2 * Copyright (c) 2021 Omar Polo <op@omarpolo.com>
4 * Permission to use, copy, modify, and distribute this software for any
5 * purpose with or without fee is hereby granted, provided that the above
6 * copyright notice and this permission notice appear in all copies.
8 * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
9 * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
10 * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
11 * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
12 * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
13 * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
14 * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
18 * A streaming gemtext parser.
25 #include "telescope.h"
31 static int gemtext_parse(struct parser*, const char*, size_t);
32 static int gemtext_foreach_line(struct parser*, const char*, size_t);
33 static int gemtext_free(struct parser*);
35 static int parse_text(struct parser*, enum line_type, const char*, size_t);
36 static int parse_link(struct parser*, enum line_type, const char*, size_t);
37 static int parse_title(struct parser*, enum line_type, const char*, size_t);
38 static int parse_item(struct parser*, enum line_type, const char*, size_t);
39 static int parse_quote(struct parser*, enum line_type, const char*, size_t);
40 static int parse_pre_start(struct parser*, enum line_type, const char*, size_t);
41 static int parse_pre_cnt(struct parser*, enum line_type, const char*, size_t);
42 static int parse_pre_end(struct parser*, enum line_type, const char*, size_t);
43 static void search_title(struct parser*, enum line_type);
45 typedef int (parselinefn)(struct parser*, enum line_type, const char*, size_t);
47 static parselinefn *parsers[] = {
48 parse_text, /* LINE_TEXT */
49 parse_link, /* LINE_LINK */
50 parse_title, /* LINE_TITLE_1 */
51 parse_title, /* LINE_TITLE_2 */
52 parse_title, /* LINE_TITLE_3 */
53 parse_item, /* LINE_ITEM */
54 parse_quote, /* LINE_QUOTE */
55 parse_pre_start, /* LINE_PRE_START */
56 parse_pre_cnt, /* LINE_PRE_CONTENT */
57 parse_pre_end, /* LINE_PRE_END */
61 gemtext_initparser(struct parser *p)
63 memset(p, 0, sizeof(*p));
65 p->name = "text/gemini";
66 p->parse = &gemtext_parse;
67 p->free = &gemtext_free;
71 emit_line(struct parser *p, enum line_type type, char *line, char *alt)
75 if ((l = calloc(1, sizeof(*l))) == NULL)
82 if (TAILQ_EMPTY(&p->head))
83 TAILQ_INSERT_HEAD(&p->head, l, lines);
85 TAILQ_INSERT_TAIL(&p->head, l, lines);
91 parse_text(struct parser *p, enum line_type t, const char *buf, size_t len)
95 if ((l = calloc(1, len+1)) == NULL)
98 return emit_line(p, t, l, NULL);
102 parse_link(struct parser *p, enum line_type t, const char *buf, size_t len)
105 const char *url_start;
108 return emit_line(p, t, NULL, NULL);
112 while (len > 0 && isspace(buf[0])) {
118 return emit_line(p, t, NULL, NULL);
121 while (len > 0 && !isspace(buf[0])) {
126 if ((u = calloc(1, buf - url_start + 1)) == NULL)
128 memcpy(u, url_start, buf - url_start);
133 while (len > 0 && isspace(buf[0])) {
141 if ((l = calloc(1, len + 1)) == NULL)
145 return emit_line(p, t, l, u);
148 if ((l = strdup(u)) == NULL)
150 return emit_line(p, t, l, u);
154 parse_title(struct parser *p, enum line_type t, const char *buf, size_t len)
161 return emit_line(p, t, NULL, NULL);
167 return emit_line(p, t, NULL, NULL);
173 return emit_line(p, t, NULL, NULL);
182 while (len > 0 && isspace(buf[0])) {
188 return emit_line(p, t, NULL, NULL);
190 if (t == LINE_TITLE_1 && *p->title == '\0')
191 strncpy(p->title, buf, MIN(sizeof(p->title)-1, len));
193 if ((l = calloc(1, len+1)) == NULL)
196 return emit_line(p, t, l, NULL);
200 parse_item(struct parser *p, enum line_type t, const char *buf, size_t len)
205 return emit_line(p, t, NULL, NULL);
210 while (len > 0 && isspace(buf[0])) {
216 return emit_line(p, t, NULL, NULL);
218 if ((l = calloc(1, len+1)) == NULL)
221 return emit_line(p, t, l, NULL);
225 parse_quote(struct parser *p, enum line_type t, const char *buf, size_t len)
230 return emit_line(p, t, NULL, NULL);
235 while (len > 0 && isspace(buf[0])) {
241 return emit_line(p, t, NULL, NULL);
243 if ((l = calloc(1, len+1)) == NULL)
246 return emit_line(p, t, l, NULL);
250 parse_pre_start(struct parser *p, enum line_type t, const char *buf, size_t len)
255 return emit_line(p, t, NULL, NULL);
260 while (len > 0 && isspace(buf[0])) {
266 return emit_line(p, t, NULL, NULL);
268 if ((l = calloc(1, len+1)) == NULL)
272 return emit_line(p, t, l, NULL);
276 parse_pre_cnt(struct parser *p, enum line_type t, const char *buf, size_t len)
281 return emit_line(p, t, NULL, NULL);
283 if ((l = calloc(1, len+1)) == NULL)
286 return emit_line(p, t, l, NULL);
290 parse_pre_end(struct parser *p, enum line_type t, const char *buf, size_t len)
292 return emit_line(p, t, NULL, NULL);
295 static inline enum line_type
296 detect_line_type(const char *buf, size_t len, int in_pre)
303 buf[0] == '`' && buf[1] == '`' && buf[2] == '`')
306 return LINE_PRE_CONTENT;
310 case '*': return LINE_ITEM;
311 case '>': return LINE_QUOTE;
313 if (len >= 1 && buf[1] == '>')
331 if (buf[0] == '`' && buf[1] == '`' && buf[2] == '`')
332 return LINE_PRE_START;
340 gemtext_parse(struct parser *p, const char *buf, size_t size)
342 return parser_foreach_line(p, buf, size, gemtext_foreach_line);
346 gemtext_foreach_line(struct parser *p, const char *line, size_t linelen)
350 t = detect_line_type(line, linelen, p->flags);
351 if (t == LINE_PRE_START)
353 if (t == LINE_PRE_END)
355 return parsers[t](p, t, line, linelen);
359 gemtext_free(struct parser *p)
363 /* flush the buffer */
365 t = detect_line_type(p->buf, p->len, p->flags);
366 if (!parsers[t](p, t, p->buf, p->len))
368 if (p->flags && !emit_line(p, LINE_PRE_END, NULL, NULL))
375 * use the first level 2 or 3 header as page title if none
378 if (*p->title == '\0')
379 search_title(p, LINE_TITLE_2);
380 if (*p->title == '\0')
381 search_title(p, LINE_TITLE_3);
387 search_title(struct parser *p, enum line_type level)
391 TAILQ_FOREACH(l, &p->head, lines) {
392 if (l->type == level) {
395 strlcpy(p->title, l->line, sizeof(p->title));