1 // Copyright 2013 The Chromium Authors
2 // Use of this source code is governed by a BSD-style license that can be
3 // found in the LICENSE file.
7 #include "testing/gtest/include/gtest/gtest.h"
8 #include "url/third_party/mozilla/url_parse.h"
10 // Interesting IE file:isms...
12 // file:/foo/bar file:///foo/bar
13 // The result here seems totally invalid!?!? This isn't UNC.
16 // file:// or any other number of slashes
17 // IE6 doesn't do anything at all if you click on this link. No error:
18 // nothing. IE6's history system seems to always color this link, so I'm
19 // guessing that it maps internally to the empty URL.
23 // /foo file:///C:/foo
24 // Interestingly, IE treats "/" as an alias for "c:\", which makes sense,
25 // but is weird to think about on Windows.
27 // file:foo/ file:foo/ (invalid?!?!?)
28 // file:/foo/ file:///foo/ (invalid?!?!?)
29 // file://foo/ file://foo/ (UNC to server "foo")
30 // file:///foo/ file:///foo/ (invalid)
31 // file:////foo/ file://foo/ (UNC to server "foo")
32 // Any more than four slashes is also treated as UNC.
34 // file:C:/ file://C:/
35 // file:/C:/ file://C:/
36 // The number of slashes after "file:" don't matter if the thing following
37 // it looks like an absolute drive path. Also, slashes and backslashes are
38 // equally valid here.
43 // Used for regular URL parse cases.
57 // Simpler version of URLParseCase for testing path URLs.
58 struct PathURLParseCase {
65 // Simpler version of URLParseCase for testing mailto URLs.
66 struct MailtoURLParseCase {
74 // More complicated version of URLParseCase for testing filesystem URLs.
75 struct FileSystemURLParseCase {
78 const char* inner_scheme;
79 const char* inner_username;
80 const char* inner_password;
81 const char* inner_host;
83 const char* inner_path;
89 bool ComponentMatches(const char* input,
90 const char* reference,
91 const Component& component) {
92 // Check that the -1 sentinel is the only allowed negative value.
93 EXPECT_TRUE(component.is_valid() || component.len == -1);
95 // Begin should be valid.
96 EXPECT_LE(0, component.begin);
98 // A NULL reference means the component should be nonexistent.
100 return component.len == -1;
101 if (!component.is_valid())
102 return false; // Reference is not NULL but we don't have anything
104 if (strlen(reference) != static_cast<size_t>(component.len))
105 return false; // Lengths don't match
107 // Now check the actual characters.
108 return strncmp(reference, &input[component.begin], component.len) == 0;
111 void ExpectInvalidComponent(const Component& component) {
112 EXPECT_EQ(0, component.begin);
113 EXPECT_EQ(-1, component.len);
116 // Parsed ----------------------------------------------------------------------
118 TEST(URLParser, Length) {
119 const char* length_cases[] = {
120 // One with everything in it.
121 "http://user:pass@host:99/foo?bar#baz",
122 // One with nothing in it.
124 // Working backwards, let's start taking off stuff from the full one.
125 "http://user:pass@host:99/foo?bar#",
126 "http://user:pass@host:99/foo?bar",
127 "http://user:pass@host:99/foo?",
128 "http://user:pass@host:99/foo",
129 "http://user:pass@host:99/",
130 "http://user:pass@host:99",
131 "http://user:pass@host:",
132 "http://user:pass@host",
137 for (const char* length_case : length_cases) {
138 int true_length = static_cast<int>(strlen(length_case));
141 ParseStandardURL(length_case, true_length, &parsed);
143 EXPECT_EQ(true_length, parsed.Length());
147 TEST(URLParser, CountCharactersBefore) {
150 Parsed::ComponentType component;
151 bool include_delimiter;
154 // Test each possibility in the case where all components are present.
156 // 0123456789012345678901
157 {"http://u:p@h:8/p?q#r", Parsed::SCHEME, true, 0},
158 {"http://u:p@h:8/p?q#r", Parsed::SCHEME, false, 0},
159 {"http://u:p@h:8/p?q#r", Parsed::USERNAME, true, 7},
160 {"http://u:p@h:8/p?q#r", Parsed::USERNAME, false, 7},
161 {"http://u:p@h:8/p?q#r", Parsed::PASSWORD, true, 9},
162 {"http://u:p@h:8/p?q#r", Parsed::PASSWORD, false, 9},
163 {"http://u:p@h:8/p?q#r", Parsed::HOST, true, 11},
164 {"http://u:p@h:8/p?q#r", Parsed::HOST, false, 11},
165 {"http://u:p@h:8/p?q#r", Parsed::PORT, true, 12},
166 {"http://u:p@h:8/p?q#r", Parsed::PORT, false, 13},
167 {"http://u:p@h:8/p?q#r", Parsed::PATH, false, 14},
168 {"http://u:p@h:8/p?q#r", Parsed::PATH, true, 14},
169 {"http://u:p@h:8/p?q#r", Parsed::QUERY, true, 16},
170 {"http://u:p@h:8/p?q#r", Parsed::QUERY, false, 17},
171 {"http://u:p@h:8/p?q#r", Parsed::REF, true, 18},
172 {"http://u:p@h:8/p?q#r", Parsed::REF, false, 19},
173 // Now test when the requested component is missing.
174 {"http://u:p@h:8/p?", Parsed::REF, true, 17},
175 {"http://u:p@h:8/p?q", Parsed::REF, true, 18},
176 {"http://u:p@h:8/p#r", Parsed::QUERY, true, 16},
177 {"http://u:p@h:8#r", Parsed::PATH, true, 14},
178 {"http://u:p@h/", Parsed::PORT, true, 12},
179 {"http://u:p@/", Parsed::HOST, true, 11},
180 // This case is a little weird. It will report that the password would
181 // start where the host begins. This is arguably correct, although you
182 // could also argue that it should start at the '@' sign. Doing it
183 // starting with the '@' sign is actually harder, so we don't bother.
184 {"http://u@h/", Parsed::PASSWORD, true, 9},
185 {"http://h/", Parsed::USERNAME, true, 7},
186 {"http:", Parsed::USERNAME, true, 5},
187 {"", Parsed::SCHEME, true, 0},
188 // Make sure a random component still works when there's nothing there.
189 {"", Parsed::REF, true, 0},
190 // File URLs are special with no host, so we test those.
191 {"file:///c:/foo", Parsed::USERNAME, true, 7},
192 {"file:///c:/foo", Parsed::PASSWORD, true, 7},
193 {"file:///c:/foo", Parsed::HOST, true, 7},
194 {"file:///c:/foo", Parsed::PATH, true, 7},
196 for (const auto& count_case : count_cases) {
197 int length = static_cast<int>(strlen(count_case.url));
199 // Simple test to distinguish file and standard URLs.
201 if (length > 0 && count_case.url[0] == 'f') {
202 ParseFileURL(count_case.url, length, &parsed);
204 ParseStandardURL(count_case.url, length, &parsed);
207 int chars_before = parsed.CountCharactersBefore(
208 count_case.component, count_case.include_delimiter);
209 EXPECT_EQ(count_case.expected_count, chars_before);
213 // Standard --------------------------------------------------------------------
216 // Input Scheme Usrname Passwd Host Port Path Query Ref
217 // ------------------------------------ ------- -------- ---------- ------------ --- ---------- ------------ -----
218 static URLParseCase cases[] = {
219 // Regular URL with all the parts
220 {"http://user:pass@foo:21/bar;par?b#c", "http", "user", "pass", "foo", 21, "/bar;par","b", "c"},
222 // Known schemes should lean towards authority identification
223 {"http:foo.com", "http", nullptr, nullptr, "foo.com", -1, nullptr, nullptr, nullptr},
226 {"\t :foo.com \n", "", nullptr, nullptr, "foo.com", -1, nullptr, nullptr, nullptr},
227 {" foo.com ", nullptr,nullptr, nullptr, "foo.com", -1, nullptr, nullptr, nullptr},
228 {"a:\t foo.com", "a", nullptr, nullptr, "\t foo.com", -1, nullptr, nullptr, nullptr},
229 {"http://f:21/ b ? d # e ", "http", nullptr, nullptr, "f", 21, "/ b ", " d ", " e"},
231 // Invalid port numbers should be identified and turned into -2, empty port
232 // numbers should be -1. Spaces aren't allowed in port numbers
233 {"http://f:/c", "http", nullptr, nullptr, "f", -1, "/c", nullptr, nullptr},
234 {"http://f:0/c", "http", nullptr, nullptr, "f", 0, "/c", nullptr, nullptr},
235 {"http://f:00000000000000/c", "http", nullptr, nullptr, "f", 0, "/c", nullptr, nullptr},
236 {"http://f:00000000000000000000080/c", "http", nullptr, nullptr, "f", 80, "/c", nullptr, nullptr},
237 {"http://f:b/c", "http", nullptr, nullptr, "f", -2, "/c", nullptr, nullptr},
238 {"http://f: /c", "http", nullptr, nullptr, "f", -2, "/c", nullptr, nullptr},
239 {"http://f:\n/c", "http", nullptr, nullptr, "f", -2, "/c", nullptr, nullptr},
240 {"http://f:fifty-two/c", "http", nullptr, nullptr, "f", -2, "/c", nullptr, nullptr},
241 {"http://f:999999/c", "http", nullptr, nullptr, "f", -2, "/c", nullptr, nullptr},
242 {"http://f: 21 / b ? d # e ", "http", nullptr, nullptr, "f", -2, "/ b ", " d ", " e"},
244 // Creative URLs missing key elements
245 {"", nullptr,nullptr, nullptr, nullptr, -1, nullptr, nullptr, nullptr},
246 {" \t", nullptr,nullptr, nullptr, nullptr, -1, nullptr, nullptr, nullptr},
247 {":foo.com/", "", nullptr, nullptr, "foo.com", -1, "/", nullptr, nullptr},
248 {":foo.com\\", "", nullptr, nullptr, "foo.com", -1, "\\", nullptr, nullptr},
249 {":", "", nullptr, nullptr, nullptr, -1, nullptr, nullptr, nullptr},
250 {":a", "", nullptr, nullptr, "a", -1, nullptr, nullptr, nullptr},
251 {":/", "", nullptr, nullptr, nullptr, -1, nullptr, nullptr, nullptr},
252 {":\\", "", nullptr, nullptr, nullptr, -1, nullptr, nullptr, nullptr},
253 {":#", "", nullptr, nullptr, nullptr, -1, nullptr, nullptr, ""},
254 {"#", nullptr,nullptr, nullptr, nullptr, -1, nullptr, nullptr, ""},
255 {"#/", nullptr,nullptr, nullptr, nullptr, -1, nullptr, nullptr, "/"},
256 {"#\\", nullptr,nullptr, nullptr, nullptr, -1, nullptr, nullptr, "\\"},
257 {"#;?", nullptr,nullptr, nullptr, nullptr, -1, nullptr, nullptr, ";?"},
258 {"?", nullptr,nullptr, nullptr, nullptr, -1, nullptr, "", nullptr},
259 {"/", nullptr,nullptr, nullptr, nullptr, -1, nullptr, nullptr, nullptr},
260 {":23", "", nullptr, nullptr, "23", -1, nullptr, nullptr, nullptr},
261 {"/:23", "/", nullptr, nullptr, "23", -1, nullptr, nullptr, nullptr},
262 {"//", nullptr,nullptr, nullptr, nullptr, -1, nullptr, nullptr, nullptr},
263 {"::", "", nullptr, nullptr, nullptr, -1, nullptr, nullptr, nullptr},
264 {"::23", "", nullptr, nullptr, nullptr, 23, nullptr, nullptr, nullptr},
265 {"foo://", "foo", nullptr, nullptr, nullptr, -1, nullptr, nullptr, nullptr},
267 // Username/passwords and things that look like them
268 {"http://a:b@c:29/d", "http", "a", "b", "c", 29, "/d", nullptr, nullptr},
269 {"http::@c:29", "http", "", "", "c", 29, nullptr, nullptr, nullptr},
270 // ... "]" in the password field isn't allowed, but we tolerate it here...
271 {"http://&a:foo(b]c@d:2/", "http", "&a", "foo(b]c", "d", 2, "/", nullptr, nullptr},
272 {"http://::@c@d:2", "http", "", ":@c", "d", 2, nullptr, nullptr, nullptr},
273 {"http://foo.com:b@d/", "http", "foo.com","b", "d", -1, "/", nullptr, nullptr},
275 {"http://foo.com/\\@", "http", nullptr, nullptr, "foo.com", -1, "/\\@", nullptr, nullptr},
276 {"http:\\\\foo.com\\", "http", nullptr, nullptr, "foo.com", -1, "\\", nullptr, nullptr},
277 {"http:\\\\a\\b:c\\d@foo.com\\", "http", nullptr, nullptr, "a", -1, "\\b:c\\d@foo.com\\", nullptr,nullptr},
279 // Tolerate different numbers of slashes.
280 {"foo:/", "foo", nullptr, nullptr, nullptr, -1, nullptr, nullptr, nullptr},
281 {"foo:/bar.com/", "foo", nullptr, nullptr, "bar.com", -1, "/", nullptr, nullptr},
282 {"foo://///////", "foo", nullptr, nullptr, nullptr, -1, nullptr, nullptr, nullptr},
283 {"foo://///////bar.com/", "foo", nullptr, nullptr, "bar.com", -1, "/", nullptr, nullptr},
284 {"foo:////://///", "foo", nullptr, nullptr, nullptr, -1, "/////", nullptr, nullptr},
286 // Raw file paths on Windows aren't handled by the parser.
287 {"c:/foo", "c", nullptr, nullptr, "foo", -1, nullptr, nullptr, nullptr},
288 {"//foo/bar", nullptr,nullptr, nullptr, "foo", -1, "/bar", nullptr, nullptr},
290 // Use the first question mark for the query and the ref.
291 {"http://foo/path;a??e#f#g", "http", nullptr, nullptr, "foo", -1, "/path;a", "?e", "f#g"},
292 {"http://foo/abcd?efgh?ijkl", "http", nullptr, nullptr, "foo", -1, "/abcd", "efgh?ijkl", nullptr},
293 {"http://foo/abcd#foo?bar", "http", nullptr, nullptr, "foo", -1, "/abcd", nullptr, "foo?bar"},
295 // IPv6, check also interesting uses of colons.
296 {"[61:24:74]:98", "[61", nullptr, nullptr, "24:74]", 98, nullptr, nullptr, nullptr},
297 {"http://[61:27]:98", "http", nullptr, nullptr, "[61:27]", 98, nullptr, nullptr, nullptr},
298 {"http:[61:27]/:foo", "http", nullptr, nullptr, "[61:27]", -1, "/:foo", nullptr, nullptr},
299 {"http://[1::2]:3:4", "http", nullptr, nullptr, "[1::2]:3", 4, nullptr, nullptr, nullptr},
301 // Partially-complete IPv6 literals, and related cases.
302 {"http://2001::1", "http", nullptr, nullptr, "2001:", 1, nullptr, nullptr, nullptr},
303 {"http://[2001::1", "http", nullptr, nullptr, "[2001::1", -1, nullptr, nullptr, nullptr},
304 {"http://2001::1]", "http", nullptr, nullptr, "2001::1]", -1, nullptr, nullptr, nullptr},
305 {"http://2001::1]:80", "http", nullptr, nullptr, "2001::1]", 80, nullptr, nullptr, nullptr},
306 {"http://[2001::1]", "http", nullptr, nullptr, "[2001::1]", -1, nullptr, nullptr, nullptr},
307 {"http://[2001::1]:80", "http", nullptr, nullptr, "[2001::1]", 80, nullptr, nullptr, nullptr},
308 {"http://[[::]]", "http", nullptr, nullptr, "[[::]]", -1, nullptr, nullptr, nullptr},
313 TEST(URLParser, Standard) {
314 // Declared outside for loop to try to catch cases in init() where we forget
315 // to reset something that is reset by the constructor.
317 for (const auto& i : cases) {
318 const char* url = i.input;
319 ParseStandardURL(url, static_cast<int>(strlen(url)), &parsed);
320 int port = ParsePort(url, parsed.port);
322 EXPECT_TRUE(ComponentMatches(url, i.scheme, parsed.scheme));
323 EXPECT_TRUE(ComponentMatches(url, i.username, parsed.username));
324 EXPECT_TRUE(ComponentMatches(url, i.password, parsed.password));
325 EXPECT_TRUE(ComponentMatches(url, i.host, parsed.host));
326 EXPECT_EQ(i.port, port);
327 EXPECT_TRUE(ComponentMatches(url, i.path, parsed.path));
328 EXPECT_TRUE(ComponentMatches(url, i.query, parsed.query));
329 EXPECT_TRUE(ComponentMatches(url, i.ref, parsed.ref));
333 // PathURL --------------------------------------------------------------------
335 // Various incarnations of path URLs.
337 static PathURLParseCase path_cases[] = {
338 {"", nullptr, nullptr},
342 {" This is \\interesting// \t", nullptr, "This is \\interesting// \t"},
343 {"about:", "about", nullptr},
344 {"about:blank", "about", "blank"},
345 {" about: blank ", "about", " blank "},
346 {"javascript :alert(\"He:/l\\l#o?foo\"); ", "javascript ", "alert(\"He:/l\\l#o?foo\"); "},
350 TEST(URLParser, PathURL) {
351 // Declared outside for loop to try to catch cases in init() where we forget
352 // to reset something that is reset by the constructor.
354 for (size_t i = 0; i < std::size(path_cases); i++) {
355 const char* url = path_cases[i].input;
356 ParsePathURL(url, static_cast<int>(strlen(url)), false, &parsed);
358 EXPECT_TRUE(ComponentMatches(url, path_cases[i].scheme, parsed.scheme))
360 EXPECT_TRUE(ComponentMatches(url, path_cases[i].path, parsed.GetContent()))
363 // The remaining components are never used for path URLs.
364 ExpectInvalidComponent(parsed.username);
365 ExpectInvalidComponent(parsed.password);
366 ExpectInvalidComponent(parsed.host);
367 ExpectInvalidComponent(parsed.port);
371 // Various incarnations of file URLs.
373 static URLParseCase file_cases[] = {
375 {"file:server", "file", nullptr, nullptr, "server", -1, nullptr, nullptr, nullptr},
376 {" file: server \t", "file", nullptr, nullptr, " server",-1, nullptr, nullptr, nullptr},
377 {"FiLe:c|", "FiLe", nullptr, nullptr, nullptr, -1, "c|", nullptr, nullptr},
378 {"FILE:/\\\\/server/file", "FILE", nullptr, nullptr, "server", -1, "/file", nullptr, nullptr},
379 {"file://server/", "file", nullptr, nullptr, "server", -1, "/", nullptr, nullptr},
380 {"file://localhost/c:/", "file", nullptr, nullptr, "localhost", -1, "/c:/", nullptr, nullptr},
381 {"file://127.0.0.1/c|\\", "file", nullptr, nullptr, "127.0.0.1", -1, "/c|\\", nullptr, nullptr},
382 {"file:/", "file", nullptr, nullptr, nullptr, -1, nullptr, nullptr, nullptr},
383 {"file:", "file", nullptr, nullptr, nullptr, -1, nullptr, nullptr, nullptr},
384 // If there is a Windows drive letter, treat any number of slashes as the
386 {"file:c:\\fo\\b", "file", nullptr, nullptr, nullptr, -1, "c:\\fo\\b", nullptr, nullptr},
387 {"file:/c:\\foo/bar", "file", nullptr, nullptr, nullptr, -1, "/c:\\foo/bar",nullptr, nullptr},
388 {"file://c:/f\\b", "file", nullptr, nullptr, nullptr, -1, "/c:/f\\b", nullptr, nullptr},
389 {"file:///C:/foo", "file", nullptr, nullptr, nullptr, -1, "/C:/foo", nullptr, nullptr},
390 {"file://///\\/\\/c:\\f\\b", "file", nullptr, nullptr, nullptr, -1, "/c:\\f\\b", nullptr, nullptr},
391 // If there is not a drive letter, we should treat is as UNC EXCEPT for
392 // three slashes, which we treat as a Unix style path.
393 {"file:server/file", "file", nullptr, nullptr, "server", -1, "/file", nullptr, nullptr},
394 {"file:/server/file", "file", nullptr, nullptr, "server", -1, "/file", nullptr, nullptr},
395 {"file://server/file", "file", nullptr, nullptr, "server", -1, "/file", nullptr, nullptr},
396 {"file:///server/file", "file", nullptr, nullptr, nullptr, -1, "/server/file",nullptr, nullptr},
397 {"file://\\server/file", "file", nullptr, nullptr, nullptr, -1, "\\server/file",nullptr, nullptr},
398 {"file:////server/file", "file", nullptr, nullptr, "server", -1, "/file", nullptr, nullptr},
399 // Queries and refs are valid for file URLs as well.
400 {"file:///C:/foo.html?#", "file", nullptr, nullptr, nullptr, -1, "/C:/foo.html", "", ""},
401 {"file:///C:/foo.html?query=yes#ref", "file", nullptr, nullptr, nullptr, -1, "/C:/foo.html", "query=yes", "ref"},
404 {"file:", "file", nullptr, nullptr, nullptr, -1, nullptr, nullptr, nullptr},
405 {"file:path", "file", nullptr, nullptr, nullptr, -1, "path", nullptr, nullptr},
406 {"file:path/", "file", nullptr, nullptr, nullptr, -1, "path/", nullptr, nullptr},
407 {"file:path/f.txt", "file", nullptr, nullptr, nullptr, -1, "path/f.txt", nullptr, nullptr},
409 {"file:/", "file", nullptr, nullptr, nullptr, -1, "/", nullptr, nullptr},
410 {"file:/path", "file", nullptr, nullptr, nullptr, -1, "/path", nullptr, nullptr},
411 {"file:/path/", "file", nullptr, nullptr, nullptr, -1, "/path/", nullptr, nullptr},
412 {"file:/path/f.txt", "file", nullptr, nullptr, nullptr, -1, "/path/f.txt", nullptr, nullptr},
414 {"file://", "file", nullptr, nullptr, nullptr, -1, nullptr, nullptr, nullptr},
415 {"file://server", "file", nullptr, nullptr, "server", -1, nullptr, nullptr, nullptr},
416 {"file://server/", "file", nullptr, nullptr, "server", -1, "/", nullptr, nullptr},
417 {"file://server/f.txt", "file", nullptr, nullptr, "server", -1, "/f.txt", nullptr, nullptr},
419 {"file:///", "file", nullptr, nullptr, nullptr, -1, "/", nullptr, nullptr},
420 {"file:///path", "file", nullptr, nullptr, nullptr, -1, "/path", nullptr, nullptr},
421 {"file:///path/", "file", nullptr, nullptr, nullptr, -1, "/path/", nullptr, nullptr},
422 {"file:///path/f.txt", "file", nullptr, nullptr, nullptr, -1, "/path/f.txt", nullptr, nullptr},
423 // More than three slashes.
424 {"file:////", "file", nullptr, nullptr, nullptr, -1, "/", nullptr, nullptr},
425 {"file:////path", "file", nullptr, nullptr, nullptr, -1, "/path", nullptr, nullptr},
426 {"file:////path/", "file", nullptr, nullptr, nullptr, -1, "/path/", nullptr, nullptr},
427 {"file:////path/f.txt", "file", nullptr, nullptr, nullptr, -1, "/path/f.txt", nullptr, nullptr},
429 {"path/f.txt", nullptr,nullptr, nullptr, nullptr, -1, "path/f.txt", nullptr, nullptr},
430 {"path:80/f.txt", "path", nullptr, nullptr, nullptr, -1, "80/f.txt", nullptr, nullptr},
431 {"path/f.txt:80", "path/f.txt",nullptr, nullptr, nullptr,-1,"80", nullptr, nullptr}, // Wrong.
432 {"/path/f.txt", nullptr,nullptr, nullptr, nullptr, -1, "/path/f.txt", nullptr, nullptr},
433 {"/path:80/f.txt", nullptr,nullptr, nullptr, nullptr, -1, "/path:80/f.txt",nullptr, nullptr},
434 {"/path/f.txt:80", nullptr,nullptr, nullptr, nullptr, -1, "/path/f.txt:80",nullptr, nullptr},
435 {"//server/f.txt", nullptr,nullptr, nullptr, "server", -1, "/f.txt", nullptr, nullptr},
436 {"//server:80/f.txt", nullptr,nullptr, nullptr, "server:80",-1, "/f.txt", nullptr, nullptr},
437 {"//server/f.txt:80", nullptr,nullptr, nullptr, "server", -1, "/f.txt:80", nullptr, nullptr},
438 {"///path/f.txt", nullptr,nullptr, nullptr, nullptr, -1, "/path/f.txt", nullptr, nullptr},
439 {"///path:80/f.txt", nullptr,nullptr, nullptr, nullptr, -1, "/path:80/f.txt",nullptr, nullptr},
440 {"///path/f.txt:80", nullptr,nullptr, nullptr, nullptr, -1, "/path/f.txt:80",nullptr, nullptr},
441 {"////path/f.txt", nullptr,nullptr, nullptr, nullptr, -1, "/path/f.txt", nullptr, nullptr},
442 {"////path:80/f.txt", nullptr,nullptr, nullptr, nullptr, -1, "/path:80/f.txt",nullptr, nullptr},
443 {"////path/f.txt:80", nullptr,nullptr, nullptr, nullptr, -1, "/path/f.txt:80",nullptr, nullptr},
444 // Queries and refs are valid for file URLs as well.
445 {"file:///foo.html?#", "file", nullptr, nullptr, nullptr, -1, "/foo.html", "", ""},
446 {"file:///foo.html?q=y#ref", "file", nullptr, nullptr, nullptr, -1, "/foo.html", "q=y", "ref"},
451 TEST(URLParser, ParseFileURL) {
452 // Declared outside for loop to try to catch cases in init() where we forget
453 // to reset something that is reset by the construtor.
455 for (size_t i = 0; i < std::size(file_cases); i++) {
456 const char* url = file_cases[i].input;
457 ParseFileURL(url, static_cast<int>(strlen(url)), &parsed);
458 int port = ParsePort(url, parsed.port);
460 EXPECT_TRUE(ComponentMatches(url, file_cases[i].scheme, parsed.scheme))
461 << " for case #" << i << " [" << url << "] "
462 << parsed.scheme.begin << ", " << parsed.scheme.len;
464 EXPECT_TRUE(ComponentMatches(url, file_cases[i].username, parsed.username))
465 << " for case #" << i << " [" << url << "] "
466 << parsed.username.begin << ", " << parsed.username.len;
468 EXPECT_TRUE(ComponentMatches(url, file_cases[i].password, parsed.password))
469 << " for case #" << i << " [" << url << "] "
470 << parsed.password.begin << ", " << parsed.password.len;
472 EXPECT_TRUE(ComponentMatches(url, file_cases[i].host, parsed.host))
473 << " for case #" << i << " [" << url << "] "
474 << parsed.host.begin << ", " << parsed.host.len;
476 EXPECT_EQ(file_cases[i].port, port)
477 << " for case #" << i << " [ " << url << "] " << port;
479 EXPECT_TRUE(ComponentMatches(url, file_cases[i].path, parsed.path))
480 << " for case #" << i << " [" << url << "] "
481 << parsed.path.begin << ", " << parsed.path.len;
483 EXPECT_TRUE(ComponentMatches(url, file_cases[i].query, parsed.query))
484 << " for case #" << i << " [" << url << "] "
485 << parsed.query.begin << ", " << parsed.query.len;
487 EXPECT_TRUE(ComponentMatches(url, file_cases[i].ref, parsed.ref))
488 << " for case #" << i << " [ "<< url << "] "
489 << parsed.query.begin << ", " << parsed.scheme.len;
494 TEST(URLParser, ExtractFileName) {
497 const char* expected;
498 } extract_cases[] = {
499 {"http://www.google.com", nullptr},
500 {"http://www.google.com/", ""},
501 {"http://www.google.com/search", "search"},
502 {"http://www.google.com/search/", ""},
503 {"http://www.google.com/foo/bar.html?baz=22", "bar.html"},
504 {"http://www.google.com/foo/bar.html#ref", "bar.html"},
505 {"http://www.google.com/search/;param", ""},
506 {"http://www.google.com/foo/bar.html;param#ref", "bar.html"},
507 {"http://www.google.com/foo/bar.html;foo;param#ref", "bar.html"},
508 {"http://www.google.com/foo/bar.html?query#ref", "bar.html"},
509 {"http://www.google.com/foo;/bar.html", "bar.html"},
510 {"http://www.google.com/foo;/", ""},
511 {"http://www.google.com/foo;", "foo"},
512 {"http://www.google.com/;", ""},
513 {"http://www.google.com/foo;bar;html", "foo"},
516 for (const auto& extract_case : extract_cases) {
517 const char* url = extract_case.input;
518 int len = static_cast<int>(strlen(url));
521 ParseStandardURL(url, len, &parsed);
524 ExtractFileName(url, parsed.path, &file_name);
526 EXPECT_TRUE(ComponentMatches(url, extract_case.expected, file_name));
530 // Returns true if the parameter with index |parameter| in the given URL's
531 // query string. The expected key can be NULL to indicate no such key index
532 // should exist. The parameter number is 1-based.
533 static bool NthParameterIs(const char* url,
535 const char* expected_key,
536 const char* expected_value) {
538 ParseStandardURL(url, static_cast<int>(strlen(url)), &parsed);
540 Component query = parsed.query;
542 for (int i = 1; i <= parameter; i++) {
543 Component key, value;
544 if (!ExtractQueryKeyValue(url, &query, &key, &value)) {
545 if (parameter >= i && !expected_key)
546 return true; // Expected nonexistent key, got one.
547 return false; // Not enough keys.
550 if (i == parameter) {
554 if (strncmp(&url[key.begin], expected_key, key.len) != 0)
556 if (strncmp(&url[value.begin], expected_value, value.len) != 0)
561 return expected_key == nullptr; // We didn't find that many parameters.
564 TEST(URLParser, ExtractQueryKeyValue) {
565 EXPECT_TRUE(NthParameterIs("http://www.google.com", 1, nullptr, nullptr));
568 char a[] = "http://www.google.com?arg1=1&arg2=2&bar";
569 EXPECT_TRUE(NthParameterIs(a, 1, "arg1", "1"));
570 EXPECT_TRUE(NthParameterIs(a, 2, "arg2", "2"));
571 EXPECT_TRUE(NthParameterIs(a, 3, "bar", ""));
572 EXPECT_TRUE(NthParameterIs(a, 4, nullptr, nullptr));
574 // Empty param at the end.
575 char b[] = "http://www.google.com?foo=bar&";
576 EXPECT_TRUE(NthParameterIs(b, 1, "foo", "bar"));
577 EXPECT_TRUE(NthParameterIs(b, 2, nullptr, nullptr));
579 // Empty param at the beginning.
580 char c[] = "http://www.google.com?&foo=bar";
581 EXPECT_TRUE(NthParameterIs(c, 1, "", ""));
582 EXPECT_TRUE(NthParameterIs(c, 2, "foo", "bar"));
583 EXPECT_TRUE(NthParameterIs(c, 3, nullptr, nullptr));
585 // Empty key with value.
586 char d[] = "http://www.google.com?=foo";
587 EXPECT_TRUE(NthParameterIs(d, 1, "", "foo"));
588 EXPECT_TRUE(NthParameterIs(d, 2, nullptr, nullptr));
590 // Empty value with key.
591 char e[] = "http://www.google.com?foo=";
592 EXPECT_TRUE(NthParameterIs(e, 1, "foo", ""));
593 EXPECT_TRUE(NthParameterIs(e, 2, nullptr, nullptr));
595 // Empty key and values.
596 char f[] = "http://www.google.com?&&==&=";
597 EXPECT_TRUE(NthParameterIs(f, 1, "", ""));
598 EXPECT_TRUE(NthParameterIs(f, 2, "", ""));
599 EXPECT_TRUE(NthParameterIs(f, 3, "", "="));
600 EXPECT_TRUE(NthParameterIs(f, 4, "", ""));
601 EXPECT_TRUE(NthParameterIs(f, 5, nullptr, nullptr));
604 // MailtoURL --------------------------------------------------------------------
607 static MailtoURLParseCase mailto_cases[] = {
608 //|input |scheme |path |query
609 {"mailto:foo@gmail.com", "mailto", "foo@gmail.com", nullptr},
610 {" mailto: to \t", "mailto", " to", nullptr},
611 {"mailto:addr1%2C%20addr2 ", "mailto", "addr1%2C%20addr2", nullptr},
612 {"Mailto:addr1, addr2 ", "Mailto", "addr1, addr2", nullptr},
613 {"mailto:addr1:addr2 ", "mailto", "addr1:addr2", nullptr},
614 {"mailto:?to=addr1,addr2", "mailto", nullptr, "to=addr1,addr2"},
615 {"mailto:?to=addr1%2C%20addr2", "mailto", nullptr, "to=addr1%2C%20addr2"},
616 {"mailto:addr1?to=addr2", "mailto", "addr1", "to=addr2"},
617 {"mailto:?body=#foobar#", "mailto", nullptr, "body=#foobar#",},
618 {"mailto:#?body=#foobar#", "mailto", "#", "body=#foobar#"},
622 TEST(URLParser, MailtoUrl) {
623 // Declared outside for loop to try to catch cases in init() where we forget
624 // to reset something that is reset by the constructor.
626 for (const auto& mailto_case : mailto_cases) {
627 const char* url = mailto_case.input;
628 ParseMailtoURL(url, static_cast<int>(strlen(url)), &parsed);
629 int port = ParsePort(url, parsed.port);
631 EXPECT_TRUE(ComponentMatches(url, mailto_case.scheme, parsed.scheme));
632 EXPECT_TRUE(ComponentMatches(url, mailto_case.path, parsed.path));
633 EXPECT_TRUE(ComponentMatches(url, mailto_case.query, parsed.query));
634 EXPECT_EQ(PORT_UNSPECIFIED, port);
636 // The remaining components are never used for mailto URLs.
637 ExpectInvalidComponent(parsed.username);
638 ExpectInvalidComponent(parsed.password);
639 ExpectInvalidComponent(parsed.port);
640 ExpectInvalidComponent(parsed.ref);
644 // Various incarnations of filesystem URLs.
645 static FileSystemURLParseCase filesystem_cases[] = {
646 // Regular URL with all the parts
647 {"filesystem:http://user:pass@foo:21/temporary/bar;par?b#c", "http", "user",
648 "pass", "foo", 21, "/temporary", "/bar;par", "b", "c"},
649 {"filesystem:https://foo/persistent/bar;par/", "https", nullptr, nullptr,
650 "foo", -1, "/persistent", "/bar;par/", nullptr, nullptr},
651 {"filesystem:file:///persistent/bar;par/", "file", nullptr, nullptr,
652 nullptr, -1, "/persistent", "/bar;par/", nullptr, nullptr},
653 {"filesystem:file:///persistent/bar;par/?query#ref", "file", nullptr,
654 nullptr, nullptr, -1, "/persistent", "/bar;par/", "query", "ref"},
655 {"filesystem:file:///persistent", "file", nullptr, nullptr, nullptr, -1,
656 "/persistent", "", nullptr, nullptr},
659 TEST(URLParser, FileSystemURL) {
660 // Declared outside for loop to try to catch cases in init() where we forget
661 // to reset something that is reset by the constructor.
663 for (const auto& filesystem_case : filesystem_cases) {
664 const char* url = filesystem_case.input;
665 ParseFileSystemURL(url, static_cast<int>(strlen(url)), &parsed);
667 EXPECT_TRUE(ComponentMatches(url, "filesystem", parsed.scheme));
668 EXPECT_EQ(!filesystem_case.inner_scheme, !parsed.inner_parsed());
669 // Only check the inner_parsed if there is one.
670 if (parsed.inner_parsed()) {
671 EXPECT_TRUE(ComponentMatches(url, filesystem_case.inner_scheme,
672 parsed.inner_parsed()->scheme));
673 EXPECT_TRUE(ComponentMatches(url, filesystem_case.inner_username,
674 parsed.inner_parsed()->username));
675 EXPECT_TRUE(ComponentMatches(url, filesystem_case.inner_password,
676 parsed.inner_parsed()->password));
677 EXPECT_TRUE(ComponentMatches(url, filesystem_case.inner_host,
678 parsed.inner_parsed()->host));
679 int port = ParsePort(url, parsed.inner_parsed()->port);
680 EXPECT_EQ(filesystem_case.inner_port, port);
682 // The remaining components are never used for filesystem URLs.
683 ExpectInvalidComponent(parsed.inner_parsed()->query);
684 ExpectInvalidComponent(parsed.inner_parsed()->ref);
687 EXPECT_TRUE(ComponentMatches(url, filesystem_case.path, parsed.path));
688 EXPECT_TRUE(ComponentMatches(url, filesystem_case.query, parsed.query));
689 EXPECT_TRUE(ComponentMatches(url, filesystem_case.ref, parsed.ref));
691 // The remaining components are never used for filesystem URLs.
692 ExpectInvalidComponent(parsed.username);
693 ExpectInvalidComponent(parsed.password);
694 ExpectInvalidComponent(parsed.host);
695 ExpectInvalidComponent(parsed.port);