1 /* $Id: mdoc.c,v 1.274 2018/12/31 07:46:07 schwarze Exp $ */
3 * Copyright (c) 2008, 2009, 2010, 2011 Kristaps Dzonsons <kristaps@bsd.lv>
4 * Copyright (c) 2010, 2012-2018 Ingo Schwarze <schwarze@openbsd.org>
6 * Permission to use, copy, modify, and distribute this software for any
7 * purpose with or without fee is hereby granted, provided that the above
8 * copyright notice and this permission notice appear in all copies.
10 * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHORS DISCLAIM ALL WARRANTIES
11 * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
12 * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHORS BE LIABLE FOR
13 * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
14 * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
15 * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
16 * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
20 #include <sys/types.h>
30 #include "mandoc_aux.h"
34 #include "libmandoc.h"
38 const char *const __mdoc_argnames[MDOC_ARG_MAX] = {
39 "split", "nosplit", "ragged",
40 "unfilled", "literal", "file",
41 "offset", "bullet", "dash",
42 "hyphen", "item", "enum",
43 "tag", "diag", "hang",
44 "ohang", "inset", "column",
45 "width", "compact", "std",
46 "filled", "words", "emphasis",
47 "symbolic", "nested", "centered"
49 const char * const *mdoc_argnames = __mdoc_argnames;
51 static int mdoc_ptext(struct roff_man *, int, char *, int);
52 static int mdoc_pmacro(struct roff_man *, int, char *, int);
56 * Main parse routine. Parses a single line -- really just hands off to
57 * the macro (mdoc_pmacro()) or text parser (mdoc_ptext()).
60 mdoc_parseln(struct roff_man *mdoc, int ln, char *buf, int offs)
63 if (mdoc->last->type != ROFFT_EQN || ln > mdoc->last->line)
64 mdoc->flags |= MDOC_NEWLINE;
67 * Let the roff nS register switch SYNOPSIS mode early,
68 * such that the parser knows at all times
69 * whether this mode is on or off.
70 * Note that this mode is also switched by the Sh macro.
72 if (roff_getreg(mdoc->roff, "nS"))
73 mdoc->flags |= MDOC_SYNOPSIS;
75 mdoc->flags &= ~MDOC_SYNOPSIS;
77 return roff_getcontrol(mdoc->roff, buf, &offs) ?
78 mdoc_pmacro(mdoc, ln, buf, offs) :
79 mdoc_ptext(mdoc, ln, buf, offs);
83 mdoc_tail_alloc(struct roff_man *mdoc, int line, int pos, enum roff_tok tok)
87 p = roff_node_alloc(mdoc, line, pos, ROFFT_TAIL, tok);
88 roff_node_append(mdoc, p);
89 mdoc->next = ROFF_NEXT_CHILD;
93 mdoc_endbody_alloc(struct roff_man *mdoc, int line, int pos,
94 enum roff_tok tok, struct roff_node *body)
98 body->flags |= NODE_ENDED;
99 body->parent->flags |= NODE_ENDED;
100 p = roff_node_alloc(mdoc, line, pos, ROFFT_BODY, tok);
102 p->norm = body->norm;
103 p->end = ENDBODY_SPACE;
104 roff_node_append(mdoc, p);
105 mdoc->next = ROFF_NEXT_SIBLING;
110 mdoc_block_alloc(struct roff_man *mdoc, int line, int pos,
111 enum roff_tok tok, struct mdoc_arg *args)
115 p = roff_node_alloc(mdoc, line, pos, ROFFT_BLOCK, tok);
126 p->norm = mandoc_calloc(1, sizeof(union mdoc_data));
131 roff_node_append(mdoc, p);
132 mdoc->next = ROFF_NEXT_CHILD;
137 mdoc_elem_alloc(struct roff_man *mdoc, int line, int pos,
138 enum roff_tok tok, struct mdoc_arg *args)
142 p = roff_node_alloc(mdoc, line, pos, ROFFT_ELEM, tok);
149 p->norm = mandoc_calloc(1, sizeof(union mdoc_data));
154 roff_node_append(mdoc, p);
155 mdoc->next = ROFF_NEXT_CHILD;
159 * Parse free-form text, that is, a line that does not begin with the
163 mdoc_ptext(struct roff_man *mdoc, int line, char *buf, int offs)
172 * If a column list contains plain text, assume an implicit item
173 * macro. This can happen one or more times at the beginning
174 * of such a list, intermixed with non-It mdoc macros and with
175 * nodes generated on the roff level, for example by tbl.
178 if ((n->tok == MDOC_Bl && n->type == ROFFT_BODY &&
179 n->end == ENDBODY_NOT && n->norm->Bl.type == LIST_column) ||
180 (n->parent != NULL && n->parent->tok == MDOC_Bl &&
181 n->parent->norm->Bl.type == LIST_column)) {
182 mdoc->flags |= MDOC_FREECOL;
183 (*mdoc_macro(MDOC_It)->fp)(mdoc, MDOC_It,
184 line, offs, &offs, buf);
189 * Search for the beginning of unescaped trailing whitespace (ws)
190 * and for the first character not to be output (end).
193 /* FIXME: replace with strcspn(). */
195 for (c = end = buf + offs; *c; c++) {
203 * Always warn about trailing tabs,
204 * even outside literal context,
205 * where they should be put on the next line.
210 * Strip trailing tabs in literal context only;
211 * outside, they affect the next line.
213 if (mdoc->flags & ROFF_NOFILL)
217 /* Skip the escaped character, too, if any. */
230 mandoc_msg(MANDOCERR_SPACE_EOL, line, (int)(ws - buf), NULL);
233 * Blank lines are allowed in no-fill mode
234 * and cancel preceding \c,
235 * but add a single vertical space elsewhere.
238 if (buf[offs] == '\0' && (mdoc->flags & ROFF_NOFILL) == 0) {
239 switch (mdoc->last->type) {
241 sp = mdoc->last->string;
242 cp = end = strchr(sp, '\0') - 2;
243 if (cp < sp || cp[0] != '\\' || cp[1] != 'c')
245 while (cp > sp && cp[-1] == '\\')
254 mandoc_msg(MANDOCERR_FI_BLANK, line, (int)(c - buf), NULL);
255 roff_elem_alloc(mdoc, line, offs, ROFF_sp);
256 mdoc->last->flags |= NODE_VALID | NODE_ENDED;
257 mdoc->next = ROFF_NEXT_SIBLING;
261 roff_word_alloc(mdoc, line, offs, buf+offs);
263 if (mdoc->flags & ROFF_NOFILL)
267 * End-of-sentence check. If the last character is an unescaped
268 * EOS character, then flag the node as being the end of a
269 * sentence. The front-end will know how to interpret this.
274 if (mandoc_eos(buf+offs, (size_t)(end-buf-offs)))
275 mdoc->last->flags |= NODE_EOS;
277 for (c = buf + offs; c != NULL; c = strchr(c + 1, '.')) {
278 if (c - buf < offs + 2)
283 isalnum((unsigned char)c[-2]) == 0 ||
284 isalnum((unsigned char)c[-1]) == 0 ||
285 (c[-2] == 'n' && c[-1] == 'c') ||
286 (c[-2] == 'v' && c[-1] == 's'))
293 if (isupper((unsigned char)(*c)))
294 mandoc_msg(MANDOCERR_EOS, line, (int)(c - buf), NULL);
301 * Parse a macro line, that is, a line beginning with the control
305 mdoc_pmacro(struct roff_man *mdoc, int ln, char *buf, int offs)
313 /* Determine the line macro. */
317 for (sz = 0; sz < 4 && strchr(" \t\\", buf[offs]) == NULL; sz++)
319 if (sz == 2 || sz == 3)
320 tok = roffhash_find(mdoc->mdocmac, buf + sv, sz);
321 if (tok == TOKEN_NONE) {
322 mandoc_msg(MANDOCERR_MACRO, ln, sv, "%s", buf + sv - 1);
326 /* Skip a leading escape sequence or tab. */
331 mandoc_escape(&cp, NULL, NULL);
341 /* Jump to the next non-whitespace word. */
343 while (buf[offs] == ' ')
347 * Trailing whitespace. Note that tabs are allowed to be passed
348 * into the parser as "text", so we only warn about spaces here.
351 if ('\0' == buf[offs] && ' ' == buf[offs - 1])
352 mandoc_msg(MANDOCERR_SPACE_EOL, ln, offs - 1, NULL);
355 * If an initial macro or a list invocation, divert directly
356 * into macro processing.
360 if (n == NULL || tok == MDOC_It || tok == MDOC_El) {
361 (*mdoc_macro(tok)->fp)(mdoc, tok, ln, sv, &offs, buf);
366 * If a column list contains a non-It macro, assume an implicit
367 * item macro. This can happen one or more times at the
368 * beginning of such a list, intermixed with text lines and
369 * with nodes generated on the roff level, for example by tbl.
372 if ((n->tok == MDOC_Bl && n->type == ROFFT_BODY &&
373 n->end == ENDBODY_NOT && n->norm->Bl.type == LIST_column) ||
374 (n->parent != NULL && n->parent->tok == MDOC_Bl &&
375 n->parent->norm->Bl.type == LIST_column)) {
376 mdoc->flags |= MDOC_FREECOL;
377 (*mdoc_macro(MDOC_It)->fp)(mdoc, MDOC_It, ln, sv, &sv, buf);
381 /* Normal processing of a macro. */
383 (*mdoc_macro(tok)->fp)(mdoc, tok, ln, sv, &offs, buf);
385 /* In quick mode (for mandocdb), abort after the NAME section. */
387 if (mdoc->quick && MDOC_Sh == tok &&
388 SEC_NAME != mdoc->last->sec)
395 mdoc_isdelim(const char *p)
424 if (0 == strcmp(p + 1, "."))
426 if (0 == strcmp(p + 1, "fR|\\fP"))