xref: /netbsd-src/usr.bin/fmt/fmt.c (revision ae1bfcddc410612bc8c58b807e1830becb69a24c)
1 /*
2  * Copyright (c) 1980 Regents of the University of California.
3  * All rights reserved.
4  *
5  * Redistribution and use in source and binary forms, with or without
6  * modification, are permitted provided that the following conditions
7  * are met:
8  * 1. Redistributions of source code must retain the above copyright
9  *    notice, this list of conditions and the following disclaimer.
10  * 2. Redistributions in binary form must reproduce the above copyright
11  *    notice, this list of conditions and the following disclaimer in the
12  *    documentation and/or other materials provided with the distribution.
13  * 3. All advertising materials mentioning features or use of this software
14  *    must display the following acknowledgement:
15  *	This product includes software developed by the University of
16  *	California, Berkeley and its contributors.
17  * 4. Neither the name of the University nor the names of its contributors
18  *    may be used to endorse or promote products derived from this software
19  *    without specific prior written permission.
20  *
21  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
22  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
23  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
24  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
25  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
26  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
27  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
28  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
29  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
30  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
31  * SUCH DAMAGE.
32  */
33 
34 #ifndef lint
35 char copyright[] =
36 "@(#) Copyright (c) 1980 Regents of the University of California.\n\
37  All rights reserved.\n";
38 #endif /* not lint */
39 
40 #ifndef lint
41 /*static char sccsid[] = "from: @(#)fmt.c	5.10 (Berkeley) 6/1/90";*/
42 static char rcsid[] = "$Id: fmt.c,v 1.2 1993/08/01 18:15:53 mycroft Exp $";
43 #endif /* not lint */
44 
45 #include <stdio.h>
46 #include <ctype.h>
47 
48 /*
49  * fmt -- format the concatenation of input files or standard input
50  * onto standard output.  Designed for use with Mail ~|
51  *
52  * Syntax : fmt [ goal [ max ] ] [ name ... ]
53  * Authors: Kurt Shoens (UCB) 12/7/78;
54  *          Liz Allen (UMCP) 2/24/83 [Addition of goal length concept].
55  */
56 
57 /* LIZ@UOM 6/18/85 -- Don't need LENGTH any more.
58  * #define	LENGTH	72		Max line length in output
59  */
60 #define	NOSTR	((char *) 0)	/* Null string pointer for lint */
61 
62 /* LIZ@UOM 6/18/85 --New variables goal_length and max_length */
63 #define GOAL_LENGTH 65
64 #define MAX_LENGTH 75
65 int	goal_length;		/* Target or goal line length in output */
66 int	max_length;		/* Max line length in output */
67 int	pfx;			/* Current leading blank count */
68 int	lineno;			/* Current input line */
69 int	mark;			/* Last place we saw a head line */
70 
71 char	*malloc();		/* for lint . . . */
72 char	*headnames[] = {"To", "Subject", "Cc", 0};
73 
74 /*
75  * Drive the whole formatter by managing input files.  Also,
76  * cause initialization of the output stuff and flush it out
77  * at the end.
78  */
79 
80 main(argc, argv)
81 	int argc;
82 	char **argv;
83 {
84 	register FILE *fi;
85 	register int errs = 0;
86 	int number;		/* LIZ@UOM 6/18/85 */
87 
88 	goal_length = GOAL_LENGTH;
89 	max_length = MAX_LENGTH;
90 	setout();
91 	lineno = 1;
92 	mark = -10;
93 	/*
94 	 * LIZ@UOM 6/18/85 -- Check for goal and max length arguments
95 	 */
96 	if (argc > 1 && (1 == (sscanf(argv[1], "%d", &number)))) {
97 		argv++;
98 		argc--;
99 		goal_length = number;
100 		if (argc > 1 && (1 == (sscanf(argv[1], "%d", &number)))) {
101 			argv++;
102 			argc--;
103 			max_length = number;
104 		}
105 	}
106 	if (max_length <= goal_length) {
107 		fprintf(stderr, "Max length must be greater than %s\n",
108 			"goal length");
109 		exit(1);
110 	}
111 	if (argc < 2) {
112 		fmt(stdin);
113 		oflush();
114 		exit(0);
115 	}
116 	while (--argc) {
117 		if ((fi = fopen(*++argv, "r")) == NULL) {
118 			perror(*argv);
119 			errs++;
120 			continue;
121 		}
122 		fmt(fi);
123 		fclose(fi);
124 	}
125 	oflush();
126 	exit(errs);
127 }
128 
129 /*
130  * Read up characters from the passed input file, forming lines,
131  * doing ^H processing, expanding tabs, stripping trailing blanks,
132  * and sending each line down for analysis.
133  */
134 fmt(fi)
135 	FILE *fi;
136 {
137 	char linebuf[BUFSIZ], canonb[BUFSIZ];
138 	register char *cp, *cp2;
139 	register int c, col;
140 
141 	c = getc(fi);
142 	while (c != EOF) {
143 		/*
144 		 * Collect a line, doing ^H processing.
145 		 * Leave tabs for now.
146 		 */
147 		cp = linebuf;
148 		while (c != '\n' && c != EOF && cp-linebuf < BUFSIZ-1) {
149 			if (c == '\b') {
150 				if (cp > linebuf)
151 					cp--;
152 				c = getc(fi);
153 				continue;
154 			}
155 			if ((c < ' ' || c >= 0177) && c != '\t') {
156 				c = getc(fi);
157 				continue;
158 			}
159 			*cp++ = c;
160 			c = getc(fi);
161 		}
162 		*cp = '\0';
163 
164 		/*
165 		 * Toss anything remaining on the input line.
166 		 */
167 		while (c != '\n' && c != EOF)
168 			c = getc(fi);
169 
170 		/*
171 		 * Expand tabs on the way to canonb.
172 		 */
173 		col = 0;
174 		cp = linebuf;
175 		cp2 = canonb;
176 		while (c = *cp++) {
177 			if (c != '\t') {
178 				col++;
179 				if (cp2-canonb < BUFSIZ-1)
180 					*cp2++ = c;
181 				continue;
182 			}
183 			do {
184 				if (cp2-canonb < BUFSIZ-1)
185 					*cp2++ = ' ';
186 				col++;
187 			} while ((col & 07) != 0);
188 		}
189 
190 		/*
191 		 * Swipe trailing blanks from the line.
192 		 */
193 		for (cp2--; cp2 >= canonb && *cp2 == ' '; cp2--)
194 			;
195 		*++cp2 = '\0';
196 		prefix(canonb);
197 		if (c != EOF)
198 			c = getc(fi);
199 	}
200 }
201 
202 /*
203  * Take a line devoid of tabs and other garbage and determine its
204  * blank prefix.  If the indent changes, call for a linebreak.
205  * If the input line is blank, echo the blank line on the output.
206  * Finally, if the line minus the prefix is a mail header, try to keep
207  * it on a line by itself.
208  */
209 prefix(line)
210 	char line[];
211 {
212 	register char *cp, **hp;
213 	register int np, h;
214 
215 	if (strlen(line) == 0) {
216 		oflush();
217 		putchar('\n');
218 		return;
219 	}
220 	for (cp = line; *cp == ' '; cp++)
221 		;
222 	np = cp - line;
223 
224 	/*
225 	 * The following horrible expression attempts to avoid linebreaks
226 	 * when the indent changes due to a paragraph.
227 	 */
228 	if (np != pfx && (np > pfx || abs(pfx-np) > 8))
229 		oflush();
230 	if (h = ishead(cp))
231 		oflush(), mark = lineno;
232 	if (lineno - mark < 3 && lineno - mark > 0)
233 		for (hp = &headnames[0]; *hp != (char *) 0; hp++)
234 			if (ispref(*hp, cp)) {
235 				h = 1;
236 				oflush();
237 				break;
238 			}
239 	if (!h && (h = (*cp == '.')))
240 		oflush();
241 	pfx = np;
242 	if (h)
243 		pack(cp);
244 	else	split(cp);
245 	if (h)
246 		oflush();
247 	lineno++;
248 }
249 
250 /*
251  * Split up the passed line into output "words" which are
252  * maximal strings of non-blanks with the blank separation
253  * attached at the end.  Pass these words along to the output
254  * line packer.
255  */
256 split(line)
257 	char line[];
258 {
259 	register char *cp, *cp2;
260 	char word[BUFSIZ];
261 	int wordl;		/* LIZ@UOM 6/18/85 */
262 
263 	cp = line;
264 	while (*cp) {
265 		cp2 = word;
266 		wordl = 0;	/* LIZ@UOM 6/18/85 */
267 
268 		/*
269 		 * Collect a 'word,' allowing it to contain escaped white
270 		 * space.
271 		 */
272 		while (*cp && *cp != ' ') {
273 			if (*cp == '\\' && isspace(cp[1]))
274 				*cp2++ = *cp++;
275 			*cp2++ = *cp++;
276 			wordl++;/* LIZ@UOM 6/18/85 */
277 		}
278 
279 		/*
280 		 * Guarantee a space at end of line. Two spaces after end of
281 		 * sentence punctuation.
282 		 */
283 		if (*cp == '\0') {
284 			*cp2++ = ' ';
285 			if (index(".:!", cp[-1]))
286 				*cp2++ = ' ';
287 		}
288 		while (*cp == ' ')
289 			*cp2++ = *cp++;
290 		*cp2 = '\0';
291 		/*
292 		 * LIZ@UOM 6/18/85 pack(word);
293 		 */
294 		pack(word, wordl);
295 	}
296 }
297 
298 /*
299  * Output section.
300  * Build up line images from the words passed in.  Prefix
301  * each line with correct number of blanks.  The buffer "outbuf"
302  * contains the current partial line image, including prefixed blanks.
303  * "outp" points to the next available space therein.  When outp is NOSTR,
304  * there ain't nothing in there yet.  At the bottom of this whole mess,
305  * leading tabs are reinserted.
306  */
307 char	outbuf[BUFSIZ];			/* Sandbagged output line image */
308 char	*outp;				/* Pointer in above */
309 
310 /*
311  * Initialize the output section.
312  */
313 setout()
314 {
315 	outp = NOSTR;
316 }
317 
318 /*
319  * Pack a word onto the output line.  If this is the beginning of
320  * the line, push on the appropriately-sized string of blanks first.
321  * If the word won't fit on the current line, flush and begin a new
322  * line.  If the word is too long to fit all by itself on a line,
323  * just give it its own and hope for the best.
324  *
325  * LIZ@UOM 6/18/85 -- If the new word will fit in at less than the
326  *	goal length, take it.  If not, then check to see if the line
327  *	will be over the max length; if so put the word on the next
328  *	line.  If not, check to see if the line will be closer to the
329  *	goal length with or without the word and take it or put it on
330  *	the next line accordingly.
331  */
332 
333 /*
334  * LIZ@UOM 6/18/85 -- pass in the length of the word as well
335  * pack(word)
336  *	char word[];
337  */
338 pack(word,wl)
339 	char word[];
340 	int wl;
341 {
342 	register char *cp;
343 	register int s, t;
344 
345 	if (outp == NOSTR)
346 		leadin();
347 	/*
348 	 * LIZ@UOM 6/18/85 -- change condition to check goal_length; s is the
349 	 * length of the line before the word is added; t is now the length
350 	 * of the line after the word is added
351 	 *	t = strlen(word);
352 	 *	if (t+s <= LENGTH)
353 	 */
354 	s = outp - outbuf;
355 	t = wl + s;
356 	if ((t <= goal_length) ||
357 	    ((t <= max_length) && (t - goal_length <= goal_length - s))) {
358 		/*
359 		 * In like flint!
360 		 */
361 		for (cp = word; *cp; *outp++ = *cp++);
362 		return;
363 	}
364 	if (s > pfx) {
365 		oflush();
366 		leadin();
367 	}
368 	for (cp = word; *cp; *outp++ = *cp++);
369 }
370 
371 /*
372  * If there is anything on the current output line, send it on
373  * its way.  Set outp to NOSTR to indicate the absence of the current
374  * line prefix.
375  */
376 oflush()
377 {
378 	if (outp == NOSTR)
379 		return;
380 	*outp = '\0';
381 	tabulate(outbuf);
382 	outp = NOSTR;
383 }
384 
385 /*
386  * Take the passed line buffer, insert leading tabs where possible, and
387  * output on standard output (finally).
388  */
389 tabulate(line)
390 	char line[];
391 {
392 	register char *cp;
393 	register int b, t;
394 
395 	/*
396 	 * Toss trailing blanks in the output line.
397 	 */
398 	cp = line + strlen(line) - 1;
399 	while (cp >= line && *cp == ' ')
400 		cp--;
401 	*++cp = '\0';
402 
403 	/*
404 	 * Count the leading blank space and tabulate.
405 	 */
406 	for (cp = line; *cp == ' '; cp++)
407 		;
408 	b = cp-line;
409 	t = b >> 3;
410 	b &= 07;
411 	if (t > 0)
412 		do
413 			putc('\t', stdout);
414 		while (--t);
415 	if (b > 0)
416 		do
417 			putc(' ', stdout);
418 		while (--b);
419 	while (*cp)
420 		putc(*cp++, stdout);
421 	putc('\n', stdout);
422 }
423 
424 /*
425  * Initialize the output line with the appropriate number of
426  * leading blanks.
427  */
428 leadin()
429 {
430 	register int b;
431 	register char *cp;
432 
433 	for (b = 0, cp = outbuf; b < pfx; b++)
434 		*cp++ = ' ';
435 	outp = cp;
436 }
437 
438 /*
439  * Save a string in dynamic space.
440  * This little goodie is needed for
441  * a headline detector in head.c
442  */
443 char *
444 savestr(str)
445 	char str[];
446 {
447 	register char *top;
448 
449 	top = malloc(strlen(str) + 1);
450 	if (top == NOSTR) {
451 		fprintf(stderr, "fmt:  Ran out of memory\n");
452 		exit(1);
453 	}
454 	strcpy(top, str);
455 	return (top);
456 }
457 
458 /*
459  * Is s1 a prefix of s2??
460  */
461 ispref(s1, s2)
462 	register char *s1, *s2;
463 {
464 
465 	while (*s1++ == *s2)
466 		;
467 	return (*s1 == '\0');
468 }
469