bin/mandoc/tbl_data.c

   1 /*      $Id: tbl_data.c,v 1.52 2019/02/09 16:00:39 schwarze Exp $ */
   2 /*
   3  * Copyright (c) 2009, 2010, 2011 Kristaps Dzonsons <kristaps@bsd.lv>
   4  * Copyright (c) 2011,2015,2017,2018,2019 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 <stdio.h>
  25 #include <stdlib.h>
  26 #include <string.h>
  27 #include <time.h>
  28
  29 #include "mandoc_aux.h"
  30 #include "mandoc.h"
  31 #include "tbl.h"
  32 #include "libmandoc.h"
  33 #include "tbl_int.h"
  34
  35 static  void             getdata(struct tbl_node *, struct tbl_span *,
  36                                 int, const char *, int *);
  37 static  struct tbl_span *newspan(struct tbl_node *, int,
  38                                 struct tbl_row *);
  39
  40
  41 static void
  42 getdata(struct tbl_node *tbl, struct tbl_span *dp,
  43                 int ln, const char *p, int *pos)
  44 {
  45         struct tbl_dat  *dat, *pdat;
  46         struct tbl_cell *cp;
  47         struct tbl_span *pdp;
  48         int              sv;
  49
  50         /*
  51          * Determine the length of the string in the cell
  52          * and advance the parse point to the end of the cell.
  53          */
  54
  55         sv = *pos;
  56         while (p[*pos] != '\0' && p[*pos] != tbl->opts.tab)
  57                 (*pos)++;
  58
  59         /* Advance to the next layout cell, skipping spanners. */
  60
  61         cp = dp->last == NULL ? dp->layout->first : dp->last->layout->next;
  62         while (cp != NULL && cp->pos == TBL_CELL_SPAN)
  63                 cp = cp->next;
  64
  65         /*
  66          * If the current layout row is out of cells, allocate
  67          * a new cell if another row of the table has at least
  68          * this number of columns, or discard the input if we
  69          * are beyond the last column of the table as a whole.
  70          */
  71
  72         if (cp == NULL) {
  73                 if (dp->layout->last->col + 1 < dp->opts->cols) {
  74                         cp = mandoc_calloc(1, sizeof(*cp));
  75                         cp->pos = TBL_CELL_LEFT;
  76                         dp->layout->last->next = cp;
  77                         cp->col = dp->layout->last->col + 1;
  78                         dp->layout->last = cp;
  79                 } else {
  80                         mandoc_msg(MANDOCERR_TBLDATA_EXTRA,
  81                             ln, sv, "%s", p + sv);
  82                         while (p[*pos] != '\0')
  83                                 (*pos)++;
  84                         return;
  85                 }
  86         }
  87
  88         dat = mandoc_malloc(sizeof(*dat));
  89         dat->layout = cp;
  90         dat->next = NULL;
  91         dat->string = NULL;
  92         dat->hspans = 0;
  93         dat->vspans = 0;
  94         dat->block = 0;
  95         dat->pos = TBL_DATA_NONE;
  96
  97         /*
  98          * Increment the number of vertical spans in a data cell above,
  99          * if this cell vertically extends one or more cells above.
 100          * The iteration must be done over data rows,
 101          * not over layout rows, because one layout row
 102          * can be reused for more than one data row.
 103          */
 104
 105         if (cp->pos == TBL_CELL_DOWN ||
 106             (*pos - sv == 2 && p[sv] == '\\' && p[sv + 1] == '^')) {
 107                 pdp = dp;
 108                 while ((pdp = pdp->prev) != NULL) {
 109                         pdat = pdp->first;
 110                         while (pdat != NULL &&
 111                             pdat->layout->col < dat->layout->col)
 112                                 pdat = pdat->next;
 113                         if (pdat == NULL)
 114                                 break;
 115                         if (pdat->layout->pos != TBL_CELL_DOWN &&
 116                             strcmp(pdat->string, "\\^") != 0) {
 117                                 pdat->vspans++;
 118                                 break;
 119                         }
 120                 }
 121         }
 122
 123         /*
 124          * Count the number of horizontal spans to the right of this cell.
 125          * This is purely a matter of the layout, independent of the data.
 126          */
 127
 128         for (cp = cp->next; cp != NULL; cp = cp->next)
 129                 if (cp->pos == TBL_CELL_SPAN)
 130                         dat->hspans++;
 131                 else
 132                         break;
 133
 134         /* Append the new data cell to the data row. */
 135
 136         if (dp->last == NULL)
 137                 dp->first = dat;
 138         else
 139                 dp->last->next = dat;
 140         dp->last = dat;
 141
 142         /*
 143          * Check for a continued-data scope opening.  This consists of a
 144          * trailing `T{' at the end of the line.  Subsequent lines,
 145          * until a standalone `T}', are included in our cell.
 146          */
 147
 148         if (*pos - sv == 2 && p[sv] == 'T' && p[sv + 1] == '{') {
 149                 tbl->part = TBL_PART_CDATA;
 150                 return;
 151         }
 152
 153         dat->string = mandoc_strndup(p + sv, *pos - sv);
 154
 155         if (p[*pos] != '\0')
 156                 (*pos)++;
 157
 158         if ( ! strcmp(dat->string, "_"))
 159                 dat->pos = TBL_DATA_HORIZ;
 160         else if ( ! strcmp(dat->string, "="))
 161                 dat->pos = TBL_DATA_DHORIZ;
 162         else if ( ! strcmp(dat->string, "\\_"))
 163                 dat->pos = TBL_DATA_NHORIZ;
 164         else if ( ! strcmp(dat->string, "\\="))
 165                 dat->pos = TBL_DATA_NDHORIZ;
 166         else
 167                 dat->pos = TBL_DATA_DATA;
 168
 169         if ((dat->layout->pos == TBL_CELL_HORIZ ||
 170             dat->layout->pos == TBL_CELL_DHORIZ ||
 171             dat->layout->pos == TBL_CELL_DOWN) &&
 172             dat->pos == TBL_DATA_DATA && *dat->string != '\0')
 173                 mandoc_msg(MANDOCERR_TBLDATA_SPAN,
 174                     ln, sv, "%s", dat->string);
 175 }
 176
 177 void
 178 tbl_cdata(struct tbl_node *tbl, int ln, const char *p, int pos)
 179 {
 180         struct tbl_dat  *dat;
 181         size_t           sz;
 182
 183         dat = tbl->last_span->last;
 184
 185         if (p[pos] == 'T' && p[pos + 1] == '}') {
 186                 pos += 2;
 187                 if (p[pos] == tbl->opts.tab) {
 188                         tbl->part = TBL_PART_DATA;
 189                         pos++;
 190                         while (p[pos] != '\0')
 191                                 getdata(tbl, tbl->last_span, ln, p, &pos);
 192                         return;
 193                 } else if (p[pos] == '\0') {
 194                         tbl->part = TBL_PART_DATA;
 195                         return;
 196                 }
 197
 198                 /* Fallthrough: T} is part of a word. */
 199         }
 200
 201         dat->pos = TBL_DATA_DATA;
 202         dat->block = 1;
 203
 204         if (dat->string != NULL) {
 205                 sz = strlen(p + pos) + strlen(dat->string) + 2;
 206                 dat->string = mandoc_realloc(dat->string, sz);
 207                 (void)strlcat(dat->string, " ", sz);
 208                 (void)strlcat(dat->string, p + pos, sz);
 209         } else
 210                 dat->string = mandoc_strdup(p + pos);
 211
 212         if (dat->layout->pos == TBL_CELL_DOWN)
 213                 mandoc_msg(MANDOCERR_TBLDATA_SPAN,
 214                     ln, pos, "%s", dat->string);
 215 }
 216
 217 static struct tbl_span *
 218 newspan(struct tbl_node *tbl, int line, struct tbl_row *rp)
 219 {
 220         struct tbl_span *dp;
 221
 222         dp = mandoc_calloc(1, sizeof(*dp));
 223         dp->line = line;
 224         dp->opts = &tbl->opts;
 225         dp->layout = rp;
 226         dp->prev = tbl->last_span;
 227
 228         if (dp->prev == NULL) {
 229                 tbl->first_span = dp;
 230                 tbl->current_span = NULL;
 231         } else
 232                 dp->prev->next = dp;
 233         tbl->last_span = dp;
 234
 235         return dp;
 236 }
 237
 238 void
 239 tbl_data(struct tbl_node *tbl, int ln, const char *p, int pos)
 240 {
 241         struct tbl_row  *rp;
 242         struct tbl_cell *cp;
 243         struct tbl_span *sp;
 244
 245         rp = (sp = tbl->last_span) == NULL ? tbl->first_row :
 246             sp->pos == TBL_SPAN_DATA && sp->layout->next != NULL ?
 247             sp->layout->next : sp->layout;
 248
 249         assert(rp != NULL);
 250
 251         if (p[1] == '\0') {
 252                 switch (p[0]) {
 253                 case '.':
 254                         /*
 255                          * Empty request lines must be handled here
 256                          * and cannot be discarded in roff_parseln()
 257                          * because in the layout section, they
 258                          * are significant and end the layout.
 259                          */
 260                         return;
 261                 case '_':
 262                         sp = newspan(tbl, ln, rp);
 263                         sp->pos = TBL_SPAN_HORIZ;
 264                         return;
 265                 case '=':
 266                         sp = newspan(tbl, ln, rp);
 267                         sp->pos = TBL_SPAN_DHORIZ;
 268                         return;
 269                 default:
 270                         break;
 271                 }
 272         }
 273
 274         /*
 275          * If the layout row contains nothing but horizontal lines,
 276          * allocate an empty span for it and assign the current span
 277          * to the next layout row accepting data.
 278          */
 279
 280         while (rp->next != NULL) {
 281                 if (rp->last->col + 1 < tbl->opts.cols)
 282                         break;
 283                 for (cp = rp->first; cp != NULL; cp = cp->next)
 284                         if (cp->pos != TBL_CELL_HORIZ &&
 285                             cp->pos != TBL_CELL_DHORIZ)
 286                                 break;
 287                 if (cp != NULL)
 288                         break;
 289                 sp = newspan(tbl, ln, rp);
 290                 sp->pos = TBL_SPAN_DATA;
 291                 rp = rp->next;
 292         }
 293
 294         /* Process a real data row. */
 295
 296         sp = newspan(tbl, ln, rp);
 297         sp->pos = TBL_SPAN_DATA;
 298         while (p[pos] != '\0')
 299                 getdata(tbl, sp, ln, p, &pos);
 300 }