contrib/mdocml/tbl_layout.c

   1 /*      $Id: tbl_layout.c,v 1.23 2012/05/27 17:54:54 schwarze Exp $ */
   2 /*
   3  * Copyright (c) 2009, 2010, 2011 Kristaps Dzonsons <kristaps@bsd.lv>
   4  * Copyright (c) 2012 Ingo Schwarze <schwarze@openbsd.org>
   5  *
   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.
   9  *
  10  * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
  11  * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
  12  * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR 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.
  17  */
  18 #ifdef HAVE_CONFIG_H
  19 #include "config.h"
  20 #endif
  21
  22 #include <assert.h>
  23 #include <ctype.h>
  24 #include <stdlib.h>
  25 #include <string.h>
  26 #include <time.h>
  27
  28 #include "mandoc.h"
  29 #include "libmandoc.h"
  30 #include "libroff.h"
  31
  32 struct  tbl_phrase {
  33         char             name;
  34         enum tbl_cellt   key;
  35 };
  36
  37 /*
  38  * FIXME: we can make this parse a lot nicer by, when an error is
  39  * encountered in a layout key, bailing to the next key (i.e. to the
  40  * next whitespace then continuing).
  41  */
  42
  43 #define KEYS_MAX         11
  44
  45 static  const struct tbl_phrase keys[KEYS_MAX] = {
  46         { 'c',           TBL_CELL_CENTRE },
  47         { 'r',           TBL_CELL_RIGHT },
  48         { 'l',           TBL_CELL_LEFT },
  49         { 'n',           TBL_CELL_NUMBER },
  50         { 's',           TBL_CELL_SPAN },
  51         { 'a',           TBL_CELL_LONG },
  52         { '^',           TBL_CELL_DOWN },
  53         { '-',           TBL_CELL_HORIZ },
  54         { '_',           TBL_CELL_HORIZ },
  55         { '=',           TBL_CELL_DHORIZ }
  56 };
  57
  58 static  int              mods(struct tbl_node *, struct tbl_cell *,
  59                                 int, const char *, int *);
  60 static  int              cell(struct tbl_node *, struct tbl_row *,
  61                                 int, const char *, int *);
  62 static  void             row(struct tbl_node *, int, const char *, int *);
  63 static  struct tbl_cell *cell_alloc(struct tbl_node *, struct tbl_row *,
  64                                 enum tbl_cellt, int vert);
  65
  66 static int
  67 mods(struct tbl_node *tbl, struct tbl_cell *cp,
  68                 int ln, const char *p, int *pos)
  69 {
  70         char             buf[5];
  71         int              i;
  72
  73         /* Not all types accept modifiers. */
  74
  75         switch (cp->pos) {
  76         case (TBL_CELL_DOWN):
  77                 /* FALLTHROUGH */
  78         case (TBL_CELL_HORIZ):
  79                 /* FALLTHROUGH */
  80         case (TBL_CELL_DHORIZ):
  81                 return(1);
  82         default:
  83                 break;
  84         }
  85
  86 mod:
  87         /*
  88          * XXX: since, at least for now, modifiers are non-conflicting
  89          * (are separable by value, regardless of position), we let
  90          * modifiers come in any order.  The existing tbl doesn't let
  91          * this happen.
  92          */
  93         switch (p[*pos]) {
  94         case ('\0'):
  95                 /* FALLTHROUGH */
  96         case (' '):
  97                 /* FALLTHROUGH */
  98         case ('\t'):
  99                 /* FALLTHROUGH */
 100         case (','):
 101                 /* FALLTHROUGH */
 102         case ('.'):
 103                 return(1);
 104         default:
 105                 break;
 106         }
 107
 108         /* Throw away parenthesised expression. */
 109
 110         if ('(' == p[*pos]) {
 111                 (*pos)++;
 112                 while (p[*pos] && ')' != p[*pos])
 113                         (*pos)++;
 114                 if (')' == p[*pos]) {
 115                         (*pos)++;
 116                         goto mod;
 117                 }
 118                 mandoc_msg(MANDOCERR_TBLLAYOUT,
 119                                 tbl->parse, ln, *pos, NULL);
 120                 return(0);
 121         }
 122
 123         /* Parse numerical spacing from modifier string. */
 124
 125         if (isdigit((unsigned char)p[*pos])) {
 126                 for (i = 0; i < 4; i++) {
 127                         if ( ! isdigit((unsigned char)p[*pos + i]))
 128                                 break;
 129                         buf[i] = p[*pos + i];
 130                 }
 131                 buf[i] = '\0';
 132
 133                 /* No greater than 4 digits. */
 134
 135                 if (4 == i) {
 136                         mandoc_msg(MANDOCERR_TBLLAYOUT, tbl->parse,
 137                                         ln, *pos, NULL);
 138                         return(0);
 139                 }
 140
 141                 *pos += i;
 142                 cp->spacing = (size_t)atoi(buf);
 143
 144                 goto mod;
 145                 /* NOTREACHED */
 146         }
 147
 148         /* TODO: GNU has many more extensions. */
 149
 150         switch (tolower((unsigned char)p[(*pos)++])) {
 151         case ('z'):
 152                 cp->flags |= TBL_CELL_WIGN;
 153                 goto mod;
 154         case ('u'):
 155                 cp->flags |= TBL_CELL_UP;
 156                 goto mod;
 157         case ('e'):
 158                 cp->flags |= TBL_CELL_EQUAL;
 159                 goto mod;
 160         case ('t'):
 161                 cp->flags |= TBL_CELL_TALIGN;
 162                 goto mod;
 163         case ('d'):
 164                 cp->flags |= TBL_CELL_BALIGN;
 165                 goto mod;
 166         case ('w'):  /* XXX for now, ignore minimal column width */
 167                 goto mod;
 168         case ('f'):
 169                 break;
 170         case ('r'):
 171                 /* FALLTHROUGH */
 172         case ('b'):
 173                 /* FALLTHROUGH */
 174         case ('i'):
 175                 (*pos)--;
 176                 break;
 177         default:
 178                 mandoc_msg(MANDOCERR_TBLLAYOUT, tbl->parse,
 179                                 ln, *pos - 1, NULL);
 180                 return(0);
 181         }
 182
 183         switch (tolower((unsigned char)p[(*pos)++])) {
 184         case ('3'):
 185                 /* FALLTHROUGH */
 186         case ('b'):
 187                 cp->flags |= TBL_CELL_BOLD;
 188                 goto mod;
 189         case ('2'):
 190                 /* FALLTHROUGH */
 191         case ('i'):
 192                 cp->flags |= TBL_CELL_ITALIC;
 193                 goto mod;
 194         case ('1'):
 195                 /* FALLTHROUGH */
 196         case ('r'):
 197                 goto mod;
 198         default:
 199                 break;
 200         }
 201
 202         mandoc_msg(MANDOCERR_TBLLAYOUT, tbl->parse,
 203                         ln, *pos - 1, NULL);
 204         return(0);
 205 }
 206
 207 static int
 208 cell(struct tbl_node *tbl, struct tbl_row *rp,
 209                 int ln, const char *p, int *pos)
 210 {
 211         int              vert, i;
 212         enum tbl_cellt   c;
 213
 214         /* Handle vertical lines. */
 215
 216         for (vert = 0; '|' == p[*pos]; ++*pos)
 217                 vert++;
 218         while (' ' == p[*pos])
 219                 (*pos)++;
 220
 221         /* Parse the column position (`c', `l', `r', ...). */
 222
 223         for (i = 0; i < KEYS_MAX; i++)
 224                 if (tolower((unsigned char)p[*pos]) == keys[i].name)
 225                         break;
 226
 227         if (KEYS_MAX == i) {
 228                 mandoc_msg(MANDOCERR_TBLLAYOUT, tbl->parse,
 229                                 ln, *pos, NULL);
 230                 return(0);
 231         }
 232
 233         c = keys[i].key;
 234
 235         /*
 236          * If a span cell is found first, raise a warning and abort the
 237          * parse.  If a span cell is found and the last layout element
 238          * isn't a "normal" layout, bail.
 239          *
 240          * FIXME: recover from this somehow?
 241          */
 242
 243         if (TBL_CELL_SPAN == c) {
 244                 if (NULL == rp->first) {
 245                         mandoc_msg(MANDOCERR_TBLLAYOUT, tbl->parse,
 246                                         ln, *pos, NULL);
 247                         return(0);
 248                 } else if (rp->last)
 249                         switch (rp->last->pos) {
 250                         case (TBL_CELL_HORIZ):
 251                         case (TBL_CELL_DHORIZ):
 252                                 mandoc_msg(MANDOCERR_TBLLAYOUT, tbl->parse,
 253                                                 ln, *pos, NULL);
 254                                 return(0);
 255                         default:
 256                                 break;
 257                         }
 258         }
 259
 260         /*
 261          * If a vertical spanner is found, we may not be in the first
 262          * row.
 263          */
 264
 265         if (TBL_CELL_DOWN == c && rp == tbl->first_row) {
 266                 mandoc_msg(MANDOCERR_TBLLAYOUT, tbl->parse, ln, *pos, NULL);
 267                 return(0);
 268         }
 269
 270         (*pos)++;
 271
 272         /* Disallow adjacent spacers. */
 273
 274         if (vert > 2) {
 275                 mandoc_msg(MANDOCERR_TBLLAYOUT, tbl->parse, ln, *pos - 1, NULL);
 276                 return(0);
 277         }
 278
 279         /* Allocate cell then parse its modifiers. */
 280
 281         return(mods(tbl, cell_alloc(tbl, rp, c, vert), ln, p, pos));
 282 }
 283
 284
 285 static void
 286 row(struct tbl_node *tbl, int ln, const char *p, int *pos)
 287 {
 288         struct tbl_row  *rp;
 289
 290 row:    /*
 291          * EBNF describing this section:
 292          *
 293          * row          ::= row_list [:space:]* [.]?[\n]
 294          * row_list     ::= [:space:]* row_elem row_tail
 295          * row_tail     ::= [:space:]*[,] row_list |
 296          *                  epsilon
 297          * row_elem     ::= [\t\ ]*[:alpha:]+
 298          */
 299
 300         rp = mandoc_calloc(1, sizeof(struct tbl_row));
 301         if (tbl->last_row)
 302                 tbl->last_row->next = rp;
 303         else
 304                 tbl->first_row = rp;
 305         tbl->last_row = rp;
 306
 307 cell:
 308         while (isspace((unsigned char)p[*pos]))
 309                 (*pos)++;
 310
 311         /* Safely exit layout context. */
 312
 313         if ('.' == p[*pos]) {
 314                 tbl->part = TBL_PART_DATA;
 315                 if (NULL == tbl->first_row)
 316                         mandoc_msg(MANDOCERR_TBLNOLAYOUT, tbl->parse,
 317                                         ln, *pos, NULL);
 318                 (*pos)++;
 319                 return;
 320         }
 321
 322         /* End (and possibly restart) a row. */
 323
 324         if (',' == p[*pos]) {
 325                 (*pos)++;
 326                 goto row;
 327         } else if ('\0' == p[*pos])
 328                 return;
 329
 330         if ( ! cell(tbl, rp, ln, p, pos))
 331                 return;
 332
 333         goto cell;
 334         /* NOTREACHED */
 335 }
 336
 337 int
 338 tbl_layout(struct tbl_node *tbl, int ln, const char *p)
 339 {
 340         int              pos;
 341
 342         pos = 0;
 343         row(tbl, ln, p, &pos);
 344
 345         /* Always succeed. */
 346         return(1);
 347 }
 348
 349 static struct tbl_cell *
 350 cell_alloc(struct tbl_node *tbl, struct tbl_row *rp, enum tbl_cellt pos,
 351                 int vert)
 352 {
 353         struct tbl_cell *p, *pp;
 354         struct tbl_head *h, *hp;
 355
 356         p = mandoc_calloc(1, sizeof(struct tbl_cell));
 357
 358         if (NULL != (pp = rp->last)) {
 359                 pp->next = p;
 360                 h = pp->head->next;
 361         } else {
 362                 rp->first = p;
 363                 h = tbl->first_head;
 364         }
 365         rp->last = p;
 366
 367         p->pos = pos;
 368         p->vert = vert;
 369
 370         /* Re-use header. */
 371
 372         if (h) {
 373                 p->head = h;
 374                 return(p);
 375         }
 376
 377         hp = mandoc_calloc(1, sizeof(struct tbl_head));
 378         hp->ident = tbl->opts.cols++;
 379         hp->vert = vert;
 380
 381         if (tbl->last_head) {
 382                 hp->prev = tbl->last_head;
 383                 tbl->last_head->next = hp;
 384         } else
 385                 tbl->first_head = hp;
 386         tbl->last_head = hp;
 387
 388         p->head = hp;
 389         return(p);
 390 }