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_3] = parse_title,
54 [LINE_QUOTE] = parse_quote,
55 [LINE_PRE_START] = parse_pre_start,
56 [LINE_PRE_END] = parse_pre_end,
60 gemtext_initparser(struct parser *p)
62 memset(p, 0, sizeof(*p));
64 p->name = "text/gemini";
65 p->parse = &gemtext_parse;
66 p->free = &gemtext_free;
70 emit_line(struct parser *p, enum line_type type, char *line, char *alt)
74 if ((l = calloc(1, sizeof(*l))) == NULL)
87 case LINE_PRE_CONTENT:
96 if (TAILQ_EMPTY(&p->head))
97 TAILQ_INSERT_HEAD(&p->head, l, lines);
99 TAILQ_INSERT_TAIL(&p->head, l, lines);
105 parse_text(struct parser *p, enum line_type t, const char *buf, size_t len)
109 if ((l = calloc(1, len+1)) == NULL)
112 return emit_line(p, t, l, NULL);
116 parse_link(struct parser *p, enum line_type t, const char *buf, size_t len)
119 const char *url_start;
122 return emit_line(p, t, NULL, NULL);
126 while (len > 0 && isspace(buf[0])) {
132 return emit_line(p, t, NULL, NULL);
135 while (len > 0 && !isspace(buf[0])) {
140 if ((u = calloc(1, buf - url_start + 1)) == NULL)
142 memcpy(u, url_start, buf - url_start);
147 while (len > 0 && isspace(buf[0])) {
155 if ((l = calloc(1, len + 1)) == NULL)
159 return emit_line(p, t, l, u);
162 if ((l = strdup(u)) == NULL)
164 return emit_line(p, t, l, u);
168 parse_title(struct parser *p, enum line_type t, const char *buf, size_t len)
175 return emit_line(p, t, NULL, NULL);
181 return emit_line(p, t, NULL, NULL);
187 return emit_line(p, t, NULL, NULL);
196 while (len > 0 && isspace(buf[0])) {
202 return emit_line(p, t, NULL, NULL);
204 if (t == LINE_TITLE_1 && *p->title == '\0')
205 strncpy(p->title, buf, MIN(sizeof(p->title)-1, len));
207 if ((l = calloc(1, len+1)) == NULL)
210 return emit_line(p, t, l, NULL);
214 parse_item(struct parser *p, enum line_type t, const char *buf, size_t len)
219 return emit_line(p, t, NULL, NULL);
224 while (len > 0 && isspace(buf[0])) {
230 return emit_line(p, t, NULL, NULL);
232 if ((l = calloc(1, len+1)) == NULL)
235 return emit_line(p, t, l, NULL);
239 parse_quote(struct parser *p, enum line_type t, const char *buf, size_t len)
244 return emit_line(p, t, NULL, NULL);
249 while (len > 0 && isspace(buf[0])) {
255 return emit_line(p, t, NULL, NULL);
257 if ((l = calloc(1, len+1)) == NULL)
260 return emit_line(p, t, l, NULL);
264 parse_pre_start(struct parser *p, enum line_type t, const char *buf, size_t len)
269 return emit_line(p, t, NULL, NULL);
274 while (len > 0 && isspace(buf[0])) {
280 return emit_line(p, t, NULL, NULL);
282 if ((l = calloc(1, len+1)) == NULL)
286 return emit_line(p, t, l, NULL);
290 parse_pre_cnt(struct parser *p, enum line_type t, const char *buf, size_t len)
295 return emit_line(p, t, NULL, NULL);
297 if ((l = calloc(1, len+1)) == NULL)
300 return emit_line(p, t, l, NULL);
304 parse_pre_end(struct parser *p, enum line_type t, const char *buf, size_t len)
306 return emit_line(p, t, NULL, NULL);
309 static inline enum line_type
310 detect_line_type(const char *buf, size_t len, int in_pre)
314 buf[0] == '`' && buf[1] == '`' && buf[2] == '`')
317 return LINE_PRE_CONTENT;
324 case '*': return LINE_ITEM;
325 case '>': return LINE_QUOTE;
327 if (len >= 1 && buf[1] == '>')
345 if (buf[0] == '`' && buf[1] == '`' && buf[2] == '`')
346 return LINE_PRE_START;
354 gemtext_parse(struct parser *p, const char *buf, size_t size)
356 return parser_foreach_line(p, buf, size, gemtext_foreach_line);
360 gemtext_foreach_line(struct parser *p, const char *line, size_t linelen)
364 t = detect_line_type(line, linelen, p->flags & PARSER_IN_PRE);
365 if (t == LINE_PRE_START)
366 p->flags ^= PARSER_IN_PRE;
367 if (t == LINE_PRE_END)
368 p->flags ^= PARSER_IN_PRE;
369 return parsers[t](p, t, line, linelen);
373 gemtext_free(struct parser *p)
377 /* flush the buffer */
379 t = detect_line_type(p->buf, p->len, p->flags & PARSER_IN_PRE);
380 if (!parsers[t](p, t, p->buf, p->len))
382 if ((p->flags & PARSER_IN_PRE) &&
383 !emit_line(p, LINE_PRE_END, NULL, NULL))
390 * use the first level 2 or 3 header as page title if none
393 if (*p->title == '\0')
394 search_title(p, LINE_TITLE_2);
395 if (*p->title == '\0')
396 search_title(p, LINE_TITLE_3);
402 search_title(struct parser *p, enum line_type level)
406 TAILQ_FOREACH(l, &p->head, lines) {
407 if (l->type == level) {
410 strlcpy(p->title, l->line, sizeof(p->title));