OLD | NEW |
1 // Copyright 2007, Google Inc. | 1 // Copyright 2007, Google Inc. |
2 // All rights reserved. | 2 // All rights reserved. |
3 // | 3 // |
4 // Redistribution and use in source and binary forms, with or without | 4 // Redistribution and use in source and binary forms, with or without |
5 // modification, are permitted provided that the following conditions are | 5 // modification, are permitted provided that the following conditions are |
6 // met: | 6 // met: |
7 // | 7 // |
8 // * Redistributions of source code must retain the above copyright | 8 // * Redistributions of source code must retain the above copyright |
9 // notice, this list of conditions and the following disclaimer. | 9 // notice, this list of conditions and the following disclaimer. |
10 // * Redistributions in binary form must reproduce the above | 10 // * Redistributions in binary form must reproduce the above |
(...skipping 12 matching lines...) Expand all Loading... |
23 // SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT | 23 // SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT |
24 // LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, | 24 // LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, |
25 // DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY | 25 // DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY |
26 // THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT | 26 // THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT |
27 // (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE | 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. | 28 // OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. |
29 | 29 |
30 #ifndef GOOGLEURL_SRC_GURL_H__ | 30 #ifndef GOOGLEURL_SRC_GURL_H__ |
31 #define GOOGLEURL_SRC_GURL_H__ | 31 #define GOOGLEURL_SRC_GURL_H__ |
32 | 32 |
33 #include <iosfwd> | 33 #include "url/gurl.h" |
34 #include <string> | |
35 | |
36 #include "base/string16.h" | |
37 #include "googleurl/src/url_canon.h" | |
38 #include "googleurl/src/url_canon_stdstring.h" | |
39 #include "googleurl/src/url_common.h" | |
40 #include "googleurl/src/url_parse.h" | |
41 | |
42 class GURL { | |
43 public: | |
44 typedef url_canon::StdStringReplacements<std::string> Replacements; | |
45 typedef url_canon::StdStringReplacements<string16> ReplacementsW; | |
46 | |
47 // Creates an empty, invalid URL. | |
48 GURL_API GURL(); | |
49 | |
50 // Copy construction is relatively inexpensive, with most of the time going | |
51 // to reallocating the string. It does not re-parse. | |
52 GURL_API GURL(const GURL& other); | |
53 | |
54 // The narrow version requires the input be UTF-8. Invalid UTF-8 input will | |
55 // result in an invalid URL. | |
56 // | |
57 // The wide version should also take an encoding parameter so we know how to | |
58 // encode the query parameters. It is probably sufficient for the narrow | |
59 // version to assume the query parameter encoding should be the same as the | |
60 // input encoding. | |
61 GURL_API explicit GURL(const std::string& url_string | |
62 /*, output_param_encoding*/); | |
63 GURL_API explicit GURL(const string16& url_string | |
64 /*, output_param_encoding*/); | |
65 | |
66 // Constructor for URLs that have already been parsed and canonicalized. This | |
67 // is used for conversions from KURL, for example. The caller must supply all | |
68 // information associated with the URL, which must be correct and consistent. | |
69 GURL_API GURL(const char* canonical_spec, size_t canonical_spec_len, | |
70 const url_parse::Parsed& parsed, bool is_valid); | |
71 | |
72 GURL_API ~GURL(); | |
73 | |
74 GURL_API GURL& operator=(const GURL& other); | |
75 | |
76 // Returns true when this object represents a valid parsed URL. When not | |
77 // valid, other functions will still succeed, but you will not get canonical | |
78 // data out in the format you may be expecting. Instead, we keep something | |
79 // "reasonable looking" so that the user can see how it's busted if | |
80 // displayed to them. | |
81 bool is_valid() const { | |
82 return is_valid_; | |
83 } | |
84 | |
85 // Returns true if the URL is zero-length. Note that empty URLs are also | |
86 // invalid, and is_valid() will return false for them. This is provided | |
87 // because some users may want to treat the empty case differently. | |
88 bool is_empty() const { | |
89 return spec_.empty(); | |
90 } | |
91 | |
92 // Returns the raw spec, i.e., the full text of the URL, in canonical UTF-8, | |
93 // if the URL is valid. If the URL is not valid, this will assert and return | |
94 // the empty string (for safety in release builds, to keep them from being | |
95 // misused which might be a security problem). | |
96 // | |
97 // The URL will be ASCII except the reference fragment, which may be UTF-8. | |
98 // It is guaranteed to be valid UTF-8. | |
99 // | |
100 // The exception is for empty() URLs (which are !is_valid()) but this will | |
101 // return the empty string without asserting. | |
102 // | |
103 // Used invalid_spec() below to get the unusable spec of an invalid URL. This | |
104 // separation is designed to prevent errors that may cause security problems | |
105 // that could result from the mistaken use of an invalid URL. | |
106 GURL_API const std::string& spec() const; | |
107 | |
108 // Returns the potentially invalid spec for a the URL. This spec MUST NOT be | |
109 // modified or sent over the network. It is designed to be displayed in error | |
110 // messages to the user, as the apperance of the spec may explain the error. | |
111 // If the spec is valid, the valid spec will be returned. | |
112 // | |
113 // The returned string is guaranteed to be valid UTF-8. | |
114 const std::string& possibly_invalid_spec() const { | |
115 return spec_; | |
116 } | |
117 | |
118 // Getter for the raw parsed structure. This allows callers to locate parts | |
119 // of the URL within the spec themselves. Most callers should consider using | |
120 // the individual component getters below. | |
121 // | |
122 // The returned parsed structure will reference into the raw spec, which may | |
123 // or may not be valid. If you are using this to index into the spec, BE | |
124 // SURE YOU ARE USING possibly_invalid_spec() to get the spec, and that you | |
125 // don't do anything "important" with invalid specs. | |
126 const url_parse::Parsed& parsed_for_possibly_invalid_spec() const { | |
127 return parsed_; | |
128 } | |
129 | |
130 // Defiant equality operator! | |
131 bool operator==(const GURL& other) const { | |
132 return spec_ == other.spec_; | |
133 } | |
134 bool operator!=(const GURL& other) const { | |
135 return spec_ != other.spec_; | |
136 } | |
137 | |
138 // Allows GURL to used as a key in STL (for example, a std::set or std::map). | |
139 bool operator<(const GURL& other) const { | |
140 return spec_ < other.spec_; | |
141 } | |
142 | |
143 // Resolves a URL that's possibly relative to this object's URL, and returns | |
144 // it. Absolute URLs are also handled according to the rules of URLs on web | |
145 // pages. | |
146 // | |
147 // It may be impossible to resolve the URLs properly. If the input is not | |
148 // "standard" (SchemeIsStandard() == false) and the input looks relative, we | |
149 // can't resolve it. In these cases, the result will be an empty, invalid | |
150 // GURL. | |
151 // | |
152 // The result may also be a nonempty, invalid URL if the input has some kind | |
153 // of encoding error. In these cases, we will try to construct a "good" URL | |
154 // that may have meaning to the user, but it will be marked invalid. | |
155 // | |
156 // It is an error to resolve a URL relative to an invalid URL. The result | |
157 // will be the empty URL. | |
158 GURL_API GURL Resolve(const std::string& relative) const; | |
159 GURL_API GURL Resolve(const string16& relative) const; | |
160 | |
161 // Like Resolve() above but takes a character set encoder which will be used | |
162 // for any query text specified in the input. The charset converter parameter | |
163 // may be NULL, in which case it will be treated as UTF-8. | |
164 // | |
165 // TODO(brettw): These should be replaced with versions that take something | |
166 // more friendly than a raw CharsetConverter (maybe like an ICU character set | |
167 // name). | |
168 GURL_API GURL ResolveWithCharsetConverter( | |
169 const std::string& relative, | |
170 url_canon::CharsetConverter* charset_converter) const; | |
171 GURL_API GURL ResolveWithCharsetConverter( | |
172 const string16& relative, | |
173 url_canon::CharsetConverter* charset_converter) const; | |
174 | |
175 // Creates a new GURL by replacing the current URL's components with the | |
176 // supplied versions. See the Replacements class in url_canon.h for more. | |
177 // | |
178 // These are not particularly quick, so avoid doing mutations when possible. | |
179 // Prefer the 8-bit version when possible. | |
180 // | |
181 // It is an error to replace components of an invalid URL. The result will | |
182 // be the empty URL. | |
183 // | |
184 // Note that we use the more general url_canon::Replacements type to give | |
185 // callers extra flexibility rather than our override. | |
186 GURL_API GURL ReplaceComponents( | |
187 const url_canon::Replacements<char>& replacements) const; | |
188 GURL_API GURL ReplaceComponents( | |
189 const url_canon::Replacements<char16>& replacements) const; | |
190 | |
191 // A helper function that is equivalent to replacing the path with a slash | |
192 // and clearing out everything after that. We sometimes need to know just the | |
193 // scheme and the authority. If this URL is not a standard URL (it doesn't | |
194 // have the regular authority and path sections), then the result will be | |
195 // an empty, invalid GURL. Note that this *does* work for file: URLs, which | |
196 // some callers may want to filter out before calling this. | |
197 // | |
198 // It is an error to get an empty path on an invalid URL. The result | |
199 // will be the empty URL. | |
200 GURL_API GURL GetWithEmptyPath() const; | |
201 | |
202 // A helper function to return a GURL containing just the scheme, host, | |
203 // and port from a URL. Equivalent to clearing any username and password, | |
204 // replacing the path with a slash, and clearing everything after that. If | |
205 // this URL is not a standard URL, then the result will be an empty, | |
206 // invalid GURL. If the URL has neither username nor password, this | |
207 // degenerates to GetWithEmptyPath(). | |
208 // | |
209 // It is an error to get the origin of an invalid URL. The result | |
210 // will be the empty URL. | |
211 GURL_API GURL GetOrigin() const; | |
212 | |
213 // Returns true if the scheme for the current URL is a known "standard" | |
214 // scheme. Standard schemes have an authority and a path section. This | |
215 // includes file: and filesystem:, which some callers may want to filter out | |
216 // explicitly by calling SchemeIsFile[System]. | |
217 GURL_API bool IsStandard() const; | |
218 | |
219 // Returns true if the given parameter (should be lower-case ASCII to match | |
220 // the canonicalized scheme) is the scheme for this URL. This call is more | |
221 // efficient than getting the scheme and comparing it because no copies or | |
222 // object constructions are done. | |
223 GURL_API bool SchemeIs(const char* lower_ascii_scheme) const; | |
224 | |
225 // We often need to know if this is a file URL. File URLs are "standard", but | |
226 // are often treated separately by some programs. | |
227 bool SchemeIsFile() const { | |
228 return SchemeIs("file"); | |
229 } | |
230 | |
231 // FileSystem URLs need to be treated differently in some cases. | |
232 bool SchemeIsFileSystem() const { | |
233 return SchemeIs("filesystem"); | |
234 } | |
235 | |
236 // If the scheme indicates a secure connection | |
237 bool SchemeIsSecure() const { | |
238 return SchemeIs("https") || SchemeIs("wss") || | |
239 (SchemeIsFileSystem() && inner_url() && inner_url()->SchemeIsSecure()); | |
240 } | |
241 | |
242 // Returns true if the hostname is an IP address. Note: this function isn't | |
243 // as cheap as a simple getter because it re-parses the hostname to verify. | |
244 // This currently identifies only IPv4 addresses (bug 822685). | |
245 GURL_API bool HostIsIPAddress() const; | |
246 | |
247 // Getters for various components of the URL. The returned string will be | |
248 // empty if the component is empty or is not present. | |
249 std::string scheme() const { // Not including the colon. See also SchemeIs. | |
250 return ComponentString(parsed_.scheme); | |
251 } | |
252 std::string username() const { | |
253 return ComponentString(parsed_.username); | |
254 } | |
255 std::string password() const { | |
256 return ComponentString(parsed_.password); | |
257 } | |
258 // Note that this may be a hostname, an IPv4 address, or an IPv6 literal | |
259 // surrounded by square brackets, like "[2001:db8::1]". To exclude these | |
260 // brackets, use HostNoBrackets() below. | |
261 std::string host() const { | |
262 return ComponentString(parsed_.host); | |
263 } | |
264 std::string port() const { // Returns -1 if "default" | |
265 return ComponentString(parsed_.port); | |
266 } | |
267 std::string path() const { // Including first slash following host | |
268 return ComponentString(parsed_.path); | |
269 } | |
270 std::string query() const { // Stuff following '?' | |
271 return ComponentString(parsed_.query); | |
272 } | |
273 std::string ref() const { // Stuff following '#' | |
274 return ComponentString(parsed_.ref); | |
275 } | |
276 | |
277 // Existance querying. These functions will return true if the corresponding | |
278 // URL component exists in this URL. Note that existance is different than | |
279 // being nonempty. http://www.google.com/? has a query that just happens to | |
280 // be empty, and has_query() will return true. | |
281 bool has_scheme() const { | |
282 return parsed_.scheme.len >= 0; | |
283 } | |
284 bool has_username() const { | |
285 return parsed_.username.len >= 0; | |
286 } | |
287 bool has_password() const { | |
288 return parsed_.password.len >= 0; | |
289 } | |
290 bool has_host() const { | |
291 // Note that hosts are special, absense of host means length 0. | |
292 return parsed_.host.len > 0; | |
293 } | |
294 bool has_port() const { | |
295 return parsed_.port.len >= 0; | |
296 } | |
297 bool has_path() const { | |
298 // Note that http://www.google.com/" has a path, the path is "/". This can | |
299 // return false only for invalid or nonstandard URLs. | |
300 return parsed_.path.len >= 0; | |
301 } | |
302 bool has_query() const { | |
303 return parsed_.query.len >= 0; | |
304 } | |
305 bool has_ref() const { | |
306 return parsed_.ref.len >= 0; | |
307 } | |
308 | |
309 // Returns a parsed version of the port. Can also be any of the special | |
310 // values defined in Parsed for ExtractPort. | |
311 GURL_API int IntPort() const; | |
312 | |
313 // Returns the port number of the url, or the default port number. | |
314 // If the scheme has no concept of port (or unknown default) returns | |
315 // PORT_UNSPECIFIED. | |
316 GURL_API int EffectiveIntPort() const; | |
317 | |
318 // Extracts the filename portion of the path and returns it. The filename | |
319 // is everything after the last slash in the path. This may be empty. | |
320 GURL_API std::string ExtractFileName() const; | |
321 | |
322 // Returns the path that should be sent to the server. This is the path, | |
323 // parameter, and query portions of the URL. It is guaranteed to be ASCII. | |
324 GURL_API std::string PathForRequest() const; | |
325 | |
326 // Returns the host, excluding the square brackets surrounding IPv6 address | |
327 // literals. This can be useful for passing to getaddrinfo(). | |
328 GURL_API std::string HostNoBrackets() const; | |
329 | |
330 // Returns true if this URL's host matches or is in the same domain as | |
331 // the given input string. For example if this URL was "www.google.com", | |
332 // this would match "com", "google.com", and "www.google.com | |
333 // (input domain should be lower-case ASCII to match the canonicalized | |
334 // scheme). This call is more efficient than getting the host and check | |
335 // whether host has the specific domain or not because no copies or | |
336 // object constructions are done. | |
337 // | |
338 // If function DomainIs has parameter domain_len, which means the parameter | |
339 // lower_ascii_domain does not gurantee to terminate with NULL character. | |
340 GURL_API bool DomainIs(const char* lower_ascii_domain, int domain_len) const; | |
341 | |
342 // If function DomainIs only has parameter lower_ascii_domain, which means | |
343 // domain string should be terminate with NULL character. | |
344 bool DomainIs(const char* lower_ascii_domain) const { | |
345 return DomainIs(lower_ascii_domain, | |
346 static_cast<int>(strlen(lower_ascii_domain))); | |
347 } | |
348 | |
349 // Swaps the contents of this GURL object with the argument without doing | |
350 // any memory allocations. | |
351 GURL_API void Swap(GURL* other); | |
352 | |
353 // Returns a reference to a singleton empty GURL. This object is for callers | |
354 // who return references but don't have anything to return in some cases. | |
355 // This function may be called from any thread. | |
356 GURL_API static const GURL& EmptyGURL(); | |
357 | |
358 // Returns the inner URL of a nested URL [currently only non-null for | |
359 // filesystem: URLs]. | |
360 const GURL* inner_url() const { | |
361 return inner_url_; | |
362 } | |
363 | |
364 private: | |
365 // Returns the substring of the input identified by the given component. | |
366 std::string ComponentString(const url_parse::Component& comp) const { | |
367 if (comp.len <= 0) | |
368 return std::string(); | |
369 return std::string(spec_, comp.begin, comp.len); | |
370 } | |
371 | |
372 // The actual text of the URL, in canonical ASCII form. | |
373 std::string spec_; | |
374 | |
375 // Set when the given URL is valid. Otherwise, we may still have a spec and | |
376 // components, but they may not identify valid resources (for example, an | |
377 // invalid port number, invalid characters in the scheme, etc.). | |
378 bool is_valid_; | |
379 | |
380 // Identified components of the canonical spec. | |
381 url_parse::Parsed parsed_; | |
382 | |
383 // Used for nested schemes [currently only filesystem:]. | |
384 GURL* inner_url_; | |
385 | |
386 // TODO bug 684583: Add encoding for query params. | |
387 }; | |
388 | |
389 // Stream operator so GURL can be used in assertion statements. | |
390 GURL_API std::ostream& operator<<(std::ostream& out, const GURL& url); | |
391 | 34 |
392 #endif // GOOGLEURL_SRC_GURL_H__ | 35 #endif // GOOGLEURL_SRC_GURL_H__ |
OLD | NEW |