1 /*
2  * Copyright (c) 2022 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 <err.h>
18 #include <errno.h>
19 #include <stdint.h>
20 #include <stdio.h>
21 #include <string.h>
22 
23 #include "iri.h"
24 
25 static int
26 resolve(const char *base, const char *ref, const char *expected)
27 {
28 	static struct iri	i;
29 	char			buf[512];
30 
31 	if (iri_parse(base, ref, &i) == -1) {
32 		fprintf(stderr, "FAIL resolve(\"%s\", \"%s\") %s\n", base, ref,
33 		    strerror(errno));
34 		return (1);
35 	}
36 
37 	if (iri_unparse(&i, buf, sizeof(buf)) == -1) {
38 		fprintf(stderr, "FAIL resolve(\"%s\", \"%s\") %s\n", base, ref,
39 		    strerror(errno));
40 		return (1);
41 	}
42 
43 	if (strcmp(buf, expected) != 0) {
44 		fprintf(stderr, "FAIL resolve(\"%s\", \"%s\")\n", base, ref);
45 		fprintf(stderr, "got:\t%s\n", buf);
46 		fprintf(stderr, "want:\t%s\n", expected);
47 		return (1);
48 	}
49 
50 	fprintf(stderr, "OK resolve(\"%s\", \"%s\") -> %s\n", base, ref,
51 	    expected);
52 	return (0);
53 }
54 
55 static int
56 setquery(const char *iri, const char *query, const char *expected)
57 {
58 	static struct iri	i;
59 	char			buf[512];
60 
61 	if (iri_parse(NULL, iri, &i) == -1) {
62 		fprintf(stderr, "FAIL can't parse <%s>: %s\n", iri,
63 		    strerror(errno));
64 		return (1);
65 	}
66 
67 	if (iri_setquery(&i, query) == -1) {
68 		fprintf(stderr, "FAIL setting query \"%s\": %s\n", query,
69 		    strerror(errno));
70 		return (1);
71 	}
72 
73 	if (iri_unparse(&i, buf, sizeof(buf)) == -1) {
74 		fprintf(stderr, "FAIL unparsing <%s> with query %s\n",
75 		    iri, query);
76 		return (1);
77 	}
78 
79 	if (strcmp(buf, expected) != 0) {
80 		fprintf(stderr, "FAIL setquery(\"%s\", \"%s\")\n", iri, query);
81 		fprintf(stderr, "got:\t%s\n", buf);
82 		fprintf(stderr, "want:\t%s\n", expected);
83 		return (1);
84 	}
85 
86 	fprintf(stderr, "OK setquery(\"%s\", \"%s\") -> %s\n", iri, query,
87 	    expected);
88 	return (0);
89 }
90 
91 static int
92 urlencode(const char *str, const char *exp)
93 {
94 	char			 enc[512];
95 	char			 dec[512];
96 
97 	if (iri_urlescape(str, enc, sizeof(enc)) == -1) {
98 		fprintf(stderr, "FAIL can't percent encode %s\n", str);
99 		return (1);
100 	}
101 
102 	if (strcmp(enc, exp) != 0) {
103 		fprintf(stderr, "FAIL %%enc: expecting <%s>; got <%s>\n",
104 		    exp, enc);
105 		return (1);
106 	}
107 
108 	if (iri_urlunescape(enc, dec, sizeof(dec)) == -1) {
109 		fprintf(stderr, "FAIL can't %%decode %s\n", enc);
110 		return (1);
111 	}
112 
113 	if (strcmp(str, dec) != 0) {
114 		fprintf(stderr,
115 		    "FAIL urlencode/decode not identity: <%s> vs <%s>\n",
116 		    str, dec);
117 		return (1);
118 	}
119 
120 	if (strcmp(enc, exp) != 0) {
121 		fprintf(stderr, "FAIL %%enc: expecting <%s>; got <%s>\n",
122 		    exp, enc);
123 		return (1);
124 	}
125 
126 	fprintf(stderr, "OK urlencode(\"%s\") -> %s\n", str, enc);
127 	return (0);
128 }
129 
130 int
131 main(void)
132 {
133 	const char	*base;
134 	int		 ret = 0;
135 
136 	/* RFC 3986 tests */
137 
138 	base = "http://a/b/c/d;p?q";
139 
140 	ret |= resolve(base, "g:h", "g:h");
141 	ret |= resolve(base, "g", "http://a/b/c/g");
142 	ret |= resolve(base, "./g", "http://a/b/c/g");
143 	ret |= resolve(base, "g/", "http://a/b/c/g/");
144 	ret |= resolve(base, "/g", "http://a/g");
145 
146 	/*
147 	 * the RFC says "http://g" but we always normalize an
148 	 * empty path to "/" if there is an authority.
149 	 */
150 	ret |= resolve(base, "//g", "http://g/");
151 
152 	ret |= resolve(base, "?y", "http://a/b/c/d;p?y");
153 	ret |= resolve(base, "g?y", "http://a/b/c/g?y");
154 	ret |= resolve(base, "#s", "http://a/b/c/d;p?q#s");
155 	ret |= resolve(base, "g#s", "http://a/b/c/g#s");
156 	ret |= resolve(base, "g?y#s", "http://a/b/c/g?y#s");
157 	ret |= resolve(base, ";x", "http://a/b/c/;x");
158 	ret |= resolve(base, "g;x", "http://a/b/c/g;x");
159 	ret |= resolve(base, "g;x?y#s", "http://a/b/c/g;x?y#s");
160 	ret |= resolve(base, "", "http://a/b/c/d;p?q");
161 	ret |= resolve(base, ".", "http://a/b/c/");
162 	ret |= resolve(base, "./", "http://a/b/c/");
163 	ret |= resolve(base, "..", "http://a/b/");
164 	ret |= resolve(base, "../", "http://a/b/");
165 	ret |= resolve(base, "../g", "http://a/b/g");
166 	ret |= resolve(base, "../..", "http://a/");
167 	ret |= resolve(base, "../../", "http://a/");
168 	ret |= resolve(base, "../../g", "http://a/g");
169 
170 	ret |= resolve(base, "../../../g", "http://a/g");
171 	ret |= resolve(base, "../../../../g", "http://a/g");
172 	ret |= resolve(base, "/./g", "http://a/g");
173 	ret |= resolve(base, "/../g", "http://a/g");
174 	ret |= resolve(base, "g.", "http://a/b/c/g.");
175 	ret |= resolve(base, ".g", "http://a/b/c/.g");
176 	ret |= resolve(base, "g..", "http://a/b/c/g..");
177 	ret |= resolve(base, "..g", "http://a/b/c/..g");
178 
179 	ret |= resolve(base, "./../g", "http://a/b/g");
180 	ret |= resolve(base, "./g/.", "http://a/b/c/g/");
181 	ret |= resolve(base, "g/./h", "http://a/b/c/g/h");
182 	ret |= resolve(base, "g/../h", "http://a/b/c/h");
183 	ret |= resolve(base, "g;x=1/./y", "http://a/b/c/g;x=1/y");
184 	ret |= resolve(base, "g;x=1/../y", "http://a/b/c/y");
185 
186 	ret |= resolve(base, "g?y/./x", "http://a/b/c/g?y/./x");
187 	ret |= resolve(base, "g?y/../x", "http://a/b/c/g?y/../x");
188 	ret |= resolve(base, "g#s/./x", "http://a/b/c/g#s/./x");
189 	ret |= resolve(base, "g#s/../x", "http://a/b/c/g#s/../x");
190 
191 	ret |= resolve(base, "http:g", "http:g");
192 
193 	/* extra tests */
194 
195 	ret |= resolve(base, "gopher://b:70", "gopher://b:70/");
196 	ret |= resolve(base, "gopher://b:70/1/foo", "gopher://b:70/1/foo");
197 
198 	base = "gopher://b:70/1/bar/foo/quux";
199 	ret |= resolve(base, "/",  "gopher://b:70/");
200 	ret |= resolve(base, "..", "gopher://b:70/1/bar/");
201 
202 	base = "gemini://a/b/c";
203 	ret |= setquery(base, "hw", "gemini://a/b/c?hw");
204 	ret |= setquery(base, "h w", "gemini://a/b/c?h%20w");
205 	ret |= setquery(base, "100%", "gemini://a/b/c?100%25");
206 	ret |= setquery(base, "%20", "gemini://a/b/c?%2520");
207 
208 	base = "gemini://a/?foo+bar";
209 	ret |= setquery(base, "foo", "gemini://a/?foo");
210 	ret |= setquery(base, "%20", "gemini://a/?%2520");
211 	ret |= setquery(base, "%20%20%20", "gemini://a/?%2520%2520%2520");
212 
213 	ret |= resolve(base, "file:///tmp/foo", "file:///tmp/foo");
214 	ret |= resolve(base, "file:/tmp/foo", "file:///tmp/foo");
215 
216 	ret |= urlencode("foobar", "foobar");
217 	ret |= urlencode("foo/bar", "foo/bar");
218 	ret |= urlencode("foo bar", "foo%20bar");
219 	ret |= urlencode("/Teloschistes flavicans",
220 	    "/Teloschistes%20flavicans");
221 
222 	return (ret);
223 }