tbl_data.c

   1 /*      $Id: tbl_data.c,v 1.22 2011/02/06 22:02:58 kristaps 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 "libmandoc.h"
  30 #include "libroff.h"
  31
  32 static  int              data(struct tbl_node *, struct tbl_span *,
  33                                 int, const char *, int *);
  34 static  struct tbl_span *newspan(struct tbl_node *, int,
  35                                 struct tbl_row *);
  36
  37 static int
  38 data(struct tbl_node *tbl, struct tbl_span *dp,
  39                 int ln, const char *p, int *pos)
  40 {
  41         struct tbl_dat  *dat;
  42         struct tbl_cell *cp;
  43         int              sv, spans;
  44
  45         cp = NULL;
  46         if (dp->last && dp->last->layout)
  47                 cp = dp->last->layout->next;
  48         else if (NULL == dp->last)
  49                 cp = dp->layout->first;
  50
  51         /*
  52          * Skip over spanners and vertical lines to data formats, since
  53          * we want to match data with data layout cells in the header.
  54          */
  55
  56         while (cp && (TBL_CELL_VERT == cp->pos ||
  57                                 TBL_CELL_DVERT == cp->pos ||
  58                                 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                 TBL_MSG(tbl, MANDOCERR_TBLEXTRADAT, ln, *pos);
  68                 /* Skip to the end... */
  69                 while (p[*pos])
  70                         (*pos)++;
  71                 return(1);
  72         }
  73
  74         dat = mandoc_calloc(1, sizeof(struct tbl_dat));
  75         dat->layout = cp;
  76         dat->pos = TBL_DATA_NONE;
  77
  78         assert(TBL_CELL_SPAN != cp->pos);
  79
  80         for (spans = 0, cp = cp->next; cp; cp = cp->next)
  81                 if (TBL_CELL_SPAN == cp->pos)
  82                         spans++;
  83                 else
  84                         break;
  85
  86         dat->spans = spans;
  87
  88         if (dp->last) {
  89                 dp->last->next = dat;
  90                 dp->last = dat;
  91         } else
  92                 dp->last = dp->first = dat;
  93
  94         sv = *pos;
  95         while (p[*pos] && p[*pos] != tbl->opts.tab)
  96                 (*pos)++;
  97
  98         /*
  99          * Check for a continued-data scope opening.  This consists of a
 100          * trailing `T{' at the end of the line.  Subsequent lines,
 101          * until a standalone `T}', are included in our cell.
 102          */
 103
 104         if (*pos - sv == 2 && 'T' == p[sv] && '{' == p[sv + 1]) {
 105                 tbl->part = TBL_PART_CDATA;
 106                 return(0);
 107         }
 108
 109         dat->string = mandoc_malloc(*pos - sv + 1);
 110         memcpy(dat->string, &p[sv], *pos - sv);
 111         dat->string[*pos - sv] = '\0';
 112
 113         if (p[*pos])
 114                 (*pos)++;
 115
 116         if ( ! strcmp(dat->string, "_"))
 117                 dat->pos = TBL_DATA_HORIZ;
 118         else if ( ! strcmp(dat->string, "="))
 119                 dat->pos = TBL_DATA_DHORIZ;
 120         else if ( ! strcmp(dat->string, "\\_"))
 121                 dat->pos = TBL_DATA_NHORIZ;
 122         else if ( ! strcmp(dat->string, "\\="))
 123                 dat->pos = TBL_DATA_NDHORIZ;
 124         else
 125                 dat->pos = TBL_DATA_DATA;
 126
 127         if (TBL_CELL_HORIZ == dat->layout->pos ||
 128                         TBL_CELL_DHORIZ == dat->layout->pos ||
 129                         TBL_CELL_DOWN == dat->layout->pos)
 130                 if (TBL_DATA_DATA == dat->pos && '\0' != *dat->string)
 131                         TBL_MSG(tbl, MANDOCERR_TBLIGNDATA, ln, sv);
 132
 133         return(1);
 134 }
 135
 136 /* ARGSUSED */
 137 int
 138 tbl_cdata(struct tbl_node *tbl, int ln, const char *p)
 139 {
 140         struct tbl_dat  *dat;
 141         size_t           sz;
 142         int              pos;
 143
 144         pos = 0;
 145
 146         dat = tbl->last_span->last;
 147
 148         if (p[pos] == 'T' && p[pos + 1] == '}') {
 149                 pos += 2;
 150                 if (p[pos] == tbl->opts.tab) {
 151                         tbl->part = TBL_PART_DATA;
 152                         pos++;
 153                         return(data(tbl, tbl->last_span, ln, p, &pos));
 154                 } else if ('\0' == p[pos]) {
 155                         tbl->part = TBL_PART_DATA;
 156                         return(1);
 157                 }
 158
 159                 /* Fallthrough: T} is part of a word. */
 160         }
 161
 162         dat->pos = TBL_DATA_DATA;
 163
 164         if (dat->string) {
 165                 sz = strlen(p) + strlen(dat->string) + 2;
 166                 dat->string = mandoc_realloc(dat->string, sz);
 167                 strlcat(dat->string, " ", sz);
 168                 strlcat(dat->string, p, sz);
 169         } else
 170                 dat->string = mandoc_strdup(p);
 171
 172         if (TBL_CELL_DOWN == dat->layout->pos)
 173                 TBL_MSG(tbl, MANDOCERR_TBLIGNDATA, ln, pos);
 174
 175         return(0);
 176 }
 177
 178 static struct tbl_span *
 179 newspan(struct tbl_node *tbl, int line, struct tbl_row *rp)
 180 {
 181         struct tbl_span *dp;
 182
 183         dp = mandoc_calloc(1, sizeof(struct tbl_span));
 184         dp->line = line;
 185         dp->tbl = &tbl->opts;
 186         dp->layout = rp;
 187         dp->head = tbl->first_head;
 188
 189         if (tbl->last_span) {
 190                 tbl->last_span->next = dp;
 191                 tbl->last_span = dp;
 192         } else {
 193                 tbl->last_span = tbl->first_span = dp;
 194                 tbl->current_span = NULL;
 195                 dp->flags |= TBL_SPAN_FIRST;
 196         }
 197
 198         return(dp);
 199 }
 200
 201 int
 202 tbl_data(struct tbl_node *tbl, int ln, const char *p)
 203 {
 204         struct tbl_span *dp;
 205         struct tbl_row  *rp;
 206         int              pos;
 207
 208         pos = 0;
 209
 210         if ('\0' == p[pos]) {
 211                 TBL_MSG(tbl, MANDOCERR_TBL, ln, pos);
 212                 return(0);
 213         }
 214
 215         /*
 216          * Choose a layout row: take the one following the last parsed
 217          * span's.  If that doesn't exist, use the last parsed span's.
 218          * If there's no last parsed span, use the first row.  Lastly,
 219          * if the last span was a horizontal line, use the same layout
 220          * (it doesn't "consume" the layout).
 221          */
 222
 223         if (tbl->last_span) {
 224                 assert(tbl->last_span->layout);
 225                 if (tbl->last_span->pos == TBL_SPAN_DATA) {
 226                         for (rp = tbl->last_span->layout->next;
 227                                         rp && rp->first; rp = rp->next) {
 228                                 switch (rp->first->pos) {
 229                                 case (TBL_CELL_HORIZ):
 230                                         dp = newspan(tbl, ln, rp);
 231                                         dp->pos = TBL_SPAN_HORIZ;
 232                                         continue;
 233                                 case (TBL_CELL_DHORIZ):
 234                                         dp = newspan(tbl, ln, rp);
 235                                         dp->pos = TBL_SPAN_DHORIZ;
 236                                         continue;
 237                                 default:
 238                                         break;
 239                                 }
 240                                 break;
 241                         }
 242                 } else
 243                         rp = tbl->last_span->layout;
 244
 245                 if (NULL == rp)
 246                         rp = tbl->last_span->layout;
 247         } else
 248                 rp = tbl->first_row;
 249
 250         assert(rp);
 251
 252         dp = newspan(tbl, ln, rp);
 253
 254         if ( ! strcmp(p, "_")) {
 255                 dp->pos = TBL_SPAN_HORIZ;
 256                 return(1);
 257         } else if ( ! strcmp(p, "=")) {
 258                 dp->pos = TBL_SPAN_DHORIZ;
 259                 return(1);
 260         }
 261
 262         dp->pos = TBL_SPAN_DATA;
 263
 264         /* This returns 0 when TBL_PART_CDATA is entered. */
 265
 266         while ('\0' != p[pos])
 267                 if ( ! data(tbl, dp, ln, p, &pos))
 268                         return(0);
 269
 270         return(1);
 271 }