1 /*
2  * Copyright (c) 2021 Omar Polo <op@omarpolo.com>
3  *
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.
7  *
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.
15  */
16 
17 #include "compat.h"
18 
19 #include <stdlib.h>
20 #include <string.h>
21 
22 #include "parser.h"
23 #include "telescope.h"
24 #include "ui.h"
25 
26 /*
27  * Load a text/gemini page given the string page.  Always returns 0.
28  */
29 int
30 load_page_from_str(struct tab *tab, const char *page)
31 {
32 	parser_init(tab, gemtext_initparser);
33 	if (!tab->buffer.page.parse(&tab->buffer.page, page, strlen(page)))
34 		abort();
35 	if (!tab->buffer.page.free(&tab->buffer.page))
36 		abort();
37 	ui_on_tab_refresh(tab);
38 	ui_on_tab_loaded(tab);
39 	return 0;
40 }
41 
42 void
43 parser_init(struct tab *tab, parserfn fn)
44 {
45 	erase_buffer(&tab->buffer);
46 	fn(&tab->buffer.page);
47 }
48 
49 int
50 parser_parse(struct tab *tab, const char *chunk, size_t len)
51 {
52 	return tab->buffer.page.parse(&tab->buffer.page, chunk, len);
53 }
54 
55 int
56 parser_free(struct tab *tab)
57 {
58 	int	 r;
59 	char	*tilde, *slash;
60 
61 	r = tab->buffer.page.free(&tab->buffer.page);
62 
63 	if (*tab->buffer.page.title != '\0')
64 		return r;
65 
66 	/*
67 	 * heuristic: see if there is a "tilde user" and use that as
68 	 * page title, using the full domain name as fallback.
69 	 */
70 	if ((tilde = strstr(tab->hist_cur->h, "/~")) != NULL) {
71 		strlcpy(tab->buffer.page.title, tilde+1,
72 		    sizeof(tab->buffer.page.title));
73 
74 		if ((slash = strchr(tab->buffer.page.title, '/')) != NULL)
75 			*slash = '\0';
76 	} else
77 		strlcpy(tab->buffer.page.title, tab->uri.host,
78 		    sizeof(tab->buffer.page.title));
79 
80 	return r;
81 }
82 
83 int
84 parser_append(struct parser *p, const char *buf, size_t len)
85 {
86 	size_t newlen;
87 	char *t;
88 
89 	newlen = len + p->len;
90 	if ((t = calloc(1, newlen)) == NULL)
91 		return 0;
92 	memcpy(t, p->buf, p->len);
93 	memcpy(t + p->len, buf, len);
94 	free(p->buf);
95 	p->buf = t;
96 	p->len = newlen;
97 	return 1;
98 }
99 
100 int
101 parser_set_buf(struct parser *p, const char *buf, size_t len)
102 {
103 	char *tmp;
104 
105 	if (len == 0) {
106 		p->len = 0;
107 		free(p->buf);
108 		p->buf = NULL;
109 		return 1;
110 	}
111 
112 	/*
113 	 * p->buf and buf can (and probably almost always will)
114 	 * overlap!
115 	 */
116 
117 	if ((tmp = calloc(1, len)) == NULL)
118 		return 0;
119 	memcpy(tmp, buf, len);
120 	free(p->buf);
121 	p->buf = tmp;
122 	p->len = len;
123 	return 1;
124 }
125 
126 int
127 parser_foreach_line(struct parser *p, const char *buf, size_t size,
128     parsechunkfn fn)
129 {
130 	char		*b, *e;
131 	unsigned int	 ch;
132 	size_t		 i, l, len;
133 
134 	if (!parser_append(p, buf, size))
135 		return 0;
136 	b = p->buf;
137 	len = p->len;
138 
139 	if (!(p->flags & PARSER_IN_BODY) && len < 3)
140 		return 1;
141 
142 	if (!(p->flags & PARSER_IN_BODY)) {
143 		p->flags |= PARSER_IN_BODY;
144 
145 		/*
146 		 * drop the BOM: only UTF-8 is supported, and there
147 		 * it's useless; some editors may still add one
148 		 * though.
149 		 */
150 		if (memmem(b, len, "\xEF\xBB\xBF", 3) == b) {
151 			b += 3;
152 			len -= 3;
153 		}
154 	}
155 
156 	/* drop every "funny" ASCII character */
157 	for (i = 0; i < len; ) {
158 		ch = b[i];
159 		if ((ch >= ' ' || ch == '\n' || ch == '\t')
160 		    && ch != 127) { /* del */
161 			++i;
162 			continue;
163 		}
164 		memmove(&b[i], &b[i+1], len - i - 1);
165 		len--;
166 	}
167 
168 	while (len > 0) {
169 		if ((e = memmem((char*)b, len, "\n", 1)) == NULL)
170 			break;
171 		l = e - b;
172 
173 		if (!fn(p, b, l))
174 			return 0;
175 
176 		len -= l;
177 		b += l;
178 
179 		if (len > 0) {
180 			/* skip \n */
181 			len--;
182 			b++;
183 		}
184 	}
185 
186 	return parser_set_buf(p, b, len);
187 }