Commit | Line | Data |
---|---|---|
15637ed4 RG |
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 | ||
37 | #ifndef lint | |
38 | char copyright[] = | |
39 | "@(#) Copyright (c) 1990 The Regents of the University of California.\n\ | |
40 | All rights reserved.\n"; | |
41 | #endif /* not lint */ | |
42 | ||
43 | #ifndef lint | |
44 | static char sccsid[] = "@(#)col.c 5.3 (Berkeley) 2/2/91"; | |
45 | #endif /* not lint */ | |
46 | ||
47 | #include <errno.h> | |
48 | #include <ctype.h> | |
49 | #include <string.h> | |
50 | #include <stdio.h> | |
51 | ||
52 | #define BS '\b' /* backspace */ | |
53 | #define TAB '\t' /* tab */ | |
54 | #define SPACE ' ' /* space */ | |
55 | #define NL '\n' /* newline */ | |
56 | #define CR '\r' /* carriage return */ | |
57 | #define ESC '\033' /* escape */ | |
58 | #define SI '\017' /* shift in to normal character set */ | |
59 | #define SO '\016' /* shift out to alternate character set */ | |
60 | #define VT '\013' /* vertical tab (aka reverse line feed) */ | |
61 | #define RLF '\007' /* ESC-07 reverse line feed */ | |
62 | #define RHLF '\010' /* ESC-010 reverse half-line feed */ | |
63 | #define FHLF '\011' /* ESC-011 forward half-line feed */ | |
64 | ||
65 | /* build up at least this many lines before flushing them out */ | |
66 | #define BUFFER_MARGIN 32 | |
67 | ||
68 | typedef char CSET; | |
69 | ||
70 | typedef struct char_str { | |
71 | #define CS_NORMAL 1 | |
72 | #define CS_ALTERNATE 2 | |
73 | short c_column; /* column character is in */ | |
74 | CSET c_set; /* character set (currently only 2) */ | |
75 | char c_char; /* character in question */ | |
76 | } CHAR; | |
77 | ||
78 | typedef struct line_str LINE; | |
79 | struct line_str { | |
80 | CHAR *l_line; /* characters on the line */ | |
81 | LINE *l_prev; /* previous line */ | |
82 | LINE *l_next; /* next line */ | |
83 | int l_lsize; /* allocated sizeof l_line */ | |
84 | int l_line_len; /* strlen(l_line) */ | |
85 | int l_needs_sort; /* set if chars went in out of order */ | |
86 | int l_max_col; /* max column in the line */ | |
87 | }; | |
88 | ||
89 | LINE *alloc_line(); | |
90 | void *xmalloc(); | |
91 | ||
92 | CSET last_set; /* char_set of last char printed */ | |
93 | LINE *lines; | |
94 | int compress_spaces; /* if doing space -> tab conversion */ | |
95 | int fine; /* if `fine' resolution (half lines) */ | |
96 | int max_bufd_lines; /* max # lines to keep in memory */ | |
97 | int nblank_lines; /* # blanks after last flushed line */ | |
98 | int no_backspaces; /* if not to output any backspaces */ | |
99 | ||
100 | #define PUTC(ch) \ | |
101 | if (putchar(ch) == EOF) \ | |
102 | wrerr(); | |
103 | ||
c4ed8671 AM |
104 | /* next tabstop after col */ |
105 | #define TABSTOP(col,ts) ((col) + (ts) - (col) % (ts)) | |
106 | ||
15637ed4 RG |
107 | main(argc, argv) |
108 | int argc; | |
109 | char **argv; | |
110 | { | |
111 | extern int optind; | |
112 | extern char *optarg; | |
113 | register int ch; | |
114 | CHAR *c; | |
115 | CSET cur_set; /* current character set */ | |
116 | LINE *l; /* current line */ | |
117 | int extra_lines; /* # of lines above first line */ | |
118 | int cur_col; /* current column */ | |
119 | int cur_line; /* line number of current position */ | |
120 | int max_line; /* max value of cur_line */ | |
121 | int this_line; /* line l points to */ | |
122 | int nflushd_lines; /* number of lines that were flushed */ | |
123 | int adjust, opt, warned; | |
124 | ||
125 | max_bufd_lines = 128; | |
126 | compress_spaces = 1; /* compress spaces into tabs */ | |
127 | while ((opt = getopt(argc, argv, "bfhl:x")) != EOF) | |
128 | switch (opt) { | |
129 | case 'b': /* do not output backspaces */ | |
130 | no_backspaces = 1; | |
131 | break; | |
132 | case 'f': /* allow half forward line feeds */ | |
133 | fine = 1; | |
134 | break; | |
135 | case 'h': /* compress spaces into tabs */ | |
136 | compress_spaces = 1; | |
137 | break; | |
138 | case 'l': /* buffered line count */ | |
139 | if ((max_bufd_lines = atoi(optarg)) <= 0) { | |
140 | (void)fprintf(stderr, | |
141 | "col: bad -l argument %s.\n", optarg); | |
142 | exit(1); | |
143 | } | |
144 | break; | |
145 | case 'x': /* do not compress spaces into tabs */ | |
146 | compress_spaces = 0; | |
147 | break; | |
148 | case '?': | |
149 | default: | |
150 | usage(); | |
151 | } | |
152 | ||
153 | if (optind != argc) | |
154 | usage(); | |
155 | ||
156 | /* this value is in half lines */ | |
157 | max_bufd_lines *= 2; | |
158 | ||
159 | adjust = cur_col = extra_lines = warned = 0; | |
160 | cur_line = max_line = nflushd_lines = this_line = 0; | |
161 | cur_set = last_set = CS_NORMAL; | |
162 | lines = l = alloc_line(); | |
163 | ||
164 | while ((ch = getchar()) != EOF) { | |
165 | if (!isgraph(ch)) { | |
166 | switch (ch) { | |
167 | case BS: /* can't go back further */ | |
168 | if (cur_col == 0) | |
169 | continue; | |
170 | --cur_col; | |
171 | continue; | |
172 | case CR: | |
173 | cur_col = 0; | |
174 | continue; | |
175 | case ESC: /* just ignore EOF */ | |
176 | switch(getchar()) { | |
177 | case RLF: | |
178 | cur_line -= 2; | |
179 | break; | |
180 | case RHLF: | |
181 | cur_line--; | |
182 | break; | |
183 | case FHLF: | |
184 | cur_line++; | |
185 | if (cur_line > max_line) | |
186 | max_line = cur_line; | |
187 | } | |
188 | continue; | |
189 | case NL: | |
190 | cur_line += 2; | |
191 | if (cur_line > max_line) | |
192 | max_line = cur_line; | |
193 | cur_col = 0; | |
194 | continue; | |
195 | case SPACE: | |
196 | ++cur_col; | |
197 | continue; | |
198 | case SI: | |
199 | cur_set = CS_NORMAL; | |
200 | continue; | |
201 | case SO: | |
202 | cur_set = CS_ALTERNATE; | |
203 | continue; | |
204 | case TAB: /* adjust column */ | |
205 | cur_col |= 7; | |
206 | ++cur_col; | |
207 | continue; | |
208 | case VT: | |
209 | cur_line -= 2; | |
210 | continue; | |
211 | } | |
212 | continue; | |
213 | } | |
214 | ||
215 | /* Must stuff ch in a line - are we at the right one? */ | |
216 | if (cur_line != this_line - adjust) { | |
217 | LINE *lnew; | |
218 | int nmove; | |
219 | ||
220 | adjust = 0; | |
221 | nmove = cur_line - this_line; | |
222 | if (!fine) { | |
223 | /* round up to next line */ | |
224 | if (cur_line & 1) { | |
225 | adjust = 1; | |
226 | nmove++; | |
227 | } | |
228 | } | |
229 | if (nmove < 0) { | |
230 | for (; nmove < 0 && l->l_prev; nmove++) | |
231 | l = l->l_prev; | |
232 | if (nmove) { | |
233 | if (nflushd_lines == 0) { | |
234 | /* | |
235 | * Allow backup past first | |
236 | * line if nothing has been | |
237 | * flushed yet. | |
238 | */ | |
239 | for (; nmove < 0; nmove++) { | |
240 | lnew = alloc_line(); | |
241 | l->l_prev = lnew; | |
242 | lnew->l_next = l; | |
243 | l = lines = lnew; | |
244 | extra_lines++; | |
245 | } | |
246 | } else { | |
247 | if (!warned++) | |
248 | warn(cur_line); | |
249 | cur_line -= nmove; | |
250 | } | |
251 | } | |
252 | } else { | |
253 | /* may need to allocate here */ | |
254 | for (; nmove > 0 && l->l_next; nmove--) | |
255 | l = l->l_next; | |
256 | for (; nmove > 0; nmove--) { | |
257 | lnew = alloc_line(); | |
258 | lnew->l_prev = l; | |
259 | l->l_next = lnew; | |
260 | l = lnew; | |
261 | } | |
262 | } | |
263 | this_line = cur_line + adjust; | |
264 | nmove = this_line - nflushd_lines; | |
265 | if (nmove >= max_bufd_lines + BUFFER_MARGIN) { | |
266 | nflushd_lines += nmove - max_bufd_lines; | |
267 | flush_lines(nmove - max_bufd_lines); | |
268 | } | |
269 | } | |
270 | /* grow line's buffer? */ | |
271 | if (l->l_line_len + 1 >= l->l_lsize) { | |
272 | int need; | |
273 | ||
274 | need = l->l_lsize ? l->l_lsize * 2 : 90; | |
275 | l->l_line = (CHAR *)xmalloc((void *) l->l_line, | |
276 | (unsigned) need * sizeof(CHAR)); | |
277 | l->l_lsize = need; | |
278 | } | |
279 | c = &l->l_line[l->l_line_len++]; | |
280 | c->c_char = ch; | |
281 | c->c_set = cur_set; | |
282 | c->c_column = cur_col; | |
283 | /* | |
284 | * If things are put in out of order, they will need sorting | |
285 | * when it is flushed. | |
286 | */ | |
287 | if (cur_col < l->l_max_col) | |
288 | l->l_needs_sort = 1; | |
289 | else | |
290 | l->l_max_col = cur_col; | |
291 | cur_col++; | |
292 | } | |
293 | /* goto the last line that had a character on it */ | |
294 | for (; l->l_next; l = l->l_next) | |
295 | this_line++; | |
296 | flush_lines(this_line - nflushd_lines + extra_lines + 1); | |
297 | ||
298 | /* make sure we leave things in a sane state */ | |
299 | if (last_set != CS_NORMAL) | |
300 | PUTC('\017'); | |
301 | ||
302 | /* flush out the last few blank lines */ | |
303 | nblank_lines = max_line - this_line; | |
304 | if (max_line & 1) | |
305 | nblank_lines++; | |
306 | else if (!nblank_lines) | |
307 | /* missing a \n on the last line? */ | |
308 | nblank_lines = 2; | |
309 | flush_blanks(); | |
310 | exit(0); | |
311 | } | |
312 | ||
313 | flush_lines(nflush) | |
314 | int nflush; | |
315 | { | |
316 | LINE *l; | |
317 | ||
318 | while (--nflush >= 0) { | |
319 | l = lines; | |
320 | lines = l->l_next; | |
321 | if (l->l_line) { | |
322 | flush_blanks(); | |
323 | flush_line(l); | |
324 | } | |
325 | nblank_lines++; | |
326 | if (l->l_line) | |
327 | (void)free((void *)l->l_line); | |
328 | free_line(l); | |
329 | } | |
330 | if (lines) | |
331 | lines->l_prev = NULL; | |
332 | } | |
333 | ||
334 | /* | |
335 | * Print a number of newline/half newlines. If fine flag is set, nblank_lines | |
336 | * is the number of half line feeds, otherwise it is the number of whole line | |
337 | * feeds. | |
338 | */ | |
339 | flush_blanks() | |
340 | { | |
341 | int half, i, nb; | |
342 | ||
343 | half = 0; | |
344 | nb = nblank_lines; | |
345 | if (nb & 1) { | |
346 | if (fine) | |
347 | half = 1; | |
348 | else | |
349 | nb++; | |
350 | } | |
351 | nb /= 2; | |
352 | for (i = nb; --i >= 0;) | |
353 | PUTC('\n'); | |
354 | if (half) { | |
355 | PUTC('\033'); | |
356 | PUTC('9'); | |
357 | if (!nb) | |
358 | PUTC('\r'); | |
359 | } | |
360 | nblank_lines = 0; | |
361 | } | |
362 | ||
363 | /* | |
364 | * Write a line to stdout taking care of space to tab conversion (-h flag) | |
365 | * and character set shifts. | |
366 | */ | |
367 | flush_line(l) | |
368 | LINE *l; | |
369 | { | |
370 | CHAR *c, *endc; | |
371 | int nchars, last_col, this_col; | |
372 | ||
373 | last_col = 0; | |
374 | nchars = l->l_line_len; | |
375 | ||
376 | if (l->l_needs_sort) { | |
377 | static CHAR *sorted; | |
378 | static int count_size, *count, i, save, sorted_size, tot; | |
379 | ||
380 | /* | |
381 | * Do an O(n) sort on l->l_line by column being careful to | |
382 | * preserve the order of characters in the same column. | |
383 | */ | |
384 | if (l->l_lsize > sorted_size) { | |
385 | sorted_size = l->l_lsize; | |
386 | sorted = (CHAR *)xmalloc((void *)sorted, | |
387 | (unsigned)sizeof(CHAR) * sorted_size); | |
388 | } | |
389 | if (l->l_max_col >= count_size) { | |
390 | count_size = l->l_max_col + 1; | |
391 | count = (int *)xmalloc((void *)count, | |
392 | (unsigned)sizeof(int) * count_size); | |
393 | } | |
394 | bzero((char *)count, sizeof(int) * l->l_max_col + 1); | |
395 | for (i = nchars, c = l->l_line; --i >= 0; c++) | |
396 | count[c->c_column]++; | |
397 | ||
398 | /* | |
399 | * calculate running total (shifted down by 1) to use as | |
400 | * indices into new line. | |
401 | */ | |
402 | for (tot = 0, i = 0; i <= l->l_max_col; i++) { | |
403 | save = count[i]; | |
404 | count[i] = tot; | |
405 | tot += save; | |
406 | } | |
407 | ||
408 | for (i = nchars, c = l->l_line; --i >= 0; c++) | |
409 | sorted[count[c->c_column]++] = *c; | |
410 | c = sorted; | |
411 | } else | |
412 | c = l->l_line; | |
413 | while (nchars > 0) { | |
414 | this_col = c->c_column; | |
415 | endc = c; | |
416 | do { | |
417 | ++endc; | |
418 | } while (--nchars > 0 && this_col == endc->c_column); | |
419 | ||
420 | /* if -b only print last character */ | |
421 | if (no_backspaces) | |
422 | c = endc - 1; | |
423 | ||
424 | if (this_col > last_col) { | |
425 | int nspace = this_col - last_col; | |
426 | ||
427 | if (compress_spaces && nspace > 1) { | |
428 | int ntabs; | |
429 | ||
c4ed8671 AM |
430 | while ((ntabs = TABSTOP(last_col, 8)) <= this_col) { |
431 | PUTC('\t'); | |
432 | last_col = ntabs; | |
433 | } | |
434 | nspace = this_col - last_col; | |
435 | /* | |
15637ed4 RG |
436 | ntabs = this_col / 8 - last_col / 8; |
437 | nspace -= ntabs * 8; | |
438 | while (--ntabs >= 0) | |
439 | PUTC('\t'); | |
c4ed8671 | 440 | */ |
15637ed4 RG |
441 | } |
442 | while (--nspace >= 0) | |
443 | PUTC(' '); | |
444 | last_col = this_col; | |
445 | } | |
446 | last_col++; | |
447 | ||
448 | for (;;) { | |
449 | if (c->c_set != last_set) { | |
450 | switch (c->c_set) { | |
451 | case CS_NORMAL: | |
452 | PUTC('\017'); | |
453 | break; | |
454 | case CS_ALTERNATE: | |
455 | PUTC('\016'); | |
456 | } | |
457 | last_set = c->c_set; | |
458 | } | |
459 | PUTC(c->c_char); | |
460 | if (++c >= endc) | |
461 | break; | |
462 | PUTC('\b'); | |
463 | } | |
464 | } | |
465 | } | |
466 | ||
467 | #define NALLOC 64 | |
468 | ||
469 | static LINE *line_freelist; | |
470 | ||
471 | LINE * | |
472 | alloc_line() | |
473 | { | |
474 | LINE *l; | |
475 | int i; | |
476 | ||
477 | if (!line_freelist) { | |
478 | l = (LINE *)xmalloc((void *)NULL, sizeof(LINE) * NALLOC); | |
479 | line_freelist = l; | |
480 | for (i = 1; i < NALLOC; i++, l++) | |
481 | l->l_next = l + 1; | |
482 | l->l_next = NULL; | |
483 | } | |
484 | l = line_freelist; | |
485 | line_freelist = l->l_next; | |
486 | ||
487 | bzero(l, sizeof(LINE)); | |
488 | return(l); | |
489 | } | |
490 | ||
491 | free_line(l) | |
492 | LINE *l; | |
493 | { | |
494 | l->l_next = line_freelist; | |
495 | line_freelist = l; | |
496 | } | |
497 | ||
498 | void * | |
499 | xmalloc(p, size) | |
500 | void *p; | |
501 | size_t size; | |
502 | { | |
503 | if (!(p = (void *)realloc(p, size))) { | |
504 | (void)fprintf(stderr, "col: %s.\n", strerror(ENOMEM)); | |
505 | exit(1); | |
506 | } | |
507 | return(p); | |
508 | } | |
509 | ||
510 | usage() | |
511 | { | |
512 | (void)fprintf(stderr, "usage: col [-bfx] [-l nline]\n"); | |
513 | exit(1); | |
514 | } | |
515 | ||
516 | wrerr() | |
517 | { | |
518 | (void)fprintf(stderr, "col: write error.\n"); | |
519 | exit(1); | |
520 | } | |
521 | ||
522 | warn(line) | |
523 | int line; | |
524 | { | |
525 | (void)fprintf(stderr, | |
526 | "col: warning: can't back up %s.\n", line < 0 ? | |
527 | "past first line" : "-- line already flushed"); | |
528 | } |