• Home
  • Line#
  • Scopes#
  • Navigate#
  • Raw
  • Download
1 // Copyright 2007, Google Inc.
2 // All rights reserved.
3 //
4 // Redistribution and use in source and binary forms, with or without
5 // modification, are permitted provided that the following conditions are
6 // met:
7 //
8 //     * Redistributions of source code must retain the above copyright
9 // notice, this list of conditions and the following disclaimer.
10 //     * Redistributions in binary form must reproduce the above
11 // copyright notice, this list of conditions and the following disclaimer
12 // in the documentation and/or other materials provided with the
13 // distribution.
14 //     * Neither the name of Google Inc. nor the names of its
15 // contributors may be used to endorse or promote products derived from
16 // this software without specific prior written permission.
17 //
18 // THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
19 // "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
20 // LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
21 // A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
22 // OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
23 // SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
24 // LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
25 // DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
26 // THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
27 // (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
28 // OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
29 
30 #include "base/basictypes.h"
31 #include "googleurl/src/url_parse.h"
32 #include "testing/gtest/include/gtest/gtest.h"
33 
34 // Some implementations of base/basictypes.h may define ARRAYSIZE.
35 // If it's not defined, we define it to the ARRAYSIZE_UNSAFE macro
36 // which is in our version of basictypes.h.
37 #ifndef ARRAYSIZE
38 #define ARRAYSIZE ARRAYSIZE_UNSAFE
39 #endif
40 
41 // Interesting IE file:isms...
42 //
43 //  file:/foo/bar              file:///foo/bar
44 //      The result here seems totally invalid!?!? This isn't UNC.
45 //
46 //  file:/
47 //  file:// or any other number of slashes
48 //      IE6 doesn't do anything at all if you click on this link. No error:
49 //      nothing. IE6's history system seems to always color this link, so I'm
50 //      guessing that it maps internally to the empty URL.
51 //
52 //  C:\                        file:///C:/
53 //  /                          file:///C:/
54 //  /foo                       file:///C:/foo
55 //      Interestingly, IE treats "/" as an alias for "c:\", which makes sense,
56 //      but is weird to think about on Windows.
57 //
58 //  file:foo/                  file:foo/  (invalid?!?!?)
59 //  file:/foo/                 file:///foo/  (invalid?!?!?)
60 //  file://foo/                file://foo/   (UNC to server "foo")
61 //  file:///foo/               file:///foo/  (invalid)
62 //  file:////foo/              file://foo/   (UNC to server "foo")
63 //      Any more than four slashes is also treated as UNC.
64 //
65 //  file:C:/                   file://C:/
66 //  file:/C:/                  file://C:/
67 //      The number of slashes after "file:" don't matter if the thing following
68 //      it looks like an absolute drive path. Also, slashes and backslashes are
69 //      equally valid here.
70 
71 namespace {
72 
73 // Used for regular URL parse cases.
74 struct URLParseCase {
75   const char* input;
76 
77   const char* scheme;
78   const char* username;
79   const char* password;
80   const char* host;
81   int port;
82   const char* path;
83   const char* query;
84   const char* ref;
85 };
86 
87 // Simpler version of URLParseCase for testing path URLs.
88 struct PathURLParseCase {
89   const char* input;
90 
91   const char* scheme;
92   const char* path;
93 };
94 
95 // Simpler version of URLParseCase for testing mailto URLs.
96 struct MailtoURLParseCase {
97   const char* input;
98 
99   const char* scheme;
100   const char* path;
101   const char* query;
102 };
103 
104 
ComponentMatches(const char * input,const char * reference,const url_parse::Component & component)105 bool ComponentMatches(const char* input,
106                       const char* reference,
107                       const url_parse::Component& component) {
108   // If the component is nonexistant (length == -1), it should begin at 0.
109   EXPECT_TRUE(component.len >= 0 || component.len == -1);
110 
111   // Begin should be valid.
112   EXPECT_LE(0, component.begin);
113 
114   // A NULL reference means the component should be nonexistant.
115   if (!reference)
116     return component.len == -1;
117   if (component.len < 0)
118     return false;  // Reference is not NULL but we don't have anything
119 
120   if (strlen(reference) != static_cast<size_t>(component.len))
121     return false;  // Lengths don't match
122 
123   // Now check the actual characters.
124   return strncmp(reference, &input[component.begin], component.len) == 0;
125 }
126 
ExpectInvalidComponent(const url_parse::Component & component)127 void ExpectInvalidComponent(const url_parse::Component& component) {
128   EXPECT_EQ(0, component.begin);
129   EXPECT_EQ(-1, component.len);
130 }
131 
132 }  // namespace
133 
134 // Parsed ----------------------------------------------------------------------
135 
TEST(URLParser,Length)136 TEST(URLParser, Length) {
137   const char* length_cases[] = {
138       // One with everything in it.
139     "http://user:pass@host:99/foo?bar#baz",
140       // One with nothing in it.
141     "",
142       // Working backwards, let's start taking off stuff from the full one.
143     "http://user:pass@host:99/foo?bar#",
144     "http://user:pass@host:99/foo?bar",
145     "http://user:pass@host:99/foo?",
146     "http://user:pass@host:99/foo",
147     "http://user:pass@host:99/",
148     "http://user:pass@host:99",
149     "http://user:pass@host:",
150     "http://user:pass@host",
151     "http://host",
152     "http://user@",
153     "http:",
154   };
155   for (size_t i = 0; i < arraysize(length_cases); i++) {
156     int true_length = static_cast<int>(strlen(length_cases[i]));
157 
158     url_parse::Parsed parsed;
159     url_parse::ParseStandardURL(length_cases[i], true_length, &parsed);
160 
161     EXPECT_EQ(true_length, parsed.Length());
162   }
163 }
164 
TEST(URLParser,CountCharactersBefore)165 TEST(URLParser, CountCharactersBefore) {
166   using namespace url_parse;
167   struct CountCase {
168     const char* url;
169     Parsed::ComponentType component;
170     bool include_delimiter;
171     int expected_count;
172   } count_cases[] = {
173       // Test each possibility in the case where all components are present.
174 //    0         1         2
175 //    0123456789012345678901
176     {"http://u:p@h:8/p?q#r", Parsed::SCHEME, true, 0},
177     {"http://u:p@h:8/p?q#r", Parsed::SCHEME, false, 0},
178     {"http://u:p@h:8/p?q#r", Parsed::USERNAME, true, 7},
179     {"http://u:p@h:8/p?q#r", Parsed::USERNAME, false, 7},
180     {"http://u:p@h:8/p?q#r", Parsed::PASSWORD, true, 9},
181     {"http://u:p@h:8/p?q#r", Parsed::PASSWORD, false, 9},
182     {"http://u:p@h:8/p?q#r", Parsed::HOST, true, 11},
183     {"http://u:p@h:8/p?q#r", Parsed::HOST, false, 11},
184     {"http://u:p@h:8/p?q#r", Parsed::PORT, true, 12},
185     {"http://u:p@h:8/p?q#r", Parsed::PORT, false, 13},
186     {"http://u:p@h:8/p?q#r", Parsed::PATH, false, 14},
187     {"http://u:p@h:8/p?q#r", Parsed::PATH, true, 14},
188     {"http://u:p@h:8/p?q#r", Parsed::QUERY, true, 16},
189     {"http://u:p@h:8/p?q#r", Parsed::QUERY, false, 17},
190     {"http://u:p@h:8/p?q#r", Parsed::REF, true, 18},
191     {"http://u:p@h:8/p?q#r", Parsed::REF, false, 19},
192       // Now test when the requested component is missing.
193     {"http://u:p@h:8/p?", Parsed::REF, true, 17},
194     {"http://u:p@h:8/p?q", Parsed::REF, true, 18},
195     {"http://u:p@h:8/p#r", Parsed::QUERY, true, 16},
196     {"http://u:p@h:8#r", Parsed::PATH, true, 14},
197     {"http://u:p@h/", Parsed::PORT, true, 12},
198     {"http://u:p@/", Parsed::HOST, true, 11},
199       // This case is a little weird. It will report that the password would
200       // start where the host begins. This is arguably correct, although you
201       // could also argue that it should start at the '@' sign. Doing it
202       // starting with the '@' sign is actually harder, so we don't bother.
203     {"http://u@h/", Parsed::PASSWORD, true, 9},
204     {"http://h/", Parsed::USERNAME, true, 7},
205     {"http:", Parsed::USERNAME, true, 5},
206     {"", Parsed::SCHEME, true, 0},
207       // Make sure a random component still works when there's nothing there.
208     {"", Parsed::REF, true, 0},
209       // File URLs are special with no host, so we test those.
210     {"file:///c:/foo", Parsed::USERNAME, true, 7},
211     {"file:///c:/foo", Parsed::PASSWORD, true, 7},
212     {"file:///c:/foo", Parsed::HOST, true, 7},
213     {"file:///c:/foo", Parsed::PATH, true, 7},
214   };
215   for (size_t i = 0; i < ARRAYSIZE(count_cases); i++) {
216     int length = static_cast<int>(strlen(count_cases[i].url));
217 
218     // Simple test to distinguish file and standard URLs.
219     url_parse::Parsed parsed;
220     if (length > 0 && count_cases[i].url[0] == 'f')
221       url_parse::ParseFileURL(count_cases[i].url, length, &parsed);
222     else
223       url_parse::ParseStandardURL(count_cases[i].url, length, &parsed);
224 
225     int chars_before = parsed.CountCharactersBefore(
226         count_cases[i].component, count_cases[i].include_delimiter);
227     EXPECT_EQ(count_cases[i].expected_count, chars_before);
228   }
229 }
230 
231 // Standard --------------------------------------------------------------------
232 
233 // Input                               Scheme  Usrname Passwd     Host         Port Path       Query        Ref
234 // ------------------------------------ ------- ------- ---------- ------------ --- ---------- ------------ -----
235 static URLParseCase cases[] = {
236   // Regular URL with all the parts
237 {"http://user:pass@foo:21/bar;par?b#c", "http", "user", "pass",    "foo",       21, "/bar;par","b",          "c"},
238 
239   // Known schemes should lean towards authority identification
240 {"http:foo.com",                        "http", NULL,  NULL,      "foo.com",    -1, NULL,      NULL,        NULL},
241 
242   // Spaces!
243 {"\t   :foo.com   \n",                  "",     NULL,  NULL,      "foo.com",    -1, NULL,      NULL,        NULL},
244 {" foo.com  ",                          NULL,   NULL,  NULL,      "foo.com",    -1, NULL,      NULL,        NULL},
245 {"a:\t foo.com",                        "a",    NULL,  NULL,      "\t foo.com", -1, NULL,      NULL,        NULL},
246 {"http://f:21/ b ? d # e ",             "http", NULL,  NULL,      "f",          21, "/ b ",    " d ",       " e"},
247 
248   // Invalid port numbers should be identified and turned into -2, empty port
249   // numbers should be -1. Spaces aren't allowed in port numbers
250 {"http://f:/c",                         "http", NULL,  NULL,      "f",          -1, "/c",      NULL,        NULL},
251 {"http://f:0/c",                        "http", NULL,  NULL,      "f",           0, "/c",      NULL,        NULL},
252 {"http://f:00000000000000/c",           "http", NULL,  NULL,      "f",           0, "/c",      NULL,        NULL},
253 {"http://f:00000000000000000000080/c",  "http", NULL,  NULL,      "f",          80, "/c",      NULL,        NULL},
254 {"http://f:b/c",                        "http", NULL,  NULL,      "f",          -2, "/c",      NULL,        NULL},
255 {"http://f: /c",                        "http", NULL,  NULL,      "f",          -2, "/c",      NULL,        NULL},
256 {"http://f:\n/c",                       "http", NULL,  NULL,      "f",          -2, "/c",      NULL,        NULL},
257 {"http://f:fifty-two/c",                "http", NULL,  NULL,      "f",          -2, "/c",      NULL,        NULL},
258 {"http://f:999999/c",                   "http", NULL,  NULL,      "f",          -2, "/c",      NULL,        NULL},
259 {"http://f: 21 / b ? d # e ",           "http", NULL,  NULL,      "f",          -2, "/ b ",    " d ",       " e"},
260 
261   // Creative URLs missing key elements
262 {"",                                    NULL,   NULL,  NULL,      NULL,         -1, NULL,      NULL,        NULL},
263 {"  \t",                                NULL,   NULL,  NULL,      NULL,         -1, NULL,      NULL,        NULL},
264 {":foo.com/",                           "",     NULL,  NULL,      "foo.com",    -1, "/",       NULL,        NULL},
265 {":foo.com\\",                          "",     NULL,  NULL,      "foo.com",    -1, "\\",      NULL,        NULL},
266 {":",                                   "",     NULL,  NULL,      NULL,         -1, NULL,      NULL,        NULL},
267 {":a",                                  "",     NULL,  NULL,      "a",          -1, NULL,      NULL,        NULL},
268 {":/",                                  "",     NULL,  NULL,      NULL,         -1, NULL,      NULL,        NULL},
269 {":\\",                                 "",     NULL,  NULL,      NULL,         -1, NULL,      NULL,        NULL},
270 {":#",                                  "",     NULL,  NULL,      NULL,         -1, NULL,      NULL,        ""},
271 {"#",                                   NULL,   NULL,  NULL,      NULL,         -1, NULL,      NULL,        ""},
272 {"#/",                                  NULL,   NULL,  NULL,      NULL,         -1, NULL,      NULL,        "/"},
273 {"#\\",                                 NULL,   NULL,  NULL,      NULL,         -1, NULL,      NULL,        "\\"},
274 {"#;?",                                 NULL,   NULL,  NULL,      NULL,         -1, NULL,      NULL,        ";?"},
275 {"?",                                   NULL,   NULL,  NULL,      NULL,         -1, NULL,      "",          NULL},
276 {"/",                                   NULL,   NULL,  NULL,      NULL,         -1, NULL,      NULL,        NULL},
277 {":23",                                 "",     NULL,  NULL,      "23",         -1, NULL,      NULL,        NULL},
278 {"/:23",                                "/",    NULL,  NULL,      "23",         -1, NULL,      NULL,        NULL},
279 {"//",                                  NULL,   NULL,  NULL,      NULL,         -1, NULL,      NULL,        NULL},
280 {"::",                                  "",     NULL,  NULL,      NULL,         -1, NULL,      NULL,        NULL},
281 {"::23",                                "",     NULL,  NULL,      NULL,         23, NULL,      NULL,        NULL},
282 {"foo://",                              "foo",  NULL,  NULL,      NULL,         -1, NULL,      NULL,        NULL},
283 
284   // Username/passwords and things that look like them
285 {"http://a:b@c:29/d",                   "http", "a",   "b",       "c",          29, "/d",      NULL,        NULL},
286 {"http::@c:29",                         "http", "",    "",        "c",          29, NULL,      NULL,        NULL},
287   // ... "]" in the password field isn't allowed, but we tolerate it here...
288 {"http://&a:foo(b]c@d:2/",              "http", "&a",  "foo(b]c", "d",           2, "/",       NULL,        NULL},
289 {"http://::@c@d:2",                     "http", "",    ":@c",     "d",           2, NULL,      NULL,        NULL},
290 {"http://foo.com:b@d/",                 "http", "foo.com", "b",   "d",          -1, "/",       NULL,        NULL},
291 
292 {"http://foo.com/\\@",                  "http", NULL,  NULL,      "foo.com",    -1, "/\\@",    NULL,        NULL},
293 {"http:\\\\foo.com\\",                  "http", NULL,  NULL,      "foo.com",    -1, "\\",      NULL,        NULL},
294 {"http:\\\\a\\b:c\\d@foo.com\\",        "http", NULL,  NULL,      "a",          -1, "\\b:c\\d@foo.com\\", NULL,   NULL},
295 
296   // Tolerate different numbers of slashes.
297 {"foo:/",                               "foo",  NULL,  NULL,      NULL,         -1, NULL,      NULL,        NULL},
298 {"foo:/bar.com/",                       "foo",  NULL,  NULL,      "bar.com",    -1, "/",       NULL,        NULL},
299 {"foo://///////",                       "foo",  NULL,  NULL,      NULL,         -1, NULL,      NULL,        NULL},
300 {"foo://///////bar.com/",               "foo",  NULL,  NULL,      "bar.com",    -1, "/",       NULL,        NULL},
301 {"foo:////://///",                      "foo",  NULL,  NULL,      NULL,         -1, "/////",   NULL,        NULL},
302 
303   // Raw file paths on Windows aren't handled by the parser.
304 {"c:/foo",                              "c",    NULL,  NULL,      "foo",        -1, NULL,      NULL,        NULL},
305 {"//foo/bar",                           NULL,   NULL,  NULL,      "foo",        -1, "/bar",    NULL,        NULL},
306 
307   // Use the first question mark for the query and the ref.
308 {"http://foo/path;a??e#f#g",            "http", NULL,  NULL,      "foo",        -1, "/path;a", "?e",      "f#g"},
309 {"http://foo/abcd?efgh?ijkl",           "http", NULL,  NULL,      "foo",        -1, "/abcd",   "efgh?ijkl", NULL},
310 {"http://foo/abcd#foo?bar",             "http", NULL,  NULL,      "foo",        -1, "/abcd",   NULL,        "foo?bar"},
311 
312   // IPv6, check also interesting uses of colons.
313 {"[61:24:74]:98",                       "[61",  NULL,  NULL,      "24:74]",     98, NULL,      NULL,        NULL},
314 {"http://[61:27]:98",                   "http", NULL,  NULL,      "[61:27]",    98, NULL,      NULL,        NULL},
315 {"http:[61:27]/:foo",                   "http", NULL,  NULL,      "[61:27]",    -1, "/:foo",   NULL,        NULL},
316 {"http://[1::2]:3:4",                   "http", NULL,  NULL,      "[1::2]:3",    4, NULL,      NULL,        NULL},
317 
318   // Partially-complete IPv6 literals, and related cases.
319 {"http://2001::1",                      "http", NULL,  NULL,      "2001:",       1, NULL,      NULL,        NULL},
320 {"http://[2001::1",                     "http", NULL,  NULL,      "[2001::1",   -1, NULL,      NULL,        NULL},
321 {"http://2001::1]",                     "http", NULL,  NULL,      "2001::1]",   -1, NULL,      NULL,        NULL},
322 {"http://2001::1]:80",                  "http", NULL,  NULL,      "2001::1]",   80, NULL,      NULL,        NULL},
323 {"http://[2001::1]",                    "http", NULL,  NULL,      "[2001::1]",  -1, NULL,      NULL,        NULL},
324 {"http://[2001::1]:80",                 "http", NULL,  NULL,      "[2001::1]",  80, NULL,      NULL,        NULL},
325 {"http://[[::]]",                       "http", NULL,  NULL,      "[[::]]",     -1, NULL,      NULL,        NULL},
326 
327 };
328 
TEST(URLParser,Standard)329 TEST(URLParser, Standard) {
330   // Declared outside for loop to try to catch cases in init() where we forget
331   // to reset something that is reset by the construtor.
332   url_parse::Parsed parsed;
333   for (size_t i = 0; i < arraysize(cases); i++) {
334     const char* url = cases[i].input;
335     url_parse::ParseStandardURL(url, static_cast<int>(strlen(url)), &parsed);
336     int port = url_parse::ParsePort(url, parsed.port);
337 
338     EXPECT_TRUE(ComponentMatches(url, cases[i].scheme, parsed.scheme));
339     EXPECT_TRUE(ComponentMatches(url, cases[i].username, parsed.username));
340     EXPECT_TRUE(ComponentMatches(url, cases[i].password, parsed.password));
341     EXPECT_TRUE(ComponentMatches(url, cases[i].host, parsed.host));
342     EXPECT_EQ(cases[i].port, port);
343     EXPECT_TRUE(ComponentMatches(url, cases[i].path, parsed.path));
344     EXPECT_TRUE(ComponentMatches(url, cases[i].query, parsed.query));
345     EXPECT_TRUE(ComponentMatches(url, cases[i].ref, parsed.ref));
346   }
347 }
348 
349 // PathURL --------------------------------------------------------------------
350 
351 // Various incarnations of path URLs.
352 static PathURLParseCase path_cases[] = {
353 {"",                                        NULL,          NULL},
354 {":",                                       "",            NULL},
355 {":/",                                      "",            "/"},
356 {"/",                                       NULL,          "/"},
357 {" This is \\interesting// \t",             NULL,          "This is \\interesting//"},
358 {"about:",                                  "about",       NULL},
359 {"about:blank",                             "about",       "blank"},
360 {"  about: blank ",                         "about",       " blank"},
361 {"javascript :alert(\"He:/l\\l#o?foo\"); ", "javascript ", "alert(\"He:/l\\l#o?foo\");"},
362 };
363 
TEST(URLParser,PathURL)364 TEST(URLParser, PathURL) {
365   // Declared outside for loop to try to catch cases in init() where we forget
366   // to reset something that is reset by the construtor.
367   url_parse::Parsed parsed;
368   for (size_t i = 0; i < arraysize(path_cases); i++) {
369     const char* url = path_cases[i].input;
370     url_parse::ParsePathURL(url, static_cast<int>(strlen(url)), &parsed);
371 
372     EXPECT_TRUE(ComponentMatches(url, path_cases[i].scheme, parsed.scheme));
373     EXPECT_TRUE(ComponentMatches(url, path_cases[i].path, parsed.path));
374 
375     // The remaining components are never used for path urls.
376     ExpectInvalidComponent(parsed.username);
377     ExpectInvalidComponent(parsed.password);
378     ExpectInvalidComponent(parsed.host);
379     ExpectInvalidComponent(parsed.port);
380     ExpectInvalidComponent(parsed.query);
381     ExpectInvalidComponent(parsed.ref);
382   }
383 }
384 
385 #ifdef WIN32
386 
387 // WindowsFile ----------------------------------------------------------------
388 
389 // Various incarnations of file URLs. These are for Windows only.
390 static URLParseCase file_cases[] = {
391 {"file:server",              "file", NULL, NULL, "server", -1, NULL,          NULL, NULL},
392 {"  file: server  \t",       "file", NULL, NULL, " server",-1, NULL,          NULL, NULL},
393 {"FiLe:c|",                  "FiLe", NULL, NULL, NULL,     -1, "c|",          NULL, NULL},
394 {"FILE:/\\\\/server/file",   "FILE", NULL, NULL, "server", -1, "/file",       NULL, NULL},
395 {"file://server/",           "file", NULL, NULL, "server", -1, "/",           NULL, NULL},
396 {"file://localhost/c:/",     "file", NULL, NULL, NULL,     -1, "/c:/",        NULL, NULL},
397 {"file://127.0.0.1/c|\\",    "file", NULL, NULL, NULL,     -1, "/c|\\",       NULL, NULL},
398 {"file:/",                   "file", NULL, NULL, NULL,     -1, NULL,          NULL, NULL},
399 {"file:",                    "file", NULL, NULL, NULL,     -1, NULL,          NULL, NULL},
400   // If there is a Windows drive letter, treat any number of slashes as the
401   // path part.
402 {"file:c:\\fo\\b",           "file", NULL, NULL, NULL,     -1, "c:\\fo\\b",   NULL, NULL},
403 {"file:/c:\\foo/bar",        "file", NULL, NULL, NULL,     -1, "/c:\\foo/bar",NULL, NULL},
404 {"file://c:/f\\b",           "file", NULL, NULL, NULL,     -1, "/c:/f\\b",    NULL, NULL},
405 {"file:///C:/foo",           "file", NULL, NULL, NULL,     -1, "/C:/foo",     NULL, NULL},
406 {"file://///\\/\\/c:\\f\\b", "file", NULL, NULL, NULL,     -1, "/c:\\f\\b",   NULL, NULL},
407   // If there is not a drive letter, we should treat is as UNC EXCEPT for
408   // three slashes, which we treat as a Unix style path.
409 {"file:server/file",         "file", NULL, NULL, "server", -1, "/file",       NULL, NULL},
410 {"file:/server/file",        "file", NULL, NULL, "server", -1, "/file",       NULL, NULL},
411 {"file://server/file",       "file", NULL, NULL, "server", -1, "/file",       NULL, NULL},
412 {"file:///server/file",      "file", NULL, NULL, NULL,     -1, "/server/file",NULL, NULL},
413 {"file://\\server/file",     "file", NULL, NULL, NULL,     -1, "\\server/file",NULL, NULL},
414 {"file:////server/file",     "file", NULL, NULL, "server", -1, "/file",       NULL, NULL},
415   // Queries and refs are valid for file URLs as well.
416 {"file:///C:/foo.html?#",   "file", NULL, NULL,  NULL,     -1, "/C:/foo.html",  "",   ""},
417 {"file:///C:/foo.html?query=yes#ref", "file", NULL, NULL, NULL, -1, "/C:/foo.html", "query=yes", "ref"},
418 };
419 
TEST(URLParser,WindowsFile)420 TEST(URLParser, WindowsFile) {
421   // Declared outside for loop to try to catch cases in init() where we forget
422   // to reset something that is reset by the construtor.
423   url_parse::Parsed parsed;
424   for (int i = 0; i < arraysize(file_cases); i++) {
425     const char* url = file_cases[i].input;
426     url_parse::ParseFileURL(url, static_cast<int>(strlen(url)), &parsed);
427     int port = url_parse::ParsePort(url, parsed.port);
428 
429     EXPECT_TRUE(ComponentMatches(url, file_cases[i].scheme, parsed.scheme));
430     EXPECT_TRUE(ComponentMatches(url, file_cases[i].username, parsed.username));
431     EXPECT_TRUE(ComponentMatches(url, file_cases[i].password, parsed.password));
432     EXPECT_TRUE(ComponentMatches(url, file_cases[i].host, parsed.host));
433     EXPECT_EQ(file_cases[i].port, port);
434     EXPECT_TRUE(ComponentMatches(url, file_cases[i].path, parsed.path));
435     EXPECT_TRUE(ComponentMatches(url, file_cases[i].query, parsed.query));
436     EXPECT_TRUE(ComponentMatches(url, file_cases[i].ref, parsed.ref));
437   }
438 }
439 
440 #endif  // WIN32
441 
TEST(URLParser,ExtractFileName)442 TEST(URLParser, ExtractFileName) {
443   struct FileCase {
444     const char* input;
445     const char* expected;
446   } file_cases[] = {
447     {"http://www.google.com", NULL},
448     {"http://www.google.com/", ""},
449     {"http://www.google.com/search", "search"},
450     {"http://www.google.com/search/", ""},
451     {"http://www.google.com/foo/bar.html?baz=22", "bar.html"},
452     {"http://www.google.com/foo/bar.html#ref", "bar.html"},
453     {"http://www.google.com/search/;param", ""},
454     {"http://www.google.com/foo/bar.html;param#ref", "bar.html"},
455     {"http://www.google.com/foo/bar.html;foo;param#ref", "bar.html;foo"},
456     {"http://www.google.com/foo/bar.html?query#ref", "bar.html"},
457   };
458 
459   for (size_t i = 0; i < ARRAYSIZE(file_cases); i++) {
460     const char* url = file_cases[i].input;
461     int len = static_cast<int>(strlen(url));
462 
463     url_parse::Parsed parsed;
464     url_parse::ParseStandardURL(url, len, &parsed);
465 
466     url_parse::Component file_name;
467     url_parse::ExtractFileName(url, parsed.path, &file_name);
468 
469     EXPECT_TRUE(ComponentMatches(url, file_cases[i].expected, file_name));
470   }
471 }
472 
473 // Returns true if the parameter with index |parameter| in the given URL's
474 // query string. The expected key can be NULL to indicate no such key index
475 // should exist. The parameter number is 1-based.
NthParameterIs(const char * url,int parameter,const char * expected_key,const char * expected_value)476 static bool NthParameterIs(const char* url,
477                            int parameter,
478                            const char* expected_key,
479                            const char* expected_value) {
480   url_parse::Parsed parsed;
481   url_parse::ParseStandardURL(url, static_cast<int>(strlen(url)), &parsed);
482 
483   url_parse::Component query = parsed.query;
484 
485   for (int i = 1; i <= parameter; i++) {
486     url_parse::Component key, value;
487     if (!url_parse::ExtractQueryKeyValue(url, &query, &key, &value)) {
488       if (parameter >= i && !expected_key)
489         return true;  // Expected nonexistant key, got one.
490       return false;  // Not enough keys.
491     }
492 
493     if (i == parameter) {
494       if (!expected_key)
495         return false;
496 
497       if (strncmp(&url[key.begin], expected_key, key.len) != 0)
498         return false;
499       if (strncmp(&url[value.begin], expected_value, value.len) != 0)
500         return false;
501       return true;
502     }
503   }
504   return expected_key == NULL;  // We didn't find that many parameters.
505 }
506 
TEST(URLParser,ExtractQueryKeyValue)507 TEST(URLParser, ExtractQueryKeyValue) {
508   EXPECT_TRUE(NthParameterIs("http://www.google.com", 1, NULL, NULL));
509 
510   // Basic case.
511   char a[] = "http://www.google.com?arg1=1&arg2=2&bar";
512   EXPECT_TRUE(NthParameterIs(a, 1, "arg1", "1"));
513   EXPECT_TRUE(NthParameterIs(a, 2, "arg2", "2"));
514   EXPECT_TRUE(NthParameterIs(a, 3, "bar", ""));
515   EXPECT_TRUE(NthParameterIs(a, 4, NULL, NULL));
516 
517   // Empty param at the end.
518   char b[] = "http://www.google.com?foo=bar&";
519   EXPECT_TRUE(NthParameterIs(b, 1, "foo", "bar"));
520   EXPECT_TRUE(NthParameterIs(b, 2, NULL, NULL));
521 
522   // Empty param at the beginning.
523   char c[] = "http://www.google.com?&foo=bar";
524   EXPECT_TRUE(NthParameterIs(c, 1, "", ""));
525   EXPECT_TRUE(NthParameterIs(c, 2, "foo", "bar"));
526   EXPECT_TRUE(NthParameterIs(c, 3, NULL, NULL));
527 
528   // Empty key with value.
529   char d[] = "http://www.google.com?=foo";
530   EXPECT_TRUE(NthParameterIs(d, 1, "", "foo"));
531   EXPECT_TRUE(NthParameterIs(d, 2, NULL, NULL));
532 
533   // Empty value with key.
534   char e[] = "http://www.google.com?foo=";
535   EXPECT_TRUE(NthParameterIs(e, 1, "foo", ""));
536   EXPECT_TRUE(NthParameterIs(e, 2, NULL, NULL));
537 
538   // Empty key and values.
539   char f[] = "http://www.google.com?&&==&=";
540   EXPECT_TRUE(NthParameterIs(f, 1, "", ""));
541   EXPECT_TRUE(NthParameterIs(f, 2, "", ""));
542   EXPECT_TRUE(NthParameterIs(f, 3, "", "="));
543   EXPECT_TRUE(NthParameterIs(f, 4, "", ""));
544   EXPECT_TRUE(NthParameterIs(f, 5, NULL, NULL));
545 }
546 
547 // MailtoURL --------------------------------------------------------------------
548 
549 static MailtoURLParseCase mailto_cases[] = {
550 //|input                       |scheme   |path               |query
551 {"mailto:foo@gmail.com",        "mailto", "foo@gmail.com",    NULL},
552 {"  mailto: to  \t",            "mailto", " to",              NULL},
553 {"mailto:addr1%2C%20addr2 ",    "mailto", "addr1%2C%20addr2", NULL},
554 {"Mailto:addr1, addr2 ",        "Mailto", "addr1, addr2",     NULL},
555 {"mailto:addr1:addr2 ",         "mailto", "addr1:addr2",      NULL},
556 {"mailto:?to=addr1,addr2",      "mailto", NULL,               "to=addr1,addr2"},
557 {"mailto:?to=addr1%2C%20addr2", "mailto", NULL,               "to=addr1%2C%20addr2"},
558 {"mailto:addr1?to=addr2",       "mailto", "addr1",            "to=addr2"},
559 {"mailto:?body=#foobar#",       "mailto", NULL,               "body=#foobar#",},
560 {"mailto:#?body=#foobar#",      "mailto", "#",                "body=#foobar#"},
561 };
562 
TEST(URLParser,MailtoUrl)563 TEST(URLParser, MailtoUrl) {
564   // Declared outside for loop to try to catch cases in init() where we forget
565   // to reset something that is reset by the construtor.
566   url_parse::Parsed parsed;
567   for (size_t i = 0; i < arraysize(mailto_cases); ++i) {
568     const char* url = mailto_cases[i].input;
569     url_parse::ParseMailtoURL(url, static_cast<int>(strlen(url)), &parsed);
570     int port = url_parse::ParsePort(url, parsed.port);
571 
572     EXPECT_TRUE(ComponentMatches(url, mailto_cases[i].scheme, parsed.scheme));
573     EXPECT_TRUE(ComponentMatches(url, mailto_cases[i].path, parsed.path));
574     EXPECT_TRUE(ComponentMatches(url, mailto_cases[i].query, parsed.query));
575     EXPECT_EQ(url_parse::PORT_UNSPECIFIED, port);
576 
577     // The remaining components are never used for mailto urls.
578     ExpectInvalidComponent(parsed.username);
579     ExpectInvalidComponent(parsed.password);
580     ExpectInvalidComponent(parsed.port);
581     ExpectInvalidComponent(parsed.ref);
582   }
583 }
584