WebCore/KURL.h

   1 /*
   2  * Copyright (C) 2003, 2004, 2005, 2006, 2007, 2008 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 COMPUTER, 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 COMPUTER, 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 KURL_h
  27 #define KURL_h
  28
  29 #include "PlatformString.h"
  30
  31 #if PLATFORM(CF)
  32 typedef const struct __CFURL* CFURLRef;
  33 #endif
  34
  35 #if PLATFORM(MAC)
  36 #ifdef __OBJC__
  37 @class NSURL;
  38 #else
  39 class NSURL;
  40 #endif
  41 #endif
  42
  43 #if PLATFORM(QT)
  44 QT_BEGIN_NAMESPACE
  45 class QUrl;
  46 QT_END_NAMESPACE
  47 #endif
  48
  49 #if USE(GOOGLEURL)
  50 #include "GoogleURLPrivate.h"
  51 #endif
  52
  53 namespace WebCore {
  54
  55 class TextEncoding;
  56 struct KURLHash;
  57
  58 // FIXME: Our terminology here is a bit inconsistent. We refer to the part
  59 // after the "#" as the "fragment" in some places and the "ref" in others.
  60 // We should fix the terminology to match the URL and URI RFCs as closely
  61 // as possible to resolve this.
  62
  63 class KURL {
  64 public:
  65     // Generates a URL which contains a null string.
  66     KURL() { invalidate(); }
  67
  68     // The argument is an absolute URL string. The string is assumed to be
  69     // an already encoded (ASCII-only) valid absolute URL.
  70     explicit KURL(const char*);
  71     explicit KURL(const String&);
  72
  73     // Resolves the relative URL with the given base URL. If provided, the
  74     // TextEncoding is used to encode non-ASCII characers. The base URL can be
  75     // null or empty, in which case the relative URL will be interpreted as
  76     // absolute.
  77     // FIXME: If the base URL is invalid, this always creates an invalid
  78     // URL. Instead I think it would be better to treat all invalid base URLs
  79     // the same way we treate null and empty base URLs.
  80     KURL(const KURL& base, const String& relative);
  81     KURL(const KURL& base, const String& relative, const TextEncoding&);
  82
  83 #if USE(GOOGLEURL)
  84     // For conversions for other structures that have already parsed and
  85     // canonicalized the URL. The input must be exactly what KURL would have
  86     // done with the same input.
  87     KURL(const char* canonicalSpec, int canonicalSpecLen,
  88          const url_parse::Parsed& parsed, bool isValid);
  89 #endif
  90
  91     // FIXME: The above functions should be harmonized so that passing a
  92     // base of null or the empty string gives the same result as the
  93     // standard String constructor.
  94
  95     // Makes a deep copy. Helpful only if you need to use a KURL on another
  96     // thread.  Since the underlying StringImpl objects are immutable, there's
  97     // no other reason to ever prefer copy() over plain old assignment.
  98     KURL copy() const;
  99
 100     bool isNull() const;
 101     bool isEmpty() const;
 102     bool isValid() const;
 103
 104     // Returns true if this URL has a path. Note that "http://foo.com/" has a
 105     // path of "/", so this function will return true. Only invalid or
 106     // non-hierarchical (like "javascript:") URLs will have no path.
 107     bool hasPath() const;
 108
 109 #if USE(GOOGLEURL)
 110     const String& string() const { return m_url.string(); }
 111 #else
 112     const String& string() const { return m_string; }
 113 #endif
 114
 115     String protocol() const;
 116     String host() const;
 117     unsigned short port() const;
 118     String user() const;
 119     String pass() const;
 120     String path() const;
 121     String lastPathComponent() const;
 122     String query() const; // Includes the "?".
 123     String ref() const; // Does *not* include the "#".
 124     bool hasRef() const;
 125
 126     String prettyURL() const;
 127     String fileSystemPath() const;
 128
 129     // Returns true if the current URL's protocol is the same as the null-
 130     // terminated ASCII argument. The argument must be lower-case.
 131     bool protocolIs(const char*) const;
 132     bool isLocalFile() const;
 133
 134     void setProtocol(const String&);
 135     void setHost(const String&);
 136
 137     // Setting the port to 0 will clear any port from the URL.
 138     void setPort(unsigned short);
 139
 140     // Input is like "foo.com" or "foo.com:8000".
 141     void setHostAndPort(const String&);
 142
 143     void setUser(const String&);
 144     void setPass(const String&);
 145
 146     // If you pass an empty path for HTTP or HTTPS URLs, the resulting path
 147     // will be "/".
 148     void setPath(const String&);
 149
 150     // The query may begin with a question mark, or, if not, one will be added
 151     // for you. Setting the query to the empty string will leave a "?" in the
 152     // URL (with nothing after it). To clear the query, pass a null string.
 153     void setQuery(const String&);
 154
 155     void setRef(const String&);
 156     void removeRef();
 157
 158     friend bool equalIgnoringRef(const KURL&, const KURL&);
 159
 160     friend bool protocolHostAndPortAreEqual(const KURL&, const KURL&);
 161
 162     unsigned hostStart() const;
 163     unsigned hostEnd() const;
 164
 165     unsigned pathStart() const;
 166     unsigned pathEnd() const;
 167     unsigned pathAfterLastSlash() const;
 168     operator const String&() const { return string(); }
 169 #if USE(JSC)
 170     operator JSC::UString() const { return string(); }
 171 #endif
 172
 173 #if PLATFORM(CF)
 174     KURL(CFURLRef);
 175     CFURLRef createCFURL() const;
 176 #endif
 177
 178 #if PLATFORM(MAC)
 179     KURL(NSURL*);
 180     operator NSURL*() const;
 181 #endif
 182 #ifdef __OBJC__
 183     operator NSString*() const { return string(); }
 184 #endif
 185
 186 #if PLATFORM(QT)
 187     KURL(const QUrl&);
 188     operator QUrl() const;
 189 #endif
 190
 191 #if USE(GOOGLEURL)
 192     // Getters for the parsed structure and its corresponding 8-bit string.
 193     const url_parse::Parsed& parsed() const { return m_url.m_parsed; }
 194     const CString& utf8String() const { return m_url.utf8String(); }
 195 #endif
 196
 197 #ifndef NDEBUG
 198     void print() const;
 199 #endif
 200
 201 private:
 202     void invalidate();
 203     bool isHierarchical() const;
 204     static bool protocolIs(const String&, const char*);
 205 #if USE(GOOGLEURL)
 206     friend class GoogleURLPrivate;
 207     void parse(const char* url, const String* originalString);  // KURLMac calls this.
 208     void copyToBuffer(Vector<char, 512>& buffer) const;  // KURLCFNet uses this.
 209     GoogleURLPrivate m_url;
 210 #else  // !USE(GOOGLEURL)
 211     void init(const KURL&, const String&, const TextEncoding&);
 212     void copyToBuffer(Vector<char, 512>& buffer) const;
 213
 214     // Parses the given URL. The originalString parameter allows for an
 215     // optimization: When the source is the same as the fixed-up string,
 216     // it will use the passed-in string instead of allocating a new one.
 217     void parse(const String&);
 218     void parse(const char* url, const String* originalString);
 219
 220     String m_string;
 221     bool m_isValid;
 222     int m_schemeEnd;
 223     int m_userStart;
 224     int m_userEnd;
 225     int m_passwordEnd;
 226     int m_hostEnd;
 227     int m_portEnd;
 228     int m_pathAfterLastSlash;
 229     int m_pathEnd;
 230     int m_queryEnd;
 231     int m_fragmentEnd;
 232 #endif
 233 };
 234
 235 bool operator==(const KURL&, const KURL&);
 236 bool operator==(const KURL&, const String&);
 237 bool operator==(const String&, const KURL&);
 238 bool operator!=(const KURL&, const KURL&);
 239 bool operator!=(const KURL&, const String&);
 240 bool operator!=(const String&, const KURL&);
 241
 242 bool equalIgnoringRef(const KURL&, const KURL&);
 243 bool protocolHostAndPortAreEqual(const KURL&, const KURL&);
 244
 245 const KURL& blankURL();
 246
 247 // Functions to do URL operations on strings.
 248 // These are operations that aren't faster on a parsed URL.
 249
 250 bool protocolIs(const String& url, const char* protocol);
 251
 252 String mimeTypeFromDataURL(const String& url);
 253
 254 // Unescapes the given string using URL escaping rules, given an optional
 255 // encoding (defaulting to UTF-8 otherwise). DANGER: If the URL has "%00"
 256 // in it, the resulting string will have embedded null characters!
 257 String decodeURLEscapeSequences(const String&);
 258 String decodeURLEscapeSequences(const String&, const TextEncoding&);
 259
 260 String encodeWithURLEscapeSequences(const String&);
 261
 262 // Inlines.
 263
 264 inline bool operator==(const KURL& a, const KURL& b)
 265 {
 266     return a.string() == b.string();
 267 }
 268
 269 inline bool operator==(const KURL& a, const String& b)
 270 {
 271     return a.string() == b;
 272 }
 273
 274 inline bool operator==(const String& a, const KURL& b)
 275 {
 276     return a == b.string();
 277 }
 278
 279 inline bool operator!=(const KURL& a, const KURL& b)
 280 {
 281     return a.string() != b.string();
 282 }
 283
 284 inline bool operator!=(const KURL& a, const String& b)
 285 {
 286     return a.string() != b;
 287 }
 288
 289 inline bool operator!=(const String& a, const KURL& b)
 290 {
 291     return a != b.string();
 292 }
 293
 294 #if !USE(GOOGLEURL)
 295
 296 // Inline versions of some non-GoogleURL functions so we can get inlining
 297 // without having to have a lot of ugly ifdefs in the class definition.
 298
 299 inline bool KURL::isNull() const
 300 {
 301     return m_string.isNull();
 302 }
 303
 304 inline bool KURL::isEmpty() const
 305 {
 306     return m_string.isEmpty();
 307 }
 308
 309 inline bool KURL::isValid() const
 310 {
 311     return m_isValid;
 312 }
 313
 314 inline unsigned KURL::hostStart() const
 315 {
 316     return (m_passwordEnd == m_userStart) ? m_passwordEnd : m_passwordEnd + 1;
 317 }
 318
 319 inline unsigned KURL::hostEnd() const
 320 {
 321     return m_hostEnd;
 322 }
 323
 324 inline unsigned KURL::pathStart() const
 325 {
 326     return m_portEnd;
 327 }
 328
 329 inline unsigned KURL::pathEnd() const
 330 {
 331     return m_pathEnd;
 332 }
 333
 334 inline unsigned KURL::pathAfterLastSlash() const
 335 {
 336     return m_pathAfterLastSlash;
 337 }
 338
 339 #endif  // !USE(GOOGLEURL)
 340
 341 } // namespace WebCore
 342
 343 namespace WTF {
 344
 345     // KURLHash is the default hash for String
 346     template<typename T> struct DefaultHash;
 347     template<> struct DefaultHash<WebCore::KURL> {
 348         typedef WebCore::KURLHash Hash;
 349     };
 350
 351 } // namespace WTF
 352
 353 #endif // KURL_h