Commit | Line | Data |
---|---|---|
c6d46f5f | 1 | /* Tags file maker to go with GNU Emacs |
f470f9bd KH |
2 | Copyright (C) 1984, 87, 88, 89, 93, 94, 95 |
3 | Free Software Foundation, Inc. and Ken Arnold | |
ea6cd314 | 4 | This file is not considered part of GNU Emacs. |
c6d46f5f | 5 | |
ea6cd314 | 6 | This program is free software; you can redistribute it and/or modify |
c6d46f5f | 7 | it under the terms of the GNU General Public License as published by |
ea6cd314 RS |
8 | the Free Software Foundation; either version 2 of the License, or |
9 | (at your option) any later version. | |
c6d46f5f | 10 | |
ea6cd314 | 11 | This program is distributed in the hope that it will be useful, |
c6d46f5f JB |
12 | but WITHOUT ANY WARRANTY; without even the implied warranty of |
13 | MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the | |
14 | GNU General Public License for more details. | |
15 | ||
16 | You should have received a copy of the GNU General Public License | |
ea6cd314 RS |
17 | along with this program; if not, write to the Free Software |
18 | Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. */ | |
c6d46f5f JB |
19 | |
20 | /* | |
21 | * Authors: | |
22 | * Ctags originally by Ken Arnold. | |
6dd5561c | 23 | * Fortran added by Jim Kleckner. |
c6d46f5f JB |
24 | * Ed Pelegri-Llopart added C typedefs. |
25 | * Gnu Emacs TAGS format and modifications by RMS? | |
26 | * Sam Kendall added C++. | |
b9755a12 | 27 | * Francesco Potorti` reorganised C and C++ based on work by Joe Wells. |
b9755a12 | 28 | * Regexp tags by Tom Tromey. |
31d4b314 | 29 | * |
46c145db | 30 | * Francesco Potorti` (pot@cnuce.cnr.it) is the current maintainer. |
c6d46f5f JB |
31 | */ |
32 | ||
d8913c1c | 33 | char pot_etags_version[] = "@(#) pot revision number is 11.53"; |
75bdbc6a FP |
34 | |
35 | #define TRUE 1 | |
36 | #define FALSE 0 | |
d8913c1c | 37 | |
75bdbc6a FP |
38 | #ifndef DEBUG |
39 | # define DEBUG FALSE | |
40 | #endif | |
46c145db | 41 | |
c6880c90 RS |
42 | #ifdef MSDOS |
43 | #include <fcntl.h> | |
88f125fc | 44 | #include <sys/param.h> |
c6880c90 RS |
45 | #endif /* MSDOS */ |
46 | ||
c05b6df5 RS |
47 | #ifdef WINDOWSNT |
48 | #include <stdlib.h> | |
49 | #include <fcntl.h> | |
50 | #include <string.h> | |
51 | #define MAXPATHLEN _MAX_PATH | |
52 | #endif | |
53 | ||
72a339d7 | 54 | #ifdef HAVE_CONFIG_H |
b9755a12 | 55 | #include <config.h> |
aab1fdae FP |
56 | /* On some systems, Emacs defines static as nothing for the sake |
57 | of unexec. We don't want that here since we don't use unexec. */ | |
1ddff51c | 58 | #undef static |
1e134a5f RM |
59 | #endif |
60 | ||
918f9ad1 JB |
61 | #include <stdio.h> |
62 | #include <ctype.h> | |
dcc89e63 FP |
63 | #include <errno.h> |
64 | #ifndef errno | |
65 | extern int errno; | |
66 | #endif | |
918f9ad1 JB |
67 | #include <sys/types.h> |
68 | #include <sys/stat.h> | |
69 | ||
2b878b4c FP |
70 | #if !defined (S_ISREG) && defined (S_IFREG) |
71 | # define S_ISREG(m) (((m) & S_IFMT) == S_IFREG) | |
72 | #endif | |
73 | ||
b9755a12 FP |
74 | #include <getopt.h> |
75 | ||
76 | #ifdef ETAGS_REGEXPS | |
77 | #include <regex.h> | |
78 | #endif /* ETAGS_REGEXPS */ | |
918f9ad1 | 79 | |
32daa216 FP |
80 | /* Define CTAGS to make the program "ctags" compatible with the usual one. |
81 | Let it undefined to make the program "etags", which makes emacs-style | |
82 | tag tables and tags typedefs, #defines and struct/union/enum by default. */ | |
83 | #ifdef CTAGS | |
84 | # undef CTAGS | |
85 | # define CTAGS TRUE | |
86 | #else | |
87 | # define CTAGS FALSE | |
c6d46f5f JB |
88 | #endif |
89 | ||
90 | /* Exit codes for success and failure. */ | |
91 | #ifdef VMS | |
32daa216 FP |
92 | #define GOOD 1 |
93 | #define BAD 0 | |
c6d46f5f | 94 | #else |
32daa216 FP |
95 | #define GOOD 0 |
96 | #define BAD 1 | |
c6d46f5f JB |
97 | #endif |
98 | ||
55597f90 FP |
99 | /* C extensions. */ |
100 | #define C_PLPL 0x00001 /* C++ */ | |
101 | #define C_STAR 0x00003 /* C* */ | |
102 | #define YACC 0x10000 /* yacc file */ | |
c6d46f5f | 103 | |
d8913c1c FP |
104 | #define streq(s,t) ((DEBUG &&!(s)&&!(t)&&(abort(),1)) || !strcmp(s,t)) |
105 | #define strneq(s,t,n) ((DEBUG &&!(s)&&!(t)&&(abort(),1)) || !strncmp(s,t,n)) | |
c6d46f5f | 106 | |
c5007f46 | 107 | #define lowcase(c) tolower ((unsigned char)c) |
79263656 | 108 | |
c6d46f5f JB |
109 | #define iswhite(arg) (_wht[arg]) /* T if char is white */ |
110 | #define begtoken(arg) (_btk[arg]) /* T if char can start token */ | |
111 | #define intoken(arg) (_itk[arg]) /* T if char can be in token */ | |
112 | #define endtoken(arg) (_etk[arg]) /* T if char ends tokens */ | |
c6d46f5f | 113 | |
b2db879b FP |
114 | #ifdef DOS_NT |
115 | # define absolutefn(fn) (fn[0] == '/' || (isalpha (fn[0]) && fn[1] == ':')) | |
116 | #else | |
117 | # define absolutefn(fn) (fn[0] == '/') | |
118 | #endif | |
119 | ||
120 | ||
55597f90 FP |
121 | /* |
122 | * xnew -- allocate storage | |
123 | * | |
124 | * SYNOPSIS: Type *xnew (int n, Type); | |
125 | */ | |
126 | #define xnew(n,Type) ((Type *) xmalloc ((n) * sizeof (Type))) | |
c6d46f5f | 127 | |
aab1fdae FP |
128 | typedef int logical; |
129 | ||
55597f90 | 130 | typedef struct nd_st |
d8913c1c | 131 | { /* sorting structure */ |
c6d46f5f JB |
132 | char *name; /* function or type name */ |
133 | char *file; /* file name */ | |
134 | logical is_func; /* use pattern or line no */ | |
c6d46f5f JB |
135 | logical been_warned; /* set if noticed dup */ |
136 | int lno; /* line number tag is on */ | |
137 | long cno; /* character number line starts on */ | |
138 | char *pat; /* search pattern */ | |
139 | struct nd_st *left, *right; /* left and right sons */ | |
55597f90 | 140 | } NODE; |
c6d46f5f | 141 | |
55597f90 | 142 | extern char *getenv (); |
c6d46f5f JB |
143 | |
144 | char *concat (); | |
46c145db | 145 | char *savenstr (), *savestr (); |
b02c5fea FP |
146 | char *etags_strchr (), *etags_strrchr (); |
147 | char *etags_getcwd (); | |
46c145db | 148 | char *relative_filename (), *absolute_filename (), *absolute_dirname (); |
03cdafdf | 149 | long *xmalloc (), *xrealloc (); |
b9755a12 FP |
150 | |
151 | typedef void Lang_function (); | |
cdc1f6a7 | 152 | #if FALSE /* many compilers barf on this */ |
b9755a12 FP |
153 | Lang_function Asm_labels; |
154 | Lang_function default_C_entries; | |
155 | Lang_function C_entries; | |
156 | Lang_function Cplusplus_entries; | |
157 | Lang_function Cstar_entries; | |
158 | Lang_function Fortran_functions; | |
159 | Lang_function Yacc_entries; | |
160 | Lang_function Lisp_functions; | |
161 | Lang_function Pascal_functions; | |
1f638249 | 162 | Lang_function Perl_functions; |
b9755a12 FP |
163 | Lang_function Prolog_functions; |
164 | Lang_function Scheme_functions; | |
165 | Lang_function TeX_functions; | |
166 | Lang_function just_read_file; | |
aab1fdae FP |
167 | #else /* so let's write it this way */ |
168 | void Asm_labels (); | |
aab1fdae | 169 | void C_entries (); |
79263656 FP |
170 | void default_C_entries (); |
171 | void plain_C_entries (); | |
aab1fdae FP |
172 | void Cplusplus_entries (); |
173 | void Cstar_entries (); | |
174 | void Fortran_functions (); | |
175 | void Yacc_entries (); | |
176 | void Lisp_functions (); | |
177 | void Pascal_functions (); | |
1f638249 | 178 | void Perl_functions (); |
aab1fdae FP |
179 | void Prolog_functions (); |
180 | void Scheme_functions (); | |
181 | void TeX_functions (); | |
182 | void just_read_file (); | |
183 | #endif | |
b9755a12 | 184 | |
1f638249 FP |
185 | Lang_function *get_language_from_name (); |
186 | Lang_function *get_language_from_interpreter (); | |
187 | Lang_function *get_language_from_suffix (); | |
c6d46f5f | 188 | int total_size_of_entries (); |
c6d46f5f | 189 | long readline (); |
b9755a12 FP |
190 | long readline_internal (); |
191 | #ifdef ETAGS_REGEXPS | |
192 | void add_regex (); | |
193 | #endif | |
c6d46f5f JB |
194 | void add_node (); |
195 | void error (); | |
d8913c1c | 196 | void suggest_asking_for_help (); |
cdc1f6a7 | 197 | void fatal (), pfatal (); |
55597f90 | 198 | void find_entries (); |
c6d46f5f JB |
199 | void free_tree (); |
200 | void getit (); | |
c6d46f5f JB |
201 | void init (); |
202 | void initbuffer (); | |
c6d46f5f JB |
203 | void pfnote (); |
204 | void process_file (); | |
205 | void put_entries (); | |
206 | void takeprec (); | |
207 | ||
c6d46f5f | 208 | \f |
55597f90 | 209 | char searchar = '/'; /* use /.../ searches */ |
c6d46f5f | 210 | |
55597f90 FP |
211 | int lineno; /* line number of current line */ |
212 | long charno; /* current character number */ | |
c6d46f5f | 213 | |
55597f90 FP |
214 | long linecharno; /* charno of start of line; not used by C, |
215 | but by every other language. */ | |
c6d46f5f | 216 | |
55597f90 FP |
217 | char *curfile; /* current input file name */ |
218 | char *tagfile; /* output file */ | |
219 | char *progname; /* name this program was invoked with */ | |
220 | char *cwd; /* current working directory */ | |
221 | char *tagfiledir; /* directory of tagfile */ | |
c6d46f5f | 222 | |
55597f90 FP |
223 | FILE *tagf; /* ioptr for tags file */ |
224 | NODE *head; /* the head of the binary tree of tags */ | |
c6d46f5f | 225 | |
55597f90 FP |
226 | /* |
227 | * A `struct linebuffer' is a structure which holds a line of text. | |
228 | * `readline' reads a line from a stream into a linebuffer and works | |
229 | * regardless of the length of the line. | |
230 | */ | |
d8913c1c FP |
231 | #define GROW_LINEBUFFER(buf,toksize) \ |
232 | while (buf.size < toksize) \ | |
233 | buf.buffer = (char *) xrealloc (buf.buffer, buf.size *= 2) | |
55597f90 FP |
234 | struct linebuffer |
235 | { | |
236 | long size; | |
237 | char *buffer; | |
238 | }; | |
c6d46f5f | 239 | |
55597f90 | 240 | struct linebuffer lb; /* the current line */ |
d8913c1c | 241 | struct linebuffer token_name; /* used by C_entries as a temporary area */ |
55597f90 FP |
242 | struct |
243 | { | |
244 | long linepos; | |
245 | struct linebuffer lb; /* used by C_entries instead of lb */ | |
246 | } lbs[2]; | |
c6d46f5f | 247 | |
55597f90 FP |
248 | /* boolean "functions" (see init) */ |
249 | logical _wht[0177], _etk[0177], _itk[0177], _btk[0177]; | |
250 | char | |
d8913c1c FP |
251 | /* white chars */ |
252 | *white = " \f\t\n\013", | |
253 | /* token ending chars */ | |
254 | *endtk = " \t\n\013\"'#()[]{}=-+%*/&|^~!<>;,.:?", | |
255 | /* token starting chars */ | |
256 | *begtk = "ABCDEFGHIJKLMNOPQRSTUVWXYZ_abcdefghijklmnopqrstuvwxyz$~@", | |
257 | /* valid in-token chars */ | |
258 | *intk = "ABCDEFGHIJKLMNOPQRSTUVWXYZ_abcdefghijklmnopqrstuvwxyz$0123456789"; | |
c6d46f5f | 259 | |
55597f90 FP |
260 | logical append_to_tagfile; /* -a: append to tags */ |
261 | /* The following three default to TRUE for etags, but to FALSE for ctags. */ | |
262 | logical typedefs; /* -t: create tags for typedefs */ | |
263 | logical typedefs_and_cplusplus; /* -T: create tags for typedefs, level */ | |
c6d46f5f | 264 | /* 0 struct/enum/union decls, and C++ */ |
32daa216 | 265 | /* member functions. */ |
55597f90 | 266 | logical constantypedefs; /* -d: create tags for C #define and enum */ |
32daa216 | 267 | /* constants. Enum consts not implemented. */ |
c6d46f5f | 268 | /* -D: opposite of -d. Default under ctags. */ |
55597f90 FP |
269 | logical update; /* -u: update tags */ |
270 | logical vgrind_style; /* -v: create vgrind style index output */ | |
271 | logical no_warnings; /* -w: suppress warnings */ | |
272 | logical cxref_style; /* -x: create cxref style output */ | |
273 | logical cplusplus; /* .[hc] means C++, not C */ | |
201f9f2b | 274 | logical noindentypedefs; /* -I: ignore indentation in C */ |
c6d46f5f | 275 | |
7537186d FP |
276 | struct option longopts[] = |
277 | { | |
4746118a | 278 | { "append", no_argument, NULL, 'a' }, |
c5007f46 | 279 | { "backward-search", no_argument, NULL, 'B' }, |
4746118a JB |
280 | { "c++", no_argument, NULL, 'C' }, |
281 | { "cxref", no_argument, NULL, 'x' }, | |
282 | { "defines", no_argument, NULL, 'd' }, | |
7537186d | 283 | { "help", no_argument, NULL, 'h' }, |
4746118a | 284 | { "help", no_argument, NULL, 'H' }, |
201f9f2b | 285 | { "ignore-indentation", no_argument, NULL, 'I' }, |
4746118a | 286 | { "include", required_argument, NULL, 'i' }, |
b9755a12 | 287 | { "language", required_argument, NULL, 'l' }, |
4746118a | 288 | { "no-defines", no_argument, NULL, 'D' }, |
b9755a12 | 289 | { "no-regex", no_argument, NULL, 'R' }, |
4746118a JB |
290 | { "no-warn", no_argument, NULL, 'w' }, |
291 | { "output", required_argument, NULL, 'o' }, | |
b9755a12 | 292 | { "regex", required_argument, NULL, 'r' }, |
4746118a JB |
293 | { "typedefs", no_argument, NULL, 't' }, |
294 | { "typedefs-and-c++", no_argument, NULL, 'T' }, | |
c5007f46 | 295 | { "update", no_argument, NULL, 'u' }, |
4746118a | 296 | { "version", no_argument, NULL, 'V' }, |
c5007f46 | 297 | { "vgrind", no_argument, NULL, 'v' }, |
4746118a JB |
298 | { 0 } |
299 | }; | |
300 | ||
b9755a12 | 301 | #ifdef ETAGS_REGEXPS |
c5007f46 | 302 | /* Structure defining a regular expression. Elements are |
b9755a12 FP |
303 | the compiled pattern, and the name string. */ |
304 | struct pattern | |
305 | { | |
306 | struct re_pattern_buffer *pattern; | |
307 | struct re_registers regs; | |
308 | char *name_pattern; | |
309 | logical error_signaled; | |
310 | }; | |
311 | ||
312 | /* Number of regexps found. */ | |
313 | int num_patterns = 0; | |
314 | ||
315 | /* Array of all regexps. */ | |
316 | struct pattern *patterns = NULL; | |
317 | #endif /* ETAGS_REGEXPS */ | |
318 | ||
1f638249 FP |
319 | /* |
320 | * Language stuff. | |
321 | */ | |
b9755a12 | 322 | |
1f638249 FP |
323 | /* Non-NULL if language fixed. */ |
324 | Lang_function *lang_func = NULL; | |
325 | ||
326 | /* Assembly code */ | |
327 | char *Asm_suffixes [] = { "a", /* Unix assembler */ | |
328 | "asm", /* Microcontroller assembly */ | |
329 | "def", /* BSO/Tasking definition includes */ | |
330 | "inc", /* Microcontroller include files */ | |
331 | "ins", /* Microcontroller include files */ | |
332 | "s", "sa", /* Unix assembler */ | |
333 | "src", /* BSO/Tasking C compiler output */ | |
334 | NULL | |
335 | }; | |
336 | ||
337 | /* Note that .c and .h can be considered C++, if the --c++ flag was | |
338 | given. That is why default_C_entries is called here. */ | |
339 | char *default_C_suffixes [] = | |
340 | { "c", "h", NULL }; | |
341 | ||
d8913c1c | 342 | /* .M is for Objective C++ files. */ |
1f638249 | 343 | char *Cplusplus_suffixes [] = |
d8913c1c | 344 | { "C", "H", "c++", "cc", "cpp", "cxx", "h++", "hh", "hpp", "hxx", "M", NULL}; |
1f638249 | 345 | |
1f638249 FP |
346 | char *Cstar_suffixes [] = |
347 | { "cs", "hs", NULL }; | |
348 | ||
1f638249 FP |
349 | char *Fortran_suffixes [] = |
350 | { "F", "f", "f90", "for", NULL }; | |
351 | ||
1f638249 FP |
352 | char *Lisp_suffixes [] = |
353 | { "cl", "clisp", "el", "l", "lisp", "lsp", "ml", NULL }; | |
354 | ||
1f638249 FP |
355 | char *Pascal_suffixes [] = |
356 | { "p", "pas", NULL }; | |
357 | ||
1f638249 FP |
358 | char *Perl_suffixes [] = |
359 | { "pl", "pm", NULL }; | |
360 | char *Perl_interpreters [] = | |
d8913c1c | 361 | { "perl", "@PERL@", NULL }; |
1f638249 | 362 | |
1f638249 | 363 | char *plain_C_suffixes [] = |
d8913c1c FP |
364 | { "pc", /* Pro*C file */ |
365 | "m", /* Objective C file */ | |
366 | "lm", /* Objective lex file */ | |
367 | NULL }; | |
1f638249 | 368 | |
1f638249 FP |
369 | char *Prolog_suffixes [] = |
370 | { "prolog", NULL }; | |
371 | ||
d8913c1c | 372 | /* Can't do the `SCM' or `scm' prefix with a version number. */ |
1f638249 FP |
373 | char *Scheme_suffixes [] = |
374 | { "SCM", "SM", "oak", "sch", "scheme", "scm", "sm", "t", NULL }; | |
375 | ||
1f638249 | 376 | char *TeX_suffixes [] = |
d8913c1c | 377 | { "TeX", "bib", "clo", "cls", "ltx", "sty", "tex", NULL }; |
1f638249 | 378 | |
1f638249 | 379 | char *Yacc_suffixes [] = |
d8913c1c | 380 | { "y", "ym", NULL }; /* .ym is Objective yacc file */ |
1f638249 FP |
381 | |
382 | /* Table of language names and corresponding functions, file suffixes | |
383 | and interpreter names. | |
384 | It is ok for a given function to be listed under more than one | |
b9755a12 | 385 | name. I just didn't. */ |
1f638249 | 386 | struct lang_entry |
b9755a12 | 387 | { |
1f638249 FP |
388 | char *name; |
389 | Lang_function *function; | |
390 | char **suffixes; | |
391 | char **interpreters; | |
b9755a12 FP |
392 | }; |
393 | ||
1f638249 | 394 | struct lang_entry lang_names [] = |
b9755a12 | 395 | { |
d8913c1c FP |
396 | { "asm", Asm_labels, Asm_suffixes, NULL }, |
397 | { "c", default_C_entries, default_C_suffixes, NULL }, | |
398 | { "c++", Cplusplus_entries, Cplusplus_suffixes, NULL }, | |
399 | { "c*", Cstar_entries, Cstar_suffixes, NULL }, | |
400 | { "fortran", Fortran_functions, Fortran_suffixes, NULL }, | |
401 | { "lisp", Lisp_functions, Lisp_suffixes, NULL }, | |
402 | { "pascal", Pascal_functions, Pascal_suffixes, NULL }, | |
403 | { "perl", Perl_functions, Perl_suffixes, Perl_interpreters }, | |
404 | { "proc", plain_C_entries, plain_C_suffixes, NULL }, | |
405 | { "prolog", Prolog_functions, Prolog_suffixes, NULL }, | |
406 | { "scheme", Scheme_functions, Scheme_suffixes, NULL }, | |
407 | { "tex", TeX_functions, TeX_suffixes, NULL }, | |
408 | { "yacc", Yacc_entries, Yacc_suffixes, NULL }, | |
409 | { "auto", NULL }, /* default guessing scheme */ | |
410 | { "none", just_read_file }, /* regexp matching only */ | |
411 | { NULL, NULL } /* end of list */ | |
b9755a12 FP |
412 | }; |
413 | ||
c6d46f5f | 414 | \f |
b9755a12 FP |
415 | void |
416 | print_language_names () | |
417 | { | |
1f638249 FP |
418 | struct lang_entry *lang; |
419 | char **ext; | |
b9755a12 FP |
420 | |
421 | puts ("\nThese are the currently supported languages, along with the\n\ | |
79263656 | 422 | default file name suffixes:"); |
1f638249 | 423 | for (lang = lang_names; lang->name != NULL; lang++) |
b9755a12 | 424 | { |
1f638249 FP |
425 | printf ("\t%s\t", lang->name); |
426 | if (lang->suffixes != NULL) | |
427 | for (ext = lang->suffixes; *ext != NULL; ext++) | |
428 | printf (" .%s", *ext); | |
b9755a12 FP |
429 | puts (""); |
430 | } | |
79263656 FP |
431 | puts ("Where `auto' means use default language for files based on file\n\ |
432 | name suffix, and `none' means only do regexp processing on files.\n\ | |
433 | If no language is specified and no matching suffix is found,\n\ | |
1f638249 FP |
434 | the first line of the file is read for a sharp-bang (#!) sequence\n\ |
435 | followed by the name of an interpreter. If no such sequence is found,\n\ | |
b9755a12 FP |
436 | Fortran is tried first; if no tags are found, C is tried next."); |
437 | } | |
438 | ||
c5007f46 FP |
439 | #ifndef VERSION |
440 | # define VERSION "19" | |
441 | #endif | |
4746118a JB |
442 | void |
443 | print_version () | |
444 | { | |
c5007f46 | 445 | printf ("%s for Emacs version %s\n", (CTAGS) ? "ctags" : "etags", VERSION); |
4746118a | 446 | |
1a0d8c80 | 447 | exit (GOOD); |
4746118a JB |
448 | } |
449 | ||
450 | void | |
451 | print_help () | |
452 | { | |
453 | printf ("These are the options accepted by %s. You may use unambiguous\n\ | |
b9755a12 | 454 | abbreviations for the long option names. A - as file name means read\n\ |
1a0d8c80 | 455 | names from stdin.\n\n", progname); |
4746118a | 456 | |
52cc7c59 JB |
457 | puts ("-a, --append\n\ |
458 | Append tag entries to existing tags file."); | |
1a0d8c80 | 459 | |
32daa216 FP |
460 | if (CTAGS) |
461 | puts ("-B, --backward-search\n\ | |
1a0d8c80 | 462 | Write the search commands for the tag entries using '?', the\n\ |
3ad2882c | 463 | backward-search command instead of '/', the forward-search command."); |
1a0d8c80 | 464 | |
52cc7c59 | 465 | puts ("-C, --c++\n\ |
79263656 | 466 | Treat files whose name suffix defaults to C language as C++ files."); |
4746118a | 467 | |
32daa216 FP |
468 | if (CTAGS) |
469 | puts ("-d, --defines\n\ | |
ee70dba5 | 470 | Create tag entries for constant C #defines, too."); |
32daa216 FP |
471 | else |
472 | puts ("-D, --no-defines\n\ | |
ee70dba5 FP |
473 | Don't create tag entries for constant C #defines. This makes\n\ |
474 | the tags file smaller."); | |
4746118a | 475 | |
32daa216 | 476 | if (!CTAGS) |
b9755a12 FP |
477 | { |
478 | puts ("-i FILE, --include=FILE\n\ | |
1a0d8c80 FP |
479 | Include a note in tag file indicating that, when searching for\n\ |
480 | a tag, one should also consult the tags file FILE after\n\ | |
481 | checking the current file."); | |
b9755a12 FP |
482 | puts ("-l LANG, --language=LANG\n\ |
483 | Force the following files to be considered as written in the\n\ | |
484 | named language up to the next --language=LANG option."); | |
7537186d FP |
485 | } |
486 | ||
b9755a12 | 487 | #ifdef ETAGS_REGEXPS |
7537186d | 488 | puts ("-r /REGEXP/, --regex=/REGEXP/\n\ |
b9755a12 FP |
489 | Make a tag for each line matching pattern REGEXP in the\n\ |
490 | following files. REGEXP is anchored (as if preceded by ^).\n\ | |
491 | The form /REGEXP/NAME/ creates a named tag. For example Tcl\n\ | |
492 | named tags can be created with:\n\ | |
493 | --regex=/proc[ \\t]+\\([^ \\t]+\\)/\\1/."); | |
7537186d | 494 | puts ("-R, --no-regex\n\ |
b9755a12 FP |
495 | Don't create tags from regexps for the following files."); |
496 | #endif /* ETAGS_REGEXPS */ | |
1a0d8c80 FP |
497 | puts ("-o FILE, --output=FILE\n\ |
498 | Write the tags to FILE."); | |
201f9f2b | 499 | puts ("-I, --ignore-indentation\n\ |
4746118a JB |
500 | Don't rely on indentation quite as much as normal. Currently,\n\ |
501 | this means not to assume that a closing brace in the first\n\ | |
502 | column is the final brace of a function or structure\n\ | |
32daa216 | 503 | definition in C and C++."); |
4746118a | 504 | |
32daa216 FP |
505 | if (CTAGS) |
506 | { | |
507 | puts ("-t, --typedefs\n\ | |
508 | Generate tag entries for C typedefs."); | |
509 | puts ("-T, --typedefs-and-c++\n\ | |
510 | Generate tag entries for C typedefs, C struct/enum/union tags,\n\ | |
511 | and C++ member functions."); | |
32daa216 | 512 | puts ("-u, --update\n\ |
4746118a JB |
513 | Update the tag entries for the given files, leaving tag\n\ |
514 | entries for other files in place. Currently, this is\n\ | |
515 | implemented by deleting the existing entries for the given\n\ | |
516 | files and then rewriting the new entries at the end of the\n\ | |
517 | tags file. It is often faster to simply rebuild the entire\n\ | |
52cc7c59 | 518 | tag file than to use this."); |
32daa216 | 519 | puts ("-v, --vgrind\n\ |
4746118a JB |
520 | Generates an index of items intended for human consumption,\n\ |
521 | similar to the output of vgrind. The index is sorted, and\n\ | |
52cc7c59 | 522 | gives the page number of each item."); |
b9755a12 FP |
523 | puts ("-w, --no-warn\n\ |
524 | Suppress warning messages about entries defined in multiple\n\ | |
525 | files."); | |
32daa216 | 526 | puts ("-x, --cxref\n\ |
4746118a JB |
527 | Like --vgrind, but in the style of cxref, rather than vgrind.\n\ |
528 | The output uses line numbers instead of page numbers, but\n\ | |
529 | beyond that the differences are cosmetic; try both to see\n\ | |
52cc7c59 | 530 | which you like."); |
32daa216 | 531 | } |
4746118a JB |
532 | |
533 | puts ("-V, --version\n\ | |
534 | Print the version of the program.\n\ | |
7537186d | 535 | -h, --help\n\ |
4746118a JB |
536 | Print this help message."); |
537 | ||
b9755a12 FP |
538 | print_language_names (); |
539 | ||
1a0d8c80 | 540 | exit (GOOD); |
4746118a JB |
541 | } |
542 | ||
543 | \f | |
b9755a12 FP |
544 | enum argument_type |
545 | { | |
546 | at_language, | |
547 | at_regexp, | |
548 | at_filename | |
549 | }; | |
550 | ||
551 | /* This structure helps us allow mixing of --lang and filenames. */ | |
552 | typedef struct | |
553 | { | |
554 | enum argument_type arg_type; | |
555 | char *what; | |
556 | Lang_function *function; | |
1f638249 | 557 | } argument; |
b9755a12 FP |
558 | |
559 | #ifdef VMS /* VMS specific functions */ | |
560 | ||
561 | #define EOS '\0' | |
562 | ||
563 | /* This is a BUG! ANY arbitrary limit is a BUG! | |
564 | Won't someone please fix this? */ | |
565 | #define MAX_FILE_SPEC_LEN 255 | |
566 | typedef struct { | |
567 | short curlen; | |
568 | char body[MAX_FILE_SPEC_LEN + 1]; | |
569 | } vspec; | |
570 | ||
571 | /* | |
572 | v1.05 nmm 26-Jun-86 fn_exp - expand specification of list of file names | |
573 | returning in each successive call the next filename matching the input | |
574 | spec. The function expects that each in_spec passed | |
575 | to it will be processed to completion; in particular, up to and | |
576 | including the call following that in which the last matching name | |
577 | is returned, the function ignores the value of in_spec, and will | |
c5007f46 | 578 | only start processing a new spec with the following call. |
b9755a12 FP |
579 | If an error occurs, on return out_spec contains the value |
580 | of in_spec when the error occurred. | |
581 | ||
582 | With each successive filename returned in out_spec, the | |
583 | function's return value is one. When there are no more matching | |
584 | names the function returns zero. If on the first call no file | |
c5007f46 | 585 | matches in_spec, or there is any other error, -1 is returned. |
b9755a12 FP |
586 | */ |
587 | ||
588 | #include <rmsdef.h> | |
589 | #include <descrip.h> | |
590 | #define OUTSIZE MAX_FILE_SPEC_LEN | |
591 | short | |
592 | fn_exp (out, in) | |
593 | vspec *out; | |
594 | char *in; | |
595 | { | |
596 | static long context = 0; | |
597 | static struct dsc$descriptor_s o; | |
598 | static struct dsc$descriptor_s i; | |
599 | static logical pass1 = TRUE; | |
600 | long status; | |
601 | short retval; | |
602 | ||
603 | if (pass1) | |
604 | { | |
605 | pass1 = FALSE; | |
606 | o.dsc$a_pointer = (char *) out; | |
607 | o.dsc$w_length = (short)OUTSIZE; | |
608 | i.dsc$a_pointer = in; | |
609 | i.dsc$w_length = (short)strlen(in); | |
610 | i.dsc$b_dtype = DSC$K_DTYPE_T; | |
611 | i.dsc$b_class = DSC$K_CLASS_S; | |
612 | o.dsc$b_dtype = DSC$K_DTYPE_VT; | |
613 | o.dsc$b_class = DSC$K_CLASS_VS; | |
614 | } | |
615 | if ((status = lib$find_file(&i, &o, &context, 0, 0)) == RMS$_NORMAL) | |
616 | { | |
617 | out->body[out->curlen] = EOS; | |
618 | return 1; | |
619 | } | |
620 | else if (status == RMS$_NMF) | |
621 | retval = 0; | |
622 | else | |
623 | { | |
624 | strcpy(out->body, in); | |
625 | retval = -1; | |
626 | } | |
627 | lib$find_file_end(&context); | |
628 | pass1 = TRUE; | |
629 | return retval; | |
c5007f46 | 630 | } |
b9755a12 FP |
631 | |
632 | /* | |
c5007f46 | 633 | v1.01 nmm 19-Aug-85 gfnames - return in successive calls the |
b9755a12 FP |
634 | name of each file specified by the provided arg expanding wildcards. |
635 | */ | |
636 | char * | |
637 | gfnames (arg, p_error) | |
638 | char *arg; | |
639 | logical *p_error; | |
640 | { | |
641 | static vspec filename = {MAX_FILE_SPEC_LEN, "\0"}; | |
642 | ||
643 | switch (fn_exp (&filename, arg)) | |
644 | { | |
645 | case 1: | |
646 | *p_error = FALSE; | |
647 | return filename.body; | |
648 | case 0: | |
649 | *p_error = FALSE; | |
650 | return NULL; | |
651 | default: | |
652 | *p_error = TRUE; | |
653 | return filename.body; | |
654 | } | |
655 | } | |
656 | ||
657 | #ifndef OLD /* Newer versions of VMS do provide `system'. */ | |
658 | system (cmd) | |
659 | char *cmd; | |
660 | { | |
661 | fprintf (stderr, "system() function not implemented under VMS\n"); | |
662 | } | |
663 | #endif | |
664 | ||
665 | #define VERSION_DELIM ';' | |
666 | char *massage_name (s) | |
667 | char *s; | |
668 | { | |
c5007f46 | 669 | char *start = s; |
b9755a12 FP |
670 | |
671 | for ( ; *s; s++) | |
672 | if (*s == VERSION_DELIM) | |
673 | { | |
674 | *s = EOS; | |
675 | break; | |
676 | } | |
677 | else | |
c5007f46 | 678 | *s = lowcase (*s); |
b9755a12 FP |
679 | return start; |
680 | } | |
681 | #endif /* VMS */ | |
682 | ||
683 | \f | |
c6d46f5f JB |
684 | void |
685 | main (argc, argv) | |
686 | int argc; | |
687 | char *argv[]; | |
688 | { | |
c6d46f5f | 689 | int i; |
1e134a5f | 690 | unsigned int nincluded_files = 0; |
72a339d7 | 691 | char **included_files = xnew (argc, char *); |
c6d46f5f | 692 | char *this_file; |
1f638249 | 693 | argument *argbuffer; |
b9755a12 | 694 | int current_arg = 0, file_count = 0; |
55597f90 | 695 | struct linebuffer filename_lb; |
c6d46f5f | 696 | #ifdef VMS |
b9755a12 | 697 | logical got_err; |
c6d46f5f | 698 | #endif |
c5007f46 | 699 | |
c05b6df5 | 700 | #ifdef DOS_NT |
42680d3c | 701 | _fmode = O_BINARY; /* all of files are treated as binary files */ |
c05b6df5 | 702 | #endif /* DOS_NT */ |
c6880c90 | 703 | |
c6d46f5f JB |
704 | progname = argv[0]; |
705 | ||
b9755a12 FP |
706 | /* Allocate enough no matter what happens. Overkill, but each one |
707 | is small. */ | |
1f638249 | 708 | argbuffer = xnew (argc, argument); |
b9755a12 FP |
709 | |
710 | #ifdef ETAGS_REGEXPS | |
711 | /* Set syntax for regular expression routines. */ | |
712 | re_set_syntax (RE_SYNTAX_EMACS); | |
713 | #endif /* ETAGS_REGEXPS */ | |
714 | ||
c6d46f5f JB |
715 | /* |
716 | * If etags, always find typedefs and structure tags. Why not? | |
717 | * Also default is to find macro constants. | |
718 | */ | |
32daa216 | 719 | if (!CTAGS) |
55597f90 | 720 | typedefs = typedefs_and_cplusplus = constantypedefs = TRUE; |
c6d46f5f | 721 | |
dcc89e63 | 722 | while (1) |
c6d46f5f | 723 | { |
b9755a12 | 724 | int opt = getopt_long (argc, argv, |
201f9f2b | 725 | "-aCdDf:Il:o:r:RStTi:BuvxwVhH", longopts, 0); |
4746118a JB |
726 | |
727 | if (opt == EOF) | |
728 | break; | |
729 | ||
730 | switch (opt) | |
c6d46f5f | 731 | { |
b02c5fea FP |
732 | case 0: |
733 | /* If getopt returns 0, then it has already processed a | |
4746118a JB |
734 | long-named option. We should do nothing. */ |
735 | break; | |
736 | ||
b9755a12 FP |
737 | case 1: |
738 | /* This means that a filename has been seen. Record it. */ | |
739 | argbuffer[current_arg].arg_type = at_filename; | |
740 | argbuffer[current_arg].what = optarg; | |
741 | ++current_arg; | |
742 | ++file_count; | |
743 | break; | |
744 | ||
4746118a JB |
745 | /* Common options. */ |
746 | case 'a': | |
55597f90 | 747 | append_to_tagfile = TRUE; |
4746118a JB |
748 | break; |
749 | case 'C': | |
55597f90 | 750 | cplusplus = TRUE; |
4746118a JB |
751 | break; |
752 | case 'd': | |
55597f90 | 753 | constantypedefs = TRUE; |
4746118a JB |
754 | break; |
755 | case 'D': | |
55597f90 | 756 | constantypedefs = FALSE; |
4746118a | 757 | break; |
32daa216 | 758 | case 'f': /* for compatibility with old makefiles */ |
4746118a | 759 | case 'o': |
6dd5561c | 760 | if (tagfile) |
c6d46f5f | 761 | { |
201f9f2b FP |
762 | fprintf (stderr, "%s: -%c option may only be given once.\n", |
763 | progname, opt); | |
d8913c1c | 764 | suggest_asking_for_help (); |
c6d46f5f | 765 | } |
6dd5561c | 766 | tagfile = optarg; |
4746118a | 767 | break; |
201f9f2b FP |
768 | case 'I': |
769 | case 'S': /* for backward compatibility */ | |
770 | noindentypedefs = TRUE; | |
771 | break; | |
b9755a12 | 772 | case 'l': |
1f638249 | 773 | argbuffer[current_arg].function = get_language_from_name (optarg); |
b9755a12 FP |
774 | argbuffer[current_arg].arg_type = at_language; |
775 | ++current_arg; | |
776 | break; | |
777 | #ifdef ETAGS_REGEXPS | |
778 | case 'r': | |
779 | argbuffer[current_arg].arg_type = at_regexp; | |
780 | argbuffer[current_arg].what = optarg; | |
781 | ++current_arg; | |
782 | break; | |
783 | case 'R': | |
784 | argbuffer[current_arg].arg_type = at_regexp; | |
785 | argbuffer[current_arg].what = NULL; | |
786 | ++current_arg; | |
787 | break; | |
788 | #endif /* ETAGS_REGEXPS */ | |
4746118a JB |
789 | case 'V': |
790 | print_version (); | |
791 | break; | |
7537186d | 792 | case 'h': |
4746118a JB |
793 | case 'H': |
794 | print_help (); | |
795 | break; | |
0c1fd2e3 | 796 | case 't': |
55597f90 | 797 | typedefs = TRUE; |
0c1fd2e3 FP |
798 | break; |
799 | case 'T': | |
55597f90 | 800 | typedefs = typedefs_and_cplusplus = TRUE; |
0c1fd2e3 | 801 | break; |
b02c5fea | 802 | #if (!CTAGS) |
4746118a JB |
803 | /* Etags options */ |
804 | case 'i': | |
4746118a JB |
805 | included_files[nincluded_files++] = optarg; |
806 | break; | |
b02c5fea | 807 | #else /* CTAGS */ |
4746118a JB |
808 | /* Ctags options. */ |
809 | case 'B': | |
810 | searchar = '?'; | |
4746118a | 811 | break; |
4746118a | 812 | case 'u': |
55597f90 | 813 | update = TRUE; |
4746118a JB |
814 | break; |
815 | case 'v': | |
55597f90 | 816 | vgrind_style = TRUE; |
4746118a JB |
817 | /*FALLTHRU*/ |
818 | case 'x': | |
55597f90 | 819 | cxref_style = TRUE; |
4746118a JB |
820 | break; |
821 | case 'w': | |
55597f90 | 822 | no_warnings = TRUE; |
4746118a | 823 | break; |
b02c5fea | 824 | #endif /* CTAGS */ |
4746118a | 825 | default: |
d8913c1c | 826 | suggest_asking_for_help (); |
c6d46f5f | 827 | } |
c6d46f5f JB |
828 | } |
829 | ||
b9755a12 FP |
830 | for (; optind < argc; ++optind) |
831 | { | |
832 | argbuffer[current_arg].arg_type = at_filename; | |
833 | argbuffer[current_arg].what = argv[optind]; | |
834 | ++current_arg; | |
835 | ++file_count; | |
836 | } | |
837 | ||
838 | if (nincluded_files == 0 && file_count == 0) | |
c6d46f5f | 839 | { |
4746118a | 840 | fprintf (stderr, "%s: No input files specified.\n", progname); |
d8913c1c | 841 | suggest_asking_for_help (); |
c6d46f5f JB |
842 | } |
843 | ||
6dd5561c | 844 | if (tagfile == NULL) |
c6d46f5f | 845 | { |
6dd5561c | 846 | tagfile = CTAGS ? "tags" : "TAGS"; |
c6d46f5f | 847 | } |
b02c5fea | 848 | cwd = etags_getcwd (); /* the current working directory */ |
46c145db | 849 | strcat (cwd, "/"); |
6dd5561c | 850 | if (streq (tagfile, "-")) |
46c145db | 851 | { |
6dd5561c | 852 | tagfiledir = cwd; |
46c145db FP |
853 | } |
854 | else | |
855 | { | |
6dd5561c | 856 | tagfiledir = absolute_dirname (tagfile, cwd); |
46c145db | 857 | } |
c6d46f5f | 858 | |
b9755a12 | 859 | init (); /* set up boolean "functions" */ |
c6d46f5f JB |
860 | |
861 | initbuffer (&lb); | |
75bdbc6a | 862 | initbuffer (&token_name); |
13fde0cd RS |
863 | initbuffer (&lbs[0].lb); |
864 | initbuffer (&lbs[1].lb); | |
c6d46f5f | 865 | initbuffer (&filename_lb); |
b9755a12 | 866 | |
32daa216 | 867 | if (!CTAGS) |
c6d46f5f | 868 | { |
6dd5561c FP |
869 | if (streq (tagfile, "-")) |
870 | tagf = stdout; | |
c6d46f5f | 871 | else |
6dd5561c FP |
872 | tagf = fopen (tagfile, append_to_tagfile ? "a" : "w"); |
873 | if (tagf == NULL) | |
cdc1f6a7 | 874 | pfatal (tagfile); |
c6d46f5f JB |
875 | } |
876 | ||
b9755a12 FP |
877 | /* |
878 | * Loop through files finding functions. | |
879 | */ | |
880 | for (i = 0; i < current_arg; ++i) | |
c6d46f5f | 881 | { |
b9755a12 | 882 | switch (argbuffer[i].arg_type) |
c6d46f5f | 883 | { |
b9755a12 FP |
884 | case at_language: |
885 | lang_func = argbuffer[i].function; | |
886 | break; | |
887 | #ifdef ETAGS_REGEXPS | |
888 | case at_regexp: | |
889 | add_regex (argbuffer[i].what); | |
890 | break; | |
c6d46f5f | 891 | #endif |
b9755a12 FP |
892 | case at_filename: |
893 | #ifdef VMS | |
894 | while ((this_file = gfnames (argbuffer[i].what, &got_err)) != NULL) | |
895 | { | |
896 | if (got_err) | |
897 | { | |
898 | error ("Can't find file %s\n", this_file); | |
899 | argc--, argv++; | |
900 | } | |
901 | else | |
902 | { | |
903 | this_file = massage_name (this_file); | |
904 | } | |
c6d46f5f | 905 | #else |
b9755a12 | 906 | this_file = argbuffer[i].what; |
c6d46f5f | 907 | #endif |
b9755a12 FP |
908 | /* Input file named "-" means read file names from stdin |
909 | and use them. */ | |
910 | if (streq (this_file, "-")) | |
9cb0aa73 FP |
911 | while (readline_internal (&filename_lb, stdin) > 0) |
912 | process_file (filename_lb.buffer); | |
b9755a12 FP |
913 | else |
914 | process_file (this_file); | |
915 | #ifdef VMS | |
c6d46f5f | 916 | } |
b9755a12 FP |
917 | #endif |
918 | break; | |
c6d46f5f | 919 | } |
46c145db | 920 | } |
9cb0aa73 | 921 | |
32daa216 | 922 | if (!CTAGS) |
c6d46f5f | 923 | { |
1e134a5f | 924 | while (nincluded_files-- > 0) |
6dd5561c | 925 | fprintf (tagf, "\f\n%s,include\n", *included_files++); |
1e134a5f | 926 | |
55597f90 | 927 | fclose (tagf); |
1a0d8c80 | 928 | exit (GOOD); |
c6d46f5f JB |
929 | } |
930 | ||
55597f90 FP |
931 | /* If CTAGS, we are here. process_file did not write the tags yet, |
932 | because we want them ordered. Let's do it now. */ | |
c6d46f5f JB |
933 | if (cxref_style) |
934 | { | |
55597f90 FP |
935 | tagf = fopen (tagfile, append_to_tagfile ? "a" : "w"); |
936 | if (tagf == NULL) | |
937 | pfatal (tagfile); | |
c6d46f5f JB |
938 | put_entries (head); |
939 | exit (GOOD); | |
940 | } | |
55597f90 | 941 | |
4746118a | 942 | if (update) |
c6d46f5f | 943 | { |
55597f90 | 944 | char cmd[BUFSIZ]; |
b9755a12 | 945 | for (i = 0; i < current_arg; ++i) |
c6d46f5f | 946 | { |
55597f90 | 947 | if (argbuffer[i].arg_type != at_filename) |
b9755a12 | 948 | continue; |
c6d46f5f JB |
949 | sprintf (cmd, |
950 | "mv %s OTAGS;fgrep -v '\t%s\t' OTAGS >%s;rm OTAGS", | |
b9755a12 | 951 | tagfile, argbuffer[i].what, tagfile); |
55597f90 FP |
952 | if (system (cmd) != GOOD) |
953 | fatal ("failed to execute shell command"); | |
c6d46f5f | 954 | } |
55597f90 | 955 | append_to_tagfile = TRUE; |
c6d46f5f | 956 | } |
55597f90 | 957 | |
6dd5561c FP |
958 | tagf = fopen (tagfile, append_to_tagfile ? "a" : "w"); |
959 | if (tagf == NULL) | |
55597f90 | 960 | pfatal (tagfile); |
c6d46f5f | 961 | put_entries (head); |
55597f90 FP |
962 | fclose (tagf); |
963 | ||
c6d46f5f JB |
964 | if (update) |
965 | { | |
55597f90 | 966 | char cmd[BUFSIZ]; |
6dd5561c | 967 | sprintf (cmd, "sort %s -o %s", tagfile, tagfile); |
55597f90 | 968 | exit (system (cmd)); |
c6d46f5f JB |
969 | } |
970 | exit (GOOD); | |
971 | } | |
972 | ||
973 | ||
b9755a12 | 974 | /* |
1f638249 | 975 | * Return a Lang_function given the name. |
b9755a12 | 976 | */ |
1f638249 FP |
977 | Lang_function * |
978 | get_language_from_name (name) | |
979 | char *name; | |
b9755a12 FP |
980 | { |
981 | struct lang_entry *lang; | |
982 | ||
d8913c1c FP |
983 | if (name != NULL) |
984 | for (lang = lang_names; lang->name != NULL; lang++) | |
985 | { | |
986 | if (streq (name, lang->name)) | |
987 | return lang->function; | |
988 | } | |
b9755a12 | 989 | |
d8913c1c FP |
990 | fprintf (stderr, "%s: language \"%s\" not recognized.\n", |
991 | progname, optarg); | |
992 | suggest_asking_for_help (); | |
993 | ||
994 | /* This point should never be reached. The function should either | |
995 | return a function pointer or never return. Note that a NULL | |
996 | pointer cannot be considered as an error, as it means that the | |
997 | language has not been explicitely imposed by the user ("auto"). */ | |
998 | return NULL; /* avoid warnings from compiler */ | |
1f638249 FP |
999 | } |
1000 | ||
1001 | ||
1002 | /* | |
1003 | * Return a Lang_function given the interpreter name. | |
1004 | */ | |
1005 | Lang_function * | |
1006 | get_language_from_interpreter (interpreter) | |
1007 | char *interpreter; | |
1008 | { | |
1009 | struct lang_entry *lang; | |
1010 | char **iname; | |
1011 | ||
1012 | if (interpreter == NULL) | |
1013 | return NULL; | |
1014 | for (lang = lang_names; lang->name != NULL; lang++) | |
1015 | if (lang->interpreters != NULL) | |
1016 | for (iname = lang->interpreters; *iname != NULL; iname++) | |
1017 | if (streq (*iname, interpreter)) | |
1018 | return lang->function; | |
1019 | ||
1020 | return NULL; | |
1021 | } | |
1022 | ||
1023 | ||
1024 | ||
1025 | /* | |
1026 | * Return a Lang_function given the file suffix. | |
1027 | */ | |
1028 | Lang_function * | |
1029 | get_language_from_suffix (suffix) | |
1030 | char *suffix; | |
1031 | { | |
1032 | struct lang_entry *lang; | |
1033 | char **ext; | |
1034 | ||
1035 | if (suffix == NULL) | |
1036 | return NULL; | |
1037 | for (lang = lang_names; lang->name != NULL; lang++) | |
1038 | if (lang->suffixes != NULL) | |
1039 | for (ext = lang->suffixes; *ext != NULL; ext++) | |
1040 | if (streq (*ext, suffix)) | |
1041 | return lang->function; | |
1042 | ||
1043 | return NULL; | |
b9755a12 FP |
1044 | } |
1045 | ||
1046 | ||
c6d46f5f JB |
1047 | /* |
1048 | * This routine is called on each file argument. | |
1049 | */ | |
1050 | void | |
1051 | process_file (file) | |
1052 | char *file; | |
1053 | { | |
1054 | struct stat stat_buf; | |
55597f90 | 1055 | FILE *inf; |
c6d46f5f | 1056 | |
42680d3c | 1057 | if (stat (file, &stat_buf) == 0 && !S_ISREG (stat_buf.st_mode)) |
c6d46f5f JB |
1058 | { |
1059 | fprintf (stderr, "Skipping %s: it is not a regular file.\n", file); | |
1060 | return; | |
1061 | } | |
6dd5561c | 1062 | if (streq (file, tagfile) && !streq (tagfile, "-")) |
c6d46f5f JB |
1063 | { |
1064 | fprintf (stderr, "Skipping inclusion of %s in self.\n", file); | |
1065 | return; | |
1066 | } | |
55597f90 FP |
1067 | inf = fopen (file, "r"); |
1068 | if (inf == NULL) | |
42680d3c | 1069 | { |
55597f90 | 1070 | perror (file); |
42680d3c FP |
1071 | return; |
1072 | } | |
55597f90 FP |
1073 | |
1074 | find_entries (file, inf); | |
1075 | ||
32daa216 | 1076 | if (!CTAGS) |
c6d46f5f | 1077 | { |
46c145db FP |
1078 | char *filename; |
1079 | ||
b2db879b | 1080 | if (absolutefn (file)) |
46c145db FP |
1081 | { |
1082 | /* file is an absolute filename. Canonicalise it. */ | |
1083 | filename = absolute_filename (file, cwd); | |
1084 | } | |
1085 | else | |
1086 | { | |
1087 | /* file is a filename relative to cwd. Make it relative | |
1088 | to the directory of the tags file. */ | |
6dd5561c | 1089 | filename = relative_filename (file, tagfiledir); |
46c145db | 1090 | } |
6dd5561c | 1091 | fprintf (tagf, "\f\n%s,%d\n", filename, total_size_of_entries (head)); |
a8d9bd4b | 1092 | free (filename); |
c6d46f5f JB |
1093 | put_entries (head); |
1094 | free_tree (head); | |
1095 | head = NULL; | |
1096 | } | |
1097 | } | |
1098 | ||
1099 | /* | |
eb8c3be9 | 1100 | * This routine sets up the boolean pseudo-functions which work |
99e0a2e0 | 1101 | * by setting boolean flags dependent upon the corresponding character |
c6d46f5f JB |
1102 | * Every char which is NOT in that string is not a white char. Therefore, |
1103 | * all of the array "_wht" is set to FALSE, and then the elements | |
1104 | * subscripted by the chars in "white" are set to TRUE. Thus "_wht" | |
1105 | * of a char is TRUE if it is the string "white", else FALSE. | |
1106 | */ | |
1107 | void | |
1108 | init () | |
1109 | { | |
13fde0cd RS |
1110 | register char *sp; |
1111 | register int i; | |
c6d46f5f JB |
1112 | |
1113 | for (i = 0; i < 0177; i++) | |
13fde0cd | 1114 | _wht[i] = _etk[i] = _itk[i] = _btk[i] = FALSE; |
c6d46f5f JB |
1115 | for (sp = white; *sp; sp++) |
1116 | _wht[*sp] = TRUE; | |
1117 | for (sp = endtk; *sp; sp++) | |
1118 | _etk[*sp] = TRUE; | |
1119 | for (sp = intk; *sp; sp++) | |
1120 | _itk[*sp] = TRUE; | |
1121 | for (sp = begtk; *sp; sp++) | |
1122 | _btk[*sp] = TRUE; | |
c6d46f5f JB |
1123 | _wht[0] = _wht['\n']; |
1124 | _etk[0] = _etk['\n']; | |
1125 | _btk[0] = _btk['\n']; | |
1126 | _itk[0] = _itk['\n']; | |
c6d46f5f JB |
1127 | } |
1128 | ||
1129 | /* | |
1130 | * This routine opens the specified file and calls the function | |
1131 | * which finds the function and type definitions. | |
1132 | */ | |
55597f90 FP |
1133 | void |
1134 | find_entries (file, inf) | |
c6d46f5f | 1135 | char *file; |
55597f90 | 1136 | FILE *inf; |
c6d46f5f | 1137 | { |
b9755a12 | 1138 | char *cp; |
1f638249 | 1139 | Lang_function *function; |
b9755a12 FP |
1140 | NODE *old_last_node; |
1141 | extern NODE *last_node; | |
c6d46f5f | 1142 | |
d8913c1c | 1143 | |
c5007f46 FP |
1144 | /* Memory leakage here: the memory block pointed by curfile is never |
1145 | released. The amount of memory leaked here is the sum of the | |
1146 | lengths of the input file names. */ | |
c6d46f5f | 1147 | curfile = savestr (file); |
c6d46f5f | 1148 | |
b9755a12 | 1149 | /* If user specified a language, use it. */ |
1f638249 FP |
1150 | function = lang_func; |
1151 | if (function != NULL) | |
13fde0cd | 1152 | { |
1f638249 | 1153 | function (inf); |
b9755a12 | 1154 | fclose (inf); |
55597f90 | 1155 | return; |
13fde0cd | 1156 | } |
b9755a12 | 1157 | |
1f638249 FP |
1158 | cp = etags_strrchr (file, '.'); |
1159 | if (cp != NULL) | |
1160 | { | |
1161 | cp += 1; | |
1162 | function = get_language_from_suffix (cp); | |
1163 | if (function != NULL) | |
1164 | { | |
1165 | function (inf); | |
1166 | fclose (inf); | |
1167 | return; | |
1168 | } | |
1169 | } | |
1170 | ||
1171 | /* Look for sharp-bang as the first two characters. */ | |
1172 | if (readline_internal (&lb, inf) > 2 | |
1173 | && lb.buffer[0] == '#' | |
1174 | && lb.buffer[1] == '!') | |
c6d46f5f | 1175 | { |
1f638249 FP |
1176 | char *lp; |
1177 | ||
1178 | /* Set lp to point at the first char after the last slash in the | |
1179 | line or, if no slashes, at the first nonblank. Then set cp to | |
1180 | the first successive blank and terminate the string. */ | |
1181 | lp = etags_strrchr (lb.buffer+2, '/'); | |
1182 | if (lp != NULL) | |
1183 | lp += 1; | |
1184 | else | |
1185 | for (lp = lb.buffer+2; *lp != '\0' && isspace (*lp); lp++) | |
1186 | continue; | |
1187 | for (cp = lp; *cp != '\0' && !isspace (*cp); cp++) | |
1188 | continue; | |
1189 | *cp = '\0'; | |
1190 | ||
1191 | if (strlen (lp) > 0) | |
b9755a12 | 1192 | { |
1f638249 FP |
1193 | function = get_language_from_interpreter (lp); |
1194 | if (function != NULL) | |
b9755a12 | 1195 | { |
1f638249 | 1196 | function (inf); |
b9755a12 | 1197 | fclose (inf); |
55597f90 | 1198 | return; |
b9755a12 FP |
1199 | } |
1200 | } | |
c6d46f5f | 1201 | } |
1f638249 | 1202 | rewind (inf); |
c6d46f5f | 1203 | |
b9755a12 FP |
1204 | /* Try Fortran. */ |
1205 | old_last_node = last_node; | |
1206 | Fortran_functions (inf); | |
c6d46f5f | 1207 | |
b9755a12 FP |
1208 | /* No Fortran entries found. Try C. */ |
1209 | if (old_last_node == last_node) | |
b2db879b FP |
1210 | { |
1211 | rewind (inf); | |
1212 | default_C_entries (inf); | |
1213 | } | |
b9755a12 | 1214 | fclose (inf); |
1f638249 | 1215 | return; |
c6d46f5f JB |
1216 | } |
1217 | \f | |
1218 | /* Record a tag. */ | |
c6d46f5f | 1219 | void |
108c932a | 1220 | pfnote (name, is_func, linestart, linelen, lno, cno) |
d8913c1c | 1221 | char *name; /* tag name, or NULL if unnamed */ |
55597f90 | 1222 | logical is_func; /* tag is a function */ |
55597f90 FP |
1223 | char *linestart; /* start of the line where tag is */ |
1224 | int linelen; /* length of the line where tag is */ | |
1225 | int lno; /* line number */ | |
1226 | long cno; /* character number */ | |
1227 | { | |
d8913c1c FP |
1228 | register NODE *np; |
1229 | ||
1230 | if (CTAGS && name == NULL) | |
1231 | return; | |
1232 | ||
1233 | np = xnew (1, NODE); | |
c6d46f5f | 1234 | |
c6d46f5f | 1235 | /* If ctags mode, change name "main" to M<thisfilename>. */ |
32daa216 | 1236 | if (CTAGS && !cxref_style && streq (name, "main")) |
c6d46f5f | 1237 | { |
108c932a | 1238 | register char *fp = etags_strrchr (curfile, '/'); |
55597f90 FP |
1239 | np->name = concat ("M", fp == 0 ? curfile : fp + 1, ""); |
1240 | fp = etags_strrchr (np->name, '.'); | |
c6d46f5f | 1241 | if (fp && fp[1] != '\0' && fp[2] == '\0') |
55597f90 | 1242 | fp[0] = 0; |
55597f90 FP |
1243 | } |
1244 | else | |
108c932a | 1245 | np->name = name; |
4b533b5b | 1246 | np->been_warned = FALSE; |
c6d46f5f JB |
1247 | np->file = curfile; |
1248 | np->is_func = is_func; | |
c6d46f5f | 1249 | np->lno = lno; |
aab1fdae FP |
1250 | /* Our char numbers are 0-base, because of C language tradition? |
1251 | ctags compatibility? old versions compatibility? I don't know. | |
15294654 | 1252 | Anyway, since emacs's are 1-base we expect etags.el to take care |
aab1fdae FP |
1253 | of the difference. If we wanted to have 1-based numbers, we would |
1254 | uncomment the +1 below. */ | |
1255 | np->cno = cno /* + 1 */ ; | |
55597f90 | 1256 | np->left = np->right = NULL; |
d8913c1c FP |
1257 | if (CTAGS && !cxref_style) |
1258 | { | |
1259 | if (strlen (linestart) < 50) | |
1260 | np->pat = concat (linestart, "$", ""); | |
1261 | else | |
1262 | np->pat = savenstr (linestart, 50); | |
1263 | } | |
1264 | else | |
1265 | np->pat = savenstr (linestart, linelen); | |
c6d46f5f JB |
1266 | |
1267 | add_node (np, &head); | |
1268 | } | |
1269 | ||
1270 | /* | |
1271 | * free_tree () | |
1272 | * recurse on left children, iterate on right children. | |
1273 | */ | |
1274 | void | |
1275 | free_tree (node) | |
1276 | register NODE *node; | |
1277 | { | |
1278 | while (node) | |
1279 | { | |
1280 | register NODE *node_right = node->right; | |
1281 | free_tree (node->left); | |
108c932a | 1282 | if (node->name != NULL) |
55597f90 | 1283 | free (node->name); |
c6d46f5f JB |
1284 | free (node->pat); |
1285 | free ((char *) node); | |
1286 | node = node_right; | |
1287 | } | |
1288 | } | |
1289 | ||
1290 | /* | |
1291 | * add_node () | |
1292 | * Adds a node to the tree of nodes. In etags mode, we don't keep | |
1293 | * it sorted; we just keep a linear list. In ctags mode, maintain | |
1294 | * an ordered tree, with no attempt at balancing. | |
1295 | * | |
1296 | * add_node is the only function allowed to add nodes, so it can | |
1297 | * maintain state. | |
1298 | */ | |
6dd5561c | 1299 | NODE *last_node = NULL; |
c6d46f5f JB |
1300 | void |
1301 | add_node (node, cur_node_p) | |
1302 | NODE *node, **cur_node_p; | |
1303 | { | |
1304 | register int dif; | |
1305 | register NODE *cur_node = *cur_node_p; | |
c6d46f5f JB |
1306 | |
1307 | if (cur_node == NULL) | |
1308 | { | |
1309 | *cur_node_p = node; | |
1310 | last_node = node; | |
1311 | return; | |
1312 | } | |
1313 | ||
32daa216 | 1314 | if (!CTAGS) |
c6d46f5f JB |
1315 | { |
1316 | /* Etags Mode */ | |
1a0d8c80 FP |
1317 | if (last_node == NULL) |
1318 | fatal ("internal error in add_node", 0); | |
c6d46f5f JB |
1319 | last_node->right = node; |
1320 | last_node = node; | |
1321 | } | |
1322 | else | |
1323 | { | |
1324 | /* Ctags Mode */ | |
1325 | dif = strcmp (node->name, cur_node->name); | |
1326 | ||
1327 | /* | |
1328 | * If this tag name matches an existing one, then | |
1329 | * do not add the node, but maybe print a warning. | |
1330 | */ | |
1331 | if (!dif) | |
1332 | { | |
108c932a | 1333 | if (streq (node->file, cur_node->file)) |
c6d46f5f JB |
1334 | { |
1335 | if (!no_warnings) | |
1336 | { | |
1337 | fprintf (stderr, "Duplicate entry in file %s, line %d: %s\n", | |
1338 | node->file, lineno, node->name); | |
1339 | fprintf (stderr, "Second entry ignored\n"); | |
1340 | } | |
c6d46f5f | 1341 | } |
c5007f46 | 1342 | else if (!cur_node->been_warned && !no_warnings) |
c6d46f5f | 1343 | { |
c5007f46 FP |
1344 | fprintf |
1345 | (stderr, | |
1346 | "Duplicate entry in files %s and %s: %s (Warning only)\n", | |
1347 | node->file, cur_node->file, node->name); | |
1348 | cur_node->been_warned = TRUE; | |
c6d46f5f | 1349 | } |
c6d46f5f JB |
1350 | return; |
1351 | } | |
1352 | ||
c6d46f5f JB |
1353 | /* Actually add the node */ |
1354 | add_node (node, dif < 0 ? &cur_node->left : &cur_node->right); | |
1355 | } | |
1356 | } | |
1357 | \f | |
1358 | void | |
1359 | put_entries (node) | |
13fde0cd | 1360 | register NODE *node; |
c6d46f5f | 1361 | { |
13fde0cd | 1362 | register char *sp; |
c6d46f5f JB |
1363 | |
1364 | if (node == NULL) | |
1365 | return; | |
1366 | ||
1367 | /* Output subentries that precede this one */ | |
1368 | put_entries (node->left); | |
1369 | ||
1370 | /* Output this entry */ | |
1371 | ||
32daa216 | 1372 | if (!CTAGS) |
c6d46f5f | 1373 | { |
108c932a FP |
1374 | if (node->name != NULL) |
1375 | fprintf (tagf, "%s\177%s\001%d,%d\n", | |
1376 | node->pat, node->name, node->lno, node->cno); | |
c6d46f5f | 1377 | else |
108c932a FP |
1378 | fprintf (tagf, "%s\177%d,%d\n", |
1379 | node->pat, node->lno, node->cno); | |
c6d46f5f | 1380 | } |
d8913c1c | 1381 | else |
c6d46f5f | 1382 | { |
d8913c1c FP |
1383 | if (node->name == NULL) |
1384 | error ("internal error: NULL name in ctags mode.", 0); | |
c6d46f5f | 1385 | |
d8913c1c FP |
1386 | if (cxref_style) |
1387 | { | |
1388 | if (vgrind_style) | |
1389 | fprintf (stdout, "%s %s %d\n", | |
1390 | node->name, node->file, (node->lno + 63) / 64); | |
1391 | else | |
1392 | fprintf (stdout, "%-16s %3d %-16s %s\n", | |
1393 | node->name, node->lno, node->file, node->pat); | |
c6d46f5f JB |
1394 | } |
1395 | else | |
d8913c1c FP |
1396 | { |
1397 | fprintf (tagf, "%s\t%s\t", node->name, node->file); | |
1398 | ||
1399 | if (node->is_func) | |
1400 | { /* a function */ | |
1401 | putc (searchar, tagf); | |
1402 | putc ('^', tagf); | |
1403 | ||
1404 | for (sp = node->pat; *sp; sp++) | |
1405 | { | |
1406 | if (*sp == '\\' || *sp == searchar) | |
1407 | putc ('\\', tagf); | |
1408 | putc (*sp, tagf); | |
1409 | } | |
1410 | putc (searchar, tagf); | |
1411 | } | |
1412 | else | |
1413 | { /* a typedef; text pattern inadequate */ | |
1414 | fprintf (tagf, "%d", node->lno); | |
1415 | } | |
1416 | putc ('\n', tagf); | |
c6d46f5f | 1417 | } |
c6d46f5f | 1418 | } |
c6d46f5f JB |
1419 | |
1420 | /* Output subentries that follow this one */ | |
1421 | put_entries (node->right); | |
1422 | } | |
1423 | ||
1424 | /* Length of a number's decimal representation. */ | |
1425 | int | |
1426 | number_len (num) | |
1427 | long num; | |
1428 | { | |
1429 | int len = 0; | |
1430 | if (!num) | |
1431 | return 1; | |
1432 | for (; num; num /= 10) | |
1433 | ++len; | |
1434 | return len; | |
1435 | } | |
1436 | ||
1437 | /* | |
1438 | * Return total number of characters that put_entries will output for | |
32daa216 FP |
1439 | * the nodes in the subtree of the specified node. Works only if |
1440 | * we are not ctags, but called only in that case. This count | |
1441 | * is irrelevant with the new tags.el, but is still supplied for | |
1442 | * backward compatibility. | |
c6d46f5f JB |
1443 | */ |
1444 | int | |
1445 | total_size_of_entries (node) | |
13fde0cd | 1446 | register NODE *node; |
c6d46f5f | 1447 | { |
13fde0cd | 1448 | register int total; |
c6d46f5f JB |
1449 | |
1450 | if (node == NULL) | |
1451 | return 0; | |
1452 | ||
1453 | total = 0; | |
1454 | for (; node; node = node->right) | |
1455 | { | |
1456 | /* Count left subentries. */ | |
1457 | total += total_size_of_entries (node->left); | |
1458 | ||
1459 | /* Count this entry */ | |
1460 | total += strlen (node->pat) + 1; | |
1461 | total += number_len ((long) node->lno) + 1 + number_len (node->cno) + 1; | |
108c932a | 1462 | if (node->name != NULL) |
c6d46f5f JB |
1463 | total += 1 + strlen (node->name); /* \001name */ |
1464 | } | |
1465 | ||
1466 | return total; | |
1467 | } | |
1468 | \f | |
1469 | /* | |
1470 | * The C symbol tables. | |
1471 | */ | |
55597f90 FP |
1472 | enum sym_type |
1473 | { | |
d8913c1c FP |
1474 | st_none, st_C_objprot, st_C_objimpl, st_C_objend, st_C_gnumacro, |
1475 | st_C_struct, st_C_enum, st_C_define, st_C_typedef, st_C_typespec, | |
55597f90 | 1476 | }; |
c6d46f5f | 1477 | |
42680d3c FP |
1478 | /* Feed stuff between (but not including) %[ and %] lines to: |
1479 | gperf -c -k1,3 -o -p -r -t | |
1480 | %[ | |
1481 | struct C_stab_entry { char *name; int c_ext; enum sym_type type; } | |
1482 | %% | |
d8913c1c FP |
1483 | @interface, 0, st_C_objprot |
1484 | @protocol, 0, st_C_objprot | |
1485 | @implementation,0, st_C_objimpl | |
1486 | @end, 0, st_C_objend | |
42680d3c FP |
1487 | class, C_PLPL, st_C_struct |
1488 | domain, C_STAR, st_C_struct | |
1489 | union, 0, st_C_struct | |
1490 | struct, 0, st_C_struct | |
1491 | enum, 0, st_C_enum | |
1492 | typedef, 0, st_C_typedef | |
1493 | define, 0, st_C_define | |
1494 | long, 0, st_C_typespec | |
1495 | short, 0, st_C_typespec | |
1496 | int, 0, st_C_typespec | |
1497 | char, 0, st_C_typespec | |
1498 | float, 0, st_C_typespec | |
1499 | double, 0, st_C_typespec | |
1500 | signed, 0, st_C_typespec | |
1501 | unsigned, 0, st_C_typespec | |
1502 | auto, 0, st_C_typespec | |
1503 | void, 0, st_C_typespec | |
1504 | extern, 0, st_C_typespec | |
1505 | static, 0, st_C_typespec | |
1506 | const, 0, st_C_typespec | |
1507 | volatile, 0, st_C_typespec | |
d8913c1c FP |
1508 | # DEFUN used in emacs, the next three used in glibc (SYSCALL only for mach). |
1509 | DEFUN, 0, st_C_gnumacro | |
1510 | SYSCALL, 0, st_C_gnumacro | |
1511 | ENTRY, 0, st_C_gnumacro | |
1512 | PSEUDO, 0, st_C_gnumacro | |
1513 | # These are defined inside C functions, so currently they are not met. | |
1514 | # EXFUN used in glibc, DEFVAR_* in emacs. | |
1515 | #EXFUN, 0, st_C_gnumacro | |
1516 | #DEFVAR_, 0, st_C_gnumacro | |
42680d3c FP |
1517 | %] |
1518 | and replace lines between %< and %> with its output. */ | |
1519 | /*%<*/ | |
1520 | /* C code produced by gperf version 1.8.1 (K&R C version) */ | |
1521 | /* Command-line: gperf -c -k1,3 -o -p -r -t */ | |
1522 | ||
1523 | ||
1524 | struct C_stab_entry { char *name; int c_ext; enum sym_type type; }; | |
1525 | ||
1526 | #define MIN_WORD_LENGTH 3 | |
d8913c1c FP |
1527 | #define MAX_WORD_LENGTH 15 |
1528 | #define MIN_HASH_VALUE 7 | |
1529 | #define MAX_HASH_VALUE 63 | |
c6d46f5f | 1530 | /* |
d8913c1c FP |
1531 | 29 keywords |
1532 | 57 is the maximum key range | |
42680d3c FP |
1533 | */ |
1534 | ||
1535 | static int | |
1536 | hash (str, len) | |
1537 | register char *str; | |
1538 | register int len; | |
1539 | { | |
1540 | static unsigned char hash_table[] = | |
1541 | { | |
d8913c1c FP |
1542 | 63, 63, 63, 63, 63, 63, 63, 63, 63, 63, |
1543 | 63, 63, 63, 63, 63, 63, 63, 63, 63, 63, | |
1544 | 63, 63, 63, 63, 63, 63, 63, 63, 63, 63, | |
1545 | 63, 63, 63, 63, 63, 63, 63, 63, 63, 63, | |
1546 | 63, 63, 63, 63, 63, 63, 63, 63, 63, 63, | |
1547 | 63, 63, 63, 63, 63, 63, 63, 63, 63, 63, | |
1548 | 63, 63, 63, 63, 17, 63, 63, 63, 4, 14, | |
1549 | 4, 63, 63, 63, 63, 63, 63, 63, 63, 63, | |
1550 | 8, 63, 63, 0, 23, 63, 63, 63, 63, 63, | |
1551 | 63, 63, 63, 63, 63, 63, 63, 28, 63, 28, | |
1552 | 10, 31, 27, 18, 63, 6, 63, 63, 26, 1, | |
1553 | 11, 2, 29, 63, 29, 16, 26, 13, 15, 63, | |
1554 | 63, 63, 63, 63, 63, 63, 63, 63, | |
42680d3c FP |
1555 | }; |
1556 | return len + hash_table[str[2]] + hash_table[str[0]]; | |
1557 | } | |
c6d46f5f | 1558 | |
42680d3c FP |
1559 | struct C_stab_entry * |
1560 | in_word_set (str, len) | |
1561 | register char *str; | |
1562 | register int len; | |
1563 | { | |
1564 | ||
1565 | static struct C_stab_entry wordlist[] = | |
1566 | { | |
d8913c1c FP |
1567 | {"",}, {"",}, {"",}, {"",}, {"",}, {"",}, {"",}, |
1568 | {"SYSCALL", 0, st_C_gnumacro}, | |
1569 | {"",}, {"",}, {"",}, {"",}, {"",}, | |
1570 | {"DEFUN", 0, st_C_gnumacro}, | |
c5007f46 | 1571 | {"",}, {"",}, {"",}, |
42680d3c | 1572 | {"domain", C_STAR, st_C_struct}, |
d8913c1c FP |
1573 | {"",}, {"",}, {"",}, {"",}, {"",}, |
1574 | {"short", 0, st_C_typespec}, | |
1575 | {"union", 0, st_C_struct}, | |
1576 | {"void", 0, st_C_typespec}, | |
1577 | {"",}, {"",}, | |
1578 | {"PSEUDO", 0, st_C_gnumacro}, | |
42680d3c | 1579 | {"double", 0, st_C_typespec}, |
d8913c1c FP |
1580 | {"",}, {"",}, |
1581 | {"@end", 0, st_C_objend}, | |
1582 | {"@implementation", 0, st_C_objimpl}, | |
1583 | {"float", 0, st_C_typespec}, | |
42680d3c | 1584 | {"int", 0, st_C_typespec}, |
c5007f46 | 1585 | {"",}, |
d8913c1c FP |
1586 | {"unsigned", 0, st_C_typespec}, |
1587 | {"@interface", 0, st_C_objprot}, | |
1588 | {"",}, | |
1589 | {"signed", 0, st_C_typespec}, | |
1590 | {"long", 0, st_C_typespec}, | |
1591 | {"ENTRY", 0, st_C_gnumacro}, | |
1592 | {"define", 0, st_C_define}, | |
1593 | {"const", 0, st_C_typespec}, | |
c5007f46 | 1594 | {"",}, {"",}, {"",}, |
d8913c1c FP |
1595 | {"enum", 0, st_C_enum}, |
1596 | {"volatile", 0, st_C_typespec}, | |
1597 | {"static", 0, st_C_typespec}, | |
42680d3c | 1598 | {"struct", 0, st_C_struct}, |
d8913c1c FP |
1599 | {"",}, {"",}, {"",}, |
1600 | {"@protocol", 0, st_C_objprot}, | |
c5007f46 | 1601 | {"",}, {"",}, |
d8913c1c FP |
1602 | {"auto", 0, st_C_typespec}, |
1603 | {"",}, | |
1604 | {"char", 0, st_C_typespec}, | |
1605 | {"class", C_PLPL, st_C_struct}, | |
1606 | {"typedef", 0, st_C_typedef}, | |
1607 | {"extern", 0, st_C_typespec}, | |
42680d3c FP |
1608 | }; |
1609 | ||
1610 | if (len <= MAX_WORD_LENGTH && len >= MIN_WORD_LENGTH) | |
1611 | { | |
1612 | register int key = hash (str, len); | |
1613 | ||
1614 | if (key <= MAX_HASH_VALUE && key >= MIN_HASH_VALUE) | |
1615 | { | |
1616 | register char *s = wordlist[key].name; | |
1617 | ||
d8913c1c | 1618 | if (*s == *str && !strncmp (str + 1, s + 1, len - 1)) |
42680d3c FP |
1619 | return &wordlist[key]; |
1620 | } | |
1621 | } | |
1622 | return 0; | |
c6d46f5f | 1623 | } |
42680d3c | 1624 | /*%>*/ |
c6d46f5f | 1625 | |
42680d3c FP |
1626 | enum sym_type |
1627 | C_symtype(str, len, c_ext) | |
1628 | char *str; | |
1629 | int len; | |
c6d46f5f JB |
1630 | int c_ext; |
1631 | { | |
42680d3c | 1632 | register struct C_stab_entry *se = in_word_set(str, len); |
c6d46f5f | 1633 | |
42680d3c FP |
1634 | if (se == NULL || (se->c_ext && !(c_ext & se->c_ext))) |
1635 | return st_none; | |
1636 | return se->type; | |
c6d46f5f JB |
1637 | } |
1638 | \f | |
13fde0cd | 1639 | /* |
13fde0cd RS |
1640 | * C functions are recognized using a simple finite automaton. |
1641 | * funcdef is its state variable. | |
1642 | */ | |
d8913c1c | 1643 | enum |
13fde0cd | 1644 | { |
31d4b314 FP |
1645 | fnone, /* nothing seen */ |
1646 | ftagseen, /* function-like tag seen */ | |
b12756c8 | 1647 | fstartlist, /* just after open parenthesis */ |
31d4b314 FP |
1648 | finlist, /* in parameter list */ |
1649 | flistseen, /* after parameter list */ | |
46e4cb76 | 1650 | fignore /* before open brace */ |
d8913c1c | 1651 | } funcdef; |
13fde0cd RS |
1652 | |
1653 | ||
46c145db FP |
1654 | /* |
1655 | * typedefs are recognized using a simple finite automaton. | |
15294654 | 1656 | * typdef is its state variable. |
13fde0cd | 1657 | */ |
d8913c1c | 1658 | enum |
13fde0cd | 1659 | { |
31d4b314 FP |
1660 | tnone, /* nothing seen */ |
1661 | ttypedseen, /* typedef keyword seen */ | |
1662 | tinbody, /* inside typedef body */ | |
46c145db FP |
1663 | tend, /* just before typedef tag */ |
1664 | tignore /* junk after typedef tag */ | |
d8913c1c | 1665 | } typdef; |
13fde0cd RS |
1666 | |
1667 | ||
c5007f46 | 1668 | /* |
46c145db FP |
1669 | * struct-like structures (enum, struct and union) are recognized |
1670 | * using another simple finite automaton. `structdef' is its state | |
1671 | * variable. | |
13fde0cd | 1672 | */ |
d8913c1c | 1673 | enum |
13fde0cd RS |
1674 | { |
1675 | snone, /* nothing seen yet */ | |
1676 | skeyseen, /* struct-like keyword seen */ | |
1677 | stagseen, /* struct-like tag seen */ | |
1678 | scolonseen, /* colon seen after struct-like tag */ | |
46e4cb76 | 1679 | sinbody /* in struct body: recognize member func defs*/ |
d8913c1c | 1680 | } structdef; |
46c145db | 1681 | |
13fde0cd RS |
1682 | /* |
1683 | * When structdef is stagseen, scolonseen, or sinbody, structtag is the | |
c5007f46 | 1684 | * struct tag, and structtype is the type of the preceding struct-like |
42680d3c | 1685 | * keyword. |
13fde0cd | 1686 | */ |
55597f90 | 1687 | char *structtag = "<uninited>"; |
42680d3c | 1688 | enum sym_type structtype; |
13fde0cd | 1689 | |
d8913c1c FP |
1690 | /* |
1691 | * When objdef is different from onone, objtag is the name of the class. | |
1692 | */ | |
1693 | char *objtag = "<uninited>"; | |
1694 | ||
13fde0cd RS |
1695 | /* |
1696 | * Yet another little state machine to deal with preprocessor lines. | |
1697 | */ | |
d8913c1c | 1698 | enum |
13fde0cd RS |
1699 | { |
1700 | dnone, /* nothing seen */ | |
1701 | dsharpseen, /* '#' seen as first char on line */ | |
1702 | ddefineseen, /* '#' and 'define' seen */ | |
46e4cb76 | 1703 | dignorerest /* ignore rest of line */ |
d8913c1c FP |
1704 | } definedef; |
1705 | ||
1706 | /* | |
1707 | * State machine for Objective C protocols and implementations. | |
1708 | */ | |
1709 | enum | |
1710 | { | |
1711 | onone, /* nothing seen */ | |
1712 | oprotocol, /* @interface or @protocol seen */ | |
1713 | oimplementation, /* @implementations seen */ | |
1714 | otagseen, /* class name seen */ | |
1715 | oparenseen, /* parenthesis before category seen */ | |
1716 | ocatseen, /* category name seen */ | |
1717 | oinbody, /* in @implementation body */ | |
1718 | omethodsign, /* in @implementation body, after +/- */ | |
1719 | omethodtag, /* after method name */ | |
1720 | omethodcolon, /* after method colon */ | |
1721 | omethodparm, /* after method parameter */ | |
1722 | oignore, /* wait for @end */ | |
1723 | } objdef; | |
13fde0cd RS |
1724 | |
1725 | /* | |
1726 | * Set this to TRUE, and the next token considered is called a function. | |
cdc1f6a7 | 1727 | * Used only for GNU emacs's function-defining macros. |
13fde0cd RS |
1728 | */ |
1729 | logical next_token_is_func; | |
1730 | ||
1731 | /* | |
1732 | * TRUE in the rules part of a yacc file, FALSE outside (parse as C). | |
1733 | */ | |
1734 | logical yacc_rules; | |
1735 | ||
d8913c1c FP |
1736 | /* |
1737 | * methodlen is the length of the method name stored in token_name. | |
1738 | */ | |
1739 | int methodlen; | |
1740 | ||
6dd5561c FP |
1741 | /* |
1742 | * consider_token () | |
1743 | * checks to see if the current token is at the start of a | |
1744 | * function, or corresponds to a typedef, or is a struct/union/enum | |
1745 | * tag. | |
1746 | * | |
1747 | * *IS_FUNC gets TRUE iff the token is a function or macro with args. | |
1748 | * C_EXT is which language we are looking at. | |
1749 | * | |
1750 | * In the future we will need some way to adjust where the end of | |
1751 | * the token is; for instance, implementing the C++ keyword | |
1752 | * `operator' properly will adjust the end of the token to be after | |
1753 | * whatever follows `operator'. | |
1754 | * | |
1755 | * Globals | |
1756 | * funcdef IN OUT | |
1757 | * structdef IN OUT | |
1758 | * definedef IN OUT | |
1759 | * typdef IN OUT | |
d8913c1c | 1760 | * objdef IN OUT |
6dd5561c FP |
1761 | * next_token_is_func IN OUT |
1762 | */ | |
1763 | ||
1764 | logical | |
d8913c1c | 1765 | consider_token (str, len, c, c_ext, cblev, parlev, is_func) |
55597f90 FP |
1766 | register char *str; /* IN: token pointer */ |
1767 | register int len; /* IN: token length */ | |
6dd5561c | 1768 | register char c; /* IN: first char after the token */ |
6dd5561c FP |
1769 | int c_ext; /* IN: C extensions mask */ |
1770 | int cblev; /* IN: curly brace level */ | |
d8913c1c | 1771 | int parlev; /* IN: parenthesis level */ |
715b6f8c | 1772 | logical *is_func; /* OUT: function found */ |
6dd5561c | 1773 | { |
55597f90 | 1774 | enum sym_type toktype = C_symtype (str, len, c_ext); |
6dd5561c FP |
1775 | |
1776 | /* | |
1777 | * Advance the definedef state machine. | |
1778 | */ | |
1779 | switch (definedef) | |
1780 | { | |
1781 | case dnone: | |
1782 | /* We're not on a preprocessor line. */ | |
1783 | break; | |
1784 | case dsharpseen: | |
1785 | if (toktype == st_C_define) | |
1786 | { | |
1787 | definedef = ddefineseen; | |
1788 | } | |
1789 | else | |
1790 | { | |
1791 | definedef = dignorerest; | |
1792 | } | |
b9755a12 | 1793 | return FALSE; |
6dd5561c FP |
1794 | case ddefineseen: |
1795 | /* | |
ee70dba5 FP |
1796 | * Make a tag for any macro, unless it is a constant |
1797 | * and constantypedefs is FALSE. | |
6dd5561c FP |
1798 | */ |
1799 | definedef = dignorerest; | |
1800 | *is_func = (c == '('); | |
1801 | if (!*is_func && !constantypedefs) | |
b9755a12 | 1802 | return FALSE; |
6dd5561c | 1803 | else |
b9755a12 | 1804 | return TRUE; |
6dd5561c | 1805 | case dignorerest: |
b9755a12 | 1806 | return FALSE; |
6dd5561c FP |
1807 | default: |
1808 | error ("internal error: definedef value.", 0); | |
1809 | } | |
1810 | ||
1811 | /* | |
1812 | * Now typedefs | |
1813 | */ | |
1814 | switch (typdef) | |
1815 | { | |
1816 | case tnone: | |
1817 | if (toktype == st_C_typedef) | |
1818 | { | |
1819 | if (typedefs) | |
1820 | typdef = ttypedseen; | |
1821 | funcdef = fnone; | |
b9755a12 | 1822 | return FALSE; |
6dd5561c FP |
1823 | } |
1824 | break; | |
1825 | case ttypedseen: | |
1826 | switch (toktype) | |
1827 | { | |
1828 | case st_none: | |
1829 | case st_C_typespec: | |
1830 | typdef = tend; | |
1831 | break; | |
1832 | case st_C_struct: | |
1833 | case st_C_enum: | |
1834 | break; | |
1835 | } | |
1836 | /* Do not return here, so the structdef stuff has a chance. */ | |
1837 | break; | |
1838 | case tend: | |
1839 | switch (toktype) | |
1840 | { | |
1841 | case st_C_typespec: | |
1842 | case st_C_struct: | |
1843 | case st_C_enum: | |
b9755a12 | 1844 | return FALSE; |
6dd5561c | 1845 | } |
b9755a12 | 1846 | return TRUE; |
6dd5561c FP |
1847 | } |
1848 | ||
1849 | /* | |
1850 | * This structdef business is currently only invoked when cblev==0. | |
1851 | * It should be recursively invoked whatever the curly brace level, | |
1852 | * and a stack of states kept, to allow for definitions of structs | |
1853 | * within structs. | |
1854 | * | |
1855 | * This structdef business is NOT invoked when we are ctags and the | |
1856 | * file is plain C. This is because a struct tag may have the same | |
1857 | * name as another tag, and this loses with ctags. | |
1858 | * | |
c5007f46 | 1859 | * This if statement deals with the typdef state machine as |
6dd5561c | 1860 | * follows: if typdef==ttypedseen and token is struct/union/class/enum, |
c5007f46 | 1861 | * return FALSE. All the other code here is for the structdef |
6dd5561c FP |
1862 | * state machine. |
1863 | */ | |
1864 | switch (toktype) | |
1865 | { | |
1866 | case st_C_struct: | |
1867 | case st_C_enum: | |
1868 | if (typdef == ttypedseen | |
1869 | || (typedefs_and_cplusplus && cblev == 0 && structdef == snone)) | |
1870 | { | |
1871 | structdef = skeyseen; | |
1872 | structtype = toktype; | |
1873 | } | |
b9755a12 | 1874 | return FALSE; |
6dd5561c FP |
1875 | } |
1876 | if (structdef == skeyseen) | |
1877 | { | |
55597f90 FP |
1878 | /* Save the tag for struct/union/class, for functions that may be |
1879 | defined inside. */ | |
6dd5561c | 1880 | if (structtype == st_C_struct) |
55597f90 | 1881 | structtag = savenstr (str, len); |
6dd5561c | 1882 | else |
55597f90 | 1883 | structtag = "<enum>"; |
6dd5561c | 1884 | structdef = stagseen; |
b9755a12 | 1885 | return TRUE; |
6dd5561c FP |
1886 | } |
1887 | ||
1888 | /* Avoid entering funcdef stuff if typdef is going on. */ | |
1889 | if (typdef != tnone) | |
1890 | { | |
1891 | definedef = dnone; | |
b9755a12 | 1892 | return FALSE; |
6dd5561c FP |
1893 | } |
1894 | ||
715b6f8c | 1895 | /* Detect GNU macros. */ |
d8913c1c FP |
1896 | if (definedef == dnone && toktype == st_C_gnumacro) |
1897 | { | |
1898 | next_token_is_func = TRUE; | |
1899 | return FALSE; | |
1900 | } | |
6dd5561c FP |
1901 | if (next_token_is_func) |
1902 | { | |
1903 | next_token_is_func = FALSE; | |
715b6f8c FP |
1904 | funcdef = fignore; |
1905 | *is_func = TRUE; | |
b9755a12 | 1906 | return TRUE; |
6dd5561c FP |
1907 | } |
1908 | ||
d8913c1c FP |
1909 | /* |
1910 | * Detecting Objective C constructs. | |
1911 | */ | |
1912 | switch (objdef) | |
1913 | { | |
1914 | case onone: | |
1915 | switch (toktype) | |
1916 | { | |
1917 | case st_C_objprot: | |
1918 | objdef = oprotocol; | |
1919 | return FALSE; | |
1920 | case st_C_objimpl: | |
1921 | objdef = oimplementation; | |
1922 | return FALSE; | |
1923 | } | |
1924 | break; | |
1925 | case oimplementation: | |
1926 | /* Save the class tag for functions that may be defined inside. */ | |
1927 | objtag = savenstr (str, len); | |
1928 | objdef = oinbody; | |
1929 | return FALSE; | |
1930 | case oprotocol: | |
1931 | /* Save the class tag for categories. */ | |
1932 | objtag = savenstr (str, len); | |
1933 | objdef = otagseen; | |
1934 | *is_func = TRUE; | |
1935 | return TRUE; | |
1936 | case oparenseen: | |
1937 | objdef = ocatseen; | |
1938 | *is_func = TRUE; | |
1939 | return TRUE; | |
1940 | case oinbody: | |
1941 | break; | |
1942 | case omethodsign: | |
1943 | if (parlev == 0) | |
1944 | { | |
1945 | objdef = omethodtag; | |
1946 | methodlen = len; | |
1947 | GROW_LINEBUFFER (token_name, methodlen+1); | |
1948 | strncpy (token_name.buffer, str, len); | |
1949 | token_name.buffer[methodlen] = '\0'; | |
1950 | return TRUE; | |
1951 | } | |
1952 | return FALSE; | |
1953 | case omethodcolon: | |
1954 | if (parlev == 0) | |
1955 | objdef = omethodparm; | |
1956 | return FALSE; | |
1957 | case omethodparm: | |
1958 | if (parlev == 0) | |
1959 | { | |
1960 | objdef = omethodtag; | |
1961 | methodlen =+ len; | |
1962 | GROW_LINEBUFFER (token_name, methodlen+1); | |
1963 | strncat (token_name.buffer, str, len); | |
1964 | return TRUE; | |
1965 | } | |
1966 | return FALSE; | |
1967 | case oignore: | |
1968 | if (toktype == st_C_objend) | |
1969 | { | |
1970 | /* Memory leakage here: the string pointed by objtag is | |
1971 | never released, because many tests would be needed to | |
1972 | avoid breaking on incorrect input code. The amount of | |
15294654 | 1973 | memory leaked here is the sum of the lengths of the |
d8913c1c FP |
1974 | class tags. |
1975 | free (objtag); */ | |
1976 | objdef = onone; | |
1977 | } | |
1978 | return FALSE; | |
1979 | } | |
1980 | ||
6dd5561c FP |
1981 | /* A function? */ |
1982 | switch (toktype) | |
1983 | { | |
1984 | case st_C_typespec: | |
1985 | if (funcdef != finlist && funcdef != fignore) | |
1986 | funcdef = fnone; /* should be useless */ | |
b9755a12 | 1987 | return FALSE; |
6dd5561c FP |
1988 | default: |
1989 | if (funcdef == fnone) | |
1990 | { | |
1991 | funcdef = ftagseen; | |
1992 | *is_func = TRUE; | |
b9755a12 | 1993 | return TRUE; |
6dd5561c FP |
1994 | } |
1995 | } | |
1996 | ||
b9755a12 | 1997 | return FALSE; |
6dd5561c FP |
1998 | } |
1999 | ||
c6d46f5f JB |
2000 | /* |
2001 | * C_entries () | |
13fde0cd RS |
2002 | * This routine finds functions, typedefs, #define's and |
2003 | * struct/union/enum definitions in C syntax and adds them | |
c6d46f5f JB |
2004 | * to the list. |
2005 | */ | |
55597f90 FP |
2006 | typedef struct |
2007 | { | |
75bdbc6a | 2008 | logical valid; |
55597f90 FP |
2009 | char *str; |
2010 | logical named; | |
2011 | int linelen; | |
2012 | int lineno; | |
2bd88040 FP |
2013 | long linepos; |
2014 | char *buffer; | |
55597f90 FP |
2015 | } TOKEN; |
2016 | ||
2017 | #define current_lb_is_new (newndx == curndx) | |
2018 | #define switch_line_buffers() (curndx = 1 - curndx) | |
c6d46f5f | 2019 | |
13fde0cd RS |
2020 | #define curlb (lbs[curndx].lb) |
2021 | #define othlb (lbs[1-curndx].lb) | |
2022 | #define newlb (lbs[newndx].lb) | |
2023 | #define curlinepos (lbs[curndx].linepos) | |
2024 | #define othlinepos (lbs[1-curndx].linepos) | |
2025 | #define newlinepos (lbs[newndx].linepos) | |
2026 | ||
c6d46f5f | 2027 | #define CNL_SAVE_DEFINEDEF \ |
13fde0cd | 2028 | do { \ |
55597f90 | 2029 | curlinepos = charno; \ |
c6d46f5f | 2030 | lineno++; \ |
13fde0cd RS |
2031 | charno += readline (&curlb, inf); \ |
2032 | lp = curlb.buffer; \ | |
2033 | quotednl = FALSE; \ | |
2034 | newndx = curndx; \ | |
b9755a12 | 2035 | } while (0) |
c6d46f5f JB |
2036 | |
2037 | #define CNL \ | |
13fde0cd | 2038 | do { \ |
c6d46f5f | 2039 | CNL_SAVE_DEFINEDEF; \ |
75bdbc6a | 2040 | if (savetok.valid) \ |
55597f90 FP |
2041 | { \ |
2042 | tok = savetok; \ | |
75bdbc6a | 2043 | savetok.valid = FALSE; \ |
55597f90 | 2044 | } \ |
c6d46f5f | 2045 | definedef = dnone; \ |
b9755a12 | 2046 | } while (0) |
13fde0cd | 2047 | |
d8913c1c FP |
2048 | /* Ideally this macro should never be called wihen tok.valid is FALSE, |
2049 | but this would mean that the state machines always guess right. */ | |
75bdbc6a | 2050 | #define make_tag(isfun) do \ |
d8913c1c FP |
2051 | if (tok.valid) { \ |
2052 | char *name = NULL; \ | |
2053 | if (CTAGS || tok.named) \ | |
2054 | name = savestr (token_name.buffer); \ | |
2055 | pfnote (name, isfun, tok.buffer, tok.linelen, tok.lineno, tok.linepos); \ | |
75bdbc6a FP |
2056 | tok.valid = FALSE; \ |
2057 | } while (0) | |
c6d46f5f JB |
2058 | |
2059 | void | |
6dd5561c | 2060 | C_entries (c_ext, inf) |
b9755a12 FP |
2061 | int c_ext; /* extension of C */ |
2062 | FILE *inf; /* input file */ | |
c6d46f5f | 2063 | { |
13fde0cd | 2064 | register char c; /* latest char read; '\0' for end of line */ |
c6d46f5f | 2065 | register char *lp; /* pointer one beyond the character `c' */ |
13fde0cd | 2066 | int curndx, newndx; /* indices for current and new lb */ |
55597f90 FP |
2067 | TOKEN tok; /* latest token read */ |
2068 | register int tokoff; /* offset in line of start of current token */ | |
2069 | register int toklen; /* length of current token */ | |
591fa824 | 2070 | int cblev; /* current curly brace level */ |
b12756c8 | 2071 | int parlev; /* current parenthesis level */ |
13fde0cd RS |
2072 | logical incomm, inquote, inchar, quotednl, midtoken; |
2073 | logical cplpl; | |
55597f90 | 2074 | TOKEN savetok; /* token saved during preprocessor handling */ |
c6d46f5f | 2075 | |
75bdbc6a | 2076 | |
13fde0cd | 2077 | curndx = newndx = 0; |
c6d46f5f JB |
2078 | lineno = 0; |
2079 | charno = 0; | |
13fde0cd | 2080 | lp = curlb.buffer; |
c6d46f5f JB |
2081 | *lp = 0; |
2082 | ||
d8913c1c FP |
2083 | funcdef = fnone; typdef = tnone; structdef = snone; |
2084 | definedef = dnone; objdef = onone; | |
75bdbc6a | 2085 | next_token_is_func = yacc_rules = FALSE; |
13fde0cd | 2086 | midtoken = inquote = inchar = incomm = quotednl = FALSE; |
75bdbc6a | 2087 | tok.valid = savetok.valid = FALSE; |
591fa824 | 2088 | cblev = 0; |
b12756c8 | 2089 | parlev = 0; |
13fde0cd | 2090 | cplpl = c_ext & C_PLPL; |
c6d46f5f | 2091 | |
c6d46f5f JB |
2092 | while (!feof (inf)) |
2093 | { | |
2094 | c = *lp++; | |
c6d46f5f JB |
2095 | if (c == '\\') |
2096 | { | |
4746118a JB |
2097 | /* If we're at the end of the line, the next character is a |
2098 | '\0'; don't skip it, because it's the thing that tells us | |
2099 | to read the next line. */ | |
13fde0cd | 2100 | if (*lp == '\0') |
99e0a2e0 | 2101 | { |
13fde0cd | 2102 | quotednl = TRUE; |
99e0a2e0 RS |
2103 | continue; |
2104 | } | |
1e134a5f | 2105 | lp++; |
c6d46f5f JB |
2106 | c = ' '; |
2107 | } | |
2108 | else if (incomm) | |
2109 | { | |
13fde0cd | 2110 | switch (c) |
c6d46f5f | 2111 | { |
13fde0cd RS |
2112 | case '*': |
2113 | if (*lp == '/') | |
2114 | { | |
2115 | c = *lp++; | |
2116 | incomm = FALSE; | |
2117 | } | |
2118 | break; | |
2119 | case '\0': | |
2120 | /* Newlines inside comments do not end macro definitions in | |
2121 | traditional cpp. */ | |
2122 | CNL_SAVE_DEFINEDEF; | |
2123 | break; | |
c6d46f5f | 2124 | } |
13fde0cd | 2125 | continue; |
c6d46f5f JB |
2126 | } |
2127 | else if (inquote) | |
2128 | { | |
13fde0cd RS |
2129 | switch (c) |
2130 | { | |
2131 | case '"': | |
2132 | inquote = FALSE; | |
2133 | break; | |
2134 | case '\0': | |
42680d3c | 2135 | /* Newlines inside strings do not end macro definitions |
13fde0cd RS |
2136 | in traditional cpp, even though compilers don't |
2137 | usually accept them. */ | |
2138 | CNL_SAVE_DEFINEDEF; | |
2139 | break; | |
2140 | } | |
2141 | continue; | |
c6d46f5f JB |
2142 | } |
2143 | else if (inchar) | |
2144 | { | |
42680d3c FP |
2145 | switch (c) |
2146 | { | |
2147 | case '\0': | |
2148 | /* Hmmm, something went wrong. */ | |
2149 | CNL; | |
2150 | /* FALLTHRU */ | |
2151 | case '\'': | |
46c145db | 2152 | inchar = FALSE; |
42680d3c FP |
2153 | break; |
2154 | } | |
c6d46f5f JB |
2155 | continue; |
2156 | } | |
c5007f46 | 2157 | else |
c6d46f5f JB |
2158 | switch (c) |
2159 | { | |
2160 | case '"': | |
2161 | inquote = TRUE; | |
b12756c8 FP |
2162 | if (funcdef != finlist && funcdef != fignore) |
2163 | funcdef = fnone; | |
c6d46f5f JB |
2164 | continue; |
2165 | case '\'': | |
2166 | inchar = TRUE; | |
b12756c8 FP |
2167 | if (funcdef != finlist && funcdef != fignore) |
2168 | funcdef = fnone; | |
c6d46f5f JB |
2169 | continue; |
2170 | case '/': | |
2171 | if (*lp == '*') | |
2172 | { | |
2173 | lp++; | |
2174 | incomm = TRUE; | |
13fde0cd | 2175 | continue; |
c6d46f5f | 2176 | } |
d8913c1c | 2177 | else if (/* cplpl && */ *lp == '/') |
c6d46f5f | 2178 | { |
d8913c1c | 2179 | c = '\0'; |
daa37602 | 2180 | break; |
c6d46f5f | 2181 | } |
b12756c8 FP |
2182 | else |
2183 | break; | |
13fde0cd RS |
2184 | case '%': |
2185 | if ((c_ext & YACC) && *lp == '%') | |
2186 | { | |
2187 | /* entering or exiting rules section in yacc file */ | |
2188 | lp++; | |
2189 | definedef = dnone; funcdef = fnone; | |
46c145db | 2190 | typdef = tnone; structdef = snone; |
13fde0cd RS |
2191 | next_token_is_func = FALSE; |
2192 | midtoken = inquote = inchar = incomm = quotednl = FALSE; | |
591fa824 | 2193 | cblev = 0; |
13fde0cd RS |
2194 | yacc_rules = !yacc_rules; |
2195 | continue; | |
591fa824 | 2196 | } |
b12756c8 FP |
2197 | else |
2198 | break; | |
c6d46f5f | 2199 | case '#': |
ee70dba5 FP |
2200 | if (definedef == dnone) |
2201 | { | |
2202 | char *cp; | |
2203 | logical cpptoken = TRUE; | |
2204 | ||
2205 | /* Look back on this line. If all blanks, or nonblanks | |
2206 | followed by an end of comment, this is a preprocessor | |
2207 | token. */ | |
2208 | for (cp = newlb.buffer; cp < lp-1; cp++) | |
2209 | if (!iswhite (*cp)) | |
2210 | { | |
2211 | if (*cp == '*' && *(cp+1) == '/') | |
2212 | { | |
2213 | cp++; | |
2214 | cpptoken = TRUE; | |
2215 | } | |
2216 | else | |
2217 | cpptoken = FALSE; | |
2218 | } | |
2219 | if (cpptoken) | |
2220 | definedef = dsharpseen; | |
2221 | } /* if (definedef == dnone) */ | |
2222 | ||
c6d46f5f | 2223 | continue; |
13fde0cd | 2224 | } /* switch (c) */ |
c6d46f5f | 2225 | |
c6d46f5f | 2226 | |
591fa824 | 2227 | /* Consider token only if some complicated conditions are satisfied. */ |
ee70dba5 FP |
2228 | if ((definedef != dnone |
2229 | || (cblev == 0 && structdef != scolonseen) | |
591fa824 | 2230 | || (cblev == 1 && cplpl && structdef == sinbody)) |
46c145db | 2231 | && typdef != tignore |
13fde0cd | 2232 | && definedef != dignorerest |
ee70dba5 | 2233 | && funcdef != finlist) |
c6d46f5f JB |
2234 | { |
2235 | if (midtoken) | |
2236 | { | |
2237 | if (endtoken (c)) | |
2238 | { | |
d8913c1c | 2239 | if (c == ':' && cplpl && *lp == ':' && begtoken(*(lp + 1))) |
c6d46f5f JB |
2240 | { |
2241 | /* | |
ee70dba5 FP |
2242 | * This handles :: in the middle, but not at the |
2243 | * beginning of an identifier. | |
c6d46f5f JB |
2244 | */ |
2245 | lp += 2; | |
2246 | toklen += 3; | |
2247 | } | |
2248 | else | |
2249 | { | |
fe0b3356 | 2250 | logical is_func = FALSE; |
c6d46f5f | 2251 | |
13fde0cd | 2252 | if (yacc_rules |
d8913c1c FP |
2253 | || consider_token (newlb.buffer + tokoff, toklen, c, |
2254 | c_ext, cblev, parlev, &is_func)) | |
c6d46f5f | 2255 | { |
99e0a2e0 | 2256 | if (structdef == sinbody |
fe0b3356 FP |
2257 | && definedef == dnone |
2258 | && is_func) | |
2259 | /* function defined in C++ class body */ | |
2260 | { | |
d8913c1c FP |
2261 | GROW_LINEBUFFER (token_name, |
2262 | strlen(structtag)+2+toklen+1); | |
75bdbc6a FP |
2263 | strcpy (token_name.buffer, structtag); |
2264 | strcat (token_name.buffer, "::"); | |
2265 | strncat (token_name.buffer, | |
2bd88040 | 2266 | newlb.buffer+tokoff, toklen); |
ee70dba5 | 2267 | tok.named = TRUE; |
c6d46f5f | 2268 | } |
d8913c1c FP |
2269 | else if (objdef == ocatseen) |
2270 | /* Objective C category */ | |
2271 | { | |
2272 | GROW_LINEBUFFER (token_name, | |
2273 | strlen(objtag)+2+toklen+1); | |
2274 | strcpy (token_name.buffer, objtag); | |
2275 | strcat (token_name.buffer, "("); | |
2276 | strncat (token_name.buffer, | |
2277 | newlb.buffer+tokoff, toklen); | |
2278 | strcat (token_name.buffer, ")"); | |
2279 | tok.named = TRUE; | |
2280 | } | |
2281 | else if (objdef == omethodtag | |
2282 | || objdef == omethodparm) | |
2283 | /* Objective C method */ | |
2284 | { | |
2285 | tok.named = TRUE; | |
2286 | } | |
c6d46f5f JB |
2287 | else |
2288 | { | |
d8913c1c | 2289 | GROW_LINEBUFFER (token_name, toklen+1); |
75bdbc6a | 2290 | strncpy (token_name.buffer, |
2bd88040 | 2291 | newlb.buffer+tokoff, toklen); |
75bdbc6a | 2292 | token_name.buffer[toklen] = '\0'; |
55597f90 FP |
2293 | if (structdef == stagseen |
2294 | || typdef == tend | |
2295 | || (is_func | |
2296 | && definedef == dignorerest)) /* macro */ | |
2297 | tok.named = TRUE; | |
2298 | else | |
2299 | tok.named = FALSE; | |
c6d46f5f | 2300 | } |
55597f90 FP |
2301 | tok.lineno = lineno; |
2302 | tok.linelen = tokoff + toklen + 1; | |
2bd88040 FP |
2303 | tok.buffer = newlb.buffer; |
2304 | tok.linepos = newlinepos; | |
75bdbc6a | 2305 | tok.valid = TRUE; |
fe0b3356 | 2306 | |
b12756c8 FP |
2307 | if (definedef == dnone |
2308 | && (funcdef == ftagseen | |
2309 | || structdef == stagseen | |
d8913c1c FP |
2310 | || typdef == tend |
2311 | || objdef != onone)) | |
13fde0cd | 2312 | { |
55597f90 FP |
2313 | if (current_lb_is_new) |
2314 | switch_line_buffers (); | |
13fde0cd RS |
2315 | } |
2316 | else | |
2bd88040 | 2317 | make_tag (is_func); |
c6d46f5f JB |
2318 | } |
2319 | midtoken = FALSE; | |
2320 | } | |
13fde0cd | 2321 | } /* if (endtoken (c)) */ |
c6d46f5f | 2322 | else if (intoken (c)) |
13fde0cd RS |
2323 | { |
2324 | toklen++; | |
2325 | continue; | |
2326 | } | |
2327 | } /* if (midtoken) */ | |
c6d46f5f JB |
2328 | else if (begtoken (c)) |
2329 | { | |
b12756c8 | 2330 | switch (definedef) |
13fde0cd | 2331 | { |
b12756c8 FP |
2332 | case dnone: |
2333 | switch (funcdef) | |
2334 | { | |
2335 | case fstartlist: | |
2336 | funcdef = finlist; | |
2337 | continue; | |
2338 | case flistseen: | |
2bd88040 | 2339 | make_tag (TRUE); |
b12756c8 FP |
2340 | funcdef = fignore; |
2341 | break; | |
2342 | case ftagseen: | |
2343 | funcdef = fnone; | |
2344 | break; | |
2345 | } | |
2346 | if (structdef == stagseen) | |
2347 | structdef = snone; | |
13fde0cd | 2348 | break; |
b12756c8 | 2349 | case dsharpseen: |
4b533b5b | 2350 | savetok = tok; |
13fde0cd | 2351 | } |
13fde0cd RS |
2352 | if (!yacc_rules || lp == newlb.buffer + 1) |
2353 | { | |
2354 | tokoff = lp - 1 - newlb.buffer; | |
2355 | toklen = 1; | |
2356 | midtoken = TRUE; | |
2357 | } | |
2358 | continue; | |
4b533b5b | 2359 | } /* if (begtoken) */ |
13fde0cd RS |
2360 | } /* if must look at token */ |
2361 | ||
2362 | ||
2363 | /* Detect end of line, colon, comma, semicolon and various braces | |
b12756c8 | 2364 | after having handled a token.*/ |
13fde0cd | 2365 | switch (c) |
1e134a5f | 2366 | { |
13fde0cd | 2367 | case ':': |
b12756c8 FP |
2368 | if (definedef != dnone) |
2369 | break; | |
d8913c1c FP |
2370 | switch (objdef) |
2371 | { | |
2372 | case otagseen: | |
2373 | objdef = oignore; | |
2374 | make_tag (TRUE); | |
2375 | break; | |
2376 | case omethodtag: | |
2377 | case omethodparm: | |
2378 | objdef = omethodcolon; | |
2379 | methodlen += 1; | |
2380 | GROW_LINEBUFFER (token_name, methodlen+1); | |
2381 | strcat (token_name.buffer, ":"); | |
2382 | break; | |
2383 | } | |
13fde0cd RS |
2384 | if (structdef == stagseen) |
2385 | structdef = scolonseen; | |
b12756c8 FP |
2386 | else |
2387 | switch (funcdef) | |
2388 | { | |
2389 | case ftagseen: | |
2390 | if (yacc_rules) | |
2391 | { | |
2bd88040 | 2392 | make_tag (FALSE); |
b12756c8 FP |
2393 | funcdef = fignore; |
2394 | } | |
2395 | break; | |
2396 | case fstartlist: | |
2397 | funcdef = fnone; | |
2398 | break; | |
2399 | } | |
13fde0cd RS |
2400 | break; |
2401 | case ';': | |
b12756c8 FP |
2402 | if (definedef != dnone) |
2403 | break; | |
46c145db FP |
2404 | if (cblev == 0) |
2405 | switch (typdef) | |
2406 | { | |
2407 | case tend: | |
2bd88040 | 2408 | make_tag (FALSE); |
46c145db FP |
2409 | /* FALLTHRU */ |
2410 | default: | |
2411 | typdef = tnone; | |
2412 | } | |
31d4b314 | 2413 | if (funcdef != fignore) |
1f638249 FP |
2414 | { |
2415 | funcdef = fnone; | |
2416 | /* The following instruction invalidates the token. | |
2417 | Probably the token should be invalidated in all | |
2418 | other cases where some state machine is reset. */ | |
2419 | tok.valid = FALSE; | |
2420 | } | |
46c145db FP |
2421 | if (structdef == stagseen) |
2422 | structdef = snone; | |
2423 | break; | |
13fde0cd | 2424 | case ',': |
46c145db FP |
2425 | if (definedef != dnone) |
2426 | break; | |
d8913c1c FP |
2427 | switch (objdef) |
2428 | { | |
2429 | case omethodtag: | |
2430 | case omethodparm: | |
2431 | make_tag (TRUE); | |
2432 | objdef = oinbody; | |
2433 | break; | |
2434 | } | |
46c145db FP |
2435 | if (funcdef != finlist && funcdef != fignore) |
2436 | funcdef = fnone; | |
2437 | if (structdef == stagseen) | |
2438 | structdef = snone; | |
2439 | break; | |
13fde0cd | 2440 | case '[': |
b12756c8 FP |
2441 | if (definedef != dnone) |
2442 | break; | |
46c145db FP |
2443 | if (cblev == 0 && typdef == tend) |
2444 | { | |
2445 | typdef = tignore; | |
2bd88040 | 2446 | make_tag (FALSE); |
46c145db FP |
2447 | break; |
2448 | } | |
31d4b314 | 2449 | if (funcdef != finlist && funcdef != fignore) |
13fde0cd RS |
2450 | funcdef = fnone; |
2451 | if (structdef == stagseen) | |
2452 | structdef = snone; | |
2453 | break; | |
2454 | case '(': | |
b12756c8 FP |
2455 | if (definedef != dnone) |
2456 | break; | |
d8913c1c FP |
2457 | if (objdef == otagseen && parlev == 0) |
2458 | objdef = oparenseen; | |
13fde0cd | 2459 | switch (funcdef) |
57e83cfe | 2460 | { |
ee70dba5 FP |
2461 | case fnone: |
2462 | switch (typdef) | |
2463 | { | |
2464 | case ttypedseen: | |
2465 | case tend: | |
2466 | /* Make sure that the next char is not a '*'. | |
2467 | This handles constructs like: | |
2468 | typedef void OperatorFun (int fun); */ | |
2469 | if (*lp != '*') | |
2470 | { | |
2471 | typdef = tignore; | |
2bd88040 | 2472 | make_tag (FALSE); |
ee70dba5 FP |
2473 | } |
2474 | break; | |
2475 | } /* switch (typdef) */ | |
2476 | break; | |
13fde0cd | 2477 | case ftagseen: |
b12756c8 | 2478 | funcdef = fstartlist; |
13fde0cd | 2479 | break; |
13fde0cd | 2480 | case flistseen: |
b12756c8 | 2481 | funcdef = finlist; |
13fde0cd | 2482 | break; |
57e83cfe | 2483 | } |
b12756c8 | 2484 | parlev++; |
13fde0cd RS |
2485 | break; |
2486 | case ')': | |
b12756c8 FP |
2487 | if (definedef != dnone) |
2488 | break; | |
d8913c1c FP |
2489 | if (objdef == ocatseen && parlev == 1) |
2490 | { | |
2491 | make_tag (TRUE); | |
2492 | objdef = oignore; | |
2493 | } | |
b12756c8 FP |
2494 | if (--parlev == 0) |
2495 | { | |
2496 | switch (funcdef) | |
2497 | { | |
2498 | case fstartlist: | |
2499 | case finlist: | |
2500 | funcdef = flistseen; | |
2501 | break; | |
2502 | } | |
46c145db FP |
2503 | if (cblev == 0 && typdef == tend) |
2504 | { | |
2505 | typdef = tignore; | |
2bd88040 | 2506 | make_tag (FALSE); |
46c145db | 2507 | } |
b12756c8 FP |
2508 | } |
2509 | else if (parlev < 0) /* can happen due to ill-conceived #if's. */ | |
2510 | parlev = 0; | |
13fde0cd RS |
2511 | break; |
2512 | case '{': | |
b12756c8 FP |
2513 | if (definedef != dnone) |
2514 | break; | |
13fde0cd RS |
2515 | if (typdef == ttypedseen) |
2516 | typdef = tinbody; | |
2517 | switch (structdef) | |
2518 | { | |
2519 | case skeyseen: /* unnamed struct */ | |
55597f90 | 2520 | structtag = "_anonymous_"; |
13fde0cd RS |
2521 | structdef = sinbody; |
2522 | break; | |
2523 | case stagseen: | |
2524 | case scolonseen: /* named struct */ | |
2525 | structdef = sinbody; | |
2bd88040 | 2526 | make_tag (FALSE); |
13fde0cd RS |
2527 | break; |
2528 | } | |
31d4b314 FP |
2529 | switch (funcdef) |
2530 | { | |
2531 | case flistseen: | |
2bd88040 | 2532 | make_tag (TRUE); |
31d4b314 FP |
2533 | /* FALLTHRU */ |
2534 | case fignore: | |
2535 | funcdef = fnone; | |
46c145db FP |
2536 | break; |
2537 | case fnone: | |
d8913c1c FP |
2538 | switch (objdef) |
2539 | { | |
2540 | case otagseen: | |
2541 | make_tag (TRUE); | |
2542 | objdef = oignore; | |
2543 | break; | |
2544 | case omethodtag: | |
2545 | case omethodparm: | |
2546 | make_tag (TRUE); | |
2547 | objdef = oinbody; | |
2548 | break; | |
2549 | default: | |
2550 | /* Neutralize `extern "C" {' grot and look inside structs. */ | |
2551 | if (cblev == 0 && structdef == snone && typdef == tnone) | |
2552 | cblev = -1; | |
2553 | } | |
31d4b314 | 2554 | } |
591fa824 | 2555 | cblev++; |
31d4b314 | 2556 | break; |
13fde0cd | 2557 | case '*': |
b12756c8 FP |
2558 | if (definedef != dnone) |
2559 | break; | |
2560 | if (funcdef == fstartlist) | |
2561 | funcdef = fnone; /* avoid tagging `foo' in `foo (*bar()) ()' */ | |
13fde0cd RS |
2562 | break; |
2563 | case '}': | |
b12756c8 FP |
2564 | if (definedef != dnone) |
2565 | break; | |
13fde0cd | 2566 | if (!noindentypedefs && lp == newlb.buffer + 1) |
b12756c8 FP |
2567 | { |
2568 | cblev = 0; /* reset curly brace level if first column */ | |
2569 | parlev = 0; /* also reset paren level, just in case... */ | |
2570 | } | |
591fa824 RS |
2571 | else if (cblev > 0) |
2572 | cblev--; | |
2573 | if (cblev == 0) | |
13fde0cd RS |
2574 | { |
2575 | if (typdef == tinbody) | |
2576 | typdef = tend; | |
c5007f46 FP |
2577 | /* Memory leakage here: the string pointed by structtag is |
2578 | never released, because I fear to miss something and | |
2579 | break things while freeing the area. The amount of | |
15294654 | 2580 | memory leaked here is the sum of the lengths of the |
c5007f46 FP |
2581 | struct tags. |
2582 | if (structdef == sinbody) | |
2583 | free (structtag); */ | |
9cb0aa73 | 2584 | |
13fde0cd | 2585 | structdef = snone; |
55597f90 | 2586 | structtag = "<error>"; |
13fde0cd RS |
2587 | } |
2588 | break; | |
d8913c1c FP |
2589 | case '+': |
2590 | case '-': | |
2591 | if (objdef == oinbody && cblev == 0) | |
2592 | { | |
2593 | objdef = omethodsign; | |
2594 | break; | |
2595 | } | |
2596 | /* FALLTHRU */ | |
2597 | case '=': case '#': case '~': case '&': case '%': case '/': | |
42680d3c | 2598 | case '|': case '^': case '!': case '<': case '>': case '.': case '?': |
b12756c8 FP |
2599 | if (definedef != dnone) |
2600 | break; | |
2601 | /* These surely cannot follow a function tag. */ | |
2602 | if (funcdef != finlist && funcdef != fignore) | |
2603 | funcdef = fnone; | |
2604 | break; | |
13fde0cd | 2605 | case '\0': |
d8913c1c FP |
2606 | if (objdef == otagseen) |
2607 | { | |
2608 | make_tag (TRUE); | |
2609 | objdef = oignore; | |
2610 | } | |
13fde0cd RS |
2611 | /* If a macro spans multiple lines don't reset its state. */ |
2612 | if (quotednl) | |
2613 | CNL_SAVE_DEFINEDEF; | |
2614 | else | |
2615 | CNL; | |
2616 | break; | |
2617 | } /* switch (c) */ | |
2618 | ||
2619 | } /* while not eof */ | |
c6d46f5f | 2620 | } |
b9755a12 FP |
2621 | |
2622 | /* | |
2623 | * Process either a C++ file or a C file depending on the setting | |
2624 | * of a global flag. | |
2625 | */ | |
2626 | void | |
2627 | default_C_entries (inf) | |
2628 | FILE *inf; | |
2629 | { | |
2630 | C_entries (cplusplus ? C_PLPL : 0, inf); | |
2631 | } | |
2632 | ||
79263656 FP |
2633 | /* Always do plain ANSI C. */ |
2634 | void | |
2635 | plain_C_entries (inf) | |
2636 | FILE *inf; | |
2637 | { | |
2638 | C_entries (0, inf); | |
2639 | } | |
2640 | ||
b9755a12 FP |
2641 | /* Always do C++. */ |
2642 | void | |
2643 | Cplusplus_entries (inf) | |
2644 | FILE *inf; | |
2645 | { | |
2646 | C_entries (C_PLPL, inf); | |
2647 | } | |
2648 | ||
2649 | /* Always do C*. */ | |
2650 | void | |
2651 | Cstar_entries (inf) | |
2652 | FILE *inf; | |
2653 | { | |
2654 | C_entries (C_STAR, inf); | |
2655 | } | |
2656 | ||
2657 | /* Always do Yacc. */ | |
2658 | void | |
2659 | Yacc_entries (inf) | |
2660 | FILE *inf; | |
2661 | { | |
2662 | C_entries (YACC, inf); | |
2663 | } | |
6dd5561c FP |
2664 | \f |
2665 | /* Fortran parsing */ | |
c6d46f5f | 2666 | |
6dd5561c | 2667 | char *dbp; |
c6d46f5f JB |
2668 | |
2669 | logical | |
6dd5561c FP |
2670 | tail (cp) |
2671 | char *cp; | |
c6d46f5f | 2672 | { |
6dd5561c | 2673 | register int len = 0; |
c6d46f5f | 2674 | |
79263656 | 2675 | while (*cp && lowcase(*cp) == lowcase(dbp[len])) |
6dd5561c | 2676 | cp++, len++; |
108c932a | 2677 | if (*cp == '\0' && !intoken(dbp[len])) |
c6d46f5f | 2678 | { |
6dd5561c | 2679 | dbp += len; |
b9755a12 | 2680 | return TRUE; |
c6d46f5f | 2681 | } |
b9755a12 | 2682 | return FALSE; |
6dd5561c | 2683 | } |
13fde0cd | 2684 | |
6dd5561c FP |
2685 | void |
2686 | takeprec () | |
2687 | { | |
2688 | while (isspace (*dbp)) | |
2689 | dbp++; | |
2690 | if (*dbp != '*') | |
2691 | return; | |
2692 | dbp++; | |
2693 | while (isspace (*dbp)) | |
2694 | dbp++; | |
79263656 FP |
2695 | if (strneq (dbp, "(*)", 3)) |
2696 | { | |
2697 | dbp += 3; | |
2698 | return; | |
2699 | } | |
6dd5561c | 2700 | if (!isdigit (*dbp)) |
c6d46f5f | 2701 | { |
6dd5561c FP |
2702 | --dbp; /* force failure */ |
2703 | return; | |
c6d46f5f | 2704 | } |
6dd5561c FP |
2705 | do |
2706 | dbp++; | |
2707 | while (isdigit (*dbp)); | |
2708 | } | |
13fde0cd | 2709 | |
6dd5561c FP |
2710 | void |
2711 | getit (inf) | |
2712 | FILE *inf; | |
2713 | { | |
2714 | register char *cp; | |
13fde0cd | 2715 | |
6dd5561c FP |
2716 | while (isspace (*dbp)) |
2717 | dbp++; | |
2718 | if (*dbp == '\0') | |
c6d46f5f | 2719 | { |
6dd5561c FP |
2720 | lineno++; |
2721 | linecharno = charno; | |
2722 | charno += readline (&lb, inf); | |
2723 | dbp = lb.buffer; | |
2724 | if (dbp[5] != '&') | |
2725 | return; | |
2726 | dbp += 6; | |
2727 | while (isspace (*dbp)) | |
2728 | dbp++; | |
c6d46f5f | 2729 | } |
6dd5561c FP |
2730 | if (!isalpha (*dbp) |
2731 | && *dbp != '_' | |
2732 | && *dbp != '$') | |
2733 | return; | |
2734 | for (cp = dbp + 1; | |
2735 | (*cp | |
2736 | && (isalpha (*cp) || isdigit (*cp) || (*cp == '_') || (*cp == '$'))); | |
2737 | cp++) | |
2738 | continue; | |
d8913c1c FP |
2739 | pfnote ((CTAGS) ? savenstr (dbp, cp-dbp) : NULL, TRUE, |
2740 | lb.buffer, cp - lb.buffer + 1, lineno, linecharno); | |
c6d46f5f | 2741 | } |
c6d46f5f | 2742 | |
b9755a12 | 2743 | void |
6dd5561c FP |
2744 | Fortran_functions (inf) |
2745 | FILE *inf; | |
c6d46f5f JB |
2746 | { |
2747 | lineno = 0; | |
2748 | charno = 0; | |
c6d46f5f | 2749 | |
6dd5561c | 2750 | while (!feof (inf)) |
c6d46f5f JB |
2751 | { |
2752 | lineno++; | |
2753 | linecharno = charno; | |
6dd5561c | 2754 | charno += readline (&lb, inf); |
c6d46f5f JB |
2755 | dbp = lb.buffer; |
2756 | if (*dbp == '%') | |
2757 | dbp++; /* Ratfor escape to fortran */ | |
2758 | while (isspace (*dbp)) | |
2759 | dbp++; | |
108c932a | 2760 | if (*dbp == '\0') |
c6d46f5f | 2761 | continue; |
79263656 | 2762 | switch (lowcase (*dbp)) |
c6d46f5f JB |
2763 | { |
2764 | case 'i': | |
2765 | if (tail ("integer")) | |
2766 | takeprec (); | |
2767 | break; | |
2768 | case 'r': | |
2769 | if (tail ("real")) | |
2770 | takeprec (); | |
2771 | break; | |
2772 | case 'l': | |
2773 | if (tail ("logical")) | |
2774 | takeprec (); | |
2775 | break; | |
2776 | case 'c': | |
2777 | if (tail ("complex") || tail ("character")) | |
2778 | takeprec (); | |
2779 | break; | |
2780 | case 'd': | |
2781 | if (tail ("double")) | |
2782 | { | |
2783 | while (isspace (*dbp)) | |
2784 | dbp++; | |
108c932a | 2785 | if (*dbp == '\0') |
c6d46f5f JB |
2786 | continue; |
2787 | if (tail ("precision")) | |
2788 | break; | |
2789 | continue; | |
2790 | } | |
2791 | break; | |
2792 | } | |
2793 | while (isspace (*dbp)) | |
2794 | dbp++; | |
108c932a | 2795 | if (*dbp == '\0') |
c6d46f5f | 2796 | continue; |
79263656 | 2797 | switch (lowcase (*dbp)) |
c6d46f5f JB |
2798 | { |
2799 | case 'f': | |
2800 | if (tail ("function")) | |
6dd5561c | 2801 | getit (inf); |
c6d46f5f JB |
2802 | continue; |
2803 | case 's': | |
2804 | if (tail ("subroutine")) | |
6dd5561c | 2805 | getit (inf); |
c6d46f5f | 2806 | continue; |
8a6c8bcf RS |
2807 | case 'e': |
2808 | if (tail ("entry")) | |
6dd5561c | 2809 | getit (inf); |
8a6c8bcf | 2810 | continue; |
c6d46f5f JB |
2811 | case 'p': |
2812 | if (tail ("program")) | |
2813 | { | |
6dd5561c | 2814 | getit (inf); |
c6d46f5f JB |
2815 | continue; |
2816 | } | |
2817 | if (tail ("procedure")) | |
6dd5561c | 2818 | getit (inf); |
c6d46f5f JB |
2819 | continue; |
2820 | } | |
2821 | } | |
c6d46f5f | 2822 | } |
6dd5561c FP |
2823 | \f |
2824 | /* | |
2825 | * Bob Weiner, Motorola Inc., 4/3/94 | |
2826 | * Unix and microcontroller assembly tag handling | |
2827 | * look for '^[a-zA-Z_.$][a-zA_Z0-9_.$]*[: ^I^J]' | |
2828 | */ | |
c6d46f5f | 2829 | void |
6dd5561c FP |
2830 | Asm_labels (inf) |
2831 | FILE *inf; | |
c6d46f5f JB |
2832 | { |
2833 | register char *cp; | |
c6d46f5f JB |
2834 | |
2835 | lineno = 0; | |
2836 | charno = 0; | |
c6d46f5f | 2837 | |
6dd5561c | 2838 | while (!feof (inf)) |
c6d46f5f JB |
2839 | { |
2840 | lineno++; | |
2841 | linecharno = charno; | |
6dd5561c FP |
2842 | charno += readline (&lb, inf); |
2843 | cp = lb.buffer; | |
2844 | ||
2845 | /* If first char is alphabetic or one of [_.$], test for colon | |
2846 | following identifier. */ | |
2847 | if (isalpha (*cp) || *cp == '_' || *cp == '.' || *cp == '$') | |
2848 | { | |
2849 | /* Read past label. */ | |
2850 | cp++; | |
2851 | while (isalnum (*cp) || *cp == '_' || *cp == '.' || *cp == '$') | |
2852 | cp++; | |
2853 | if (*cp == ':' || isspace (*cp)) | |
2854 | { | |
2855 | /* Found end of label, so copy it and add it to the table. */ | |
d8913c1c | 2856 | pfnote ((CTAGS) ? savenstr(lb.buffer, cp-lb.buffer) : NULL, TRUE, |
55597f90 | 2857 | lb.buffer, cp - lb.buffer + 1, lineno, linecharno); |
6dd5561c FP |
2858 | } |
2859 | } | |
c6d46f5f JB |
2860 | } |
2861 | } | |
2862 | \f | |
1f638249 FP |
2863 | /* |
2864 | * Perl support by Bart Robinson <lomew@cs.utah.edu> | |
2865 | * Perl sub names: look for /^sub[ \t\n]+[^ \t\n{]+/ | |
2866 | */ | |
2867 | void | |
2868 | Perl_functions (inf) | |
2869 | FILE *inf; | |
2870 | { | |
2871 | register char *cp; | |
2872 | ||
2873 | lineno = 0; | |
2874 | charno = 0; | |
2875 | ||
2876 | while (!feof (inf)) | |
2877 | { | |
2878 | lineno++; | |
2879 | linecharno = charno; | |
2880 | charno += readline (&lb, inf); | |
2881 | cp = lb.buffer; | |
2882 | ||
2883 | if (*cp++ == 's' && *cp++ == 'u' && *cp++ == 'b' && isspace(*cp++)) | |
2884 | { | |
2885 | while (*cp && isspace(*cp)) | |
2886 | cp++; | |
2887 | while (*cp && ! isspace(*cp) && *cp != '{') | |
2888 | cp++; | |
d8913c1c | 2889 | pfnote ((CTAGS) ? savenstr (lb.buffer, cp-lb.buffer) : NULL, TRUE, |
1f638249 FP |
2890 | lb.buffer, cp - lb.buffer + 1, lineno, linecharno); |
2891 | } | |
2892 | } | |
2893 | } | |
2894 | \f | |
c6d46f5f JB |
2895 | /* Added by Mosur Mohan, 4/22/88 */ |
2896 | /* Pascal parsing */ | |
2897 | ||
2898 | #define GET_NEW_LINE \ | |
2899 | { \ | |
2900 | linecharno = charno; lineno++; \ | |
2901 | charno += 1 + readline (&lb, inf); \ | |
2902 | dbp = lb.buffer; \ | |
2903 | } | |
2904 | ||
aab1fdae FP |
2905 | /* |
2906 | * Locates tags for procedures & functions. Doesn't do any type- or | |
2907 | * var-definitions. It does look for the keyword "extern" or | |
2908 | * "forward" immediately following the procedure statement; if found, | |
2909 | * the tag is skipped. | |
c6d46f5f | 2910 | */ |
c6d46f5f | 2911 | void |
6dd5561c FP |
2912 | Pascal_functions (inf) |
2913 | FILE *inf; | |
c6d46f5f JB |
2914 | { |
2915 | struct linebuffer tline; /* mostly copied from C_entries */ | |
2916 | long save_lcno; | |
108c932a | 2917 | int save_lineno, save_len; |
d8913c1c | 2918 | char c, *cp, *namebuf; |
c6d46f5f JB |
2919 | |
2920 | logical /* each of these flags is TRUE iff: */ | |
b9755a12 | 2921 | incomment, /* point is inside a comment */ |
c6d46f5f | 2922 | inquote, /* point is inside '..' string */ |
108c932a FP |
2923 | get_tagname, /* point is after PROCEDURE/FUNCTION |
2924 | keyword, so next item = potential tag */ | |
c6d46f5f JB |
2925 | found_tag, /* point is after a potential tag */ |
2926 | inparms, /* point is within parameter-list */ | |
108c932a FP |
2927 | verify_tag; /* point has passed the parm-list, so the |
2928 | next token will determine whether this | |
2929 | is a FORWARD/EXTERN to be ignored, or | |
2930 | whether it is a real tag */ | |
c6d46f5f JB |
2931 | |
2932 | lineno = 0; | |
2933 | charno = 0; | |
2934 | dbp = lb.buffer; | |
108c932a FP |
2935 | *dbp = '\0'; |
2936 | save_len = 0; | |
c6d46f5f JB |
2937 | initbuffer (&tline); |
2938 | ||
b9755a12 | 2939 | incomment = inquote = FALSE; |
c6d46f5f JB |
2940 | found_tag = FALSE; /* have a proc name; check if extern */ |
2941 | get_tagname = FALSE; /* have found "procedure" keyword */ | |
2942 | inparms = FALSE; /* found '(' after "proc" */ | |
2943 | verify_tag = FALSE; /* check if "extern" is ahead */ | |
2944 | ||
2945 | /* long main loop to get next char */ | |
6dd5561c | 2946 | while (!feof (inf)) |
c6d46f5f JB |
2947 | { |
2948 | c = *dbp++; | |
55597f90 | 2949 | if (c == '\0') /* if end of line */ |
c6d46f5f JB |
2950 | { |
2951 | GET_NEW_LINE; | |
55597f90 | 2952 | if (*dbp == '\0') |
c6d46f5f JB |
2953 | continue; |
2954 | if (!((found_tag && verify_tag) || | |
2955 | get_tagname)) | |
108c932a FP |
2956 | c = *dbp++; /* only if don't need *dbp pointing |
2957 | to the beginning of the name of | |
2958 | the procedure or function */ | |
c6d46f5f | 2959 | } |
b9755a12 | 2960 | if (incomment) |
c6d46f5f | 2961 | { |
108c932a | 2962 | if (c == '}') /* within { } comments */ |
b9755a12 | 2963 | incomment = FALSE; |
108c932a | 2964 | else if (c == '*' && *dbp == ')') /* within (* *) comments */ |
c6d46f5f | 2965 | { |
b9755a12 FP |
2966 | dbp++; |
2967 | incomment = FALSE; | |
c6d46f5f JB |
2968 | } |
2969 | continue; | |
2970 | } | |
2971 | else if (inquote) | |
2972 | { | |
2973 | if (c == '\'') | |
2974 | inquote = FALSE; | |
2975 | continue; | |
2976 | } | |
55597f90 | 2977 | else |
c6d46f5f JB |
2978 | switch (c) |
2979 | { | |
2980 | case '\'': | |
2981 | inquote = TRUE; /* found first quote */ | |
2982 | continue; | |
108c932a | 2983 | case '{': /* found open { comment */ |
b9755a12 | 2984 | incomment = TRUE; |
c6d46f5f JB |
2985 | continue; |
2986 | case '(': | |
108c932a | 2987 | if (*dbp == '*') /* found open (* comment */ |
c6d46f5f | 2988 | { |
b9755a12 | 2989 | incomment = TRUE; |
c6d46f5f JB |
2990 | dbp++; |
2991 | } | |
2992 | else if (found_tag) /* found '(' after tag, i.e., parm-list */ | |
2993 | inparms = TRUE; | |
2994 | continue; | |
2995 | case ')': /* end of parms list */ | |
2996 | if (inparms) | |
2997 | inparms = FALSE; | |
2998 | continue; | |
2999 | case ';': | |
108c932a | 3000 | if (found_tag && !inparms) /* end of proc or fn stmt */ |
c6d46f5f JB |
3001 | { |
3002 | verify_tag = TRUE; | |
3003 | break; | |
3004 | } | |
3005 | continue; | |
3006 | } | |
108c932a | 3007 | if (found_tag && verify_tag && (*dbp != ' ')) |
c6d46f5f JB |
3008 | { |
3009 | /* check if this is an "extern" declaration */ | |
108c932a | 3010 | if (*dbp == '\0') |
c6d46f5f | 3011 | continue; |
108c932a | 3012 | if (lowcase (*dbp == 'e')) |
c6d46f5f JB |
3013 | { |
3014 | if (tail ("extern")) /* superfluous, really! */ | |
3015 | { | |
3016 | found_tag = FALSE; | |
3017 | verify_tag = FALSE; | |
3018 | } | |
3019 | } | |
108c932a | 3020 | else if (lowcase (*dbp) == 'f') |
c6d46f5f JB |
3021 | { |
3022 | if (tail ("forward")) /* check for forward reference */ | |
3023 | { | |
3024 | found_tag = FALSE; | |
3025 | verify_tag = FALSE; | |
3026 | } | |
3027 | } | |
108c932a | 3028 | if (found_tag && verify_tag) /* not external proc, so make tag */ |
c6d46f5f JB |
3029 | { |
3030 | found_tag = FALSE; | |
3031 | verify_tag = FALSE; | |
d8913c1c | 3032 | pfnote (namebuf, TRUE, |
108c932a | 3033 | tline.buffer, save_len, save_lineno, save_lcno); |
c6d46f5f JB |
3034 | continue; |
3035 | } | |
3036 | } | |
3037 | if (get_tagname) /* grab name of proc or fn */ | |
3038 | { | |
108c932a | 3039 | if (*dbp == '\0') |
c6d46f5f JB |
3040 | continue; |
3041 | ||
3042 | /* save all values for later tagging */ | |
d8913c1c | 3043 | GROW_LINEBUFFER (tline, strlen (lb.buffer) + 1); |
c6d46f5f JB |
3044 | strcpy (tline.buffer, lb.buffer); |
3045 | save_lineno = lineno; | |
3046 | save_lcno = linecharno; | |
3047 | ||
3048 | /* grab block name */ | |
d8913c1c | 3049 | for (cp = dbp + 1; *cp && (!endtoken (*cp)); cp++) |
c6d46f5f | 3050 | continue; |
d8913c1c FP |
3051 | namebuf = (CTAGS) ? savenstr (dbp, cp-dbp) : NULL; |
3052 | dbp = cp; /* set dbp to e-o-token */ | |
108c932a | 3053 | save_len = dbp - lb.buffer + 1; |
c6d46f5f JB |
3054 | get_tagname = FALSE; |
3055 | found_tag = TRUE; | |
3056 | continue; | |
3057 | ||
3058 | /* and proceed to check for "extern" */ | |
3059 | } | |
55597f90 | 3060 | else if (!incomment && !inquote && !found_tag) |
c6d46f5f JB |
3061 | { |
3062 | /* check for proc/fn keywords */ | |
79263656 | 3063 | switch (lowcase (c)) |
c6d46f5f JB |
3064 | { |
3065 | case 'p': | |
3066 | if (tail ("rocedure")) /* c = 'p', dbp has advanced */ | |
3067 | get_tagname = TRUE; | |
3068 | continue; | |
3069 | case 'f': | |
3070 | if (tail ("unction")) | |
3071 | get_tagname = TRUE; | |
3072 | continue; | |
3073 | } | |
3074 | } | |
6dd5561c | 3075 | } /* while not eof */ |
c5007f46 | 3076 | |
108c932a | 3077 | free (tline.buffer); |
c6d46f5f JB |
3078 | } |
3079 | \f | |
3080 | /* | |
3081 | * lisp tag functions | |
55597f90 | 3082 | * look for (def or (DEF, quote or QUOTE |
c6d46f5f | 3083 | */ |
c6d46f5f | 3084 | int |
55597f90 FP |
3085 | L_isdef (strp) |
3086 | register char *strp; | |
c6d46f5f | 3087 | { |
55597f90 FP |
3088 | return ((strp[1] == 'd' || strp[1] == 'D') |
3089 | && (strp[2] == 'e' || strp[2] == 'E') | |
3090 | && (strp[3] == 'f' || strp[3] == 'F')); | |
31d4b314 FP |
3091 | } |
3092 | ||
3093 | int | |
55597f90 FP |
3094 | L_isquote (strp) |
3095 | register char *strp; | |
3096 | { | |
3097 | return ((*(++strp) == 'q' || *strp == 'Q') | |
3098 | && (*(++strp) == 'u' || *strp == 'U') | |
3099 | && (*(++strp) == 'o' || *strp == 'O') | |
3100 | && (*(++strp) == 't' || *strp == 'T') | |
3101 | && (*(++strp) == 'e' || *strp == 'E') | |
3102 | && isspace(*(++strp))); | |
c6d46f5f JB |
3103 | } |
3104 | ||
3105 | void | |
3106 | L_getit () | |
3107 | { | |
3108 | register char *cp; | |
c6d46f5f | 3109 | |
31d4b314 FP |
3110 | if (*dbp == '\'') /* Skip prefix quote */ |
3111 | dbp++; | |
3112 | else if (*dbp == '(' && L_isquote (dbp)) /* Skip "(quote " */ | |
3113 | { | |
3114 | dbp += 7; | |
3115 | while (isspace(*dbp)) | |
3116 | dbp++; | |
3117 | } | |
55597f90 FP |
3118 | for (cp = dbp /*+1*/; |
3119 | *cp && *cp != '(' && *cp != ' ' && *cp != ')'; | |
3120 | cp++) | |
c6d46f5f | 3121 | continue; |
31d4b314 FP |
3122 | if (cp == dbp) |
3123 | return; | |
c5007f46 | 3124 | |
d8913c1c FP |
3125 | pfnote ((CTAGS) ? savenstr (dbp, cp-dbp) : NULL, TRUE, |
3126 | lb.buffer, cp - lb.buffer + 1, lineno, linecharno); | |
c6d46f5f | 3127 | } |
6dd5561c FP |
3128 | |
3129 | void | |
3130 | Lisp_functions (inf) | |
3131 | FILE *inf; | |
3132 | { | |
3133 | lineno = 0; | |
3134 | charno = 0; | |
6dd5561c FP |
3135 | |
3136 | while (!feof (inf)) | |
3137 | { | |
3138 | lineno++; | |
3139 | linecharno = charno; | |
3140 | charno += readline (&lb, inf); | |
3141 | dbp = lb.buffer; | |
3142 | if (dbp[0] == '(') | |
3143 | { | |
3144 | if (L_isdef (dbp)) | |
3145 | { | |
3146 | while (!isspace (*dbp)) | |
3147 | dbp++; | |
3148 | while (isspace (*dbp)) | |
3149 | dbp++; | |
3150 | L_getit (); | |
3151 | } | |
3152 | else | |
3153 | { | |
3154 | /* Check for (foo::defmumble name-defined ... */ | |
3155 | do | |
3156 | dbp++; | |
3157 | while (*dbp && !isspace (*dbp) | |
3158 | && *dbp != ':' && *dbp != '(' && *dbp != ')'); | |
3159 | if (*dbp == ':') | |
3160 | { | |
3161 | do | |
3162 | dbp++; | |
3163 | while (*dbp == ':'); | |
3164 | ||
3165 | if (L_isdef (dbp - 1)) | |
3166 | { | |
3167 | while (!isspace (*dbp)) | |
3168 | dbp++; | |
3169 | while (isspace (*dbp)) | |
3170 | dbp++; | |
3171 | L_getit (); | |
3172 | } | |
3173 | } | |
3174 | } | |
3175 | } | |
3176 | } | |
3177 | } | |
c6d46f5f JB |
3178 | \f |
3179 | /* | |
3180 | * Scheme tag functions | |
3181 | * look for (def... xyzzy | |
3182 | * look for (def... (xyzzy | |
3183 | * look for (def ... ((...(xyzzy .... | |
3184 | * look for (set! xyzzy | |
3185 | */ | |
3186 | ||
6dd5561c | 3187 | void get_scheme (); |
c6d46f5f JB |
3188 | |
3189 | void | |
6dd5561c FP |
3190 | Scheme_functions (inf) |
3191 | FILE *inf; | |
c6d46f5f JB |
3192 | { |
3193 | lineno = 0; | |
3194 | charno = 0; | |
c6d46f5f | 3195 | |
6dd5561c | 3196 | while (!feof (inf)) |
c6d46f5f JB |
3197 | { |
3198 | lineno++; | |
3199 | linecharno = charno; | |
6dd5561c | 3200 | charno += readline (&lb, inf); |
c6d46f5f JB |
3201 | dbp = lb.buffer; |
3202 | if (dbp[0] == '(' && | |
3203 | (dbp[1] == 'D' || dbp[1] == 'd') && | |
3204 | (dbp[2] == 'E' || dbp[2] == 'e') && | |
3205 | (dbp[3] == 'F' || dbp[3] == 'f')) | |
3206 | { | |
3207 | while (!isspace (*dbp)) | |
3208 | dbp++; | |
3209 | /* Skip over open parens and white space */ | |
3210 | while (*dbp && (isspace (*dbp) || *dbp == '(')) | |
3211 | dbp++; | |
3212 | get_scheme (); | |
3213 | } | |
3214 | if (dbp[0] == '(' && | |
3215 | (dbp[1] == 'S' || dbp[1] == 's') && | |
3216 | (dbp[2] == 'E' || dbp[2] == 'e') && | |
3217 | (dbp[3] == 'T' || dbp[3] == 't') && | |
3218 | (dbp[4] == '!' || dbp[4] == '!') && | |
3219 | (isspace (dbp[5]))) | |
3220 | { | |
3221 | while (!isspace (*dbp)) | |
3222 | dbp++; | |
3223 | /* Skip over white space */ | |
3224 | while (isspace (*dbp)) | |
3225 | dbp++; | |
3226 | get_scheme (); | |
3227 | } | |
3228 | } | |
3229 | } | |
3230 | ||
6dd5561c | 3231 | void |
c6d46f5f JB |
3232 | get_scheme () |
3233 | { | |
3234 | register char *cp; | |
c6d46f5f | 3235 | |
108c932a | 3236 | if (*dbp == '\0') |
c6d46f5f JB |
3237 | return; |
3238 | /* Go till you get to white space or a syntactic break */ | |
55597f90 FP |
3239 | for (cp = dbp + 1; |
3240 | *cp && *cp != '(' && *cp != ')' && !isspace (*cp); | |
3241 | cp++) | |
c6d46f5f | 3242 | continue; |
d8913c1c FP |
3243 | pfnote ((CTAGS) ? savenstr (dbp, cp-dbp) : NULL, TRUE, |
3244 | lb.buffer, cp - lb.buffer + 1, lineno, linecharno); | |
c6d46f5f JB |
3245 | } |
3246 | \f | |
3247 | /* Find tags in TeX and LaTeX input files. */ | |
3248 | ||
3249 | /* TEX_toktab is a table of TeX control sequences that define tags. | |
3250 | Each TEX_tabent records one such control sequence. | |
3251 | CONVERT THIS TO USE THE Stab TYPE!! */ | |
c6d46f5f JB |
3252 | struct TEX_tabent |
3253 | { | |
3254 | char *name; | |
3255 | int len; | |
3256 | }; | |
3257 | ||
3258 | struct TEX_tabent *TEX_toktab = NULL; /* Table with tag tokens */ | |
3259 | ||
3260 | /* Default set of control sequences to put into TEX_toktab. | |
3261 | The value of environment var TEXTAGS is prepended to this. */ | |
3262 | ||
6dd5561c | 3263 | char *TEX_defenv = "\ |
c5007f46 FP |
3264 | :chapter:section:subsection:subsubsection:eqno:label:ref:cite:bibitem\ |
3265 | :part:appendix:entry:index"; | |
c6d46f5f JB |
3266 | |
3267 | void TEX_mode (); | |
3268 | struct TEX_tabent *TEX_decode_env (); | |
c6d46f5f | 3269 | int TEX_Token (); |
108c932a FP |
3270 | #if TeX_named_tokens |
3271 | void TEX_getit (); | |
3272 | #endif | |
c6d46f5f | 3273 | |
6dd5561c FP |
3274 | char TEX_esc = '\\'; |
3275 | char TEX_opgrp = '{'; | |
3276 | char TEX_clgrp = '}'; | |
c6d46f5f JB |
3277 | |
3278 | /* | |
3279 | * TeX/LaTeX scanning loop. | |
3280 | */ | |
c6d46f5f | 3281 | void |
6dd5561c FP |
3282 | TeX_functions (inf) |
3283 | FILE *inf; | |
c6d46f5f JB |
3284 | { |
3285 | char *lasthit; | |
3286 | ||
3287 | lineno = 0; | |
3288 | charno = 0; | |
c6d46f5f JB |
3289 | |
3290 | /* Select either \ or ! as escape character. */ | |
6dd5561c | 3291 | TEX_mode (inf); |
c6d46f5f JB |
3292 | |
3293 | /* Initialize token table once from environment. */ | |
3294 | if (!TEX_toktab) | |
3295 | TEX_toktab = TEX_decode_env ("TEXTAGS", TEX_defenv); | |
3296 | ||
6dd5561c | 3297 | while (!feof (inf)) |
d2729198 | 3298 | { /* Scan each line in file */ |
c6d46f5f JB |
3299 | lineno++; |
3300 | linecharno = charno; | |
6dd5561c | 3301 | charno += readline (&lb, inf); |
c6d46f5f JB |
3302 | dbp = lb.buffer; |
3303 | lasthit = dbp; | |
b02c5fea | 3304 | while (dbp = etags_strchr (dbp, TEX_esc)) /* Look at each esc in line */ |
8a6c8bcf RS |
3305 | { |
3306 | register int i; | |
c6d46f5f | 3307 | |
8a6c8bcf RS |
3308 | if (!*(++dbp)) |
3309 | break; | |
3310 | linecharno += dbp - lasthit; | |
c6d46f5f | 3311 | lasthit = dbp; |
8a6c8bcf RS |
3312 | i = TEX_Token (lasthit); |
3313 | if (0 <= i) | |
c6d46f5f | 3314 | { |
108c932a FP |
3315 | pfnote (NULL, TRUE, |
3316 | lb.buffer, strlen (lb.buffer), lineno, linecharno); | |
3317 | #if TeX_named_tokens | |
8a6c8bcf | 3318 | TEX_getit (lasthit, TEX_toktab[i].len); |
108c932a | 3319 | #endif |
d2729198 | 3320 | break; /* We only save a line once */ |
c6d46f5f JB |
3321 | } |
3322 | } | |
3323 | } | |
3324 | } | |
3325 | ||
3326 | #define TEX_LESC '\\' | |
3327 | #define TEX_SESC '!' | |
3328 | #define TEX_cmt '%' | |
3329 | ||
aab1fdae FP |
3330 | /* Figure out whether TeX's escapechar is '\\' or '!' and set grouping |
3331 | chars accordingly. */ | |
c6d46f5f | 3332 | void |
6dd5561c FP |
3333 | TEX_mode (inf) |
3334 | FILE *inf; | |
c6d46f5f JB |
3335 | { |
3336 | int c; | |
3337 | ||
6dd5561c | 3338 | while ((c = getc (inf)) != EOF) |
c6d46f5f JB |
3339 | { |
3340 | /* Skip to next line if we hit the TeX comment char. */ | |
3341 | if (c == TEX_cmt) | |
3342 | while (c != '\n') | |
6dd5561c | 3343 | c = getc (inf); |
c6d46f5f JB |
3344 | else if (c == TEX_LESC || c == TEX_SESC ) |
3345 | break; | |
3346 | } | |
3347 | ||
3348 | if (c == TEX_LESC) | |
3349 | { | |
3350 | TEX_esc = TEX_LESC; | |
3351 | TEX_opgrp = '{'; | |
3352 | TEX_clgrp = '}'; | |
3353 | } | |
3354 | else | |
3355 | { | |
3356 | TEX_esc = TEX_SESC; | |
3357 | TEX_opgrp = '<'; | |
3358 | TEX_clgrp = '>'; | |
3359 | } | |
6dd5561c | 3360 | rewind (inf); |
c6d46f5f JB |
3361 | } |
3362 | ||
aab1fdae FP |
3363 | /* Read environment and prepend it to the default string. |
3364 | Build token table. */ | |
c6d46f5f JB |
3365 | struct TEX_tabent * |
3366 | TEX_decode_env (evarname, defenv) | |
3367 | char *evarname; | |
3368 | char *defenv; | |
3369 | { | |
3370 | register char *env, *p; | |
c6d46f5f JB |
3371 | |
3372 | struct TEX_tabent *tab; | |
3373 | int size, i; | |
3374 | ||
3375 | /* Append default string to environment. */ | |
3376 | env = getenv (evarname); | |
3377 | if (!env) | |
3378 | env = defenv; | |
3379 | else | |
3380 | env = concat (env, defenv, ""); | |
3381 | ||
3382 | /* Allocate a token table */ | |
3383 | for (size = 1, p = env; p;) | |
b02c5fea | 3384 | if ((p = etags_strchr (p, ':')) && *(++p)) |
c6d46f5f | 3385 | size++; |
8a6c8bcf RS |
3386 | /* Add 1 to leave room for null terminator. */ |
3387 | tab = xnew (size + 1, struct TEX_tabent); | |
c6d46f5f JB |
3388 | |
3389 | /* Unpack environment string into token table. Be careful about */ | |
3390 | /* zero-length strings (leading ':', "::" and trailing ':') */ | |
3391 | for (i = 0; *env;) | |
3392 | { | |
b02c5fea | 3393 | p = etags_strchr (env, ':'); |
c6d46f5f JB |
3394 | if (!p) /* End of environment string. */ |
3395 | p = env + strlen (env); | |
3396 | if (p - env > 0) | |
3397 | { /* Only non-zero strings. */ | |
3398 | tab[i].name = savenstr (env, p - env); | |
3399 | tab[i].len = strlen (tab[i].name); | |
3400 | i++; | |
3401 | } | |
3402 | if (*p) | |
3403 | env = p + 1; | |
3404 | else | |
3405 | { | |
3406 | tab[i].name = NULL; /* Mark end of table. */ | |
3407 | tab[i].len = 0; | |
3408 | break; | |
3409 | } | |
3410 | } | |
3411 | return tab; | |
3412 | } | |
3413 | ||
108c932a | 3414 | #if TeX_named_tokens |
c6d46f5f JB |
3415 | /* Record a tag defined by a TeX command of length LEN and starting at NAME. |
3416 | The name being defined actually starts at (NAME + LEN + 1). | |
3417 | But we seem to include the TeX command in the tag name. */ | |
c6d46f5f JB |
3418 | void |
3419 | TEX_getit (name, len) | |
3420 | char *name; | |
3421 | int len; | |
3422 | { | |
3423 | char *p = name + len; | |
c6d46f5f | 3424 | |
108c932a | 3425 | if (*name == '\0') |
c6d46f5f JB |
3426 | return; |
3427 | ||
3428 | /* Let tag name extend to next group close (or end of line) */ | |
3429 | while (*p && *p != TEX_clgrp) | |
3430 | p++; | |
108c932a FP |
3431 | pfnote (savenstr (name, p-name), TRUE, |
3432 | lb.buffer, strlen (lb.buffer), lineno, linecharno); | |
c6d46f5f | 3433 | } |
108c932a | 3434 | #endif |
c6d46f5f JB |
3435 | |
3436 | /* If the text at CP matches one of the tag-defining TeX command names, | |
b02c5fea | 3437 | return the pointer to the first occurrence of that command in TEX_toktab. |
aab1fdae FP |
3438 | Otherwise return -1. |
3439 | Keep the capital `T' in `Token' for dumb truncating compilers | |
c6d46f5f JB |
3440 | (this distinguishes it from `TEX_toktab' */ |
3441 | int | |
3442 | TEX_Token (cp) | |
3443 | char *cp; | |
3444 | { | |
3445 | int i; | |
3446 | ||
3447 | for (i = 0; TEX_toktab[i].len > 0; i++) | |
1a0d8c80 | 3448 | if (strneq (TEX_toktab[i].name, cp, TEX_toktab[i].len)) |
c6d46f5f JB |
3449 | return i; |
3450 | return -1; | |
3451 | } | |
3452 | \f | |
3453 | /* Support for Prolog. */ | |
3454 | ||
aab1fdae | 3455 | /* Whole head (not only functor, but also arguments) |
c6d46f5f | 3456 | is gotten in compound term. */ |
c6d46f5f | 3457 | void |
b9755a12 | 3458 | prolog_getit (s) |
c6d46f5f | 3459 | char *s; |
c6d46f5f | 3460 | { |
55597f90 | 3461 | char *save_s; |
c6d46f5f JB |
3462 | int insquote, npar; |
3463 | ||
3464 | save_s = s; | |
3465 | insquote = FALSE; | |
3466 | npar = 0; | |
3467 | while (1) | |
3468 | { | |
55597f90 | 3469 | if (s[0] == '\0') /* syntax error. */ |
c6d46f5f | 3470 | return; |
55597f90 | 3471 | else if (insquote && s[0] == '\'' && s[1] == '\'') |
c6d46f5f | 3472 | s += 2; |
55597f90 | 3473 | else if (s[0] == '\'') |
c6d46f5f JB |
3474 | { |
3475 | insquote = !insquote; | |
3476 | s++; | |
3477 | } | |
55597f90 | 3478 | else if (!insquote && s[0] == '(') |
c6d46f5f JB |
3479 | { |
3480 | npar++; | |
3481 | s++; | |
3482 | } | |
55597f90 | 3483 | else if (!insquote && s[0] == ')') |
c6d46f5f JB |
3484 | { |
3485 | npar--; | |
3486 | s++; | |
3487 | if (npar == 0) | |
3488 | break; | |
3489 | else if (npar < 0) /* syntax error. */ | |
3490 | return; | |
3491 | } | |
55597f90 FP |
3492 | else if (!insquote && s[0] == '.' |
3493 | && (isspace (s[1]) || s[1] == '\0')) | |
c6d46f5f JB |
3494 | { /* fullstop. */ |
3495 | if (npar != 0) /* syntax error. */ | |
3496 | return; | |
3497 | s++; | |
3498 | break; | |
3499 | } | |
3500 | else | |
3501 | s++; | |
3502 | } | |
d8913c1c FP |
3503 | pfnote ((CTAGS) ? savenstr (save_s, s-save_s) : NULL, TRUE, |
3504 | save_s, s-save_s, lineno, linecharno); | |
c6d46f5f JB |
3505 | } |
3506 | ||
3507 | /* It is assumed that prolog predicate starts from column 0. */ | |
c6d46f5f | 3508 | void |
6dd5561c FP |
3509 | Prolog_functions (inf) |
3510 | FILE *inf; | |
c6d46f5f JB |
3511 | { |
3512 | void skip_comment (), prolog_getit (); | |
3513 | ||
3514 | lineno = linecharno = charno = 0; | |
6dd5561c | 3515 | while (!feof (inf)) |
c6d46f5f JB |
3516 | { |
3517 | lineno++; | |
3518 | linecharno += charno; | |
6dd5561c | 3519 | charno = readline (&lb, inf) + 1; /* 1 for newline. */ |
c6d46f5f JB |
3520 | dbp = lb.buffer; |
3521 | if (isspace (dbp[0])) /* not predicate header. */ | |
3522 | continue; | |
3523 | else if (dbp[0] == '%') /* comment. */ | |
3524 | continue; | |
3525 | else if (dbp[0] == '/' && dbp[1] == '*') /* comment. */ | |
6dd5561c | 3526 | skip_comment (&lb, inf, &lineno, &linecharno); |
c6d46f5f | 3527 | else /* found. */ |
b9755a12 | 3528 | prolog_getit (dbp); |
c6d46f5f JB |
3529 | } |
3530 | } | |
3531 | ||
3532 | void | |
6dd5561c | 3533 | skip_comment (plb, inf, plineno, plinecharno) |
c6d46f5f | 3534 | struct linebuffer *plb; |
6dd5561c | 3535 | FILE *inf; |
c6d46f5f JB |
3536 | int *plineno; /* result */ |
3537 | long *plinecharno; /* result */ | |
3538 | { | |
b9755a12 FP |
3539 | char *cp; |
3540 | ||
3541 | do | |
c6d46f5f | 3542 | { |
b9755a12 FP |
3543 | for (cp = plb->buffer; *cp != '\0'; cp++) |
3544 | if (cp[0] == '*' && cp[1] == '/') | |
3545 | return; | |
c6d46f5f | 3546 | (*plineno)++; |
b9755a12 FP |
3547 | *plinecharno += readline (plb, inf) + 1; /* 1 for newline. */ |
3548 | } | |
3549 | while (!feof(inf)); | |
c6d46f5f | 3550 | } |
b9755a12 FP |
3551 | \f |
3552 | #ifdef ETAGS_REGEXPS | |
3553 | /* Take a string like "/blah/" and turn it into "blah", making sure | |
3554 | that the first and last characters are the same, and handling | |
15294654 | 3555 | quoted separator characters. Actually, stops on the occurrence of |
b9755a12 FP |
3556 | an unquoted separator. Also turns "\t" into a Tab character. |
3557 | Returns pointer to terminating separator. Works in place. Null | |
3558 | terminates name string. */ | |
3559 | char * | |
3560 | scan_separators (name) | |
3561 | char *name; | |
3562 | { | |
3563 | char sep = name[0]; | |
3564 | char *copyto = name; | |
3565 | logical quoted = FALSE; | |
3566 | ||
3567 | for (++name; *name != '\0'; ++name) | |
3568 | { | |
3569 | if (quoted) | |
3570 | { | |
3571 | if (*name == 't') | |
3572 | *copyto++ = '\t'; | |
3573 | else if (*name == sep) | |
3574 | *copyto++ = sep; | |
3575 | else | |
3576 | { | |
3577 | /* Something else is quoted, so preserve the quote. */ | |
3578 | *copyto++ = '\\'; | |
3579 | *copyto++ = *name; | |
3580 | } | |
3581 | quoted = FALSE; | |
3582 | } | |
3583 | else if (*name == '\\') | |
3584 | quoted = TRUE; | |
3585 | else if (*name == sep) | |
3586 | break; | |
3587 | else | |
3588 | *copyto++ = *name; | |
3589 | } | |
c6d46f5f | 3590 | |
b9755a12 FP |
3591 | /* Terminate copied string. */ |
3592 | *copyto = '\0'; | |
3593 | return name; | |
3594 | } | |
c6d46f5f | 3595 | |
b9755a12 FP |
3596 | /* Turn a name, which is an ed-style (but Emacs syntax) regular |
3597 | expression, into a real regular expression by compiling it. */ | |
3598 | void | |
3599 | add_regex (regexp_pattern) | |
3600 | char *regexp_pattern; | |
c6d46f5f | 3601 | { |
b9755a12 FP |
3602 | char *name; |
3603 | const char *err; | |
3604 | struct re_pattern_buffer *patbuf; | |
c6d46f5f | 3605 | |
b9755a12 FP |
3606 | if (regexp_pattern == NULL) |
3607 | { | |
3608 | /* Remove existing regexps. */ | |
3609 | num_patterns = 0; | |
3610 | patterns = NULL; | |
3611 | return; | |
3612 | } | |
c6d46f5f | 3613 | |
b9755a12 FP |
3614 | if (regexp_pattern[0] == '\0') |
3615 | { | |
3616 | error ("missing regexp", 0); | |
3617 | return; | |
3618 | } | |
3619 | if (regexp_pattern[strlen(regexp_pattern)-1] != regexp_pattern[0]) | |
3620 | { | |
3621 | error ("%s: unterminated regexp", regexp_pattern); | |
3622 | return; | |
3623 | } | |
3624 | name = scan_separators (regexp_pattern); | |
3625 | if (regexp_pattern[0] == '\0') | |
3626 | { | |
3627 | error ("null regexp", 0); | |
3628 | return; | |
3629 | } | |
3630 | (void) scan_separators (name); | |
3631 | ||
3632 | patbuf = xnew (1, struct re_pattern_buffer); | |
3633 | patbuf->translate = NULL; | |
3634 | patbuf->fastmap = NULL; | |
3635 | patbuf->buffer = NULL; | |
3636 | patbuf->allocated = 0; | |
3637 | ||
3638 | err = re_compile_pattern (regexp_pattern, strlen (regexp_pattern), patbuf); | |
3639 | if (err != NULL) | |
3640 | { | |
3641 | error ("%s while compiling pattern", err); | |
3642 | return; | |
3643 | } | |
3644 | ||
3645 | num_patterns += 1; | |
3646 | if (num_patterns == 1) | |
3647 | patterns = xnew (1, struct pattern); | |
c6d46f5f | 3648 | else |
b9755a12 FP |
3649 | patterns = ((struct pattern *) |
3650 | xrealloc (patterns, | |
3651 | (num_patterns * sizeof (struct pattern)))); | |
3652 | patterns[num_patterns - 1].pattern = patbuf; | |
3653 | patterns[num_patterns - 1].name_pattern = savestr (name); | |
3654 | patterns[num_patterns - 1].error_signaled = FALSE; | |
3655 | } | |
3656 | ||
3657 | /* | |
c5007f46 | 3658 | * Do the substitutions indicated by the regular expression and |
b9755a12 FP |
3659 | * arguments. |
3660 | */ | |
3661 | char * | |
3662 | substitute (in, out, regs) | |
3663 | char *in, *out; | |
3664 | struct re_registers *regs; | |
3665 | { | |
3666 | char *result = NULL, *t; | |
3667 | int size = 0; | |
3668 | ||
3669 | /* Pass 1: figure out how much size to allocate. */ | |
3670 | for (t = out; *t; ++t) | |
3671 | { | |
3672 | if (*t == '\\') | |
3673 | { | |
3674 | ++t; | |
3675 | if (!*t) | |
3676 | { | |
15294654 | 3677 | fprintf (stderr, "%s: pattern substitution ends prematurely\n", |
b9755a12 FP |
3678 | progname); |
3679 | return NULL; | |
3680 | } | |
3681 | if (isdigit (*t)) | |
3682 | { | |
3683 | int dig = *t - '0'; | |
3684 | size += regs->end[dig] - regs->start[dig]; | |
3685 | } | |
3686 | } | |
3687 | } | |
3688 | ||
3689 | /* Allocate space and do the substitutions. */ | |
3690 | result = xnew (size + 1, char); | |
3691 | size = 0; | |
3692 | for (; *out; ++out) | |
3693 | { | |
3694 | if (*out == '\\') | |
3695 | { | |
3696 | ++out; | |
3697 | if (isdigit (*out)) | |
3698 | { | |
3699 | /* Using "dig2" satisfies my debugger. Bleah. */ | |
3700 | int dig2 = *out - '0'; | |
3701 | strncpy (result + size, in + regs->start[dig2], | |
3702 | regs->end[dig2] - regs->start[dig2]); | |
3703 | size += regs->end[dig2] - regs->start[dig2]; | |
3704 | } | |
3705 | else | |
c5007f46 | 3706 | result[size++] = *out; |
b9755a12 FP |
3707 | } |
3708 | else | |
3709 | result[size++] = *out; | |
3710 | } | |
3711 | result[size] = '\0'; | |
3712 | ||
3713 | return result; | |
c6d46f5f JB |
3714 | } |
3715 | \f | |
b9755a12 | 3716 | #endif /* ETAGS_REGEXPS */ |
c6d46f5f | 3717 | /* Initialize a linebuffer for use */ |
c6d46f5f JB |
3718 | void |
3719 | initbuffer (linebuffer) | |
3720 | struct linebuffer *linebuffer; | |
3721 | { | |
3722 | linebuffer->size = 200; | |
3723 | linebuffer->buffer = xnew (200, char); | |
3724 | } | |
3725 | ||
3726 | /* | |
3727 | * Read a line of text from `stream' into `linebuffer'. | |
3728 | * Return the number of characters read from `stream', | |
3729 | * which is the length of the line including the newline, if any. | |
3730 | */ | |
3731 | long | |
b9755a12 | 3732 | readline_internal (linebuffer, stream) |
c6d46f5f JB |
3733 | struct linebuffer *linebuffer; |
3734 | register FILE *stream; | |
3735 | { | |
3736 | char *buffer = linebuffer->buffer; | |
3737 | register char *p = linebuffer->buffer; | |
3738 | register char *pend; | |
aab1fdae | 3739 | int chars_deleted; |
c6d46f5f | 3740 | |
eb8c3be9 | 3741 | pend = p + linebuffer->size; /* Separate to avoid 386/IX compiler bug. */ |
c6d46f5f JB |
3742 | |
3743 | while (1) | |
3744 | { | |
3745 | register int c = getc (stream); | |
3746 | if (p == pend) | |
3747 | { | |
3748 | linebuffer->size *= 2; | |
3749 | buffer = (char *) xrealloc (buffer, linebuffer->size); | |
3750 | p += buffer - linebuffer->buffer; | |
3751 | pend = buffer + linebuffer->size; | |
3752 | linebuffer->buffer = buffer; | |
3753 | } | |
aab1fdae | 3754 | if (c == EOF) |
c6d46f5f | 3755 | { |
aab1fdae FP |
3756 | chars_deleted = 0; |
3757 | break; | |
3758 | } | |
3759 | if (c == '\n') | |
3760 | { | |
a8d9bd4b | 3761 | if (p > buffer && p[-1] == '\r') |
aab1fdae FP |
3762 | { |
3763 | *--p = '\0'; | |
3764 | chars_deleted = 2; | |
3765 | } | |
3766 | else | |
3767 | { | |
3768 | *p = '\0'; | |
3769 | chars_deleted = 1; | |
3770 | } | |
c6d46f5f JB |
3771 | break; |
3772 | } | |
3773 | *p++ = c; | |
3774 | } | |
3775 | ||
aab1fdae | 3776 | return p - buffer + chars_deleted; |
c6d46f5f | 3777 | } |
b9755a12 FP |
3778 | |
3779 | /* | |
3780 | * Like readline_internal, above, but try to match the input | |
3781 | * line against any existing regular expressions. | |
3782 | */ | |
3783 | long | |
3784 | readline (linebuffer, stream) | |
3785 | struct linebuffer *linebuffer; | |
3786 | FILE *stream; | |
3787 | { | |
3788 | /* Read new line. */ | |
b9755a12 | 3789 | long result = readline_internal (linebuffer, stream); |
b9755a12 | 3790 | #ifdef ETAGS_REGEXPS |
d8913c1c FP |
3791 | int i; |
3792 | ||
b9755a12 FP |
3793 | /* Match against all listed patterns. */ |
3794 | for (i = 0; i < num_patterns; ++i) | |
3795 | { | |
3796 | int match = re_match (patterns[i].pattern, linebuffer->buffer, | |
3797 | (int)result, 0, &patterns[i].regs); | |
3798 | switch (match) | |
3799 | { | |
3800 | case -2: | |
3801 | /* Some error. */ | |
3802 | if (!patterns[i].error_signaled) | |
3803 | { | |
3804 | error ("error while matching pattern %d", i); | |
3805 | patterns[i].error_signaled = TRUE; | |
3806 | } | |
3807 | break; | |
3808 | case -1: | |
3809 | /* No match. */ | |
3810 | break; | |
3811 | default: | |
3812 | /* Match occurred. Construct a tag. */ | |
3813 | if (patterns[i].name_pattern[0] != '\0') | |
3814 | { | |
3815 | /* Make a named tag. */ | |
3816 | char *name = substitute (linebuffer->buffer, | |
3817 | patterns[i].name_pattern, | |
3818 | &patterns[i].regs); | |
3819 | if (name != NULL) | |
108c932a FP |
3820 | pfnote (name, TRUE, |
3821 | linebuffer->buffer, match, lineno, linecharno); | |
b9755a12 FP |
3822 | } |
3823 | else | |
3824 | { | |
3825 | /* Make an unnamed tag. */ | |
108c932a FP |
3826 | pfnote (NULL, TRUE, |
3827 | linebuffer->buffer, match, lineno, linecharno); | |
b9755a12 FP |
3828 | } |
3829 | break; | |
3830 | } | |
3831 | } | |
3832 | #endif /* ETAGS_REGEXPS */ | |
3833 | ||
3834 | return result; | |
3835 | } | |
3836 | ||
3837 | /* | |
3838 | * Read a file, but do no processing. This is used to do regexp | |
3839 | * matching on files that have no language defined. | |
3840 | */ | |
3841 | void | |
3842 | just_read_file (inf) | |
3843 | FILE *inf; | |
3844 | { | |
3845 | while (!feof (inf)) | |
3846 | { | |
3847 | ++lineno; | |
3848 | linecharno = charno; | |
3849 | charno += readline (&lb, inf) + 1; | |
3850 | } | |
3851 | } | |
3852 | ||
c6d46f5f | 3853 | \f |
55597f90 FP |
3854 | /* |
3855 | * Return a pointer to a space of size strlen(cp)+1 allocated | |
3856 | * with xnew where the string CP has been copied. | |
3857 | */ | |
c6d46f5f JB |
3858 | char * |
3859 | savestr (cp) | |
3860 | char *cp; | |
3861 | { | |
3862 | return savenstr (cp, strlen (cp)); | |
3863 | } | |
3864 | ||
55597f90 FP |
3865 | /* |
3866 | * Return a pointer to a space of size LEN+1 allocated with xnew where | |
3867 | * the string CP has been copied for at most the first LEN characters. | |
3868 | */ | |
c6d46f5f JB |
3869 | char * |
3870 | savenstr (cp, len) | |
3871 | char *cp; | |
3872 | int len; | |
3873 | { | |
3874 | register char *dp; | |
3875 | ||
3876 | dp = xnew (len + 1, char); | |
1a0d8c80 | 3877 | strncpy (dp, cp, len); |
c6d46f5f JB |
3878 | dp[len] = '\0'; |
3879 | return dp; | |
3880 | } | |
3881 | ||
c6d46f5f JB |
3882 | /* |
3883 | * Return the ptr in sp at which the character c last | |
3884 | * appears; NULL if not found | |
3885 | * | |
b02c5fea | 3886 | * Identical to System V strrchr, included for portability. |
c6d46f5f | 3887 | */ |
c6d46f5f | 3888 | char * |
b02c5fea | 3889 | etags_strrchr (sp, c) |
c6d46f5f JB |
3890 | register char *sp, c; |
3891 | { | |
3892 | register char *r; | |
3893 | ||
3894 | r = NULL; | |
3895 | do | |
3896 | { | |
3897 | if (*sp == c) | |
3898 | r = sp; | |
3899 | } while (*sp++); | |
b9755a12 | 3900 | return r; |
c6d46f5f JB |
3901 | } |
3902 | ||
9d7ad1b3 | 3903 | |
c6d46f5f JB |
3904 | /* |
3905 | * Return the ptr in sp at which the character c first | |
3906 | * appears; NULL if not found | |
3907 | * | |
b02c5fea | 3908 | * Identical to System V strchr, included for portability. |
c6d46f5f | 3909 | */ |
c6d46f5f | 3910 | char * |
b02c5fea | 3911 | etags_strchr (sp, c) |
c6d46f5f JB |
3912 | register char *sp, c; |
3913 | { | |
3914 | do | |
3915 | { | |
3916 | if (*sp == c) | |
b9755a12 FP |
3917 | return sp; |
3918 | } while (*sp++); | |
3919 | return NULL; | |
c6d46f5f JB |
3920 | } |
3921 | ||
c6d46f5f | 3922 | /* Print error message and exit. */ |
c6d46f5f JB |
3923 | void |
3924 | fatal (s1, s2) | |
3925 | char *s1, *s2; | |
3926 | { | |
3927 | error (s1, s2); | |
1a0d8c80 | 3928 | exit (BAD); |
c6d46f5f JB |
3929 | } |
3930 | ||
cdc1f6a7 FP |
3931 | void |
3932 | pfatal (s1) | |
3933 | char *s1; | |
3934 | { | |
3935 | perror (s1); | |
3936 | exit (BAD); | |
3937 | } | |
3938 | ||
d8913c1c FP |
3939 | void |
3940 | suggest_asking_for_help () | |
3941 | { | |
3942 | fprintf (stderr, "\tTry `%s --help' for a complete list of options.\n", | |
3943 | progname); | |
3944 | exit (BAD); | |
3945 | } | |
3946 | ||
c6d46f5f | 3947 | /* Print error message. `s1' is printf control string, `s2' is arg for it. */ |
c6d46f5f JB |
3948 | void |
3949 | error (s1, s2) | |
3950 | char *s1, *s2; | |
3951 | { | |
3952 | fprintf (stderr, "%s: ", progname); | |
3953 | fprintf (stderr, s1, s2); | |
3954 | fprintf (stderr, "\n"); | |
3955 | } | |
3956 | ||
46c145db FP |
3957 | /* Return a newly-allocated string whose contents |
3958 | concatenate those of s1, s2, s3. */ | |
c6d46f5f JB |
3959 | char * |
3960 | concat (s1, s2, s3) | |
3961 | char *s1, *s2, *s3; | |
3962 | { | |
3963 | int len1 = strlen (s1), len2 = strlen (s2), len3 = strlen (s3); | |
3964 | char *result = xnew (len1 + len2 + len3 + 1, char); | |
3965 | ||
1a0d8c80 FP |
3966 | strcpy (result, s1); |
3967 | strcpy (result + len1, s2); | |
3968 | strcpy (result + len1 + len2, s3); | |
46c145db | 3969 | result[len1 + len2 + len3] = '\0'; |
c6d46f5f JB |
3970 | |
3971 | return result; | |
3972 | } | |
b02c5fea | 3973 | \f |
cdc1f6a7 | 3974 | /* Does the same work as the system V getcwd, but does not need to |
c5007f46 | 3975 | guess the buffer size in advance. */ |
88f125fc RS |
3976 | char * |
3977 | etags_getcwd () | |
915e30c8 | 3978 | { |
0ff74d81 | 3979 | #ifdef DOS_NT |
cdc1f6a7 FP |
3980 | char *p, path[MAXPATHLEN + 1]; /* Fixed size is safe on MSDOS. */ |
3981 | ||
3982 | getwd (path); | |
3983 | p = path; | |
88f125fc RS |
3984 | while (*p) |
3985 | if (*p == '\\') | |
3986 | *p++ = '/'; | |
3987 | else | |
c5007f46 | 3988 | *p++ = lowcase (*p); |
cdc1f6a7 FP |
3989 | |
3990 | return strdup (path); | |
0ff74d81 FP |
3991 | #else /* not DOS_NT */ |
3992 | #if HAVE_GETCWD | |
cdc1f6a7 FP |
3993 | int bufsize = 200; |
3994 | char *path = xnew (bufsize, char); | |
c6d46f5f | 3995 | |
cdc1f6a7 | 3996 | while (getcwd (path, bufsize) == NULL) |
b02c5fea | 3997 | { |
dcc89e63 | 3998 | if (errno != ERANGE) |
0f394065 | 3999 | pfatal ("getcwd"); |
5e9c8296 | 4000 | bufsize *= 2; |
cdc1f6a7 | 4001 | path = xnew (bufsize, char); |
5e9c8296 | 4002 | } |
b02c5fea | 4003 | |
cdc1f6a7 | 4004 | return path; |
cdc1f6a7 | 4005 | #else /* not DOS_NT and not HAVE_GETCWD */ |
cdc1f6a7 FP |
4006 | struct linebuffer path; |
4007 | FILE *pipe; | |
b02c5fea | 4008 | |
cdc1f6a7 | 4009 | initbuffer (&path); |
915e30c8 | 4010 | pipe = (FILE *) popen ("pwd 2>/dev/null", "r"); |
cdc1f6a7 | 4011 | if (pipe == NULL || readline_internal (&path, pipe) == 0) |
915e30c8 | 4012 | pfatal ("pwd"); |
cdc1f6a7 | 4013 | pclose (pipe); |
b02c5fea | 4014 | |
cdc1f6a7 | 4015 | return path.buffer; |
0ff74d81 FP |
4016 | #endif /* not HAVE_GETCWD */ |
4017 | #endif /* not DOS_NT */ | |
b02c5fea FP |
4018 | } |
4019 | ||
4020 | /* Return a newly allocated string containing the filename | |
4021 | of FILE relative to the absolute directory DIR (which | |
4022 | should end with a slash). */ | |
46c145db FP |
4023 | char * |
4024 | relative_filename (file, dir) | |
4025 | char *file, *dir; | |
4026 | { | |
108c932a | 4027 | char *fp, *dp, *abs, *res; |
46c145db FP |
4028 | |
4029 | /* Find the common root of file and dir. */ | |
108c932a FP |
4030 | abs = absolute_filename (file, cwd); |
4031 | fp = abs; | |
46c145db FP |
4032 | dp = dir; |
4033 | while (*fp++ == *dp++) | |
4034 | continue; | |
4035 | do | |
4036 | { | |
4037 | fp--; | |
4038 | dp--; | |
4039 | } | |
4040 | while (*fp != '/'); | |
4041 | ||
4042 | /* Build a sequence of "../" strings for the resulting relative filename. */ | |
b02c5fea | 4043 | for (dp = etags_strchr (dp + 1, '/'), res = ""; |
46c145db | 4044 | dp != NULL; |
b02c5fea | 4045 | dp = etags_strchr (dp + 1, '/')) |
46c145db FP |
4046 | { |
4047 | res = concat (res, "../", ""); | |
4048 | } | |
4049 | ||
4050 | /* Add the filename relative to the common root of file and dir. */ | |
4051 | res = concat (res, fp + 1, ""); | |
108c932a | 4052 | free (abs); |
46c145db | 4053 | |
108c932a | 4054 | return res; |
46c145db FP |
4055 | } |
4056 | ||
4057 | /* Return a newly allocated string containing the | |
b02c5fea FP |
4058 | absolute filename of FILE given CWD (which should |
4059 | end with a slash). */ | |
46c145db FP |
4060 | char * |
4061 | absolute_filename (file, cwd) | |
4062 | char *file, *cwd; | |
4063 | { | |
4064 | char *slashp, *cp, *res; | |
4065 | ||
b2db879b | 4066 | if (absolutefn (file)) |
46c145db FP |
4067 | res = concat (file, "", ""); |
4068 | else | |
4069 | res = concat (cwd, file, ""); | |
4070 | ||
4071 | /* Delete the "/dirname/.." and "/." substrings. */ | |
b02c5fea | 4072 | slashp = etags_strchr (res, '/'); |
46c145db FP |
4073 | while (slashp != NULL && slashp[0] != '\0') |
4074 | { | |
4075 | if (slashp[1] == '.') | |
4076 | { | |
4077 | if (slashp[2] == '.' | |
4078 | && (slashp[3] == '/' || slashp[3] == '\0')) | |
4079 | { | |
4080 | cp = slashp; | |
4081 | do | |
4082 | cp--; | |
4083 | while (cp >= res && *cp != '/'); | |
4084 | if (*cp == '/') | |
4085 | { | |
4086 | strcpy (cp, slashp + 3); | |
4087 | } | |
4088 | else /* else (cp == res) */ | |
4089 | { | |
1875d994 | 4090 | if (slashp[3] != '\0') |
46c145db FP |
4091 | strcpy (cp, slashp + 4); |
4092 | else | |
4093 | return "."; | |
4094 | } | |
4095 | slashp = cp; | |
e9b2b94c | 4096 | continue; |
46c145db FP |
4097 | } |
4098 | else if (slashp[2] == '/' || slashp[2] == '\0') | |
4099 | { | |
4100 | strcpy (slashp, slashp + 2); | |
e9b2b94c | 4101 | continue; |
46c145db FP |
4102 | } |
4103 | } | |
e9b2b94c FP |
4104 | |
4105 | slashp = etags_strchr (slashp + 1, '/'); | |
46c145db FP |
4106 | } |
4107 | ||
4108 | return res; | |
4109 | } | |
4110 | ||
b02c5fea FP |
4111 | /* Return a newly allocated string containing the absolute |
4112 | filename of dir where FILE resides given CWD (which should | |
4113 | end with a slash). */ | |
46c145db FP |
4114 | char * |
4115 | absolute_dirname (file, cwd) | |
4116 | char *file, *cwd; | |
4117 | { | |
4118 | char *slashp, *res; | |
4119 | char save; | |
4120 | ||
b02c5fea | 4121 | slashp = etags_strrchr (file, '/'); |
46c145db FP |
4122 | if (slashp == NULL) |
4123 | return cwd; | |
4124 | save = slashp[1]; | |
4125 | slashp[1] = '\0'; | |
4126 | res = absolute_filename (file, cwd); | |
4127 | slashp[1] = save; | |
4128 | ||
4129 | return res; | |
4130 | } | |
4131 | ||
c6d46f5f | 4132 | /* Like malloc but get fatal error if memory is exhausted. */ |
03cdafdf | 4133 | long * |
c6d46f5f | 4134 | xmalloc (size) |
42680d3c | 4135 | unsigned int size; |
c6d46f5f | 4136 | { |
03cdafdf | 4137 | long *result = (long *) malloc (size); |
1a0d8c80 | 4138 | if (result == NULL) |
c6d46f5f JB |
4139 | fatal ("virtual memory exhausted", 0); |
4140 | return result; | |
4141 | } | |
4142 | ||
03cdafdf | 4143 | long * |
c6d46f5f JB |
4144 | xrealloc (ptr, size) |
4145 | char *ptr; | |
42680d3c | 4146 | unsigned int size; |
c6d46f5f | 4147 | { |
108c932a | 4148 | long *result = (long *) realloc (ptr, size); |
1a0d8c80 | 4149 | if (result == NULL) |
c6d46f5f JB |
4150 | fatal ("virtual memory exhausted"); |
4151 | return result; | |
4152 | } |