tbl_data.c

   1 /*      $Id: tbl_data.c,v 1.55 2021/05/18 13:22:43 schwarze Exp $ */
   2 /*
   3  * Copyright (c) 2009, 2010, 2011 Kristaps Dzonsons <kristaps@bsd.lv>
   4  * Copyright (c) 2011,2015,2017-2019,2021 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 #include "config.h"
  19
  20 #include <sys/types.h>
  21
  22 #include <assert.h>
  23 #include <ctype.h>
  24 #include <stdint.h>
  25 #include <stdio.h>
  26 #include <stdlib.h>
  27 #include <string.h>
  28 #include <time.h>
  29
  30 #include "mandoc_aux.h"
  31 #include "mandoc.h"
  32 #include "tbl.h"
  33 #include "libmandoc.h"
  34 #include "tbl_int.h"
  35
  36 static  void             getdata(struct tbl_node *, struct tbl_span *,
  37                                 int, const char *, int *);
  38 static  struct tbl_span *newspan(struct tbl_node *, int,
  39                                 struct tbl_row *);
  40
  41
  42 static void
  43 getdata(struct tbl_node *tbl, struct tbl_span *dp,
  44                 int ln, const char *p, int *pos)
  45 {
  46         struct tbl_dat  *dat, *pdat;
  47         struct tbl_cell *cp;
  48         struct tbl_span *pdp;
  49         const char      *ccp;
  50         int              sv;
  51
  52         /*
  53          * Determine the length of the string in the cell
  54          * and advance the parse point to the end of the cell.
  55          */
  56
  57         sv = *pos;
  58         ccp = p + sv;
  59         while (*ccp != '\0' && *ccp != tbl->opts.tab)
  60                 if (*ccp++ == '\\')
  61                         mandoc_escape(&ccp, NULL, NULL);
  62         *pos = ccp - p;
  63
  64         /* Advance to the next layout cell, skipping spanners. */
  65
  66         cp = dp->last == NULL ? dp->layout->first : dp->last->layout->next;
  67         while (cp != NULL && cp->pos == TBL_CELL_SPAN)
  68                 cp = cp->next;
  69
  70         /*
  71          * If the current layout row is out of cells, allocate
  72          * a new cell if another row of the table has at least
  73          * this number of columns, or discard the input if we
  74          * are beyond the last column of the table as a whole.
  75          */
  76
  77         if (cp == NULL) {
  78                 if (dp->layout->last->col + 1 < dp->opts->cols) {
  79                         cp = mandoc_calloc(1, sizeof(*cp));
  80                         cp->pos = TBL_CELL_LEFT;
  81                         cp->spacing = SIZE_MAX;
  82                         dp->layout->last->next = cp;
  83                         cp->col = dp->layout->last->col + 1;
  84                         dp->layout->last = cp;
  85                 } else {
  86                         mandoc_msg(MANDOCERR_TBLDATA_EXTRA,
  87                             ln, sv, "%s", p + sv);
  88                         while (p[*pos] != '\0')
  89                                 (*pos)++;
  90                         return;
  91                 }
  92         }
  93
  94         dat = mandoc_malloc(sizeof(*dat));
  95         dat->layout = cp;
  96         dat->next = NULL;
  97         dat->string = NULL;
  98         dat->hspans = 0;
  99         dat->vspans = 0;
 100         dat->block = 0;
 101         dat->pos = TBL_DATA_NONE;
 102
 103         /*
 104          * Increment the number of vertical spans in a data cell above,
 105          * if this cell vertically extends one or more cells above.
 106          * The iteration must be done over data rows,
 107          * not over layout rows, because one layout row
 108          * can be reused for more than one data row.
 109          */
 110
 111         if (cp->pos == TBL_CELL_DOWN ||
 112             (*pos - sv == 2 && p[sv] == '\\' && p[sv + 1] == '^')) {
 113                 pdp = dp;
 114                 while ((pdp = pdp->prev) != NULL) {
 115                         pdat = pdp->first;
 116                         while (pdat != NULL &&
 117                             pdat->layout->col < dat->layout->col)
 118                                 pdat = pdat->next;
 119                         if (pdat == NULL)
 120                                 break;
 121                         if (pdat->layout->pos != TBL_CELL_DOWN &&
 122                             strcmp(pdat->string, "\\^") != 0) {
 123                                 pdat->vspans++;
 124                                 break;
 125                         }
 126                 }
 127         }
 128
 129         /*
 130          * Count the number of horizontal spans to the right of this cell.
 131          * This is purely a matter of the layout, independent of the data.
 132          */
 133
 134         for (cp = cp->next; cp != NULL; cp = cp->next)
 135                 if (cp->pos == TBL_CELL_SPAN)
 136                         dat->hspans++;
 137                 else
 138                         break;
 139
 140         /* Append the new data cell to the data row. */
 141
 142         if (dp->last == NULL)
 143                 dp->first = dat;
 144         else
 145                 dp->last->next = dat;
 146         dp->last = dat;
 147
 148         /*
 149          * Check for a continued-data scope opening.  This consists of a
 150          * trailing `T{' at the end of the line.  Subsequent lines,
 151          * until a standalone `T}', are included in our cell.
 152          */
 153
 154         if (*pos - sv == 2 && p[sv] == 'T' && p[sv + 1] == '{') {
 155                 tbl->part = TBL_PART_CDATA;
 156                 return;
 157         }
 158
 159         dat->string = mandoc_strndup(p + sv, *pos - sv);
 160
 161         if (p[*pos] != '\0')
 162                 (*pos)++;
 163
 164         if ( ! strcmp(dat->string, "_"))
 165                 dat->pos = TBL_DATA_HORIZ;
 166         else if ( ! strcmp(dat->string, "="))
 167                 dat->pos = TBL_DATA_DHORIZ;
 168         else if ( ! strcmp(dat->string, "\\_"))
 169                 dat->pos = TBL_DATA_NHORIZ;
 170         else if ( ! strcmp(dat->string, "\\="))
 171                 dat->pos = TBL_DATA_NDHORIZ;
 172         else
 173                 dat->pos = TBL_DATA_DATA;
 174
 175         if ((dat->layout->pos == TBL_CELL_HORIZ ||
 176             dat->layout->pos == TBL_CELL_DHORIZ ||
 177             dat->layout->pos == TBL_CELL_DOWN) &&
 178             dat->pos == TBL_DATA_DATA && *dat->string != '\0')
 179                 mandoc_msg(MANDOCERR_TBLDATA_SPAN,
 180                     ln, sv, "%s", dat->string);
 181 }
 182
 183 void
 184 tbl_cdata(struct tbl_node *tbl, int ln, const char *p, int pos)
 185 {
 186         struct tbl_dat  *dat;
 187         size_t           sz;
 188
 189         dat = tbl->last_span->last;
 190
 191         if (p[pos] == 'T' && p[pos + 1] == '}') {
 192                 pos += 2;
 193                 if (p[pos] == tbl->opts.tab) {
 194                         tbl->part = TBL_PART_DATA;
 195                         pos++;
 196                         while (p[pos] != '\0')
 197                                 getdata(tbl, tbl->last_span, ln, p, &pos);
 198                         return;
 199                 } else if (p[pos] == '\0') {
 200                         tbl->part = TBL_PART_DATA;
 201                         return;
 202                 }
 203
 204                 /* Fallthrough: T} is part of a word. */
 205         }
 206
 207         dat->pos = TBL_DATA_DATA;
 208         dat->block = 1;
 209
 210         if (dat->string != NULL) {
 211                 sz = strlen(p + pos) + strlen(dat->string) + 2;
 212                 dat->string = mandoc_realloc(dat->string, sz);
 213                 (void)strlcat(dat->string, " ", sz);
 214                 (void)strlcat(dat->string, p + pos, sz);
 215         } else
 216                 dat->string = mandoc_strdup(p + pos);
 217
 218         if (dat->layout->pos == TBL_CELL_DOWN)
 219                 mandoc_msg(MANDOCERR_TBLDATA_SPAN,
 220                     ln, pos, "%s", dat->string);
 221 }
 222
 223 static struct tbl_span *
 224 newspan(struct tbl_node *tbl, int line, struct tbl_row *rp)
 225 {
 226         struct tbl_span *dp;
 227
 228         dp = mandoc_calloc(1, sizeof(*dp));
 229         dp->line = line;
 230         dp->opts = &tbl->opts;
 231         dp->layout = rp;
 232         dp->prev = tbl->last_span;
 233
 234         if (dp->prev == NULL) {
 235                 tbl->first_span = dp;
 236                 tbl->current_span = NULL;
 237         } else
 238                 dp->prev->next = dp;
 239         tbl->last_span = dp;
 240
 241         return dp;
 242 }
 243
 244 void
 245 tbl_data(struct tbl_node *tbl, int ln, const char *p, int pos)
 246 {
 247         struct tbl_row  *rp;
 248         struct tbl_cell *cp;
 249         struct tbl_span *sp;
 250
 251         for (sp = tbl->last_span; sp != NULL; sp = sp->prev)
 252                 if (sp->pos == TBL_SPAN_DATA)
 253                         break;
 254         rp = sp == NULL ? tbl->first_row :
 255             sp->layout->next == NULL ? sp->layout : sp->layout->next;
 256         assert(rp != NULL);
 257
 258         if (p[1] == '\0') {
 259                 switch (p[0]) {
 260                 case '.':
 261                         /*
 262                          * Empty request lines must be handled here
 263                          * and cannot be discarded in roff_parseln()
 264                          * because in the layout section, they
 265                          * are significant and end the layout.
 266                          */
 267                         return;
 268                 case '_':
 269                         sp = newspan(tbl, ln, rp);
 270                         sp->pos = TBL_SPAN_HORIZ;
 271                         return;
 272                 case '=':
 273                         sp = newspan(tbl, ln, rp);
 274                         sp->pos = TBL_SPAN_DHORIZ;
 275                         return;
 276                 default:
 277                         break;
 278                 }
 279         }
 280
 281         /*
 282          * If the layout row contains nothing but horizontal lines,
 283          * allocate an empty span for it and assign the current span
 284          * to the next layout row accepting data.
 285          */
 286
 287         while (rp->next != NULL) {
 288                 if (rp->last->col + 1 < tbl->opts.cols)
 289                         break;
 290                 for (cp = rp->first; cp != NULL; cp = cp->next)
 291                         if (cp->pos != TBL_CELL_HORIZ &&
 292                             cp->pos != TBL_CELL_DHORIZ)
 293                                 break;
 294                 if (cp != NULL)
 295                         break;
 296                 sp = newspan(tbl, ln, rp);
 297                 sp->pos = TBL_SPAN_DATA;
 298                 rp = rp->next;
 299         }
 300
 301         /* Process a real data row. */
 302
 303         sp = newspan(tbl, ln, rp);
 304         sp->pos = TBL_SPAN_DATA;
 305         while (p[pos] != '\0')
 306                 getdata(tbl, sp, ln, p, &pos);
 307 }