url/url_util.h

   1 // Copyright 2013 The Chromium Authors. All rights reserved.
   2 // Use of this source code is governed by a BSD-style license that can be
   3 // found in the LICENSE file.
   4
   5 #ifndef URL_URL_UTIL_H_
   6 #define URL_URL_UTIL_H_
   7
   8 #include <string>
   9
  10 #include "base/strings/string16.h"
  11 #include "url/url_canon.h"
  12 #include "url/url_export.h"
  13 #include "url/url_parse.h"
  14
  15 namespace url_util {
  16
  17 // Init ------------------------------------------------------------------------
  18
  19 // Initialization is NOT required, it will be implicitly initialized when first
  20 // used. However, this implicit initialization is NOT threadsafe. If you are
  21 // using this library in a threaded environment and don't have a consistent
  22 // "first call" (an example might be calling "AddStandardScheme" with your
  23 // special application-specific schemes) then you will want to call initialize
  24 // before spawning any threads.
  25 //
  26 // It is OK to call this function more than once, subsequent calls will simply
  27 // "noop", unless Shutdown() was called in the mean time. This will also be a
  28 // "noop" if other calls to the library have forced an initialization
  29 // beforehand.
  30 URL_EXPORT void Initialize();
  31
  32 // Cleanup is not required, except some strings may leak. For most user
  33 // applications, this is fine. If you're using it in a library that may get
  34 // loaded and unloaded, you'll want to unload to properly clean up your
  35 // library.
  36 URL_EXPORT void Shutdown();
  37
  38 // Schemes --------------------------------------------------------------------
  39
  40 // Adds an application-defined scheme to the internal list of "standard" URL
  41 // schemes. This function is not threadsafe and can not be called concurrently
  42 // with any other url_util function. It will assert if the list of standard
  43 // schemes has been locked (see LockStandardSchemes).
  44 URL_EXPORT void AddStandardScheme(const char* new_scheme);
  45
  46 // Sets a flag to prevent future calls to AddStandardScheme from succeeding.
  47 //
  48 // This is designed to help prevent errors for multithreaded applications.
  49 // Normal usage would be to call AddStandardScheme for your custom schemes at
  50 // the beginning of program initialization, and then LockStandardSchemes. This
  51 // prevents future callers from mistakenly calling AddStandardScheme when the
  52 // program is running with multiple threads, where such usage would be
  53 // dangerous.
  54 //
  55 // We could have had AddStandardScheme use a lock instead, but that would add
  56 // some platform-specific dependencies we don't otherwise have now, and is
  57 // overkill considering the normal usage is so simple.
  58 URL_EXPORT void LockStandardSchemes();
  59
  60 // Locates the scheme in the given string and places it into |found_scheme|,
  61 // which may be NULL to indicate the caller does not care about the range.
  62 //
  63 // Returns whether the given |compare| scheme matches the scheme found in the
  64 // input (if any). The |compare| scheme must be a valid canonical scheme or
  65 // the result of the comparison is undefined.
  66 URL_EXPORT bool FindAndCompareScheme(const char* str,
  67                                      int str_len,
  68                                      const char* compare,
  69                                      url_parse::Component* found_scheme);
  70 URL_EXPORT bool FindAndCompareScheme(const base::char16* str,
  71                                      int str_len,
  72                                      const char* compare,
  73                                      url_parse::Component* found_scheme);
  74 inline bool FindAndCompareScheme(const std::string& str,
  75                                  const char* compare,
  76                                  url_parse::Component* found_scheme) {
  77   return FindAndCompareScheme(str.data(), static_cast<int>(str.size()),
  78                               compare, found_scheme);
  79 }
  80 inline bool FindAndCompareScheme(const base::string16& str,
  81                                  const char* compare,
  82                                  url_parse::Component* found_scheme) {
  83   return FindAndCompareScheme(str.data(), static_cast<int>(str.size()),
  84                               compare, found_scheme);
  85 }
  86
  87 // Returns true if the given string represents a standard URL. This means that
  88 // either the scheme is in the list of known standard schemes.
  89 URL_EXPORT bool IsStandard(const char* spec,
  90                            const url_parse::Component& scheme);
  91 URL_EXPORT bool IsStandard(const base::char16* spec,
  92                            const url_parse::Component& scheme);
  93
  94 // TODO(brettw) remove this. This is a temporary compatibility hack to avoid
  95 // breaking the WebKit build when this version is synced via Chrome.
  96 inline bool IsStandard(const char* spec, int spec_len,
  97                        const url_parse::Component& scheme) {
  98   return IsStandard(spec, scheme);
  99 }
 100
 101 // URL library wrappers -------------------------------------------------------
 102
 103 // Parses the given spec according to the extracted scheme type. Normal users
 104 // should use the URL object, although this may be useful if performance is
 105 // critical and you don't want to do the heap allocation for the std::string.
 106 //
 107 // As with the url_canon::Canonicalize* functions, the charset converter can
 108 // be NULL to use UTF-8 (it will be faster in this case).
 109 //
 110 // Returns true if a valid URL was produced, false if not. On failure, the
 111 // output and parsed structures will still be filled and will be consistent,
 112 // but they will not represent a loadable URL.
 113 URL_EXPORT bool Canonicalize(const char* spec,
 114                              int spec_len,
 115                              url_canon::CharsetConverter* charset_converter,
 116                              url_canon::CanonOutput* output,
 117                              url_parse::Parsed* output_parsed);
 118 URL_EXPORT bool Canonicalize(const base::char16* spec,
 119                              int spec_len,
 120                              url_canon::CharsetConverter* charset_converter,
 121                              url_canon::CanonOutput* output,
 122                              url_parse::Parsed* output_parsed);
 123
 124 // Resolves a potentially relative URL relative to the given parsed base URL.
 125 // The base MUST be valid. The resulting canonical URL and parsed information
 126 // will be placed in to the given out variables.
 127 //
 128 // The relative need not be relative. If we discover that it's absolute, this
 129 // will produce a canonical version of that URL. See Canonicalize() for more
 130 // about the charset_converter.
 131 //
 132 // Returns true if the output is valid, false if the input could not produce
 133 // a valid URL.
 134 URL_EXPORT bool ResolveRelative(const char* base_spec,
 135                                 int base_spec_len,
 136                                 const url_parse::Parsed& base_parsed,
 137                                 const char* relative,
 138                                 int relative_length,
 139                                 url_canon::CharsetConverter* charset_converter,
 140                                 url_canon::CanonOutput* output,
 141                                 url_parse::Parsed* output_parsed);
 142 URL_EXPORT bool ResolveRelative(const char* base_spec,
 143                                 int base_spec_len,
 144                                 const url_parse::Parsed& base_parsed,
 145                                 const base::char16* relative,
 146                                 int relative_length,
 147                                 url_canon::CharsetConverter* charset_converter,
 148                                 url_canon::CanonOutput* output,
 149                                 url_parse::Parsed* output_parsed);
 150
 151 // Replaces components in the given VALID input url. The new canonical URL info
 152 // is written to output and out_parsed.
 153 //
 154 // Returns true if the resulting URL is valid.
 155 URL_EXPORT bool ReplaceComponents(
 156     const char* spec,
 157     int spec_len,
 158     const url_parse::Parsed& parsed,
 159     const url_canon::Replacements<char>& replacements,
 160     url_canon::CharsetConverter* charset_converter,
 161     url_canon::CanonOutput* output,
 162     url_parse::Parsed* out_parsed);
 163 URL_EXPORT bool ReplaceComponents(
 164     const char* spec,
 165     int spec_len,
 166     const url_parse::Parsed& parsed,
 167     const url_canon::Replacements<base::char16>& replacements,
 168     url_canon::CharsetConverter* charset_converter,
 169     url_canon::CanonOutput* output,
 170     url_parse::Parsed* out_parsed);
 171
 172 // String helper functions ----------------------------------------------------
 173
 174 // Compare the lower-case form of the given string against the given ASCII
 175 // string.  This is useful for doing checking if an input string matches some
 176 // token, and it is optimized to avoid intermediate string copies.
 177 //
 178 // The versions of this function that don't take a b_end assume that the b
 179 // string is NULL terminated.
 180 URL_EXPORT bool LowerCaseEqualsASCII(const char* a_begin,
 181                                      const char* a_end,
 182                                      const char* b);
 183 URL_EXPORT bool LowerCaseEqualsASCII(const char* a_begin,
 184                                      const char* a_end,
 185                                      const char* b_begin,
 186                                      const char* b_end);
 187 URL_EXPORT bool LowerCaseEqualsASCII(const base::char16* a_begin,
 188                                      const base::char16* a_end,
 189                                      const char* b);
 190
 191 // Unescapes the given string using URL escaping rules.
 192 URL_EXPORT void DecodeURLEscapeSequences(const char* input, int length,
 193                                          url_canon::CanonOutputW* output);
 194
 195 // Escapes the given string as defined by the JS method encodeURIComponent.  See
 196 // https://developer.mozilla.org/en/JavaScript/Reference/Global_Objects/encodeURIComponent
 197 URL_EXPORT void EncodeURIComponent(const char* input, int length,
 198                                    url_canon::CanonOutput* output);
 199
 200
 201 }  // namespace url_util
 202
 203 #endif  // URL_URL_UTIL_H_