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 "hist.h"
23 #include "parser.h"
24 #include "telescope.h"
25 #include "xwrapper.h"
26 
27 static int parser_foreach_line(struct buffer *, const char *, size_t);
28 
29 void
30 parser_init(struct buffer *buffer, const struct parser *p)
31 {
32 	erase_buffer(buffer);
33 
34 	memset(buffer->title, 0, sizeof(buffer->title));
35 	buffer->parser = p;
36 	buffer->mode = p->name;
37 	buffer->parser_flags = p->initflags;
38 }
39 
40 int
41 parser_parse(struct buffer *buffer, const char *chunk, size_t len)
42 {
43 	const struct parser *p = buffer->parser;
44 
45 	if (p->parse)
46 		return p->parse(buffer, chunk, len);
47 	return parser_foreach_line(buffer, chunk, len);
48 }
49 
50 int
51 parser_parsef(struct buffer *buffer, const char *fmt, ...)
52 {
53 	char *s;
54 	va_list ap;
55 	int r;
56 
57 	va_start(ap, fmt);
58 	r = vasprintf(&s, fmt, ap);
59 	va_end(ap);
60 
61 	if (r == -1)
62 		return 0;
63 
64 	r = parser_parse(buffer, s, strlen(s));
65 	free(s);
66 	return r;
67 }
68 
69 int
70 parser_free(struct tab *tab)
71 {
72 	struct buffer		*buffer = &tab->buffer;
73 	const struct parser	*p = buffer->parser;
74 	int			 r = 1;
75 	char			*tilde, *slash;
76 
77 	if (p->free) {
78 		r = p->free(buffer);
79 	} else if (buffer->len != 0) {
80 		if (p->parse)
81 			r = p->parse(buffer, buffer->buf, buffer->len);
82 		else
83 			r = parser_foreach_line(buffer, buffer->buf,
84 			    buffer->len);
85 	}
86 
87 	free(buffer->buf);
88 	buffer->buf = NULL;
89 	buffer->len = 0;
90 
91 	if (*buffer->title != '\0')
92 		return r;
93 
94 	/*
95 	 * heuristic: see if there is a "tilde user" and use that as
96 	 * page title, using the full domain name as fallback.
97 	 */
98 	if ((tilde = strstr(hist_cur(tab->hist), "/~")) != NULL) {
99 		strlcpy(buffer->title, tilde+1, sizeof(buffer->title));
100 
101 		if ((slash = strchr(buffer->title, '/')) != NULL)
102 			*slash = '\0';
103 	} else
104 		strlcpy(buffer->title, tab->iri.iri_host,
105 		    sizeof(buffer->title));
106 
107 	return r;
108 }
109 
110 int
111 parser_serialize(struct buffer *b, FILE *fp)
112 {
113 	const struct parser	*p = b->parser;
114 	struct line		*line;
115 	const char		*text;
116 	int			 r;
117 
118 	if (p->serialize != NULL)
119 		return p->serialize(b, fp);
120 
121 	/* a default implementation good enough for plain text */
122 	TAILQ_FOREACH(line, &b->head, lines) {
123 		if ((text = line->line) == NULL)
124 			text = "";
125 
126 		r = fprintf(fp, "%s\n", text);
127 		if (r == -1)
128 			return 0;
129 	}
130 
131 	return 1;
132 }
133 
134 static int
135 parser_append(struct buffer *b, const char *buf, size_t len)
136 {
137 	size_t newlen;
138 	char *t;
139 
140 	if (len == 0)
141 		return (1);
142 
143 	newlen = len + b->len;
144 	t = xcalloc(1, newlen);
145 	memcpy(t, b->buf, b->len);
146 	memcpy(t + b->len, buf, len);
147 	free(b->buf);
148 	b->buf = t;
149 	b->len = newlen;
150 	return 1;
151 }
152 
153 static int
154 parser_set_buf(struct buffer *b, const char *buf, size_t len)
155 {
156 	char *tmp;
157 
158 	if (len == 0) {
159 		b->len = 0;
160 		free(b->buf);
161 		b->buf = NULL;
162 		return 1;
163 	}
164 
165 	/*
166 	 * p->buf and buf can (and probably almost always will)
167 	 * overlap!
168 	 */
169 
170 	tmp = xcalloc(1, len);
171 	memcpy(tmp, buf, len);
172 	free(b->buf);
173 	b->buf = tmp;
174 	b->len = len;
175 	return 1;
176 }
177 
178 static int
179 parser_foreach_line(struct buffer *b, const char *buf, size_t size)
180 {
181 	const struct parser	*p = b->parser;
182 	char			*beg, *end;
183 	unsigned int		 ch;
184 	size_t			 i, l, len;
185 
186 	if (!parser_append(b, buf, size))
187 		return 0;
188 	beg = b->buf;
189 	len = b->len;
190 
191 	if (!(b->parser_flags & PARSER_IN_BODY) && len < 3)
192 		return 1;
193 
194 	if (!(b->parser_flags & PARSER_IN_BODY)) {
195 		b->parser_flags |= PARSER_IN_BODY;
196 
197 		/*
198 		 * drop the BOM: only UTF-8 is supported, and there
199 		 * it's useless; some editors may still add one
200 		 * though.
201 		 */
202 		if (memmem(beg, len, "\xEF\xBB\xBF", 3) == beg) {
203 			b += 3;
204 			len -= 3;
205 		}
206 	}
207 
208 	/* drop every "funny" ASCII character */
209 	for (i = 0; i < len; ) {
210 		ch = beg[i];
211 		if ((ch >= ' ' || ch == '\n' || ch == '\t')
212 		    && ch != 127) { /* del */
213 			++i;
214 			continue;
215 		}
216 		memmove(&beg[i], &beg[i+1], len - i - 1);
217 		len--;
218 	}
219 
220 	while (len > 0) {
221 		if ((end = memmem((char*)beg, len, "\n", 1)) == NULL)
222 			break;
223 		l = end - beg;
224 
225 		if (!p->parseline(b, beg, l))
226 			return 0;
227 
228 		len -= l;
229 		beg += l;
230 
231 		if (len > 0) {
232 			/* skip \n */
233 			len--;
234 			beg++;
235 		}
236 	}
237 
238 	return parser_set_buf(b, beg, len);
239 }