tbl_data.c

   1 /*      $Id: tbl_data.c,v 1.49 2018/12/13 02:06:07 schwarze Exp $ */
   2 /*
   3  * Copyright (c) 2009, 2010, 2011 Kristaps Dzonsons <kristaps@bsd.lv>
   4  * Copyright (c) 2011, 2015, 2017, 2018 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 <stdlib.h>
  25 #include <string.h>
  26 #include <time.h>
  27
  28 #include "mandoc_aux.h"
  29 #include "mandoc.h"
  30 #include "tbl.h"
  31 #include "libmandoc.h"
  32 #include "tbl_int.h"
  33
  34 static  void             getdata(struct tbl_node *, struct tbl_span *,
  35                                 int, const char *, int *);
  36 static  struct tbl_span *newspan(struct tbl_node *, int,
  37                                 struct tbl_row *);
  38
  39
  40 static void
  41 getdata(struct tbl_node *tbl, struct tbl_span *dp,
  42                 int ln, const char *p, int *pos)
  43 {
  44         struct tbl_dat  *dat, *pdat;
  45         struct tbl_cell *cp;
  46         struct tbl_span *pdp;
  47         int              sv;
  48
  49         /*
  50          * Determine the length of the string in the cell
  51          * and advance the parse point to the end of the cell.
  52          */
  53
  54         sv = *pos;
  55         while (p[*pos] != '\0' && p[*pos] != tbl->opts.tab)
  56                 (*pos)++;
  57
  58         /* Advance to the next layout cell, skipping spanners. */
  59
  60         cp = dp->last == NULL ? dp->layout->first : dp->last->layout->next;
  61         while (cp != NULL && cp->pos == TBL_CELL_SPAN)
  62                 cp = cp->next;
  63
  64         /*
  65          * If the current layout row is out of cells, allocate
  66          * a new cell if another row of the table has at least
  67          * this number of columns, or discard the input if we
  68          * are beyond the last column of the table as a whole.
  69          */
  70
  71         if (cp == NULL) {
  72                 if (dp->layout->last->col + 1 < dp->opts->cols) {
  73                         cp = mandoc_calloc(1, sizeof(*cp));
  74                         cp->pos = TBL_CELL_LEFT;
  75                         dp->layout->last->next = cp;
  76                         cp->col = dp->layout->last->col + 1;
  77                         dp->layout->last = cp;
  78                 } else {
  79                         mandoc_msg(MANDOCERR_TBLDATA_EXTRA, tbl->parse,
  80                             ln, sv, p + sv);
  81                         while (p[*pos] != '\0')
  82                                 (*pos)++;
  83                         return;
  84                 }
  85         }
  86
  87         dat = mandoc_malloc(sizeof(*dat));
  88         dat->layout = cp;
  89         dat->next = NULL;
  90         dat->string = NULL;
  91         dat->hspans = 0;
  92         dat->vspans = 0;
  93         dat->block = 0;
  94         dat->pos = TBL_DATA_NONE;
  95
  96         /*
  97          * Increment the number of vertical spans in a data cell above,
  98          * if this cell vertically extends one or more cells above.
  99          * The iteration must be done over data rows,
 100          * not over layout rows, because one layout row
 101          * can be reused for more than one data row.
 102          */
 103
 104         if (cp->pos == TBL_CELL_DOWN ||
 105             (*pos - sv == 2 && p[sv] == '\\' && p[sv + 1] == '^')) {
 106                 pdp = dp;
 107                 while ((pdp = pdp->prev) != NULL) {
 108                         pdat = pdp->first;
 109                         while (pdat != NULL &&
 110                             pdat->layout->col < dat->layout->col)
 111                                 pdat = pdat->next;
 112                         if (pdat == NULL)
 113                                 break;
 114                         if (pdat->layout->pos != TBL_CELL_DOWN &&
 115                             strcmp(pdat->string, "\\^") != 0) {
 116                                 pdat->vspans++;
 117                                 break;
 118                         }
 119                 }
 120         }
 121
 122         /*
 123          * Count the number of horizontal spans to the right of this cell.
 124          * This is purely a matter of the layout, independent of the data.
 125          */
 126
 127         for (cp = cp->next; cp != NULL; cp = cp->next)
 128                 if (cp->pos == TBL_CELL_SPAN)
 129                         dat->hspans++;
 130                 else
 131                         break;
 132
 133         /* Append the new data cell to the data row. */
 134
 135         if (dp->last == NULL)
 136                 dp->first = dat;
 137         else
 138                 dp->last->next = dat;
 139         dp->last = dat;
 140
 141         /*
 142          * Check for a continued-data scope opening.  This consists of a
 143          * trailing `T{' at the end of the line.  Subsequent lines,
 144          * until a standalone `T}', are included in our cell.
 145          */
 146
 147         if (*pos - sv == 2 && p[sv] == 'T' && p[sv + 1] == '{') {
 148                 tbl->part = TBL_PART_CDATA;
 149                 return;
 150         }
 151
 152         dat->string = mandoc_strndup(p + sv, *pos - sv);
 153
 154         if (p[*pos] != '\0')
 155                 (*pos)++;
 156
 157         if ( ! strcmp(dat->string, "_"))
 158                 dat->pos = TBL_DATA_HORIZ;
 159         else if ( ! strcmp(dat->string, "="))
 160                 dat->pos = TBL_DATA_DHORIZ;
 161         else if ( ! strcmp(dat->string, "\\_"))
 162                 dat->pos = TBL_DATA_NHORIZ;
 163         else if ( ! strcmp(dat->string, "\\="))
 164                 dat->pos = TBL_DATA_NDHORIZ;
 165         else
 166                 dat->pos = TBL_DATA_DATA;
 167
 168         if ((dat->layout->pos == TBL_CELL_HORIZ ||
 169             dat->layout->pos == TBL_CELL_DHORIZ ||
 170             dat->layout->pos == TBL_CELL_DOWN) &&
 171             dat->pos == TBL_DATA_DATA && *dat->string != '\0')
 172                 mandoc_msg(MANDOCERR_TBLDATA_SPAN,
 173                     tbl->parse, ln, sv, dat->string);
 174 }
 175
 176 void
 177 tbl_cdata(struct tbl_node *tbl, int ln, const char *p, int pos)
 178 {
 179         struct tbl_dat  *dat;
 180         size_t           sz;
 181
 182         dat = tbl->last_span->last;
 183
 184         if (p[pos] == 'T' && p[pos + 1] == '}') {
 185                 pos += 2;
 186                 if (p[pos] == tbl->opts.tab) {
 187                         tbl->part = TBL_PART_DATA;
 188                         pos++;
 189                         while (p[pos] != '\0')
 190                                 getdata(tbl, tbl->last_span, ln, p, &pos);
 191                         return;
 192                 } else if (p[pos] == '\0') {
 193                         tbl->part = TBL_PART_DATA;
 194                         return;
 195                 }
 196
 197                 /* Fallthrough: T} is part of a word. */
 198         }
 199
 200         dat->pos = TBL_DATA_DATA;
 201         dat->block = 1;
 202
 203         if (dat->string != NULL) {
 204                 sz = strlen(p + pos) + strlen(dat->string) + 2;
 205                 dat->string = mandoc_realloc(dat->string, sz);
 206                 (void)strlcat(dat->string, " ", sz);
 207                 (void)strlcat(dat->string, p + pos, sz);
 208         } else
 209                 dat->string = mandoc_strdup(p + pos);
 210
 211         if (dat->layout->pos == TBL_CELL_DOWN)
 212                 mandoc_msg(MANDOCERR_TBLDATA_SPAN, tbl->parse,
 213                     ln, pos, dat->string);
 214 }
 215
 216 static struct tbl_span *
 217 newspan(struct tbl_node *tbl, int line, struct tbl_row *rp)
 218 {
 219         struct tbl_span *dp;
 220
 221         dp = mandoc_calloc(1, sizeof(*dp));
 222         dp->line = line;
 223         dp->opts = &tbl->opts;
 224         dp->layout = rp;
 225         dp->prev = tbl->last_span;
 226
 227         if (dp->prev == NULL) {
 228                 tbl->first_span = dp;
 229                 tbl->current_span = NULL;
 230         } else
 231                 dp->prev->next = dp;
 232         tbl->last_span = dp;
 233
 234         return dp;
 235 }
 236
 237 void
 238 tbl_data(struct tbl_node *tbl, int ln, const char *p, int pos)
 239 {
 240         struct tbl_row  *rp;
 241         struct tbl_cell *cp;
 242         struct tbl_span *sp;
 243
 244         rp = (sp = tbl->last_span) == NULL ? tbl->first_row :
 245             sp->pos == TBL_SPAN_DATA && sp->layout->next != NULL ?
 246             sp->layout->next : sp->layout;
 247
 248         assert(rp != NULL);
 249
 250         if ( ! strcmp(p, "_")) {
 251                 sp = newspan(tbl, ln, rp);
 252                 sp->pos = TBL_SPAN_HORIZ;
 253                 return;
 254         } else if ( ! strcmp(p, "=")) {
 255                 sp = newspan(tbl, ln, rp);
 256                 sp->pos = TBL_SPAN_DHORIZ;
 257                 return;
 258         }
 259
 260         /*
 261          * If the layout row contains nothing but horizontal lines,
 262          * allocate an empty span for it and assign the current span
 263          * to the next layout row accepting data.
 264          */
 265
 266         while (rp->next != NULL) {
 267                 if (rp->last->col + 1 < tbl->opts.cols)
 268                         break;
 269                 for (cp = rp->first; cp != NULL; cp = cp->next)
 270                         if (cp->pos != TBL_CELL_HORIZ &&
 271                             cp->pos != TBL_CELL_DHORIZ)
 272                                 break;
 273                 if (cp != NULL)
 274                         break;
 275                 sp = newspan(tbl, ln, rp);
 276                 sp->pos = TBL_SPAN_DATA;
 277                 rp = rp->next;
 278         }
 279
 280         /* Process a real data row. */
 281
 282         sp = newspan(tbl, ln, rp);
 283         sp->pos = TBL_SPAN_DATA;
 284         while (p[pos] != '\0')
 285                 getdata(tbl, sp, ln, p, &pos);
 286 }