1 /*
2  * Copyright (c) 2020 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 <stdio.h>
19 #include <string.h>
20 
21 #include "gmid.h"
22 
23 #define TEST(iri, fail, exp, descr)				\
24 	if (!run_test(iri, fail, exp)) {			\
25 		fprintf(stderr, "%s:%d: error: %s\n",		\
26 		    __FILE__, __LINE__, descr);			\
27 		exit(1);					\
28 	}
29 
30 #define IRI(schema, host, port, path, query, frag)		\
31 	((struct iri){schema, host, port, 0, path, query, frag})
32 
33 #define DIFF(wanted, got, field)					\
34 	if (wanted->field == NULL || got->field == NULL ||		\
35 	    strcmp(wanted->field, got->field)) {			\
36 		fprintf(stderr, #field ":\n\tgot: %s\n\twanted: %s\n",	\
37 		    got->field, wanted->field);				\
38 		return 0;						\
39 	}
40 
41 #define PASS 0
42 #define FAIL 1
43 
44 int
45 diff_iri(struct iri *p, struct iri *exp)
46 {
47         DIFF(p, exp, schema);
48         DIFF(p, exp, host);
49         DIFF(p, exp, port);
50         DIFF(p, exp, path);
51         DIFF(p, exp, query);
52         DIFF(p, exp, fragment);
53 	return 1;
54 }
55 
56 int
57 run_test(const char *iri, int should_fail, struct iri expected)
58 {
59 	int failed, ok = 1;
60 	char *iri_copy;
61 	struct iri parsed;
62 	const char *error;
63 
64 	if ((iri_copy = strdup(iri)) == NULL)
65 		err(1, "strdup");
66 
67 	fprintf(stderr, "=> %s\n", iri);
68 	failed = !parse_iri(iri_copy, &parsed, &error);
69 
70 	if (failed && should_fail)
71 		goto done;
72 
73 	if (error != NULL)
74 		fprintf(stderr, "> %s\n", error);
75 
76 	ok = !failed && !should_fail;
77 	if (ok)
78 		ok = diff_iri(&expected, &parsed);
79 
80 done:
81 	free(iri_copy);
82 	return ok;
83 }
84 
85 int
86 main(void)
87 {
88 	struct iri empty = {"", "", "", PASS, "", "", ""};
89 
90 	TEST("http://omarpolo.com",
91 	    PASS,
92 	    IRI("http", "omarpolo.com", "", "", "", ""),
93 	    "can parse iri with empty path");
94 
95 	/* schema */
96 	TEST("omarpolo.com", FAIL, empty, "FAIL when the schema is missing");
97 	TEST("gemini:/omarpolo.com", FAIL, empty, "FAIL with invalid marker");
98 	TEST("gemini//omarpolo.com", FAIL, empty, "FAIL with invalid marker");
99 	TEST("h!!p://omarpolo.com", FAIL, empty, "FAIL with invalid schema");
100 
101 	/* authority */
102 	TEST("gemini://omarpolo.com",
103 	    PASS,
104 	    IRI("gemini", "omarpolo.com", "", "", "", ""),
105 	    "can parse authority with empty path");
106 	TEST("gemini://omarpolo.com/",
107 	    PASS,
108 	    IRI("gemini", "omarpolo.com", "", "", "", ""),
109 	    "can parse authority with empty path (alt)")
110 	TEST("gemini://omarpolo.com:1965",
111 	    PASS,
112 	    IRI("gemini", "omarpolo.com", "1965", "", "", ""),
113 	    "can parse with port and empty path");
114 	TEST("gemini://omarpolo.com:1965/",
115 	    PASS,
116 	    IRI("gemini", "omarpolo.com", "1965", "", "", ""),
117 	    "can parse with port and empty path")
118 	TEST("gemini://omarpolo.com:196s",
119 	    FAIL,
120 	    empty,
121 	    "FAIL with invalid port number");
122 
123 	/* path */
124 	TEST("gemini://omarpolo.com/foo/bar/baz",
125 	    PASS,
126 	    IRI("gemini", "omarpolo.com", "", "foo/bar/baz", "", ""),
127 	    "parse simple paths");
128 	TEST("gemini://omarpolo.com/foo//bar///baz",
129 	    PASS,
130 	    IRI("gemini", "omarpolo.com", "", "foo/bar/baz", "", ""),
131 	    "parse paths with multiple slashes");
132 	TEST("gemini://omarpolo.com/foo/./bar/./././baz",
133 	    PASS,
134 	    IRI("gemini", "omarpolo.com", "", "foo/bar/baz", "", ""),
135 	    "parse paths with . elements");
136 	TEST("gemini://omarpolo.com/foo/bar/../bar/baz",
137 	    PASS,
138 	    IRI("gemini", "omarpolo.com", "", "foo/bar/baz", "", ""),
139 	    "parse paths with .. elements");
140 	TEST("gemini://omarpolo.com/foo/../foo/bar/../bar/baz/../baz",
141 	    PASS,
142 	    IRI("gemini", "omarpolo.com", "", "foo/bar/baz", "", ""),
143 	    "parse paths with multiple .. elements");
144 	TEST("gemini://omarpolo.com/foo/..",
145 	    PASS,
146             IRI("gemini", "omarpolo.com", "", "", "", ""),
147 	    "parse paths with a trailing ..");
148 	TEST("gemini://omarpolo.com/foo/../",
149 	    PASS,
150             IRI("gemini", "omarpolo.com", "", "", "", ""),
151 	    "parse paths with a trailing ..");
152 	TEST("gemini://omarpolo.com/foo/../..",
153 	    FAIL,
154             empty,
155 	    "reject paths that would escape the root");
156 	TEST("gemini://omarpolo.com/foo/../foo/../././/bar/baz/.././.././/",
157 	    PASS,
158             IRI("gemini", "omarpolo.com", "", "", "", ""),
159 	    "parse path with lots of cleaning available");
160 
161 	/* query */
162 	TEST("foo://example.com/foo/?gne",
163 	    PASS,
164 	    IRI("foo", "example.com", "", "foo/", "gne", ""),
165 	    "parse query strings");
166 	TEST("foo://example.com/foo/?gne&foo",
167 	    PASS,
168 	    IRI("foo", "example.com", "", "foo/", "gne&foo", ""),
169 	    "parse query strings");
170 	TEST("foo://example.com/foo/?gne%2F",
171 	    PASS,
172 	    IRI("foo", "example.com", "", "foo/", "gne/", ""),
173 	    "parse query strings");
174 
175 	/* fragment */
176 	TEST("foo://bar.co/#foo",
177 	    PASS,
178 	    IRI("foo", "bar.co", "", "", "", "foo"),
179 	    "can recognize fragments");
180 
181 	/* percent encoding */
182 	TEST("foo://bar.com/caf%C3%A8.gmi",
183 	    PASS,
184 	    IRI("foo", "bar.com", "", "cafè.gmi", "", ""),
185 	    "can decode");
186 	TEST("foo://bar.com/caff%C3%A8%20macchiato.gmi",
187 	    PASS,
188 	    IRI("foo", "bar.com", "", "caffè macchiato.gmi", "", ""),
189 	    "can decode");
190 	TEST("foo://bar.com/caff%C3%A8+macchiato.gmi",
191 	    PASS,
192 	    IRI("foo", "bar.com", "", "caffè+macchiato.gmi", "", ""),
193 	    "can decode");
194 	TEST("foo://bar.com/foo%2F..%2F..",
195 	    FAIL,
196 	    empty,
197 	    "conversion and checking are done in the correct order");
198 	TEST("foo://bar.com/foo%00?baz",
199 	    FAIL,
200 	    empty,
201 	    "rejects %00");
202 
203 	/* IRI */
204         TEST("foo://bar.com/cafè.gmi",
205 	    PASS,
206 	    IRI("foo", "bar.com", "", "cafè.gmi", "" , ""),
207 	    "decode IRI (with a 2-byte utf8 seq)");
208 	TEST("foo://bar.com/世界.gmi",
209 	    PASS,
210 	    IRI("foo", "bar.com", "", "世界.gmi", "" , ""),
211 	    "decode IRI");
212 	TEST("foo://bar.com/😼.gmi",
213 	    PASS,
214 	    IRI("foo", "bar.com", "", "😼.gmi", "" , ""),
215 	    "decode IRI (with a 3-byte utf8 seq)");
216 	TEST("foo://bar.com/😼/𤭢.gmi",
217 	    PASS,
218 	    IRI("foo", "bar.com", "", "😼/𤭢.gmi", "" , ""),
219 	    "decode IRI (with a 3-byte and a 4-byte utf8 seq)");
220 	TEST("foo://bar.com/世界/\xC0\x80",
221 	    FAIL,
222 	    empty,
223 	    "reject invalid sequence (overlong NUL)");
224 
225 	return 0;
226 }