1 /* $Id: mdoc.c,v 1.275 2020/04/06 10:16:17 schwarze Exp $ */
3 * Copyright (c) 2010, 2012-2018, 2020 Ingo Schwarze <schwarze@openbsd.org>
4 * Copyright (c) 2008, 2009, 2010, 2011 Kristaps Dzonsons <kristaps@bsd.lv>
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.
18 * Top level and utility functions of the mdoc(7) parser for mandoc(1).
22 #include <sys/types.h>
32 #include "mandoc_aux.h"
36 #include "libmandoc.h"
40 const char *const __mdoc_argnames[MDOC_ARG_MAX] = {
41 "split", "nosplit", "ragged",
42 "unfilled", "literal", "file",
43 "offset", "bullet", "dash",
44 "hyphen", "item", "enum",
45 "tag", "diag", "hang",
46 "ohang", "inset", "column",
47 "width", "compact", "std",
48 "filled", "words", "emphasis",
49 "symbolic", "nested", "centered"
51 const char * const *mdoc_argnames = __mdoc_argnames;
53 static int mdoc_ptext(struct roff_man *, int, char *, int);
54 static int mdoc_pmacro(struct roff_man *, int, char *, int);
58 * Main parse routine. Parses a single line -- really just hands off to
59 * the macro (mdoc_pmacro()) or text parser (mdoc_ptext()).
62 mdoc_parseln(struct roff_man *mdoc, int ln, char *buf, int offs)
65 if (mdoc->last->type != ROFFT_EQN || ln > mdoc->last->line)
66 mdoc->flags |= MDOC_NEWLINE;
69 * Let the roff nS register switch SYNOPSIS mode early,
70 * such that the parser knows at all times
71 * whether this mode is on or off.
72 * Note that this mode is also switched by the Sh macro.
74 if (roff_getreg(mdoc->roff, "nS"))
75 mdoc->flags |= MDOC_SYNOPSIS;
77 mdoc->flags &= ~MDOC_SYNOPSIS;
79 return roff_getcontrol(mdoc->roff, buf, &offs) ?
80 mdoc_pmacro(mdoc, ln, buf, offs) :
81 mdoc_ptext(mdoc, ln, buf, offs);
85 mdoc_tail_alloc(struct roff_man *mdoc, int line, int pos, enum roff_tok tok)
89 p = roff_node_alloc(mdoc, line, pos, ROFFT_TAIL, tok);
90 roff_node_append(mdoc, p);
91 mdoc->next = ROFF_NEXT_CHILD;
95 mdoc_endbody_alloc(struct roff_man *mdoc, int line, int pos,
96 enum roff_tok tok, struct roff_node *body)
100 body->flags |= NODE_ENDED;
101 body->parent->flags |= NODE_ENDED;
102 p = roff_node_alloc(mdoc, line, pos, ROFFT_BODY, tok);
104 p->norm = body->norm;
105 p->end = ENDBODY_SPACE;
106 roff_node_append(mdoc, p);
107 mdoc->next = ROFF_NEXT_SIBLING;
112 mdoc_block_alloc(struct roff_man *mdoc, int line, int pos,
113 enum roff_tok tok, struct mdoc_arg *args)
117 p = roff_node_alloc(mdoc, line, pos, ROFFT_BLOCK, tok);
128 p->norm = mandoc_calloc(1, sizeof(union mdoc_data));
133 roff_node_append(mdoc, p);
134 mdoc->next = ROFF_NEXT_CHILD;
139 mdoc_elem_alloc(struct roff_man *mdoc, int line, int pos,
140 enum roff_tok tok, struct mdoc_arg *args)
144 p = roff_node_alloc(mdoc, line, pos, ROFFT_ELEM, tok);
151 p->norm = mandoc_calloc(1, sizeof(union mdoc_data));
156 roff_node_append(mdoc, p);
157 mdoc->next = ROFF_NEXT_CHILD;
161 * Parse free-form text, that is, a line that does not begin with the
165 mdoc_ptext(struct roff_man *mdoc, int line, char *buf, int offs)
174 * If a column list contains plain text, assume an implicit item
175 * macro. This can happen one or more times at the beginning
176 * of such a list, intermixed with non-It mdoc macros and with
177 * nodes generated on the roff level, for example by tbl.
180 if ((n->tok == MDOC_Bl && n->type == ROFFT_BODY &&
181 n->end == ENDBODY_NOT && n->norm->Bl.type == LIST_column) ||
182 (n->parent != NULL && n->parent->tok == MDOC_Bl &&
183 n->parent->norm->Bl.type == LIST_column)) {
184 mdoc->flags |= MDOC_FREECOL;
185 (*mdoc_macro(MDOC_It)->fp)(mdoc, MDOC_It,
186 line, offs, &offs, buf);
191 * Search for the beginning of unescaped trailing whitespace (ws)
192 * and for the first character not to be output (end).
195 /* FIXME: replace with strcspn(). */
197 for (c = end = buf + offs; *c; c++) {
205 * Always warn about trailing tabs,
206 * even outside literal context,
207 * where they should be put on the next line.
212 * Strip trailing tabs in literal context only;
213 * outside, they affect the next line.
215 if (mdoc->flags & ROFF_NOFILL)
219 /* Skip the escaped character, too, if any. */
232 mandoc_msg(MANDOCERR_SPACE_EOL, line, (int)(ws - buf), NULL);
235 * Blank lines are allowed in no-fill mode
236 * and cancel preceding \c,
237 * but add a single vertical space elsewhere.
240 if (buf[offs] == '\0' && (mdoc->flags & ROFF_NOFILL) == 0) {
241 switch (mdoc->last->type) {
243 sp = mdoc->last->string;
244 cp = end = strchr(sp, '\0') - 2;
245 if (cp < sp || cp[0] != '\\' || cp[1] != 'c')
247 while (cp > sp && cp[-1] == '\\')
256 mandoc_msg(MANDOCERR_FI_BLANK, line, (int)(c - buf), NULL);
257 roff_elem_alloc(mdoc, line, offs, ROFF_sp);
258 mdoc->last->flags |= NODE_VALID | NODE_ENDED;
259 mdoc->next = ROFF_NEXT_SIBLING;
263 roff_word_alloc(mdoc, line, offs, buf+offs);
265 if (mdoc->flags & ROFF_NOFILL)
269 * End-of-sentence check. If the last character is an unescaped
270 * EOS character, then flag the node as being the end of a
271 * sentence. The front-end will know how to interpret this.
276 if (mandoc_eos(buf+offs, (size_t)(end-buf-offs)))
277 mdoc->last->flags |= NODE_EOS;
279 for (c = buf + offs; c != NULL; c = strchr(c + 1, '.')) {
280 if (c - buf < offs + 2)
285 isalnum((unsigned char)c[-2]) == 0 ||
286 isalnum((unsigned char)c[-1]) == 0 ||
287 (c[-2] == 'n' && c[-1] == 'c') ||
288 (c[-2] == 'v' && c[-1] == 's'))
295 if (isupper((unsigned char)(*c)))
296 mandoc_msg(MANDOCERR_EOS, line, (int)(c - buf), NULL);
303 * Parse a macro line, that is, a line beginning with the control
307 mdoc_pmacro(struct roff_man *mdoc, int ln, char *buf, int offs)
315 /* Determine the line macro. */
319 for (sz = 0; sz < 4 && strchr(" \t\\", buf[offs]) == NULL; sz++)
321 if (sz == 2 || sz == 3)
322 tok = roffhash_find(mdoc->mdocmac, buf + sv, sz);
323 if (tok == TOKEN_NONE) {
324 mandoc_msg(MANDOCERR_MACRO, ln, sv, "%s", buf + sv - 1);
328 /* Skip a leading escape sequence or tab. */
333 mandoc_escape(&cp, NULL, NULL);
343 /* Jump to the next non-whitespace word. */
345 while (buf[offs] == ' ')
349 * Trailing whitespace. Note that tabs are allowed to be passed
350 * into the parser as "text", so we only warn about spaces here.
353 if ('\0' == buf[offs] && ' ' == buf[offs - 1])
354 mandoc_msg(MANDOCERR_SPACE_EOL, ln, offs - 1, NULL);
357 * If an initial or transparent macro or a list invocation,
358 * divert directly into macro processing.
362 if (n == NULL || tok == MDOC_It || tok == MDOC_El ||
363 roff_tok_transparent(tok)) {
364 (*mdoc_macro(tok)->fp)(mdoc, tok, ln, sv, &offs, buf);
369 * If a column list contains a non-It macro, assume an implicit
370 * item macro. This can happen one or more times at the
371 * beginning of such a list, intermixed with text lines and
372 * with nodes generated on the roff level, for example by tbl.
375 if ((n->tok == MDOC_Bl && n->type == ROFFT_BODY &&
376 n->end == ENDBODY_NOT && n->norm->Bl.type == LIST_column) ||
377 (n->parent != NULL && n->parent->tok == MDOC_Bl &&
378 n->parent->norm->Bl.type == LIST_column)) {
379 mdoc->flags |= MDOC_FREECOL;
380 (*mdoc_macro(MDOC_It)->fp)(mdoc, MDOC_It, ln, sv, &sv, buf);
384 /* Normal processing of a macro. */
386 (*mdoc_macro(tok)->fp)(mdoc, tok, ln, sv, &offs, buf);
388 /* In quick mode (for mandocdb), abort after the NAME section. */
390 if (mdoc->quick && MDOC_Sh == tok &&
391 SEC_NAME != mdoc->last->sec)
398 mdoc_isdelim(const char *p)
427 if (0 == strcmp(p + 1, "."))
429 if (0 == strcmp(p + 1, "fR|\\fP"))