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.
5
#ifndef URL_URL_UTIL_H_
6
#define URL_URL_UTIL_H_
10
#include "base/strings/string16.h"
11
#include "url/url_canon.h"
12
#include "url/url_export.h"
13
#include "url/url_parse.h"
17
// Init ------------------------------------------------------------------------
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.
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
30
URL_EXPORT void Initialize();
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
36
URL_EXPORT void Shutdown();
38
// Schemes --------------------------------------------------------------------
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);
46
// Sets a flag to prevent future calls to AddStandardScheme from succeeding.
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
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();
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.
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,
69
url_parse::Component* found_scheme);
70
URL_EXPORT bool FindAndCompareScheme(const base::char16* str,
73
url_parse::Component* found_scheme);
74
inline bool FindAndCompareScheme(const std::string& str,
76
url_parse::Component* found_scheme) {
77
return FindAndCompareScheme(str.data(), static_cast<int>(str.size()),
78
compare, found_scheme);
80
inline bool FindAndCompareScheme(const base::string16& str,
82
url_parse::Component* found_scheme) {
83
return FindAndCompareScheme(str.data(), static_cast<int>(str.size()),
84
compare, found_scheme);
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);
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);
101
// URL library wrappers -------------------------------------------------------
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.
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).
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,
116
url_canon::CharsetConverter* charset_converter,
117
url_canon::CanonOutput* output,
118
url_parse::Parsed* output_parsed);
119
URL_EXPORT bool Canonicalize(const base::char16* spec,
122
url_canon::CharsetConverter* charset_converter,
123
url_canon::CanonOutput* output,
124
url_parse::Parsed* output_parsed);
126
// Resolves a potentially relative URL relative to the given parsed base URL.
127
// The base MUST be valid. The resulting canonical URL and parsed information
128
// will be placed in to the given out variables.
130
// The relative need not be relative. If we discover that it's absolute, this
131
// will produce a canonical version of that URL. See Canonicalize() for more
132
// about the charset_converter.
134
// Returns true if the output is valid, false if the input could not produce
136
URL_EXPORT bool ResolveRelative(const char* base_spec,
138
const url_parse::Parsed& base_parsed,
139
const char* relative,
141
url_canon::CharsetConverter* charset_converter,
142
url_canon::CanonOutput* output,
143
url_parse::Parsed* output_parsed);
144
URL_EXPORT bool ResolveRelative(const char* base_spec,
146
const url_parse::Parsed& base_parsed,
147
const base::char16* relative,
149
url_canon::CharsetConverter* charset_converter,
150
url_canon::CanonOutput* output,
151
url_parse::Parsed* output_parsed);
153
// Replaces components in the given VALID input url. The new canonical URL info
154
// is written to output and out_parsed.
156
// Returns true if the resulting URL is valid.
157
URL_EXPORT bool ReplaceComponents(
160
const url_parse::Parsed& parsed,
161
const url_canon::Replacements<char>& replacements,
162
url_canon::CharsetConverter* charset_converter,
163
url_canon::CanonOutput* output,
164
url_parse::Parsed* out_parsed);
165
URL_EXPORT bool ReplaceComponents(
168
const url_parse::Parsed& parsed,
169
const url_canon::Replacements<base::char16>& replacements,
170
url_canon::CharsetConverter* charset_converter,
171
url_canon::CanonOutput* output,
172
url_parse::Parsed* out_parsed);
174
// String helper functions ----------------------------------------------------
176
// Compare the lower-case form of the given string against the given ASCII
177
// string. This is useful for doing checking if an input string matches some
178
// token, and it is optimized to avoid intermediate string copies.
180
// The versions of this function that don't take a b_end assume that the b
181
// string is NULL terminated.
182
URL_EXPORT bool LowerCaseEqualsASCII(const char* a_begin,
185
URL_EXPORT bool LowerCaseEqualsASCII(const char* a_begin,
189
URL_EXPORT bool LowerCaseEqualsASCII(const base::char16* a_begin,
190
const base::char16* a_end,
193
// Unescapes the given string using URL escaping rules.
194
URL_EXPORT void DecodeURLEscapeSequences(const char* input, int length,
195
url_canon::CanonOutputW* output);
197
// Escapes the given string as defined by the JS method encodeURIComponent. See
198
// https://developer.mozilla.org/en/JavaScript/Reference/Global_Objects/encodeURIComponent
199
URL_EXPORT void EncodeURIComponent(const char* input, int length,
200
url_canon::CanonOutput* output);
203
} // namespace url_util
205
#endif // URL_URL_UTIL_H_