f638249 | 713 | |
988e88ab | 714 | static const char *Texinfo_suffixes [] = |
89d57763 | 715 | { "texi", "texinfo", "txi", NULL }; |
988e88ab | 716 | static const char Texinfo_help [] = |
4fdb4553 | 717 | "for texinfo files, lines starting with @node are tagged."; |
97052c63 | 718 | |
988e88ab | 719 | static const char *Yacc_suffixes [] = |
89d57763 | 720 | { "y", "y++", "ym", "yxx", "yy", NULL }; /* .ym is Objective yacc file */ |
988e88ab | 721 | static const char Yacc_help [] = |
4fdb4553 FP |
722 | "In Bison or Yacc input files, each rule defines as a tag the\n\ |
723 | nonterminal it constructs. The portions of the file that contain\n\ | |
724 | C code are parsed as C code (use --help --lang=c --lang=yacc\n\ | |
725 | for full help)."; | |
726 | ||
988e88ab | 727 | static const char auto_help [] = |
4fdb4553 FP |
728 | "`auto' is not a real language, it indicates to use\n\ |
729 | a default language for files base on file name suffix and file contents."; | |
730 | ||
988e88ab | 731 | static const char none_help [] = |
4fdb4553 FP |
732 | "`none' is not a real language, it indicates to only do\n\ |
733 | regexp processing on files."; | |
734 | ||
988e88ab | 735 | static const char no_lang_help [] = |
4fdb4553 FP |
736 | "No detailed help available for this language."; |
737 | ||
1f638249 | 738 | |
93b7ac65 FP |
739 | /* |
740 | * Table of languages. | |
741 | * | |
742 | * It is ok for a given function to be listed under more than one | |
743 | * name. I just didn't. | |
744 | */ | |
b9755a12 | 745 | |
db590582 | 746 | static language lang_names [] = |
b9755a12 | 747 | { |
4fdb4553 FP |
748 | { "ada", Ada_help, Ada_funcs, Ada_suffixes }, |
749 | { "asm", Asm_help, Asm_labels, Asm_suffixes }, | |
750 | { "c", default_C_help, default_C_entries, default_C_suffixes }, | |
751 | { "c++", Cplusplus_help, Cplusplus_entries, Cplusplus_suffixes }, | |
752 | { "c*", no_lang_help, Cstar_entries, Cstar_suffixes }, | |
753 | { "cobol", Cobol_help, Cobol_paragraphs, Cobol_suffixes }, | |
754 | { "erlang", Erlang_help, Erlang_functions, Erlang_suffixes }, | |
55102b5d | 755 | { "forth", Forth_help, Forth_words, Forth_suffixes }, |
4fdb4553 FP |
756 | { "fortran", Fortran_help, Fortran_functions, Fortran_suffixes }, |
757 | { "html", HTML_help, HTML_labels, HTML_suffixes }, | |
758 | { "java", Cjava_help, Cjava_entries, Cjava_suffixes }, | |
759 | { "lisp", Lisp_help, Lisp_functions, Lisp_suffixes }, | |
05d9a399 | 760 | { "lua", Lua_help, Lua_functions, Lua_suffixes }, |
4fdb4553 FP |
761 | { "makefile", Makefile_help,Makefile_targets,NULL,Makefile_filenames}, |
762 | { "objc", Objc_help, plain_C_entries, Objc_suffixes }, | |
763 | { "pascal", Pascal_help, Pascal_functions, Pascal_suffixes }, | |
764 | { "perl",Perl_help,Perl_functions,Perl_suffixes,NULL,Perl_interpreters}, | |
765 | { "php", PHP_help, PHP_functions, PHP_suffixes }, | |
766 | { "postscript",PS_help, PS_functions, PS_suffixes }, | |
767 | { "proc", no_lang_help, plain_C_entries, plain_C_suffixes }, | |
768 | { "prolog", Prolog_help, Prolog_functions, Prolog_suffixes }, | |
769 | { "python", Python_help, Python_functions, Python_suffixes }, | |
770 | { "scheme", Scheme_help, Scheme_functions, Scheme_suffixes }, | |
771 | { "tex", TeX_help, TeX_commands, TeX_suffixes }, | |
772 | { "texinfo", Texinfo_help, Texinfo_nodes, Texinfo_suffixes }, | |
773 | { "yacc", Yacc_help,Yacc_entries,Yacc_suffixes,NULL,NULL,TRUE}, | |
774 | { "auto", auto_help }, /* default guessing scheme */ | |
775 | { "none", none_help, just_read_file }, /* regexp matching only */ | |
776 | { NULL } /* end of list */ | |
b9755a12 | 777 | }; |
b2521b0a | 778 | |
c6d46f5f | 779 | \f |
944bdd72 | 780 | static void |
873fbd0b | 781 | print_language_names (void) |
b9755a12 | 782 | { |
93b7ac65 | 783 | language *lang; |
988e88ab | 784 | const char **name, **ext; |
b9755a12 FP |
785 | |
786 | puts ("\nThese are the currently supported languages, along with the\n\ | |
83be933c | 787 | default file names and dot suffixes:"); |
1f638249 | 788 | for (lang = lang_names; lang->name != NULL; lang++) |
b9755a12 | 789 | { |
83be933c FP |
790 | printf (" %-*s", 10, lang->name); |
791 | if (lang->filenames != NULL) | |
792 | for (name = lang->filenames; *name != NULL; name++) | |
793 | printf (" %s", *name); | |
1f638249 FP |
794 | if (lang->suffixes != NULL) |
795 | for (ext = lang->suffixes; *ext != NULL; ext++) | |
796 | printf (" .%s", *ext); | |
b9755a12 FP |
797 | puts (""); |
798 | } | |
4fdb4553 | 799 | puts ("where `auto' means use default language for files based on file\n\ |
79263656 FP |
800 | name suffix, and `none' means only do regexp processing on files.\n\ |
801 | If no language is specified and no matching suffix is found,\n\ | |
1f638249 FP |
802 | the first line of the file is read for a sharp-bang (#!) sequence\n\ |
803 | followed by the name of an interpreter. If no such sequence is found,\n\ | |
93b7ac65 | 804 | Fortran is tried first; if no tags are found, C is tried next.\n\ |
4fdb4553 FP |
805 | When parsing any C file, a \"class\" or \"template\" keyword\n\ |
806 | switches to C++."); | |
4c964351 | 807 | puts ("Compressed files are supported using gzip, bzip2, and xz.\n\ |
4fdb4553 FP |
808 | \n\ |
809 | For detailed help on a given language use, for example,\n\ | |
810 | etags --help --lang=ada."); | |
b9755a12 FP |
811 | } |
812 | ||
944bdd72 | 813 | #ifndef EMACS_NAME |
24dbe96a | 814 | # define EMACS_NAME "standalone" |
944bdd72 | 815 | #endif |
c5007f46 | 816 | #ifndef VERSION |
d3706fa9 | 817 | # define VERSION "17.38.1.4" |
c5007f46 | 818 | #endif |
944bdd72 | 819 | static void |
873fbd0b | 820 | print_version (void) |
4746118a | 821 | { |
78f83752 | 822 | char emacs_copyright[] = COPYRIGHT; |
d16c1140 | 823 | |
944bdd72 | 824 | printf ("%s (%s %s)\n", (CTAGS) ? "ctags" : "etags", EMACS_NAME, VERSION); |
d16c1140 | 825 | puts (emacs_copyright); |
b7e1b21b | 826 | puts ("This program is distributed under the terms in ETAGS.README"); |
4746118a | 827 | |
3f0656ff | 828 | exit (EXIT_SUCCESS); |
4746118a JB |
829 | } |
830 | ||
172aa4c1 FP |
831 | #ifndef PRINT_UNDOCUMENTED_OPTIONS_HELP |
832 | # define PRINT_UNDOCUMENTED_OPTIONS_HELP FALSE | |
833 | #endif | |
834 | ||
944bdd72 | 835 | static void |
873fbd0b | 836 | print_help (argument *argbuffer) |
4746118a | 837 | { |
4fdb4553 FP |
838 | bool help_for_lang = FALSE; |
839 | ||
840 | for (; argbuffer->arg_type != at_end; argbuffer++) | |
841 | if (argbuffer->arg_type == at_language) | |
842 | { | |
843 | if (help_for_lang) | |
844 | puts (""); | |
845 | puts (argbuffer->lang->help); | |
846 | help_for_lang = TRUE; | |
847 | } | |
848 | ||
849 | if (help_for_lang) | |
3f0656ff | 850 | exit (EXIT_SUCCESS); |
4fdb4553 | 851 | |
30903246 FP |
852 | printf ("Usage: %s [options] [[regex-option ...] file-name] ...\n\ |
853 | \n\ | |
854 | These are the options accepted by %s.\n", progname, progname); | |
9239d970 | 855 | puts ("You may use unambiguous abbreviations for the long option names."); |
9e0a3f98 FP |
856 | puts (" A - as file name means read names from stdin (one per line).\n\ |
857 | Absolute names are stored in the output file as they are.\n\ | |
858 | Relative ones are stored relative to the output file's directory.\n"); | |
4746118a | 859 | |
55102b5d | 860 | puts ("-a, --append\n\ |
52cc7c59 | 861 | Append tag entries to existing tags file."); |
1a0d8c80 | 862 | |
93b7ac65 | 863 | puts ("--packages-only\n\ |
9e0a3f98 | 864 | For Ada files, only generate tags for packages."); |
93b7ac65 | 865 | |
32daa216 FP |
866 | if (CTAGS) |
867 | puts ("-B, --backward-search\n\ | |
1a0d8c80 | 868 | Write the search commands for the tag entries using '?', the\n\ |
3ad2882c | 869 | backward-search command instead of '/', the forward-search command."); |
1a0d8c80 | 870 | |
8c463abe FP |
871 | /* This option is mostly obsolete, because etags can now automatically |
872 | detect C++. Retained for backward compatibility and for debugging and | |
873 | experimentation. In principle, we could want to tag as C++ even | |
4fdb4553 | 874 | before any "class" or "template" keyword. |
52cc7c59 | 875 | puts ("-C, --c++\n\ |
79263656 | 876 | Treat files whose name suffix defaults to C language as C++ files."); |
8c463abe | 877 | */ |
4746118a | 878 | |
93b7ac65 FP |
879 | puts ("--declarations\n\ |
880 | In C and derived languages, create tags for function declarations,"); | |
881 | if (CTAGS) | |
882 | puts ("\tand create tags for extern variables if --globals is used."); | |
883 | else | |
884 | puts | |
885 | ("\tand create tags for extern variables unless --no-globals is used."); | |
886 | ||
32daa216 FP |
887 | if (CTAGS) |
888 | puts ("-d, --defines\n\ | |
3ca80e28 | 889 | Create tag entries for C #define constants and enum constants, too."); |
32daa216 FP |
890 | else |
891 | puts ("-D, --no-defines\n\ | |
3ca80e28 FP |
892 | Don't create tag entries for C #define constants and enum constants.\n\ |
893 | This makes the tags file smaller."); | |
4746118a | 894 | |
32daa216 | 895 | if (!CTAGS) |
9e0a3f98 | 896 | puts ("-i FILE, --include=FILE\n\ |
1a0d8c80 FP |
897 | Include a note in tag file indicating that, when searching for\n\ |
898 | a tag, one should also consult the tags file FILE after\n\ | |
899 | checking the current file."); | |
9e0a3f98 FP |
900 | |
901 | puts ("-l LANG, --language=LANG\n\ | |
b9755a12 FP |
902 | Force the following files to be considered as written in the\n\ |
903 | named language up to the next --language=LANG option."); | |
7537186d | 904 | |
30903246 FP |
905 | if (CTAGS) |
906 | puts ("--globals\n\ | |
93b7ac65 | 907 | Create tag entries for global variables in some languages."); |
30903246 FP |
908 | else |
909 | puts ("--no-globals\n\ | |
93b7ac65 FP |
910 | Do not create tag entries for global variables in some\n\ |
911 | languages. This makes the tags file smaller."); | |
172aa4c1 FP |
912 | |
913 | if (PRINT_UNDOCUMENTED_OPTIONS_HELP) | |
914 | puts ("--no-line-directive\n\ | |
915 | Ignore #line preprocessor directives in C and derived languages."); | |
916 | ||
eac50785 FP |
917 | if (CTAGS) |
918 | puts ("--members\n\ | |
919 | Create tag entries for members of structures in some languages."); | |
920 | else | |
921 | puts ("--no-members\n\ | |
78064948 FP |
922 | Do not create tag entries for members of structures\n\ |
923 | in some languages."); | |
30903246 | 924 | |
89fb2be1 | 925 | puts ("-r REGEXP, --regex=REGEXP or --regex=@regexfile\n\ |
3c04a71a | 926 | Make a tag for each line matching a regular expression pattern\n\ |
89fb2be1 FP |
927 | in the following files. {LANGUAGE}REGEXP uses REGEXP for LANGUAGE\n\ |
928 | files only. REGEXFILE is a file containing one REGEXP per line.\n\ | |
929 | REGEXP takes the form /TAGREGEXP/TAGNAME/MODS, where TAGNAME/ is\n\ | |
930 | optional. The TAGREGEXP pattern is anchored (as if preceded by ^)."); | |
931 | puts (" If TAGNAME/ is present, the tags created are named.\n\ | |
93b7ac65 | 932 | For example Tcl named tags can be created with:\n\ |
89fb2be1 FP |
933 | --regex=\"/proc[ \\t]+\\([^ \\t]+\\)/\\1/.\".\n\ |
934 | MODS are optional one-letter modifiers: `i' means to ignore case,\n\ | |
935 | `m' means to allow multi-line matches, `s' implies `m' and\n\ | |
3c04a71a | 936 | causes dot to match any character, including newline."); |
172aa4c1 | 937 | |
7537186d | 938 | puts ("-R, --no-regex\n\ |
b9755a12 | 939 | Don't create tags from regexps for the following files."); |
172aa4c1 | 940 | |
201f9f2b | 941 | puts ("-I, --ignore-indentation\n\ |
3c04a71a FP |
942 | In C and C++ do not assume that a closing brace in the first\n\ |
943 | column is the final brace of a function or structure definition."); | |
172aa4c1 | 944 | |
e7d3b099 FP |
945 | puts ("-o FILE, --output=FILE\n\ |
946 | Write the tags to FILE."); | |
172aa4c1 | 947 | |
e7d3b099 FP |
948 | puts ("--parse-stdin=NAME\n\ |
949 | Read from standard input and record tags as belonging to file NAME."); | |
4746118a | 950 | |
32daa216 FP |
951 | if (CTAGS) |
952 | { | |
953 | puts ("-t, --typedefs\n\ | |
93b7ac65 | 954 | Generate tag entries for C and Ada typedefs."); |
32daa216 FP |
955 | puts ("-T, --typedefs-and-c++\n\ |
956 | Generate tag entries for C typedefs, C struct/enum/union tags,\n\ | |
957 | and C++ member functions."); | |
9e0a3f98 FP |
958 | } |
959 | ||
960 | if (CTAGS) | |
961 | puts ("-u, --update\n\ | |
4746118a JB |
962 | Update the tag entries for the given files, leaving tag\n\ |
963 | entries for other files in place. Currently, this is\n\ | |
964 | implemented by deleting the existing entries for the given\n\ | |
965 | files and then rewriting the new entries at the end of the\n\ | |
966 | tags file. It is often faster to simply rebuild the entire\n\ | |
52cc7c59 | 967 | tag file than to use this."); |
9e0a3f98 FP |
968 | |
969 | if (CTAGS) | |
970 | { | |
32daa216 | 971 | puts ("-v, --vgrind\n\ |
2e0bea68 FP |
972 | Print on the standard output an index of items intended for\n\ |
973 | human consumption, similar to the output of vgrind. The index\n\ | |
974 | is sorted, and gives the page number of each item."); | |
172aa4c1 FP |
975 | |
976 | if (PRINT_UNDOCUMENTED_OPTIONS_HELP) | |
977 | puts ("-w, --no-duplicates\n\ | |
ed8bbc0e FP |
978 | Do not create duplicate tag entries, for compatibility with\n\ |
979 | traditional ctags."); | |
172aa4c1 FP |
980 | |
981 | if (PRINT_UNDOCUMENTED_OPTIONS_HELP) | |
982 | puts ("-w, --no-warn\n\ | |
ed8bbc0e | 983 | Suppress warning messages about duplicate tag entries."); |
172aa4c1 | 984 | |
32daa216 | 985 | puts ("-x, --cxref\n\ |
4746118a JB |
986 | Like --vgrind, but in the style of cxref, rather than vgrind.\n\ |
987 | The output uses line numbers instead of page numbers, but\n\ | |
988 | beyond that the differences are cosmetic; try both to see\n\ | |
52cc7c59 | 989 | which you like."); |
32daa216 | 990 | } |
4746118a JB |
991 | |
992 | puts ("-V, --version\n\ | |
993 | Print the version of the program.\n\ | |
7537186d | 994 | -h, --help\n\ |
4fdb4553 FP |
995 | Print this help message.\n\ |
996 | Followed by one or more `--language' options prints detailed\n\ | |
997 | help about tag generation for the specified languages."); | |
4746118a | 998 | |
b9755a12 FP |
999 | print_language_names (); |
1000 | ||
b6b48846 | 1001 | puts (""); |
a5c7ef8b | 1002 | puts ("Report bugs to bug-gnu-emacs@gnu.org"); |
b6b48846 | 1003 | |
3f0656ff | 1004 | exit (EXIT_SUCCESS); |
4746118a JB |
1005 | } |
1006 | ||
1007 | \f | |
d0dff6e5 | 1008 | int |
873fbd0b | 1009 | main (int argc, char **argv) |
c6d46f5f | 1010 | { |
c6d46f5f | 1011 | int i; |
2f608d34 FP |
1012 | unsigned int nincluded_files; |
1013 | char **included_files; | |
1f638249 | 1014 | argument *argbuffer; |
2f608d34 | 1015 | int current_arg, file_count; |
93b7ac65 | 1016 | linebuffer filename_lb; |
4fdb4553 | 1017 | bool help_asked = FALSE; |
9250f758 | 1018 | ptrdiff_t len; |
97b90b0a FP |
1019 | char *optstring; |
1020 | int opt; | |
1021 | ||
c5007f46 | 1022 | |
c05b6df5 | 1023 | #ifdef DOS_NT |
42680d3c | 1024 | _fmode = O_BINARY; /* all of files are treated as binary files */ |
c05b6df5 | 1025 | #endif /* DOS_NT */ |
c6880c90 | 1026 | |
c6d46f5f | 1027 | progname = argv[0]; |
2f608d34 FP |
1028 | nincluded_files = 0; |
1029 | included_files = xnew (argc, char *); | |
1030 | current_arg = 0; | |
1031 | file_count = 0; | |
c6d46f5f | 1032 | |
b9755a12 FP |
1033 | /* Allocate enough no matter what happens. Overkill, but each one |
1034 | is small. */ | |
1f638249 | 1035 | argbuffer = xnew (argc, argument); |
b9755a12 | 1036 | |
c6d46f5f | 1037 | /* |
89d8309f | 1038 | * Always find typedefs and structure tags. |
78064948 | 1039 | * Also default to find macro constants, enum constants, struct |
89d8309f | 1040 | * members and global variables. Do it for both etags and ctags. |
c6d46f5f | 1041 | */ |
89d8309f FP |
1042 | typedefs = typedefs_or_cplusplus = constantypedefs = TRUE; |
1043 | globals = members = TRUE; | |
c6d46f5f | 1044 | |
55102b5d FP |
1045 | /* When the optstring begins with a '-' getopt_long does not rearrange the |
1046 | non-options arguments to be at the end, but leaves them alone. */ | |
9239d970 PE |
1047 | optstring = concat ("-ac:Cf:Il:o:r:RSVhH", |
1048 | (CTAGS) ? "BxdtTuvw" : "Di:", | |
1049 | ""); | |
4746118a | 1050 | |
55102b5d | 1051 | while ((opt = getopt_long (argc, argv, optstring, longopts, NULL)) != EOF) |
97b90b0a FP |
1052 | switch (opt) |
1053 | { | |
1054 | case 0: | |
1055 | /* If getopt returns 0, then it has already processed a | |
1056 | long-named option. We should do nothing. */ | |
4746118a JB |
1057 | break; |
1058 | ||
97b90b0a FP |
1059 | case 1: |
1060 | /* This means that a file name has been seen. Record it. */ | |
1061 | argbuffer[current_arg].arg_type = at_filename; | |
1062 | argbuffer[current_arg].what = optarg; | |
9250f758 PE |
1063 | len = strlen (optarg); |
1064 | if (whatlen_max < len) | |
1065 | whatlen_max = len; | |
97b90b0a FP |
1066 | ++current_arg; |
1067 | ++file_count; | |
1068 | break; | |
4746118a | 1069 | |
97b90b0a FP |
1070 | case STDIN: |
1071 | /* Parse standard input. Idea by Vivek <vivek@etla.org>. */ | |
1072 | argbuffer[current_arg].arg_type = at_stdin; | |
1073 | argbuffer[current_arg].what = optarg; | |
9250f758 PE |
1074 | len = strlen (optarg); |
1075 | if (whatlen_max < len) | |
1076 | whatlen_max = len; | |
97b90b0a FP |
1077 | ++current_arg; |
1078 | ++file_count; | |
e7d3b099 FP |
1079 | if (parsing_stdin) |
1080 | fatal ("cannot parse standard input more than once", (char *)NULL); | |
97b90b0a FP |
1081 | parsing_stdin = TRUE; |
1082 | break; | |
b9755a12 | 1083 | |
97b90b0a | 1084 | /* Common options. */ |
55102b5d | 1085 | case 'a': append_to_tagfile = TRUE; break; |
97b90b0a FP |
1086 | case 'C': cplusplus = TRUE; break; |
1087 | case 'f': /* for compatibility with old makefiles */ | |
1088 | case 'o': | |
1089 | if (tagfile) | |
93b7ac65 | 1090 | { |
db5a3003 | 1091 | error ("-o option may only be given once."); |
97b90b0a | 1092 | suggest_asking_for_help (); |
4fdb4553 | 1093 | /* NOTREACHED */ |
93b7ac65 | 1094 | } |
97b90b0a FP |
1095 | tagfile = optarg; |
1096 | break; | |
1097 | case 'I': | |
1098 | case 'S': /* for backward compatibility */ | |
3c04a71a | 1099 | ignoreindent = TRUE; |
97b90b0a FP |
1100 | break; |
1101 | case 'l': | |
1102 | { | |
1103 | language *lang = get_language_from_langname (optarg); | |
1104 | if (lang != NULL) | |
1105 | { | |
1106 | argbuffer[current_arg].lang = lang; | |
1107 | argbuffer[current_arg].arg_type = at_language; | |
1108 | ++current_arg; | |
1109 | } | |
c6d46f5f | 1110 | } |
97b90b0a | 1111 | break; |
24dbe96a FP |
1112 | case 'c': |
1113 | /* Backward compatibility: support obsolete --ignore-case-regexp. */ | |
1114 | optarg = concat (optarg, "i", ""); /* memory leak here */ | |
1115 | /* FALLTHRU */ | |
97b90b0a FP |
1116 | case 'r': |
1117 | argbuffer[current_arg].arg_type = at_regexp; | |
1118 | argbuffer[current_arg].what = optarg; | |
9250f758 PE |
1119 | len = strlen (optarg); |
1120 | if (whatlen_max < len) | |
1121 | whatlen_max = len; | |
97b90b0a FP |
1122 | ++current_arg; |
1123 | break; | |
1124 | case 'R': | |
1125 | argbuffer[current_arg].arg_type = at_regexp; | |
1126 | argbuffer[current_arg].what = NULL; | |
1127 | ++current_arg; | |
1128 | break; | |
97b90b0a FP |
1129 | case 'V': |
1130 | print_version (); | |
1131 | break; | |
1132 | case 'h': | |
1133 | case 'H': | |
4fdb4553 | 1134 | help_asked = TRUE; |
97b90b0a FP |
1135 | break; |
1136 | ||
1137 | /* Etags options */ | |
97b90b0a FP |
1138 | case 'D': constantypedefs = FALSE; break; |
1139 | case 'i': included_files[nincluded_files++] = optarg; break; | |
1140 | ||
1141 | /* Ctags options. */ | |
1142 | case 'B': searchar = '?'; break; | |
1143 | case 'd': constantypedefs = TRUE; break; | |
1144 | case 't': typedefs = TRUE; break; | |
1145 | case 'T': typedefs = typedefs_or_cplusplus = TRUE; break; | |
1146 | case 'u': update = TRUE; break; | |
1147 | case 'v': vgrind_style = TRUE; /*FALLTHRU*/ | |
1148 | case 'x': cxref_style = TRUE; break; | |
1149 | case 'w': no_warnings = TRUE; break; | |
1150 | default: | |
1151 | suggest_asking_for_help (); | |
4fdb4553 | 1152 | /* NOTREACHED */ |
97b90b0a | 1153 | } |
c6d46f5f | 1154 | |
55102b5d | 1155 | /* No more options. Store the rest of arguments. */ |
4fdb4553 | 1156 | for (; optind < argc; optind++) |
b9755a12 FP |
1157 | { |
1158 | argbuffer[current_arg].arg_type = at_filename; | |
1159 | argbuffer[current_arg].what = argv[optind]; | |
9250f758 PE |
1160 | len = strlen (argv[optind]); |
1161 | if (whatlen_max < len) | |
1162 | whatlen_max = len; | |
b9755a12 FP |
1163 | ++current_arg; |
1164 | ++file_count; | |
1165 | } | |
1166 | ||
4fdb4553 FP |
1167 | argbuffer[current_arg].arg_type = at_end; |
1168 | ||
1169 | if (help_asked) | |
1170 | print_help (argbuffer); | |
1171 | /* NOTREACHED */ | |
1172 | ||
b9755a12 | 1173 | if (nincluded_files == 0 && file_count == 0) |
c6d46f5f | 1174 | { |
db5a3003 | 1175 | error ("no input files specified."); |
d8913c1c | 1176 | suggest_asking_for_help (); |
4fdb4553 | 1177 | /* NOTREACHED */ |
c6d46f5f JB |
1178 | } |
1179 | ||
6dd5561c | 1180 | if (tagfile == NULL) |
d3706fa9 | 1181 | tagfile = savestr (CTAGS ? "tags" : "TAGS"); |
b02c5fea | 1182 | cwd = etags_getcwd (); /* the current working directory */ |
f55aa55a | 1183 | if (cwd[strlen (cwd) - 1] != '/') |
cec68fcb FP |
1184 | { |
1185 | char *oldcwd = cwd; | |
1186 | cwd = concat (oldcwd, "/", ""); | |
1187 | free (oldcwd); | |
1188 | } | |
d3706fa9 FP |
1189 | |
1190 | /* Compute base directory for relative file names. */ | |
91702d63 FP |
1191 | if (streq (tagfile, "-") |
1192 | || strneq (tagfile, "/dev/", 5)) | |
d3706fa9 | 1193 | tagfiledir = cwd; /* relative file names are relative to cwd */ |
46c145db | 1194 | else |
9817cf3f FP |
1195 | { |
1196 | canonicalize_filename (tagfile); | |
1197 | tagfiledir = absolute_dirname (tagfile, cwd); | |
1198 | } | |
c6d46f5f | 1199 | |
b9755a12 | 1200 | init (); /* set up boolean "functions" */ |
c6d46f5f | 1201 | |
61a1f6fa FP |
1202 | linebuffer_init (&lb); |
1203 | linebuffer_init (&filename_lb); | |
1204 | linebuffer_init (&filebuf); | |
1205 | linebuffer_init (&token_name); | |
b9755a12 | 1206 | |
32daa216 | 1207 | if (!CTAGS) |
c6d46f5f | 1208 | { |
6dd5561c | 1209 | if (streq (tagfile, "-")) |
ce7f6d62 RS |
1210 | { |
1211 | tagf = stdout; | |
1212 | #ifdef DOS_NT | |
1213 | /* Switch redirected `stdout' to binary mode (setting `_fmode' | |
cf347d3c | 1214 | doesn't take effect until after `stdout' is already open). */ |
ce7f6d62 RS |
1215 | if (!isatty (fileno (stdout))) |
1216 | setmode (fileno (stdout), O_BINARY); | |
1217 | #endif /* DOS_NT */ | |
1218 | } | |
c6d46f5f | 1219 | else |
6dd5561c FP |
1220 | tagf = fopen (tagfile, append_to_tagfile ? "a" : "w"); |
1221 | if (tagf == NULL) | |
cdc1f6a7 | 1222 | pfatal (tagfile); |
c6d46f5f JB |
1223 | } |
1224 | ||
b9755a12 FP |
1225 | /* |
1226 | * Loop through files finding functions. | |
1227 | */ | |
4fdb4553 | 1228 | for (i = 0; i < current_arg; i++) |
c6d46f5f | 1229 | { |
9e0a3f98 FP |
1230 | static language *lang; /* non-NULL if language is forced */ |
1231 | char *this_file; | |
1232 | ||
b9755a12 | 1233 | switch (argbuffer[i].arg_type) |
c6d46f5f | 1234 | { |
b9755a12 | 1235 | case at_language: |
9e0a3f98 | 1236 | lang = argbuffer[i].lang; |
b9755a12 | 1237 | break; |
b9755a12 | 1238 | case at_regexp: |
24dbe96a | 1239 | analyse_regex (argbuffer[i].what); |
b9755a12 | 1240 | break; |
b9755a12 | 1241 | case at_filename: |
b9755a12 | 1242 | this_file = argbuffer[i].what; |
b9755a12 | 1243 | /* Input file named "-" means read file names from stdin |
93b7ac65 | 1244 | (one per line) and use them. */ |
b9755a12 | 1245 | if (streq (this_file, "-")) |
97b90b0a FP |
1246 | { |
1247 | if (parsing_stdin) | |
1248 | fatal ("cannot parse standard input AND read file names from it", | |
1249 | (char *)NULL); | |
1250 | while (readline_internal (&filename_lb, stdin) > 0) | |
1251 | process_file_name (filename_lb.buffer, lang); | |
1252 | } | |
b9755a12 | 1253 | else |
97b90b0a | 1254 | process_file_name (this_file, lang); |
b9755a12 | 1255 | break; |
97b90b0a FP |
1256 | case at_stdin: |
1257 | this_file = argbuffer[i].what; | |
1258 | process_file (stdin, this_file, lang); | |
1259 | break; | |
c6d46f5f | 1260 | } |
46c145db | 1261 | } |
9cb0aa73 | 1262 | |
61a1f6fa | 1263 | free_regexps (); |
61a1f6fa | 1264 | free (lb.buffer); |
24dbe96a | 1265 | free (filebuf.buffer); |
61a1f6fa | 1266 | free (token_name.buffer); |
93b7ac65 | 1267 | |
db590582 | 1268 | if (!CTAGS || cxref_style) |
c6d46f5f | 1269 | { |
2e0bea68 FP |
1270 | /* Write the remaining tags to tagf (ETAGS) or stdout (CXREF). */ |
1271 | put_entries (nodehead); | |
9e0a3f98 FP |
1272 | free_tree (nodehead); |
1273 | nodehead = NULL; | |
db590582 | 1274 | if (!CTAGS) |
61a1f6fa FP |
1275 | { |
1276 | fdesc *fdp; | |
1277 | ||
1278 | /* Output file entries that have no tags. */ | |
1279 | for (fdp = fdhead; fdp != NULL; fdp = fdp->next) | |
1280 | if (!fdp->written) | |
1281 | fprintf (tagf, "\f\n%s,0\n", fdp->taggedfname); | |
1282 | ||
1283 | while (nincluded_files-- > 0) | |
1284 | fprintf (tagf, "\f\n%s,include\n", *included_files++); | |
2e0bea68 FP |
1285 | |
1286 | if (fclose (tagf) == EOF) | |
1287 | pfatal (tagfile); | |
61a1f6fa | 1288 | } |
db590582 | 1289 | |
3f0656ff | 1290 | exit (EXIT_SUCCESS); |
c6d46f5f | 1291 | } |
55597f90 | 1292 | |
89d8309f | 1293 | /* From here on, we are in (CTAGS && !cxref_style) */ |
4746118a | 1294 | if (update) |
c6d46f5f | 1295 | { |
9250f758 PE |
1296 | char *cmd = |
1297 | xmalloc (strlen (tagfile) + whatlen_max + | |
1298 | sizeof "mv..OTAGS;fgrep -v '\t\t' OTAGS >;rm OTAGS"); | |
b9755a12 | 1299 | for (i = 0; i < current_arg; ++i) |
c6d46f5f | 1300 | { |
97b90b0a FP |
1301 | switch (argbuffer[i].arg_type) |
1302 | { | |
1303 | case at_filename: | |
1304 | case at_stdin: | |
1305 | break; | |
1306 | default: | |
1307 | continue; /* the for loop */ | |
1308 | } | |
9250f758 PE |
1309 | strcpy (cmd, "mv "); |
1310 | strcat (cmd, tagfile); | |
1311 | strcat (cmd, " OTAGS;fgrep -v '\t"); | |
1312 | strcat (cmd, argbuffer[i].what); | |
1313 | strcat (cmd, "\t' OTAGS >"); | |
1314 | strcat (cmd, tagfile); | |
1315 | strcat (cmd, ";rm OTAGS"); | |
3f0656ff | 1316 | if (system (cmd) != EXIT_SUCCESS) |
086eac13 | 1317 | fatal ("failed to execute shell command", (char *)NULL); |
c6d46f5f | 1318 | } |
9250f758 | 1319 | free (cmd); |
55597f90 | 1320 | append_to_tagfile = TRUE; |
c6d46f5f | 1321 | } |
55597f90 | 1322 | |
6dd5561c FP |
1323 | tagf = fopen (tagfile, append_to_tagfile ? "a" : "w"); |
1324 | if (tagf == NULL) | |
55597f90 | 1325 | pfatal (tagfile); |
89fb2be1 | 1326 | put_entries (nodehead); /* write all the tags (CTAGS) */ |
9e0a3f98 FP |
1327 | free_tree (nodehead); |
1328 | nodehead = NULL; | |
db590582 FP |
1329 | if (fclose (tagf) == EOF) |
1330 | pfatal (tagfile); | |
55597f90 | 1331 | |
b9509712 FP |
1332 | if (CTAGS) |
1333 | if (append_to_tagfile || update) | |
1334 | { | |
9250f758 | 1335 | char *cmd = xmalloc (2 * strlen (tagfile) + sizeof "sort -u -o.."); |
660222c7 FP |
1336 | /* Maybe these should be used: |
1337 | setenv ("LC_COLLATE", "C", 1); | |
1338 | setenv ("LC_ALL", "C", 1); */ | |
9250f758 PE |
1339 | strcpy (cmd, "sort -u -o "); |
1340 | strcat (cmd, tagfile); | |
1341 | strcat (cmd, " "); | |
1342 | strcat (cmd, tagfile); | |
b9509712 FP |
1343 | exit (system (cmd)); |
1344 | } | |
3f0656ff | 1345 | return EXIT_SUCCESS; |
c6d46f5f JB |
1346 | } |
1347 | ||
1348 | ||
93b7ac65 FP |
1349 | /* |
1350 | * Return a compressor given the file name. If EXTPTR is non-zero, | |
1351 | * return a pointer into FILE where the compressor-specific | |
1352 | * extension begins. If no compressor is found, NULL is returned | |
1353 | * and EXTPTR is not significant. | |
b2521b0a | 1354 | * Idea by Vladimir Alexiev <vladimir@cs.ualberta.ca> (1998) |
93b7ac65 | 1355 | */ |
944bdd72 | 1356 | static compressor * |
873fbd0b | 1357 | get_compressor_from_suffix (char *file, char **extptr) |
93b7ac65 FP |
1358 | { |
1359 | compressor *compr; | |
1360 | char *slash, *suffix; | |
1361 | ||
9817cf3f | 1362 | /* File has been processed by canonicalize_filename, |
93b7ac65 FP |
1363 | so we don't need to consider backslashes on DOS_NT. */ |
1364 | slash = etags_strrchr (file, '/'); | |
1365 | suffix = etags_strrchr (file, '.'); | |
1366 | if (suffix == NULL || suffix < slash) | |
1367 | return NULL; | |
1368 | if (extptr != NULL) | |
1369 | *extptr = suffix; | |
1370 | suffix += 1; | |
1371 | /* Let those poor souls who live with DOS 8+3 file name limits get | |
1372 | some solace by treating foo.cgz as if it were foo.c.gz, etc. | |
1373 | Only the first do loop is run if not MSDOS */ | |
1374 | do | |
1375 | { | |
1376 | for (compr = compressors; compr->suffix != NULL; compr++) | |
1377 | if (streq (compr->suffix, suffix)) | |
1378 | return compr; | |
944bdd72 FP |
1379 | if (!MSDOS) |
1380 | break; /* do it only once: not really a loop */ | |
93b7ac65 FP |
1381 | if (extptr != NULL) |
1382 | *extptr = ++suffix; | |
1383 | } while (*suffix != '\0'); | |
1384 | return NULL; | |
1385 | } | |
1386 | ||
1387 | ||
1388 | ||
b9755a12 | 1389 | /* |
93b7ac65 | 1390 | * Return a language given the name. |
b9755a12 | 1391 | */ |
944bdd72 | 1392 | static language * |
873fbd0b | 1393 | get_language_from_langname (const char *name) |
b9755a12 | 1394 | { |
93b7ac65 | 1395 | language *lang; |
b9755a12 | 1396 | |
93b7ac65 | 1397 | if (name == NULL) |
db5a3003 | 1398 | error ("empty language name"); |
93b7ac65 FP |
1399 | else |
1400 | { | |
1401 | for (lang = lang_names; lang->name != NULL; lang++) | |
d8913c1c | 1402 | if (streq (name, lang->name)) |
93b7ac65 FP |
1403 | return lang; |
1404 | error ("unknown language \"%s\"", name); | |
1405 | } | |
d8913c1c | 1406 | |
93b7ac65 | 1407 | return NULL; |
1f638249 FP |
1408 | } |
1409 | ||
1410 | ||
1411 | /* | |
93b7ac65 | 1412 | * Return a language given the interpreter name. |
1f638249 | 1413 | */ |
944bdd72 | 1414 | static language * |
873fbd0b | 1415 | get_language_from_interpreter (char *interpreter) |
1f638249 | 1416 | { |
93b7ac65 | 1417 | language *lang; |
988e88ab | 1418 | const char **iname; |
1f638249 FP |
1419 | |
1420 | if (interpreter == NULL) | |
1421 | return NULL; | |
1422 | for (lang = lang_names; lang->name != NULL; lang++) | |
1423 | if (lang->interpreters != NULL) | |
1424 | for (iname = lang->interpreters; *iname != NULL; iname++) | |
1425 | if (streq (*iname, interpreter)) | |
93b7ac65 | 1426 | return lang; |
1f638249 FP |
1427 | |
1428 | return NULL; | |
1429 | } | |
1430 | ||
1431 | ||
1432 | ||
1433 | /* | |
93b7ac65 | 1434 | * Return a language given the file name. |
1f638249 | 1435 | */ |
944bdd72 | 1436 | static language * |
873fbd0b | 1437 | get_language_from_filename (char *file, int case_sensitive) |
1f638249 | 1438 | { |
93b7ac65 | 1439 | language *lang; |
988e88ab | 1440 | const char **name, **ext, *suffix; |
89d57763 FP |
1441 | |
1442 | /* Try whole file name first. */ | |
1443 | for (lang = lang_names; lang->name != NULL; lang++) | |
1444 | if (lang->filenames != NULL) | |
1445 | for (name = lang->filenames; *name != NULL; name++) | |
a5363890 FP |
1446 | if ((case_sensitive) |
1447 | ? streq (*name, file) | |
1448 | : strcaseeq (*name, file)) | |
89d57763 | 1449 | return lang; |
1f638249 | 1450 | |
89d57763 | 1451 | /* If not found, try suffix after last dot. */ |
93b7ac65 | 1452 | suffix = etags_strrchr (file, '.'); |
1f638249 FP |
1453 | if (suffix == NULL) |
1454 | return NULL; | |
93b7ac65 | 1455 | suffix += 1; |
1f638249 FP |
1456 | for (lang = lang_names; lang->name != NULL; lang++) |
1457 | if (lang->suffixes != NULL) | |
1458 | for (ext = lang->suffixes; *ext != NULL; ext++) | |
a5363890 FP |
1459 | if ((case_sensitive) |
1460 | ? streq (*ext, suffix) | |
1461 | : strcaseeq (*ext, suffix)) | |
93b7ac65 | 1462 | return lang; |
1f638249 | 1463 | return NULL; |
b9755a12 FP |
1464 | } |
1465 | ||
9e0a3f98 | 1466 | \f |
c6d46f5f JB |
1467 | /* |
1468 | * This routine is called on each file argument. | |
1469 | */ | |
944bdd72 | 1470 | static void |
873fbd0b | 1471 | process_file_name (char *file, language *lang) |
c6d46f5f JB |
1472 | { |
1473 | struct stat stat_buf; | |
55597f90 | 1474 | FILE *inf; |
9e0a3f98 | 1475 | fdesc *fdp; |
93b7ac65 FP |
1476 | compressor *compr; |
1477 | char *compressed_name, *uncompressed_name; | |
1478 | char *ext, *real_name; | |
db590582 | 1479 | int retval; |
ce7f6d62 | 1480 | |
93b7ac65 | 1481 | canonicalize_filename (file); |
6dd5561c | 1482 | if (streq (file, tagfile) && !streq (tagfile, "-")) |
c6d46f5f | 1483 | { |
30903246 | 1484 | error ("skipping inclusion of %s in self.", file); |
c6d46f5f JB |
1485 | return; |
1486 | } | |
93b7ac65 FP |
1487 | if ((compr = get_compressor_from_suffix (file, &ext)) == NULL) |
1488 | { | |
1489 | compressed_name = NULL; | |
1490 | real_name = uncompressed_name = savestr (file); | |
1491 | } | |
1492 | else | |
1493 | { | |
1494 | real_name = compressed_name = savestr (file); | |
1495 | uncompressed_name = savenstr (file, ext - file); | |
1496 | } | |
1497 | ||
9e0a3f98 FP |
1498 | /* If the canonicalized uncompressed name |
1499 | has already been dealt with, skip it silently. */ | |
1500 | for (fdp = fdhead; fdp != NULL; fdp = fdp->next) | |
93b7ac65 | 1501 | { |
9e0a3f98 FP |
1502 | assert (fdp->infname != NULL); |
1503 | if (streq (uncompressed_name, fdp->infname)) | |
1504 | goto cleanup; | |
1505 | } | |
1506 | ||
93b7ac65 FP |
1507 | if (stat (real_name, &stat_buf) != 0) |
1508 | { | |
1509 | /* Reset real_name and try with a different name. */ | |
1510 | real_name = NULL; | |
1511 | if (compressed_name != NULL) /* try with the given suffix */ | |
1512 | { | |
1513 | if (stat (uncompressed_name, &stat_buf) == 0) | |
1514 | real_name = uncompressed_name; | |
1515 | } | |
1516 | else /* try all possible suffixes */ | |
1517 | { | |
1518 | for (compr = compressors; compr->suffix != NULL; compr++) | |
1519 | { | |
1520 | compressed_name = concat (file, ".", compr->suffix); | |
1521 | if (stat (compressed_name, &stat_buf) != 0) | |
1522 | { | |
944bdd72 | 1523 | if (MSDOS) |
93b7ac65 | 1524 | { |
944bdd72 FP |
1525 | char *suf = compressed_name + strlen (file); |
1526 | size_t suflen = strlen (compr->suffix) + 1; | |
1527 | for ( ; suf[1]; suf++, suflen--) | |
93b7ac65 | 1528 | { |
944bdd72 FP |
1529 | memmove (suf, suf + 1, suflen); |
1530 | if (stat (compressed_name, &stat_buf) == 0) | |
1531 | { | |
1532 | real_name = compressed_name; | |
1533 | break; | |
1534 | } | |
93b7ac65 | 1535 | } |
944bdd72 FP |
1536 | if (real_name != NULL) |
1537 | break; | |
1538 | } /* MSDOS */ | |
93b7ac65 FP |
1539 | free (compressed_name); |
1540 | compressed_name = NULL; | |
1541 | } | |
1542 | else | |
1543 | { | |
1544 | real_name = compressed_name; | |
1545 | break; | |
1546 | } | |
1547 | } | |
1548 | } | |
1549 | if (real_name == NULL) | |
1550 | { | |
1551 | perror (file); | |
9e0a3f98 | 1552 | goto cleanup; |
93b7ac65 FP |
1553 | } |
1554 | } /* try with a different name */ | |
1555 | ||
1556 | if (!S_ISREG (stat_buf.st_mode)) | |
1557 | { | |
1558 | error ("skipping %s: it is not a regular file.", real_name); | |
9e0a3f98 | 1559 | goto cleanup; |
93b7ac65 FP |
1560 | } |
1561 | if (real_name == compressed_name) | |
1562 | { | |
1563 | char *cmd = concat (compr->command, " ", real_name); | |
8c463abe | 1564 | inf = (FILE *) popen (cmd, "r"); |
93b7ac65 FP |
1565 | free (cmd); |
1566 | } | |
1567 | else | |
1568 | inf = fopen (real_name, "r"); | |
55597f90 | 1569 | if (inf == NULL) |
42680d3c | 1570 | { |
93b7ac65 | 1571 | perror (real_name); |
9e0a3f98 | 1572 | goto cleanup; |
42680d3c | 1573 | } |
55597f90 | 1574 | |
97b90b0a FP |
1575 | process_file (inf, uncompressed_name, lang); |
1576 | ||
1577 | if (real_name == compressed_name) | |
1578 | retval = pclose (inf); | |
1579 | else | |
1580 | retval = fclose (inf); | |
1581 | if (retval < 0) | |
1582 | pfatal (file); | |
1583 | ||
1584 | cleanup: | |
c2cd06e6 JM |
1585 | free (compressed_name); |
1586 | free (uncompressed_name); | |
97b90b0a FP |
1587 | last_node = NULL; |
1588 | curfdp = NULL; | |
1589 | return; | |
1590 | } | |
1591 | ||
1592 | static void | |
873fbd0b | 1593 | process_file (FILE *fh, char *fn, language *lang) |
97b90b0a FP |
1594 | { |
1595 | static const fdesc emptyfdesc; | |
1596 | fdesc *fdp; | |
1597 | ||
a5363890 FP |
1598 | /* Create a new input file description entry. */ |
1599 | fdp = xnew (1, fdesc); | |
1600 | *fdp = emptyfdesc; | |
1601 | fdp->next = fdhead; | |
97b90b0a | 1602 | fdp->infname = savestr (fn); |
a5363890 | 1603 | fdp->lang = lang; |
97b90b0a FP |
1604 | fdp->infabsname = absolute_filename (fn, cwd); |
1605 | fdp->infabsdir = absolute_dirname (fn, cwd); | |
1606 | if (filename_is_absolute (fn)) | |
db590582 | 1607 | { |
97b90b0a FP |
1608 | /* An absolute file name. Canonicalize it. */ |
1609 | fdp->taggedfname = absolute_filename (fn, NULL); | |
db590582 FP |
1610 | } |
1611 | else | |
1612 | { | |
97b90b0a | 1613 | /* A file name relative to cwd. Make it relative |
db590582 | 1614 | to the directory of the tags file. */ |
97b90b0a | 1615 | fdp->taggedfname = relative_filename (fn, tagfiledir); |
db590582 | 1616 | } |
a5363890 FP |
1617 | fdp->usecharno = TRUE; /* use char position when making tags */ |
1618 | fdp->prop = NULL; | |
61a1f6fa | 1619 | fdp->written = FALSE; /* not written on tags file yet */ |
9e0a3f98 | 1620 | |
a5363890 | 1621 | fdhead = fdp; |
9e0a3f98 FP |
1622 | curfdp = fdhead; /* the current file description */ |
1623 | ||
97b90b0a | 1624 | find_entries (fh); |
93b7ac65 | 1625 | |
2431364f | 1626 | /* If not Ctags, and if this is not metasource and if it contained no #line |
8378bcd3 FP |
1627 | directives, we can write the tags and free all nodes pointing to |
1628 | curfdp. */ | |
2431364f | 1629 | if (!CTAGS |
8378bcd3 | 1630 | && curfdp->usecharno /* no #line directives in this file */ |
2431364f FP |
1631 | && !curfdp->lang->metasource) |
1632 | { | |
8378bcd3 FP |
1633 | node *np, *prev; |
1634 | ||
1635 | /* Look for the head of the sublist relative to this file. See add_node | |
1636 | for the structure of the node tree. */ | |
1637 | prev = NULL; | |
1638 | for (np = nodehead; np != NULL; prev = np, np = np->left) | |
1639 | if (np->fdp == curfdp) | |
1640 | break; | |
1641 | ||
1642 | /* If we generated tags for this file, write and delete them. */ | |
1643 | if (np != NULL) | |
1644 | { | |
1645 | /* This is the head of the last sublist, if any. The following | |
1646 | instructions depend on this being true. */ | |
1647 | assert (np->left == NULL); | |
1648 | ||
1649 | assert (fdhead == curfdp); | |
1650 | assert (last_node->fdp == curfdp); | |
1651 | put_entries (np); /* write tags for file curfdp->taggedfname */ | |
1652 | free_tree (np); /* remove the written nodes */ | |
1653 | if (prev == NULL) | |
1654 | nodehead = NULL; /* no nodes left */ | |
1655 | else | |
1656 | prev->left = NULL; /* delete the pointer to the sublist */ | |
1657 | } | |
2431364f | 1658 | } |
c6d46f5f JB |
1659 | } |
1660 | ||
1661 | /* | |
eb8c3be9 | 1662 | * This routine sets up the boolean pseudo-functions which work |
93b7ac65 | 1663 | * by setting boolean flags dependent upon the corresponding character. |
c6d46f5f JB |
1664 | * Every char which is NOT in that string is not a white char. Therefore, |
1665 | * all of the array "_wht" is set to FALSE, and then the elements | |
1666 | * subscripted by the chars in "white" are set to TRUE. Thus "_wht" | |
1667 | * of a char is TRUE if it is the string "white", else FALSE. | |
1668 | */ | |
944bdd72 | 1669 | static void |
873fbd0b | 1670 | init (void) |
c6d46f5f | 1671 | { |
988e88ab | 1672 | register const char *sp; |
13fde0cd | 1673 | register int i; |
c6d46f5f | 1674 | |
30903246 | 1675 | for (i = 0; i < CHARS; i++) |
5e617bc2 | 1676 | iswhite (i) = notinname (i) = begtoken (i) = intoken (i) = endtoken (i) = FALSE; |
93b7ac65 FP |
1677 | for (sp = white; *sp != '\0'; sp++) iswhite (*sp) = TRUE; |
1678 | for (sp = nonam; *sp != '\0'; sp++) notinname (*sp) = TRUE; | |
5e617bc2 | 1679 | notinname ('\0') = notinname ('\n'); |
71cbb895 | 1680 | for (sp = begtk; *sp != '\0'; sp++) begtoken (*sp) = TRUE; |
5e617bc2 | 1681 | begtoken ('\0') = begtoken ('\n'); |
71cbb895 | 1682 | for (sp = midtk; *sp != '\0'; sp++) intoken (*sp) = TRUE; |
5e617bc2 | 1683 | intoken ('\0') = intoken ('\n'); |
71cbb895 | 1684 | for (sp = endtk; *sp != '\0'; sp++) endtoken (*sp) = TRUE; |
5e617bc2 | 1685 | endtoken ('\0') = endtoken ('\n'); |
c6d46f5f JB |
1686 | } |
1687 | ||
1688 | /* | |
1689 | * This routine opens the specified file and calls the function | |
1690 | * which finds the function and type definitions. | |
1691 | */ | |
944bdd72 | 1692 | static void |
873fbd0b | 1693 | find_entries (FILE *inf) |
c6d46f5f | 1694 | { |
b9755a12 | 1695 | char *cp; |
9e0a3f98 FP |
1696 | language *lang = curfdp->lang; |
1697 | Lang_function *parser = NULL; | |
93b7ac65 | 1698 | |
b9755a12 | 1699 | /* If user specified a language, use it. */ |
93b7ac65 | 1700 | if (lang != NULL && lang->function != NULL) |
13fde0cd | 1701 | { |
9e0a3f98 | 1702 | parser = lang->function; |
13fde0cd | 1703 | } |
b9755a12 | 1704 | |
9e0a3f98 FP |
1705 | /* Else try to guess the language given the file name. */ |
1706 | if (parser == NULL) | |
1f638249 | 1707 | { |
a5363890 | 1708 | lang = get_language_from_filename (curfdp->infname, TRUE); |
9e0a3f98 FP |
1709 | if (lang != NULL && lang->function != NULL) |
1710 | { | |
1711 | curfdp->lang = lang; | |
1712 | parser = lang->function; | |
1713 | } | |
1f638249 FP |
1714 | } |
1715 | ||
9e0a3f98 FP |
1716 | /* Else look for sharp-bang as the first two characters. */ |
1717 | if (parser == NULL | |
1718 | && readline_internal (&lb, inf) > 0 | |
e0903a92 | 1719 | && lb.len >= 2 |
1f638249 FP |
1720 | && lb.buffer[0] == '#' |
1721 | && lb.buffer[1] == '!') | |
c6d46f5f | 1722 | { |
1f638249 FP |
1723 | char *lp; |
1724 | ||
1725 | /* Set lp to point at the first char after the last slash in the | |
1726 | line or, if no slashes, at the first nonblank. Then set cp to | |
1727 | the first successive blank and terminate the string. */ | |
1728 | lp = etags_strrchr (lb.buffer+2, '/'); | |
1729 | if (lp != NULL) | |
1730 | lp += 1; | |
1731 | else | |
93b7ac65 FP |
1732 | lp = skip_spaces (lb.buffer + 2); |
1733 | cp = skip_non_spaces (lp); | |
1f638249 FP |
1734 | *cp = '\0'; |
1735 | ||
1736 | if (strlen (lp) > 0) | |
b9755a12 | 1737 | { |
93b7ac65 FP |
1738 | lang = get_language_from_interpreter (lp); |
1739 | if (lang != NULL && lang->function != NULL) | |
b9755a12 | 1740 | { |
9e0a3f98 FP |
1741 | curfdp->lang = lang; |
1742 | parser = lang->function; | |
b9755a12 FP |
1743 | } |
1744 | } | |
c6d46f5f | 1745 | } |
9e0a3f98 | 1746 | |
a5363890 FP |
1747 | /* We rewind here, even if inf may be a pipe. We fail if the |
1748 | length of the first line is longer than the pipe block size, | |
1749 | which is unlikely. */ | |
24dbe96a | 1750 | rewind (inf); |
54ef70a2 | 1751 | |
a5363890 FP |
1752 | /* Else try to guess the language given the case insensitive file name. */ |
1753 | if (parser == NULL) | |
1754 | { | |
1755 | lang = get_language_from_filename (curfdp->infname, FALSE); | |
1756 | if (lang != NULL && lang->function != NULL) | |
1757 | { | |
1758 | curfdp->lang = lang; | |
1759 | parser = lang->function; | |
1760 | } | |
1761 | } | |
54ef70a2 | 1762 | |
24dbe96a FP |
1763 | /* Else try Fortran or C. */ |
1764 | if (parser == NULL) | |
1765 | { | |
1766 | node *old_last_node = last_node; | |
1767 | ||
1768 | curfdp->lang = get_language_from_langname ("fortran"); | |
1769 | find_entries (inf); | |
1770 | ||
1771 | if (old_last_node == last_node) | |
1772 | /* No Fortran entries found. Try C. */ | |
1773 | { | |
1774 | /* We do not tag if rewind fails. | |
1775 | Only the file name will be recorded in the tags file. */ | |
1776 | rewind (inf); | |
1777 | curfdp->lang = get_language_from_langname (cplusplus ? "c++" : "c"); | |
1778 | find_entries (inf); | |
1779 | } | |
1780 | return; | |
1781 | } | |
1782 | ||
9e0a3f98 FP |
1783 | if (!no_line_directive |
1784 | && curfdp->lang != NULL && curfdp->lang->metasource) | |
a5363890 | 1785 | /* It may be that this is a bingo.y file, and we already parsed a bingo.c |
9e0a3f98 | 1786 | file, or anyway we parsed a file that is automatically generated from |
a5363890 | 1787 | this one. If this is the case, the bingo.c file contained #line |
9e0a3f98 FP |
1788 | directives that generated tags pointing to this file. Let's delete |
1789 | them all before parsing this file, which is the real source. */ | |
1790 | { | |
1791 | fdesc **fdpp = &fdhead; | |
1792 | while (*fdpp != NULL) | |
1793 | if (*fdpp != curfdp | |
1794 | && streq ((*fdpp)->taggedfname, curfdp->taggedfname)) | |
1795 | /* We found one of those! We must delete both the file description | |
1796 | and all tags referring to it. */ | |
1797 | { | |
1798 | fdesc *badfdp = *fdpp; | |
1799 | ||
9c485bbe FP |
1800 | /* Delete the tags referring to badfdp->taggedfname |
1801 | that were obtained from badfdp->infname. */ | |
8378bcd3 | 1802 | invalidate_nodes (badfdp, &nodehead); |
9e0a3f98 | 1803 | |
2431364f | 1804 | *fdpp = badfdp->next; /* remove the bad description from the list */ |
8378bcd3 | 1805 | free_fdesc (badfdp); |
9e0a3f98 FP |
1806 | } |
1807 | else | |
1808 | fdpp = &(*fdpp)->next; /* advance the list pointer */ | |
1809 | } | |
1810 | ||
24dbe96a | 1811 | assert (parser != NULL); |
e7d3b099 | 1812 | |
e1dbe924 | 1813 | /* Generic initializations before reading from file. */ |
61a1f6fa | 1814 | linebuffer_setlen (&filebuf, 0); /* reset the file buffer */ |
9e0a3f98 | 1815 | |
e1dbe924 | 1816 | /* Generic initializations before parsing file with readline. */ |
24dbe96a FP |
1817 | lineno = 0; /* reset global line number */ |
1818 | charno = 0; /* reset global char number */ | |
1819 | linecharno = 0; /* reset global char number of line start */ | |
c6d46f5f | 1820 | |
24dbe96a FP |
1821 | parser (inf); |
1822 | ||
24dbe96a | 1823 | regex_tag_multiline (); |
c6d46f5f | 1824 | } |
b2521b0a | 1825 | |
c6d46f5f | 1826 | \f |
b2521b0a | 1827 | /* |
2201e3dc FP |
1828 | * Check whether an implicitly named tag should be created, |
1829 | * then call `pfnote'. | |
1830 | * NAME is a string that is internally copied by this function. | |
1831 | * | |
b2521b0a FP |
1832 | * TAGS format specification |
1833 | * Idea by Sam Kendall <kendall@mv.mv.com> (1997) | |
081b7c70 | 1834 | * The following is explained in some more detail in etc/ETAGS.EBNF. |
30903246 | 1835 | * |
2201e3dc FP |
1836 | * make_tag creates tags with "implicit tag names" (unnamed tags) |
1837 | * if the following are all true, assuming NONAM=" \f\t\n\r()=,;": | |
1838 | * 1. NAME does not contain any of the characters in NONAM; | |
1839 | * 2. LINESTART contains name as either a rightmost, or rightmost but | |
30903246 | 1840 | * one character, substring; |
2201e3dc FP |
1841 | * 3. the character, if any, immediately before NAME in LINESTART must |
1842 | * be a character in NONAM; | |
1843 | * 4. the character, if any, immediately after NAME in LINESTART must | |
1844 | * also be a character in NONAM. | |
93b7ac65 | 1845 | * |
e1dbe924 | 1846 | * The implementation uses the notinname() macro, which recognizes the |
2201e3dc FP |
1847 | * characters stored in the string `nonam'. |
1848 | * etags.el needs to use the same characters that are in NONAM. | |
30903246 | 1849 | */ |
0dacfc4b | 1850 | static void |
988e88ab J |
1851 | make_tag (const char *name, /* tag name, or NULL if unnamed */ |
1852 | int namelen, /* tag length */ | |
1853 | int is_func, /* tag is a function */ | |
1854 | char *linestart, /* start of the line where tag is */ | |
1855 | int linelen, /* length of the line where tag is */ | |
1856 | int lno, /* line number */ | |
1857 | long int cno) /* character number */ | |
30903246 | 1858 | { |
61a1f6fa | 1859 | bool named = (name != NULL && namelen > 0); |
988e88ab | 1860 | char *nname = NULL; |
30903246 | 1861 | |
61a1f6fa FP |
1862 | if (!CTAGS && named) /* maybe set named to false */ |
1863 | /* Let's try to make an implicit tag name, that is, create an unnamed tag | |
1864 | such that etags.el can guess a name from it. */ | |
30903246 | 1865 | { |
74032cc3 | 1866 | int i; |
988e88ab | 1867 | register const char *cp = name; |
74032cc3 FP |
1868 | |
1869 | for (i = 0; i < namelen; i++) | |
1870 | if (notinname (*cp++)) | |
1871 | break; | |
1872 | if (i == namelen) /* rule #1 */ | |
30903246 FP |
1873 | { |
1874 | cp = linestart + linelen - namelen; | |
1875 | if (notinname (linestart[linelen-1])) | |
1876 | cp -= 1; /* rule #4 */ | |
1877 | if (cp >= linestart /* rule #2 */ | |
1878 | && (cp == linestart | |
1879 | || notinname (cp[-1])) /* rule #3 */ | |
1880 | && strneq (name, cp, namelen)) /* rule #2 */ | |
2201e3dc | 1881 | named = FALSE; /* use implicit tag name */ |
30903246 FP |
1882 | } |
1883 | } | |
93b7ac65 | 1884 | |
30903246 | 1885 | if (named) |
988e88ab J |
1886 | nname = savenstr (name, namelen); |
1887 | ||
1888 | pfnote (nname, is_func, linestart, linelen, lno, cno); | |
30903246 FP |
1889 | } |
1890 | ||
081b7c70 FP |
1891 | /* Record a tag. */ |
1892 | static void | |
873fbd0b DN |
1893 | pfnote (char *name, int is_func, char *linestart, int linelen, int lno, long int cno) |
1894 | /* tag name, or NULL if unnamed */ | |
1895 | /* tag is a function */ | |
1896 | /* start of the line where tag is */ | |
1897 | /* length of the line where tag is */ | |
1898 | /* line number */ | |
1899 | /* character number */ | |
081b7c70 FP |
1900 | { |
1901 | register node *np; | |
1902 | ||
a127d423 | 1903 | assert (name == NULL || name[0] != '\0'); |
081b7c70 FP |
1904 | if (CTAGS && name == NULL) |
1905 | return; | |
1906 | ||
1907 | np = xnew (1, node); | |
1908 | ||
1909 | /* If ctags mode, change name "main" to M<thisfilename>. */ | |
1910 | if (CTAGS && !cxref_style && streq (name, "main")) | |
1911 | { | |
1912 | register char *fp = etags_strrchr (curfdp->taggedfname, '/'); | |
1913 | np->name = concat ("M", fp == NULL ? curfdp->taggedfname : fp + 1, ""); | |
1914 | fp = etags_strrchr (np->name, '.'); | |
1915 | if (fp != NULL && fp[1] != '\0' && fp[2] == '\0') | |
1916 | fp[0] = '\0'; | |
1917 | } | |
1918 | else | |
1919 | np->name = name; | |
1920 | np->valid = TRUE; | |
1921 | np->been_warned = FALSE; | |
1922 | np->fdp = curfdp; | |
1923 | np->is_func = is_func; | |
1924 | np->lno = lno; | |
1925 | if (np->fdp->usecharno) | |
1926 | /* Our char numbers are 0-base, because of C language tradition? | |
1927 | ctags compatibility? old versions compatibility? I don't know. | |
1928 | Anyway, since emacs's are 1-base we expect etags.el to take care | |
1929 | of the difference. If we wanted to have 1-based numbers, we would | |
1930 | uncomment the +1 below. */ | |
1931 | np->cno = cno /* + 1 */ ; | |
1932 | else | |
1933 | np->cno = invalidcharno; | |
1934 | np->left = np->right = NULL; | |
1935 | if (CTAGS && !cxref_style) | |
1936 | { | |
1937 | if (strlen (linestart) < 50) | |
61a1f6fa | 1938 | np->regex = concat (linestart, "$", ""); |
081b7c70 | 1939 | else |
61a1f6fa | 1940 | np->regex = savenstr (linestart, 50); |
081b7c70 FP |
1941 | } |
1942 | else | |
61a1f6fa | 1943 | np->regex = savenstr (linestart, linelen); |
081b7c70 FP |
1944 | |
1945 | add_node (np, &nodehead); | |
1946 | } | |
1947 | ||
c6d46f5f JB |
1948 | /* |
1949 | * free_tree () | |
1950 | * recurse on left children, iterate on right children. | |
1951 | */ | |
944bdd72 | 1952 | static void |
873fbd0b | 1953 | free_tree (register node *np) |
c6d46f5f | 1954 | { |
93b7ac65 | 1955 | while (np) |
c6d46f5f | 1956 | { |
93b7ac65 FP |
1957 | register node *node_right = np->right; |
1958 | free_tree (np->left); | |
c2cd06e6 | 1959 | free (np->name); |
61a1f6fa | 1960 | free (np->regex); |
93b7ac65 FP |
1961 | free (np); |
1962 | np = node_right; | |
c6d46f5f JB |
1963 | } |
1964 | } | |
1965 | ||
8378bcd3 FP |
1966 | /* |
1967 | * free_fdesc () | |
1968 | * delete a file description | |
1969 | */ | |
1970 | static void | |
873fbd0b | 1971 | free_fdesc (register fdesc *fdp) |
8378bcd3 | 1972 | { |
c2cd06e6 JM |
1973 | free (fdp->infname); |
1974 | free (fdp->infabsname); | |
1975 | free (fdp->infabsdir); | |
1976 | free (fdp->taggedfname); | |
1977 | free (fdp->prop); | |
8378bcd3 FP |
1978 | free (fdp); |
1979 | } | |
1980 | ||
c6d46f5f JB |
1981 | /* |
1982 | * add_node () | |
db590582 FP |
1983 | * Adds a node to the tree of nodes. In etags mode, sort by file |
1984 | * name. In ctags mode, sort by tag name. Make no attempt at | |
1985 | * balancing. | |
c6d46f5f JB |
1986 | * |
1987 | * add_node is the only function allowed to add nodes, so it can | |
1988 | * maintain state. | |
1989 | */ | |
944bdd72 | 1990 | static void |
873fbd0b | 1991 | add_node (node *np, node **cur_node_p) |
c6d46f5f JB |
1992 | { |
1993 | register int dif; | |
93b7ac65 | 1994 | register node *cur_node = *cur_node_p; |
c6d46f5f JB |
1995 | |
1996 | if (cur_node == NULL) | |
1997 | { | |
93b7ac65 FP |
1998 | *cur_node_p = np; |
1999 | last_node = np; | |
c6d46f5f JB |
2000 | return; |
2001 | } | |
2002 | ||
32daa216 | 2003 | if (!CTAGS) |
8378bcd3 | 2004 | /* Etags Mode */ |
c6d46f5f | 2005 | { |
db590582 FP |
2006 | /* For each file name, tags are in a linked sublist on the right |
2007 | pointer. The first tags of different files are a linked list | |
2008 | on the left pointer. last_node points to the end of the last | |
2009 | used sublist. */ | |
8378bcd3 | 2010 | if (last_node != NULL && last_node->fdp == np->fdp) |
db590582 FP |
2011 | { |
2012 | /* Let's use the same sublist as the last added node. */ | |
8378bcd3 | 2013 | assert (last_node->right == NULL); |
db590582 FP |
2014 | last_node->right = np; |
2015 | last_node = np; | |
2016 | } | |
9e0a3f98 | 2017 | else if (cur_node->fdp == np->fdp) |
db590582 FP |
2018 | { |
2019 | /* Scanning the list we found the head of a sublist which is | |
2020 | good for us. Let's scan this sublist. */ | |
2021 | add_node (np, &cur_node->right); | |
2022 | } | |
2023 | else | |
2024 | /* The head of this sublist is not good for us. Let's try the | |
2025 | next one. */ | |
2026 | add_node (np, &cur_node->left); | |
8378bcd3 FP |
2027 | } /* if ETAGS mode */ |
2028 | ||
c6d46f5f JB |
2029 | else |
2030 | { | |
2031 | /* Ctags Mode */ | |
93b7ac65 | 2032 | dif = strcmp (np->name, cur_node->name); |
c6d46f5f JB |
2033 | |
2034 | /* | |
2035 | * If this tag name matches an existing one, then | |
2036 | * do not add the node, but maybe print a warning. | |
2037 | */ | |
ed8bbc0e | 2038 | if (no_duplicates && !dif) |
c6d46f5f | 2039 | { |
9e0a3f98 | 2040 | if (np->fdp == cur_node->fdp) |
c6d46f5f JB |
2041 | { |
2042 | if (!no_warnings) | |
2043 | { | |
2044 | fprintf (stderr, "Duplicate entry in file %s, line %d: %s\n", | |
9e0a3f98 | 2045 | np->fdp->infname, lineno, np->name); |
c6d46f5f JB |
2046 | fprintf (stderr, "Second entry ignored\n"); |
2047 | } | |
c6d46f5f | 2048 | } |
c5007f46 | 2049 | else if (!cur_node->been_warned && !no_warnings) |
c6d46f5f | 2050 | { |
c5007f46 FP |
2051 | fprintf |
2052 | (stderr, | |
2053 | "Duplicate entry in files %s and %s: %s (Warning only)\n", | |
9e0a3f98 | 2054 | np->fdp->infname, cur_node->fdp->infname, np->name); |
c5007f46 | 2055 | cur_node->been_warned = TRUE; |
c6d46f5f | 2056 | } |
c6d46f5f JB |
2057 | return; |
2058 | } | |
2059 | ||
c6d46f5f | 2060 | /* Actually add the node */ |
93b7ac65 | 2061 | add_node (np, dif < 0 ? &cur_node->left : &cur_node->right); |
8378bcd3 | 2062 | } /* if CTAGS mode */ |
c6d46f5f | 2063 | } |
b2521b0a | 2064 | |
9e0a3f98 FP |
2065 | /* |
2066 | * invalidate_nodes () | |
2067 | * Scan the node tree and invalidate all nodes pointing to the | |
8378bcd3 | 2068 | * given file description (CTAGS case) or free them (ETAGS case). |
9e0a3f98 FP |
2069 | */ |
2070 | static void | |
873fbd0b | 2071 | invalidate_nodes (fdesc *badfdp, node **npp) |
9e0a3f98 | 2072 | { |
8378bcd3 FP |
2073 | node *np = *npp; |
2074 | ||
2075 | if (np == NULL) | |
2076 | return; | |
2077 | ||
2078 | if (CTAGS) | |
2079 | { | |
2080 | if (np->left != NULL) | |
2081 | invalidate_nodes (badfdp, &np->left); | |
2082 | if (np->fdp == badfdp) | |
24dbe96a | 2083 | np->valid = FALSE; |
8378bcd3 FP |
2084 | if (np->right != NULL) |
2085 | invalidate_nodes (badfdp, &np->right); | |
2086 | } | |
2087 | else | |
2088 | { | |
89fb2be1 | 2089 | assert (np->fdp != NULL); |
8378bcd3 FP |
2090 | if (np->fdp == badfdp) |
2091 | { | |
89fb2be1 | 2092 | *npp = np->left; /* detach the sublist from the list */ |
8378bcd3 FP |
2093 | np->left = NULL; /* isolate it */ |
2094 | free_tree (np); /* free it */ | |
89fb2be1 | 2095 | invalidate_nodes (badfdp, npp); |
8378bcd3 | 2096 | } |
89fb2be1 FP |
2097 | else |
2098 | invalidate_nodes (badfdp, &np->left); | |
8378bcd3 | 2099 | } |
9e0a3f98 FP |
2100 | } |
2101 | ||
c6d46f5f | 2102 | \f |
f57e2426 | 2103 | static int total_size_of_entries (node *); |
5994c183 | 2104 | static int number_len (long) ATTRIBUTE_CONST; |
db590582 | 2105 | |
9e0a3f98 | 2106 | /* Length of a non-negative number's decimal representation. */ |
db590582 | 2107 | static int |
873fbd0b | 2108 | number_len (long int num) |
db590582 FP |
2109 | { |
2110 | int len = 1; | |
2111 | while ((num /= 10) > 0) | |
2112 | len += 1; | |
2113 | return len; | |
2114 | } | |
2115 | ||
2116 | /* | |
2117 | * Return total number of characters that put_entries will output for | |
2118 | * the nodes in the linked list at the right of the specified node. | |
2119 | * This count is irrelevant with etags.el since emacs 19.34 at least, | |
2120 | * but is still supplied for backward compatibility. | |
2121 | */ | |
2122 | static int | |
873fbd0b | 2123 | total_size_of_entries (register node *np) |
db590582 FP |
2124 | { |
2125 | register int total = 0; | |
2126 | ||
2127 | for (; np != NULL; np = np->right) | |
61a1f6fa FP |
2128 | if (np->valid) |
2129 | { | |
2130 | total += strlen (np->regex) + 1; /* pat\177 */ | |
2131 | if (np->name != NULL) | |
2132 | total += strlen (np->name) + 1; /* name\001 */ | |
2133 | total += number_len ((long) np->lno) + 1; /* lno, */ | |
2134 | if (np->cno != invalidcharno) /* cno */ | |
2135 | total += number_len (np->cno); | |
2136 | total += 1; /* newline */ | |
2137 | } | |
db590582 FP |
2138 | |
2139 | return total; | |
2140 | } | |
db590582 | 2141 | |
944bdd72 | 2142 | static void |
873fbd0b | 2143 | put_entries (register node *np) |
c6d46f5f | 2144 | { |
13fde0cd | 2145 | register char *sp; |
9e0a3f98 | 2146 | static fdesc *fdp = NULL; |
c6d46f5f | 2147 | |
93b7ac65 | 2148 | if (np == NULL) |
c6d46f5f JB |
2149 | return; |
2150 | ||
2151 | /* Output subentries that precede this one */ | |
db590582 FP |
2152 | if (CTAGS) |
2153 | put_entries (np->left); | |
c6d46f5f JB |
2154 | |
2155 | /* Output this entry */ | |
9e0a3f98 | 2156 | if (np->valid) |
c6d46f5f | 2157 | { |
9e0a3f98 | 2158 | if (!CTAGS) |
db590582 | 2159 | { |
9e0a3f98 FP |
2160 | /* Etags mode */ |
2161 | if (fdp != np->fdp) | |
2162 | { | |
2163 | fdp = np->fdp; | |
2164 | fprintf (tagf, "\f\n%s,%d\n", | |
2165 | fdp->taggedfname, total_size_of_entries (np)); | |
61a1f6fa | 2166 | fdp->written = TRUE; |
9e0a3f98 | 2167 | } |
61a1f6fa | 2168 | fputs (np->regex, tagf); |
9e0a3f98 FP |
2169 | fputc ('\177', tagf); |
2170 | if (np->name != NULL) | |
2171 | { | |
2172 | fputs (np->name, tagf); | |
2173 | fputc ('\001', tagf); | |
2174 | } | |
2175 | fprintf (tagf, "%d,", np->lno); | |
2176 | if (np->cno != invalidcharno) | |
2177 | fprintf (tagf, "%ld", np->cno); | |
2178 | fputs ("\n", tagf); | |
db590582 | 2179 | } |
c6d46f5f | 2180 | else |
d8913c1c | 2181 | { |
9e0a3f98 FP |
2182 | /* Ctags mode */ |
2183 | if (np->name == NULL) | |
db5a3003 | 2184 | error ("internal error: NULL name in ctags mode."); |
9e0a3f98 FP |
2185 | |
2186 | if (cxref_style) | |
2187 | { | |
2188 | if (vgrind_style) | |
2189 | fprintf (stdout, "%s %s %d\n", | |
2190 | np->name, np->fdp->taggedfname, (np->lno + 63) / 64); | |
2191 | else | |
2192 | fprintf (stdout, "%-16s %3d %-16s %s\n", | |
61a1f6fa | 2193 | np->name, np->lno, np->fdp->taggedfname, np->regex); |
9e0a3f98 | 2194 | } |
d8913c1c | 2195 | else |
9e0a3f98 FP |
2196 | { |
2197 | fprintf (tagf, "%s\t%s\t", np->name, np->fdp->taggedfname); | |
d8913c1c | 2198 | |
9e0a3f98 FP |
2199 | if (np->is_func) |
2200 | { /* function or #define macro with args */ | |
2201 | putc (searchar, tagf); | |
2202 | putc ('^', tagf); | |
d8913c1c | 2203 | |
61a1f6fa | 2204 | for (sp = np->regex; *sp; sp++) |
9e0a3f98 FP |
2205 | { |
2206 | if (*sp == '\\' || *sp == searchar) | |
2207 | putc ('\\', tagf); | |
2208 | putc (*sp, tagf); | |
2209 | } | |
2210 | putc (searchar, tagf); | |
d8913c1c | 2211 | } |
9e0a3f98 FP |
2212 | else |
2213 | { /* anything else; text pattern inadequate */ | |
2214 | fprintf (tagf, "%d", np->lno); | |
2215 | } | |
2216 | putc ('\n', tagf); | |
d8913c1c | 2217 | } |
c6d46f5f | 2218 | } |
9e0a3f98 | 2219 | } /* if this node contains a valid tag */ |
db590582 | 2220 | |
c6d46f5f | 2221 | /* Output subentries that follow this one */ |
93b7ac65 | 2222 | put_entries (np->right); |
db590582 FP |
2223 | if (!CTAGS) |
2224 | put_entries (np->left); | |
c6d46f5f | 2225 | } |
b2521b0a | 2226 | |
c6d46f5f | 2227 | \f |
b28e26be FP |
2228 | /* C extensions. */ |
2229 | #define C_EXT 0x00fff /* C extensions */ | |
2230 | #define C_PLAIN 0x00000 /* C */ | |
2231 | #define C_PLPL 0x00001 /* C++ */ | |
2232 | #define C_STAR 0x00003 /* C* */ | |
2233 | #define C_JAVA 0x00005 /* JAVA */ | |
2234 | #define C_AUTO 0x01000 /* C, but switch to C++ if `class' is met */ | |
2235 | #define YACC 0x10000 /* yacc file */ | |
2236 | ||
c6d46f5f JB |
2237 | /* |
2238 | * The C symbol tables. | |
2239 | */ | |
55597f90 FP |
2240 | enum sym_type |
2241 | { | |
30903246 FP |
2242 | st_none, |
2243 | st_C_objprot, st_C_objimpl, st_C_objend, | |
2244 | st_C_gnumacro, | |
05d9a399 | 2245 | st_C_ignore, st_C_attribute, |
30903246 | 2246 | st_C_javastruct, |
93b7ac65 | 2247 | st_C_operator, |
f55ae599 | 2248 | st_C_class, st_C_template, |
05d9a399 | 2249 | st_C_struct, st_C_extern, st_C_enum, st_C_define, st_C_typedef |
55597f90 | 2250 | }; |
c6d46f5f | 2251 | |
f57e2426 J |
2252 | static unsigned int hash (const char *, unsigned int); |
2253 | static struct C_stab_entry * in_word_set (const char *, unsigned int); | |
2254 | static enum sym_type C_symtype (char *, int, int); | |
71cbb895 | 2255 | |
42680d3c | 2256 | /* Feed stuff between (but not including) %[ and %] lines to: |
05d9a399 | 2257 | gperf -m 5 |
42680d3c | 2258 | %[ |
05d9a399 FP |
2259 | %compare-strncmp |
2260 | %enum | |
2261 | %struct-type | |
42680d3c FP |
2262 | struct C_stab_entry { char *name; int c_ext; enum sym_type type; } |
2263 | %% | |
05d9a399 FP |
2264 | if, 0, st_C_ignore |
2265 | for, 0, st_C_ignore | |
2266 | while, 0, st_C_ignore | |
2267 | switch, 0, st_C_ignore | |
2268 | return, 0, st_C_ignore | |
2269 | __attribute__, 0, st_C_attribute | |
36c51859 | 2270 | GTY, 0, st_C_attribute |
05d9a399 FP |
2271 | @interface, 0, st_C_objprot |
2272 | @protocol, 0, st_C_objprot | |
2273 | @implementation,0, st_C_objimpl | |
2274 | @end, 0, st_C_objend | |
47fd18c6 FP |
2275 | import, (C_JAVA & ~C_PLPL), st_C_ignore |
2276 | package, (C_JAVA & ~C_PLPL), st_C_ignore | |
05d9a399 | 2277 | friend, C_PLPL, st_C_ignore |
47fd18c6 FP |
2278 | extends, (C_JAVA & ~C_PLPL), st_C_javastruct |
2279 | implements, (C_JAVA & ~C_PLPL), st_C_javastruct | |
2280 | interface, (C_JAVA & ~C_PLPL), st_C_struct | |
05d9a399 FP |
2281 | class, 0, st_C_class |
2282 | namespace, C_PLPL, st_C_struct | |
2283 | domain, C_STAR, st_C_struct | |
2284 | union, 0, st_C_struct | |
2285 | struct, 0, st_C_struct | |
2286 | extern, 0, st_C_extern | |
2287 | enum, 0, st_C_enum | |
2288 | typedef, 0, st_C_typedef | |
2289 | define, 0, st_C_define | |
327891eb | 2290 | undef, 0, st_C_define |
05d9a399 FP |
2291 | operator, C_PLPL, st_C_operator |
2292 | template, 0, st_C_template | |
d8913c1c | 2293 | # DEFUN used in emacs, the next three used in glibc (SYSCALL only for mach). |
05d9a399 FP |
2294 | DEFUN, 0, st_C_gnumacro |
2295 | SYSCALL, 0, st_C_gnumacro | |
2296 | ENTRY, 0, st_C_gnumacro | |
2297 | PSEUDO, 0, st_C_gnumacro | |
d8913c1c FP |
2298 | # These are defined inside C functions, so currently they are not met. |
2299 | # EXFUN used in glibc, DEFVAR_* in emacs. | |
05d9a399 FP |
2300 | #EXFUN, 0, st_C_gnumacro |
2301 | #DEFVAR_, 0, st_C_gnumacro | |
42680d3c | 2302 | %] |
05d9a399 FP |
2303 | and replace lines between %< and %> with its output, then: |
2304 | - remove the #if characterset check | |
2305 | - make in_word_set static and not inline. */ | |
42680d3c | 2306 | /*%<*/ |
05d9a399 FP |
2307 | /* C code produced by gperf version 3.0.1 */ |
2308 | /* Command-line: gperf -m 5 */ | |
327891eb | 2309 | /* Computed positions: -k'2-3' */ |
42680d3c | 2310 | |
988e88ab | 2311 | struct C_stab_entry { const char *name; int c_ext; enum sym_type type; }; |
327891eb | 2312 | /* maximum key range = 33, duplicates = 0 */ |
42680d3c | 2313 | |
55d4c1b2 | 2314 | static inline unsigned int |
873fbd0b | 2315 | hash (register const char *str, register unsigned int len) |
42680d3c | 2316 | { |
93b7ac65 | 2317 | static unsigned char asso_values[] = |
42680d3c | 2318 | { |
327891eb FP |
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, | |
36c51859 CY |
2325 | 35, 35, 35, 35, 35, 35, 35, 35, 35, 3, |
2326 | 26, 35, 35, 35, 35, 35, 35, 35, 27, 35, | |
2327 | 35, 35, 35, 24, 0, 35, 35, 35, 35, 0, | |
327891eb FP |
2328 | 35, 35, 35, 35, 35, 1, 35, 16, 35, 6, |
2329 | 23, 0, 0, 35, 22, 0, 35, 35, 5, 0, | |
2330 | 0, 15, 1, 35, 6, 35, 8, 19, 35, 16, | |
2331 | 4, 5, 35, 35, 35, 35, 35, 35, 35, 35, | |
2332 | 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, | |
2333 | 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, | |
2334 | 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, | |
2335 | 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, | |
2336 | 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, | |
2337 | 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, | |
2338 | 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, | |
2339 | 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, | |
2340 | 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, | |
2341 | 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, | |
2342 | 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, | |
2343 | 35, 35, 35, 35, 35, 35, 35, 35, 35, 35, | |
2344 | 35, 35, 35, 35, 35, 35 | |
93b7ac65 | 2345 | }; |
327891eb FP |
2346 | register int hval = len; |
2347 | ||
2348 | switch (hval) | |
2349 | { | |
2350 | default: | |
2351 | hval += asso_values[(unsigned char)str[2]]; | |
2352 | /*FALLTHROUGH*/ | |
2353 | case 2: | |
2354 | hval += asso_values[(unsigned char)str[1]]; | |
2355 | break; | |
2356 | } | |
2357 | return hval; | |
42680d3c | 2358 | } |
c6d46f5f | 2359 | |
30526cc6 | 2360 | static struct C_stab_entry * |
873fbd0b | 2361 | in_word_set (register const char *str, register unsigned int len) |
42680d3c | 2362 | { |
05d9a399 FP |
2363 | enum |
2364 | { | |
36c51859 | 2365 | TOTAL_KEYWORDS = 33, |
05d9a399 FP |
2366 | MIN_WORD_LENGTH = 2, |
2367 | MAX_WORD_LENGTH = 15, | |
327891eb FP |
2368 | MIN_HASH_VALUE = 2, |
2369 | MAX_HASH_VALUE = 34 | |
05d9a399 FP |
2370 | }; |
2371 | ||
93b7ac65 | 2372 | static struct C_stab_entry wordlist[] = |
42680d3c | 2373 | { |
327891eb | 2374 | {""}, {""}, |
05d9a399 | 2375 | {"if", 0, st_C_ignore}, |
36c51859 | 2376 | {"GTY", 0, st_C_attribute}, |
05d9a399 | 2377 | {"@end", 0, st_C_objend}, |
327891eb | 2378 | {"union", 0, st_C_struct}, |
05d9a399 | 2379 | {"define", 0, st_C_define}, |
47fd18c6 | 2380 | {"import", (C_JAVA & ~C_PLPL), st_C_ignore}, |
05d9a399 | 2381 | {"template", 0, st_C_template}, |
327891eb FP |
2382 | {"operator", C_PLPL, st_C_operator}, |
2383 | {"@interface", 0, st_C_objprot}, | |
47fd18c6 | 2384 | {"implements", (C_JAVA & ~C_PLPL), st_C_javastruct}, |
327891eb | 2385 | {"friend", C_PLPL, st_C_ignore}, |
05d9a399 | 2386 | {"typedef", 0, st_C_typedef}, |
327891eb FP |
2387 | {"return", 0, st_C_ignore}, |
2388 | {"@implementation",0, st_C_objimpl}, | |
2389 | {"@protocol", 0, st_C_objprot}, | |
47fd18c6 | 2390 | {"interface", (C_JAVA & ~C_PLPL), st_C_struct}, |
327891eb | 2391 | {"extern", 0, st_C_extern}, |
47fd18c6 | 2392 | {"extends", (C_JAVA & ~C_PLPL), st_C_javastruct}, |
05d9a399 | 2393 | {"struct", 0, st_C_struct}, |
327891eb | 2394 | {"domain", C_STAR, st_C_struct}, |
05d9a399 | 2395 | {"switch", 0, st_C_ignore}, |
327891eb FP |
2396 | {"enum", 0, st_C_enum}, |
2397 | {"for", 0, st_C_ignore}, | |
2398 | {"namespace", C_PLPL, st_C_struct}, | |
05d9a399 | 2399 | {"class", 0, st_C_class}, |
327891eb FP |
2400 | {"while", 0, st_C_ignore}, |
2401 | {"undef", 0, st_C_define}, | |
47fd18c6 | 2402 | {"package", (C_JAVA & ~C_PLPL), st_C_ignore}, |
05d9a399 FP |
2403 | {"__attribute__", 0, st_C_attribute}, |
2404 | {"SYSCALL", 0, st_C_gnumacro}, | |
05d9a399 | 2405 | {"ENTRY", 0, st_C_gnumacro}, |
327891eb | 2406 | {"PSEUDO", 0, st_C_gnumacro}, |
05d9a399 | 2407 | {"DEFUN", 0, st_C_gnumacro} |
42680d3c FP |
2408 | }; |
2409 | ||
2410 | if (len <= MAX_WORD_LENGTH && len >= MIN_WORD_LENGTH) | |
2411 | { | |
2412 | register int key = hash (str, len); | |
2413 | ||
93b7ac65 | 2414 | if (key <= MAX_HASH_VALUE && key >= 0) |
42680d3c | 2415 | { |
93b7ac65 | 2416 | register const char *s = wordlist[key].name; |
42680d3c | 2417 | |
05d9a399 | 2418 | if (*str == *s && !strncmp (str + 1, s + 1, len - 1) && s[len] == '\0') |
42680d3c FP |
2419 | return &wordlist[key]; |
2420 | } | |
2421 | } | |
2422 | return 0; | |
c6d46f5f | 2423 | } |
42680d3c | 2424 | /*%>*/ |
c6d46f5f | 2425 | |
944bdd72 | 2426 | static enum sym_type |
873fbd0b | 2427 | C_symtype (char *str, int len, int c_ext) |
c6d46f5f | 2428 | { |
3f1c8fcd | 2429 | register struct C_stab_entry *se = in_word_set (str, len); |
c6d46f5f | 2430 | |
42680d3c FP |
2431 | if (se == NULL || (se->c_ext && !(c_ext & se->c_ext))) |
2432 | return st_none; | |
2433 | return se->type; | |
c6d46f5f | 2434 | } |
b2521b0a | 2435 | |
c6d46f5f | 2436 | \f |
05d9a399 FP |
2437 | /* |
2438 | * Ignoring __attribute__ ((list)) | |
2439 | */ | |
2440 | static bool inattribute; /* looking at an __attribute__ construct */ | |
2441 | ||
b2521b0a FP |
2442 | /* |
2443 | * C functions and variables are recognized using a simple | |
2444 | * finite automaton. fvdef is its state variable. | |
2445 | */ | |
db590582 | 2446 | static enum |
13fde0cd | 2447 | { |
30903246 | 2448 | fvnone, /* nothing seen */ |
b2521b0a FP |
2449 | fdefunkey, /* Emacs DEFUN keyword seen */ |
2450 | fdefunname, /* Emacs DEFUN name seen */ | |
93b7ac65 | 2451 | foperator, /* func: operator keyword seen (cplpl) */ |
30903246 FP |
2452 | fvnameseen, /* function or variable name seen */ |
2453 | fstartlist, /* func: just after open parenthesis */ | |
2454 | finlist, /* func: in parameter list */ | |
2455 | flistseen, /* func: after parameter list */ | |
2456 | fignore, /* func: before open brace */ | |
2457 | vignore /* var-like: ignore until ';' */ | |
2458 | } fvdef; | |
13fde0cd | 2459 | |
db590582 | 2460 | static bool fvextern; /* func or var: extern keyword seen; */ |
13fde0cd | 2461 | |
b2521b0a FP |
2462 | /* |
2463 | * typedefs are recognized using a simple finite automaton. | |
2464 | * typdef is its state variable. | |
2465 | */ | |
db590582 | 2466 | static enum |
13fde0cd | 2467 | { |
31d4b314 | 2468 | tnone, /* nothing seen */ |
93b7ac65 FP |
2469 | tkeyseen, /* typedef keyword seen */ |
2470 | ttypeseen, /* defined type seen */ | |
31d4b314 | 2471 | tinbody, /* inside typedef body */ |
46c145db FP |
2472 | tend, /* just before typedef tag */ |
2473 | tignore /* junk after typedef tag */ | |
d8913c1c | 2474 | } typdef; |
13fde0cd | 2475 | |
b2521b0a FP |
2476 | /* |
2477 | * struct-like structures (enum, struct and union) are recognized | |
2478 | * using another simple finite automaton. `structdef' is its state | |
2479 | * variable. | |
2480 | */ | |
db590582 | 2481 | static enum |
13fde0cd | 2482 | { |
8c463abe | 2483 | snone, /* nothing seen yet, |
05d9a399 | 2484 | or in struct body if bracelev > 0 */ |
13fde0cd RS |
2485 | skeyseen, /* struct-like keyword seen */ |
2486 | stagseen, /* struct-like tag seen */ | |
8c463abe | 2487 | scolonseen /* colon seen after struct-like tag */ |
d8913c1c | 2488 | } structdef; |
46c145db | 2489 | |
d8913c1c FP |
2490 | /* |
2491 | * When objdef is different from onone, objtag is the name of the class. | |
2492 | */ | |
988e88ab | 2493 | static const char *objtag = "<uninited>"; |
d8913c1c | 2494 | |
13fde0cd RS |
2495 | /* |
2496 | * Yet another little state machine to deal with preprocessor lines. | |
2497 | */ | |
db590582 | 2498 | static enum |
13fde0cd RS |
2499 | { |
2500 | dnone, /* nothing seen */ | |
2501 | dsharpseen, /* '#' seen as first char on line */ | |
2502 | ddefineseen, /* '#' and 'define' seen */ | |
46e4cb76 | 2503 | dignorerest /* ignore rest of line */ |
d8913c1c FP |
2504 | } definedef; |
2505 | ||
2506 | /* | |
2507 | * State machine for Objective C protocols and implementations. | |
b2521b0a | 2508 | * Idea by Tom R.Hageman <tom@basil.icce.rug.nl> (1995) |
d8913c1c | 2509 | */ |
db590582 | 2510 | static enum |
d8913c1c FP |
2511 | { |
2512 | onone, /* nothing seen */ | |
2513 | oprotocol, /* @interface or @protocol seen */ | |
2514 | oimplementation, /* @implementations seen */ | |
2515 | otagseen, /* class name seen */ | |
2516 | oparenseen, /* parenthesis before category seen */ | |
2517 | ocatseen, /* category name seen */ | |
2518 | oinbody, /* in @implementation body */ | |
2519 | omethodsign, /* in @implementation body, after +/- */ | |
2520 | omethodtag, /* after method name */ | |
2521 | omethodcolon, /* after method colon */ | |
2522 | omethodparm, /* after method parameter */ | |
ba1fe3a8 | 2523 | oignore /* wait for @end */ |
d8913c1c | 2524 | } objdef; |
13fde0cd | 2525 | |
30903246 FP |
2526 | |
2527 | /* | |
2528 | * Use this structure to keep info about the token read, and how it | |
2529 | * should be tagged. Used by the make_C_tag function to build a tag. | |
2530 | */ | |
db590582 | 2531 | static struct tok |
30903246 | 2532 | { |
9c485bbe FP |
2533 | char *line; /* string containing the token */ |
2534 | int offset; /* where the token starts in LINE */ | |
2535 | int length; /* token length */ | |
2536 | /* | |
2537 | The previous members can be used to pass strings around for generic | |
2538 | purposes. The following ones specifically refer to creating tags. In this | |
2539 | case the token contained here is the pattern that will be used to create a | |
2540 | tag. | |
2541 | */ | |
2542 | bool valid; /* do not create a tag; the token should be | |
2543 | invalidated whenever a state machine is | |
2544 | reset prematurely */ | |
2545 | bool named; /* create a named tag */ | |
2546 | int lineno; /* source line number of tag */ | |
2547 | long linepos; /* source char number of tag */ | |
b2521b0a | 2548 | } token; /* latest token read */ |
d8913c1c | 2549 | |
8c463abe FP |
2550 | /* |
2551 | * Variables and functions for dealing with nested structures. | |
2552 | * Idea by Mykola Dzyuba <mdzyuba@yahoo.com> (2001) | |
2553 | */ | |
f57e2426 J |
2554 | static void pushclass_above (int, char *, int); |
2555 | static void popclass_above (int); | |
988e88ab | 2556 | static void write_classname (linebuffer *, const char *qualifier); |
8c463abe | 2557 | |
db590582 | 2558 | static struct { |
8c463abe | 2559 | char **cname; /* nested class names */ |
05d9a399 | 2560 | int *bracelev; /* nested class brace level */ |
8c463abe FP |
2561 | int nl; /* class nesting level (elements used) */ |
2562 | int size; /* length of the array */ | |
2563 | } cstack; /* stack for nested declaration tags */ | |
2564 | /* Current struct nesting depth (namespace, class, struct, union, enum). */ | |
2565 | #define nestlev (cstack.nl) | |
3c04a71a | 2566 | /* After struct keyword or in struct body, not inside a nested function. */ |
8c463abe | 2567 | #define instruct (structdef == snone && nestlev > 0 \ |
05d9a399 | 2568 | && bracelev == cstack.bracelev[nestlev-1] + 1) |
8c463abe FP |
2569 | |
2570 | static void | |
873fbd0b | 2571 | pushclass_above (int bracelev, char *str, int len) |
8c463abe FP |
2572 | { |
2573 | int nl; | |
2574 | ||
05d9a399 | 2575 | popclass_above (bracelev); |
8c463abe FP |
2576 | nl = cstack.nl; |
2577 | if (nl >= cstack.size) | |
2578 | { | |
2579 | int size = cstack.size *= 2; | |
2580 | xrnew (cstack.cname, size, char *); | |
05d9a399 | 2581 | xrnew (cstack.bracelev, size, int); |
8c463abe | 2582 | } |
05d9a399 | 2583 | assert (nl == 0 || cstack.bracelev[nl-1] < bracelev); |
8c463abe | 2584 | cstack.cname[nl] = (str == NULL) ? NULL : savenstr (str, len); |
05d9a399 | 2585 | cstack.bracelev[nl] = bracelev; |
8c463abe FP |
2586 | cstack.nl = nl + 1; |
2587 | } | |
2588 | ||
2589 | static void | |
873fbd0b | 2590 | popclass_above (int bracelev) |
8c463abe FP |
2591 | { |
2592 | int nl; | |
2593 | ||
2594 | for (nl = cstack.nl - 1; | |
05d9a399 | 2595 | nl >= 0 && cstack.bracelev[nl] >= bracelev; |
8c463abe FP |
2596 | nl--) |
2597 | { | |
c2cd06e6 | 2598 | free (cstack.cname[nl]); |
8c463abe FP |
2599 | cstack.nl = nl; |
2600 | } | |
2601 | } | |
2602 | ||
2603 | static void | |
988e88ab | 2604 | write_classname (linebuffer *cn, const char *qualifier) |
8c463abe FP |
2605 | { |
2606 | int i, len; | |
2607 | int qlen = strlen (qualifier); | |
2608 | ||
2609 | if (cstack.nl == 0 || cstack.cname[0] == NULL) | |
2610 | { | |
2611 | len = 0; | |
2612 | cn->len = 0; | |
2613 | cn->buffer[0] = '\0'; | |
2614 | } | |
2615 | else | |
2616 | { | |
2617 | len = strlen (cstack.cname[0]); | |
2618 | linebuffer_setlen (cn, len); | |
2619 | strcpy (cn->buffer, cstack.cname[0]); | |
2620 | } | |
2621 | for (i = 1; i < cstack.nl; i++) | |
2622 | { | |
e99a530f | 2623 | char *s = cstack.cname[i]; |
8c463abe FP |
2624 | if (s == NULL) |
2625 | continue; | |
e99a530f PE |
2626 | linebuffer_setlen (cn, len + qlen + strlen (s)); |
2627 | len += sprintf (cn->buffer + len, "%s%s", qualifier, s); | |
8c463abe FP |
2628 | } |
2629 | } | |
2630 | ||
2631 | \f | |
f57e2426 J |
2632 | static bool consider_token (char *, int, int, int *, int, int, bool *); |
2633 | static void make_C_tag (bool); | |
71cbb895 | 2634 | |
6dd5561c FP |
2635 | /* |
2636 | * consider_token () | |
2637 | * checks to see if the current token is at the start of a | |
30903246 FP |
2638 | * function or variable, or corresponds to a typedef, or |
2639 | * is a struct/union/enum tag, or #define, or an enum constant. | |
6dd5561c | 2640 | * |
79f94cd0 | 2641 | * *IS_FUNC gets TRUE if the token is a function or #define macro |
8c463abe | 2642 | * with args. C_EXTP points to which language we are looking at. |
6dd5561c | 2643 | * |
6dd5561c | 2644 | * Globals |
30903246 | 2645 | * fvdef IN OUT |
6dd5561c FP |
2646 | * structdef IN OUT |
2647 | * definedef IN OUT | |
2648 | * typdef IN OUT | |
d8913c1c | 2649 | * objdef IN OUT |
6dd5561c FP |
2650 | */ |
2651 | ||
944bdd72 | 2652 | static bool |
873fbd0b DN |
2653 | consider_token (register char *str, register int len, register int c, int *c_extp, int bracelev, int parlev, int *is_func_or_var) |
2654 | /* IN: token pointer */ | |
2655 | /* IN: token length */ | |
2656 | /* IN: first char after the token */ | |
2657 | /* IN, OUT: C extensions mask */ | |
2658 | /* IN: brace level */ | |
2659 | /* IN: parenthesis level */ | |
2660 | /* OUT: function or variable found */ | |
6dd5561c | 2661 | { |
05d9a399 | 2662 | /* When structdef is stagseen, scolonseen, or snone with bracelev > 0, |
8c463abe | 2663 | structtype is the type of the preceding struct-like keyword, and |
05d9a399 | 2664 | structbracelev is the brace level where it has been seen. */ |
b2521b0a | 2665 | static enum sym_type structtype; |
05d9a399 | 2666 | static int structbracelev; |
b2521b0a FP |
2667 | static enum sym_type toktype; |
2668 | ||
2669 | ||
8c463abe | 2670 | toktype = C_symtype (str, len, *c_extp); |
6dd5561c FP |
2671 | |
2672 | /* | |
05d9a399 | 2673 | * Skip __attribute__ |
6dd5561c | 2674 | */ |
05d9a399 | 2675 | if (toktype == st_C_attribute) |
6dd5561c | 2676 | { |
05d9a399 | 2677 | inattribute = TRUE; |
cec68fcb | 2678 | return FALSE; |
05d9a399 FP |
2679 | } |
2680 | ||
2681 | /* | |
2682 | * Advance the definedef state machine. | |
2683 | */ | |
2684 | switch (definedef) | |
2685 | { | |
2686 | case dnone: | |
2687 | /* We're not on a preprocessor line. */ | |
2688 | if (toktype == st_C_gnumacro) | |
2689 | { | |
2690 | fvdef = fdefunkey; | |
2691 | return FALSE; | |
2692 | } | |
2693 | break; | |
2694 | case dsharpseen: | |
2695 | if (toktype == st_C_define) | |
2696 | { | |
2697 | definedef = ddefineseen; | |
2698 | } | |
2699 | else | |
2700 | { | |
2701 | definedef = dignorerest; | |
2702 | } | |
2703 | return FALSE; | |
2704 | case ddefineseen: | |
2705 | /* | |
2706 | * Make a tag for any macro, unless it is a constant | |
2707 | * and constantypedefs is FALSE. | |
2708 | */ | |
2709 | definedef = dignorerest; | |
2710 | *is_func_or_var = (c == '('); | |
2711 | if (!*is_func_or_var && !constantypedefs) | |
2712 | return FALSE; | |
2713 | else | |
2714 | return TRUE; | |
2715 | case dignorerest: | |
2716 | return FALSE; | |
2717 | default: | |
db5a3003 | 2718 | error ("internal error: definedef value."); |
05d9a399 FP |
2719 | } |
2720 | ||
2721 | /* | |
2722 | * Now typedefs | |
2723 | */ | |
2724 | switch (typdef) | |
2725 | { | |
2726 | case tnone: | |
2727 | if (toktype == st_C_typedef) | |
2728 | { | |
2729 | if (typedefs) | |
2730 | typdef = tkeyseen; | |
2731 | fvextern = FALSE; | |
2732 | fvdef = fvnone; | |
2733 | return FALSE; | |
2734 | } | |
2735 | break; | |
2736 | case tkeyseen: | |
2737 | switch (toktype) | |
2738 | { | |
2739 | case st_none: | |
2740 | case st_C_class: | |
2741 | case st_C_struct: | |
2742 | case st_C_enum: | |
2743 | typdef = ttypeseen; | |
2744 | } | |
2745 | break; | |
2746 | case ttypeseen: | |
2747 | if (structdef == snone && fvdef == fvnone) | |
2748 | { | |
2749 | fvdef = fvnameseen; | |
2750 | return TRUE; | |
2751 | } | |
2752 | break; | |
2753 | case tend: | |
2754 | switch (toktype) | |
2755 | { | |
2756 | case st_C_class: | |
2757 | case st_C_struct: | |
2758 | case st_C_enum: | |
2759 | return FALSE; | |
2760 | } | |
2761 | return TRUE; | |
2762 | } | |
2763 | ||
05d9a399 FP |
2764 | switch (toktype) |
2765 | { | |
2766 | case st_C_javastruct: | |
2767 | if (structdef == stagseen) | |
2768 | structdef = scolonseen; | |
2769 | return FALSE; | |
2770 | case st_C_template: | |
2771 | case st_C_class: | |
2772 | if ((*c_extp & C_AUTO) /* automatic detection of C++ language */ | |
2773 | && bracelev == 0 | |
2774 | && definedef == dnone && structdef == snone | |
2775 | && typdef == tnone && fvdef == fvnone) | |
2776 | *c_extp = (*c_extp | C_PLPL) & ~C_AUTO; | |
2777 | if (toktype == st_C_template) | |
2778 | break; | |
2779 | /* FALLTHRU */ | |
2780 | case st_C_struct: | |
2781 | case st_C_enum: | |
2782 | if (parlev == 0 | |
2783 | && fvdef != vignore | |
2784 | && (typdef == tkeyseen | |
2785 | || (typedefs_or_cplusplus && structdef == snone))) | |
2786 | { | |
2787 | structdef = skeyseen; | |
2788 | structtype = toktype; | |
2789 | structbracelev = bracelev; | |
2790 | if (fvdef == fvnameseen) | |
2791 | fvdef = fvnone; | |
2792 | } | |
2793 | return FALSE; | |
2794 | } | |
2795 | ||
2796 | if (structdef == skeyseen) | |
2797 | { | |
2798 | structdef = stagseen; | |
2799 | return TRUE; | |
2800 | } | |
2801 | ||
2802 | if (typdef != tnone) | |
2803 | definedef = dnone; | |
2804 | ||
2805 | /* Detect Objective C constructs. */ | |
2806 | switch (objdef) | |
2807 | { | |
2808 | case onone: | |
2809 | switch (toktype) | |
2810 | { | |
2811 | case st_C_objprot: | |
2812 | objdef = oprotocol; | |
2813 | return FALSE; | |
2814 | case st_C_objimpl: | |
2815 | objdef = oimplementation; | |
2816 | return FALSE; | |
2817 | } | |
2818 | break; | |
2819 | case oimplementation: | |
2820 | /* Save the class tag for functions or variables defined inside. */ | |
2821 | objtag = savenstr (str, len); | |
2822 | objdef = oinbody; | |
2823 | return FALSE; | |
2824 | case oprotocol: | |
2825 | /* Save the class tag for categories. */ | |
2826 | objtag = savenstr (str, len); | |
2827 | objdef = otagseen; | |
2828 | *is_func_or_var = TRUE; | |
2829 | return TRUE; | |
2830 | case oparenseen: | |
2831 | objdef = ocatseen; | |
2832 | *is_func_or_var = TRUE; | |
2833 | return TRUE; | |
2834 | case oinbody: | |
2835 | break; | |
2836 | case omethodsign: | |
2837 | if (parlev == 0) | |
2838 | { | |
2839 | fvdef = fvnone; | |
2840 | objdef = omethodtag; | |
2841 | linebuffer_setlen (&token_name, len); | |
e99a530f | 2842 | memcpy (token_name.buffer, str, len); |
05d9a399 FP |
2843 | token_name.buffer[len] = '\0'; |
2844 | return TRUE; | |
2845 | } | |
2846 | return FALSE; | |
2847 | case omethodcolon: | |
2848 | if (parlev == 0) | |
2849 | objdef = omethodparm; | |
2850 | return FALSE; | |
2851 | case omethodparm: | |
2852 | if (parlev == 0) | |
2853 | { | |
e99a530f | 2854 | int oldlen = token_name.len; |
05d9a399 FP |
2855 | fvdef = fvnone; |
2856 | objdef = omethodtag; | |
e99a530f PE |
2857 | linebuffer_setlen (&token_name, oldlen + len); |
2858 | memcpy (token_name.buffer + oldlen, str, len); | |
5bf64749 | 2859 | token_name.buffer[oldlen + len] = '\0'; |
05d9a399 FP |
2860 | return TRUE; |
2861 | } | |
2862 | return FALSE; | |
2863 | case oignore: | |
2864 | if (toktype == st_C_objend) | |
2865 | { | |
2866 | /* Memory leakage here: the string pointed by objtag is | |
2867 | never released, because many tests would be needed to | |
2868 | avoid breaking on incorrect input code. The amount of | |
2869 | memory leaked here is the sum of the lengths of the | |
2870 | class tags. | |
2871 | free (objtag); */ | |
2872 | objdef = onone; | |
2873 | } | |
2874 | return FALSE; | |
2875 | } | |
2876 | ||
2877 | /* A function, variable or enum constant? */ | |
2878 | switch (toktype) | |
2879 | { | |
2880 | case st_C_extern: | |
2881 | fvextern = TRUE; | |
2882 | switch (fvdef) | |
2883 | { | |
2884 | case finlist: | |
2885 | case flistseen: | |
2886 | case fignore: | |
2887 | case vignore: | |
2888 | break; | |
2889 | default: | |
2890 | fvdef = fvnone; | |
2891 | } | |
2892 | return FALSE; | |
2893 | case st_C_ignore: | |
2894 | fvextern = FALSE; | |
2895 | fvdef = vignore; | |
2896 | return FALSE; | |
2897 | case st_C_operator: | |
2898 | fvdef = foperator; | |
2899 | *is_func_or_var = TRUE; | |
2900 | return TRUE; | |
2901 | case st_none: | |
2902 | if (constantypedefs | |
2903 | && structdef == snone | |
2904 | && structtype == st_C_enum && bracelev > structbracelev) | |
2905 | return TRUE; /* enum constant */ | |
2906 | switch (fvdef) | |
2907 | { | |
2908 | case fdefunkey: | |
2909 | if (bracelev > 0) | |
2910 | break; | |
2911 | fvdef = fdefunname; /* GNU macro */ | |
2912 | *is_func_or_var = TRUE; | |
2913 | return TRUE; | |
2914 | case fvnone: | |
2915 | switch (typdef) | |
2916 | { | |
2917 | case ttypeseen: | |
2918 | return FALSE; | |
2919 | case tnone: | |
2920 | if ((strneq (str, "asm", 3) && endtoken (str[3])) | |
2921 | || (strneq (str, "__asm__", 7) && endtoken (str[7]))) | |
2922 | { | |
2923 | fvdef = vignore; | |
2924 | return FALSE; | |
2925 | } | |
2926 | break; | |
2927 | } | |
2928 | /* FALLTHRU */ | |
2929 | case fvnameseen: | |
2bc2e3fc | 2930 | if (len >= 10 && strneq (str+len-10, "::operator", 10)) |
8c463abe | 2931 | { |
9c485bbe FP |
2932 | if (*c_extp & C_AUTO) /* automatic detection of C++ */ |
2933 | *c_extp = (*c_extp | C_PLPL) & ~C_AUTO; | |
8c463abe FP |
2934 | fvdef = foperator; |
2935 | *is_func_or_var = TRUE; | |
2936 | return TRUE; | |
2937 | } | |
05d9a399 | 2938 | if (bracelev > 0 && !instruct) |
8c463abe | 2939 | break; |
30903246 FP |
2940 | fvdef = fvnameseen; /* function or variable */ |
2941 | *is_func_or_var = TRUE; | |
b9755a12 | 2942 | return TRUE; |
6dd5561c | 2943 | } |
93b7ac65 | 2944 | break; |
6dd5561c FP |
2945 | } |
2946 | ||
b9755a12 | 2947 | return FALSE; |
6dd5561c FP |
2948 | } |
2949 | ||
b2521b0a | 2950 | \f |
c6d46f5f | 2951 | /* |
b2521b0a FP |
2952 | * C_entries often keeps pointers to tokens or lines which are older than |
2953 | * the line currently read. By keeping two line buffers, and switching | |
2954 | * them at end of line, it is possible to use those pointers. | |
c6d46f5f | 2955 | */ |
db590582 | 2956 | static struct |
b2521b0a FP |
2957 | { |
2958 | long linepos; | |
2959 | linebuffer lb; | |
2960 | } lbs[2]; | |
2961 | ||
55597f90 FP |
2962 | #define current_lb_is_new (newndx == curndx) |
2963 | #define switch_line_buffers() (curndx = 1 - curndx) | |
c6d46f5f | 2964 | |
13fde0cd | 2965 | #define curlb (lbs[curndx].lb) |
13fde0cd RS |
2966 | #define newlb (lbs[newndx].lb) |
2967 | #define curlinepos (lbs[curndx].linepos) | |
13fde0cd RS |
2968 | #define newlinepos (lbs[newndx].linepos) |
2969 | ||
9c485bbe FP |
2970 | #define plainc ((c_ext & C_EXT) == C_PLAIN) |
2971 | #define cplpl (c_ext & C_PLPL) | |
3c04a71a FP |
2972 | #define cjava ((c_ext & C_JAVA) == C_JAVA) |
2973 | ||
93b7ac65 | 2974 | #define CNL_SAVE_DEFINEDEF() \ |
13fde0cd | 2975 | do { \ |
55597f90 | 2976 | curlinepos = charno; \ |
e7d3b099 | 2977 | readline (&curlb, inf); \ |
13fde0cd RS |
2978 | lp = curlb.buffer; \ |
2979 | quotednl = FALSE; \ | |
2980 | newndx = curndx; \ | |
b9755a12 | 2981 | } while (0) |
c6d46f5f | 2982 | |
93b7ac65 | 2983 | #define CNL() \ |
13fde0cd | 2984 | do { \ |
93b7ac65 | 2985 | CNL_SAVE_DEFINEDEF(); \ |
b2521b0a | 2986 | if (savetoken.valid) \ |
55597f90 | 2987 | { \ |
b2521b0a FP |
2988 | token = savetoken; \ |
2989 | savetoken.valid = FALSE; \ | |
55597f90 | 2990 | } \ |
c6d46f5f | 2991 | definedef = dnone; \ |
b9755a12 | 2992 | } while (0) |
13fde0cd | 2993 | |
086eac13 | 2994 | |
944bdd72 | 2995 | static void |
873fbd0b | 2996 | make_C_tag (int isfun) |
086eac13 | 2997 | { |
f6880bf3 | 2998 | /* This function is never called when token.valid is FALSE, but |
086eac13 | 2999 | we must protect against invalid input or internal errors. */ |
9c485bbe | 3000 | if (token.valid) |
2201e3dc FP |
3001 | make_tag (token_name.buffer, token_name.len, isfun, token.line, |
3002 | token.offset+token.length+1, token.lineno, token.linepos); | |
89d8309f | 3003 | else if (DEBUG) |
e1dbe924 | 3004 | { /* this branch is optimized away if !DEBUG */ |
89d8309f FP |
3005 | make_tag (concat ("INVALID TOKEN:-->", token_name.buffer, ""), |
3006 | token_name.len + 17, isfun, token.line, | |
3007 | token.offset+token.length+1, token.lineno, token.linepos); | |
db5a3003 | 3008 | error ("INVALID TOKEN"); |
89d8309f | 3009 | } |
2201e3dc FP |
3010 | |
3011 | token.valid = FALSE; | |
086eac13 FP |
3012 | } |
3013 | ||
c6d46f5f | 3014 | |
b2521b0a FP |
3015 | /* |
3016 | * C_entries () | |
3017 | * This routine finds functions, variables, typedefs, | |
3018 | * #define's, enum constants and struct/union/enum definitions in | |
3019 | * C syntax and adds them to the list. | |
3020 | */ | |
0dacfc4b | 3021 | static void |
873fbd0b DN |
3022 | C_entries (int c_ext, FILE *inf) |
3023 | /* extension of C */ | |
3024 | /* input file */ | |
c6d46f5f | 3025 | { |
13fde0cd | 3026 | register char c; /* latest char read; '\0' for end of line */ |
c6d46f5f | 3027 | register char *lp; /* pointer one beyond the character `c' */ |
13fde0cd | 3028 | int curndx, newndx; /* indices for current and new lb */ |
55597f90 FP |
3029 | register int tokoff; /* offset in line of start of current token */ |
3030 | register int toklen; /* length of current token */ | |
988e88ab | 3031 | const char *qualifier; /* string used to qualify names */ |
30903246 | 3032 | int qlen; /* length of qualifier */ |
05d9a399 FP |
3033 | int bracelev; /* current brace level */ |
3034 | int bracketlev; /* current bracket level */ | |
b12756c8 | 3035 | int parlev; /* current parenthesis level */ |
05d9a399 FP |
3036 | int attrparlev; /* __attribute__ parenthesis level */ |
3037 | int templatelev; /* current template level */ | |
3038 | int typdefbracelev; /* bracelev where a typedef struct body begun */ | |
30903246 | 3039 | bool incomm, inquote, inchar, quotednl, midtoken; |
b2521b0a | 3040 | bool yacc_rules; /* in the rules part of a yacc file */ |
35a4c758 | 3041 | struct tok savetoken = {0}; /* token saved during preprocessor handling */ |
c6d46f5f | 3042 | |
75bdbc6a | 3043 | |
61a1f6fa FP |
3044 | linebuffer_init (&lbs[0].lb); |
3045 | linebuffer_init (&lbs[1].lb); | |
8c463abe FP |
3046 | if (cstack.size == 0) |
3047 | { | |
3048 | cstack.size = (DEBUG) ? 1 : 4; | |
3049 | cstack.nl = 0; | |
3050 | cstack.cname = xnew (cstack.size, char *); | |
05d9a399 | 3051 | cstack.bracelev = xnew (cstack.size, int); |
8c463abe | 3052 | } |
b2521b0a | 3053 | |
05d9a399 | 3054 | tokoff = toklen = typdefbracelev = 0; /* keep compiler quiet */ |
13fde0cd | 3055 | curndx = newndx = 0; |
13fde0cd | 3056 | lp = curlb.buffer; |
c6d46f5f JB |
3057 | *lp = 0; |
3058 | ||
93b7ac65 FP |
3059 | fvdef = fvnone; fvextern = FALSE; typdef = tnone; |
3060 | structdef = snone; definedef = dnone; objdef = onone; | |
b2521b0a | 3061 | yacc_rules = FALSE; |
13fde0cd | 3062 | midtoken = inquote = inchar = incomm = quotednl = FALSE; |
b2521b0a | 3063 | token.valid = savetoken.valid = FALSE; |
05d9a399 | 3064 | bracelev = bracketlev = parlev = attrparlev = templatelev = 0; |
30903246 FP |
3065 | if (cjava) |
3066 | { qualifier = "."; qlen = 1; } | |
3067 | else | |
3068 | { qualifier = "::"; qlen = 2; } | |
c6d46f5f | 3069 | |
b2521b0a | 3070 | |
c6d46f5f JB |
3071 | while (!feof (inf)) |
3072 | { | |
3073 | c = *lp++; | |
c6d46f5f JB |
3074 | if (c == '\\') |
3075 | { | |
05d9a399 FP |
3076 | /* If we are at the end of the line, the next character is a |
3077 | '\0'; do not skip it, because it is what tells us | |
4746118a | 3078 | to read the next line. */ |
13fde0cd | 3079 | if (*lp == '\0') |
99e0a2e0 | 3080 | { |
13fde0cd | 3081 | quotednl = TRUE; |
99e0a2e0 RS |
3082 | continue; |
3083 | } | |
1e134a5f | 3084 | lp++; |
c6d46f5f JB |
3085 | c = ' '; |
3086 | } | |
3087 | else if (incomm) | |
3088 | { | |
13fde0cd | 3089 | switch (c) |
c6d46f5f | 3090 | { |
13fde0cd RS |
3091 | case '*': |
3092 | if (*lp == '/') | |
3093 | { | |
3094 | c = *lp++; | |
3095 | incomm = FALSE; | |
3096 | } | |
3097 | break; | |
3098 | case '\0': | |
3099 | /* Newlines inside comments do not end macro definitions in | |
3100 | traditional cpp. */ | |
93b7ac65 | 3101 | CNL_SAVE_DEFINEDEF (); |
13fde0cd | 3102 | break; |
c6d46f5f | 3103 | } |
13fde0cd | 3104 | continue; |
c6d46f5f JB |
3105 | } |
3106 | else if (inquote) | |
3107 | { | |
13fde0cd RS |
3108 | switch (c) |
3109 | { | |
3110 | case '"': | |
3111 | inquote = FALSE; | |
3112 | break; | |
3113 | case '\0': | |
42680d3c | 3114 | /* Newlines inside strings do not end macro definitions |
13fde0cd RS |
3115 | in traditional cpp, even though compilers don't |
3116 | usually accept them. */ | |
93b7ac65 | 3117 | CNL_SAVE_DEFINEDEF (); |
13fde0cd RS |
3118 | break; |
3119 | } | |
3120 | continue; | |
c6d46f5f JB |
3121 | } |
3122 | else if (inchar) | |
3123 | { | |
42680d3c FP |
3124 | switch (c) |
3125 | { | |
3126 | case '\0': | |
3127 | /* Hmmm, something went wrong. */ | |
93b7ac65 | 3128 | CNL (); |
42680d3c FP |
3129 | /* FALLTHRU */ |
3130 | case '\'': | |
46c145db | 3131 | inchar = FALSE; |
42680d3c FP |
3132 | break; |
3133 | } | |
c6d46f5f JB |
3134 | continue; |
3135 | } | |
05d9a399 FP |
3136 | else switch (c) |
3137 | { | |
3138 | case '"': | |
3139 | inquote = TRUE; | |
28796b3a AS |
3140 | if (bracketlev > 0) |
3141 | continue; | |
05d9a399 FP |
3142 | if (inattribute) |
3143 | break; | |
3144 | switch (fvdef) | |
3145 | { | |
3146 | case fdefunkey: | |
3147 | case fstartlist: | |
3148 | case finlist: | |
3149 | case fignore: | |
3150 | case vignore: | |
3151 | break; | |
3152 | default: | |
3153 | fvextern = FALSE; | |
3154 | fvdef = fvnone; | |
3155 | } | |
3156 | continue; | |
3157 | case '\'': | |
3158 | inchar = TRUE; | |
28796b3a AS |
3159 | if (bracketlev > 0) |
3160 | continue; | |
05d9a399 FP |
3161 | if (inattribute) |
3162 | break; | |
28796b3a | 3163 | if (fvdef != finlist && fvdef != fignore && fvdef != vignore) |
05d9a399 FP |
3164 | { |
3165 | fvextern = FALSE; | |
3166 | fvdef = fvnone; | |
3167 | } | |
3168 | continue; | |
3169 | case '/': | |
3170 | if (*lp == '*') | |
3171 | { | |
05d9a399 | 3172 | incomm = TRUE; |
eac50785 FP |
3173 | lp++; |
3174 | c = ' '; | |
28796b3a AS |
3175 | if (bracketlev > 0) |
3176 | continue; | |
05d9a399 FP |
3177 | } |
3178 | else if (/* cplpl && */ *lp == '/') | |
3179 | { | |
3180 | c = '\0'; | |
05d9a399 | 3181 | } |
eac50785 | 3182 | break; |
05d9a399 FP |
3183 | case '%': |
3184 | if ((c_ext & YACC) && *lp == '%') | |
3185 | { | |
3186 | /* Entering or exiting rules section in yacc file. */ | |
3187 | lp++; | |
3188 | definedef = dnone; fvdef = fvnone; fvextern = FALSE; | |
3189 | typdef = tnone; structdef = snone; | |
3190 | midtoken = inquote = inchar = incomm = quotednl = FALSE; | |
3191 | bracelev = 0; | |
3192 | yacc_rules = !yacc_rules; | |
3193 | continue; | |
3194 | } | |
3195 | else | |
3196 | break; | |
3197 | case '#': | |
3198 | if (definedef == dnone) | |
3199 | { | |
3200 | char *cp; | |
3201 | bool cpptoken = TRUE; | |
3202 | ||
3203 | /* Look back on this line. If all blanks, or nonblanks | |
3204 | followed by an end of comment, this is a preprocessor | |
3205 | token. */ | |
3206 | for (cp = newlb.buffer; cp < lp-1; cp++) | |
3207 | if (!iswhite (*cp)) | |
3208 | { | |
28796b3a | 3209 | if (*cp == '*' && cp[1] == '/') |
05d9a399 FP |
3210 | { |
3211 | cp++; | |
3212 | cpptoken = TRUE; | |
3213 | } | |
3214 | else | |
3215 | cpptoken = FALSE; | |
3216 | } | |
3217 | if (cpptoken) | |
3218 | definedef = dsharpseen; | |
3219 | } /* if (definedef == dnone) */ | |
3220 | continue; | |
3221 | case '[': | |
3222 | bracketlev++; | |
28796b3a AS |
3223 | continue; |
3224 | default: | |
3225 | if (bracketlev > 0) | |
3226 | { | |
3227 | if (c == ']') | |
3228 | --bracketlev; | |
3229 | else if (c == '\0') | |
3230 | CNL_SAVE_DEFINEDEF (); | |
3231 | continue; | |
3232 | } | |
3233 | break; | |
05d9a399 | 3234 | } /* switch (c) */ |
c6d46f5f | 3235 | |
c6d46f5f | 3236 | |
8c463abe | 3237 | /* Consider token only if some involved conditions are satisfied. */ |
b2521b0a | 3238 | if (typdef != tignore |
13fde0cd | 3239 | && definedef != dignorerest |
b2521b0a | 3240 | && fvdef != finlist |
05d9a399 | 3241 | && templatelev == 0 |
b2521b0a | 3242 | && (definedef != dnone |
05d9a399 FP |
3243 | || structdef != scolonseen) |
3244 | && !inattribute) | |
c6d46f5f JB |
3245 | { |
3246 | if (midtoken) | |
3247 | { | |
3248 | if (endtoken (c)) | |
3249 | { | |
9c485bbe FP |
3250 | if (c == ':' && *lp == ':' && begtoken (lp[1])) |
3251 | /* This handles :: in the middle, | |
3252 | but not at the beginning of an identifier. | |
e1dbe924 | 3253 | Also, space-separated :: is not recognized. */ |
c6d46f5f | 3254 | { |
9c485bbe FP |
3255 | if (c_ext & C_AUTO) /* automatic detection of C++ */ |
3256 | c_ext = (c_ext | C_PLPL) & ~C_AUTO; | |
c6d46f5f | 3257 | lp += 2; |
93b7ac65 FP |
3258 | toklen += 2; |
3259 | c = lp[-1]; | |
8f79fe72 | 3260 | goto still_in_token; |
c6d46f5f JB |
3261 | } |
3262 | else | |
3263 | { | |
b2521b0a FP |
3264 | bool funorvar = FALSE; |
3265 | ||
13fde0cd | 3266 | if (yacc_rules |
d8913c1c | 3267 | || consider_token (newlb.buffer + tokoff, toklen, c, |
05d9a399 FP |
3268 | &c_ext, bracelev, parlev, |
3269 | &funorvar)) | |
c6d46f5f | 3270 | { |
93b7ac65 FP |
3271 | if (fvdef == foperator) |
3272 | { | |
3273 | char *oldlp = lp; | |
3274 | lp = skip_spaces (lp-1); | |
3275 | if (*lp != '\0') | |
3276 | lp += 1; | |
3277 | while (*lp != '\0' | |
71cbb895 | 3278 | && !iswhite (*lp) && *lp != '(') |
93b7ac65 FP |
3279 | lp += 1; |
3280 | c = *lp++; | |
3281 | toklen += lp - oldlp; | |
3282 | } | |
b2521b0a | 3283 | token.named = FALSE; |
9c485bbe | 3284 | if (!plainc |
8c463abe FP |
3285 | && nestlev > 0 && definedef == dnone) |
3286 | /* in struct body */ | |
fe0b3356 | 3287 | { |
e99a530f | 3288 | int len; |
8c463abe | 3289 | write_classname (&token_name, qualifier); |
e99a530f PE |
3290 | len = token_name.len; |
3291 | linebuffer_setlen (&token_name, len+qlen+toklen); | |
3292 | sprintf (token_name.buffer + len, "%s%.*s", | |
3293 | qualifier, toklen, newlb.buffer + tokoff); | |
b2521b0a | 3294 | token.named = TRUE; |
c6d46f5f | 3295 | } |
d8913c1c FP |
3296 | else if (objdef == ocatseen) |
3297 | /* Objective C category */ | |
3298 | { | |
30903246 | 3299 | int len = strlen (objtag) + 2 + toklen; |
b2521b0a | 3300 | linebuffer_setlen (&token_name, len); |
e99a530f PE |
3301 | sprintf (token_name.buffer, "%s(%.*s)", |
3302 | objtag, toklen, newlb.buffer + tokoff); | |
b2521b0a | 3303 | token.named = TRUE; |
d8913c1c FP |
3304 | } |
3305 | else if (objdef == omethodtag | |
3306 | || objdef == omethodparm) | |
3307 | /* Objective C method */ | |
3308 | { | |
b2521b0a FP |
3309 | token.named = TRUE; |
3310 | } | |
3311 | else if (fvdef == fdefunname) | |
8c463abe | 3312 | /* GNU DEFUN and similar macros */ |
b2521b0a FP |
3313 | { |
3314 | bool defun = (newlb.buffer[tokoff] == 'F'); | |
3315 | int off = tokoff; | |
3316 | int len = toklen; | |
3317 | ||
3318 | /* Rewrite the tag so that emacs lisp DEFUNs | |
3319 | can be found by their elisp name */ | |
3320 | if (defun) | |
3321 | { | |
3322 | off += 1; | |
3323 | len -= 1; | |
3324 | } | |
b2521b0a | 3325 | linebuffer_setlen (&token_name, len); |
e99a530f PE |
3326 | memcpy (token_name.buffer, |
3327 | newlb.buffer + off, len); | |
b2521b0a FP |
3328 | token_name.buffer[len] = '\0'; |
3329 | if (defun) | |
3330 | while (--len >= 0) | |
3331 | if (token_name.buffer[len] == '_') | |
3332 | token_name.buffer[len] = '-'; | |
3333 | token.named = defun; | |
d8913c1c | 3334 | } |
c6d46f5f JB |
3335 | else |
3336 | { | |
b2521b0a | 3337 | linebuffer_setlen (&token_name, toklen); |
e99a530f PE |
3338 | memcpy (token_name.buffer, |
3339 | newlb.buffer + tokoff, toklen); | |
75bdbc6a | 3340 | token_name.buffer[toklen] = '\0'; |
93b7ac65 | 3341 | /* Name macros and members. */ |
b2521b0a FP |
3342 | token.named = (structdef == stagseen |
3343 | || typdef == ttypeseen | |
3344 | || typdef == tend | |
3345 | || (funorvar | |
3346 | && definedef == dignorerest) | |
3347 | || (funorvar | |
3348 | && definedef == dnone | |
8c463abe | 3349 | && structdef == snone |
05d9a399 | 3350 | && bracelev > 0)); |
c6d46f5f | 3351 | } |
b2521b0a | 3352 | token.lineno = lineno; |
8c463abe FP |
3353 | token.offset = tokoff; |
3354 | token.length = toklen; | |
b2521b0a FP |
3355 | token.line = newlb.buffer; |
3356 | token.linepos = newlinepos; | |
3357 | token.valid = TRUE; | |
fe0b3356 | 3358 | |
b12756c8 | 3359 | if (definedef == dnone |
30903246 | 3360 | && (fvdef == fvnameseen |
93b7ac65 | 3361 | || fvdef == foperator |
b12756c8 | 3362 | || structdef == stagseen |
d8913c1c | 3363 | || typdef == tend |
8c463abe | 3364 | || typdef == ttypeseen |
d8913c1c | 3365 | || objdef != onone)) |
13fde0cd | 3366 | { |
55597f90 FP |
3367 | if (current_lb_is_new) |
3368 | switch_line_buffers (); | |
13fde0cd | 3369 | } |
8c463abe FP |
3370 | else if (definedef != dnone |
3371 | || fvdef == fdefunname | |
3372 | || instruct) | |
30903246 | 3373 | make_C_tag (funorvar); |
c6d46f5f | 3374 | } |
05d9a399 FP |
3375 | else /* not yacc and consider_token failed */ |
3376 | { | |
3377 | if (inattribute && fvdef == fignore) | |
3378 | { | |
3379 | /* We have just met __attribute__ after a | |
3380 | function parameter list: do not tag the | |
3381 | function again. */ | |
3382 | fvdef = fvnone; | |
3383 | } | |
3384 | } | |
c6d46f5f JB |
3385 | midtoken = FALSE; |
3386 | } | |
13fde0cd | 3387 | } /* if (endtoken (c)) */ |
c6d46f5f | 3388 | else if (intoken (c)) |
8f79fe72 | 3389 | still_in_token: |
13fde0cd RS |
3390 | { |
3391 | toklen++; | |
3392 | continue; | |
3393 | } | |
3394 | } /* if (midtoken) */ | |
c6d46f5f JB |
3395 | else if (begtoken (c)) |
3396 | { | |
b12756c8 | 3397 | switch (definedef) |
13fde0cd | 3398 | { |
b12756c8 | 3399 | case dnone: |
30903246 | 3400 | switch (fvdef) |
b12756c8 FP |
3401 | { |
3402 | case fstartlist: | |
05d9a399 FP |
3403 | /* This prevents tagging fb in |
3404 | void (__attribute__((noreturn)) *fb) (void); | |
3405 | Fixing this is not easy and not very important. */ | |
30903246 | 3406 | fvdef = finlist; |
b12756c8 FP |
3407 | continue; |
3408 | case flistseen: | |
9c485bbe FP |
3409 | if (plainc || declarations) |
3410 | { | |
3411 | make_C_tag (TRUE); /* a function */ | |
3412 | fvdef = fignore; | |
3413 | } | |
b12756c8 | 3414 | break; |
b12756c8 | 3415 | } |
cec68fcb | 3416 | if (structdef == stagseen && !cjava) |
8c463abe | 3417 | { |
05d9a399 | 3418 | popclass_above (bracelev); |
8c463abe FP |
3419 | structdef = snone; |
3420 | } | |
13fde0cd | 3421 | break; |
b12756c8 | 3422 | case dsharpseen: |
b2521b0a | 3423 | savetoken = token; |
f6566f90 | 3424 | break; |
13fde0cd | 3425 | } |
13fde0cd RS |
3426 | if (!yacc_rules || lp == newlb.buffer + 1) |
3427 | { | |
3428 | tokoff = lp - 1 - newlb.buffer; | |
3429 | toklen = 1; | |
3430 | midtoken = TRUE; | |
3431 | } | |
3432 | continue; | |
4b533b5b | 3433 | } /* if (begtoken) */ |
13fde0cd RS |
3434 | } /* if must look at token */ |
3435 | ||
3436 | ||
3437 | /* Detect end of line, colon, comma, semicolon and various braces | |
b12756c8 | 3438 | after having handled a token.*/ |
13fde0cd | 3439 | switch (c) |
1e134a5f | 3440 | { |
13fde0cd | 3441 | case ':': |
05d9a399 FP |
3442 | if (inattribute) |
3443 | break; | |
8c463abe FP |
3444 | if (yacc_rules && token.offset == 0 && token.valid) |
3445 | { | |
3446 | make_C_tag (FALSE); /* a yacc function */ | |
3447 | break; | |
3448 | } | |
b12756c8 FP |
3449 | if (definedef != dnone) |
3450 | break; | |
d8913c1c FP |
3451 | switch (objdef) |
3452 | { | |
3453 | case otagseen: | |
3454 | objdef = oignore; | |
30903246 | 3455 | make_C_tag (TRUE); /* an Objective C class */ |
d8913c1c FP |
3456 | break; |
3457 | case omethodtag: | |
3458 | case omethodparm: | |
3459 | objdef = omethodcolon; | |
b2521b0a | 3460 | linebuffer_setlen (&token_name, token_name.len + 1); |
d8913c1c FP |
3461 | strcat (token_name.buffer, ":"); |
3462 | break; | |
3463 | } | |
13fde0cd | 3464 | if (structdef == stagseen) |
3c04a71a FP |
3465 | { |
3466 | structdef = scolonseen; | |
3467 | break; | |
3468 | } | |
9c485bbe | 3469 | /* Should be useless, but may be work as a safety net. */ |
3c04a71a FP |
3470 | if (cplpl && fvdef == flistseen) |
3471 | { | |
3472 | make_C_tag (TRUE); /* a function */ | |
3473 | fvdef = fignore; | |
3474 | break; | |
3475 | } | |
13fde0cd RS |
3476 | break; |
3477 | case ';': | |
05d9a399 | 3478 | if (definedef != dnone || inattribute) |
b12756c8 | 3479 | break; |
8c463abe | 3480 | switch (typdef) |
1f638249 | 3481 | { |
8c463abe FP |
3482 | case tend: |
3483 | case ttypeseen: | |
3484 | make_C_tag (FALSE); /* a typedef */ | |
3485 | typdef = tnone; | |
93b7ac65 | 3486 | fvdef = fvnone; |
93b7ac65 | 3487 | break; |
8c463abe FP |
3488 | case tnone: |
3489 | case tinbody: | |
3490 | case tignore: | |
3491 | switch (fvdef) | |
3492 | { | |
3493 | case fignore: | |
3c04a71a | 3494 | if (typdef == tignore || cplpl) |
8c463abe FP |
3495 | fvdef = fvnone; |
3496 | break; | |
3497 | case fvnameseen: | |
05d9a399 | 3498 | if ((globals && bracelev == 0 && (!fvextern || declarations)) |
8c463abe FP |
3499 | || (members && instruct)) |
3500 | make_C_tag (FALSE); /* a variable */ | |
3501 | fvextern = FALSE; | |
3502 | fvdef = fvnone; | |
3503 | token.valid = FALSE; | |
3504 | break; | |
3505 | case flistseen: | |
05d9a399 FP |
3506 | if ((declarations |
3507 | && (cplpl || !instruct) | |
3508 | && (typdef == tnone || (typdef != tignore && instruct))) | |
3509 | || (members | |
3510 | && plainc && instruct)) | |
3511 | make_C_tag (TRUE); /* a function */ | |
8c463abe FP |
3512 | /* FALLTHRU */ |
3513 | default: | |
3514 | fvextern = FALSE; | |
3515 | fvdef = fvnone; | |
3516 | if (declarations | |
9c485bbe | 3517 | && cplpl && structdef == stagseen) |
8c463abe FP |
3518 | make_C_tag (FALSE); /* forward declaration */ |
3519 | else | |
8c463abe FP |
3520 | token.valid = FALSE; |
3521 | } /* switch (fvdef) */ | |
30903246 FP |
3522 | /* FALLTHRU */ |
3523 | default: | |
8c463abe | 3524 | if (!instruct) |
b2521b0a | 3525 | typdef = tnone; |
8c463abe | 3526 | } |
46c145db FP |
3527 | if (structdef == stagseen) |
3528 | structdef = snone; | |
3529 | break; | |
13fde0cd | 3530 | case ',': |
05d9a399 | 3531 | if (definedef != dnone || inattribute) |
46c145db | 3532 | break; |
d8913c1c FP |
3533 | switch (objdef) |
3534 | { | |
3535 | case omethodtag: | |
3536 | case omethodparm: | |
30903246 | 3537 | make_C_tag (TRUE); /* an Objective C method */ |
d8913c1c FP |
3538 | objdef = oinbody; |
3539 | break; | |
3540 | } | |
30903246 FP |
3541 | switch (fvdef) |
3542 | { | |
b2521b0a | 3543 | case fdefunkey: |
93b7ac65 | 3544 | case foperator: |
8c463abe | 3545 | case fstartlist: |
30903246 FP |
3546 | case finlist: |
3547 | case fignore: | |
3548 | case vignore: | |
3549 | break; | |
b2521b0a FP |
3550 | case fdefunname: |
3551 | fvdef = fignore; | |
30903246 | 3552 | break; |
05d9a399 FP |
3553 | case fvnameseen: |
3554 | if (parlev == 0 | |
3555 | && ((globals | |
3556 | && bracelev == 0 | |
3557 | && templatelev == 0 | |
3558 | && (!fvextern || declarations)) | |
3559 | || (members && instruct))) | |
3560 | make_C_tag (FALSE); /* a variable */ | |
8c463abe | 3561 | break; |
05d9a399 | 3562 | case flistseen: |
8c463abe FP |
3563 | if ((declarations && typdef == tnone && !instruct) |
3564 | || (members && typdef != tignore && instruct)) | |
b2521b0a | 3565 | { |
05d9a399 | 3566 | make_C_tag (TRUE); /* a function */ |
8c463abe | 3567 | fvdef = fvnameseen; |
b2521b0a | 3568 | } |
8c463abe FP |
3569 | else if (!declarations) |
3570 | fvdef = fvnone; | |
3571 | token.valid = FALSE; | |
3572 | break; | |
30903246 FP |
3573 | default: |
3574 | fvdef = fvnone; | |
3575 | } | |
46c145db FP |
3576 | if (structdef == stagseen) |
3577 | structdef = snone; | |
3578 | break; | |
05d9a399 FP |
3579 | case ']': |
3580 | if (definedef != dnone || inattribute) | |
b12756c8 | 3581 | break; |
8c463abe FP |
3582 | if (structdef == stagseen) |
3583 | structdef = snone; | |
3584 | switch (typdef) | |
46c145db | 3585 | { |
8c463abe FP |
3586 | case ttypeseen: |
3587 | case tend: | |
46c145db | 3588 | typdef = tignore; |
30903246 | 3589 | make_C_tag (FALSE); /* a typedef */ |
46c145db | 3590 | break; |
8c463abe FP |
3591 | case tnone: |
3592 | case tinbody: | |
3593 | switch (fvdef) | |
3594 | { | |
3595 | case foperator: | |
3596 | case finlist: | |
3597 | case fignore: | |
3598 | case vignore: | |
3599 | break; | |
3600 | case fvnameseen: | |
05d9a399 FP |
3601 | if ((members && bracelev == 1) |
3602 | || (globals && bracelev == 0 | |
8c463abe FP |
3603 | && (!fvextern || declarations))) |
3604 | make_C_tag (FALSE); /* a variable */ | |
3605 | /* FALLTHRU */ | |
3606 | default: | |
3607 | fvdef = fvnone; | |
3608 | } | |
30903246 | 3609 | break; |
30903246 | 3610 | } |
13fde0cd RS |
3611 | break; |
3612 | case '(': | |
05d9a399 FP |
3613 | if (inattribute) |
3614 | { | |
3615 | attrparlev++; | |
3616 | break; | |
3617 | } | |
b12756c8 FP |
3618 | if (definedef != dnone) |
3619 | break; | |
d8913c1c FP |
3620 | if (objdef == otagseen && parlev == 0) |
3621 | objdef = oparenseen; | |
30903246 | 3622 | switch (fvdef) |
57e83cfe | 3623 | { |
30903246 | 3624 | case fvnameseen: |
93b7ac65 | 3625 | if (typdef == ttypeseen |
93b7ac65 | 3626 | && *lp != '*' |
8c463abe | 3627 | && !instruct) |
93b7ac65 FP |
3628 | { |
3629 | /* This handles constructs like: | |
3630 | typedef void OperatorFun (int fun); */ | |
3631 | make_C_tag (FALSE); | |
3632 | typdef = tignore; | |
8c463abe FP |
3633 | fvdef = fignore; |
3634 | break; | |
93b7ac65 FP |
3635 | } |
3636 | /* FALLTHRU */ | |
3637 | case foperator: | |
3638 | fvdef = fstartlist; | |
13fde0cd | 3639 | break; |
13fde0cd | 3640 | case flistseen: |
30903246 | 3641 | fvdef = finlist; |
13fde0cd | 3642 | break; |
57e83cfe | 3643 | } |
b12756c8 | 3644 | parlev++; |
13fde0cd RS |
3645 | break; |
3646 | case ')': | |
05d9a399 FP |
3647 | if (inattribute) |
3648 | { | |
3649 | if (--attrparlev == 0) | |
3650 | inattribute = FALSE; | |
3651 | break; | |
3652 | } | |
b12756c8 FP |
3653 | if (definedef != dnone) |
3654 | break; | |
d8913c1c FP |
3655 | if (objdef == ocatseen && parlev == 1) |
3656 | { | |
30903246 | 3657 | make_C_tag (TRUE); /* an Objective C category */ |
d8913c1c FP |
3658 | objdef = oignore; |
3659 | } | |
b12756c8 FP |
3660 | if (--parlev == 0) |
3661 | { | |
30903246 | 3662 | switch (fvdef) |
b12756c8 FP |
3663 | { |
3664 | case fstartlist: | |
3665 | case finlist: | |
30903246 | 3666 | fvdef = flistseen; |
b12756c8 FP |
3667 | break; |
3668 | } | |
8c463abe FP |
3669 | if (!instruct |
3670 | && (typdef == tend | |
3671 | || typdef == ttypeseen)) | |
46c145db FP |
3672 | { |
3673 | typdef = tignore; | |
30903246 | 3674 | make_C_tag (FALSE); /* a typedef */ |
46c145db | 3675 | } |
b12756c8 FP |
3676 | } |
3677 | else if (parlev < 0) /* can happen due to ill-conceived #if's. */ | |
3678 | parlev = 0; | |
13fde0cd RS |
3679 | break; |
3680 | case '{': | |
b12756c8 FP |
3681 | if (definedef != dnone) |
3682 | break; | |
93b7ac65 | 3683 | if (typdef == ttypeseen) |
8c463abe | 3684 | { |
f55ae599 | 3685 | /* Whenever typdef is set to tinbody (currently only |
05d9a399 | 3686 | here), typdefbracelev should be set to bracelev. */ |
8c463abe | 3687 | typdef = tinbody; |
05d9a399 | 3688 | typdefbracelev = bracelev; |
8c463abe | 3689 | } |
30903246 | 3690 | switch (fvdef) |
31d4b314 FP |
3691 | { |
3692 | case flistseen: | |
b2521b0a | 3693 | make_C_tag (TRUE); /* a function */ |
31d4b314 FP |
3694 | /* FALLTHRU */ |
3695 | case fignore: | |
30903246 | 3696 | fvdef = fvnone; |
46c145db | 3697 | break; |
30903246 | 3698 | case fvnone: |
d8913c1c FP |
3699 | switch (objdef) |
3700 | { | |
3701 | case otagseen: | |
30903246 | 3702 | make_C_tag (TRUE); /* an Objective C class */ |
d8913c1c FP |
3703 | objdef = oignore; |
3704 | break; | |
3705 | case omethodtag: | |
3706 | case omethodparm: | |
30903246 | 3707 | make_C_tag (TRUE); /* an Objective C method */ |
d8913c1c FP |
3708 | objdef = oinbody; |
3709 | break; | |
3710 | default: | |
3ca80e28 | 3711 | /* Neutralize `extern "C" {' grot. */ |
05d9a399 | 3712 | if (bracelev == 0 && structdef == snone && nestlev == 0 |
8c463abe | 3713 | && typdef == tnone) |
05d9a399 | 3714 | bracelev = -1; |
d8913c1c | 3715 | } |
f6566f90 | 3716 | break; |
31d4b314 | 3717 | } |
b2521b0a FP |
3718 | switch (structdef) |
3719 | { | |
3720 | case skeyseen: /* unnamed struct */ | |
05d9a399 | 3721 | pushclass_above (bracelev, NULL, 0); |
8c463abe | 3722 | structdef = snone; |
b2521b0a | 3723 | break; |
8c463abe FP |
3724 | case stagseen: /* named struct or enum */ |
3725 | case scolonseen: /* a class */ | |
05d9a399 | 3726 | pushclass_above (bracelev,token.line+token.offset, token.length); |
8c463abe FP |
3727 | structdef = snone; |
3728 | make_C_tag (FALSE); /* a struct or enum */ | |
b2521b0a FP |
3729 | break; |
3730 | } | |
89d8309f | 3731 | bracelev += 1; |
31d4b314 | 3732 | break; |
13fde0cd | 3733 | case '*': |
b12756c8 FP |
3734 | if (definedef != dnone) |
3735 | break; | |
30903246 | 3736 | if (fvdef == fstartlist) |
9c485bbe FP |
3737 | { |
3738 | fvdef = fvnone; /* avoid tagging `foo' in `foo (*bar()) ()' */ | |
3739 | token.valid = FALSE; | |
3740 | } | |
13fde0cd RS |
3741 | break; |
3742 | case '}': | |
b12756c8 FP |
3743 | if (definedef != dnone) |
3744 | break; | |
89d8309f | 3745 | bracelev -= 1; |
3c04a71a | 3746 | if (!ignoreindent && lp == newlb.buffer + 1) |
b12756c8 | 3747 | { |
05d9a399 | 3748 | if (bracelev != 0) |
89d8309f | 3749 | token.valid = FALSE; /* unexpected value, token unreliable */ |
05d9a399 | 3750 | bracelev = 0; /* reset brace level if first column */ |
b12756c8 FP |
3751 | parlev = 0; /* also reset paren level, just in case... */ |
3752 | } | |
89d8309f | 3753 | else if (bracelev < 0) |
172aa4c1 | 3754 | { |
89d8309f FP |
3755 | token.valid = FALSE; /* something gone amiss, token unreliable */ |
3756 | bracelev = 0; | |
172aa4c1 | 3757 | } |
f14a3ad4 FP |
3758 | if (bracelev == 0 && fvdef == vignore) |
3759 | fvdef = fvnone; /* end of function */ | |
05d9a399 | 3760 | popclass_above (bracelev); |
8c463abe | 3761 | structdef = snone; |
05d9a399 FP |
3762 | /* Only if typdef == tinbody is typdefbracelev significant. */ |
3763 | if (typdef == tinbody && bracelev <= typdefbracelev) | |
13fde0cd | 3764 | { |
05d9a399 | 3765 | assert (bracelev == typdefbracelev); |
8c463abe | 3766 | typdef = tend; |
13fde0cd RS |
3767 | } |
3768 | break; | |
30903246 FP |
3769 | case '=': |
3770 | if (definedef != dnone) | |
3771 | break; | |
3772 | switch (fvdef) | |
3773 | { | |
93b7ac65 | 3774 | case foperator: |
30903246 FP |
3775 | case finlist: |
3776 | case fignore: | |
3777 | case vignore: | |
3778 | break; | |
3779 | case fvnameseen: | |
05d9a399 FP |
3780 | if ((members && bracelev == 1) |
3781 | || (globals && bracelev == 0 && (!fvextern || declarations))) | |
30903246 FP |
3782 | make_C_tag (FALSE); /* a variable */ |
3783 | /* FALLTHRU */ | |
3784 | default: | |
3785 | fvdef = vignore; | |
3786 | } | |
3787 | break; | |
8c463abe | 3788 | case '<': |
05d9a399 FP |
3789 | if (cplpl |
3790 | && (structdef == stagseen || fvdef == fvnameseen)) | |
8c463abe | 3791 | { |
05d9a399 | 3792 | templatelev++; |
8c463abe FP |
3793 | break; |
3794 | } | |
3795 | goto resetfvdef; | |
3796 | case '>': | |
05d9a399 | 3797 | if (templatelev > 0) |
8c463abe | 3798 | { |
05d9a399 | 3799 | templatelev--; |
8c463abe FP |
3800 | break; |
3801 | } | |
3802 | goto resetfvdef; | |
d8913c1c FP |
3803 | case '+': |
3804 | case '-': | |
05d9a399 | 3805 | if (objdef == oinbody && bracelev == 0) |
d8913c1c FP |
3806 | { |
3807 | objdef = omethodsign; | |
3808 | break; | |
3809 | } | |
3810 | /* FALLTHRU */ | |
8c463abe | 3811 | resetfvdef: |
05d9a399 FP |
3812 | case '#': case '~': case '&': case '%': case '/': |
3813 | case '|': case '^': case '!': case '.': case '?': | |
b12756c8 FP |
3814 | if (definedef != dnone) |
3815 | break; | |
93b7ac65 FP |
3816 | /* These surely cannot follow a function tag in C. */ |
3817 | switch (fvdef) | |
3818 | { | |
3819 | case foperator: | |
3820 | case finlist: | |
3821 | case fignore: | |
3822 | case vignore: | |
3823 | break; | |
3824 | default: | |
3825 | fvdef = fvnone; | |
3826 | } | |
b12756c8 | 3827 | break; |
13fde0cd | 3828 | case '\0': |
d8913c1c FP |
3829 | if (objdef == otagseen) |
3830 | { | |
30903246 | 3831 | make_C_tag (TRUE); /* an Objective C class */ |
d8913c1c FP |
3832 | objdef = oignore; |
3833 | } | |
13fde0cd RS |
3834 | /* If a macro spans multiple lines don't reset its state. */ |
3835 | if (quotednl) | |
93b7ac65 | 3836 | CNL_SAVE_DEFINEDEF (); |
13fde0cd | 3837 | else |
93b7ac65 | 3838 | CNL (); |
13fde0cd RS |
3839 | break; |
3840 | } /* switch (c) */ | |
3841 | ||
3842 | } /* while not eof */ | |
b2521b0a | 3843 | |
b2521b0a FP |
3844 | free (lbs[0].lb.buffer); |
3845 | free (lbs[1].lb.buffer); | |
c6d46f5f | 3846 | } |
b9755a12 FP |
3847 | |
3848 | /* | |
3849 | * Process either a C++ file or a C file depending on the setting | |
3850 | * of a global flag. | |
3851 | */ | |
944bdd72 | 3852 | static void |
873fbd0b | 3853 | default_C_entries (FILE *inf) |
b9755a12 | 3854 | { |
8c463abe | 3855 | C_entries (cplusplus ? C_PLPL : C_AUTO, inf); |
b9755a12 FP |
3856 | } |
3857 | ||
8c463abe | 3858 | /* Always do plain C. */ |
944bdd72 | 3859 | static void |
873fbd0b | 3860 | plain_C_entries (FILE *inf) |
79263656 FP |
3861 | { |
3862 | C_entries (0, inf); | |
3863 | } | |
3864 | ||
b9755a12 | 3865 | /* Always do C++. */ |
944bdd72 | 3866 | static void |
873fbd0b | 3867 | Cplusplus_entries (FILE *inf) |
b9755a12 FP |
3868 | { |
3869 | C_entries (C_PLPL, inf); | |
3870 | } | |
3871 | ||
cec68fcb | 3872 | /* Always do Java. */ |
944bdd72 | 3873 | static void |
873fbd0b | 3874 | Cjava_entries (FILE *inf) |
cec68fcb FP |
3875 | { |
3876 | C_entries (C_JAVA, inf); | |
3877 | } | |
3878 | ||
b9755a12 | 3879 | /* Always do C*. */ |
944bdd72 | 3880 | static void |
873fbd0b | 3881 | Cstar_entries (FILE *inf) |
b9755a12 FP |
3882 | { |
3883 | C_entries (C_STAR, inf); | |
3884 | } | |
3885 | ||
3886 | /* Always do Yacc. */ | |
944bdd72 | 3887 | static void |
873fbd0b | 3888 | Yacc_entries (FILE *inf) |
b9755a12 FP |
3889 | { |
3890 | C_entries (YACC, inf); | |
3891 | } | |
b2521b0a | 3892 | |
6dd5561c | 3893 | \f |
62aec606 | 3894 | /* Useful macros. */ |
93b7ac65 | 3895 | #define LOOP_ON_INPUT_LINES(file_pointer, line_buffer, char_pointer) \ |
e7d3b099 | 3896 | for (; /* loop initialization */ \ |
93b7ac65 | 3897 | !feof (file_pointer) /* loop test */ \ |
2b749964 FP |
3898 | && /* instructions at start of loop */ \ |
3899 | (readline (&line_buffer, file_pointer), \ | |
3900 | char_pointer = line_buffer.buffer, \ | |
93b7ac65 FP |
3901 | TRUE); \ |
3902 | ) | |
55102b5d FP |
3903 | |
3904 | #define LOOKING_AT(cp, kw) /* kw is the keyword, a literal string */ \ | |
5e617bc2 JB |
3905 | ((assert ("" kw), TRUE) /* syntax error if not a literal string */ \ |
3906 | && strneq ((cp), kw, sizeof (kw)-1) /* cp points at kw */ \ | |
3907 | && notinname ((cp)[sizeof (kw)-1]) /* end of kw */ \ | |
3908 | && ((cp) = skip_spaces ((cp)+sizeof (kw)-1))) /* skip spaces */ | |
55102b5d FP |
3909 | |
3910 | /* Similar to LOOKING_AT but does not use notinname, does not skip */ | |
3911 | #define LOOKING_AT_NOCASE(cp, kw) /* the keyword is a literal string */ \ | |
5e617bc2 JB |
3912 | ((assert ("" kw), TRUE) /* syntax error if not a literal string */ \ |
3913 | && strncaseeq ((cp), kw, sizeof (kw)-1) /* cp points at kw */ \ | |
3914 | && ((cp) += sizeof (kw)-1)) /* skip spaces */ | |
93b7ac65 FP |
3915 | |
3916 | /* | |
3917 | * Read a file, but do no processing. This is used to do regexp | |
3918 | * matching on files that have no language defined. | |
3919 | */ | |
944bdd72 | 3920 | static void |
873fbd0b | 3921 | just_read_file (FILE *inf) |
93b7ac65 | 3922 | { |
8c422c30 PE |
3923 | while (!feof (inf)) |
3924 | readline (&lb, inf); | |
93b7ac65 | 3925 | } |
b2521b0a | 3926 | |
93b7ac65 FP |
3927 | \f |
3928 | /* Fortran parsing */ | |
c6d46f5f | 3929 | |
f57e2426 J |
3930 | static void F_takeprec (void); |
3931 | static void F_getit (FILE *); | |
13fde0cd | 3932 | |
944bdd72 | 3933 | static void |
873fbd0b | 3934 | F_takeprec (void) |
6dd5561c | 3935 | { |
93b7ac65 | 3936 | dbp = skip_spaces (dbp); |
6dd5561c FP |
3937 | if (*dbp != '*') |
3938 | return; | |
3939 | dbp++; | |
93b7ac65 | 3940 | dbp = skip_spaces (dbp); |
79263656 FP |
3941 | if (strneq (dbp, "(*)", 3)) |
3942 | { | |
3943 | dbp += 3; | |
3944 | return; | |
3945 | } | |
af03e6ab | 3946 | if (!ISDIGIT (*dbp)) |
c6d46f5f | 3947 | { |
6dd5561c FP |
3948 | --dbp; /* force failure */ |
3949 | return; | |
c6d46f5f | 3950 | } |
6dd5561c FP |
3951 | do |
3952 | dbp++; | |
af03e6ab | 3953 | while (ISDIGIT (*dbp)); |
6dd5561c | 3954 | } |
13fde0cd | 3955 | |
944bdd72 | 3956 | static void |
873fbd0b | 3957 | F_getit (FILE *inf) |
6dd5561c FP |
3958 | { |
3959 | register char *cp; | |
13fde0cd | 3960 | |
93b7ac65 | 3961 | dbp = skip_spaces (dbp); |
6dd5561c | 3962 | if (*dbp == '\0') |
c6d46f5f | 3963 | { |
e7d3b099 | 3964 | readline (&lb, inf); |
6dd5561c FP |
3965 | dbp = lb.buffer; |
3966 | if (dbp[5] != '&') | |
3967 | return; | |
3968 | dbp += 6; | |
93b7ac65 | 3969 | dbp = skip_spaces (dbp); |
c6d46f5f | 3970 | } |
af03e6ab | 3971 | if (!ISALPHA (*dbp) && *dbp != '_' && *dbp != '$') |
6dd5561c | 3972 | return; |
93b7ac65 | 3973 | for (cp = dbp + 1; *cp != '\0' && intoken (*cp); cp++) |
6dd5561c | 3974 | continue; |
081b7c70 FP |
3975 | make_tag (dbp, cp-dbp, TRUE, |
3976 | lb.buffer, cp - lb.buffer + 1, lineno, linecharno); | |
c6d46f5f | 3977 | } |
c6d46f5f | 3978 | |
93b7ac65 | 3979 | |
944bdd72 | 3980 | static void |
873fbd0b | 3981 | Fortran_functions (FILE *inf) |
c6d46f5f | 3982 | { |
93b7ac65 | 3983 | LOOP_ON_INPUT_LINES (inf, lb, dbp) |
c6d46f5f | 3984 | { |
c6d46f5f JB |
3985 | if (*dbp == '%') |
3986 | dbp++; /* Ratfor escape to fortran */ | |
93b7ac65 | 3987 | dbp = skip_spaces (dbp); |
108c932a | 3988 | if (*dbp == '\0') |
c6d46f5f | 3989 | continue; |
9d77a91f CY |
3990 | |
3991 | if (LOOKING_AT_NOCASE (dbp, "recursive")) | |
3992 | dbp = skip_spaces (dbp); | |
3993 | ||
d0417b4c DS |
3994 | if (LOOKING_AT_NOCASE (dbp, "pure")) |
3995 | dbp = skip_spaces (dbp); | |
3996 | ||
0f054abc GM |
3997 | if (LOOKING_AT_NOCASE (dbp, "elemental")) |
3998 | dbp = skip_spaces (dbp); | |
3999 | ||
79263656 | 4000 | switch (lowcase (*dbp)) |
c6d46f5f JB |
4001 | { |
4002 | case 'i': | |
eaaacacd FP |
4003 | if (nocase_tail ("integer")) |
4004 | F_takeprec (); | |
c6d46f5f JB |
4005 | break; |
4006 | case 'r': | |
eaaacacd FP |
4007 | if (nocase_tail ("real")) |
4008 | F_takeprec (); | |
c6d46f5f JB |
4009 | break; |
4010 | case 'l': | |
eaaacacd FP |
4011 | if (nocase_tail ("logical")) |
4012 | F_takeprec (); | |
c6d46f5f JB |
4013 | break; |
4014 | case 'c': | |
eaaacacd FP |
4015 | if (nocase_tail ("complex") || nocase_tail ("character")) |
4016 | F_takeprec (); | |
c6d46f5f JB |
4017 | break; |
4018 | case 'd': | |
eaaacacd | 4019 | if (nocase_tail ("double")) |
c6d46f5f | 4020 | { |
93b7ac65 | 4021 | dbp = skip_spaces (dbp); |
108c932a | 4022 | if (*dbp == '\0') |
c6d46f5f | 4023 | continue; |
eaaacacd | 4024 | if (nocase_tail ("precision")) |
c6d46f5f JB |
4025 | break; |
4026 | continue; | |
4027 | } | |
4028 | break; | |
4029 | } | |
93b7ac65 | 4030 | dbp = skip_spaces (dbp); |
108c932a | 4031 | if (*dbp == '\0') |
c6d46f5f | 4032 | continue; |
79263656 | 4033 | switch (lowcase (*dbp)) |
c6d46f5f JB |
4034 | { |
4035 | case 'f': | |
eaaacacd FP |
4036 | if (nocase_tail ("function")) |
4037 | F_getit (inf); | |
c6d46f5f JB |
4038 | continue; |
4039 | case 's': | |
eaaacacd FP |
4040 | if (nocase_tail ("subroutine")) |
4041 | F_getit (inf); | |
c6d46f5f | 4042 | continue; |
8a6c8bcf | 4043 | case 'e': |
eaaacacd FP |
4044 | if (nocase_tail ("entry")) |
4045 | F_getit (inf); | |
8a6c8bcf | 4046 | continue; |
15d5b889 | 4047 | case 'b': |
eaaacacd | 4048 | if (nocase_tail ("blockdata") || nocase_tail ("block data")) |
15d5b889 | 4049 | { |
93b7ac65 | 4050 | dbp = skip_spaces (dbp); |
15d5b889 | 4051 | if (*dbp == '\0') /* assume un-named */ |
081b7c70 FP |
4052 | make_tag ("blockdata", 9, TRUE, |
4053 | lb.buffer, dbp - lb.buffer, lineno, linecharno); | |
15d5b889 | 4054 | else |
eaaacacd | 4055 | F_getit (inf); /* look for name */ |
15d5b889 RS |
4056 | } |
4057 | continue; | |
c6d46f5f JB |
4058 | } |
4059 | } | |
c6d46f5f | 4060 | } |
b2521b0a | 4061 | |
6dd5561c | 4062 | \f |
93b7ac65 | 4063 | /* |
93b7ac65 | 4064 | * Ada parsing |
eaaacacd | 4065 | * Original code by |
73d3586d | 4066 | * Philippe Waroquiers (1998) |
93b7ac65 | 4067 | */ |
71cbb895 | 4068 | |
93b7ac65 FP |
4069 | /* Once we are positioned after an "interesting" keyword, let's get |
4070 | the real tag value necessary. */ | |
944bdd72 | 4071 | static void |
988e88ab | 4072 | Ada_getit (FILE *inf, const char *name_qualifier) |
93b7ac65 FP |
4073 | { |
4074 | register char *cp; | |
4075 | char *name; | |
4076 | char c; | |
4077 | ||
4078 | while (!feof (inf)) | |
4079 | { | |
4080 | dbp = skip_spaces (dbp); | |
4081 | if (*dbp == '\0' | |
4082 | || (dbp[0] == '-' && dbp[1] == '-')) | |
4083 | { | |
e7d3b099 | 4084 | readline (&lb, inf); |
93b7ac65 FP |
4085 | dbp = lb.buffer; |
4086 | } | |
5e617bc2 | 4087 | switch (lowcase (*dbp)) |
93b7ac65 FP |
4088 | { |
4089 | case 'b': | |
eaaacacd | 4090 | if (nocase_tail ("body")) |
93b7ac65 FP |
4091 | { |
4092 | /* Skipping body of procedure body or package body or .... | |
4093 | resetting qualifier to body instead of spec. */ | |
4094 | name_qualifier = "/b"; | |
4095 | continue; | |
4096 | } | |
4097 | break; | |
4098 | case 't': | |
93b7ac65 | 4099 | /* Skipping type of task type or protected type ... */ |
eaaacacd | 4100 | if (nocase_tail ("type")) |
93b7ac65 FP |
4101 | continue; |
4102 | break; | |
4103 | } | |
4104 | if (*dbp == '"') | |
4105 | { | |
4106 | dbp += 1; | |
4107 | for (cp = dbp; *cp != '\0' && *cp != '"'; cp++) | |
4108 | continue; | |
4109 | } | |
4110 | else | |
4111 | { | |
4112 | dbp = skip_spaces (dbp); | |
4113 | for (cp = dbp; | |
4114 | (*cp != '\0' | |
af03e6ab | 4115 | && (ISALPHA (*cp) || ISDIGIT (*cp) || *cp == '_' || *cp == '.')); |
93b7ac65 FP |
4116 | cp++) |
4117 | continue; | |
4118 | if (cp == dbp) | |
4119 | return; | |
4120 | } | |
4121 | c = *cp; | |
4122 | *cp = '\0'; | |
4123 | name = concat (dbp, name_qualifier, ""); | |
4124 | *cp = c; | |
081b7c70 FP |
4125 | make_tag (name, strlen (name), TRUE, |
4126 | lb.buffer, cp - lb.buffer + 1, lineno, linecharno); | |
4127 | free (name); | |
93b7ac65 FP |
4128 | if (c == '"') |
4129 | dbp = cp + 1; | |
4130 | return; | |
4131 | } | |
4132 | } | |
4133 | ||
944bdd72 | 4134 | static void |
873fbd0b | 4135 | Ada_funcs (FILE *inf) |
93b7ac65 FP |
4136 | { |
4137 | bool inquote = FALSE; | |
61a1f6fa | 4138 | bool skip_till_semicolumn = FALSE; |
93b7ac65 FP |
4139 | |
4140 | LOOP_ON_INPUT_LINES (inf, lb, dbp) | |
4141 | { | |
4142 | while (*dbp != '\0') | |
4143 | { | |
4144 | /* Skip a string i.e. "abcd". */ | |
4145 | if (inquote || (*dbp == '"')) | |
4146 | { | |
58cb46fb | 4147 | dbp = etags_strchr (dbp + !inquote, '"'); |
93b7ac65 FP |
4148 | if (dbp != NULL) |
4149 | { | |
4150 | inquote = FALSE; | |
4151 | dbp += 1; | |
4152 | continue; /* advance char */ | |
4153 | } | |
4154 | else | |
4155 | { | |
4156 | inquote = TRUE; | |
4157 | break; /* advance line */ | |
4158 | } | |
4159 | } | |
4160 | ||
4161 | /* Skip comments. */ | |
4162 | if (dbp[0] == '-' && dbp[1] == '-') | |
4163 | break; /* advance line */ | |
4164 | ||
4165 | /* Skip character enclosed in single quote i.e. 'a' | |
4166 | and skip single quote starting an attribute i.e. 'Image. */ | |
4167 | if (*dbp == '\'') | |
4168 | { | |
4169 | dbp++ ; | |
4170 | if (*dbp != '\0') | |
4171 | dbp++; | |
4172 | continue; | |
4173 | } | |
4174 | ||
61a1f6fa FP |
4175 | if (skip_till_semicolumn) |
4176 | { | |
4177 | if (*dbp == ';') | |
4178 | skip_till_semicolumn = FALSE; | |
4179 | dbp++; | |
4180 | continue; /* advance char */ | |
4181 | } | |
4182 | ||
93b7ac65 FP |
4183 | /* Search for beginning of a token. */ |
4184 | if (!begtoken (*dbp)) | |
4185 | { | |
4186 | dbp++; | |
4187 | continue; /* advance char */ | |
4188 | } | |
4189 | ||
4190 | /* We are at the beginning of a token. */ | |
5e617bc2 | 4191 | switch (lowcase (*dbp)) |
93b7ac65 FP |
4192 | { |
4193 | case 'f': | |
eaaacacd FP |
4194 | if (!packages_only && nocase_tail ("function")) |
4195 | Ada_getit (inf, "/f"); | |
93b7ac65 FP |
4196 | else |
4197 | break; /* from switch */ | |
4198 | continue; /* advance char */ | |
4199 | case 'p': | |
eaaacacd FP |
4200 | if (!packages_only && nocase_tail ("procedure")) |
4201 | Ada_getit (inf, "/p"); | |
4202 | else if (nocase_tail ("package")) | |
4203 | Ada_getit (inf, "/s"); | |
4204 | else if (nocase_tail ("protected")) /* protected type */ | |
4205 | Ada_getit (inf, "/t"); | |
93b7ac65 FP |
4206 | else |
4207 | break; /* from switch */ | |
4208 | continue; /* advance char */ | |
61a1f6fa FP |
4209 | |
4210 | case 'u': | |
4211 | if (typedefs && !packages_only && nocase_tail ("use")) | |
4212 | { | |
4213 | /* when tagging types, avoid tagging use type Pack.Typename; | |
4214 | for this, we will skip everything till a ; */ | |
4215 | skip_till_semicolumn = TRUE; | |
4216 | continue; /* advance char */ | |
4217 | } | |
4218 | ||
93b7ac65 | 4219 | case 't': |
eaaacacd FP |
4220 | if (!packages_only && nocase_tail ("task")) |
4221 | Ada_getit (inf, "/k"); | |
4222 | else if (typedefs && !packages_only && nocase_tail ("type")) | |
93b7ac65 | 4223 | { |
eaaacacd | 4224 | Ada_getit (inf, "/t"); |
93b7ac65 FP |
4225 | while (*dbp != '\0') |
4226 | dbp += 1; | |
4227 | } | |
4228 | else | |
4229 | break; /* from switch */ | |
4230 | continue; /* advance char */ | |
4231 | } | |
4232 | ||
4233 | /* Look for the end of the token. */ | |
4234 | while (!endtoken (*dbp)) | |
4235 | dbp++; | |
4236 | ||
4237 | } /* advance char */ | |
4238 | } /* advance line */ | |
4239 | } | |
b2521b0a | 4240 | |
93b7ac65 | 4241 | \f |
6dd5561c | 4242 | /* |
6dd5561c | 4243 | * Unix and microcontroller assembly tag handling |
eaaacacd FP |
4244 | * Labels: /^[a-zA-Z_.$][a-zA_Z0-9_.$]*[: ^I^J]/ |
4245 | * Idea by Bob Weiner, Motorola Inc. (1994) | |
6dd5561c | 4246 | */ |
944bdd72 | 4247 | static void |
873fbd0b | 4248 | Asm_labels (FILE *inf) |
c6d46f5f JB |
4249 | { |
4250 | register char *cp; | |
c6d46f5f | 4251 | |
93b7ac65 | 4252 | LOOP_ON_INPUT_LINES (inf, lb, cp) |
c6d46f5f | 4253 | { |
6dd5561c FP |
4254 | /* If first char is alphabetic or one of [_.$], test for colon |
4255 | following identifier. */ | |
af03e6ab | 4256 | if (ISALPHA (*cp) || *cp == '_' || *cp == '.' || *cp == '$') |
6dd5561c FP |
4257 | { |
4258 | /* Read past label. */ | |
4259 | cp++; | |
af03e6ab | 4260 | while (ISALNUM (*cp) || *cp == '_' || *cp == '.' || *cp == '$') |
6dd5561c | 4261 | cp++; |
71cbb895 | 4262 | if (*cp == ':' || iswhite (*cp)) |
081b7c70 FP |
4263 | /* Found end of label, so copy it and add it to the table. */ |
4264 | make_tag (lb.buffer, cp - lb.buffer, TRUE, | |
55597f90 | 4265 | lb.buffer, cp - lb.buffer + 1, lineno, linecharno); |
6dd5561c | 4266 | } |
c6d46f5f JB |
4267 | } |
4268 | } | |
b2521b0a | 4269 | |
c6d46f5f | 4270 | \f |
1f638249 | 4271 | /* |
b2521b0a | 4272 | * Perl support |
eaaacacd | 4273 | * Perl sub names: /^sub[ \t\n]+[^ \t\n{]+/ |
00054d21 | 4274 | * /^use constant[ \t\n]+[^ \t\n{=,;]+/ |
93b7ac65 | 4275 | * Perl variable names: /^(my|local).../ |
eaaacacd FP |
4276 | * Original code by Bart Robinson <lomew@cs.utah.edu> (1995) |
4277 | * Additions by Michael Ernst <mernst@alum.mit.edu> (1997) | |
c38e0c97 | 4278 | * Ideas by Kai Großjohann <Kai.Grossjohann@CS.Uni-Dortmund.DE> (2001) |
1f638249 | 4279 | */ |
944bdd72 | 4280 | static void |
873fbd0b | 4281 | Perl_functions (FILE *inf) |
1f638249 | 4282 | { |
eaaacacd | 4283 | char *package = savestr ("main"); /* current package name */ |
1f638249 FP |
4284 | register char *cp; |
4285 | ||
93b7ac65 | 4286 | LOOP_ON_INPUT_LINES (inf, lb, cp) |
1f638249 | 4287 | { |
1797886f | 4288 | cp = skip_spaces (cp); |
eaaacacd FP |
4289 | |
4290 | if (LOOKING_AT (cp, "package")) | |
93b7ac65 | 4291 | { |
eaaacacd | 4292 | free (package); |
081b7c70 | 4293 | get_tag (cp, &package); |
eaaacacd FP |
4294 | } |
4295 | else if (LOOKING_AT (cp, "sub")) | |
4296 | { | |
00054d21 | 4297 | char *pos, *sp; |
eaaacacd | 4298 | |
00054d21 KR |
4299 | subr: |
4300 | sp = cp; | |
eaaacacd FP |
4301 | while (!notinname (*cp)) |
4302 | cp++; | |
4303 | if (cp == sp) | |
081b7c70 FP |
4304 | continue; /* nothing found */ |
4305 | if ((pos = etags_strchr (sp, ':')) != NULL | |
4306 | && pos < cp && pos[1] == ':') | |
4307 | /* The name is already qualified. */ | |
4308 | make_tag (sp, cp - sp, TRUE, | |
4309 | lb.buffer, cp - lb.buffer + 1, lineno, linecharno); | |
eaaacacd | 4310 | else |
081b7c70 FP |
4311 | /* Qualify it. */ |
4312 | { | |
4313 | char savechar, *name; | |
4314 | ||
4315 | savechar = *cp; | |
4316 | *cp = '\0'; | |
4317 | name = concat (package, "::", sp); | |
4318 | *cp = savechar; | |
5e617bc2 | 4319 | make_tag (name, strlen (name), TRUE, |
081b7c70 FP |
4320 | lb.buffer, cp - lb.buffer + 1, lineno, linecharno); |
4321 | free (name); | |
4322 | } | |
00054d21 KR |
4323 | } |
4324 | else if (LOOKING_AT (cp, "use constant") | |
4325 | || LOOKING_AT (cp, "use constant::defer")) | |
4326 | { | |
4327 | /* For hash style multi-constant like | |
4328 | use constant { FOO => 123, | |
4329 | BAR => 456 }; | |
4330 | only the first FOO is picked up. Parsing across the value | |
4331 | expressions would be difficult in general, due to possible nested | |
4332 | hashes, here-documents, etc. */ | |
4333 | if (*cp == '{') | |
4334 | cp = skip_spaces (cp+1); | |
4335 | goto subr; | |
93b7ac65 | 4336 | } |
00054d21 | 4337 | else if (globals) /* only if we are tagging global vars */ |
93b7ac65 | 4338 | { |
081b7c70 FP |
4339 | /* Skip a qualifier, if any. */ |
4340 | bool qual = LOOKING_AT (cp, "my") || LOOKING_AT (cp, "local"); | |
93b7ac65 | 4341 | /* After "my" or "local", but before any following paren or space. */ |
081b7c70 | 4342 | char *varstart = cp; |
1f638249 | 4343 | |
081b7c70 FP |
4344 | if (qual /* should this be removed? If yes, how? */ |
4345 | && (*cp == '$' || *cp == '@' || *cp == '%')) | |
93b7ac65 | 4346 | { |
081b7c70 FP |
4347 | varstart += 1; |
4348 | do | |
93b7ac65 | 4349 | cp++; |
081b7c70 | 4350 | while (ISALNUM (*cp) || *cp == '_'); |
93b7ac65 | 4351 | } |
081b7c70 | 4352 | else if (qual) |
93b7ac65 FP |
4353 | { |
4354 | /* Should be examining a variable list at this point; | |
4355 | could insist on seeing an open parenthesis. */ | |
4356 | while (*cp != '\0' && *cp != ';' && *cp != '=' && *cp != ')') | |
4357 | cp++; | |
4358 | } | |
081b7c70 FP |
4359 | else |
4360 | continue; | |
93b7ac65 | 4361 | |
081b7c70 FP |
4362 | make_tag (varstart, cp - varstart, FALSE, |
4363 | lb.buffer, cp - lb.buffer + 1, lineno, linecharno); | |
93b7ac65 FP |
4364 | } |
4365 | } | |
7e237d24 | 4366 | free (package); |
93b7ac65 | 4367 | } |
b2521b0a | 4368 | |
62aec606 | 4369 | |
93b7ac65 | 4370 | /* |
b2521b0a | 4371 | * Python support |
49adb67a | 4372 | * Look for /^[\t]*def[ \t\n]+[^ \t\n(:]+/ or /^class[ \t\n]+[^ \t\n(:]+/ |
eaaacacd | 4373 | * Idea by Eric S. Raymond <esr@thyrsus.com> (1997) |
49adb67a | 4374 | * More ideas by seb bacon <seb@jamkit.com> (2002) |
93b7ac65 | 4375 | */ |
944bdd72 | 4376 | static void |
873fbd0b | 4377 | Python_functions (FILE *inf) |
93b7ac65 FP |
4378 | { |
4379 | register char *cp; | |
4380 | ||
62aec606 | 4381 | LOOP_ON_INPUT_LINES (inf, lb, cp) |
49adb67a FP |
4382 | { |
4383 | cp = skip_spaces (cp); | |
4384 | if (LOOKING_AT (cp, "def") || LOOKING_AT (cp, "class")) | |
4385 | { | |
4386 | char *name = cp; | |
4387 | while (!notinname (*cp) && *cp != ':') | |
4388 | cp++; | |
081b7c70 FP |
4389 | make_tag (name, cp - name, TRUE, |
4390 | lb.buffer, cp - lb.buffer + 1, lineno, linecharno); | |
49adb67a FP |
4391 | } |
4392 | } | |
62aec606 FP |
4393 | } |
4394 | ||
4395 | \f | |
4396 | /* | |
4397 | * PHP support | |
4398 | * Look for: | |
4399 | * - /^[ \t]*function[ \t\n]+[^ \t\n(]+/ | |
4400 | * - /^[ \t]*class[ \t\n]+[^ \t\n]+/ | |
4401 | * - /^[ \t]*define\(\"[^\"]+/ | |
4402 | * Only with --members: | |
4403 | * - /^[ \t]*var[ \t\n]+\$[^ \t\n=;]/ | |
eaaacacd | 4404 | * Idea by Diez B. Roggisch (2001) |
62aec606 FP |
4405 | */ |
4406 | static void | |
873fbd0b | 4407 | PHP_functions (FILE *inf) |
62aec606 | 4408 | { |
49adb67a | 4409 | register char *cp, *name; |
62aec606 FP |
4410 | bool search_identifier = FALSE; |
4411 | ||
93b7ac65 FP |
4412 | LOOP_ON_INPUT_LINES (inf, lb, cp) |
4413 | { | |
62aec606 | 4414 | cp = skip_spaces (cp); |
49adb67a | 4415 | name = cp; |
62aec606 FP |
4416 | if (search_identifier |
4417 | && *cp != '\0') | |
1f638249 | 4418 | { |
eaaacacd | 4419 | while (!notinname (*cp)) |
1f638249 | 4420 | cp++; |
081b7c70 FP |
4421 | make_tag (name, cp - name, TRUE, |
4422 | lb.buffer, cp - lb.buffer + 1, lineno, linecharno); | |
62aec606 | 4423 | search_identifier = FALSE; |
93b7ac65 | 4424 | } |
62aec606 FP |
4425 | else if (LOOKING_AT (cp, "function")) |
4426 | { | |
5e617bc2 | 4427 | if (*cp == '&') |
62aec606 | 4428 | cp = skip_spaces (cp+1); |
5e617bc2 | 4429 | if (*cp != '\0') |
62aec606 | 4430 | { |
49adb67a | 4431 | name = cp; |
eaaacacd | 4432 | while (!notinname (*cp)) |
62aec606 | 4433 | cp++; |
081b7c70 FP |
4434 | make_tag (name, cp - name, TRUE, |
4435 | lb.buffer, cp - lb.buffer + 1, lineno, linecharno); | |
62aec606 FP |
4436 | } |
4437 | else | |
4438 | search_identifier = TRUE; | |
4439 | } | |
4440 | else if (LOOKING_AT (cp, "class")) | |
4441 | { | |
4442 | if (*cp != '\0') | |
4443 | { | |
49adb67a | 4444 | name = cp; |
62aec606 FP |
4445 | while (*cp != '\0' && !iswhite (*cp)) |
4446 | cp++; | |
081b7c70 FP |
4447 | make_tag (name, cp - name, FALSE, |
4448 | lb.buffer, cp - lb.buffer + 1, lineno, linecharno); | |
62aec606 FP |
4449 | } |
4450 | else | |
4451 | search_identifier = TRUE; | |
4452 | } | |
4453 | else if (strneq (cp, "define", 6) | |
4454 | && (cp = skip_spaces (cp+6)) | |
4455 | && *cp++ == '(' | |
4456 | && (*cp == '"' || *cp == '\'')) | |
93b7ac65 | 4457 | { |
62aec606 | 4458 | char quote = *cp++; |
49adb67a | 4459 | name = cp; |
62aec606 | 4460 | while (*cp != quote && *cp != '\0') |
1f638249 | 4461 | cp++; |
081b7c70 FP |
4462 | make_tag (name, cp - name, FALSE, |
4463 | lb.buffer, cp - lb.buffer + 1, lineno, linecharno); | |
62aec606 FP |
4464 | } |
4465 | else if (members | |
4466 | && LOOKING_AT (cp, "var") | |
4467 | && *cp == '$') | |
4468 | { | |
49adb67a | 4469 | name = cp; |
5e617bc2 | 4470 | while (!notinname (*cp)) |
62aec606 | 4471 | cp++; |
081b7c70 FP |
4472 | make_tag (name, cp - name, FALSE, |
4473 | lb.buffer, cp - lb.buffer + 1, lineno, linecharno); | |
1f638249 FP |
4474 | } |
4475 | } | |
4476 | } | |
b2521b0a | 4477 | |
1f638249 | 4478 | \f |
eaaacacd | 4479 | /* |
30903246 FP |
4480 | * Cobol tag functions |
4481 | * We could look for anything that could be a paragraph name. | |
4482 | * i.e. anything that starts in column 8 is one word and ends in a full stop. | |
eaaacacd | 4483 | * Idea by Corny de Souza (1993) |
30903246 | 4484 | */ |
944bdd72 | 4485 | static void |
873fbd0b | 4486 | Cobol_paragraphs (FILE *inf) |
30903246 | 4487 | { |
93b7ac65 | 4488 | register char *bp, *ep; |
30903246 | 4489 | |
93b7ac65 | 4490 | LOOP_ON_INPUT_LINES (inf, lb, bp) |
30903246 | 4491 | { |
30903246 FP |
4492 | if (lb.len < 9) |
4493 | continue; | |
93b7ac65 | 4494 | bp += 8; |
30903246 FP |
4495 | |
4496 | /* If eoln, compiler option or comment ignore whole line. */ | |
af03e6ab | 4497 | if (bp[-1] != ' ' || !ISALNUM (bp[0])) |
30903246 FP |
4498 | continue; |
4499 | ||
af03e6ab | 4500 | for (ep = bp; ISALNUM (*ep) || *ep == '-'; ep++) |
30903246 | 4501 | continue; |
93b7ac65 | 4502 | if (*ep++ == '.') |
081b7c70 FP |
4503 | make_tag (bp, ep - bp, TRUE, |
4504 | lb.buffer, ep - lb.buffer + 1, lineno, linecharno); | |
30903246 FP |
4505 | } |
4506 | } | |
b2521b0a | 4507 | |
30903246 | 4508 | \f |
89d57763 FP |
4509 | /* |
4510 | * Makefile support | |
2b749964 | 4511 | * Ideas by Assar Westerlund <assar@sics.se> (2001) |
89d57763 FP |
4512 | */ |
4513 | static void | |
873fbd0b | 4514 | Makefile_targets (FILE *inf) |
89d57763 FP |
4515 | { |
4516 | register char *bp; | |
4517 | ||
4518 | LOOP_ON_INPUT_LINES (inf, lb, bp) | |
4519 | { | |
4520 | if (*bp == '\t' || *bp == '#') | |
4521 | continue; | |
4522 | while (*bp != '\0' && *bp != '=' && *bp != ':') | |
4523 | bp++; | |
2b749964 | 4524 | if (*bp == ':' || (globals && *bp == '=')) |
6598a3d4 FP |
4525 | { |
4526 | /* We should detect if there is more than one tag, but we do not. | |
4527 | We just skip initial and final spaces. */ | |
4528 | char * namestart = skip_spaces (lb.buffer); | |
4529 | while (--bp > namestart) | |
4530 | if (!notinname (*bp)) | |
4531 | break; | |
4532 | make_tag (namestart, bp - namestart + 1, TRUE, | |
4533 | lb.buffer, bp - lb.buffer + 2, lineno, linecharno); | |
4534 | } | |
89d57763 FP |
4535 | } |
4536 | } | |
b2521b0a | 4537 | |
89d57763 | 4538 | \f |
aab1fdae | 4539 | /* |
eaaacacd FP |
4540 | * Pascal parsing |
4541 | * Original code by Mosur K. Mohan (1989) | |
4542 | * | |
aab1fdae FP |
4543 | * Locates tags for procedures & functions. Doesn't do any type- or |
4544 | * var-definitions. It does look for the keyword "extern" or | |
4545 | * "forward" immediately following the procedure statement; if found, | |
4546 | * the tag is skipped. | |
c6d46f5f | 4547 | */ |
944bdd72 | 4548 | static void |
873fbd0b | 4549 | Pascal_functions (FILE *inf) |
c6d46f5f | 4550 | { |
93b7ac65 | 4551 | linebuffer tline; /* mostly copied from C_entries */ |
c6d46f5f | 4552 | long save_lcno; |
081b7c70 FP |
4553 | int save_lineno, namelen, taglen; |
4554 | char c, *name; | |
c6d46f5f | 4555 | |
79f94cd0 | 4556 | bool /* each of these flags is TRUE if: */ |
b9755a12 | 4557 | incomment, /* point is inside a comment */ |
c6d46f5f | 4558 | inquote, /* point is inside '..' string */ |
108c932a FP |
4559 | get_tagname, /* point is after PROCEDURE/FUNCTION |
4560 | keyword, so next item = potential tag */ | |
c6d46f5f JB |
4561 | found_tag, /* point is after a potential tag */ |
4562 | inparms, /* point is within parameter-list */ | |
108c932a FP |
4563 | verify_tag; /* point has passed the parm-list, so the |
4564 | next token will determine whether this | |
4565 | is a FORWARD/EXTERN to be ignored, or | |
4566 | whether it is a real tag */ | |
c6d46f5f | 4567 | |
081b7c70 FP |
4568 | save_lcno = save_lineno = namelen = taglen = 0; /* keep compiler quiet */ |
4569 | name = NULL; /* keep compiler quiet */ | |
c6d46f5f | 4570 | dbp = lb.buffer; |
108c932a | 4571 | *dbp = '\0'; |
61a1f6fa | 4572 | linebuffer_init (&tline); |
c6d46f5f | 4573 | |
b9755a12 | 4574 | incomment = inquote = FALSE; |
c6d46f5f | 4575 | found_tag = FALSE; /* have a proc name; check if extern */ |
081b7c70 | 4576 | get_tagname = FALSE; /* found "procedure" keyword */ |
c6d46f5f JB |
4577 | inparms = FALSE; /* found '(' after "proc" */ |
4578 | verify_tag = FALSE; /* check if "extern" is ahead */ | |
4579 | ||
93b7ac65 FP |
4580 | |
4581 | while (!feof (inf)) /* long main loop to get next char */ | |
c6d46f5f JB |
4582 | { |
4583 | c = *dbp++; | |
55597f90 | 4584 | if (c == '\0') /* if end of line */ |
c6d46f5f | 4585 | { |
e7d3b099 | 4586 | readline (&lb, inf); |
cf347d3c | 4587 | dbp = lb.buffer; |
55597f90 | 4588 | if (*dbp == '\0') |
c6d46f5f | 4589 | continue; |
93b7ac65 FP |
4590 | if (!((found_tag && verify_tag) |
4591 | || get_tagname)) | |
108c932a FP |
4592 | c = *dbp++; /* only if don't need *dbp pointing |
4593 | to the beginning of the name of | |
4594 | the procedure or function */ | |
c6d46f5f | 4595 | } |
b9755a12 | 4596 | if (incomment) |
c6d46f5f | 4597 | { |
108c932a | 4598 | if (c == '}') /* within { } comments */ |
b9755a12 | 4599 | incomment = FALSE; |
108c932a | 4600 | else if (c == '*' && *dbp == ')') /* within (* *) comments */ |
c6d46f5f | 4601 | { |
b9755a12 FP |
4602 | dbp++; |
4603 | incomment = FALSE; | |
c6d46f5f JB |
4604 | } |
4605 | continue; | |
4606 | } | |
4607 | else if (inquote) | |
4608 | { | |
4609 | if (c == '\'') | |
4610 | inquote = FALSE; | |
4611 | continue; | |
4612 | } | |
55597f90 | 4613 | else |
c6d46f5f JB |
4614 | switch (c) |
4615 | { | |
4616 | case '\'': | |
4617 | inquote = TRUE; /* found first quote */ | |
4618 | continue; | |
108c932a | 4619 | case '{': /* found open { comment */ |
b9755a12 | 4620 | incomment = TRUE; |
c6d46f5f JB |
4621 | continue; |
4622 | case '(': | |
108c932a | 4623 | if (*dbp == '*') /* found open (* comment */ |
c6d46f5f | 4624 | { |
b9755a12 | 4625 | incomment = TRUE; |
c6d46f5f JB |
4626 | dbp++; |
4627 | } | |
4628 | else if (found_tag) /* found '(' after tag, i.e., parm-list */ | |
4629 | inparms = TRUE; | |
4630 | continue; | |
4631 | case ')': /* end of parms list */ | |
4632 | if (inparms) | |
4633 | inparms = FALSE; | |
4634 | continue; | |
4635 | case ';': | |
108c932a | 4636 | if (found_tag && !inparms) /* end of proc or fn stmt */ |
c6d46f5f JB |
4637 | { |
4638 | verify_tag = TRUE; | |
4639 | break; | |
4640 | } | |
4641 | continue; | |
4642 | } | |
108c932a | 4643 | if (found_tag && verify_tag && (*dbp != ' ')) |
c6d46f5f | 4644 | { |
081b7c70 | 4645 | /* Check if this is an "extern" declaration. */ |
108c932a | 4646 | if (*dbp == '\0') |
c6d46f5f | 4647 | continue; |
8453bb55 | 4648 | if (lowcase (*dbp) == 'e') |
c6d46f5f | 4649 | { |
eaaacacd | 4650 | if (nocase_tail ("extern")) /* superfluous, really! */ |
c6d46f5f JB |
4651 | { |
4652 | found_tag = FALSE; | |
4653 | verify_tag = FALSE; | |
4654 | } | |
4655 | } | |
108c932a | 4656 | else if (lowcase (*dbp) == 'f') |
c6d46f5f | 4657 | { |
61a1f6fa | 4658 | if (nocase_tail ("forward")) /* check for forward reference */ |
c6d46f5f JB |
4659 | { |
4660 | found_tag = FALSE; | |
4661 | verify_tag = FALSE; | |
4662 | } | |
4663 | } | |
108c932a | 4664 | if (found_tag && verify_tag) /* not external proc, so make tag */ |
c6d46f5f JB |
4665 | { |
4666 | found_tag = FALSE; | |
4667 | verify_tag = FALSE; | |
081b7c70 FP |
4668 | make_tag (name, namelen, TRUE, |
4669 | tline.buffer, taglen, save_lineno, save_lcno); | |
c6d46f5f JB |
4670 | continue; |
4671 | } | |
4672 | } | |
4673 | if (get_tagname) /* grab name of proc or fn */ | |
4674 | { | |
081b7c70 FP |
4675 | char *cp; |
4676 | ||
108c932a | 4677 | if (*dbp == '\0') |
c6d46f5f JB |
4678 | continue; |
4679 | ||
081b7c70 FP |
4680 | /* Find block name. */ |
4681 | for (cp = dbp + 1; *cp != '\0' && !endtoken (*cp); cp++) | |
4682 | continue; | |
4683 | ||
4684 | /* Save all values for later tagging. */ | |
b2521b0a | 4685 | linebuffer_setlen (&tline, lb.len); |
c6d46f5f JB |
4686 | strcpy (tline.buffer, lb.buffer); |
4687 | save_lineno = lineno; | |
4688 | save_lcno = linecharno; | |
081b7c70 FP |
4689 | name = tline.buffer + (dbp - lb.buffer); |
4690 | namelen = cp - dbp; | |
4691 | taglen = cp - lb.buffer + 1; | |
c6d46f5f | 4692 | |
d8913c1c | 4693 | dbp = cp; /* set dbp to e-o-token */ |
c6d46f5f JB |
4694 | get_tagname = FALSE; |
4695 | found_tag = TRUE; | |
4696 | continue; | |
4697 | ||
081b7c70 | 4698 | /* And proceed to check for "extern". */ |
c6d46f5f | 4699 | } |
55597f90 | 4700 | else if (!incomment && !inquote && !found_tag) |
c6d46f5f | 4701 | { |
081b7c70 | 4702 | /* Check for proc/fn keywords. */ |
79263656 | 4703 | switch (lowcase (c)) |
c6d46f5f JB |
4704 | { |
4705 | case 'p': | |
eaaacacd | 4706 | if (nocase_tail ("rocedure")) /* c = 'p', dbp has advanced */ |
c6d46f5f JB |
4707 | get_tagname = TRUE; |
4708 | continue; | |
4709 | case 'f': | |
eaaacacd | 4710 | if (nocase_tail ("unction")) |
c6d46f5f JB |
4711 | get_tagname = TRUE; |
4712 | continue; | |
4713 | } | |
4714 | } | |
081b7c70 | 4715 | } /* while not eof */ |
c5007f46 | 4716 | |
108c932a | 4717 | free (tline.buffer); |
c6d46f5f | 4718 | } |
b2521b0a | 4719 | |
c6d46f5f JB |
4720 | \f |
4721 | /* | |
b2521b0a | 4722 | * Lisp tag functions |
55597f90 | 4723 | * look for (def or (DEF, quote or QUOTE |
c6d46f5f | 4724 | */ |
71cbb895 | 4725 | |
f57e2426 | 4726 | static void L_getit (void); |
c6d46f5f | 4727 | |
944bdd72 | 4728 | static void |
873fbd0b | 4729 | L_getit (void) |
c6d46f5f | 4730 | { |
31d4b314 FP |
4731 | if (*dbp == '\'') /* Skip prefix quote */ |
4732 | dbp++; | |
93b7ac65 | 4733 | else if (*dbp == '(') |
31d4b314 | 4734 | { |
eaaacacd FP |
4735 | dbp++; |
4736 | /* Try to skip "(quote " */ | |
4737 | if (!LOOKING_AT (dbp, "quote") && !LOOKING_AT (dbp, "QUOTE")) | |
4738 | /* Ok, then skip "(" before name in (defstruct (foo)) */ | |
4739 | dbp = skip_spaces (dbp); | |
31d4b314 | 4740 | } |
081b7c70 | 4741 | get_tag (dbp, NULL); |
c6d46f5f | 4742 | } |
6dd5561c | 4743 | |
944bdd72 | 4744 | static void |
873fbd0b | 4745 | Lisp_functions (FILE *inf) |
6dd5561c | 4746 | { |
93b7ac65 | 4747 | LOOP_ON_INPUT_LINES (inf, lb, dbp) |
6dd5561c | 4748 | { |
eaaacacd FP |
4749 | if (dbp[0] != '(') |
4750 | continue; | |
4751 | ||
1cbaa705 KR |
4752 | /* "(defvar foo)" is a declaration rather than a definition. */ |
4753 | if (! declarations) | |
4754 | { | |
4755 | char *p = dbp + 1; | |
4756 | if (LOOKING_AT (p, "defvar")) | |
4757 | { | |
4758 | p = skip_name (p); /* past var name */ | |
4759 | p = skip_spaces (p); | |
4760 | if (*p == ')') | |
4761 | continue; | |
4762 | } | |
4763 | } | |
4764 | ||
eaaacacd | 4765 | if (strneq (dbp+1, "def", 3) || strneq (dbp+1, "DEF", 3)) |
6dd5561c | 4766 | { |
eaaacacd FP |
4767 | dbp = skip_non_spaces (dbp); |
4768 | dbp = skip_spaces (dbp); | |
4769 | L_getit (); | |
4770 | } | |
4771 | else | |
4772 | { | |
4773 | /* Check for (foo::defmumble name-defined ... */ | |
4774 | do | |
4775 | dbp++; | |
4776 | while (!notinname (*dbp) && *dbp != ':'); | |
4777 | if (*dbp == ':') | |
6dd5561c | 4778 | { |
6dd5561c FP |
4779 | do |
4780 | dbp++; | |
eaaacacd | 4781 | while (*dbp == ':'); |
6dd5561c | 4782 | |
eaaacacd FP |
4783 | if (strneq (dbp, "def", 3) || strneq (dbp, "DEF", 3)) |
4784 | { | |
4785 | dbp = skip_non_spaces (dbp); | |
4786 | dbp = skip_spaces (dbp); | |
4787 | L_getit (); | |
6dd5561c FP |
4788 | } |
4789 | } | |
4790 | } | |
4791 | } | |
4792 | } | |
b2521b0a | 4793 | |
c6d46f5f | 4794 | \f |
05d9a399 FP |
4795 | /* |
4796 | * Lua script language parsing | |
4797 | * Original code by David A. Capello <dacap@users.sourceforge.net> (2004) | |
4798 | * | |
4799 | * "function" and "local function" are tags if they start at column 1. | |
4800 | */ | |
4801 | static void | |
873fbd0b | 4802 | Lua_functions (FILE *inf) |
05d9a399 FP |
4803 | { |
4804 | register char *bp; | |
4805 | ||
4806 | LOOP_ON_INPUT_LINES (inf, lb, bp) | |
4807 | { | |
4808 | if (bp[0] != 'f' && bp[0] != 'l') | |
4809 | continue; | |
4810 | ||
21cb180b | 4811 | (void)LOOKING_AT (bp, "local"); /* skip possible "local" */ |
05d9a399 FP |
4812 | |
4813 | if (LOOKING_AT (bp, "function")) | |
4814 | get_tag (bp, NULL); | |
4815 | } | |
4816 | } | |
4817 | ||
4818 | \f | |
cec68fcb | 4819 | /* |
7877f373 | 4820 | * PostScript tags |
cec68fcb | 4821 | * Just look for lines where the first character is '/' |
93b7ac65 | 4822 | * Also look at "defineps" for PSWrap |
eaaacacd FP |
4823 | * Ideas by: |
4824 | * Richard Mlynarik <mly@adoc.xerox.com> (1997) | |
4825 | * Masatake Yamato <masata-y@is.aist-nara.ac.jp> (1999) | |
cec68fcb | 4826 | */ |
944bdd72 | 4827 | static void |
873fbd0b | 4828 | PS_functions (FILE *inf) |
cec68fcb | 4829 | { |
93b7ac65 | 4830 | register char *bp, *ep; |
30903246 | 4831 | |
93b7ac65 | 4832 | LOOP_ON_INPUT_LINES (inf, lb, bp) |
cec68fcb | 4833 | { |
93b7ac65 | 4834 | if (bp[0] == '/') |
cec68fcb | 4835 | { |
93b7ac65 FP |
4836 | for (ep = bp+1; |
4837 | *ep != '\0' && *ep != ' ' && *ep != '{'; | |
4838 | ep++) | |
cec68fcb | 4839 | continue; |
081b7c70 FP |
4840 | make_tag (bp, ep - bp, TRUE, |
4841 | lb.buffer, ep - lb.buffer + 1, lineno, linecharno); | |
93b7ac65 | 4842 | } |
eaaacacd | 4843 | else if (LOOKING_AT (bp, "defineps")) |
081b7c70 | 4844 | get_tag (bp, NULL); |
cec68fcb FP |
4845 | } |
4846 | } | |
4847 | ||
4848 | \f | |
55102b5d FP |
4849 | /* |
4850 | * Forth tags | |
4851 | * Ignore anything after \ followed by space or in ( ) | |
4852 | * Look for words defined by : | |
4853 | * Look for constant, code, create, defer, value, and variable | |
4854 | * OBP extensions: Look for buffer:, field, | |
4855 | * Ideas by Eduardo Horvath <eeh@netbsd.org> (2004) | |
4856 | */ | |
4857 | static void | |
873fbd0b | 4858 | Forth_words (FILE *inf) |
55102b5d FP |
4859 | { |
4860 | register char *bp; | |
4861 | ||
4862 | LOOP_ON_INPUT_LINES (inf, lb, bp) | |
4863 | while ((bp = skip_spaces (bp))[0] != '\0') | |
5e617bc2 | 4864 | if (bp[0] == '\\' && iswhite (bp[1])) |
55102b5d | 4865 | break; /* read next line */ |
5e617bc2 | 4866 | else if (bp[0] == '(' && iswhite (bp[1])) |
55102b5d FP |
4867 | do /* skip to ) or eol */ |
4868 | bp++; | |
4869 | while (*bp != ')' && *bp != '\0'); | |
5e617bc2 | 4870 | else if ((bp[0] == ':' && iswhite (bp[1]) && bp++) |
55102b5d FP |
4871 | || LOOKING_AT_NOCASE (bp, "constant") |
4872 | || LOOKING_AT_NOCASE (bp, "code") | |
4873 | || LOOKING_AT_NOCASE (bp, "create") | |
4874 | || LOOKING_AT_NOCASE (bp, "defer") | |
4875 | || LOOKING_AT_NOCASE (bp, "value") | |
4876 | || LOOKING_AT_NOCASE (bp, "variable") | |
4877 | || LOOKING_AT_NOCASE (bp, "buffer:") | |
4878 | || LOOKING_AT_NOCASE (bp, "field")) | |
4879 | get_tag (skip_spaces (bp), NULL); /* Yay! A definition! */ | |
4880 | else | |
4881 | bp = skip_non_spaces (bp); | |
4882 | } | |
4883 | ||
4884 | \f | |
c6d46f5f JB |
4885 | /* |
4886 | * Scheme tag functions | |
4887 | * look for (def... xyzzy | |
eaaacacd FP |
4888 | * (def... (xyzzy |
4889 | * (def ... ((...(xyzzy .... | |
4890 | * (set! xyzzy | |
4891 | * Original code by Ken Haase (1985?) | |
c6d46f5f | 4892 | */ |
944bdd72 | 4893 | static void |
873fbd0b | 4894 | Scheme_functions (FILE *inf) |
c6d46f5f | 4895 | { |
93b7ac65 | 4896 | register char *bp; |
c6d46f5f | 4897 | |
93b7ac65 | 4898 | LOOP_ON_INPUT_LINES (inf, lb, bp) |
c6d46f5f | 4899 | { |
eaaacacd | 4900 | if (strneq (bp, "(def", 4) || strneq (bp, "(DEF", 4)) |
c6d46f5f | 4901 | { |
eaaacacd | 4902 | bp = skip_non_spaces (bp+4); |
f9b84f9f CY |
4903 | /* Skip over open parens and white space. Don't continue past |
4904 | '\0'. */ | |
4905 | while (*bp && notinname (*bp)) | |
93b7ac65 | 4906 | bp++; |
081b7c70 | 4907 | get_tag (bp, NULL); |
c6d46f5f | 4908 | } |
62aec606 | 4909 | if (LOOKING_AT (bp, "(SET!") || LOOKING_AT (bp, "(set!")) |
081b7c70 | 4910 | get_tag (bp, NULL); |
c6d46f5f JB |
4911 | } |
4912 | } | |
b2521b0a | 4913 | |
c6d46f5f JB |
4914 | \f |
4915 | /* Find tags in TeX and LaTeX input files. */ | |
4916 | ||
4917 | /* TEX_toktab is a table of TeX control sequences that define tags. | |
f6566f90 FP |
4918 | * Each entry records one such control sequence. |
4919 | * | |
4920 | * Original code from who knows whom. | |
4921 | * Ideas by: | |
4922 | * Stefan Monnier (2002) | |
4923 | */ | |
c6d46f5f | 4924 | |
f6566f90 | 4925 | static linebuffer *TEX_toktab = NULL; /* Table with tag tokens */ |
c6d46f5f JB |
4926 | |
4927 | /* Default set of control sequences to put into TEX_toktab. | |
4928 | The value of environment var TEXTAGS is prepended to this. */ | |
988e88ab | 4929 | static const char *TEX_defenv = "\ |
c5007f46 | 4930 | :chapter:section:subsection:subsubsection:eqno:label:ref:cite:bibitem\ |
f6566f90 FP |
4931 | :part:appendix:entry:index:def\ |
4932 | :newcommand:renewcommand:newenvironment:renewenvironment"; | |
c6d46f5f | 4933 | |
f57e2426 | 4934 | static void TEX_mode (FILE *); |
988e88ab | 4935 | static void TEX_decode_env (const char *, const char *); |
c6d46f5f | 4936 | |
db590582 FP |
4937 | static char TEX_esc = '\\'; |
4938 | static char TEX_opgrp = '{'; | |
4939 | static char TEX_clgrp = '}'; | |
c6d46f5f JB |
4940 | |
4941 | /* | |
4942 | * TeX/LaTeX scanning loop. | |
4943 | */ | |
944bdd72 | 4944 | static void |
873fbd0b | 4945 | TeX_commands (FILE *inf) |
c6d46f5f | 4946 | { |
f6566f90 FP |
4947 | char *cp; |
4948 | linebuffer *key; | |
c6d46f5f | 4949 | |
c6d46f5f | 4950 | /* Select either \ or ! as escape character. */ |
6dd5561c | 4951 | TEX_mode (inf); |
c6d46f5f JB |
4952 | |
4953 | /* Initialize token table once from environment. */ | |
f6566f90 FP |
4954 | if (TEX_toktab == NULL) |
4955 | TEX_decode_env ("TEXTAGS", TEX_defenv); | |
c6d46f5f | 4956 | |
93b7ac65 FP |
4957 | LOOP_ON_INPUT_LINES (inf, lb, cp) |
4958 | { | |
f6566f90 FP |
4959 | /* Look at each TEX keyword in line. */ |
4960 | for (;;) | |
8a6c8bcf | 4961 | { |
f6566f90 FP |
4962 | /* Look for a TEX escape. */ |
4963 | while (*cp++ != TEX_esc) | |
4964 | if (cp[-1] == '\0' || cp[-1] == '%') | |
4965 | goto tex_next_line; | |
4966 | ||
4967 | for (key = TEX_toktab; key->buffer != NULL; key++) | |
4968 | if (strneq (cp, key->buffer, key->len)) | |
4969 | { | |
4970 | register char *p; | |
081b7c70 | 4971 | int namelen, linelen; |
f6566f90 FP |
4972 | bool opgrp = FALSE; |
4973 | ||
4974 | cp = skip_spaces (cp + key->len); | |
4975 | if (*cp == TEX_opgrp) | |
4976 | { | |
4977 | opgrp = TRUE; | |
4978 | cp++; | |
4979 | } | |
4980 | for (p = cp; | |
4981 | (!iswhite (*p) && *p != '#' && | |
4982 | *p != TEX_opgrp && *p != TEX_clgrp); | |
4983 | p++) | |
4984 | continue; | |
081b7c70 | 4985 | namelen = p - cp; |
f6566f90 FP |
4986 | linelen = lb.len; |
4987 | if (!opgrp || *p == TEX_clgrp) | |
4988 | { | |
4989 | while (*p != '\0' && *p != TEX_opgrp && *p != TEX_clgrp) | |
21cb180b | 4990 | p++; |
f6566f90 FP |
4991 | linelen = p - lb.buffer + 1; |
4992 | } | |
081b7c70 FP |
4993 | make_tag (cp, namelen, TRUE, |
4994 | lb.buffer, linelen, lineno, linecharno); | |
f6566f90 FP |
4995 | goto tex_next_line; /* We only tag a line once */ |
4996 | } | |
c6d46f5f | 4997 | } |
f6566f90 FP |
4998 | tex_next_line: |
4999 | ; | |
c6d46f5f JB |
5000 | } |
5001 | } | |
5002 | ||
5003 | #define TEX_LESC '\\' | |
5004 | #define TEX_SESC '!' | |
c6d46f5f | 5005 | |
aab1fdae FP |
5006 | /* Figure out whether TeX's escapechar is '\\' or '!' and set grouping |
5007 | chars accordingly. */ | |
944bdd72 | 5008 | static void |
873fbd0b | 5009 | TEX_mode (FILE *inf) |
c6d46f5f JB |
5010 | { |
5011 | int c; | |
5012 | ||
6dd5561c | 5013 | while ((c = getc (inf)) != EOF) |
c6d46f5f JB |
5014 | { |
5015 | /* Skip to next line if we hit the TeX comment char. */ | |
f6566f90 | 5016 | if (c == '%') |
867cc23e | 5017 | while (c != '\n' && c != EOF) |
6dd5561c | 5018 | c = getc (inf); |
c6d46f5f JB |
5019 | else if (c == TEX_LESC || c == TEX_SESC ) |
5020 | break; | |
5021 | } | |
5022 | ||
5023 | if (c == TEX_LESC) | |
5024 | { | |
5025 | TEX_esc = TEX_LESC; | |
5026 | TEX_opgrp = '{'; | |
5027 | TEX_clgrp = '}'; | |
5028 | } | |
5029 | else | |
5030 | { | |
5031 | TEX_esc = TEX_SESC; | |
5032 | TEX_opgrp = '<'; | |
5033 | TEX_clgrp = '>'; | |
5034 | } | |
93b7ac65 FP |
5035 | /* If the input file is compressed, inf is a pipe, and rewind may fail. |
5036 | No attempt is made to correct the situation. */ | |
6dd5561c | 5037 | rewind (inf); |
c6d46f5f JB |
5038 | } |
5039 | ||
aab1fdae FP |
5040 | /* Read environment and prepend it to the default string. |
5041 | Build token table. */ | |
f6566f90 | 5042 | static void |
988e88ab | 5043 | TEX_decode_env (const char *evarname, const char *defenv) |
c6d46f5f | 5044 | { |
988e88ab | 5045 | register const char *env, *p; |
f6566f90 | 5046 | int i, len; |
c6d46f5f JB |
5047 | |
5048 | /* Append default string to environment. */ | |
5049 | env = getenv (evarname); | |
5050 | if (!env) | |
5051 | env = defenv; | |
5052 | else | |
988e88ab | 5053 | env = concat (env, defenv, ""); |
c6d46f5f JB |
5054 | |
5055 | /* Allocate a token table */ | |
f6566f90 | 5056 | for (len = 1, p = env; p;) |
93b7ac65 | 5057 | if ((p = etags_strchr (p, ':')) && *++p != '\0') |
f6566f90 FP |
5058 | len++; |
5059 | TEX_toktab = xnew (len, linebuffer); | |
c6d46f5f JB |
5060 | |
5061 | /* Unpack environment string into token table. Be careful about */ | |
5062 | /* zero-length strings (leading ':', "::" and trailing ':') */ | |
f6566f90 | 5063 | for (i = 0; *env != '\0';) |
c6d46f5f | 5064 | { |
b02c5fea | 5065 | p = etags_strchr (env, ':'); |
c6d46f5f JB |
5066 | if (!p) /* End of environment string. */ |
5067 | p = env + strlen (env); | |
5068 | if (p - env > 0) | |
5069 | { /* Only non-zero strings. */ | |
f6566f90 FP |
5070 | TEX_toktab[i].buffer = savenstr (env, p - env); |
5071 | TEX_toktab[i].len = p - env; | |
c6d46f5f JB |
5072 | i++; |
5073 | } | |
5074 | if (*p) | |
5075 | env = p + 1; | |
5076 | else | |
5077 | { | |
f6566f90 FP |
5078 | TEX_toktab[i].buffer = NULL; /* Mark end of table. */ |
5079 | TEX_toktab[i].len = 0; | |
c6d46f5f JB |
5080 | break; |
5081 | } | |
5082 | } | |
c6d46f5f | 5083 | } |
b2521b0a | 5084 | |
c6d46f5f | 5085 | \f |
97052c63 DL |
5086 | /* Texinfo support. Dave Love, Mar. 2000. */ |
5087 | static void | |
873fbd0b | 5088 | Texinfo_nodes (FILE *inf) |
97052c63 DL |
5089 | { |
5090 | char *cp, *start; | |
5091 | LOOP_ON_INPUT_LINES (inf, lb, cp) | |
62aec606 FP |
5092 | if (LOOKING_AT (cp, "@node")) |
5093 | { | |
5094 | start = cp; | |
5095 | while (*cp != '\0' && *cp != ',') | |
5096 | cp++; | |
081b7c70 FP |
5097 | make_tag (start, cp - start, TRUE, |
5098 | lb.buffer, cp - lb.buffer + 1, lineno, linecharno); | |
62aec606 | 5099 | } |
97052c63 | 5100 | } |
b2521b0a | 5101 | |
97052c63 | 5102 | \f |
61a1f6fa FP |
5103 | /* |
5104 | * HTML support. | |
5105 | * Contents of <title>, <h1>, <h2>, <h3> are tags. | |
5106 | * Contents of <a name=xxx> are tags with name xxx. | |
5107 | * | |
c38e0c97 | 5108 | * Francesco Potortì, 2002. |
61a1f6fa FP |
5109 | */ |
5110 | static void | |
873fbd0b | 5111 | HTML_labels (FILE *inf) |
61a1f6fa FP |
5112 | { |
5113 | bool getnext = FALSE; /* next text outside of HTML tags is a tag */ | |
3c04a71a FP |
5114 | bool skiptag = FALSE; /* skip to the end of the current HTML tag */ |
5115 | bool intag = FALSE; /* inside an html tag, looking for ID= */ | |
5116 | bool inanchor = FALSE; /* when INTAG, is an anchor, look for NAME= */ | |
61a1f6fa FP |
5117 | char *end; |
5118 | ||
5119 | ||
5120 | linebuffer_setlen (&token_name, 0); /* no name in buffer */ | |
5121 | ||
5122 | LOOP_ON_INPUT_LINES (inf, lb, dbp) | |
3c04a71a FP |
5123 | for (;;) /* loop on the same line */ |
5124 | { | |
5125 | if (skiptag) /* skip HTML tag */ | |
61a1f6fa FP |
5126 | { |
5127 | while (*dbp != '\0' && *dbp != '>') | |
5128 | dbp++; | |
5129 | if (*dbp == '>') | |
5130 | { | |
5131 | dbp += 1; | |
3c04a71a | 5132 | skiptag = FALSE; |
61a1f6fa FP |
5133 | continue; /* look on the same line */ |
5134 | } | |
5135 | break; /* go to next line */ | |
5136 | } | |
5137 | ||
3c04a71a | 5138 | else if (intag) /* look for "name=" or "id=" */ |
61a1f6fa | 5139 | { |
3c04a71a FP |
5140 | while (*dbp != '\0' && *dbp != '>' |
5141 | && lowcase (*dbp) != 'n' && lowcase (*dbp) != 'i') | |
61a1f6fa FP |
5142 | dbp++; |
5143 | if (*dbp == '\0') | |
5144 | break; /* go to next line */ | |
5145 | if (*dbp == '>') | |
5146 | { | |
5147 | dbp += 1; | |
3c04a71a | 5148 | intag = FALSE; |
61a1f6fa FP |
5149 | continue; /* look on the same line */ |
5150 | } | |
3c04a71a FP |
5151 | if ((inanchor && LOOKING_AT_NOCASE (dbp, "name=")) |
5152 | || LOOKING_AT_NOCASE (dbp, "id=")) | |
61a1f6fa FP |
5153 | { |
5154 | bool quoted = (dbp[0] == '"'); | |
5155 | ||
5156 | if (quoted) | |
5157 | for (end = ++dbp; *end != '\0' && *end != '"'; end++) | |
5158 | continue; | |
5159 | else | |
5160 | for (end = dbp; *end != '\0' && intoken (*end); end++) | |
5161 | continue; | |
5162 | linebuffer_setlen (&token_name, end - dbp); | |
e99a530f | 5163 | memcpy (token_name.buffer, dbp, end - dbp); |
61a1f6fa FP |
5164 | token_name.buffer[end - dbp] = '\0'; |
5165 | ||
5166 | dbp = end; | |
3c04a71a FP |
5167 | intag = FALSE; /* we found what we looked for */ |
5168 | skiptag = TRUE; /* skip to the end of the tag */ | |
61a1f6fa FP |
5169 | getnext = TRUE; /* then grab the text */ |
5170 | continue; /* look on the same line */ | |
5171 | } | |
3c04a71a | 5172 | dbp += 1; |
61a1f6fa FP |
5173 | } |
5174 | ||
5175 | else if (getnext) /* grab next tokens and tag them */ | |
5176 | { | |
5177 | dbp = skip_spaces (dbp); | |
5178 | if (*dbp == '\0') | |
5179 | break; /* go to next line */ | |
5180 | if (*dbp == '<') | |
5181 | { | |
3c04a71a FP |
5182 | intag = TRUE; |
5183 | inanchor = (lowcase (dbp[1]) == 'a' && !intoken (dbp[2])); | |
61a1f6fa FP |
5184 | continue; /* look on the same line */ |
5185 | } | |
5186 | ||
5187 | for (end = dbp + 1; *end != '\0' && *end != '<'; end++) | |
5188 | continue; | |
5189 | make_tag (token_name.buffer, token_name.len, TRUE, | |
5190 | dbp, end - dbp, lineno, linecharno); | |
5191 | linebuffer_setlen (&token_name, 0); /* no name in buffer */ | |
5192 | getnext = FALSE; | |
5193 | break; /* go to next line */ | |
5194 | } | |
5195 | ||
5196 | else /* look for an interesting HTML tag */ | |
5197 | { | |
5198 | while (*dbp != '\0' && *dbp != '<') | |
5199 | dbp++; | |
5200 | if (*dbp == '\0') | |
5201 | break; /* go to next line */ | |
3c04a71a FP |
5202 | intag = TRUE; |
5203 | if (lowcase (dbp[1]) == 'a' && !intoken (dbp[2])) | |
61a1f6fa | 5204 | { |
3c04a71a FP |
5205 | inanchor = TRUE; |
5206 | continue; /* look on the same line */ | |
5207 | } | |
5208 | else if (LOOKING_AT_NOCASE (dbp, "<title>") | |
5209 | || LOOKING_AT_NOCASE (dbp, "<h1>") | |
5210 | || LOOKING_AT_NOCASE (dbp, "<h2>") | |
5211 | || LOOKING_AT_NOCASE (dbp, "<h3>")) | |
5212 | { | |
5213 | intag = FALSE; | |
61a1f6fa FP |
5214 | getnext = TRUE; |
5215 | continue; /* look on the same line */ | |
5216 | } | |
3c04a71a | 5217 | dbp += 1; |
61a1f6fa | 5218 | } |
3c04a71a | 5219 | } |
61a1f6fa FP |
5220 | } |
5221 | ||
5222 | \f | |
8dc7496c | 5223 | /* |
eaaacacd | 5224 | * Prolog support |
8dc7496c | 5225 | * |
eaaacacd FP |
5226 | * Assumes that the predicate or rule starts at column 0. |
5227 | * Only the first clause of a predicate or rule is added. | |
5228 | * Original code by Sunichirou Sugou (1989) | |
5229 | * Rewritten by Anders Lindgren (1996) | |
8dc7496c | 5230 | */ |
3ef271f2 | 5231 | static size_t prolog_pr (char *, char *); |
f57e2426 | 5232 | static void prolog_skip_comment (linebuffer *, FILE *); |
3ef271f2 | 5233 | static size_t prolog_atom (char *, size_t); |
30903246 | 5234 | |
944bdd72 | 5235 | static void |
873fbd0b | 5236 | Prolog_functions (FILE *inf) |
8dc7496c | 5237 | { |
93b7ac65 | 5238 | char *cp, *last; |
3ef271f2 PE |
5239 | size_t len; |
5240 | size_t allocated; | |
8dc7496c FP |
5241 | |
5242 | allocated = 0; | |
5243 | len = 0; | |
5244 | last = NULL; | |
5245 | ||
93b7ac65 | 5246 | LOOP_ON_INPUT_LINES (inf, lb, cp) |
8dc7496c | 5247 | { |
93b7ac65 | 5248 | if (cp[0] == '\0') /* Empty line */ |
8dc7496c | 5249 | continue; |
71cbb895 | 5250 | else if (iswhite (cp[0])) /* Not a predicate */ |
8dc7496c | 5251 | continue; |
93b7ac65 | 5252 | else if (cp[0] == '/' && cp[1] == '*') /* comment. */ |
3f1c8fcd | 5253 | prolog_skip_comment (&lb, inf); |
62aec606 | 5254 | else if ((len = prolog_pr (cp, last)) > 0) |
8dc7496c | 5255 | { |
eaaacacd FP |
5256 | /* Predicate or rule. Store the function name so that we |
5257 | only generate a tag for the first clause. */ | |
8dc7496c | 5258 | if (last == NULL) |
5e617bc2 | 5259 | last = xnew (len + 1, char); |
8dc7496c | 5260 | else if (len + 1 > allocated) |
b2521b0a | 5261 | xrnew (last, len + 1, char); |
8dc7496c | 5262 | allocated = len + 1; |
e99a530f | 5263 | memcpy (last, cp, len); |
8dc7496c FP |
5264 | last[len] = '\0'; |
5265 | } | |
5266 | } | |
c2cd06e6 | 5267 | free (last); |
8dc7496c FP |
5268 | } |
5269 | ||
c6d46f5f | 5270 | |
944bdd72 | 5271 | static void |
873fbd0b | 5272 | prolog_skip_comment (linebuffer *plb, FILE *inf) |
8dc7496c FP |
5273 | { |
5274 | char *cp; | |
5275 | ||
5276 | do | |
5277 | { | |
5278 | for (cp = plb->buffer; *cp != '\0'; cp++) | |
5279 | if (cp[0] == '*' && cp[1] == '/') | |
5280 | return; | |
e7d3b099 | 5281 | readline (plb, inf); |
8dc7496c | 5282 | } |
5e617bc2 | 5283 | while (!feof (inf)); |
8dc7496c FP |
5284 | } |
5285 | ||
5286 | /* | |
62aec606 | 5287 | * A predicate or rule definition is added if it matches: |
8dc7496c | 5288 | * <beginning of line><Prolog Atom><whitespace>( |
62aec606 | 5289 | * or <beginning of line><Prolog Atom><whitespace>:- |
8dc7496c FP |
5290 | * |
5291 | * It is added to the tags database if it doesn't match the | |
5292 | * name of the previous clause header. | |
5293 | * | |
62aec606 FP |
5294 | * Return the size of the name of the predicate or rule, or 0 if no |
5295 | * header was found. | |
8dc7496c | 5296 | */ |
3ef271f2 | 5297 | static size_t |
873fbd0b | 5298 | prolog_pr (char *s, char *last) |
f68c809d | 5299 | |
873fbd0b | 5300 | /* Name of last clause. */ |
c6d46f5f | 5301 | { |
3ef271f2 PE |
5302 | size_t pos; |
5303 | size_t len; | |
8dc7496c | 5304 | |
30903246 | 5305 | pos = prolog_atom (s, 0); |
3ef271f2 | 5306 | if (! pos) |
8dc7496c FP |
5307 | return 0; |
5308 | ||
5309 | len = pos; | |
93b7ac65 | 5310 | pos = skip_spaces (s + pos) - s; |
8dc7496c | 5311 | |
62aec606 FP |
5312 | if ((s[pos] == '.' |
5313 | || (s[pos] == '(' && (pos += 1)) | |
5314 | || (s[pos] == ':' && s[pos + 1] == '-' && (pos += 2))) | |
5315 | && (last == NULL /* save only the first clause */ | |
3ef271f2 | 5316 | || len != strlen (last) |
62aec606 | 5317 | || !strneq (s, last, len))) |
c6d46f5f | 5318 | { |
081b7c70 | 5319 | make_tag (s, len, TRUE, s, pos, lineno, linecharno); |
8dc7496c | 5320 | return len; |
c6d46f5f | 5321 | } |
62aec606 FP |
5322 | else |
5323 | return 0; | |
8dc7496c FP |
5324 | } |
5325 | ||
5326 | /* | |
5327 | * Consume a Prolog atom. | |
3ef271f2 | 5328 | * Return the number of bytes consumed, or 0 if there was an error. |
8dc7496c FP |
5329 | * |
5330 | * A prolog atom, in this context, could be one of: | |
5331 | * - An alphanumeric sequence, starting with a lower case letter. | |
5332 | * - A quoted arbitrary string. Single quotes can escape themselves. | |
5333 | * Backslash quotes everything. | |
5334 | */ | |
3ef271f2 PE |
5335 | static size_t |
5336 | prolog_atom (char *s, size_t pos) | |
8dc7496c | 5337 | { |
3ef271f2 | 5338 | size_t origpos; |
8dc7496c FP |
5339 | |
5340 | origpos = pos; | |
5341 | ||
5e617bc2 | 5342 | if (ISLOWER (s[pos]) || (s[pos] == '_')) |
8dc7496c FP |
5343 | { |
5344 | /* The atom is unquoted. */ | |
5345 | pos++; | |
5e617bc2 | 5346 | while (ISALNUM (s[pos]) || (s[pos] == '_')) |
c6d46f5f | 5347 | { |
8dc7496c | 5348 | pos++; |
c6d46f5f | 5349 | } |
8dc7496c FP |
5350 | return pos - origpos; |
5351 | } | |
5352 | else if (s[pos] == '\'') | |
5353 | { | |
5354 | pos++; | |
5355 | ||
97b90b0a | 5356 | for (;;) |
c6d46f5f | 5357 | { |
8dc7496c FP |
5358 | if (s[pos] == '\'') |
5359 | { | |
5360 | pos++; | |
5361 | if (s[pos] != '\'') | |
5362 | break; | |
5363 | pos++; /* A double quote */ | |
5364 | } | |
5365 | else if (s[pos] == '\0') | |
5366 | /* Multiline quoted atoms are ignored. */ | |
3ef271f2 | 5367 | return 0; |
8dc7496c FP |
5368 | else if (s[pos] == '\\') |
5369 | { | |
5370 | if (s[pos+1] == '\0') | |
3ef271f2 | 5371 | return 0; |
8dc7496c FP |
5372 | pos += 2; |
5373 | } | |
5374 | else | |
5375 | pos++; | |
c6d46f5f | 5376 | } |
8dc7496c | 5377 | return pos - origpos; |
c6d46f5f | 5378 | } |
8dc7496c | 5379 | else |
3ef271f2 | 5380 | return 0; |
c6d46f5f | 5381 | } |
b2521b0a | 5382 | |
8dc7496c | 5383 | \f |
93b7ac65 | 5384 | /* |
eaaacacd | 5385 | * Support for Erlang |
8dc7496c FP |
5386 | * |
5387 | * Generates tags for functions, defines, and records. | |
8dc7496c | 5388 | * Assumes that Erlang functions start at column 0. |
eaaacacd | 5389 | * Original code by Anders Lindgren (1996) |
8dc7496c | 5390 | */ |
f57e2426 J |
5391 | static int erlang_func (char *, char *); |
5392 | static void erlang_attribute (char *); | |
5393 | static int erlang_atom (char *); | |
30903246 | 5394 | |
944bdd72 | 5395 | static void |
873fbd0b | 5396 | Erlang_functions (FILE *inf) |
c6d46f5f | 5397 | { |
93b7ac65 | 5398 | char *cp, *last; |
8dc7496c FP |
5399 | int len; |
5400 | int allocated; | |
5401 | ||
5402 | allocated = 0; | |
5403 | len = 0; | |
5404 | last = NULL; | |
5405 | ||
93b7ac65 | 5406 | LOOP_ON_INPUT_LINES (inf, lb, cp) |
c6d46f5f | 5407 | { |
93b7ac65 | 5408 | if (cp[0] == '\0') /* Empty line */ |
c6d46f5f | 5409 | continue; |
71cbb895 | 5410 | else if (iswhite (cp[0])) /* Not function nor attribute */ |
c6d46f5f | 5411 | continue; |
93b7ac65 | 5412 | else if (cp[0] == '%') /* comment */ |
8dc7496c | 5413 | continue; |
93b7ac65 | 5414 | else if (cp[0] == '"') /* Sometimes, strings start in column one */ |
8dc7496c | 5415 | continue; |
93b7ac65 | 5416 | else if (cp[0] == '-') /* attribute, e.g. "-define" */ |
8dc7496c | 5417 | { |
93b7ac65 | 5418 | erlang_attribute (cp); |
7e237d24 FP |
5419 | if (last != NULL) |
5420 | { | |
5421 | free (last); | |
5422 | last = NULL; | |
5423 | } | |
8dc7496c | 5424 | } |
93b7ac65 | 5425 | else if ((len = erlang_func (cp, last)) > 0) |
8dc7496c | 5426 | { |
93b7ac65 | 5427 | /* |
8dc7496c FP |
5428 | * Function. Store the function name so that we only |
5429 | * generates a tag for the first clause. | |
5430 | */ | |
5431 | if (last == NULL) | |
30903246 | 5432 | last = xnew (len + 1, char); |
8dc7496c | 5433 | else if (len + 1 > allocated) |
b2521b0a | 5434 | xrnew (last, len + 1, char); |
8dc7496c | 5435 | allocated = len + 1; |
e99a530f | 5436 | memcpy (last, cp, len); |
8dc7496c FP |
5437 | last[len] = '\0'; |
5438 | } | |
5439 | } | |
c2cd06e6 | 5440 | free (last); |
8dc7496c FP |
5441 | } |
5442 | ||
5443 | ||
5444 | /* | |
5445 | * A function definition is added if it matches: | |
5446 | * <beginning of line><Erlang Atom><whitespace>( | |
5447 | * | |
5448 | * It is added to the tags database if it doesn't match the | |
5449 | * name of the previous clause header. | |
5450 | * | |
5451 | * Return the size of the name of the function, or 0 if no function | |
5452 | * was found. | |
5453 | */ | |
0dacfc4b | 5454 | static int |
873fbd0b | 5455 | erlang_func (char *s, char *last) |
f68c809d | 5456 | |
873fbd0b | 5457 | /* Name of last clause. */ |
8dc7496c | 5458 | { |
8dc7496c FP |
5459 | int pos; |
5460 | int len; | |
5461 | ||
89fb2be1 | 5462 | pos = erlang_atom (s); |
8dc7496c FP |
5463 | if (pos < 1) |
5464 | return 0; | |
5465 | ||
5466 | len = pos; | |
93b7ac65 | 5467 | pos = skip_spaces (s + pos) - s; |
8dc7496c | 5468 | |
30903246 FP |
5469 | /* Save only the first clause. */ |
5470 | if (s[pos++] == '(' | |
5471 | && (last == NULL | |
93b7ac65 | 5472 | || len != (int)strlen (last) |
30903246 | 5473 | || !strneq (s, last, len))) |
8dc7496c | 5474 | { |
081b7c70 | 5475 | make_tag (s, len, TRUE, s, pos, lineno, linecharno); |
8dc7496c FP |
5476 | return len; |
5477 | } | |
30903246 | 5478 | |
8dc7496c | 5479 | return 0; |
c6d46f5f JB |
5480 | } |
5481 | ||
8dc7496c FP |
5482 | |
5483 | /* | |
93b7ac65 | 5484 | * Handle attributes. Currently, tags are generated for defines |
8dc7496c FP |
5485 | * and records. |
5486 | * | |
5487 | * They are on the form: | |
5488 | * -define(foo, bar). | |
5489 | * -define(Foo(M, N), M+N). | |
5490 | * -record(graph, {vtab = notable, cyclic = true}). | |
5491 | */ | |
944bdd72 | 5492 | static void |
873fbd0b | 5493 | erlang_attribute (char *s) |
c6d46f5f | 5494 | { |
89fb2be1 | 5495 | char *cp = s; |
8dc7496c | 5496 | |
89fb2be1 FP |
5497 | if ((LOOKING_AT (cp, "-define") || LOOKING_AT (cp, "-record")) |
5498 | && *cp++ == '(') | |
c6d46f5f | 5499 | { |
89fb2be1 FP |
5500 | int len = erlang_atom (skip_spaces (cp)); |
5501 | if (len > 0) | |
081b7c70 | 5502 | make_tag (cp, len, TRUE, s, cp + len - s, lineno, linecharno); |
b9755a12 | 5503 | } |
8dc7496c FP |
5504 | return; |
5505 | } | |
5506 | ||
5507 | ||
5508 | /* | |
5509 | * Consume an Erlang atom (or variable). | |
5510 | * Return the number of bytes consumed, or -1 if there was an error. | |
5511 | */ | |
944bdd72 | 5512 | static int |
873fbd0b | 5513 | erlang_atom (char *s) |
8dc7496c | 5514 | { |
89fb2be1 | 5515 | int pos = 0; |
8dc7496c | 5516 | |
af03e6ab | 5517 | if (ISALPHA (s[pos]) || s[pos] == '_') |
8dc7496c FP |
5518 | { |
5519 | /* The atom is unquoted. */ | |
89fb2be1 | 5520 | do |
8dc7496c | 5521 | pos++; |
89fb2be1 | 5522 | while (ISALNUM (s[pos]) || s[pos] == '_'); |
8dc7496c FP |
5523 | } |
5524 | else if (s[pos] == '\'') | |
5525 | { | |
89fb2be1 FP |
5526 | for (pos++; s[pos] != '\''; pos++) |
5527 | if (s[pos] == '\0' /* multiline quoted atoms are ignored */ | |
5528 | || (s[pos] == '\\' && s[++pos] == '\0')) | |
5529 | return 0; | |
8dc7496c | 5530 | pos++; |
8dc7496c | 5531 | } |
89fb2be1 FP |
5532 | |
5533 | return pos; | |
8dc7496c | 5534 | } |
b2521b0a | 5535 | |
b9755a12 | 5536 | \f |
f57e2426 J |
5537 | static char *scan_separators (char *); |
5538 | static void add_regex (char *, language *); | |
5539 | static char *substitute (char *, char *, struct re_registers *); | |
71cbb895 | 5540 | |
24dbe96a FP |
5541 | /* |
5542 | * Take a string like "/blah/" and turn it into "blah", verifying | |
5543 | * that the first and last characters are the same, and handling | |
5544 | * quoted separator characters. Actually, stops on the occurrence of | |
5545 | * an unquoted separator. Also process \t, \n, etc. and turn into | |
5546 | * appropriate characters. Works in place. Null terminates name string. | |
5547 | * Returns pointer to terminating separator, or NULL for | |
5548 | * unterminated regexps. | |
5549 | */ | |
944bdd72 | 5550 | static char * |
873fbd0b | 5551 | scan_separators (char *name) |
b9755a12 FP |
5552 | { |
5553 | char sep = name[0]; | |
5554 | char *copyto = name; | |
30903246 | 5555 | bool quoted = FALSE; |
b9755a12 FP |
5556 | |
5557 | for (++name; *name != '\0'; ++name) | |
5558 | { | |
5559 | if (quoted) | |
5560 | { | |
e7d3b099 | 5561 | switch (*name) |
b9755a12 | 5562 | { |
24dbe96a FP |
5563 | case 'a': *copyto++ = '\007'; break; /* BEL (bell) */ |
5564 | case 'b': *copyto++ = '\b'; break; /* BS (back space) */ | |
5565 | case 'd': *copyto++ = 0177; break; /* DEL (delete) */ | |
5566 | case 'e': *copyto++ = 033; break; /* ESC (delete) */ | |
5567 | case 'f': *copyto++ = '\f'; break; /* FF (form feed) */ | |
5568 | case 'n': *copyto++ = '\n'; break; /* NL (new line) */ | |
5569 | case 'r': *copyto++ = '\r'; break; /* CR (carriage return) */ | |
5570 | case 't': *copyto++ = '\t'; break; /* TAB (horizontal tab) */ | |
5571 | case 'v': *copyto++ = '\v'; break; /* VT (vertical tab) */ | |
e7d3b099 FP |
5572 | default: |
5573 | if (*name == sep) | |
5574 | *copyto++ = sep; | |
5575 | else | |
5576 | { | |
5577 | /* Something else is quoted, so preserve the quote. */ | |
5578 | *copyto++ = '\\'; | |
5579 | *copyto++ = *name; | |
5580 | } | |
5581 | break; | |
b9755a12 FP |
5582 | } |
5583 | quoted = FALSE; | |
5584 | } | |
5585 | else if (*name == '\\') | |
5586 | quoted = TRUE; | |
5587 | else if (*name == sep) | |
5588 | break; | |
5589 | else | |
5590 | *copyto++ = *name; | |
5591 | } | |
24dbe96a FP |
5592 | if (*name != sep) |
5593 | name = NULL; /* signal unterminated regexp */ | |
c6d46f5f | 5594 | |
b9755a12 FP |
5595 | /* Terminate copied string. */ |
5596 | *copyto = '\0'; | |
5597 | return name; | |
5598 | } | |
c6d46f5f | 5599 | |
cec68fcb | 5600 | /* Look at the argument of --regex or --no-regex and do the right |
93b7ac65 | 5601 | thing. Same for each line of a regexp file. */ |
944bdd72 | 5602 | static void |
873fbd0b | 5603 | analyse_regex (char *regex_arg) |
c6d46f5f | 5604 | { |
cec68fcb | 5605 | if (regex_arg == NULL) |
76fecdff | 5606 | { |
61a1f6fa | 5607 | free_regexps (); /* --no-regex: remove existing regexps */ |
76fecdff FP |
5608 | return; |
5609 | } | |
93b7ac65 FP |
5610 | |
5611 | /* A real --regexp option or a line in a regexp file. */ | |
5612 | switch (regex_arg[0]) | |
b9755a12 | 5613 | { |
93b7ac65 FP |
5614 | /* Comments in regexp file or null arg to --regex. */ |
5615 | case '\0': | |
5616 | case ' ': | |
5617 | case '\t': | |
5618 | break; | |
cec68fcb | 5619 | |
93b7ac65 FP |
5620 | /* Read a regex file. This is recursive and may result in a |
5621 | loop, which will stop when the file descriptors are exhausted. */ | |
5622 | case '@': | |
5623 | { | |
5624 | FILE *regexfp; | |
5625 | linebuffer regexbuf; | |
5626 | char *regexfile = regex_arg + 1; | |
5627 | ||
5628 | /* regexfile is a file containing regexps, one per line. */ | |
5629 | regexfp = fopen (regexfile, "r"); | |
5630 | if (regexfp == NULL) | |
d9df6f40 | 5631 | pfatal (regexfile); |
61a1f6fa | 5632 | linebuffer_init (®exbuf); |
93b7ac65 | 5633 | while (readline_internal (®exbuf, regexfp) > 0) |
24dbe96a | 5634 | analyse_regex (regexbuf.buffer); |
93b7ac65 FP |
5635 | free (regexbuf.buffer); |
5636 | fclose (regexfp); | |
5637 | } | |
5638 | break; | |
5639 | ||
5640 | /* Regexp to be used for a specific language only. */ | |
5641 | case '{': | |
5642 | { | |
5643 | language *lang; | |
5644 | char *lang_name = regex_arg + 1; | |
5645 | char *cp; | |
5646 | ||
5647 | for (cp = lang_name; *cp != '}'; cp++) | |
5648 | if (*cp == '\0') | |
5649 | { | |
5650 | error ("unterminated language name in regex: %s", regex_arg); | |
5651 | return; | |
5652 | } | |
24dbe96a | 5653 | *cp++ = '\0'; |
89d57763 | 5654 | lang = get_language_from_langname (lang_name); |
93b7ac65 | 5655 | if (lang == NULL) |
cec68fcb | 5656 | return; |
24dbe96a | 5657 | add_regex (cp, lang); |
93b7ac65 FP |
5658 | } |
5659 | break; | |
5660 | ||
5661 | /* Regexp to be used for any language. */ | |
5662 | default: | |
24dbe96a | 5663 | add_regex (regex_arg, NULL); |
93b7ac65 | 5664 | break; |
cec68fcb FP |
5665 | } |
5666 | } | |
5667 | ||
61a1f6fa FP |
5668 | /* Separate the regexp pattern, compile it, |
5669 | and care for optional name and modifiers. */ | |
944bdd72 | 5670 | static void |
873fbd0b | 5671 | add_regex (char *regexp_pattern, language *lang) |
cec68fcb | 5672 | { |
7e88eb25 | 5673 | static struct re_pattern_buffer zeropattern; |
24dbe96a | 5674 | char sep, *pat, *name, *modifiers; |
db5a3003 | 5675 | char empty = '\0'; |
cec68fcb FP |
5676 | const char *err; |
5677 | struct re_pattern_buffer *patbuf; | |
61a1f6fa FP |
5678 | regexp *rp; |
5679 | bool | |
5680 | force_explicit_name = TRUE, /* do not use implicit tag names */ | |
5681 | ignore_case = FALSE, /* case is significant */ | |
5682 | multi_line = FALSE, /* matches are done one line at a time */ | |
5683 | single_line = FALSE; /* dot does not match newline */ | |
cec68fcb FP |
5684 | |
5685 | ||
5e617bc2 | 5686 | if (strlen (regexp_pattern) < 3) |
b9755a12 | 5687 | { |
db5a3003 | 5688 | error ("null regexp"); |
b9755a12 FP |
5689 | return; |
5690 | } | |
24dbe96a | 5691 | sep = regexp_pattern[0]; |
b9755a12 | 5692 | name = scan_separators (regexp_pattern); |
24dbe96a | 5693 | if (name == NULL) |
b9755a12 | 5694 | { |
24dbe96a FP |
5695 | error ("%s: unterminated regexp", regexp_pattern); |
5696 | return; | |
5697 | } | |
5698 | if (name[1] == sep) | |
5699 | { | |
5700 | error ("null name for regexp \"%s\"", regexp_pattern); | |
b9755a12 FP |
5701 | return; |
5702 | } | |
24dbe96a FP |
5703 | modifiers = scan_separators (name); |
5704 | if (modifiers == NULL) /* no terminating separator --> no name */ | |
5705 | { | |
5706 | modifiers = name; | |
db5a3003 | 5707 | name = ∅ |
24dbe96a FP |
5708 | } |
5709 | else | |
5710 | modifiers += 1; /* skip separator */ | |
5711 | ||
5712 | /* Parse regex modifiers. */ | |
24dbe96a FP |
5713 | for (; modifiers[0] != '\0'; modifiers++) |
5714 | switch (modifiers[0]) | |
5715 | { | |
61a1f6fa FP |
5716 | case 'N': |
5717 | if (modifiers == name) | |
db5a3003 | 5718 | error ("forcing explicit tag name but no name, ignoring"); |
61a1f6fa FP |
5719 | force_explicit_name = TRUE; |
5720 | break; | |
24dbe96a FP |
5721 | case 'i': |
5722 | ignore_case = TRUE; | |
5723 | break; | |
5724 | case 's': | |
5725 | single_line = TRUE; | |
5726 | /* FALLTHRU */ | |
5727 | case 'm': | |
5728 | multi_line = TRUE; | |
5729 | need_filebuf = TRUE; | |
5730 | break; | |
5731 | default: | |
db5a3003 | 5732 | error ("invalid regexp modifier `%c', ignoring", modifiers[0]); |
2b749964 | 5733 | break; |
24dbe96a | 5734 | } |
b9755a12 FP |
5735 | |
5736 | patbuf = xnew (1, struct re_pattern_buffer); | |
7e88eb25 FP |
5737 | *patbuf = zeropattern; |
5738 | if (ignore_case) | |
24dbe96a FP |
5739 | { |
5740 | static char lc_trans[CHARS]; | |
5741 | int i; | |
5742 | for (i = 0; i < CHARS; i++) | |
5743 | lc_trans[i] = lowcase (i); | |
5744 | patbuf->translate = lc_trans; /* translation table to fold case */ | |
5745 | } | |
5746 | ||
5747 | if (multi_line) | |
5748 | pat = concat ("^", regexp_pattern, ""); /* anchor to beginning of line */ | |
5749 | else | |
5750 | pat = regexp_pattern; | |
b9755a12 | 5751 | |
24dbe96a FP |
5752 | if (single_line) |
5753 | re_set_syntax (RE_SYNTAX_EMACS | RE_DOT_NEWLINE); | |
5754 | else | |
5755 | re_set_syntax (RE_SYNTAX_EMACS); | |
5756 | ||
3d864325 | 5757 | err = re_compile_pattern (pat, strlen (pat), patbuf); |
24dbe96a FP |
5758 | if (multi_line) |
5759 | free (pat); | |
b9755a12 FP |
5760 | if (err != NULL) |
5761 | { | |
5762 | error ("%s while compiling pattern", err); | |
5763 | return; | |
5764 | } | |
5765 | ||
61a1f6fa FP |
5766 | rp = p_head; |
5767 | p_head = xnew (1, regexp); | |
5768 | p_head->pattern = savestr (regexp_pattern); | |
5769 | p_head->p_next = rp; | |
eaaacacd FP |
5770 | p_head->lang = lang; |
5771 | p_head->pat = patbuf; | |
61a1f6fa | 5772 | p_head->name = savestr (name); |
93b7ac65 | 5773 | p_head->error_signaled = FALSE; |
61a1f6fa | 5774 | p_head->force_explicit_name = force_explicit_name; |
9e0a3f98 | 5775 | p_head->ignore_case = ignore_case; |
24dbe96a | 5776 | p_head->multi_line = multi_line; |
b9755a12 FP |
5777 | } |
5778 | ||
5779 | /* | |
c5007f46 | 5780 | * Do the substitutions indicated by the regular expression and |
b9755a12 FP |
5781 | * arguments. |
5782 | */ | |
944bdd72 | 5783 | static char * |
873fbd0b | 5784 | substitute (char *in, char *out, struct re_registers *regs) |
b9755a12 | 5785 | { |
2f608d34 | 5786 | char *result, *t; |
e0903a92 | 5787 | int size, dig, diglen; |
2f608d34 FP |
5788 | |
5789 | result = NULL; | |
5790 | size = strlen (out); | |
b9755a12 | 5791 | |
e0903a92 FP |
5792 | /* Pass 1: figure out how much to allocate by finding all \N strings. */ |
5793 | if (out[size - 1] == '\\') | |
5794 | fatal ("pattern error in \"%s\"", out); | |
5795 | for (t = etags_strchr (out, '\\'); | |
5796 | t != NULL; | |
5797 | t = etags_strchr (t + 2, '\\')) | |
af03e6ab | 5798 | if (ISDIGIT (t[1])) |
2f608d34 | 5799 | { |
e0903a92 FP |
5800 | dig = t[1] - '0'; |
5801 | diglen = regs->end[dig] - regs->start[dig]; | |
5802 | size += diglen - 2; | |
2f608d34 | 5803 | } |
e0903a92 FP |
5804 | else |
5805 | size -= 1; | |
b9755a12 FP |
5806 | |
5807 | /* Allocate space and do the substitutions. */ | |
a127d423 | 5808 | assert (size >= 0); |
b9755a12 | 5809 | result = xnew (size + 1, char); |
e0903a92 FP |
5810 | |
5811 | for (t = result; *out != '\0'; out++) | |
af03e6ab | 5812 | if (*out == '\\' && ISDIGIT (*++out)) |
e0903a92 | 5813 | { |
e0903a92 FP |
5814 | dig = *out - '0'; |
5815 | diglen = regs->end[dig] - regs->start[dig]; | |
e99a530f | 5816 | memcpy (t, in + regs->start[dig], diglen); |
e0903a92 FP |
5817 | t += diglen; |
5818 | } | |
5819 | else | |
5820 | *t++ = *out; | |
5821 | *t = '\0'; | |
5822 | ||
a127d423 FP |
5823 | assert (t <= result + size); |
5824 | assert (t - result == (int)strlen (result)); | |
b9755a12 FP |
5825 | |
5826 | return result; | |
c6d46f5f | 5827 | } |
93b7ac65 | 5828 | |
61a1f6fa | 5829 | /* Deallocate all regexps. */ |
944bdd72 | 5830 | static void |
873fbd0b | 5831 | free_regexps (void) |
93b7ac65 | 5832 | { |
61a1f6fa | 5833 | regexp *rp; |
93b7ac65 FP |
5834 | while (p_head != NULL) |
5835 | { | |
61a1f6fa FP |
5836 | rp = p_head->p_next; |
5837 | free (p_head->pattern); | |
5838 | free (p_head->name); | |
93b7ac65 | 5839 | free (p_head); |
61a1f6fa | 5840 | p_head = rp; |
93b7ac65 FP |
5841 | } |
5842 | return; | |
5843 | } | |
24dbe96a FP |
5844 | |
5845 | /* | |
5846 | * Reads the whole file as a single string from `filebuf' and looks for | |
5847 | * multi-line regular expressions, creating tags on matches. | |
5848 | * readline already dealt with normal regexps. | |
5849 | * | |
5850 | * Idea by Ben Wing <ben@666.com> (2002). | |
5851 | */ | |
5852 | static void | |
873fbd0b | 5853 | regex_tag_multiline (void) |
24dbe96a FP |
5854 | { |
5855 | char *buffer = filebuf.buffer; | |
61a1f6fa FP |
5856 | regexp *rp; |
5857 | char *name; | |
24dbe96a | 5858 | |
61a1f6fa | 5859 | for (rp = p_head; rp != NULL; rp = rp->p_next) |
24dbe96a FP |
5860 | { |
5861 | int match = 0; | |
5862 | ||
61a1f6fa | 5863 | if (!rp->multi_line) |
24dbe96a FP |
5864 | continue; /* skip normal regexps */ |
5865 | ||
e1dbe924 | 5866 | /* Generic initializations before parsing file from memory. */ |
24dbe96a FP |
5867 | lineno = 1; /* reset global line number */ |
5868 | charno = 0; /* reset global char number */ | |
5869 | linecharno = 0; /* reset global char number of line start */ | |
5870 | ||
5871 | /* Only use generic regexps or those for the current language. */ | |
61a1f6fa | 5872 | if (rp->lang != NULL && rp->lang != curfdp->lang) |
24dbe96a FP |
5873 | continue; |
5874 | ||
5875 | while (match >= 0 && match < filebuf.len) | |
5876 | { | |
61a1f6fa FP |
5877 | match = re_search (rp->pat, buffer, filebuf.len, charno, |
5878 | filebuf.len - match, &rp->regs); | |
24dbe96a FP |
5879 | switch (match) |
5880 | { | |
5881 | case -2: | |
5882 | /* Some error. */ | |
61a1f6fa | 5883 | if (!rp->error_signaled) |
24dbe96a FP |
5884 | { |
5885 | error ("regexp stack overflow while matching \"%s\"", | |
61a1f6fa FP |
5886 | rp->pattern); |
5887 | rp->error_signaled = TRUE; | |
24dbe96a FP |
5888 | } |
5889 | break; | |
5890 | case -1: | |
5891 | /* No match. */ | |
5892 | break; | |
5893 | default: | |
61a1f6fa | 5894 | if (match == rp->regs.end[0]) |
24dbe96a | 5895 | { |
61a1f6fa | 5896 | if (!rp->error_signaled) |
24dbe96a FP |
5897 | { |
5898 | error ("regexp matches the empty string: \"%s\"", | |
61a1f6fa FP |
5899 | rp->pattern); |
5900 | rp->error_signaled = TRUE; | |
24dbe96a FP |
5901 | } |
5902 | match = -3; /* exit from while loop */ | |
5903 | break; | |
5904 | } | |
5905 | ||
5906 | /* Match occurred. Construct a tag. */ | |
61a1f6fa | 5907 | while (charno < rp->regs.end[0]) |
24dbe96a FP |
5908 | if (buffer[charno++] == '\n') |
5909 | lineno++, linecharno = charno; | |
61a1f6fa | 5910 | name = rp->name; |
a127d423 FP |
5911 | if (name[0] == '\0') |
5912 | name = NULL; | |
5913 | else /* make a named tag */ | |
61a1f6fa FP |
5914 | name = substitute (buffer, rp->name, &rp->regs); |
5915 | if (rp->force_explicit_name) | |
5916 | /* Force explicit tag name, if a name is there. */ | |
5917 | pfnote (name, TRUE, buffer + linecharno, | |
5918 | charno - linecharno + 1, lineno, linecharno); | |
24dbe96a | 5919 | else |
61a1f6fa | 5920 | make_tag (name, strlen (name), TRUE, buffer + linecharno, |
24dbe96a | 5921 | charno - linecharno + 1, lineno, linecharno); |
24dbe96a FP |
5922 | break; |
5923 | } | |
5924 | } | |
5925 | } | |
5926 | } | |
5927 | ||
c6d46f5f | 5928 | \f |
eaaacacd | 5929 | static bool |
988e88ab | 5930 | nocase_tail (const char *cp) |
eaaacacd FP |
5931 | { |
5932 | register int len = 0; | |
5933 | ||
5934 | while (*cp != '\0' && lowcase (*cp) == lowcase (dbp[len])) | |
5935 | cp++, len++; | |
5936 | if (*cp == '\0' && !intoken (dbp[len])) | |
5937 | { | |
5938 | dbp += len; | |
5939 | return TRUE; | |
5940 | } | |
5941 | return FALSE; | |
5942 | } | |
5943 | ||
081b7c70 | 5944 | static void |
873fbd0b | 5945 | get_tag (register char *bp, char **namepp) |
93b7ac65 | 5946 | { |
081b7c70 | 5947 | register char *cp = bp; |
93b7ac65 | 5948 | |
081b7c70 FP |
5949 | if (*bp != '\0') |
5950 | { | |
5951 | /* Go till you get to white space or a syntactic break */ | |
5952 | for (cp = bp + 1; !notinname (*cp); cp++) | |
5953 | continue; | |
5954 | make_tag (bp, cp - bp, TRUE, | |
5955 | lb.buffer, cp - lb.buffer + 1, lineno, linecharno); | |
5956 | } | |
5957 | ||
5958 | if (namepp != NULL) | |
5959 | *namepp = savenstr (bp, cp - bp); | |
93b7ac65 FP |
5960 | } |
5961 | ||
c6d46f5f | 5962 | /* |
93b7ac65 FP |
5963 | * Read a line of text from `stream' into `lbp', excluding the |
5964 | * newline or CR-NL, if any. Return the number of characters read from | |
5965 | * `stream', which is the length of the line including the newline. | |
5966 | * | |
24dbe96a FP |
5967 | * On DOS or Windows we do not count the CR character, if any before the |
5968 | * NL, in the returned length; this mirrors the behavior of Emacs on those | |
93b7ac65 FP |
5969 | * platforms (for text files, it translates CR-NL to NL as it reads in the |
5970 | * file). | |
24dbe96a FP |
5971 | * |
5972 | * If multi-line regular expressions are requested, each line read is | |
5973 | * appended to `filebuf'. | |
c6d46f5f | 5974 | */ |
944bdd72 | 5975 | static long |
873fbd0b | 5976 | readline_internal (linebuffer *lbp, register FILE *stream) |
c6d46f5f | 5977 | { |
93b7ac65 FP |
5978 | char *buffer = lbp->buffer; |
5979 | register char *p = lbp->buffer; | |
c6d46f5f | 5980 | register char *pend; |
aab1fdae | 5981 | int chars_deleted; |
c6d46f5f | 5982 | |
93b7ac65 | 5983 | pend = p + lbp->size; /* Separate to avoid 386/IX compiler bug. */ |
c6d46f5f | 5984 | |
97b90b0a | 5985 | for (;;) |
c6d46f5f JB |
5986 | { |
5987 | register int c = getc (stream); | |
5988 | if (p == pend) | |
5989 | { | |
93b7ac65 FP |
5990 | /* We're at the end of linebuffer: expand it. */ |
5991 | lbp->size *= 2; | |
b2521b0a | 5992 | xrnew (buffer, lbp->size, char); |
93b7ac65 FP |
5993 | p += buffer - lbp->buffer; |
5994 | pend = buffer + lbp->size; | |
5995 | lbp->buffer = buffer; | |
c6d46f5f | 5996 | } |
aab1fdae | 5997 | if (c == EOF) |
c6d46f5f | 5998 | { |
8dc7496c | 5999 | *p = '\0'; |
aab1fdae FP |
6000 | chars_deleted = 0; |
6001 | break; | |
6002 | } | |
6003 | if (c == '\n') | |
6004 | { | |
a8d9bd4b | 6005 | if (p > buffer && p[-1] == '\r') |
aab1fdae | 6006 | { |
e0903a92 | 6007 | p -= 1; |
70611080 | 6008 | #ifdef DOS_NT |
b6b48846 FP |
6009 | /* Assume CRLF->LF translation will be performed by Emacs |
6010 | when loading this file, so CRs won't appear in the buffer. | |
6011 | It would be cleaner to compensate within Emacs; | |
6012 | however, Emacs does not know how many CRs were deleted | |
6013 | before any given point in the file. */ | |
70611080 RS |
6014 | chars_deleted = 1; |
6015 | #else | |
aab1fdae | 6016 | chars_deleted = 2; |
70611080 | 6017 | #endif |
aab1fdae FP |
6018 | } |
6019 | else | |
6020 | { | |
aab1fdae FP |
6021 | chars_deleted = 1; |
6022 | } | |
e0903a92 | 6023 | *p = '\0'; |
c6d46f5f JB |
6024 | break; |
6025 | } | |
6026 | *p++ = c; | |
6027 | } | |
93b7ac65 | 6028 | lbp->len = p - buffer; |
c6d46f5f | 6029 | |
24dbe96a FP |
6030 | if (need_filebuf /* we need filebuf for multi-line regexps */ |
6031 | && chars_deleted > 0) /* not at EOF */ | |
6032 | { | |
6033 | while (filebuf.size <= filebuf.len + lbp->len + 1) /* +1 for \n */ | |
6034 | { | |
6035 | /* Expand filebuf. */ | |
6036 | filebuf.size *= 2; | |
6037 | xrnew (filebuf.buffer, filebuf.size, char); | |
6038 | } | |
e99a530f | 6039 | memcpy (filebuf.buffer + filebuf.len, lbp->buffer, lbp->len); |
24dbe96a FP |
6040 | filebuf.len += lbp->len; |
6041 | filebuf.buffer[filebuf.len++] = '\n'; | |
6042 | filebuf.buffer[filebuf.len] = '\0'; | |
6043 | } | |
6044 | ||
93b7ac65 | 6045 | return lbp->len + chars_deleted; |
c6d46f5f | 6046 | } |
b9755a12 FP |
6047 | |
6048 | /* | |
e0903a92 | 6049 | * Like readline_internal, above, but in addition try to match the |
24dbe96a FP |
6050 | * input line against relevant regular expressions and manage #line |
6051 | * directives. | |
b9755a12 | 6052 | */ |
e7d3b099 | 6053 | static void |
873fbd0b | 6054 | readline (linebuffer *lbp, FILE *stream) |
b9755a12 | 6055 | { |
e7d3b099 FP |
6056 | long result; |
6057 | ||
6058 | linecharno = charno; /* update global char number of line start */ | |
6059 | result = readline_internal (lbp, stream); /* read line */ | |
6060 | lineno += 1; /* increment global line number */ | |
6061 | charno += result; /* increment global char number */ | |
db590582 | 6062 | |
f6b1b0a8 | 6063 | /* Honor #line directives. */ |
9e0a3f98 | 6064 | if (!no_line_directive) |
db590582 | 6065 | { |
9e0a3f98 | 6066 | static bool discard_until_line_directive; |
db590582 | 6067 | |
9e0a3f98 FP |
6068 | /* Check whether this is a #line directive. */ |
6069 | if (result > 12 && strneq (lbp->buffer, "#line ", 6)) | |
db590582 | 6070 | { |
21cb180b FP |
6071 | unsigned int lno; |
6072 | int start = 0; | |
db590582 | 6073 | |
21cb180b FP |
6074 | if (sscanf (lbp->buffer, "#line %u \"%n", &lno, &start) >= 1 |
6075 | && start > 0) /* double quote character found */ | |
db590582 | 6076 | { |
21cb180b | 6077 | char *endp = lbp->buffer + start; |
9e0a3f98 | 6078 | |
9e0a3f98 FP |
6079 | while ((endp = etags_strchr (endp, '"')) != NULL |
6080 | && endp[-1] == '\\') | |
6081 | endp++; | |
6082 | if (endp != NULL) | |
6083 | /* Ok, this is a real #line directive. Let's deal with it. */ | |
49adb67a | 6084 | { |
9e0a3f98 FP |
6085 | char *taggedabsname; /* absolute name of original file */ |
6086 | char *taggedfname; /* name of original file as given */ | |
6087 | char *name; /* temp var */ | |
6088 | ||
6089 | discard_until_line_directive = FALSE; /* found it */ | |
6090 | name = lbp->buffer + start; | |
6091 | *endp = '\0'; | |
9817cf3f | 6092 | canonicalize_filename (name); |
6598a3d4 | 6093 | taggedabsname = absolute_filename (name, tagfiledir); |
9e0a3f98 FP |
6094 | if (filename_is_absolute (name) |
6095 | || filename_is_absolute (curfdp->infname)) | |
6096 | taggedfname = savestr (taggedabsname); | |
6097 | else | |
6098 | taggedfname = relative_filename (taggedabsname,tagfiledir); | |
db590582 | 6099 | |
9e0a3f98 FP |
6100 | if (streq (curfdp->taggedfname, taggedfname)) |
6101 | /* The #line directive is only a line number change. We | |
6102 | deal with this afterwards. */ | |
6103 | free (taggedfname); | |
6104 | else | |
6105 | /* The tags following this #line directive should be | |
6106 | attributed to taggedfname. In order to do this, set | |
6107 | curfdp accordingly. */ | |
6108 | { | |
6109 | fdesc *fdp; /* file description pointer */ | |
6110 | ||
6111 | /* Go look for a file description already set up for the | |
6112 | file indicated in the #line directive. If there is | |
6113 | one, use it from now until the next #line | |
6114 | directive. */ | |
6115 | for (fdp = fdhead; fdp != NULL; fdp = fdp->next) | |
6116 | if (streq (fdp->infname, curfdp->infname) | |
6117 | && streq (fdp->taggedfname, taggedfname)) | |
6118 | /* If we remove the second test above (after the &&) | |
6119 | then all entries pertaining to the same file are | |
6120 | coalesced in the tags file. If we use it, then | |
6121 | entries pertaining to the same file but generated | |
6122 | from different files (via #line directives) will | |
6123 | go into separate sections in the tags file. These | |
6124 | alternatives look equivalent. The first one | |
6125 | destroys some apparently useless information. */ | |
6126 | { | |
6127 | curfdp = fdp; | |
6128 | free (taggedfname); | |
6129 | break; | |
6130 | } | |
6131 | /* Else, if we already tagged the real file, skip all | |
6132 | input lines until the next #line directive. */ | |
6133 | if (fdp == NULL) /* not found */ | |
6134 | for (fdp = fdhead; fdp != NULL; fdp = fdp->next) | |
6135 | if (streq (fdp->infabsname, taggedabsname)) | |
6136 | { | |
6137 | discard_until_line_directive = TRUE; | |
6138 | free (taggedfname); | |
6139 | break; | |
6140 | } | |
6141 | /* Else create a new file description and use that from | |
6142 | now on, until the next #line directive. */ | |
6143 | if (fdp == NULL) /* not found */ | |
6144 | { | |
6145 | fdp = fdhead; | |
6146 | fdhead = xnew (1, fdesc); | |
6147 | *fdhead = *curfdp; /* copy curr. file description */ | |
6148 | fdhead->next = fdp; | |
6149 | fdhead->infname = savestr (curfdp->infname); | |
6150 | fdhead->infabsname = savestr (curfdp->infabsname); | |
6151 | fdhead->infabsdir = savestr (curfdp->infabsdir); | |
6152 | fdhead->taggedfname = taggedfname; | |
6153 | fdhead->usecharno = FALSE; | |
61a1f6fa FP |
6154 | fdhead->prop = NULL; |
6155 | fdhead->written = FALSE; | |
9e0a3f98 FP |
6156 | curfdp = fdhead; |
6157 | } | |
6158 | } | |
6159 | free (taggedabsname); | |
e7d3b099 FP |
6160 | lineno = lno - 1; |
6161 | readline (lbp, stream); | |
6162 | return; | |
9e0a3f98 | 6163 | } /* if a real #line directive */ |
045b9da7 | 6164 | } /* if #line is followed by a number */ |
9e0a3f98 FP |
6165 | } /* if line begins with "#line " */ |
6166 | ||
6167 | /* If we are here, no #line directive was found. */ | |
6168 | if (discard_until_line_directive) | |
6169 | { | |
6170 | if (result > 0) | |
e7d3b099 | 6171 | { |
24dbe96a FP |
6172 | /* Do a tail recursion on ourselves, thus discarding the contents |
6173 | of the line buffer. */ | |
e7d3b099 FP |
6174 | readline (lbp, stream); |
6175 | return; | |
6176 | } | |
9e0a3f98 FP |
6177 | /* End of file. */ |
6178 | discard_until_line_directive = FALSE; | |
e7d3b099 | 6179 | return; |
db590582 | 6180 | } |
9e0a3f98 | 6181 | } /* if #line directives should be considered */ |
49adb67a | 6182 | |
db590582 FP |
6183 | { |
6184 | int match; | |
61a1f6fa FP |
6185 | regexp *rp; |
6186 | char *name; | |
d8913c1c | 6187 | |
61a1f6fa | 6188 | /* Match against relevant regexps. */ |
db590582 | 6189 | if (lbp->len > 0) |
61a1f6fa | 6190 | for (rp = p_head; rp != NULL; rp = rp->p_next) |
db590582 | 6191 | { |
24dbe96a FP |
6192 | /* Only use generic regexps or those for the current language. |
6193 | Also do not use multiline regexps, which is the job of | |
6194 | regex_tag_multiline. */ | |
61a1f6fa FP |
6195 | if ((rp->lang != NULL && rp->lang != fdhead->lang) |
6196 | || rp->multi_line) | |
db590582 | 6197 | continue; |
93b7ac65 | 6198 | |
61a1f6fa | 6199 | match = re_match (rp->pat, lbp->buffer, lbp->len, 0, &rp->regs); |
db590582 FP |
6200 | switch (match) |
6201 | { | |
6202 | case -2: | |
6203 | /* Some error. */ | |
61a1f6fa | 6204 | if (!rp->error_signaled) |
db590582 | 6205 | { |
24dbe96a | 6206 | error ("regexp stack overflow while matching \"%s\"", |
61a1f6fa FP |
6207 | rp->pattern); |
6208 | rp->error_signaled = TRUE; | |
db590582 FP |
6209 | } |
6210 | break; | |
6211 | case -1: | |
6212 | /* No match. */ | |
6213 | break; | |
24dbe96a FP |
6214 | case 0: |
6215 | /* Empty string matched. */ | |
61a1f6fa | 6216 | if (!rp->error_signaled) |
24dbe96a | 6217 | { |
61a1f6fa FP |
6218 | error ("regexp matches the empty string: \"%s\"", rp->pattern); |
6219 | rp->error_signaled = TRUE; | |
24dbe96a FP |
6220 | } |
6221 | break; | |
db590582 FP |
6222 | default: |
6223 | /* Match occurred. Construct a tag. */ | |
61a1f6fa | 6224 | name = rp->name; |
a127d423 FP |
6225 | if (name[0] == '\0') |
6226 | name = NULL; | |
6227 | else /* make a named tag */ | |
61a1f6fa FP |
6228 | name = substitute (lbp->buffer, rp->name, &rp->regs); |
6229 | if (rp->force_explicit_name) | |
6230 | /* Force explicit tag name, if a name is there. */ | |
6231 | pfnote (name, TRUE, lbp->buffer, match, lineno, linecharno); | |
db590582 | 6232 | else |
61a1f6fa | 6233 | make_tag (name, strlen (name), TRUE, |
db590582 | 6234 | lbp->buffer, match, lineno, linecharno); |
db590582 FP |
6235 | break; |
6236 | } | |
6237 | } | |
6238 | } | |
b9755a12 | 6239 | } |
b2521b0a | 6240 | |
c6d46f5f | 6241 | \f |
55597f90 FP |
6242 | /* |
6243 | * Return a pointer to a space of size strlen(cp)+1 allocated | |
6244 | * with xnew where the string CP has been copied. | |
6245 | */ | |
944bdd72 | 6246 | static char * |
988e88ab | 6247 | savestr (const char *cp) |
c6d46f5f JB |
6248 | { |
6249 | return savenstr (cp, strlen (cp)); | |
6250 | } | |
6251 | ||
55597f90 FP |
6252 | /* |
6253 | * Return a pointer to a space of size LEN+1 allocated with xnew where | |
6254 | * the string CP has been copied for at most the first LEN characters. | |
6255 | */ | |
944bdd72 | 6256 | static char * |
988e88ab | 6257 | savenstr (const char *cp, int len) |
c6d46f5f JB |
6258 | { |
6259 | register char *dp; | |
6260 | ||
6261 | dp = xnew (len + 1, char); | |
e99a530f | 6262 | memcpy (dp, cp, len); |
c6d46f5f JB |
6263 | dp[len] = '\0'; |
6264 | return dp; | |
6265 | } | |
6266 | ||
c6d46f5f JB |
6267 | /* |
6268 | * Return the ptr in sp at which the character c last | |
6269 | * appears; NULL if not found | |
6270 | * | |
944bdd72 | 6271 | * Identical to POSIX strrchr, included for portability. |
c6d46f5f | 6272 | */ |
944bdd72 | 6273 | static char * |
873fbd0b | 6274 | etags_strrchr (register const char *sp, register int c) |
c6d46f5f | 6275 | { |
944bdd72 | 6276 | register const char *r; |
c6d46f5f JB |
6277 | |
6278 | r = NULL; | |
6279 | do | |
6280 | { | |
6281 | if (*sp == c) | |
6282 | r = sp; | |
6283 | } while (*sp++); | |
944bdd72 | 6284 | return (char *)r; |
c6d46f5f JB |
6285 | } |
6286 | ||
6287 | /* | |
6288 | * Return the ptr in sp at which the character c first | |
6289 | * appears; NULL if not found | |
6290 | * | |
944bdd72 | 6291 | * Identical to POSIX strchr, included for portability. |
c6d46f5f | 6292 | */ |
944bdd72 | 6293 | static char * |
873fbd0b | 6294 | etags_strchr (register const char *sp, register int c) |
c6d46f5f JB |
6295 | { |
6296 | do | |
6297 | { | |
6298 | if (*sp == c) | |
944bdd72 | 6299 | return (char *)sp; |
b9755a12 FP |
6300 | } while (*sp++); |
6301 | return NULL; | |
c6d46f5f JB |
6302 | } |
6303 | ||
55102b5d | 6304 | /* Skip spaces (end of string is not space), return new pointer. */ |
944bdd72 | 6305 | static char * |
873fbd0b | 6306 | skip_spaces (char *cp) |
93b7ac65 | 6307 | { |
71cbb895 | 6308 | while (iswhite (*cp)) |
93b7ac65 FP |
6309 | cp++; |
6310 | return cp; | |
6311 | } | |
6312 | ||
55102b5d | 6313 | /* Skip non spaces, except end of string, return new pointer. */ |
944bdd72 | 6314 | static char * |
873fbd0b | 6315 | skip_non_spaces (char *cp) |
93b7ac65 | 6316 | { |
71cbb895 | 6317 | while (*cp != '\0' && !iswhite (*cp)) |
93b7ac65 FP |
6318 | cp++; |
6319 | return cp; | |
6320 | } | |
6321 | ||
1cbaa705 KR |
6322 | /* Skip any chars in the "name" class.*/ |
6323 | static char * | |
6324 | skip_name (char *cp) | |
6325 | { | |
6326 | /* '\0' is a notinname() so loop stops there too */ | |
6327 | while (! notinname (*cp)) | |
6328 | cp++; | |
6329 | return cp; | |
6330 | } | |
6331 | ||
c6d46f5f | 6332 | /* Print error message and exit. */ |
b2521b0a | 6333 | void |
988e88ab | 6334 | fatal (const char *s1, const char *s2) |
c6d46f5f JB |
6335 | { |
6336 | error (s1, s2); | |
3f0656ff | 6337 | exit (EXIT_FAILURE); |
c6d46f5f JB |
6338 | } |
6339 | ||
944bdd72 | 6340 | static void |
988e88ab | 6341 | pfatal (const char *s1) |
cdc1f6a7 FP |
6342 | { |
6343 | perror (s1); | |
3f0656ff | 6344 | exit (EXIT_FAILURE); |
cdc1f6a7 FP |
6345 | } |
6346 | ||
944bdd72 | 6347 | static void |
873fbd0b | 6348 | suggest_asking_for_help (void) |
d8913c1c | 6349 | { |
9239d970 PE |
6350 | fprintf (stderr, "\tTry `%s --help' for a complete list of options.\n", |
6351 | progname); | |
3f0656ff | 6352 | exit (EXIT_FAILURE); |
d8913c1c FP |
6353 | } |
6354 | ||
db5a3003 | 6355 | /* Output a diagnostic with printf-style FORMAT and args. */ |
944bdd72 | 6356 | static void |
db5a3003 | 6357 | error (const char *format, ...) |
c6d46f5f | 6358 | { |
db5a3003 PE |
6359 | va_list ap; |
6360 | va_start (ap, format); | |
c6d46f5f | 6361 | fprintf (stderr, "%s: ", progname); |
db5a3003 | 6362 | vfprintf (stderr, format, ap); |
c6d46f5f | 6363 | fprintf (stderr, "\n"); |
db5a3003 | 6364 | va_end (ap); |
c6d46f5f JB |
6365 | } |
6366 | ||
46c145db FP |
6367 | /* Return a newly-allocated string whose contents |
6368 | concatenate those of s1, s2, s3. */ | |
944bdd72 | 6369 | static char * |
988e88ab | 6370 | concat (const char *s1, const char *s2, const char *s3) |
c6d46f5f JB |
6371 | { |
6372 | int len1 = strlen (s1), len2 = strlen (s2), len3 = strlen (s3); | |
6373 | char *result = xnew (len1 + len2 + len3 + 1, char); | |
6374 | ||
1a0d8c80 FP |
6375 | strcpy (result, s1); |
6376 | strcpy (result + len1, s2); | |
6377 | strcpy (result + len1 + len2, s3); | |
46c145db | 6378 | result[len1 + len2 + len3] = '\0'; |
c6d46f5f JB |
6379 | |
6380 | return result; | |
6381 | } | |
b2521b0a | 6382 | |
b02c5fea | 6383 | \f |
cdc1f6a7 | 6384 | /* Does the same work as the system V getcwd, but does not need to |
c5007f46 | 6385 | guess the buffer size in advance. */ |
944bdd72 | 6386 | static char * |
873fbd0b | 6387 | etags_getcwd (void) |
915e30c8 | 6388 | { |
cdc1f6a7 FP |
6389 | int bufsize = 200; |
6390 | char *path = xnew (bufsize, char); | |
c6d46f5f | 6391 | |
cdc1f6a7 | 6392 | while (getcwd (path, bufsize) == NULL) |
b02c5fea | 6393 | { |
dcc89e63 | 6394 | if (errno != ERANGE) |
0f394065 | 6395 | pfatal ("getcwd"); |
5e9c8296 | 6396 | bufsize *= 2; |
cec68fcb | 6397 | free (path); |
cdc1f6a7 | 6398 | path = xnew (bufsize, char); |
5e9c8296 | 6399 | } |
b02c5fea | 6400 | |
93b7ac65 | 6401 | canonicalize_filename (path); |
cdc1f6a7 | 6402 | return path; |
b02c5fea FP |
6403 | } |
6404 | ||
93b7ac65 FP |
6405 | /* Return a newly allocated string containing the file name of FILE |
6406 | relative to the absolute directory DIR (which should end with a slash). */ | |
944bdd72 | 6407 | static char * |
873fbd0b | 6408 | relative_filename (char *file, char *dir) |
46c145db | 6409 | { |
93b7ac65 | 6410 | char *fp, *dp, *afn, *res; |
2f608d34 | 6411 | int i; |
46c145db | 6412 | |
3ca80e28 | 6413 | /* Find the common root of file and dir (with a trailing slash). */ |
93b7ac65 FP |
6414 | afn = absolute_filename (file, cwd); |
6415 | fp = afn; | |
46c145db FP |
6416 | dp = dir; |
6417 | while (*fp++ == *dp++) | |
6418 | continue; | |
3f1c8fcd | 6419 | fp--, dp--; /* back to the first differing char */ |
93b7ac65 FP |
6420 | #ifdef DOS_NT |
6421 | if (fp == afn && afn[0] != '/') /* cannot build a relative name */ | |
6422 | return afn; | |
6423 | #endif | |
6424 | do /* look at the equal chars until '/' */ | |
3ca80e28 | 6425 | fp--, dp--; |
93b7ac65 | 6426 | while (*fp != '/'); |
46c145db | 6427 | |
2f608d34 FP |
6428 | /* Build a sequence of "../" strings for the resulting relative file name. */ |
6429 | i = 0; | |
6430 | while ((dp = etags_strchr (dp + 1, '/')) != NULL) | |
9817cf3f | 6431 | i += 1; |
2f608d34 FP |
6432 | res = xnew (3*i + strlen (fp + 1) + 1, char); |
6433 | res[0] = '\0'; | |
6434 | while (i-- > 0) | |
6435 | strcat (res, "../"); | |
6436 | ||
6437 | /* Add the file name relative to the common root of file and dir. */ | |
6438 | strcat (res, fp + 1); | |
93b7ac65 | 6439 | free (afn); |
46c145db | 6440 | |
108c932a | 6441 | return res; |
46c145db FP |
6442 | } |
6443 | ||
93b7ac65 FP |
6444 | /* Return a newly allocated string containing the absolute file name |
6445 | of FILE given DIR (which should end with a slash). */ | |
944bdd72 | 6446 | static char * |
873fbd0b | 6447 | absolute_filename (char *file, char *dir) |
46c145db FP |
6448 | { |
6449 | char *slashp, *cp, *res; | |
6450 | ||
93b7ac65 | 6451 | if (filename_is_absolute (file)) |
2f608d34 | 6452 | res = savestr (file); |
ce7f6d62 | 6453 | #ifdef DOS_NT |
2f608d34 | 6454 | /* We don't support non-absolute file names with a drive |
ce7f6d62 RS |
6455 | letter, like `d:NAME' (it's too much hassle). */ |
6456 | else if (file[1] == ':') | |
2f608d34 | 6457 | fatal ("%s: relative file names with drive letters not supported", file); |
ce7f6d62 | 6458 | #endif |
46c145db | 6459 | else |
93b7ac65 | 6460 | res = concat (dir, file, ""); |
46c145db FP |
6461 | |
6462 | /* Delete the "/dirname/.." and "/." substrings. */ | |
b02c5fea | 6463 | slashp = etags_strchr (res, '/'); |
46c145db FP |
6464 | while (slashp != NULL && slashp[0] != '\0') |
6465 | { | |
6466 | if (slashp[1] == '.') | |
6467 | { | |
6468 | if (slashp[2] == '.' | |
6469 | && (slashp[3] == '/' || slashp[3] == '\0')) | |
6470 | { | |
6471 | cp = slashp; | |
6472 | do | |
6473 | cp--; | |
93b7ac65 | 6474 | while (cp >= res && !filename_is_absolute (cp)); |
2f608d34 FP |
6475 | if (cp < res) |
6476 | cp = slashp; /* the absolute name begins with "/.." */ | |
ce7f6d62 RS |
6477 | #ifdef DOS_NT |
6478 | /* Under MSDOS and NT we get `d:/NAME' as absolute | |
2f608d34 | 6479 | file name, so the luser could say `d:/../NAME'. |
ce7f6d62 | 6480 | We silently treat this as `d:/NAME'. */ |
2f608d34 FP |
6481 | else if (cp[0] != '/') |
6482 | cp = slashp; | |
ce7f6d62 | 6483 | #endif |
46f3381a | 6484 | memmove (cp, slashp + 3, strlen (slashp + 2)); |
46c145db | 6485 | slashp = cp; |
e9b2b94c | 6486 | continue; |
46c145db FP |
6487 | } |
6488 | else if (slashp[2] == '/' || slashp[2] == '\0') | |
6489 | { | |
46f3381a | 6490 | memmove (slashp, slashp + 2, strlen (slashp + 1)); |
e9b2b94c | 6491 | continue; |
46c145db FP |
6492 | } |
6493 | } | |
e9b2b94c FP |
6494 | |
6495 | slashp = etags_strchr (slashp + 1, '/'); | |
46c145db | 6496 | } |
7ebca6a6 | 6497 | |
327891eb FP |
6498 | if (res[0] == '\0') /* just a safety net: should never happen */ |
6499 | { | |
6500 | free (res); | |
6501 | return savestr ("/"); | |
6502 | } | |
2f608d34 FP |
6503 | else |
6504 | return res; | |
46c145db FP |
6505 | } |
6506 | ||
b02c5fea | 6507 | /* Return a newly allocated string containing the absolute |
93b7ac65 | 6508 | file name of dir where FILE resides given DIR (which should |
b02c5fea | 6509 | end with a slash). */ |
944bdd72 | 6510 | static char * |
873fbd0b | 6511 | absolute_dirname (char *file, char *dir) |
46c145db FP |
6512 | { |
6513 | char *slashp, *res; | |
6514 | char save; | |
6515 | ||
b02c5fea | 6516 | slashp = etags_strrchr (file, '/'); |
46c145db | 6517 | if (slashp == NULL) |
93b7ac65 | 6518 | return savestr (dir); |
46c145db FP |
6519 | save = slashp[1]; |
6520 | slashp[1] = '\0'; | |
93b7ac65 | 6521 | res = absolute_filename (file, dir); |
46c145db FP |
6522 | slashp[1] = save; |
6523 | ||
6524 | return res; | |
6525 | } | |
6526 | ||
93b7ac65 FP |
6527 | /* Whether the argument string is an absolute file name. The argument |
6528 | string must have been canonicalized with canonicalize_filename. */ | |
944bdd72 | 6529 | static bool |
873fbd0b | 6530 | filename_is_absolute (char *fn) |
93b7ac65 FP |
6531 | { |
6532 | return (fn[0] == '/' | |
6533 | #ifdef DOS_NT | |
5e617bc2 | 6534 | || (ISALPHA (fn[0]) && fn[1] == ':' && fn[2] == '/') |
93b7ac65 FP |
6535 | #endif |
6536 | ); | |
6537 | } | |
6538 | ||
e5075711 | 6539 | /* Downcase DOS drive letter and collapse separators into single slashes. |
9817cf3f | 6540 | Works in place. */ |
944bdd72 | 6541 | static void |
873fbd0b | 6542 | canonicalize_filename (register char *fn) |
93b7ac65 | 6543 | { |
9817cf3f FP |
6544 | register char* cp; |
6545 | char sep = '/'; | |
6546 | ||
93b7ac65 | 6547 | #ifdef DOS_NT |
c8b04710 | 6548 | /* Canonicalize drive letter case. */ |
5e617bc2 | 6549 | # define ISUPPER(c) isupper (CHAR (c)) |
e5075711 | 6550 | if (fn[0] != '\0' && fn[1] == ':' && ISUPPER (fn[0])) |
32e793fa | 6551 | fn[0] = lowcase (fn[0]); |
9817cf3f FP |
6552 | |
6553 | sep = '\\'; | |
93b7ac65 | 6554 | #endif |
9817cf3f FP |
6555 | |
6556 | /* Collapse multiple separators into a single slash. */ | |
6557 | for (cp = fn; *cp != '\0'; cp++, fn++) | |
6558 | if (*cp == sep) | |
6559 | { | |
6560 | *fn = '/'; | |
6561 | while (cp[1] == sep) | |
6562 | cp++; | |
6563 | } | |
6564 | else | |
6565 | *fn = *cp; | |
6566 | *fn = '\0'; | |
93b7ac65 FP |
6567 | } |
6568 | ||
61a1f6fa | 6569 | \f |
9817cf3f | 6570 | /* Initialize a linebuffer for use. */ |
61a1f6fa | 6571 | static void |
873fbd0b | 6572 | linebuffer_init (linebuffer *lbp) |
61a1f6fa FP |
6573 | { |
6574 | lbp->size = (DEBUG) ? 3 : 200; | |
6575 | lbp->buffer = xnew (lbp->size, char); | |
6576 | lbp->buffer[0] = '\0'; | |
6577 | lbp->len = 0; | |
6578 | } | |
6579 | ||
b2521b0a | 6580 | /* Set the minimum size of a string contained in a linebuffer. */ |
944bdd72 | 6581 | static void |
873fbd0b | 6582 | linebuffer_setlen (linebuffer *lbp, int toksize) |
3f1c8fcd | 6583 | { |
b2521b0a FP |
6584 | while (lbp->size <= toksize) |
6585 | { | |
6586 | lbp->size *= 2; | |
6587 | xrnew (lbp->buffer, lbp->size, char); | |
6588 | } | |
6589 | lbp->len = toksize; | |
3f1c8fcd FP |
6590 | } |
6591 | ||
61a1f6fa | 6592 | /* Like malloc but get fatal error if memory is exhausted. */ |
261cb4bb | 6593 | static void * |
9250f758 | 6594 | xmalloc (size_t size) |
c6d46f5f | 6595 | { |
261cb4bb | 6596 | void *result = malloc (size); |
1a0d8c80 | 6597 | if (result == NULL) |
086eac13 | 6598 | fatal ("virtual memory exhausted", (char *)NULL); |
c6d46f5f JB |
6599 | return result; |
6600 | } | |
6601 | ||
261cb4bb | 6602 | static void * |
9250f758 | 6603 | xrealloc (char *ptr, size_t size) |
c6d46f5f | 6604 | { |
261cb4bb | 6605 | void *result = realloc (ptr, size); |
1a0d8c80 | 6606 | if (result == NULL) |
086eac13 | 6607 | fatal ("virtual memory exhausted", (char *)NULL); |
c6d46f5f JB |
6608 | return result; |
6609 | } | |
62aec606 FP |
6610 | |
6611 | /* | |
6612 | * Local Variables: | |
62aec606 FP |
6613 | * indent-tabs-mode: t |
6614 | * tab-width: 8 | |
9e0a3f98 | 6615 | * fill-column: 79 |
61a1f6fa | 6616 | * c-font-lock-extra-types: ("FILE" "bool" "language" "linebuffer" "fdesc" "node" "regexp") |
a423dfde | 6617 | * c-file-style: "gnu" |
62aec606 FP |
6618 | * End: |
6619 | */ | |
ab5796a9 | 6620 | |
3f0656ff | 6621 | /* etags.c ends here */ |