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_free(struct parser*);
34 static int parse_text(struct parser*, enum line_type, const char*, size_t);
35 static int parse_link(struct parser*, enum line_type, const char*, size_t);
36 static int parse_title(struct parser*, enum line_type, const char*, size_t);
37 static int parse_item(struct parser*, enum line_type, const char*, size_t);
38 static int parse_quote(struct parser*, enum line_type, const char*, size_t);
39 static int parse_pre_start(struct parser*, enum line_type, const char*, size_t);
40 static int parse_pre_cnt(struct parser*, enum line_type, const char*, size_t);
41 static int parse_pre_end(struct parser*, enum line_type, const char*, size_t);
43 typedef int (parselinefn)(struct parser*, enum line_type, const char*, size_t);
45 static parselinefn *parsers[] = {
46 parse_text, /* LINE_TEXT */
47 parse_link, /* LINE_LINK */
48 parse_title, /* LINE_TITLE_1 */
49 parse_title, /* LINE_TITLE_2 */
50 parse_title, /* LINE_TITLE_3 */
51 parse_item, /* LINE_ITEM */
52 parse_quote, /* LINE_QUOTE */
53 parse_pre_start, /* LINE_PRE_START */
54 parse_pre_cnt, /* LINE_PRE_CONTENT */
55 parse_pre_end, /* LINE_PRE_END */
59 gemtext_initparser(struct parser *p)
61 memset(p, 0, sizeof(*p));
63 p->parse = &gemtext_parse;
64 p->free = &gemtext_free;
68 emit_line(struct parser *p, enum line_type type, char *line, char *alt)
72 if ((l = calloc(1, sizeof(*l))) == NULL)
79 if (TAILQ_EMPTY(&p->head))
80 TAILQ_INSERT_HEAD(&p->head, l, lines);
82 TAILQ_INSERT_TAIL(&p->head, l, lines);
88 parse_text(struct parser *p, enum line_type t, const char *buf, size_t len)
92 if ((l = calloc(1, len+1)) == NULL)
95 return emit_line(p, t, l, NULL);
99 parse_link(struct parser *p, enum line_type t, const char *buf, size_t len)
102 const char *url_start;
105 return emit_line(p, t, NULL, NULL);
109 while (len > 0 && isspace(buf[0])) {
115 return emit_line(p, t, NULL, NULL);
118 while (len > 0 && !isspace(buf[0])) {
123 if ((u = calloc(1, buf - url_start + 1)) == NULL)
125 memcpy(u, url_start, buf - url_start);
130 while (len > 0 && isspace(buf[0])) {
138 if ((l = calloc(1, len + 1)) == NULL)
142 return emit_line(p, t, l, u);
145 if ((l = strdup(u)) == NULL)
147 return emit_line(p, t, l, u);
151 parse_title(struct parser *p, enum line_type t, const char *buf, size_t len)
158 return emit_line(p, t, NULL, NULL);
164 return emit_line(p, t, NULL, NULL);
170 return emit_line(p, t, NULL, NULL);
179 while (len > 0 && isspace(buf[0])) {
185 return emit_line(p, t, NULL, NULL);
187 if ((l = calloc(1, len+1)) == NULL)
190 return emit_line(p, t, l, NULL);
194 parse_item(struct parser *p, enum line_type t, const char *buf, size_t len)
199 return emit_line(p, t, NULL, NULL);
204 while (len > 0 && isspace(buf[0])) {
210 return emit_line(p, t, NULL, NULL);
212 if ((l = calloc(1, len+1)) == NULL)
215 return emit_line(p, t, l, NULL);
219 parse_quote(struct parser *p, enum line_type t, const char *buf, size_t len)
224 return emit_line(p, t, NULL, NULL);
229 while (len > 0 && isspace(buf[0])) {
235 return emit_line(p, t, NULL, NULL);
237 if ((l = calloc(1, len+1)) == NULL)
240 return emit_line(p, t, l, NULL);
244 parse_pre_start(struct parser *p, enum line_type t, const char *buf, size_t len)
249 return emit_line(p, t, NULL, NULL);
254 while (len > 0 && isspace(buf[0])) {
260 return emit_line(p, t, NULL, NULL);
262 if ((l = calloc(1, len+1)) == NULL)
266 return emit_line(p, t, NULL, l);
270 parse_pre_cnt(struct parser *p, enum line_type t, const char *buf, size_t len)
275 return emit_line(p, t, NULL, NULL);
277 if ((l = calloc(1, len+1)) == NULL)
280 return emit_line(p, t, l, NULL);
284 parse_pre_end(struct parser *p, enum line_type t, const char *buf, size_t len)
286 return emit_line(p, t, NULL, NULL);
289 static inline enum line_type
290 detect_line_type(const char *buf, size_t len, int in_pre)
299 buf[0] == '`' && buf[1] == '`' && buf[2] == '`')
302 return LINE_PRE_CONTENT;
306 case '*': return LINE_ITEM;
307 case '>': return LINE_QUOTE;
309 if (len >= 1 && buf[1] == '>')
327 if (buf[0] == '`' && buf[1] == '`' && buf[2] == '`')
328 return LINE_PRE_START;
336 append(struct parser *p, const char *buf, size_t len)
341 newlen = len + p->len;
342 if ((t = calloc(1, newlen)) == NULL)
344 memcpy(t, p->buf, p->len);
345 memcpy(t + p->len, buf, len);
353 set_buf(struct parser *p, const char *buf, size_t len)
363 if ((p->buf = calloc(1, len)) == NULL)
365 memcpy(p->buf, buf, len);
371 gemtext_parse(struct parser *p, const char *buf, size_t size)
381 if (!append(p, buf, size))
388 if ((e = telescope_strnchr((char*)b, '\n', len)) == NULL)
391 t = detect_line_type(b, l, p->flags);
392 if (t == LINE_PRE_START)
394 if (t == LINE_PRE_END)
396 if (!parsers[t](p, t, b, l))
409 return set_buf(p, b, len);
413 gemtext_free(struct parser *p)
417 /* flush the buffer */
419 t = detect_line_type(p->buf, p->len, p->flags);
420 if (!parsers[t](p, t, p->buf, p->len))
422 if (p->flags && !emit_line(p, LINE_PRE_END, NULL, NULL))