1 | #define COPYRIGHT_NOTICE "Copyright (C) 2003 Free Software Foundation, Inc."
|
---|
2 | #define BUG_ADDRESS "bonzini@gnu.org"
|
---|
3 |
|
---|
4 | /* GNU SED, a batch stream editor.
|
---|
5 | Copyright (C) 1989,90,91,92,93,94,95,98,99,2002,2003
|
---|
6 | Free Software Foundation, Inc.
|
---|
7 |
|
---|
8 | This program is free software; you can redistribute it and/or modify
|
---|
9 | it under the terms of the GNU General Public License as published by
|
---|
10 | the Free Software Foundation; either version 2, or (at your option)
|
---|
11 | any later version.
|
---|
12 |
|
---|
13 | This program is distributed in the hope that it will be useful,
|
---|
14 | but WITHOUT ANY WARRANTY; without even the implied warranty of
|
---|
15 | MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
---|
16 | GNU General Public License for more details.
|
---|
17 |
|
---|
18 | You should have received a copy of the GNU General Public License
|
---|
19 | along with this program; if not, write to the Free Software
|
---|
20 | Foundation, 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */
|
---|
21 |
|
---|
22 |
|
---|
23 | #include "sed.h"
|
---|
24 |
|
---|
25 |
|
---|
26 | #include <stdio.h>
|
---|
27 | #ifdef HAVE_STRINGS_H
|
---|
28 | # include <strings.h>
|
---|
29 | #else
|
---|
30 | # include <string.h>
|
---|
31 | #endif /*HAVE_STRINGS_H*/
|
---|
32 | #ifdef HAVE_MEMORY_H
|
---|
33 | # include <memory.h>
|
---|
34 | #endif
|
---|
35 |
|
---|
36 | #ifndef HAVE_STRCHR
|
---|
37 | # define strchr index
|
---|
38 | # define strrchr rindex
|
---|
39 | #endif
|
---|
40 |
|
---|
41 | #ifdef HAVE_STDLIB_H
|
---|
42 | # include <stdlib.h>
|
---|
43 | #endif
|
---|
44 |
|
---|
45 | #ifdef HAVE_SYS_TYPES_H
|
---|
46 | # include <sys/types.h>
|
---|
47 | #endif
|
---|
48 | #include "getopt.h"
|
---|
49 |
|
---|
50 | #ifndef BOOTSTRAP
|
---|
51 | #ifndef HAVE_STDLIB_H
|
---|
52 | extern char *getenv P_((const char *));
|
---|
53 | #endif
|
---|
54 | #endif
|
---|
55 |
|
---|
56 | #ifndef HAVE_STRTOUL
|
---|
57 | # define ATOI(x) atoi(x)
|
---|
58 | #else
|
---|
59 | # define ATOI(x) strtoul(x, NULL, 0)
|
---|
60 | #endif
|
---|
61 |
|
---|
62 | int extended_regexp_flags = 0;
|
---|
63 |
|
---|
64 | #ifndef CONFIG_WITHOUT_O_OPT
|
---|
65 | /* The output file, defaults to stdout but can be overridden
|
---|
66 | by the -o or --output option. main sets this to avoid problems. */
|
---|
67 | FILE *sed_stdout = NULL;
|
---|
68 | #endif
|
---|
69 |
|
---|
70 | /* If set, fflush(stdout) on every line output. */
|
---|
71 | bool unbuffered_output = false;
|
---|
72 |
|
---|
73 | /* If set, don't write out the line unless explicitly told to */
|
---|
74 | bool no_default_output = false;
|
---|
75 |
|
---|
76 | /* If set, reset line counts on every new file. */
|
---|
77 | bool separate_files = false;
|
---|
78 |
|
---|
79 | /* How do we edit files in-place? (we don't if NULL) */
|
---|
80 | char *in_place_extension = NULL;
|
---|
81 |
|
---|
82 | /* Do we need to be pedantically POSIX compliant? */
|
---|
83 | enum posixicity_types posixicity;
|
---|
84 |
|
---|
85 | /* How long should the `l' command's output line be? */
|
---|
86 | countT lcmd_out_line_len = 70;
|
---|
87 |
|
---|
88 | /* The complete compiled SED program that we are going to run: */
|
---|
89 | static struct vector *the_program = NULL;
|
---|
90 |
|
---|
91 | static void usage P_((int));
|
---|
92 | static void
|
---|
93 | usage(status)
|
---|
94 | int status;
|
---|
95 | {
|
---|
96 | FILE *out = status ? stderr : stdout;
|
---|
97 |
|
---|
98 | #ifdef REG_PERL
|
---|
99 | #define PERL_HELP _(" -R, --regexp-perl\n use Perl 5's regular expressions syntax in the script.\n")
|
---|
100 | #else
|
---|
101 | #define PERL_HELP ""
|
---|
102 | #endif
|
---|
103 |
|
---|
104 | fprintf(out, _("\
|
---|
105 | Usage: %s [OPTION]... {script-only-if-no-other-script} [input-file]...\n\
|
---|
106 | \n"), myname);
|
---|
107 |
|
---|
108 | fprintf(out, _(" -n, --quiet, --silent\n\
|
---|
109 | suppress automatic printing of pattern space\n"));
|
---|
110 | fprintf(out, _(" -e script, --expression=script\n\
|
---|
111 | add the script to the commands to be executed\n"));
|
---|
112 | fprintf(out, _(" -f script-file, --file=script-file\n\
|
---|
113 | add the contents of script-file to the commands to be executed\n"));
|
---|
114 | fprintf(out, _(" -i[SUFFIX], --in-place[=SUFFIX]\n\
|
---|
115 | edit files in place (makes backup if extension supplied)\n"));
|
---|
116 | fprintf(out, _(" -l N, --line-length=N\n\
|
---|
117 | specify the desired line-wrap length for the `l' command\n"));
|
---|
118 | #ifndef CONFIG_WITHOUT_O_LANG_C
|
---|
119 | fprintf(out, _(" --codepage=N\n\
|
---|
120 | switches the locale to the given codepage, affecting how\n\
|
---|
121 | input files are treated and outputted\n\
|
---|
122 | windows only, ignored elsewhere\n"));
|
---|
123 | # if _MSC_VER >= 1900
|
---|
124 | fprintf(out, _(" --utf8\n\
|
---|
125 | alias for --codepage=.UTF-8\n"));
|
---|
126 | # endif
|
---|
127 | fprintf(out, _(" --lang_c\n\
|
---|
128 | specify C locale\n"));
|
---|
129 | #endif
|
---|
130 | fprintf(out, _(" --posix\n\
|
---|
131 | disable all GNU extensions.\n"));
|
---|
132 | #ifndef CONFIG_WITHOUT_O_OPT
|
---|
133 | fprintf(out, _(" -o, --output=file, --append=file, --output-text=file,\n"));
|
---|
134 | fprintf(out, _(" --output-binary=file, --append-text=file, append-binary=file\n\
|
---|
135 | use the specified file instead of stdout; the first\n\
|
---|
136 | three uses the default text/binary mode.\n"));
|
---|
137 | #endif
|
---|
138 | fprintf(out, _(" -r, --regexp-extended\n\
|
---|
139 | use extended regular expressions in the script.\n"));
|
---|
140 | fprintf(out, "%s", PERL_HELP);
|
---|
141 | fprintf(out, _(" -s, --separate\n\
|
---|
142 | consider files as separate rather than as a single continuous\n\
|
---|
143 | long stream.\n"));
|
---|
144 | fprintf(out, _(" -u, --unbuffered\n\
|
---|
145 | load minimal amounts of data from the input files and flush\n\
|
---|
146 | the output buffers more often\n"));
|
---|
147 | fprintf(out, _(" --help display this help and exit\n"));
|
---|
148 | fprintf(out, _(" --version output version information and exit\n"));
|
---|
149 | fprintf(out, _("\n\
|
---|
150 | If no -e, --expression, -f, or --file option is given, then the first\n\
|
---|
151 | non-option argument is taken as the sed script to interpret. All\n\
|
---|
152 | remaining arguments are names of input files; if no input files are\n\
|
---|
153 | specified, then the standard input is read.\n\
|
---|
154 | \n"));
|
---|
155 | fprintf(out, _("E-mail bug reports to: %s .\n\
|
---|
156 | Be sure to include the word ``%s'' somewhere in the ``Subject:'' field.\n"),
|
---|
157 | BUG_ADDRESS, PACKAGE);
|
---|
158 |
|
---|
159 | ck_fclose (NULL);
|
---|
160 | exit (status);
|
---|
161 | }
|
---|
162 |
|
---|
163 | int
|
---|
164 | main(argc, argv)
|
---|
165 | int argc;
|
---|
166 | char **argv;
|
---|
167 | {
|
---|
168 | #ifdef REG_PERL
|
---|
169 | # ifndef CONFIG_WITHOUT_O_OPT
|
---|
170 | #define SHORTOPTS "snrRue:f:l:i::o:V:"
|
---|
171 | # else
|
---|
172 | #define SHORTOPTS "snrRue:f:l:i::V:"
|
---|
173 | # endif
|
---|
174 | #else
|
---|
175 | # ifndef CONFIG_WITHOUT_O_OPT
|
---|
176 | #define SHORTOPTS "snrue:f:l:i::o:V:"
|
---|
177 | # else
|
---|
178 | #define SHORTOPTS "snrue:f:l:i::V:"
|
---|
179 | # endif
|
---|
180 | #endif
|
---|
181 |
|
---|
182 | static struct option longopts[] = {
|
---|
183 | {"regexp-extended", 0, NULL, 'r'},
|
---|
184 | #ifdef REG_PERL
|
---|
185 | {"regexp-perl", 0, NULL, 'R'},
|
---|
186 | #endif
|
---|
187 | {"expression", 1, NULL, 'e'},
|
---|
188 | {"file", 1, NULL, 'f'},
|
---|
189 | {"in-place", 2, NULL, 'i'},
|
---|
190 | {"line-length", 1, NULL, 'l'},
|
---|
191 | #ifndef CONFIG_WITHOUT_O_LANG_C
|
---|
192 | {"lang_c", 0, NULL, 'L'},
|
---|
193 | {"codepage", 1, NULL, 305},
|
---|
194 | {"utf8", 0, NULL, 306},
|
---|
195 | #endif
|
---|
196 | {"quiet", 0, NULL, 'n'},
|
---|
197 | {"posix", 0, NULL, 'p'},
|
---|
198 | {"silent", 0, NULL, 'n'},
|
---|
199 | {"separate", 0, NULL, 's'},
|
---|
200 | {"unbuffered", 0, NULL, 'u'},
|
---|
201 | #ifndef CONFIG_WITHOUT_O_OPT
|
---|
202 | {"output", 1, NULL, 'o'},
|
---|
203 | {"output-binary", 1, NULL, 300},
|
---|
204 | {"output-text", 1, NULL, 301},
|
---|
205 | {"append", 1, NULL, 302},
|
---|
206 | {"append-binary", 1, NULL, 303},
|
---|
207 | {"append-text", 1, NULL, 304},
|
---|
208 | #endif
|
---|
209 | {"version", 0, NULL, 'v'},
|
---|
210 | {"help", 0, NULL, 'h'},
|
---|
211 | {NULL, 0, NULL, 0}
|
---|
212 | };
|
---|
213 |
|
---|
214 | int opt;
|
---|
215 | int return_code;
|
---|
216 | const char *cols = getenv("COLS");
|
---|
217 | #ifdef KBUILD_OS_WINDOWS
|
---|
218 | const char *locale;
|
---|
219 | #endif
|
---|
220 |
|
---|
221 | initialize_main (&argc, &argv);
|
---|
222 | #ifndef CONFIG_WITHOUT_O_OPT
|
---|
223 | sed_stdout = stdout;
|
---|
224 | #endif
|
---|
225 | #if HAVE_SETLOCALE
|
---|
226 | # ifdef KBUILD_OS_WINDOWS
|
---|
227 | locale = setlocale (LC_ALL, "");
|
---|
228 | if (getenv("KMK_SED_CODEPAGE_DEBUG"))
|
---|
229 | fprintf (stderr, "kmk_sed: codepage=%u locale=%s ACP=%u\n",
|
---|
230 | get_crt_codepage(), locale, get_ansi_codepage());
|
---|
231 | # else
|
---|
232 | setlocale (LC_ALL, "");
|
---|
233 | # endif
|
---|
234 | #endif
|
---|
235 | initialize_mbcs ();
|
---|
236 |
|
---|
237 | #if ENABLE_NLS
|
---|
238 |
|
---|
239 | /* Tell program which translations to use and where to find. */
|
---|
240 | bindtextdomain (PACKAGE, LOCALEDIR);
|
---|
241 | textdomain (PACKAGE);
|
---|
242 | #endif
|
---|
243 |
|
---|
244 | if (getenv("POSIXLY_CORRECT") != NULL)
|
---|
245 | posixicity = POSIXLY_CORRECT;
|
---|
246 | else
|
---|
247 | posixicity = POSIXLY_EXTENDED;
|
---|
248 |
|
---|
249 | /* If environment variable `COLS' is set, use its value for
|
---|
250 | the baseline setting of `lcmd_out_line_len'. The "-1"
|
---|
251 | is to avoid gratuitous auto-line-wrap on ttys.
|
---|
252 | */
|
---|
253 | if (cols)
|
---|
254 | {
|
---|
255 | countT t = ATOI(cols);
|
---|
256 | if (t > 1)
|
---|
257 | lcmd_out_line_len = t-1;
|
---|
258 | }
|
---|
259 |
|
---|
260 | myname = *argv;
|
---|
261 | while ((opt = getopt_long(argc, argv, SHORTOPTS, longopts, NULL)) != EOF)
|
---|
262 | {
|
---|
263 | switch (opt)
|
---|
264 | {
|
---|
265 | case 'n':
|
---|
266 | no_default_output = true;
|
---|
267 | break;
|
---|
268 | case 'e':
|
---|
269 | the_program = compile_string(the_program, optarg, strlen(optarg));
|
---|
270 | break;
|
---|
271 | case 'f':
|
---|
272 | the_program = compile_file(the_program, optarg);
|
---|
273 | break;
|
---|
274 |
|
---|
275 | case 'i':
|
---|
276 | separate_files = true;
|
---|
277 | if (optarg == NULL)
|
---|
278 | /* use no backups */
|
---|
279 | in_place_extension = ck_strdup ("*");
|
---|
280 |
|
---|
281 | else if (strchr(optarg, '*') != NULL)
|
---|
282 | in_place_extension = ck_strdup(optarg);
|
---|
283 |
|
---|
284 | else
|
---|
285 | {
|
---|
286 | in_place_extension = MALLOC (strlen(optarg) + 2, char);
|
---|
287 | in_place_extension[0] = '*';
|
---|
288 | strcpy (in_place_extension + 1, optarg);
|
---|
289 | }
|
---|
290 |
|
---|
291 | break;
|
---|
292 |
|
---|
293 | case 'l':
|
---|
294 | lcmd_out_line_len = ATOI(optarg);
|
---|
295 | break;
|
---|
296 |
|
---|
297 | #ifndef CONFIG_WITHOUT_O_LANG_C
|
---|
298 | case 'L':
|
---|
299 | # ifdef KBUILD_OS_WINDOWS
|
---|
300 | locale = setlocale (LC_ALL, "C");
|
---|
301 | if (getenv("KMK_SED_CODEPAGE_DEBUG"))
|
---|
302 | fprintf (stderr, "kmk_sed: codepage=%u locale=%s ACP=%u\n",
|
---|
303 | get_crt_codepage(), locale, get_ansi_codepage());
|
---|
304 | # else
|
---|
305 | setlocale (LC_ALL, "C");
|
---|
306 | # endif
|
---|
307 | initialize_mbcs ();
|
---|
308 | # if ENABLE_NLS
|
---|
309 | bindtextdomain (PACKAGE, LOCALEDIR);
|
---|
310 | textdomain (PACKAGE);
|
---|
311 | # endif
|
---|
312 | break;
|
---|
313 |
|
---|
314 | case 306: /* --codepage=N */
|
---|
315 | # if _MSC_VER < 1900
|
---|
316 | break; /* does not work, so ignore */
|
---|
317 | # else
|
---|
318 | optarg = ".UTF-8";
|
---|
319 | # endif
|
---|
320 | /* fall through */
|
---|
321 | case 305: /* --codepage=N */
|
---|
322 | {
|
---|
323 | # ifdef KBUILD_OS_WINDOWS
|
---|
324 | char szTmp[64];
|
---|
325 | if (optarg[0] != '.')
|
---|
326 | optarg = strncat (strcpy (szTmp, "."), optarg, sizeof (szTmp) - 1);
|
---|
327 | locale = setlocale (LC_ALL, optarg);
|
---|
328 | if (locale == NULL)
|
---|
329 | fprintf (stderr,
|
---|
330 | _("%s: warning: setlocale (LC_ALL, \"%s\") failed: %d\n"),
|
---|
331 | myname, optarg, errno);
|
---|
332 | else if (getenv("KMK_SED_CODEPAGE_DEBUG"))
|
---|
333 | fprintf (stderr, "kmk_sed: codepage=%u locale=%s ACP=%u\n",
|
---|
334 | get_crt_codepage(), locale, get_ansi_codepage());
|
---|
335 | initialize_mbcs();
|
---|
336 | # if ENABLE_NLS
|
---|
337 | bindtextdomain (PACKAGE, LOCALEDIR);
|
---|
338 | textdomain (PACKAGE);
|
---|
339 | # endif
|
---|
340 | # endif
|
---|
341 | break;
|
---|
342 | }
|
---|
343 | #endif
|
---|
344 |
|
---|
345 | #ifndef CONFIG_WITHOUT_O_OPT
|
---|
346 | case 'o':
|
---|
347 | sed_stdout = ck_fopen (optarg, "w", true /* fail on error */);
|
---|
348 | break;
|
---|
349 |
|
---|
350 | case 300:
|
---|
351 | sed_stdout = ck_fopen (optarg, "wb", true /* fail on error */);
|
---|
352 | break;
|
---|
353 |
|
---|
354 | case 301:
|
---|
355 | sed_stdout = ck_fopen (optarg, "wt", true /* fail on error */);
|
---|
356 | break;
|
---|
357 |
|
---|
358 | case 302:
|
---|
359 | sed_stdout = ck_fopen (optarg, "a", true /* fail on error */);
|
---|
360 | break;
|
---|
361 |
|
---|
362 | case 303:
|
---|
363 | sed_stdout = ck_fopen (optarg, "ab", true /* fail on error */);
|
---|
364 | break;
|
---|
365 |
|
---|
366 | case 304:
|
---|
367 | sed_stdout = ck_fopen (optarg, "at", true /* fail on error */);
|
---|
368 | break;
|
---|
369 | #endif
|
---|
370 |
|
---|
371 | case 'p':
|
---|
372 | posixicity = POSIXLY_BASIC;
|
---|
373 | break;
|
---|
374 |
|
---|
375 | case 'r':
|
---|
376 | if (extended_regexp_flags)
|
---|
377 | usage(4);
|
---|
378 | extended_regexp_flags = REG_EXTENDED;
|
---|
379 | break;
|
---|
380 |
|
---|
381 | #ifdef REG_PERL
|
---|
382 | case 'R':
|
---|
383 | if (extended_regexp_flags)
|
---|
384 | usage(4);
|
---|
385 | extended_regexp_flags = REG_PERL;
|
---|
386 | break;
|
---|
387 | #endif
|
---|
388 |
|
---|
389 | case 's':
|
---|
390 | separate_files = true;
|
---|
391 | break;
|
---|
392 |
|
---|
393 | case 'u':
|
---|
394 | unbuffered_output = true;
|
---|
395 | break;
|
---|
396 |
|
---|
397 | case 'v':
|
---|
398 | #ifdef KBUILD_VERSION_MAJOR
|
---|
399 | fprintf(stdout, _("kmk_sed - kBuild version %d.%d.%d\n"
|
---|
400 | "\n"
|
---|
401 | "Based on "),
|
---|
402 | KBUILD_VERSION_MAJOR, KBUILD_VERSION_MINOR, KBUILD_VERSION_PATCH);
|
---|
403 | #endif
|
---|
404 | #ifdef REG_PERL
|
---|
405 | fprintf(stdout, _("super-sed version %s\n"), VERSION);
|
---|
406 | fprintf(stdout, _("based on GNU sed version %s\n\n"), SED_FEATURE_VERSION);
|
---|
407 | #else
|
---|
408 | fprintf(stdout, _("GNU sed version %s\n"), VERSION);
|
---|
409 | #endif
|
---|
410 | fprintf(stdout, _("%s\n\
|
---|
411 | This is free software; see the source for copying conditions. There is NO\n\
|
---|
412 | warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE,\n\
|
---|
413 | to the extent permitted by law.\n\
|
---|
414 | "), COPYRIGHT_NOTICE);
|
---|
415 |
|
---|
416 | ck_fclose (NULL);
|
---|
417 | exit (0);
|
---|
418 | case 'h':
|
---|
419 | usage(0);
|
---|
420 | break;
|
---|
421 | default:
|
---|
422 | usage(4);
|
---|
423 | }
|
---|
424 | }
|
---|
425 |
|
---|
426 | if (!the_program)
|
---|
427 | {
|
---|
428 | if (optind < argc)
|
---|
429 | {
|
---|
430 | char *arg = argv[optind++];
|
---|
431 | the_program = compile_string(the_program, arg, strlen(arg));
|
---|
432 | }
|
---|
433 | else
|
---|
434 | usage(4);
|
---|
435 | }
|
---|
436 | check_final_program(the_program);
|
---|
437 |
|
---|
438 | return_code = process_files(the_program, argv+optind);
|
---|
439 |
|
---|
440 | finish_program(the_program);
|
---|
441 | ck_fclose(NULL);
|
---|
442 |
|
---|
443 | return return_code;
|
---|
444 | }
|
---|
445 |
|
---|
446 | #ifdef __HAIKU__ /* mbrtowc is busted, just stub it and pray the input won't ever acutally be multibyte... */
|
---|
447 | size_t mbrtowc(wchar_t *pwc, const char *pch, size_t n, mbstate_t *ps)
|
---|
448 | {
|
---|
449 | if (!n)
|
---|
450 | return 0;
|
---|
451 | if (pwc)
|
---|
452 | *pwc = *pch;
|
---|
453 | return 1;
|
---|
454 | }
|
---|
455 | #endif
|
---|