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.
27 #include "telescope.h"
33 static int gemtext_parse(struct parser*, const char*, size_t);
34 static int gemtext_foreach_line(struct parser*, const char*, size_t);
35 static int gemtext_free(struct parser*);
37 static int parse_text(struct parser*, enum line_type, const char*, size_t);
38 static int parse_link(struct parser*, enum line_type, const char*, size_t);
39 static int parse_title(struct parser*, enum line_type, const char*, size_t);
40 static int parse_item(struct parser*, enum line_type, const char*, size_t);
41 static int parse_quote(struct parser*, enum line_type, const char*, size_t);
42 static int parse_pre_start(struct parser*, enum line_type, const char*, size_t);
43 static int parse_pre_cnt(struct parser*, enum line_type, const char*, size_t);
44 static int parse_pre_end(struct parser*, enum line_type, const char*, size_t);
45 static void search_title(struct parser*, enum line_type);
47 typedef int (parselinefn)(struct parser*, enum line_type, const char*, size_t);
49 static parselinefn *parsers[] = {
50 [LINE_TEXT] = parse_text,
51 [LINE_LINK] = parse_link,
52 [LINE_TITLE_1] = parse_title,
53 [LINE_TITLE_2] = parse_title,
54 [LINE_TITLE_3] = parse_title,
55 [LINE_ITEM] = parse_item,
56 [LINE_QUOTE] = parse_quote,
57 [LINE_PRE_START] = parse_pre_start,
58 [LINE_PRE_CONTENT] = parse_pre_cnt,
59 [LINE_PRE_END] = parse_pre_end,
63 gemtext_initparser(struct parser *p)
65 memset(p, 0, sizeof(*p));
67 p->name = "text/gemini";
68 p->parse = &gemtext_parse;
69 p->free = &gemtext_free;
73 emit_line(struct parser *p, enum line_type type, char *line, char *alt)
77 if ((l = calloc(1, sizeof(*l))) == NULL)
90 case LINE_PRE_CONTENT:
99 if (TAILQ_EMPTY(&p->head))
100 TAILQ_INSERT_HEAD(&p->head, l, lines);
102 TAILQ_INSERT_TAIL(&p->head, l, lines);
108 parse_text(struct parser *p, enum line_type t, const char *buf, size_t len)
112 if ((l = calloc(1, len+1)) == NULL)
115 return emit_line(p, t, l, NULL);
119 parse_link(struct parser *p, enum line_type t, const char *buf, size_t len)
122 const char *url_start;
125 return emit_line(p, t, NULL, NULL);
129 while (len > 0 && isspace(buf[0])) {
135 return emit_line(p, t, NULL, NULL);
138 while (len > 0 && !isspace(buf[0])) {
143 if ((u = calloc(1, buf - url_start + 1)) == NULL)
145 memcpy(u, url_start, buf - url_start);
150 while (len > 0 && isspace(buf[0])) {
158 if ((l = calloc(1, len + 1)) == NULL)
162 return emit_line(p, t, l, u);
165 if ((l = strdup(u)) == NULL)
167 return emit_line(p, t, l, u);
171 parse_title(struct parser *p, enum line_type t, const char *buf, size_t len)
178 return emit_line(p, t, NULL, NULL);
184 return emit_line(p, t, NULL, NULL);
190 return emit_line(p, t, NULL, NULL);
199 while (len > 0 && isspace(buf[0])) {
205 return emit_line(p, t, NULL, NULL);
207 if (t == LINE_TITLE_1 && *p->title == '\0')
208 strncpy(p->title, buf, MIN(sizeof(p->title)-1, len));
210 if ((l = calloc(1, len+1)) == NULL)
213 return emit_line(p, t, l, NULL);
217 parse_item(struct parser *p, enum line_type t, const char *buf, size_t len)
222 return emit_line(p, t, NULL, NULL);
227 while (len > 0 && isspace(buf[0])) {
233 return emit_line(p, t, NULL, NULL);
235 if ((l = calloc(1, len+1)) == NULL)
238 return emit_line(p, t, l, NULL);
242 parse_quote(struct parser *p, enum line_type t, const char *buf, size_t len)
247 return emit_line(p, t, NULL, NULL);
252 while (len > 0 && isspace(buf[0])) {
258 return emit_line(p, t, NULL, NULL);
260 if ((l = calloc(1, len+1)) == NULL)
263 return emit_line(p, t, l, NULL);
267 parse_pre_start(struct parser *p, enum line_type t, const char *buf, size_t len)
272 return emit_line(p, t, NULL, NULL);
277 while (len > 0 && isspace(buf[0])) {
283 return emit_line(p, t, NULL, NULL);
285 if ((l = calloc(1, len+1)) == NULL)
289 return emit_line(p, t, l, NULL);
293 parse_pre_cnt(struct parser *p, enum line_type t, const char *buf, size_t len)
298 return emit_line(p, t, NULL, NULL);
300 if ((l = calloc(1, len+1)) == NULL)
303 return emit_line(p, t, l, NULL);
307 parse_pre_end(struct parser *p, enum line_type t, const char *buf, size_t len)
309 return emit_line(p, t, NULL, NULL);
312 static inline enum line_type
313 detect_line_type(const char *buf, size_t len, int in_pre)
317 buf[0] == '`' && buf[1] == '`' && buf[2] == '`')
320 return LINE_PRE_CONTENT;
327 case '*': return LINE_ITEM;
328 case '>': return LINE_QUOTE;
330 if (len >= 1 && buf[1] == '>')
348 if (buf[0] == '`' && buf[1] == '`' && buf[2] == '`')
349 return LINE_PRE_START;
357 gemtext_parse(struct parser *p, const char *buf, size_t size)
359 return parser_foreach_line(p, buf, size, gemtext_foreach_line);
363 gemtext_foreach_line(struct parser *p, const char *line, size_t linelen)
367 t = detect_line_type(line, linelen, p->flags & PARSER_IN_PRE);
368 if (t == LINE_PRE_START)
369 p->flags ^= PARSER_IN_PRE;
370 if (t == LINE_PRE_END)
371 p->flags ^= PARSER_IN_PRE;
372 return parsers[t](p, t, line, linelen);
376 gemtext_free(struct parser *p)
380 /* flush the buffer */
382 t = detect_line_type(p->buf, p->len, p->flags & PARSER_IN_PRE);
383 if (!parsers[t](p, t, p->buf, p->len))
385 if ((p->flags & PARSER_IN_PRE) &&
386 !emit_line(p, LINE_PRE_END, NULL, NULL))
393 * use the first level 2 or 3 header as page title if none
396 if (*p->title == '\0')
397 search_title(p, LINE_TITLE_2);
398 if (*p->title == '\0')
399 search_title(p, LINE_TITLE_3);
405 search_title(struct parser *p, enum line_type level)
409 TAILQ_FOREACH(l, &p->head, lines) {
410 if (l->type == level) {
413 strlcpy(p->title, l->line, sizeof(p->title));