Line data Source code
1 : /* Copyright (C) 1989-2025 Free Software Foundation, Inc.
2 : Written by James Clark (jjc@jclark.com)
3 :
4 : This file is part of groff, the GNU roff typesetting system.
5 :
6 : groff is free software; you can redistribute it and/or modify it under
7 : the terms of the GNU General Public License as published by the Free
8 : Software Foundation, either version 3 of the License, or
9 : (at your option) any later version.
10 :
11 : groff is distributed in the hope that it will be useful, but WITHOUT ANY
12 : WARRANTY; without even the implied warranty of MERCHANTABILITY or
13 : FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
14 : for more details.
15 :
16 : You should have received a copy of the GNU General Public License
17 : along with this program. If not, see <http://www.gnu.org/licenses/>. */
18 :
19 : #ifdef HAVE_CONFIG_H
20 : #include <config.h>
21 : #endif
22 :
23 : #include <assert.h>
24 : #include <ctype.h>
25 : #include <errno.h>
26 : #include <stdlib.h>
27 :
28 : #include <getopt.h> // getopt_long()
29 :
30 : #include "lib.h"
31 :
32 : #include "errarg.h"
33 : #include "error.h"
34 : #include "stringclass.h"
35 : #include "nonposix.h"
36 : #include "searchpath.h"
37 : #include "lf.h"
38 :
39 : // Initialize inclusion search path with only the current directory.
40 : static search_path include_search_path(0 /* nullptr */, 0 /* nullptr */,
41 : 0, 1);
42 :
43 : bool want_att_compat = false;
44 : bool want_raw_output = false;
45 : bool want_tex_output = false;
46 :
47 : extern "C" const char *Version_string;
48 :
49 : // forward declaration
50 : static bool do_file(const char *);
51 :
52 0 : void usage(FILE *stream)
53 : {
54 0 : fprintf(stream, "usage: %s [-Crt] [-I dir] [input-file ...]\n"
55 : "usage: %s {-v | --version}\n"
56 : "usage: %s --help\n",
57 : program_name, program_name, program_name);
58 0 : if (stdout == stream)
59 0 : fputs("\n"
60 : "GNU soelim eliminates source requests in roff(7) and other text\n"
61 : "files; it replaces lines of the form \".so includedâfile\" within\n"
62 : "each text input-file with the contents of included-file recursively,\n"
63 : "flattening a tree of documents. By default, it writes roff \"lf\"\n"
64 : "requests as well to record the name and line number of each\n"
65 : "input-file and included-file. Use the -t option to produce TeX\n"
66 : "comments instead of roff requests. Use the -r option to write\n"
67 : "neither. See the soelim(1) manual page.\n",
68 : stream);
69 0 : }
70 :
71 77 : int main(int argc, char **argv)
72 : {
73 77 : program_name = argv[0];
74 : int opt;
75 : static const struct option long_options[] = {
76 : { "help", no_argument, 0 /* nullptr */, CHAR_MAX + 1 },
77 : { "version", no_argument, 0 /* nullptr */, 'v' },
78 : { 0 /* nullptr */, 0, 0 /* nullptr */, 0 }
79 : };
80 107 : while ((opt = getopt_long(argc, argv, ":CI:rtv", long_options,
81 : 0 /* nullptr */))
82 107 : != EOF)
83 30 : switch (opt) {
84 0 : case 'v':
85 0 : printf("GNU soelim (groff) version %s\n", Version_string);
86 0 : exit(EXIT_SUCCESS);
87 : break;
88 2 : case 'C':
89 2 : want_att_compat = true;
90 2 : break;
91 27 : case 'I':
92 27 : include_search_path.command_line_dir(optarg);
93 27 : break;
94 1 : case 'r':
95 1 : want_raw_output = true;
96 1 : break;
97 0 : case 't':
98 0 : want_tex_output = true;
99 0 : break;
100 0 : case CHAR_MAX + 1: // --help
101 0 : usage(stdout);
102 0 : exit(EXIT_SUCCESS);
103 : break;
104 0 : case '?':
105 0 : if (optopt != 0)
106 0 : error("unrecognized command-line option '%1'", char(optopt));
107 : else
108 0 : error("unrecognized command-line option '%1'",
109 0 : argv[(optind - 1)]);
110 0 : usage(stderr);
111 0 : exit(2);
112 : break;
113 0 : case ':':
114 0 : error("command-line option '%1' requires an argument",
115 0 : char(optopt));
116 0 : usage(stderr);
117 0 : exit(2);
118 : break;
119 0 : default:
120 0 : assert(0 == "unhandled getopt_long return value");
121 : }
122 77 : int nbad = 0;
123 77 : if (optind >= argc)
124 63 : nbad += !do_file("-");
125 : else
126 28 : for (int i = optind; i < argc; i++)
127 14 : nbad += !do_file(argv[i]);
128 77 : if (ferror(stdout))
129 0 : fatal("error status on standard output stream");
130 77 : if (fflush(stdout) < 0)
131 0 : fatal("cannot flush standard output stream: %1", strerror(errno));
132 77 : return (nbad != 0);
133 : }
134 :
135 79 : void set_location()
136 : {
137 79 : if (!want_raw_output) {
138 76 : if (!want_tex_output)
139 76 : printf(".lf %d %s%s\n", current_lineno,
140 76 : ('"' == current_filename[0]) ? "" : "\"", current_filename);
141 : else
142 : // XXX: Should we quote the file name? What's TeX-conventional?
143 0 : printf("%% file %s, line %d\n", current_filename, current_lineno);
144 : }
145 79 : }
146 :
147 1 : void do_so(const char *line)
148 : {
149 1 : const char *p = line;
150 2 : while (*p == ' ')
151 1 : p++;
152 1 : string filename;
153 1 : bool is_filename_valid = true;
154 1 : const char *q = p;
155 1 : if ('"' == *q)
156 0 : q++;
157 14 : for (; is_filename_valid && (*q != '\0') && (*q != '\n'); q++)
158 13 : if (*q == '\\') {
159 0 : switch (*++q) {
160 0 : case 'e':
161 : case '\\':
162 0 : filename += '\\';
163 0 : break;
164 0 : case ' ':
165 0 : warning("escaping a space is unnecessary and not compatible"
166 : " with troff syntax");
167 0 : filename += ' ';
168 : // TODO: groff 1.24.0 release + 2 years?
169 : // is_filename_valid = false; // or fall through
170 0 : break;
171 0 : default:
172 0 : is_filename_valid = false;
173 0 : break;
174 : }
175 : }
176 : else
177 13 : filename += char(*q);
178 1 : if (is_filename_valid && (filename.length() > 0)) {
179 1 : filename += '\0';
180 1 : const char *fn = current_filename;
181 1 : int ln = current_lineno;
182 1 : current_lineno--;
183 1 : if (do_file(filename.contents())) {
184 1 : current_filename = fn;
185 1 : current_lineno = ln;
186 1 : set_location();
187 1 : return;
188 : }
189 0 : current_lineno++;
190 : }
191 0 : fputs(".so", stdout);
192 0 : fputs(line, stdout);
193 : }
194 :
195 78 : static bool do_file(const char *filename)
196 : {
197 78 : char *file_name_in_path = 0 /* nullptr */;
198 78 : FILE *fp = include_search_path.open_file_cautiously(filename,
199 : &file_name_in_path);
200 78 : int err = errno;
201 156 : string whole_filename(filename);
202 78 : if (strcmp(filename, "-") && file_name_in_path != 0 /* nullptr */)
203 15 : whole_filename = file_name_in_path;
204 78 : whole_filename += '\0';
205 78 : free(file_name_in_path);
206 78 : if (0 /* nullptr */ == fp) {
207 0 : error("cannot open '%1': %2", whole_filename.contents(),
208 0 : strerror(err));
209 0 : return false;
210 : }
211 78 : normalize_file_name_for_lf_request(whole_filename);
212 78 : current_filename = whole_filename.contents();
213 78 : current_lineno = 1;
214 78 : set_location();
215 78 : enum { START, MIDDLE, HAD_DOT, HAD_s, HAD_so, HAD_l, HAD_lf } state
216 : = START;
217 : for (;;) {
218 652074 : int c = getc(fp);
219 652074 : if (c == EOF)
220 78 : break;
221 651996 : switch (state) {
222 27448 : case START:
223 27448 : if (c == '.')
224 14266 : state = HAD_DOT;
225 : else {
226 13182 : putchar(c);
227 13182 : if (c == '\n') {
228 1587 : current_lineno++;
229 1587 : state = START;
230 : }
231 : else
232 11595 : state = MIDDLE;
233 : }
234 27448 : break;
235 605603 : case MIDDLE:
236 605603 : putchar(c);
237 605603 : if (c == '\n') {
238 23804 : current_lineno++;
239 23804 : state = START;
240 : }
241 605603 : break;
242 14266 : case HAD_DOT:
243 14266 : if (c == 's')
244 2917 : state = HAD_s;
245 11349 : else if (c == 'l')
246 881 : state = HAD_l;
247 : else {
248 10468 : putchar('.');
249 10468 : putchar(c);
250 10468 : if (c == '\n') {
251 1185 : current_lineno++;
252 1185 : state = START;
253 : }
254 : else
255 9283 : state = MIDDLE;
256 : }
257 14266 : break;
258 2917 : case HAD_s:
259 2917 : if (c == 'o')
260 12 : state = HAD_so;
261 : else {
262 2905 : putchar('.');
263 2905 : putchar('s');
264 2905 : putchar(c);
265 2905 : if (c == '\n') {
266 0 : current_lineno++;
267 0 : state = START;
268 : }
269 : else
270 2905 : state = MIDDLE;
271 : }
272 2917 : break;
273 12 : case HAD_so:
274 12 : if (c == ' ' || c == '\n' || want_att_compat) {
275 1 : string line;
276 15 : for (; c != EOF && c != '\n'; c = getc(fp))
277 14 : line += c;
278 1 : current_lineno++;
279 1 : line += '\n';
280 1 : line += '\0';
281 1 : do_so(line.contents());
282 1 : state = START;
283 : }
284 : else {
285 11 : fputs(".so", stdout);
286 11 : putchar(c);
287 11 : state = MIDDLE;
288 : }
289 12 : break;
290 881 : case HAD_l:
291 881 : if (c == 'f')
292 869 : state = HAD_lf;
293 : else {
294 12 : putchar('.');
295 12 : putchar('l');
296 12 : putchar(c);
297 12 : if (c == '\n') {
298 0 : current_lineno++;
299 0 : state = START;
300 : }
301 : else
302 12 : state = MIDDLE;
303 : }
304 881 : break;
305 869 : case HAD_lf:
306 869 : if (c == ' ' || c == '\n' || want_att_compat) {
307 869 : string line;
308 6201 : for (; c != EOF && c != '\n'; c = getc(fp))
309 5332 : line += c;
310 869 : current_lineno++;
311 869 : line += '\n';
312 869 : line += '\0';
313 869 : interpret_lf_request_arguments(line.contents());
314 869 : printf(".lf%s", line.contents());
315 869 : state = START;
316 : }
317 : else {
318 0 : fputs(".lf", stdout);
319 0 : putchar(c);
320 0 : state = MIDDLE;
321 : }
322 869 : break;
323 0 : default:
324 0 : assert(0 == "unhandled state in file parser");
325 : }
326 651996 : }
327 78 : switch (state) {
328 0 : case HAD_DOT:
329 0 : fputs(".\n", stdout);
330 0 : break;
331 0 : case HAD_l:
332 0 : fputs(".l\n", stdout);
333 0 : break;
334 0 : case HAD_s:
335 0 : fputs(".s\n", stdout);
336 0 : break;
337 0 : case HAD_lf:
338 0 : fputs(".lf\n", stdout);
339 0 : break;
340 0 : case HAD_so:
341 0 : fputs(".so\n", stdout);
342 0 : break;
343 2 : case MIDDLE:
344 2 : putc('\n', stdout);
345 2 : break;
346 76 : case START:
347 76 : break;
348 : }
349 78 : if (fp != stdin)
350 15 : if (fclose(fp) < 0)
351 0 : fatal("cannot close '%1': %2", whole_filename.contents(),
352 0 : strerror(errno));
353 78 : current_filename = 0 /* nullptr */;
354 78 : return true;
355 : }
356 :
357 : // Local Variables:
358 : // fill-column: 72
359 : // mode: C++
360 : // End:
361 : // vim: set cindent noexpandtab shiftwidth=2 textwidth=72:
|
