1/*
2 * Copyright (C) 2003, 2004, 2005, 2006, 2007, 2008, 2011, 2012, 2013 Apple Inc. 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
6 * are met:
7 * 1. Redistributions of source code must retain the above copyright
8 *    notice, this list of conditions and the following disclaimer.
9 * 2. Redistributions in binary form must reproduce the above copyright
10 *    notice, this list of conditions and the following disclaimer in the
11 *    documentation and/or other materials provided with the distribution.
12 *
13 * THIS SOFTWARE IS PROVIDED BY APPLE INC. ``AS IS'' AND ANY
14 * EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
15 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
16 * PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL APPLE INC. OR
17 * CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
18 * EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
19 * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
20 * PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY
21 * OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
22 * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
23 * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
24 */
25
26#ifndef URL_h
27#define URL_h
28
29#include <wtf/Forward.h>
30#include <wtf/HashMap.h>
31#include <wtf/RetainPtr.h>
32#include <wtf/text/WTFString.h>
33
34#if USE(CF)
35typedef const struct __CFURL* CFURLRef;
36#endif
37
38#if USE(SOUP)
39#include "GUniquePtrSoup.h"
40#endif
41
42#if USE(FOUNDATION)
43OBJC_CLASS NSURL;
44#endif
45
46namespace WebCore {
47
48class TextEncoding;
49struct URLHash;
50
51enum ParsedURLStringTag { ParsedURLString };
52
53class URL {
54public:
55    // Generates a URL which contains a null string.
56    URL() { invalidate(); }
57
58    // The argument is an absolute URL string. The string is assumed to be output of URL::string() called on a valid
59    // URL object, or indiscernible from such.
60    // It is usually best to avoid repeatedly parsing a string, unless memory saving outweigh the possible slow-downs.
61    URL(ParsedURLStringTag, const String&);
62    explicit URL(WTF::HashTableDeletedValueType) : m_string(WTF::HashTableDeletedValue) { }
63    bool isHashTableDeletedValue() const { return string().isHashTableDeletedValue(); }
64
65    // Resolves the relative URL with the given base URL. If provided, the
66    // TextEncoding is used to encode non-ASCII characers. The base URL can be
67    // null or empty, in which case the relative URL will be interpreted as
68    // absolute.
69    // FIXME: If the base URL is invalid, this always creates an invalid
70    // URL. Instead I think it would be better to treat all invalid base URLs
71    // the same way we treate null and empty base URLs.
72    URL(const URL& base, const String& relative);
73    URL(const URL& base, const String& relative, const TextEncoding&);
74
75    static URL fakeURLWithRelativePart(const String&);
76
77    String strippedForUseAsReferrer() const;
78
79    // FIXME: The above functions should be harmonized so that passing a
80    // base of null or the empty string gives the same result as the
81    // standard String constructor.
82
83    // Makes a deep copy. Helpful only if you need to use a URL on another
84    // thread.  Since the underlying StringImpl objects are immutable, there's
85    // no other reason to ever prefer copy() over plain old assignment.
86    URL copy() const;
87
88    bool isNull() const;
89    bool isEmpty() const;
90    bool isValid() const;
91
92    // Returns true if you can set the host and port for the URL.
93    // Non-hierarchical URLs don't have a host and port.
94    bool canSetHostOrPort() const { return isHierarchical(); }
95
96    bool canSetPathname() const { return isHierarchical(); }
97    bool isHierarchical() const;
98
99    const String& string() const { return m_string; }
100
101    String stringCenterEllipsizedToLength(unsigned length = 1024) const;
102
103    String protocol() const;
104    String host() const;
105    unsigned short port() const;
106    bool hasPort() const;
107    String user() const;
108    String pass() const;
109    String path() const;
110    String lastPathComponent() const;
111    String query() const;
112    String fragmentIdentifier() const;
113    bool hasFragmentIdentifier() const;
114
115    String baseAsString() const;
116
117    String fileSystemPath() const;
118
119    // Returns true if the current URL's protocol is the same as the null-
120    // terminated ASCII argument. The argument must be lower-case.
121    bool protocolIs(const char*) const;
122    bool protocolIsData() const { return protocolIs("data"); }
123    bool protocolIsInHTTPFamily() const;
124    bool isLocalFile() const;
125    bool isBlankURL() const;
126
127    bool setProtocol(const String&);
128    void setHost(const String&);
129
130    void removePort();
131    void setPort(unsigned short);
132
133    // Input is like "foo.com" or "foo.com:8000".
134    void setHostAndPort(const String&);
135
136    void setUser(const String&);
137    void setPass(const String&);
138
139    // If you pass an empty path for HTTP or HTTPS URLs, the resulting path
140    // will be "/".
141    void setPath(const String&);
142
143    // The query may begin with a question mark, or, if not, one will be added
144    // for you. Setting the query to the empty string will leave a "?" in the
145    // URL (with nothing after it). To clear the query, pass a null string.
146    void setQuery(const String&);
147
148    void setFragmentIdentifier(const String&);
149    void removeFragmentIdentifier();
150
151    friend bool equalIgnoringFragmentIdentifier(const URL&, const URL&);
152
153    friend bool protocolHostAndPortAreEqual(const URL&, const URL&);
154
155    unsigned hostStart() const;
156    unsigned hostEnd() const;
157
158    unsigned pathStart() const;
159    unsigned pathEnd() const;
160    unsigned pathAfterLastSlash() const;
161
162    operator const String&() const { return string(); }
163
164#if USE(CF)
165    URL(CFURLRef);
166    RetainPtr<CFURLRef> createCFURL() const;
167#endif
168
169#if USE(SOUP)
170    URL(SoupURI*);
171    GUniquePtr<SoupURI> createSoupURI() const;
172#endif
173
174#if USE(FOUNDATION)
175    URL(NSURL*);
176    operator NSURL*() const;
177#endif
178#ifdef __OBJC__
179    operator NSString*() const { return string(); }
180#endif
181
182    const URL* innerURL() const { return 0; }
183
184#ifndef NDEBUG
185    void print() const;
186#endif
187
188    bool isSafeToSendToAnotherThread() const;
189
190private:
191    void invalidate();
192    static bool protocolIs(const String&, const char*);
193    void init(const URL&, const String&, const TextEncoding&);
194    void copyToBuffer(Vector<char, 512>& buffer) const;
195
196    // Parses the given URL. The originalString parameter allows for an
197    // optimization: When the source is the same as the fixed-up string,
198    // it will use the passed-in string instead of allocating a new one.
199    void parse(const String&);
200    void parse(const char* url, const String* originalString = 0);
201
202    bool hasPath() const;
203
204    String m_string;
205    bool m_isValid : 1;
206    bool m_protocolIsInHTTPFamily : 1;
207
208    int m_schemeEnd;
209    int m_userStart;
210    int m_userEnd;
211    int m_passwordEnd;
212    int m_hostEnd;
213    int m_portEnd;
214    int m_pathAfterLastSlash;
215    int m_pathEnd;
216    int m_queryEnd;
217    int m_fragmentEnd;
218};
219
220bool operator==(const URL&, const URL&);
221bool operator==(const URL&, const String&);
222bool operator==(const String&, const URL&);
223bool operator!=(const URL&, const URL&);
224bool operator!=(const URL&, const String&);
225bool operator!=(const String&, const URL&);
226
227bool equalIgnoringFragmentIdentifier(const URL&, const URL&);
228bool protocolHostAndPortAreEqual(const URL&, const URL&);
229
230const URL& blankURL();
231
232// Functions to do URL operations on strings.
233// These are operations that aren't faster on a parsed URL.
234// These are also different from the URL functions in that they don't require the string to be a valid and parsable URL.
235// This is especially important because valid javascript URLs are not necessarily considered valid by URL.
236
237bool protocolIs(const String& url, const char* protocol);
238bool protocolIsJavaScript(const String& url);
239bool protocolIsInHTTPFamily(const String& url);
240
241bool isDefaultPortForProtocol(unsigned short port, const String& protocol);
242bool portAllowed(const URL&); // Blacklist ports that should never be used for Web resources.
243
244bool isValidProtocol(const String&);
245
246String mimeTypeFromDataURL(const String& url);
247String mimeTypeFromURL(const URL&);
248
249// Unescapes the given string using URL escaping rules, given an optional
250// encoding (defaulting to UTF-8 otherwise). DANGER: If the URL has "%00"
251// in it, the resulting string will have embedded null characters!
252String decodeURLEscapeSequences(const String&);
253String decodeURLEscapeSequences(const String&, const TextEncoding&);
254
255String encodeWithURLEscapeSequences(const String&);
256
257// Inlines.
258
259inline bool operator==(const URL& a, const URL& b)
260{
261    return a.string() == b.string();
262}
263
264inline bool operator==(const URL& a, const String& b)
265{
266    return a.string() == b;
267}
268
269inline bool operator==(const String& a, const URL& b)
270{
271    return a == b.string();
272}
273
274inline bool operator!=(const URL& a, const URL& b)
275{
276    return a.string() != b.string();
277}
278
279inline bool operator!=(const URL& a, const String& b)
280{
281    return a.string() != b;
282}
283
284inline bool operator!=(const String& a, const URL& b)
285{
286    return a != b.string();
287}
288
289// Inline versions of some non-GoogleURL functions so we can get inlining
290// without having to have a lot of ugly ifdefs in the class definition.
291
292inline bool URL::isNull() const
293{
294    return m_string.isNull();
295}
296
297inline bool URL::isEmpty() const
298{
299    return m_string.isEmpty();
300}
301
302inline bool URL::isValid() const
303{
304    return m_isValid;
305}
306
307inline bool URL::hasPath() const
308{
309    return m_pathEnd != m_portEnd;
310}
311
312inline bool URL::hasPort() const
313{
314    return m_hostEnd < m_portEnd;
315}
316
317inline bool URL::protocolIsInHTTPFamily() const
318{
319    return m_protocolIsInHTTPFamily;
320}
321
322inline unsigned URL::hostStart() const
323{
324    return (m_passwordEnd == m_userStart) ? m_passwordEnd : m_passwordEnd + 1;
325}
326
327inline unsigned URL::hostEnd() const
328{
329    return m_hostEnd;
330}
331
332inline unsigned URL::pathStart() const
333{
334    return m_portEnd;
335}
336
337inline unsigned URL::pathEnd() const
338{
339    return m_pathEnd;
340}
341
342inline unsigned URL::pathAfterLastSlash() const
343{
344    return m_pathAfterLastSlash;
345}
346
347#if PLATFORM(IOS)
348void enableURLSchemeCanonicalization(bool);
349#endif
350
351} // namespace WebCore
352
353namespace WTF {
354
355    // URLHash is the default hash for String
356    template<typename T> struct DefaultHash;
357    template<> struct DefaultHash<WebCore::URL> {
358        typedef WebCore::URLHash Hash;
359    };
360
361} // namespace WTF
362
363#endif // URL_h
364