]> git.cameronkatri.com Git - mandoc.git/blob - tbl_data.c
Reject the escape sequences \[uD800] to \[uDFFF] in the parser.
[mandoc.git] / tbl_data.c
1 /* $Id: tbl_data.c,v 1.41 2015/10/06 18:32:20 schwarze Exp $ */
2 /*
3 * Copyright (c) 2009, 2010, 2011 Kristaps Dzonsons <kristaps@bsd.lv>
4 * Copyright (c) 2011, 2015 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.h"
29 #include "mandoc_aux.h"
30 #include "libmandoc.h"
31 #include "libroff.h"
32
33 static void getdata(struct tbl_node *, struct tbl_span *,
34 int, const char *, int *);
35 static struct tbl_span *newspan(struct tbl_node *, int,
36 struct tbl_row *);
37
38
39 static void
40 getdata(struct tbl_node *tbl, struct tbl_span *dp,
41 int ln, const char *p, int *pos)
42 {
43 struct tbl_dat *dat;
44 struct tbl_cell *cp;
45 int sv;
46
47 /* Advance to the next layout cell, skipping spanners. */
48
49 cp = dp->last == NULL ? dp->layout->first : dp->last->layout->next;
50 while (cp != NULL && cp->pos == TBL_CELL_SPAN)
51 cp = cp->next;
52
53 /*
54 * Stop processing when we reach the end of the available layout
55 * cells. This means that we have extra input.
56 */
57
58 if (cp == NULL) {
59 mandoc_msg(MANDOCERR_TBLDATA_EXTRA, tbl->parse,
60 ln, *pos, p + *pos);
61 /* Skip to the end... */
62 while (p[*pos])
63 (*pos)++;
64 return;
65 }
66
67 dat = mandoc_calloc(1, sizeof(*dat));
68 dat->layout = cp;
69 dat->pos = TBL_DATA_NONE;
70 dat->spans = 0;
71 for (cp = cp->next; cp != NULL; cp = cp->next)
72 if (cp->pos == TBL_CELL_SPAN)
73 dat->spans++;
74 else
75 break;
76
77 if (dp->last == NULL)
78 dp->first = dat;
79 else
80 dp->last->next = dat;
81 dp->last = dat;
82
83 sv = *pos;
84 while (p[*pos] && p[*pos] != tbl->opts.tab)
85 (*pos)++;
86
87 /*
88 * Check for a continued-data scope opening. This consists of a
89 * trailing `T{' at the end of the line. Subsequent lines,
90 * until a standalone `T}', are included in our cell.
91 */
92
93 if (*pos - sv == 2 && p[sv] == 'T' && p[sv + 1] == '{') {
94 tbl->part = TBL_PART_CDATA;
95 return;
96 }
97
98 dat->string = mandoc_strndup(p + sv, *pos - sv);
99
100 if (p[*pos])
101 (*pos)++;
102
103 if ( ! strcmp(dat->string, "_"))
104 dat->pos = TBL_DATA_HORIZ;
105 else if ( ! strcmp(dat->string, "="))
106 dat->pos = TBL_DATA_DHORIZ;
107 else if ( ! strcmp(dat->string, "\\_"))
108 dat->pos = TBL_DATA_NHORIZ;
109 else if ( ! strcmp(dat->string, "\\="))
110 dat->pos = TBL_DATA_NDHORIZ;
111 else
112 dat->pos = TBL_DATA_DATA;
113
114 if ((dat->layout->pos == TBL_CELL_HORIZ ||
115 dat->layout->pos == TBL_CELL_DHORIZ ||
116 dat->layout->pos == TBL_CELL_DOWN) &&
117 dat->pos == TBL_DATA_DATA && *dat->string != '\0')
118 mandoc_msg(MANDOCERR_TBLDATA_SPAN,
119 tbl->parse, ln, sv, dat->string);
120 }
121
122 int
123 tbl_cdata(struct tbl_node *tbl, int ln, const char *p, int pos)
124 {
125 struct tbl_dat *dat;
126 size_t sz;
127
128 dat = tbl->last_span->last;
129
130 if (p[pos] == 'T' && p[pos + 1] == '}') {
131 pos += 2;
132 if (p[pos] == tbl->opts.tab) {
133 tbl->part = TBL_PART_DATA;
134 pos++;
135 while (p[pos] != '\0')
136 getdata(tbl, tbl->last_span, ln, p, &pos);
137 return 1;
138 } else if (p[pos] == '\0') {
139 tbl->part = TBL_PART_DATA;
140 return 1;
141 }
142
143 /* Fallthrough: T} is part of a word. */
144 }
145
146 dat->pos = TBL_DATA_DATA;
147
148 if (dat->string != NULL) {
149 sz = strlen(p + pos) + strlen(dat->string) + 2;
150 dat->string = mandoc_realloc(dat->string, sz);
151 (void)strlcat(dat->string, " ", sz);
152 (void)strlcat(dat->string, p + pos, sz);
153 } else
154 dat->string = mandoc_strdup(p + pos);
155
156 if (dat->layout->pos == TBL_CELL_DOWN)
157 mandoc_msg(MANDOCERR_TBLDATA_SPAN, tbl->parse,
158 ln, pos, dat->string);
159
160 return 0;
161 }
162
163 static struct tbl_span *
164 newspan(struct tbl_node *tbl, int line, struct tbl_row *rp)
165 {
166 struct tbl_span *dp;
167
168 dp = mandoc_calloc(1, sizeof(*dp));
169 dp->line = line;
170 dp->opts = &tbl->opts;
171 dp->layout = rp;
172 dp->prev = tbl->last_span;
173
174 if (dp->prev == NULL) {
175 tbl->first_span = dp;
176 tbl->current_span = NULL;
177 } else
178 dp->prev->next = dp;
179 tbl->last_span = dp;
180
181 return dp;
182 }
183
184 void
185 tbl_data(struct tbl_node *tbl, int ln, const char *p, int pos)
186 {
187 struct tbl_span *dp;
188 struct tbl_row *rp;
189
190 /*
191 * Choose a layout row: take the one following the last parsed
192 * span's. If that doesn't exist, use the last parsed span's.
193 * If there's no last parsed span, use the first row. Lastly,
194 * if the last span was a horizontal line, use the same layout
195 * (it doesn't "consume" the layout).
196 */
197
198 if (tbl->last_span != NULL) {
199 if (tbl->last_span->pos == TBL_SPAN_DATA) {
200 for (rp = tbl->last_span->layout->next;
201 rp != NULL && rp->first != NULL;
202 rp = rp->next) {
203 switch (rp->first->pos) {
204 case TBL_CELL_HORIZ:
205 dp = newspan(tbl, ln, rp);
206 dp->pos = TBL_SPAN_HORIZ;
207 continue;
208 case TBL_CELL_DHORIZ:
209 dp = newspan(tbl, ln, rp);
210 dp->pos = TBL_SPAN_DHORIZ;
211 continue;
212 default:
213 break;
214 }
215 break;
216 }
217 } else
218 rp = tbl->last_span->layout;
219
220 if (rp == NULL)
221 rp = tbl->last_span->layout;
222 } else
223 rp = tbl->first_row;
224
225 assert(rp);
226
227 dp = newspan(tbl, ln, rp);
228
229 if ( ! strcmp(p, "_")) {
230 dp->pos = TBL_SPAN_HORIZ;
231 return;
232 } else if ( ! strcmp(p, "=")) {
233 dp->pos = TBL_SPAN_DHORIZ;
234 return;
235 }
236
237 dp->pos = TBL_SPAN_DATA;
238
239 while (p[pos] != '\0')
240 getdata(tbl, dp, ln, p, &pos);
241 }