]>
Commit | Line | Data |
---|---|---|
6dbe3af9 KZ |
1 | /*- |
2 | * Copyright (c) 1990 The Regents of the University of California. | |
3 | * All rights reserved. | |
4 | * | |
5 | * This code is derived from software contributed to Berkeley by | |
6 | * Michael Rendell of the Memorial University of Newfoundland. | |
7 | * | |
8 | * Redistribution and use in source and binary forms, with or without | |
9 | * modification, are permitted provided that the following conditions | |
10 | * are met: | |
11 | * 1. Redistributions of source code must retain the above copyright | |
12 | * notice, this list of conditions and the following disclaimer. | |
13 | * 2. Redistributions in binary form must reproduce the above copyright | |
14 | * notice, this list of conditions and the following disclaimer in the | |
15 | * documentation and/or other materials provided with the distribution. | |
16 | * 3. All advertising materials mentioning features or use of this software | |
17 | * must display the following acknowledgement: | |
18 | * This product includes software developed by the University of | |
19 | * California, Berkeley and its contributors. | |
20 | * 4. Neither the name of the University nor the names of its contributors | |
21 | * may be used to endorse or promote products derived from this software | |
22 | * without specific prior written permission. | |
23 | * | |
24 | * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND | |
25 | * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE | |
26 | * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE | |
27 | * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE | |
28 | * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL | |
29 | * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS | |
30 | * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) | |
31 | * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT | |
32 | * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY | |
33 | * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF | |
34 | * SUCH DAMAGE. | |
35 | * | |
36 | * Wed Jun 22 22:15:41 1994, faith@cs.unc.edu: Added internationalization | |
37 | * patches from Andries.Brouwer@cwi.nl | |
38 | * Wed Sep 14 22:31:17 1994: patches from Carl Christofferson | |
39 | * (cchris@connected.com) | |
40 | */ | |
41 | ||
42 | #ifndef lint | |
43 | char copyright[] = | |
44 | "@(#) Copyright (c) 1990 The Regents of the University of California.\n\ | |
45 | All rights reserved.\n"; | |
46 | #endif /* not lint */ | |
47 | ||
48 | #ifndef lint | |
49 | static char sccsid[] = "@(#)col.c 5.3 (Berkeley) 2/2/91"; | |
50 | #endif /* not lint */ | |
51 | ||
fd6b7a7f KZ |
52 | #include <stdlib.h> |
53 | #include <malloc.h> | |
6dbe3af9 KZ |
54 | #include <errno.h> |
55 | #include <ctype.h> | |
56 | #include <string.h> | |
57 | #include <stdio.h> | |
fd6b7a7f | 58 | #include <unistd.h> |
6dbe3af9 KZ |
59 | #include <locale.h> |
60 | ||
61 | #define BS '\b' /* backspace */ | |
62 | #define TAB '\t' /* tab */ | |
63 | #define SPACE ' ' /* space */ | |
64 | #define NL '\n' /* newline */ | |
65 | #define CR '\r' /* carriage return */ | |
66 | #define ESC '\033' /* escape */ | |
67 | #define SI '\017' /* shift in to normal character set */ | |
68 | #define SO '\016' /* shift out to alternate character set */ | |
69 | #define VT '\013' /* vertical tab (aka reverse line feed) */ | |
70 | #define RLF '\007' /* ESC-07 reverse line feed */ | |
71 | #define RHLF '\010' /* ESC-010 reverse half-line feed */ | |
72 | #define FHLF '\011' /* ESC-011 forward half-line feed */ | |
73 | ||
74 | /* build up at least this many lines before flushing them out */ | |
75 | #define BUFFER_MARGIN 32 | |
76 | ||
77 | typedef char CSET; | |
78 | ||
79 | typedef struct char_str { | |
80 | #define CS_NORMAL 1 | |
81 | #define CS_ALTERNATE 2 | |
82 | short c_column; /* column character is in */ | |
83 | CSET c_set; /* character set (currently only 2) */ | |
84 | char c_char; /* character in question */ | |
85 | } CHAR; | |
86 | ||
87 | typedef struct line_str LINE; | |
88 | struct line_str { | |
89 | CHAR *l_line; /* characters on the line */ | |
90 | LINE *l_prev; /* previous line */ | |
91 | LINE *l_next; /* next line */ | |
92 | int l_lsize; /* allocated sizeof l_line */ | |
93 | int l_line_len; /* strlen(l_line) */ | |
94 | int l_needs_sort; /* set if chars went in out of order */ | |
95 | int l_max_col; /* max column in the line */ | |
96 | }; | |
97 | ||
fd6b7a7f KZ |
98 | void usage(void); |
99 | void wrerr(void); | |
100 | void warn(int); | |
101 | void free_line(LINE *l); | |
102 | void flush_line(LINE *l); | |
103 | void flush_lines(int); | |
104 | void flush_blanks(void); | |
105 | void *xmalloc(void *p, size_t size); | |
106 | LINE *alloc_line(void); | |
6dbe3af9 KZ |
107 | |
108 | CSET last_set; /* char_set of last char printed */ | |
109 | LINE *lines; | |
110 | int compress_spaces; /* if doing space -> tab conversion */ | |
111 | int fine; /* if `fine' resolution (half lines) */ | |
112 | int max_bufd_lines; /* max # lines to keep in memory */ | |
113 | int nblank_lines; /* # blanks after last flushed line */ | |
114 | int no_backspaces; /* if not to output any backspaces */ | |
115 | ||
116 | #define PUTC(ch) \ | |
117 | if (putchar(ch) == EOF) \ | |
118 | wrerr(); | |
119 | ||
fd6b7a7f | 120 | int main(int argc, char **argv) |
6dbe3af9 KZ |
121 | { |
122 | extern int optind; | |
123 | extern char *optarg; | |
124 | register int ch; | |
125 | CHAR *c; | |
126 | CSET cur_set; /* current character set */ | |
127 | LINE *l; /* current line */ | |
128 | int extra_lines; /* # of lines above first line */ | |
129 | int cur_col; /* current column */ | |
130 | int cur_line; /* line number of current position */ | |
131 | int max_line; /* max value of cur_line */ | |
132 | int this_line; /* line l points to */ | |
133 | int nflushd_lines; /* number of lines that were flushed */ | |
134 | int adjust, opt, warned; | |
135 | ||
136 | /* we discard characters that do not pass isgraph() */ | |
137 | setlocale(LC_CTYPE, ""); | |
138 | ||
139 | max_bufd_lines = 128; | |
140 | compress_spaces = 1; /* compress spaces into tabs */ | |
141 | while ((opt = getopt(argc, argv, "bfhl:x")) != EOF) | |
142 | switch (opt) { | |
143 | case 'b': /* do not output backspaces */ | |
144 | no_backspaces = 1; | |
145 | break; | |
146 | case 'f': /* allow half forward line feeds */ | |
147 | fine = 1; | |
148 | break; | |
149 | case 'h': /* compress spaces into tabs */ | |
150 | compress_spaces = 1; | |
151 | break; | |
152 | case 'l': /* buffered line count */ | |
153 | if ((max_bufd_lines = atoi(optarg)) <= 0) { | |
154 | (void)fprintf(stderr, | |
155 | "col: bad -l argument %s.\n", optarg); | |
156 | exit(1); | |
157 | } | |
158 | break; | |
159 | case 'x': /* do not compress spaces into tabs */ | |
160 | compress_spaces = 0; | |
161 | break; | |
162 | case '?': | |
163 | default: | |
164 | usage(); | |
165 | } | |
166 | ||
167 | if (optind != argc) | |
168 | usage(); | |
169 | ||
170 | /* this value is in half lines */ | |
171 | max_bufd_lines *= 2; | |
172 | ||
173 | adjust = cur_col = extra_lines = warned = 0; | |
174 | cur_line = max_line = nflushd_lines = this_line = 0; | |
175 | cur_set = last_set = CS_NORMAL; | |
176 | lines = l = alloc_line(); | |
177 | ||
178 | while ((ch = getchar()) != EOF) { | |
179 | if (!isgraph(ch)) { | |
180 | switch (ch) { | |
181 | case BS: /* can't go back further */ | |
182 | if (cur_col == 0) | |
183 | continue; | |
184 | --cur_col; | |
185 | continue; | |
186 | case CR: | |
187 | cur_col = 0; | |
188 | continue; | |
189 | case ESC: /* just ignore EOF */ | |
190 | switch(getchar()) { | |
191 | case RLF: | |
192 | cur_line -= 2; | |
193 | break; | |
194 | case RHLF: | |
195 | cur_line--; | |
196 | break; | |
197 | case FHLF: | |
198 | cur_line++; | |
199 | if (cur_line > max_line) | |
200 | max_line = cur_line; | |
201 | } | |
202 | continue; | |
203 | case NL: | |
204 | cur_line += 2; | |
205 | if (cur_line > max_line) | |
206 | max_line = cur_line; | |
207 | cur_col = 0; | |
208 | continue; | |
209 | case SPACE: | |
210 | ++cur_col; | |
211 | continue; | |
212 | case SI: | |
213 | cur_set = CS_NORMAL; | |
214 | continue; | |
215 | case SO: | |
216 | cur_set = CS_ALTERNATE; | |
217 | continue; | |
218 | case TAB: /* adjust column */ | |
219 | cur_col |= 7; | |
220 | ++cur_col; | |
221 | continue; | |
222 | case VT: | |
223 | cur_line -= 2; | |
224 | continue; | |
225 | } | |
226 | continue; | |
227 | } | |
228 | ||
229 | /* Must stuff ch in a line - are we at the right one? */ | |
230 | if (cur_line != this_line - adjust) { | |
231 | LINE *lnew; | |
232 | int nmove; | |
233 | ||
234 | adjust = 0; | |
235 | nmove = cur_line - this_line; | |
236 | if (!fine) { | |
237 | /* round up to next line */ | |
238 | if (cur_line & 1) { | |
239 | adjust = 1; | |
240 | nmove++; | |
241 | } | |
242 | } | |
243 | if (nmove < 0) { | |
244 | for (; nmove < 0 && l->l_prev; nmove++) | |
245 | l = l->l_prev; | |
246 | if (nmove) { | |
247 | if (nflushd_lines == 0) { | |
248 | /* | |
249 | * Allow backup past first | |
250 | * line if nothing has been | |
251 | * flushed yet. | |
252 | */ | |
253 | for (; nmove < 0; nmove++) { | |
254 | lnew = alloc_line(); | |
255 | l->l_prev = lnew; | |
256 | lnew->l_next = l; | |
257 | l = lines = lnew; | |
258 | extra_lines++; | |
259 | } | |
260 | } else { | |
261 | if (!warned++) | |
262 | warn(cur_line); | |
263 | cur_line -= nmove; | |
264 | } | |
265 | } | |
266 | } else { | |
267 | /* may need to allocate here */ | |
268 | for (; nmove > 0 && l->l_next; nmove--) | |
269 | l = l->l_next; | |
270 | for (; nmove > 0; nmove--) { | |
271 | lnew = alloc_line(); | |
272 | lnew->l_prev = l; | |
273 | l->l_next = lnew; | |
274 | l = lnew; | |
275 | } | |
276 | } | |
277 | this_line = cur_line + adjust; | |
278 | nmove = this_line - nflushd_lines; | |
279 | if (nmove >= max_bufd_lines + BUFFER_MARGIN) { | |
280 | nflushd_lines += nmove - max_bufd_lines; | |
281 | flush_lines(nmove - max_bufd_lines); | |
282 | } | |
283 | } | |
284 | /* grow line's buffer? */ | |
285 | if (l->l_line_len + 1 >= l->l_lsize) { | |
286 | int need; | |
287 | ||
288 | need = l->l_lsize ? l->l_lsize * 2 : 90; | |
289 | l->l_line = (CHAR *)xmalloc((void *) l->l_line, | |
290 | (unsigned) need * sizeof(CHAR)); | |
291 | l->l_lsize = need; | |
292 | } | |
293 | c = &l->l_line[l->l_line_len++]; | |
294 | c->c_char = ch; | |
295 | c->c_set = cur_set; | |
296 | c->c_column = cur_col; | |
297 | /* | |
298 | * If things are put in out of order, they will need sorting | |
299 | * when it is flushed. | |
300 | */ | |
301 | if (cur_col < l->l_max_col) | |
302 | l->l_needs_sort = 1; | |
303 | else | |
304 | l->l_max_col = cur_col; | |
305 | cur_col++; | |
306 | } | |
307 | /* goto the last line that had a character on it */ | |
308 | for (; l->l_next; l = l->l_next) | |
309 | this_line++; | |
310 | flush_lines(this_line - nflushd_lines + extra_lines + 1); | |
311 | ||
312 | /* make sure we leave things in a sane state */ | |
313 | if (last_set != CS_NORMAL) | |
314 | PUTC('\017'); | |
315 | ||
316 | /* flush out the last few blank lines */ | |
317 | nblank_lines = max_line - this_line; | |
318 | if (max_line & 1) | |
319 | nblank_lines++; | |
320 | else if (!nblank_lines) | |
321 | /* missing a \n on the last line? */ | |
322 | nblank_lines = 2; | |
323 | flush_blanks(); | |
fd6b7a7f | 324 | return 0; |
6dbe3af9 KZ |
325 | } |
326 | ||
fd6b7a7f | 327 | void flush_lines(int nflush) |
6dbe3af9 KZ |
328 | { |
329 | LINE *l; | |
330 | ||
331 | while (--nflush >= 0) { | |
332 | l = lines; | |
333 | lines = l->l_next; | |
334 | if (l->l_line) { | |
335 | flush_blanks(); | |
336 | flush_line(l); | |
337 | } | |
338 | nblank_lines++; | |
339 | if (l->l_line) | |
340 | (void)free((void *)l->l_line); | |
341 | free_line(l); | |
342 | } | |
343 | if (lines) | |
344 | lines->l_prev = NULL; | |
345 | } | |
346 | ||
347 | /* | |
348 | * Print a number of newline/half newlines. If fine flag is set, nblank_lines | |
349 | * is the number of half line feeds, otherwise it is the number of whole line | |
350 | * feeds. | |
351 | */ | |
fd6b7a7f | 352 | void flush_blanks() |
6dbe3af9 KZ |
353 | { |
354 | int half, i, nb; | |
355 | ||
356 | half = 0; | |
357 | nb = nblank_lines; | |
358 | if (nb & 1) { | |
359 | if (fine) | |
360 | half = 1; | |
361 | else | |
362 | nb++; | |
363 | } | |
364 | nb /= 2; | |
365 | for (i = nb; --i >= 0;) | |
366 | PUTC('\n'); | |
367 | if (half) { | |
368 | PUTC('\033'); | |
369 | PUTC('9'); | |
370 | if (!nb) | |
371 | PUTC('\r'); | |
372 | } | |
373 | nblank_lines = 0; | |
374 | } | |
375 | ||
376 | /* | |
377 | * Write a line to stdout taking care of space to tab conversion (-h flag) | |
378 | * and character set shifts. | |
379 | */ | |
fd6b7a7f | 380 | void flush_line(LINE *l) |
6dbe3af9 KZ |
381 | { |
382 | CHAR *c, *endc; | |
383 | int nchars, last_col, this_col; | |
384 | ||
385 | last_col = 0; | |
386 | nchars = l->l_line_len; | |
387 | ||
388 | if (l->l_needs_sort) { | |
389 | static CHAR *sorted; | |
390 | static int count_size, *count, i, save, sorted_size, tot; | |
391 | ||
392 | /* | |
393 | * Do an O(n) sort on l->l_line by column being careful to | |
394 | * preserve the order of characters in the same column. | |
395 | */ | |
396 | if (l->l_lsize > sorted_size) { | |
397 | sorted_size = l->l_lsize; | |
398 | sorted = (CHAR *)xmalloc((void *)sorted, | |
399 | (unsigned)sizeof(CHAR) * sorted_size); | |
400 | } | |
401 | if (l->l_max_col >= count_size) { | |
402 | count_size = l->l_max_col + 1; | |
403 | count = (int *)xmalloc((void *)count, | |
404 | (unsigned)sizeof(int) * count_size); | |
405 | } | |
406 | bzero((char *)count, sizeof(int) * l->l_max_col + 1); | |
407 | for (i = nchars, c = l->l_line; --i >= 0; c++) | |
408 | count[c->c_column]++; | |
409 | ||
410 | /* | |
411 | * calculate running total (shifted down by 1) to use as | |
412 | * indices into new line. | |
413 | */ | |
414 | for (tot = 0, i = 0; i <= l->l_max_col; i++) { | |
415 | save = count[i]; | |
416 | count[i] = tot; | |
417 | tot += save; | |
418 | } | |
419 | ||
420 | for (i = nchars, c = l->l_line; --i >= 0; c++) | |
421 | sorted[count[c->c_column]++] = *c; | |
422 | c = sorted; | |
423 | } else | |
424 | c = l->l_line; | |
425 | while (nchars > 0) { | |
426 | this_col = c->c_column; | |
427 | endc = c; | |
428 | do { | |
429 | ++endc; | |
430 | } while (--nchars > 0 && this_col == endc->c_column); | |
431 | ||
432 | /* if -b only print last character */ | |
433 | if (no_backspaces) | |
434 | c = endc - 1; | |
435 | ||
436 | if (this_col > last_col) { | |
437 | int nspace = this_col - last_col; | |
438 | ||
439 | if (compress_spaces && nspace > 1) { | |
440 | int ntabs; | |
441 | ||
442 | ntabs = this_col / 8 - last_col / 8; | |
443 | if (ntabs > 0) { | |
444 | nspace = this_col & 7; | |
445 | while (--ntabs >= 0) | |
446 | PUTC('\t'); | |
447 | } | |
448 | } | |
449 | while (--nspace >= 0) | |
450 | PUTC(' '); | |
451 | last_col = this_col; | |
452 | } | |
453 | last_col++; | |
454 | ||
455 | for (;;) { | |
456 | if (c->c_set != last_set) { | |
457 | switch (c->c_set) { | |
458 | case CS_NORMAL: | |
459 | PUTC('\017'); | |
460 | break; | |
461 | case CS_ALTERNATE: | |
462 | PUTC('\016'); | |
463 | } | |
464 | last_set = c->c_set; | |
465 | } | |
466 | PUTC(c->c_char); | |
467 | if (++c >= endc) | |
468 | break; | |
469 | PUTC('\b'); | |
470 | } | |
471 | } | |
472 | } | |
473 | ||
474 | #define NALLOC 64 | |
475 | ||
476 | static LINE *line_freelist; | |
477 | ||
478 | LINE * | |
479 | alloc_line() | |
480 | { | |
481 | LINE *l; | |
482 | int i; | |
483 | ||
484 | if (!line_freelist) { | |
485 | l = (LINE *)xmalloc((void *)NULL, sizeof(LINE) * NALLOC); | |
486 | line_freelist = l; | |
487 | for (i = 1; i < NALLOC; i++, l++) | |
488 | l->l_next = l + 1; | |
489 | l->l_next = NULL; | |
490 | } | |
491 | l = line_freelist; | |
492 | line_freelist = l->l_next; | |
493 | ||
494 | bzero(l, sizeof(LINE)); | |
495 | return(l); | |
496 | } | |
497 | ||
fd6b7a7f | 498 | void free_line(LINE *l) |
6dbe3af9 KZ |
499 | { |
500 | l->l_next = line_freelist; | |
501 | line_freelist = l; | |
502 | } | |
503 | ||
504 | void * | |
fd6b7a7f | 505 | xmalloc(void *p, size_t size) |
6dbe3af9 KZ |
506 | { |
507 | if (!(p = (void *)realloc(p, size))) { | |
508 | (void)fprintf(stderr, "col: %s.\n", strerror(ENOMEM)); | |
509 | exit(1); | |
510 | } | |
511 | return(p); | |
512 | } | |
513 | ||
fd6b7a7f | 514 | void usage() |
6dbe3af9 KZ |
515 | { |
516 | (void)fprintf(stderr, "usage: col [-bfx] [-l nline]\n"); | |
517 | exit(1); | |
518 | } | |
519 | ||
fd6b7a7f | 520 | void wrerr() |
6dbe3af9 KZ |
521 | { |
522 | (void)fprintf(stderr, "col: write error.\n"); | |
523 | exit(1); | |
524 | } | |
525 | ||
fd6b7a7f | 526 | void warn(int line) |
6dbe3af9 KZ |
527 | { |
528 | (void)fprintf(stderr, | |
529 | "col: warning: can't back up %s.\n", line < 0 ? | |
530 | "past first line" : "-- line already flushed"); | |
531 | } |