tbl_data.c

   1 /*      $Id: tbl_data.c,v 1.56 2021/08/10 12:55:04 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->font = ESCAPE_FONTROMAN;
  82                         cp->spacing = SIZE_MAX;
  83                         dp->layout->last->next = cp;
  84                         cp->col = dp->layout->last->col + 1;
  85                         dp->layout->last = cp;
  86                 } else {
  87                         mandoc_msg(MANDOCERR_TBLDATA_EXTRA,
  88                             ln, sv, "%s", p + sv);
  89                         while (p[*pos] != '\0')
  90                                 (*pos)++;
  91                         return;
  92                 }
  93         }
  94
  95         dat = mandoc_malloc(sizeof(*dat));
  96         dat->layout = cp;
  97         dat->next = NULL;
  98         dat->string = NULL;
  99         dat->hspans = 0;
 100         dat->vspans = 0;
 101         dat->block = 0;
 102         dat->pos = TBL_DATA_NONE;
 103
 104         /*
 105          * Increment the number of vertical spans in a data cell above,
 106          * if this cell vertically extends one or more cells above.
 107          * The iteration must be done over data rows,
 108          * not over layout rows, because one layout row
 109          * can be reused for more than one data row.
 110          */
 111
 112         if (cp->pos == TBL_CELL_DOWN ||
 113             (*pos - sv == 2 && p[sv] == '\\' && p[sv + 1] == '^')) {
 114                 pdp = dp;
 115                 while ((pdp = pdp->prev) != NULL) {
 116                         pdat = pdp->first;
 117                         while (pdat != NULL &&
 118                             pdat->layout->col < dat->layout->col)
 119                                 pdat = pdat->next;
 120                         if (pdat == NULL)
 121                                 break;
 122                         if (pdat->layout->pos != TBL_CELL_DOWN &&
 123                             strcmp(pdat->string, "\\^") != 0) {
 124                                 pdat->vspans++;
 125                                 break;
 126                         }
 127                 }
 128         }
 129
 130         /*
 131          * Count the number of horizontal spans to the right of this cell.
 132          * This is purely a matter of the layout, independent of the data.
 133          */
 134
 135         for (cp = cp->next; cp != NULL; cp = cp->next)
 136                 if (cp->pos == TBL_CELL_SPAN)
 137                         dat->hspans++;
 138                 else
 139                         break;
 140
 141         /* Append the new data cell to the data row. */
 142
 143         if (dp->last == NULL)
 144                 dp->first = dat;
 145         else
 146                 dp->last->next = dat;
 147         dp->last = dat;
 148
 149         /*
 150          * Check for a continued-data scope opening.  This consists of a
 151          * trailing `T{' at the end of the line.  Subsequent lines,
 152          * until a standalone `T}', are included in our cell.
 153          */
 154
 155         if (*pos - sv == 2 && p[sv] == 'T' && p[sv + 1] == '{') {
 156                 tbl->part = TBL_PART_CDATA;
 157                 return;
 158         }
 159
 160         dat->string = mandoc_strndup(p + sv, *pos - sv);
 161
 162         if (p[*pos] != '\0')
 163                 (*pos)++;
 164
 165         if ( ! strcmp(dat->string, "_"))
 166                 dat->pos = TBL_DATA_HORIZ;
 167         else if ( ! strcmp(dat->string, "="))
 168                 dat->pos = TBL_DATA_DHORIZ;
 169         else if ( ! strcmp(dat->string, "\\_"))
 170                 dat->pos = TBL_DATA_NHORIZ;
 171         else if ( ! strcmp(dat->string, "\\="))
 172                 dat->pos = TBL_DATA_NDHORIZ;
 173         else
 174                 dat->pos = TBL_DATA_DATA;
 175
 176         if ((dat->layout->pos == TBL_CELL_HORIZ ||
 177             dat->layout->pos == TBL_CELL_DHORIZ ||
 178             dat->layout->pos == TBL_CELL_DOWN) &&
 179             dat->pos == TBL_DATA_DATA && *dat->string != '\0')
 180                 mandoc_msg(MANDOCERR_TBLDATA_SPAN,
 181                     ln, sv, "%s", dat->string);
 182 }
 183
 184 void
 185 tbl_cdata(struct tbl_node *tbl, int ln, const char *p, int pos)
 186 {
 187         struct tbl_dat  *dat;
 188         size_t           sz;
 189
 190         dat = tbl->last_span->last;
 191
 192         if (p[pos] == 'T' && p[pos + 1] == '}') {
 193                 pos += 2;
 194                 if (p[pos] == tbl->opts.tab) {
 195                         tbl->part = TBL_PART_DATA;
 196                         pos++;
 197                         while (p[pos] != '\0')
 198                                 getdata(tbl, tbl->last_span, ln, p, &pos);
 199                         return;
 200                 } else if (p[pos] == '\0') {
 201                         tbl->part = TBL_PART_DATA;
 202                         return;
 203                 }
 204
 205                 /* Fallthrough: T} is part of a word. */
 206         }
 207
 208         dat->pos = TBL_DATA_DATA;
 209         dat->block = 1;
 210
 211         if (dat->string != NULL) {
 212                 sz = strlen(p + pos) + strlen(dat->string) + 2;
 213                 dat->string = mandoc_realloc(dat->string, sz);
 214                 (void)strlcat(dat->string, " ", sz);
 215                 (void)strlcat(dat->string, p + pos, sz);
 216         } else
 217                 dat->string = mandoc_strdup(p + pos);
 218
 219         if (dat->layout->pos == TBL_CELL_DOWN)
 220                 mandoc_msg(MANDOCERR_TBLDATA_SPAN,
 221                     ln, pos, "%s", dat->string);
 222 }
 223
 224 static struct tbl_span *
 225 newspan(struct tbl_node *tbl, int line, struct tbl_row *rp)
 226 {
 227         struct tbl_span *dp;
 228
 229         dp = mandoc_calloc(1, sizeof(*dp));
 230         dp->line = line;
 231         dp->opts = &tbl->opts;
 232         dp->layout = rp;
 233         dp->prev = tbl->last_span;
 234
 235         if (dp->prev == NULL) {
 236                 tbl->first_span = dp;
 237                 tbl->current_span = NULL;
 238         } else
 239                 dp->prev->next = dp;
 240         tbl->last_span = dp;
 241
 242         return dp;
 243 }
 244
 245 void
 246 tbl_data(struct tbl_node *tbl, int ln, const char *p, int pos)
 247 {
 248         struct tbl_row  *rp;
 249         struct tbl_cell *cp;
 250         struct tbl_span *sp;
 251
 252         for (sp = tbl->last_span; sp != NULL; sp = sp->prev)
 253                 if (sp->pos == TBL_SPAN_DATA)
 254                         break;
 255         rp = sp == NULL ? tbl->first_row :
 256             sp->layout->next == NULL ? sp->layout : sp->layout->next;
 257         assert(rp != NULL);
 258
 259         if (p[1] == '\0') {
 260                 switch (p[0]) {
 261                 case '.':
 262                         /*
 263                          * Empty request lines must be handled here
 264                          * and cannot be discarded in roff_parseln()
 265                          * because in the layout section, they
 266                          * are significant and end the layout.
 267                          */
 268                         return;
 269                 case '_':
 270                         sp = newspan(tbl, ln, rp);
 271                         sp->pos = TBL_SPAN_HORIZ;
 272                         return;
 273                 case '=':
 274                         sp = newspan(tbl, ln, rp);
 275                         sp->pos = TBL_SPAN_DHORIZ;
 276                         return;
 277                 default:
 278                         break;
 279                 }
 280         }
 281
 282         /*
 283          * If the layout row contains nothing but horizontal lines,
 284          * allocate an empty span for it and assign the current span
 285          * to the next layout row accepting data.
 286          */
 287
 288         while (rp->next != NULL) {
 289                 if (rp->last->col + 1 < tbl->opts.cols)
 290                         break;
 291                 for (cp = rp->first; cp != NULL; cp = cp->next)
 292                         if (cp->pos != TBL_CELL_HORIZ &&
 293                             cp->pos != TBL_CELL_DHORIZ)
 294                                 break;
 295                 if (cp != NULL)
 296                         break;
 297                 sp = newspan(tbl, ln, rp);
 298                 sp->pos = TBL_SPAN_DATA;
 299                 rp = rp->next;
 300         }
 301
 302         /* Process a real data row. */
 303
 304         sp = newspan(tbl, ln, rp);
 305         sp->pos = TBL_SPAN_DATA;
 306         while (p[pos] != '\0')
 307                 getdata(tbl, sp, ln, p, &pos);
 308 }