man.c

   1 /*      $Id: man.c,v 1.174 2017/06/03 15:55:24 schwarze Exp $ */
   2 /*
   3  * Copyright (c) 2008, 2009, 2010, 2011 Kristaps Dzonsons <kristaps@bsd.lv>
   4  * Copyright (c) 2013, 2014, 2015, 2017 Ingo Schwarze <schwarze@openbsd.org>
   5  * Copyright (c) 2011 Joerg Sonnenberger <joerg@netbsd.org>
   6  *
   7  * Permission to use, copy, modify, and distribute this software for any
   8  * purpose with or without fee is hereby granted, provided that the above
   9  * copyright notice and this permission notice appear in all copies.
  10  *
  11  * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHORS DISCLAIM ALL WARRANTIES
  12  * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
  13  * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHORS BE LIABLE FOR
  14  * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
  15  * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
  16  * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
  17  * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
  18  */
  19 #include "config.h"
  20
  21 #include <sys/types.h>
  22
  23 #include <assert.h>
  24 #include <ctype.h>
  25 #include <stdarg.h>
  26 #include <stdlib.h>
  27 #include <stdio.h>
  28 #include <string.h>
  29
  30 #include "mandoc_aux.h"
  31 #include "mandoc.h"
  32 #include "roff.h"
  33 #include "man.h"
  34 #include "libmandoc.h"
  35 #include "roff_int.h"
  36 #include "libman.h"
  37
  38 static  void             man_descope(struct roff_man *, int, int);
  39 static  int              man_ptext(struct roff_man *, int, char *, int);
  40 static  int              man_pmacro(struct roff_man *, int, char *, int);
  41
  42
  43 int
  44 man_parseln(struct roff_man *man, int ln, char *buf, int offs)
  45 {
  46
  47         if (man->last->type != ROFFT_EQN || ln > man->last->line)
  48                 man->flags |= MAN_NEWLINE;
  49
  50         return roff_getcontrol(man->roff, buf, &offs) ?
  51             man_pmacro(man, ln, buf, offs) :
  52             man_ptext(man, ln, buf, offs);
  53 }
  54
  55 static void
  56 man_descope(struct roff_man *man, int line, int offs)
  57 {
  58         /*
  59          * Co-ordinate what happens with having a next-line scope open:
  60          * first close out the element scope (if applicable), then close
  61          * out the block scope (also if applicable).
  62          */
  63
  64         if (man->flags & MAN_ELINE) {
  65                 man->flags &= ~MAN_ELINE;
  66                 man_unscope(man, man->last->parent);
  67         }
  68         if ( ! (man->flags & MAN_BLINE))
  69                 return;
  70         man->flags &= ~MAN_BLINE;
  71         man_unscope(man, man->last->parent);
  72         roff_body_alloc(man, line, offs, man->last->tok);
  73 }
  74
  75 static int
  76 man_ptext(struct roff_man *man, int line, char *buf, int offs)
  77 {
  78         int              i;
  79
  80         /* Literal free-form text whitespace is preserved. */
  81
  82         if (man->flags & MAN_LITERAL) {
  83                 roff_word_alloc(man, line, offs, buf + offs);
  84                 man_descope(man, line, offs);
  85                 return 1;
  86         }
  87
  88         for (i = offs; buf[i] == ' '; i++)
  89                 /* Skip leading whitespace. */ ;
  90
  91         /*
  92          * Blank lines are ignored in next line scope and right
  93          * after headings but add a single vertical space elsewhere.
  94          */
  95
  96         if (buf[i] == '\0') {
  97                 if (man->flags & (MAN_ELINE | MAN_BLINE))
  98                         mandoc_msg(MANDOCERR_BLK_BLANK, man->parse,
  99                             line, 0, NULL);
 100                 else if (man->last->tok != MAN_SH &&
 101                     man->last->tok != MAN_SS) {
 102                         roff_elem_alloc(man, line, offs, ROFF_sp);
 103                         man->next = ROFF_NEXT_SIBLING;
 104                 }
 105                 return 1;
 106         }
 107
 108         /*
 109          * Warn if the last un-escaped character is whitespace. Then
 110          * strip away the remaining spaces (tabs stay!).
 111          */
 112
 113         i = (int)strlen(buf);
 114         assert(i);
 115
 116         if (' ' == buf[i - 1] || '\t' == buf[i - 1]) {
 117                 if (i > 1 && '\\' != buf[i - 2])
 118                         mandoc_msg(MANDOCERR_SPACE_EOL, man->parse,
 119                             line, i - 1, NULL);
 120
 121                 for (--i; i && ' ' == buf[i]; i--)
 122                         /* Spin back to non-space. */ ;
 123
 124                 /* Jump ahead of escaped whitespace. */
 125                 i += '\\' == buf[i] ? 2 : 1;
 126
 127                 buf[i] = '\0';
 128         }
 129         roff_word_alloc(man, line, offs, buf + offs);
 130
 131         /*
 132          * End-of-sentence check.  If the last character is an unescaped
 133          * EOS character, then flag the node as being the end of a
 134          * sentence.  The front-end will know how to interpret this.
 135          */
 136
 137         assert(i);
 138         if (mandoc_eos(buf, (size_t)i))
 139                 man->last->flags |= NODE_EOS;
 140
 141         man_descope(man, line, offs);
 142         return 1;
 143 }
 144
 145 static int
 146 man_pmacro(struct roff_man *man, int ln, char *buf, int offs)
 147 {
 148         struct roff_node *n;
 149         const char      *cp;
 150         size_t           sz;
 151         enum roff_tok    tok;
 152         int              ppos;
 153         int              bline;
 154
 155         /* Determine the line macro. */
 156
 157         ppos = offs;
 158         tok = TOKEN_NONE;
 159         for (sz = 0; sz < 4 && strchr(" \t\\", buf[offs]) == NULL; sz++)
 160                 offs++;
 161         if (sz > 0 && sz < 4)
 162                 tok = roffhash_find(man->manmac, buf + ppos, sz);
 163         if (tok == TOKEN_NONE) {
 164                 mandoc_msg(MANDOCERR_MACRO, man->parse,
 165                     ln, ppos, buf + ppos - 1);
 166                 return 1;
 167         }
 168
 169         /* Skip a leading escape sequence or tab. */
 170
 171         switch (buf[offs]) {
 172         case '\\':
 173                 cp = buf + offs + 1;
 174                 mandoc_escape(&cp, NULL, NULL);
 175                 offs = cp - buf;
 176                 break;
 177         case '\t':
 178                 offs++;
 179                 break;
 180         default:
 181                 break;
 182         }
 183
 184         /* Jump to the next non-whitespace word. */
 185
 186         while (buf[offs] == ' ')
 187                 offs++;
 188
 189         /*
 190          * Trailing whitespace.  Note that tabs are allowed to be passed
 191          * into the parser as "text", so we only warn about spaces here.
 192          */
 193
 194         if (buf[offs] == '\0' && buf[offs - 1] == ' ')
 195                 mandoc_msg(MANDOCERR_SPACE_EOL, man->parse,
 196                     ln, offs - 1, NULL);
 197
 198         /*
 199          * Some macros break next-line scopes; otherwise, remember
 200          * whether we are in next-line scope for a block head.
 201          */
 202
 203         man_breakscope(man, tok);
 204         bline = man->flags & MAN_BLINE;
 205
 206         /*
 207          * If the line in next-line scope ends with \c, keep the
 208          * next-line scope open for the subsequent input line.
 209          * That is not at all portable, only groff >= 1.22.4
 210          * does it, but *if* this weird idiom occurs in a manual
 211          * page, that's very likely what the author intended.
 212          */
 213
 214         if (bline) {
 215                 cp = strchr(buf + offs, '\0') - 2;
 216                 if (cp >= buf && cp[0] == '\\' && cp[1] == 'c')
 217                         bline = 0;
 218         }
 219
 220         /* Call to handler... */
 221
 222         assert(man_macros[tok].fp);
 223         (*man_macros[tok].fp)(man, tok, ln, ppos, &offs, buf);
 224
 225         /* In quick mode (for mandocdb), abort after the NAME section. */
 226
 227         if (man->quick && tok == MAN_SH) {
 228                 n = man->last;
 229                 if (n->type == ROFFT_BODY &&
 230                     strcmp(n->prev->child->string, "NAME"))
 231                         return 2;
 232         }
 233
 234         /*
 235          * If we are in a next-line scope for a block head,
 236          * close it out now and switch to the body,
 237          * unless the next-line scope is allowed to continue.
 238          */
 239
 240         if ( ! bline || man->flags & MAN_ELINE ||
 241             man_macros[tok].flags & MAN_NSCOPED)
 242                 return 1;
 243
 244         assert(man->flags & MAN_BLINE);
 245         man->flags &= ~MAN_BLINE;
 246
 247         man_unscope(man, man->last->parent);
 248         roff_body_alloc(man, ln, ppos, man->last->tok);
 249         return 1;
 250 }
 251
 252 void
 253 man_breakscope(struct roff_man *man, int tok)
 254 {
 255         struct roff_node *n;
 256
 257         /*
 258          * An element next line scope is open,
 259          * and the new macro is not allowed inside elements.
 260          * Delete the element that is being broken.
 261          */
 262
 263         if (man->flags & MAN_ELINE && (tok < MAN_TH ||
 264             ! (man_macros[tok].flags & MAN_NSCOPED))) {
 265                 n = man->last;
 266                 assert(n->type != ROFFT_TEXT);
 267                 if (man_macros[n->tok].flags & MAN_NSCOPED)
 268                         n = n->parent;
 269
 270                 mandoc_vmsg(MANDOCERR_BLK_LINE, man->parse,
 271                     n->line, n->pos, "%s breaks %s",
 272                     roff_name[tok], roff_name[n->tok]);
 273
 274                 roff_node_delete(man, n);
 275                 man->flags &= ~MAN_ELINE;
 276         }
 277
 278         /*
 279          * Weird special case:
 280          * Switching fill mode closes section headers.
 281          */
 282
 283         if (man->flags & MAN_BLINE &&
 284             (tok == MAN_nf || tok == MAN_fi) &&
 285             (man->last->tok == MAN_SH || man->last->tok == MAN_SS)) {
 286                 n = man->last;
 287                 man_unscope(man, n);
 288                 roff_body_alloc(man, n->line, n->pos, n->tok);
 289                 man->flags &= ~MAN_BLINE;
 290         }
 291
 292         /*
 293          * A block header next line scope is open,
 294          * and the new macro is not allowed inside block headers.
 295          * Delete the block that is being broken.
 296          */
 297
 298         if (man->flags & MAN_BLINE && (tok < MAN_TH ||
 299             man_macros[tok].flags & MAN_BSCOPE)) {
 300                 n = man->last;
 301                 if (n->type == ROFFT_TEXT)
 302                         n = n->parent;
 303                 if ( ! (man_macros[n->tok].flags & MAN_BSCOPE))
 304                         n = n->parent;
 305
 306                 assert(n->type == ROFFT_HEAD);
 307                 n = n->parent;
 308                 assert(n->type == ROFFT_BLOCK);
 309                 assert(man_macros[n->tok].flags & MAN_SCOPED);
 310
 311                 mandoc_vmsg(MANDOCERR_BLK_LINE, man->parse,
 312                     n->line, n->pos, "%s breaks %s",
 313                     roff_name[tok], roff_name[n->tok]);
 314
 315                 roff_node_delete(man, n);
 316                 man->flags &= ~MAN_BLINE;
 317         }
 318 }
 319
 320 const struct mparse *
 321 man_mparse(const struct roff_man *man)
 322 {
 323
 324         assert(man && man->parse);
 325         return man->parse;
 326 }
 327
 328 void
 329 man_state(struct roff_man *man, struct roff_node *n)
 330 {
 331
 332         switch(n->tok) {
 333         case MAN_nf:
 334         case MAN_EX:
 335                 if (man->flags & MAN_LITERAL && ! (n->flags & NODE_VALID))
 336                         mandoc_msg(MANDOCERR_NF_SKIP, man->parse,
 337                             n->line, n->pos, "nf");
 338                 man->flags |= MAN_LITERAL;
 339                 break;
 340         case MAN_fi:
 341         case MAN_EE:
 342                 if ( ! (man->flags & MAN_LITERAL) &&
 343                      ! (n->flags & NODE_VALID))
 344                         mandoc_msg(MANDOCERR_FI_SKIP, man->parse,
 345                             n->line, n->pos, "fi");
 346                 man->flags &= ~MAN_LITERAL;
 347                 break;
 348         default:
 349                 break;
 350         }
 351         man->last->flags |= NODE_VALID;
 352 }
 353
 354 void
 355 man_validate(struct roff_man *man)
 356 {
 357
 358         man->last = man->first;
 359         man_node_validate(man);
 360         man->flags &= ~MAN_LITERAL;
 361 }