contrib/mdocml/tbl_data.c

   1 /*      $Id: tbl_data.c,v 1.31 2014/04/23 16:08:33 schwarze Exp $ */
   2 /*
   3  * Copyright (c) 2009, 2010, 2011 Kristaps Dzonsons <kristaps@bsd.lv>
   4  * Copyright (c) 2011 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 "mandoc_aux.h"
  30 #include "libmandoc.h"
  31 #include "libroff.h"
  32
  33 static  int              getdata(struct tbl_node *, struct tbl_span *,
  34                                 int, const char *, int *);
  35 static  struct tbl_span *newspan(struct tbl_node *, int,
  36                                 struct tbl_row *);
  37
  38
  39 static int
  40 getdata(struct tbl_node *tbl, struct tbl_span *dp,
  41                 int ln, const char *p, int *pos)
  42 {
  43         struct tbl_dat  *dat;
  44         struct tbl_cell *cp;
  45         int              sv, spans;
  46
  47         cp = NULL;
  48         if (dp->last && dp->last->layout)
  49                 cp = dp->last->layout->next;
  50         else if (NULL == dp->last)
  51                 cp = dp->layout->first;
  52
  53         /*
  54          * Skip over spanners, since
  55          * we want to match data with data layout cells in the header.
  56          */
  57
  58         while (cp && TBL_CELL_SPAN == cp->pos)
  59                 cp = cp->next;
  60
  61         /*
  62          * Stop processing when we reach the end of the available layout
  63          * cells.  This means that we have extra input.
  64          */
  65
  66         if (NULL == cp) {
  67                 mandoc_msg(MANDOCERR_TBLEXTRADAT, tbl->parse,
  68                     ln, *pos, NULL);
  69                 /* Skip to the end... */
  70                 while (p[*pos])
  71                         (*pos)++;
  72                 return(1);
  73         }
  74
  75         dat = mandoc_calloc(1, sizeof(struct tbl_dat));
  76         dat->layout = cp;
  77         dat->pos = TBL_DATA_NONE;
  78
  79         assert(TBL_CELL_SPAN != cp->pos);
  80
  81         for (spans = 0, cp = cp->next; cp; cp = cp->next)
  82                 if (TBL_CELL_SPAN == cp->pos)
  83                         spans++;
  84                 else
  85                         break;
  86
  87         dat->spans = spans;
  88
  89         if (dp->last) {
  90                 dp->last->next = dat;
  91                 dp->last = dat;
  92         } else
  93                 dp->last = dp->first = dat;
  94
  95         sv = *pos;
  96         while (p[*pos] && p[*pos] != tbl->opts.tab)
  97                 (*pos)++;
  98
  99         /*
 100          * Check for a continued-data scope opening.  This consists of a
 101          * trailing `T{' at the end of the line.  Subsequent lines,
 102          * until a standalone `T}', are included in our cell.
 103          */
 104
 105         if (*pos - sv == 2 && 'T' == p[sv] && '{' == p[sv + 1]) {
 106                 tbl->part = TBL_PART_CDATA;
 107                 return(1);
 108         }
 109
 110         assert(*pos - sv >= 0);
 111
 112         dat->string = mandoc_malloc((size_t)(*pos - sv + 1));
 113         memcpy(dat->string, &p[sv], (size_t)(*pos - sv));
 114         dat->string[*pos - sv] = '\0';
 115
 116         if (p[*pos])
 117                 (*pos)++;
 118
 119         if ( ! strcmp(dat->string, "_"))
 120                 dat->pos = TBL_DATA_HORIZ;
 121         else if ( ! strcmp(dat->string, "="))
 122                 dat->pos = TBL_DATA_DHORIZ;
 123         else if ( ! strcmp(dat->string, "\\_"))
 124                 dat->pos = TBL_DATA_NHORIZ;
 125         else if ( ! strcmp(dat->string, "\\="))
 126                 dat->pos = TBL_DATA_NDHORIZ;
 127         else
 128                 dat->pos = TBL_DATA_DATA;
 129
 130         if (TBL_CELL_HORIZ == dat->layout->pos ||
 131             TBL_CELL_DHORIZ == dat->layout->pos ||
 132             TBL_CELL_DOWN == dat->layout->pos)
 133                 if (TBL_DATA_DATA == dat->pos && '\0' != *dat->string)
 134                         mandoc_msg(MANDOCERR_TBLIGNDATA,
 135                             tbl->parse, ln, sv, NULL);
 136
 137         return(1);
 138 }
 139
 140 int
 141 tbl_cdata(struct tbl_node *tbl, int ln, const char *p)
 142 {
 143         struct tbl_dat  *dat;
 144         size_t           sz;
 145         int              pos;
 146
 147         pos = 0;
 148
 149         dat = tbl->last_span->last;
 150
 151         if (p[pos] == 'T' && p[pos + 1] == '}') {
 152                 pos += 2;
 153                 if (p[pos] == tbl->opts.tab) {
 154                         tbl->part = TBL_PART_DATA;
 155                         pos++;
 156                         return(getdata(tbl, tbl->last_span, ln, p, &pos));
 157                 } else if ('\0' == p[pos]) {
 158                         tbl->part = TBL_PART_DATA;
 159                         return(1);
 160                 }
 161
 162                 /* Fallthrough: T} is part of a word. */
 163         }
 164
 165         dat->pos = TBL_DATA_DATA;
 166
 167         if (dat->string) {
 168                 sz = strlen(p) + strlen(dat->string) + 2;
 169                 dat->string = mandoc_realloc(dat->string, sz);
 170                 (void)strlcat(dat->string, " ", sz);
 171                 (void)strlcat(dat->string, p, sz);
 172         } else
 173                 dat->string = mandoc_strdup(p);
 174
 175         if (TBL_CELL_DOWN == dat->layout->pos)
 176                 mandoc_msg(MANDOCERR_TBLIGNDATA, tbl->parse,
 177                     ln, pos, NULL);
 178
 179         return(0);
 180 }
 181
 182 static struct tbl_span *
 183 newspan(struct tbl_node *tbl, int line, struct tbl_row *rp)
 184 {
 185         struct tbl_span *dp;
 186
 187         dp = mandoc_calloc(1, sizeof(struct tbl_span));
 188         dp->line = line;
 189         dp->opts = &tbl->opts;
 190         dp->layout = rp;
 191         dp->head = tbl->first_head;
 192
 193         if (tbl->last_span) {
 194                 tbl->last_span->next = dp;
 195                 tbl->last_span = dp;
 196         } else {
 197                 tbl->last_span = tbl->first_span = dp;
 198                 tbl->current_span = NULL;
 199                 dp->flags |= TBL_SPAN_FIRST;
 200         }
 201
 202         return(dp);
 203 }
 204
 205 int
 206 tbl_data(struct tbl_node *tbl, int ln, const char *p)
 207 {
 208         struct tbl_span *dp;
 209         struct tbl_row  *rp;
 210         int              pos;
 211
 212         pos = 0;
 213
 214         if ('\0' == p[pos]) {
 215                 mandoc_msg(MANDOCERR_TBL, tbl->parse, ln, pos, NULL);
 216                 return(0);
 217         }
 218
 219         /*
 220          * Choose a layout row: take the one following the last parsed
 221          * span's.  If that doesn't exist, use the last parsed span's.
 222          * If there's no last parsed span, use the first row.  Lastly,
 223          * if the last span was a horizontal line, use the same layout
 224          * (it doesn't "consume" the layout).
 225          */
 226
 227         if (tbl->last_span) {
 228                 assert(tbl->last_span->layout);
 229                 if (tbl->last_span->pos == TBL_SPAN_DATA) {
 230                         for (rp = tbl->last_span->layout->next;
 231                                         rp && rp->first; rp = rp->next) {
 232                                 switch (rp->first->pos) {
 233                                 case TBL_CELL_HORIZ:
 234                                         dp = newspan(tbl, ln, rp);
 235                                         dp->pos = TBL_SPAN_HORIZ;
 236                                         continue;
 237                                 case TBL_CELL_DHORIZ:
 238                                         dp = newspan(tbl, ln, rp);
 239                                         dp->pos = TBL_SPAN_DHORIZ;
 240                                         continue;
 241                                 default:
 242                                         break;
 243                                 }
 244                                 break;
 245                         }
 246                 } else
 247                         rp = tbl->last_span->layout;
 248
 249                 if (NULL == rp)
 250                         rp = tbl->last_span->layout;
 251         } else
 252                 rp = tbl->first_row;
 253
 254         assert(rp);
 255
 256         dp = newspan(tbl, ln, rp);
 257
 258         if ( ! strcmp(p, "_")) {
 259                 dp->pos = TBL_SPAN_HORIZ;
 260                 return(1);
 261         } else if ( ! strcmp(p, "=")) {
 262                 dp->pos = TBL_SPAN_DHORIZ;
 263                 return(1);
 264         }
 265
 266         dp->pos = TBL_SPAN_DATA;
 267
 268         /* This returns 0 when TBL_PART_CDATA is entered. */
 269
 270         while ('\0' != p[pos])
 271                 if ( ! getdata(tbl, dp, ln, p, &pos))
 272                         return(0);
 273
 274         return(1);
 275 }