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->name = "text/gemini";
64 p->parse = &gemtext_parse;
65 p->free = &gemtext_free;
69 emit_line(struct parser *p, enum line_type type, char *line, char *alt)
73 if ((l = calloc(1, sizeof(*l))) == NULL)
80 if (TAILQ_EMPTY(&p->head))
81 TAILQ_INSERT_HEAD(&p->head, l, lines);
83 TAILQ_INSERT_TAIL(&p->head, l, lines);
89 parse_text(struct parser *p, enum line_type t, const char *buf, size_t len)
93 if ((l = calloc(1, len+1)) == NULL)
96 return emit_line(p, t, l, NULL);
100 parse_link(struct parser *p, enum line_type t, const char *buf, size_t len)
103 const char *url_start;
106 return emit_line(p, t, NULL, NULL);
110 while (len > 0 && isspace(buf[0])) {
116 return emit_line(p, t, NULL, NULL);
119 while (len > 0 && !isspace(buf[0])) {
124 if ((u = calloc(1, buf - url_start + 1)) == NULL)
126 memcpy(u, url_start, buf - url_start);
131 while (len > 0 && isspace(buf[0])) {
139 if ((l = calloc(1, len + 1)) == NULL)
143 return emit_line(p, t, l, u);
146 if ((l = strdup(u)) == NULL)
148 return emit_line(p, t, l, u);
152 parse_title(struct parser *p, enum line_type t, const char *buf, size_t len)
159 return emit_line(p, t, NULL, NULL);
165 return emit_line(p, t, NULL, NULL);
171 return emit_line(p, t, NULL, NULL);
180 while (len > 0 && isspace(buf[0])) {
186 return emit_line(p, t, NULL, NULL);
188 if (t == LINE_TITLE_1 && *p->title == '\0')
189 strncpy(p->title, buf, MIN(sizeof(p->title)-1, len));
191 if ((l = calloc(1, len+1)) == NULL)
194 return emit_line(p, t, l, NULL);
198 parse_item(struct parser *p, enum line_type t, const char *buf, size_t len)
203 return emit_line(p, t, NULL, NULL);
208 while (len > 0 && isspace(buf[0])) {
214 return emit_line(p, t, NULL, NULL);
216 if ((l = calloc(1, len+1)) == NULL)
219 return emit_line(p, t, l, NULL);
223 parse_quote(struct parser *p, enum line_type t, const char *buf, size_t len)
228 return emit_line(p, t, NULL, NULL);
233 while (len > 0 && isspace(buf[0])) {
239 return emit_line(p, t, NULL, NULL);
241 if ((l = calloc(1, len+1)) == NULL)
244 return emit_line(p, t, l, NULL);
248 parse_pre_start(struct parser *p, enum line_type t, const char *buf, size_t len)
253 return emit_line(p, t, NULL, NULL);
258 while (len > 0 && isspace(buf[0])) {
264 return emit_line(p, t, NULL, NULL);
266 if ((l = calloc(1, len+1)) == NULL)
270 return emit_line(p, t, l, NULL);
274 parse_pre_cnt(struct parser *p, enum line_type t, const char *buf, size_t len)
279 return emit_line(p, t, NULL, NULL);
281 if ((l = calloc(1, len+1)) == NULL)
284 return emit_line(p, t, l, NULL);
288 parse_pre_end(struct parser *p, enum line_type t, const char *buf, size_t len)
290 return emit_line(p, t, NULL, NULL);
293 static inline enum line_type
294 detect_line_type(const char *buf, size_t len, int in_pre)
301 buf[0] == '`' && buf[1] == '`' && buf[2] == '`')
304 return LINE_PRE_CONTENT;
308 case '*': return LINE_ITEM;
309 case '>': return LINE_QUOTE;
311 if (len >= 1 && buf[1] == '>')
329 if (buf[0] == '`' && buf[1] == '`' && buf[2] == '`')
330 return LINE_PRE_START;
338 gemtext_parse(struct parser *p, const char *buf, size_t size)
348 if (!parser_append(p, buf, size))
355 if ((e = memmem((char*)b, len, "\n", 1)) == NULL)
358 t = detect_line_type(b, l, p->flags);
359 if (t == LINE_PRE_START)
361 if (t == LINE_PRE_END)
363 if (!parsers[t](p, t, b, l))
376 return parser_set_buf(p, b, len);
380 gemtext_free(struct parser *p)
384 /* flush the buffer */
386 t = detect_line_type(p->buf, p->len, p->flags);
387 if (!parsers[t](p, t, p->buf, p->len))
389 if (p->flags && !emit_line(p, LINE_PRE_END, NULL, NULL))