1 /* $Id: tbl_data.c,v 1.24 2011/03/20 16:02:05 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 mandoc_msg(MANDOCERR_TBLEXTRADAT,
68 tbl->parse, ln, *pos, NULL);
69 /* Skip to the end... */
70 while (p[*pos])
71 (*pos)++;
72 return(1);
73 }
74
75 dat = mandoc_calloc(1, sizeof(struct tbl_dat));
76 dat->layout = cp;
77 dat->pos = TBL_DATA_NONE;
78
79 assert(TBL_CELL_SPAN != cp->pos);
80
81 for (spans = 0, cp = cp->next; cp; cp = cp->next)
82 if (TBL_CELL_SPAN == cp->pos)
83 spans++;
84 else
85 break;
86
87 dat->spans = spans;
88
89 if (dp->last) {
90 dp->last->next = dat;
91 dp->last = dat;
92 } else
93 dp->last = dp->first = dat;
94
95 sv = *pos;
96 while (p[*pos] && p[*pos] != tbl->opts.tab)
97 (*pos)++;
98
99 /*
100 * Check for a continued-data scope opening. This consists of a
101 * trailing `T{' at the end of the line. Subsequent lines,
102 * until a standalone `T}', are included in our cell.
103 */
104
105 if (*pos - sv == 2 && 'T' == p[sv] && '{' == p[sv + 1]) {
106 tbl->part = TBL_PART_CDATA;
107 return(0);
108 }
109
110 assert(*pos - sv >= 0);
111
112 dat->string = mandoc_malloc((size_t)(*pos - sv + 1));
113 memcpy(dat->string, &p[sv], (size_t)(*pos - sv));
114 dat->string[*pos - sv] = '\0';
115
116 if (p[*pos])
117 (*pos)++;
118
119 if ( ! strcmp(dat->string, "_"))
120 dat->pos = TBL_DATA_HORIZ;
121 else if ( ! strcmp(dat->string, "="))
122 dat->pos = TBL_DATA_DHORIZ;
123 else if ( ! strcmp(dat->string, "\\_"))
124 dat->pos = TBL_DATA_NHORIZ;
125 else if ( ! strcmp(dat->string, "\\="))
126 dat->pos = TBL_DATA_NDHORIZ;
127 else
128 dat->pos = TBL_DATA_DATA;
129
130 if (TBL_CELL_HORIZ == dat->layout->pos ||
131 TBL_CELL_DHORIZ == dat->layout->pos ||
132 TBL_CELL_DOWN == dat->layout->pos)
133 if (TBL_DATA_DATA == dat->pos && '\0' != *dat->string)
134 mandoc_msg(MANDOCERR_TBLIGNDATA,
135 tbl->parse, ln, sv, NULL);
136
137 return(1);
138 }
139
140 /* ARGSUSED */
141 int
142 tbl_cdata(struct tbl_node *tbl, int ln, const char *p)
143 {
144 struct tbl_dat *dat;
145 size_t sz;
146 int pos;
147
148 pos = 0;
149
150 dat = tbl->last_span->last;
151
152 if (p[pos] == 'T' && p[pos + 1] == '}') {
153 pos += 2;
154 if (p[pos] == tbl->opts.tab) {
155 tbl->part = TBL_PART_DATA;
156 pos++;
157 return(data(tbl, tbl->last_span, ln, p, &pos));
158 } else if ('\0' == p[pos]) {
159 tbl->part = TBL_PART_DATA;
160 return(1);
161 }
162
163 /* Fallthrough: T} is part of a word. */
164 }
165
166 dat->pos = TBL_DATA_DATA;
167
168 if (dat->string) {
169 sz = strlen(p) + strlen(dat->string) + 2;
170 dat->string = mandoc_realloc(dat->string, sz);
171 strlcat(dat->string, " ", sz);
172 strlcat(dat->string, p, sz);
173 } else
174 dat->string = mandoc_strdup(p);
175
176 if (TBL_CELL_DOWN == dat->layout->pos)
177 mandoc_msg(MANDOCERR_TBLIGNDATA,
178 tbl->parse, ln, pos, NULL);
179
180 return(0);
181 }
182
183 static struct tbl_span *
184 newspan(struct tbl_node *tbl, int line, struct tbl_row *rp)
185 {
186 struct tbl_span *dp;
187
188 dp = mandoc_calloc(1, sizeof(struct tbl_span));
189 dp->line = line;
190 dp->tbl = &tbl->opts;
191 dp->layout = rp;
192 dp->head = tbl->first_head;
193
194 if (tbl->last_span) {
195 tbl->last_span->next = dp;
196 tbl->last_span = dp;
197 } else {
198 tbl->last_span = tbl->first_span = dp;
199 tbl->current_span = NULL;
200 dp->flags |= TBL_SPAN_FIRST;
201 }
202
203 return(dp);
204 }
205
206 int
207 tbl_data(struct tbl_node *tbl, int ln, const char *p)
208 {
209 struct tbl_span *dp;
210 struct tbl_row *rp;
211 int pos;
212
213 pos = 0;
214
215 if ('\0' == p[pos]) {
216 mandoc_msg(MANDOCERR_TBL, tbl->parse, ln, pos, NULL);
217 return(0);
218 }
219
220 /*
221 * Choose a layout row: take the one following the last parsed
222 * span's. If that doesn't exist, use the last parsed span's.
223 * If there's no last parsed span, use the first row. Lastly,
224 * if the last span was a horizontal line, use the same layout
225 * (it doesn't "consume" the layout).
226 */
227
228 if (tbl->last_span) {
229 assert(tbl->last_span->layout);
230 if (tbl->last_span->pos == TBL_SPAN_DATA) {
231 for (rp = tbl->last_span->layout->next;
232 rp && rp->first; rp = rp->next) {
233 switch (rp->first->pos) {
234 case (TBL_CELL_HORIZ):
235 dp = newspan(tbl, ln, rp);
236 dp->pos = TBL_SPAN_HORIZ;
237 continue;
238 case (TBL_CELL_DHORIZ):
239 dp = newspan(tbl, ln, rp);
240 dp->pos = TBL_SPAN_DHORIZ;
241 continue;
242 default:
243 break;
244 }
245 break;
246 }
247 } else
248 rp = tbl->last_span->layout;
249
250 if (NULL == rp)
251 rp = tbl->last_span->layout;
252 } else
253 rp = tbl->first_row;
254
255 assert(rp);
256
257 dp = newspan(tbl, ln, rp);
258
259 if ( ! strcmp(p, "_")) {
260 dp->pos = TBL_SPAN_HORIZ;
261 return(1);
262 } else if ( ! strcmp(p, "=")) {
263 dp->pos = TBL_SPAN_DHORIZ;
264 return(1);
265 }
266
267 dp->pos = TBL_SPAN_DATA;
268
269 /* This returns 0 when TBL_PART_CDATA is entered. */
270
271 while ('\0' != p[pos])
272 if ( ! data(tbl, dp, ln, p, &pos))
273 return(0);
274
275 return(1);
276 }