2 * Copyright (c) 1980, 1993
3 * The Regents of the University of California. All rights reserved.
5 * Redistribution and use in source and binary forms, with or without
6 * modification, are permitted provided that the following conditions
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.
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
35 static const char copyright[] =
36 "@(#) Copyright (c) 1980, 1993\n\
37 The Regents of the University of California. All rights reserved.\n";
42 static char sccsid[] = "@(#)checknr.c 8.1 (Berkeley) 6/6/93";
46 #include <sys/cdefs.h>
47 __FBSDID("$FreeBSD$");
50 * checknr: check an nroff/troff input file for matching macro calls.
51 * we also attempt to match size and font changes, but only the embedded
52 * kind. These must end in \s0 and \fP resp. Maybe more sophistication
53 * later but for now think of these restrictions as contributions to
54 * structured typesetting.
61 #define MAXSTK 100 /* Stack size */
62 #define MAXBR 100 /* Max number of bracket pairs known */
63 #define MAXCMDS 500 /* Max number of commands known */
65 void addcmd __P((char *));
66 void addmac __P((const char *));
67 int binsrch __P((const char *));
68 void checkknown __P((char *));
69 void chkcmd __P((char *, char *));
70 void complain __P((int));
71 int eq __P((const char *, const char *));
72 void nomatch __P((char *));
74 void process __P((FILE *));
76 static void usage __P((void));
79 * The stack on which we remember what we've seen so far.
82 int opno; /* number of opening bracket */
83 int pl; /* '+', '-', ' ' for \s, 1 for \f, 0 for .ft */
84 int parm; /* parm to size, font, etc */
85 int lno; /* line number the thing came in in */
90 * The kinds of opening and closing brackets.
96 /* A few bare bones troff commands */
98 {"sz", "sz"}, /* also \s */
100 {"ft", "ft"}, /* also \f */
101 /* the -mm package */
114 /* the -ms package */
130 /* The -me package */
139 /* Things needed by preprocessors */
148 * All commands known to nroff, plus macro packages.
149 * Used so we can complain about unrecognized commands.
151 const char *knowncmds[MAXCMDS] = {
152 "$c", "$f", "$h", "$p", "$s", "(b", "(c", "(d", "(f", "(l", "(q", "(t",
153 "(x", "(z", ")b", ")c", ")d", ")f", ")l", ")q", ")t", ")x", ")z", "++",
154 "+c", "1C", "1c", "2C", "2c", "@(", "@)", "@C", "@D", "@F", "@I", "@M",
155 "@c", "@e", "@f", "@h", "@m", "@n", "@o", "@p", "@r", "@t", "@z", "AB",
156 "AE", "AF", "AI", "AL", "AM", "AS", "AT", "AU", "AX", "B", "B1", "B2",
157 "BD", "BE", "BG", "BL", "BS", "BT", "BX", "C1", "C2", "CD", "CM", "CT",
158 "D", "DA", "DE", "DF", "DL", "DS", "DT", "EC", "EF", "EG", "EH", "EM",
159 "EN", "EQ", "EX", "FA", "FD", "FE", "FG", "FJ", "FK", "FL", "FN", "FO",
160 "FQ", "FS", "FV", "FX", "H", "HC", "HD", "HM", "HO", "HU", "I", "ID",
161 "IE", "IH", "IM", "IP", "IX", "IZ", "KD", "KE", "KF", "KQ", "KS", "LB",
162 "LC", "LD", "LE", "LG", "LI", "LP", "MC", "ME", "MF", "MH", "ML", "MR",
163 "MT", "ND", "NE", "NH", "NL", "NP", "NS", "OF", "OH", "OK", "OP", "P",
164 "P1", "PF", "PH", "PP", "PT", "PX", "PY", "QE", "QP", "QS", "R", "RA",
165 "RC", "RE", "RL", "RP", "RQ", "RS", "RT", "S", "S0", "S2", "S3", "SA",
166 "SG", "SH", "SK", "SM", "SP", "SY", "T&", "TA", "TB", "TC", "TD", "TE",
167 "TH", "TL", "TM", "TP", "TQ", "TR", "TS", "TX", "UL", "US", "UX", "VL",
168 "WC", "WH", "XA", "XD", "XE", "XF", "XK", "XP", "XS", "[", "[-", "[0",
169 "[1", "[2", "[3", "[4", "[5", "[<", "[>", "[]", "]", "]-", "]<", "]>",
170 "][", "ab", "ac", "ad", "af", "am", "ar", "as", "b", "ba", "bc", "bd",
171 "bi", "bl", "bp", "br", "bx", "c.", "c2", "cc", "ce", "cf", "ch", "cs",
172 "ct", "cu", "da", "de", "di", "dl", "dn", "ds", "dt", "dw", "dy", "ec",
173 "ef", "eh", "el", "em", "eo", "ep", "ev", "ex", "fc", "fi", "fl", "fo",
174 "fp", "ft", "fz", "hc", "he", "hl", "hp", "ht", "hw", "hx", "hy", "i",
175 "ie", "if", "ig", "in", "ip", "it", "ix", "lc", "lg", "li", "ll", "ln",
176 "lo", "lp", "ls", "lt", "m1", "m2", "m3", "m4", "mc", "mk", "mo", "n1",
177 "n2", "na", "ne", "nf", "nh", "nl", "nm", "nn", "np", "nr", "ns", "nx",
178 "of", "oh", "os", "pa", "pc", "pi", "pl", "pm", "pn", "po", "pp", "ps",
179 "q", "r", "rb", "rd", "re", "rm", "rn", "ro", "rr", "rs", "rt", "sb",
180 "sc", "sh", "sk", "so", "sp", "ss", "st", "sv", "sz", "ta", "tc", "th",
181 "ti", "tl", "tm", "tp", "tr", "u", "uf", "uh", "ul", "vs", "wh", "xp",
185 int lineno; /* current line number in input file */
186 const char *cfilename; /* name of current file */
187 int nfiles; /* number of files to process */
188 int fflag; /* -f: ignore \f */
189 int sflag; /* -s: ignore \s */
190 int ncmds; /* size of knowncmds */
191 int slot; /* slot in knowncmds found by binsrch */
203 /* Figure out how many known commands there are */
204 while (knowncmds[ncmds])
206 while (argc > 1 && argv[1][0] == '-') {
209 /* -a: add pairs of macros */
211 i = strlen(argv[1]) - 2;
214 /* look for empty macro slots */
215 for (i=0; br[i].opbr; i++)
217 for (cp=argv[1]+3; cp[-1]; cp += 6) {
218 br[i].opbr = strncpy(malloc(3), cp, 2);
219 br[i].clbr = strncpy(malloc(3), cp+3, 2);
220 addmac(br[i].opbr); /* knows pairs are also known cmds */
226 /* -c: add known commands */
228 i = strlen(argv[1]) - 2;
231 for (cp=argv[1]+3; cp[-1]; cp += 3) {
232 if (cp[2] && cp[2] != '.')
240 /* -f: ignore font changes */
245 /* -s: ignore size changes */
258 for (i=1; i<argc; i++) {
260 f = fopen(cfilename, "r");
277 "usage: checknr [-a.xx.yy.xx.yy...] [-c.xx.xx.xx...] [-s] [-f] file\n");
286 char mac[5]; /* The current macro or nroff command */
288 static char line[256]; /* the current line */
291 for (lineno = 1; fgets(line, sizeof line, f); lineno++) {
292 if (line[0] == '.') {
294 * find and isolate the macro/command name.
296 strncpy(mac, line+1, 4);
297 if (isspace(mac[0])) {
299 printf("Empty command\n");
300 } else if (isspace(mac[1])) {
302 } else if (isspace(mac[2])) {
304 } else if (mac[0] != '\\' || mac[1] != '\"') {
306 printf("Command too long\n");
310 * Is it a known command?
324 * At this point we process the line looking
327 for (i=0; line[i]; i++)
328 if (line[i]=='\\' && (i==0 || line[i-1]!='\\')) {
329 if (!sflag && line[++i]=='s') {
336 while (isdigit(line[++i]))
337 n = 10 * n + line[i] - '0';
340 if (stk[stktop].opno == SZ) {
344 printf("unmatched \\s0\n");
347 stk[++stktop].opno = SZ;
349 stk[stktop].parm = n;
350 stk[stktop].lno = lineno;
352 } else if (!fflag && line[i]=='f') {
355 if (stk[stktop].opno == FT) {
359 printf("unmatched \\fP\n");
362 stk[++stktop].opno = FT;
364 stk[stktop].parm = n;
365 stk[stktop].lno = lineno;
371 * We've hit the end and look at all this stuff that hasn't been
372 * matched yet! Complain, complain.
374 for (i=stktop; i>=0; i--) {
384 printf("Unmatched ");
394 printf(".%s", br[stk[i].opno].opbr);
395 else switch(stk[i].opno) {
397 printf("\\s%c%d", stk[i].pl, stk[i].parm);
400 printf("\\f%c", stk[i].parm);
403 printf("Bug: stk[%d].opno = %d = .%s, .%s",
404 i, stk[i].opno, br[stk[i].opno].opbr, br[stk[i].opno].clbr);
416 * Check to see if it matches top of stack.
418 if (stktop >= 0 && eq(mac, br[stk[stktop].opno].clbr))
419 stktop--; /* OK. Pop & forget */
421 /* No. Maybe it's an opener */
422 for (i=0; br[i].opbr; i++) {
423 if (eq(mac, br[i].opbr)) {
424 /* Found. Push it. */
426 stk[stktop].opno = i;
428 stk[stktop].parm = 0;
429 stk[stktop].lno = lineno;
433 * Maybe it's an unmatched closer.
434 * NOTE: this depends on the fact
435 * that none of the closers can be
438 if (eq(mac, br[i].clbr)) {
453 * Look for a match further down on stack
454 * If we find one, it suggests that the stuff in
455 * between is supposed to match itself.
457 for (j=stktop; j>=0; j--)
458 if (eq(mac,br[stk[j].opno].clbr)) {
459 /* Found. Make a good diagnostic. */
462 * Check for special case \fx..\fR and don't
465 if (stk[j+1].opno==FT && stk[j+1].parm!='R'
466 && stk[j+2].opno==FT && stk[j+2].parm=='R') {
471 * We have two unmatched frobs. Chances are
472 * they were intended to match, so we mention
477 printf(" does not match %d: ", stk[j+2].lno);
480 } else for (i=j+1; i <= stktop; i++) {
486 /* Didn't find one. Throw this away. */
488 printf("Unmatched .%s\n", mac);
491 /* eq: are two strings equal? */
496 return (strcmp(s1, s2) == 0);
499 /* print the first part of an error message, given the line number */
505 printf("%s: ", cfilename);
506 printf("%d: ", linen);
516 if (binsrch(mac) >= 0)
518 if (mac[0] == '\\' && mac[1] == '"') /* comments */
522 printf("Unknown command: .%s\n", mac);
526 * We have a .de xx line in "line". Add xx to the list of known commands.
534 /* grab the macro being defined */
536 while (isspace(*mac))
540 printf("illegal define: %s\n", line);
544 if (isspace(mac[1]) || mac[1] == '\\')
546 if (ncmds >= MAXCMDS) {
547 printf("Only %d known commands allowed\n", MAXCMDS);
554 * Add mac to the list. We should really have some kind of tree
555 * structure here but this is a quick-and-dirty job and I just don't
556 * have time to mess with it. (I wonder if this will come back to haunt
557 * me someday?) Anyway, I claim that .de is fairly rare in user
558 * nroff programs, and the register loop below is pretty fast.
564 const char **src, **dest, **loc;
566 if (binsrch(mac) >= 0){ /* it's OK to redefine something */
568 printf("binsrch(%s) -> already in table\n", mac);
572 /* binsrch sets slot as a side effect */
574 printf("binsrch(%s) -> %d\n", mac, slot);
576 loc = &knowncmds[slot];
577 src = &knowncmds[ncmds-1];
581 *loc = strcpy(malloc(3), mac);
584 printf("after: %s %s %s %s %s, %d cmds\n", knowncmds[slot-2], knowncmds[slot-1], knowncmds[slot], knowncmds[slot+1], knowncmds[slot+2], ncmds);
589 * Do a binary search in knowncmds for mac.
590 * If found, return the index. If not, return -1.
596 const char *p; /* pointer to current cmd in list */
597 int d; /* difference if any */
598 int mid; /* mid point in binary search */
599 int top, bot; /* boundaries of bin search, inclusive */
616 slot = bot; /* place it would have gone */