uri_parser 0.0.1
Sign up to get free protection for your applications and to get access to all the features.
- data/.gitignore +6 -0
- data/.rvmrc +1 -0
- data/Gemfile +6 -0
- data/Rakefile +13 -0
- data/ext/uri_parser/basictypes.h +89 -0
- data/ext/uri_parser/extconf.h +6 -0
- data/ext/uri_parser/extconf.rb +50 -0
- data/ext/uri_parser/logging.h +5 -0
- data/ext/uri_parser/scoped_ptr.h +322 -0
- data/ext/uri_parser/string16.cc +95 -0
- data/ext/uri_parser/string16.h +194 -0
- data/ext/uri_parser/uri_parser.cc +87 -0
- data/ext/uri_parser/url_canon.h +872 -0
- data/ext/uri_parser/url_canon_etc.cc +392 -0
- data/ext/uri_parser/url_canon_fileurl.cc +215 -0
- data/ext/uri_parser/url_canon_host.cc +401 -0
- data/ext/uri_parser/url_canon_icu.cc +207 -0
- data/ext/uri_parser/url_canon_icu.h +63 -0
- data/ext/uri_parser/url_canon_internal.cc +427 -0
- data/ext/uri_parser/url_canon_internal.h +453 -0
- data/ext/uri_parser/url_canon_internal_file.h +157 -0
- data/ext/uri_parser/url_canon_ip.cc +737 -0
- data/ext/uri_parser/url_canon_ip.h +101 -0
- data/ext/uri_parser/url_canon_mailtourl.cc +137 -0
- data/ext/uri_parser/url_canon_path.cc +380 -0
- data/ext/uri_parser/url_canon_pathurl.cc +128 -0
- data/ext/uri_parser/url_canon_query.cc +189 -0
- data/ext/uri_parser/url_canon_relative.cc +572 -0
- data/ext/uri_parser/url_canon_stdstring.h +134 -0
- data/ext/uri_parser/url_canon_stdurl.cc +211 -0
- data/ext/uri_parser/url_common.h +48 -0
- data/ext/uri_parser/url_file.h +108 -0
- data/ext/uri_parser/url_parse.cc +760 -0
- data/ext/uri_parser/url_parse.h +336 -0
- data/ext/uri_parser/url_parse_file.cc +243 -0
- data/ext/uri_parser/url_parse_internal.h +112 -0
- data/ext/uri_parser/url_util.cc +553 -0
- data/ext/uri_parser/url_util.h +222 -0
- data/lib/uri_parser.rb +28 -0
- data/lib/uri_parser/version.rb +3 -0
- data/spec/spec_helper.rb +16 -0
- data/spec/uri_parser_spec.rb +54 -0
- data/uri_parser.gemspec +26 -0
- metadata +117 -0
data/.gitignore
ADDED
data/.rvmrc
ADDED
@@ -0,0 +1 @@
|
|
1
|
+
rvm use 1.9.2@uri_parser --create
|
data/Gemfile
ADDED
data/Rakefile
ADDED
@@ -0,0 +1,13 @@
|
|
1
|
+
require 'bundler'
|
2
|
+
Bundler::GemHelper.install_tasks
|
3
|
+
|
4
|
+
require 'rake/extensiontask'
|
5
|
+
Rake::ExtensionTask.new('uri_parser') do |ext|
|
6
|
+
ext.lib_dir = File.join('lib', 'uri_parser')
|
7
|
+
ext.source_pattern = "*.{c,cpp,cc,h}"
|
8
|
+
end
|
9
|
+
|
10
|
+
require "rspec/core/rake_task"
|
11
|
+
RSpec::Core::RakeTask.new(:spec => :compile)
|
12
|
+
task :default => :spec
|
13
|
+
|
@@ -0,0 +1,89 @@
|
|
1
|
+
// Copyright 2001 - 2003 Google Inc. All Rights Reserved
|
2
|
+
|
3
|
+
#ifndef BASE_BASICTYPES_H__
|
4
|
+
#define BASE_BASICTYPES_H__
|
5
|
+
|
6
|
+
typedef unsigned char uint8;
|
7
|
+
typedef unsigned short uint16;
|
8
|
+
typedef unsigned int uint32;
|
9
|
+
typedef unsigned long long uint64;
|
10
|
+
|
11
|
+
const uint8 kuint8max = (( uint8) 0xFF);
|
12
|
+
const uint32 kuint32max = ((uint32) 0xFFFFFFFF);
|
13
|
+
|
14
|
+
// The arraysize(arr) macro returns the # of elements in an array arr.
|
15
|
+
// The expression is a compile-time constant, and therefore can be
|
16
|
+
// used in defining new arrays, for example. If you use arraysize on
|
17
|
+
// a pointer by mistake, you will get a compile-time error.
|
18
|
+
//
|
19
|
+
// One caveat is that arraysize() doesn't accept any array of an
|
20
|
+
// anonymous type or a type defined inside a function. In these rare
|
21
|
+
// cases, you have to use the unsafe ARRAYSIZE() macro below. This is
|
22
|
+
// due to a limitation in C++'s template system. The limitation might
|
23
|
+
// eventually be removed, but it hasn't happened yet.
|
24
|
+
|
25
|
+
// This template function declaration is used in defining arraysize.
|
26
|
+
// Note that the function doesn't need an implementation, as we only
|
27
|
+
// use its type.
|
28
|
+
template <typename T, size_t N>
|
29
|
+
char (&ArraySizeHelper(T (&array)[N]))[N];
|
30
|
+
|
31
|
+
// That gcc wants both of these prototypes seems mysterious. VC, for
|
32
|
+
// its part, can't decide which to use (another mystery). Matching of
|
33
|
+
// template overloads: the final frontier.
|
34
|
+
#ifndef _MSC_VER
|
35
|
+
template <typename T, size_t N>
|
36
|
+
char (&ArraySizeHelper(const T (&array)[N]))[N];
|
37
|
+
#endif
|
38
|
+
|
39
|
+
#define arraysize(array) (sizeof(ArraySizeHelper(array)))
|
40
|
+
|
41
|
+
// ARRAYSIZE performs essentially the same calculation as arraysize,
|
42
|
+
// but can be used on anonymous types or types defined inside
|
43
|
+
// functions. It's less safe than arraysize as it accepts some
|
44
|
+
// (although not all) pointers. Therefore, you should use arraysize
|
45
|
+
// whenever possible.
|
46
|
+
//
|
47
|
+
// The expression ARRAYSIZE(a) is a compile-time constant of type
|
48
|
+
// size_t.
|
49
|
+
//
|
50
|
+
// ARRAYSIZE catches a few type errors. If you see a compiler error
|
51
|
+
//
|
52
|
+
// "warning: division by zero in ..."
|
53
|
+
//
|
54
|
+
// when using ARRAYSIZE, you are (wrongfully) giving it a pointer.
|
55
|
+
// You should only use ARRAYSIZE on statically allocated arrays.
|
56
|
+
//
|
57
|
+
// The following comments are on the implementation details, and can
|
58
|
+
// be ignored by the users.
|
59
|
+
//
|
60
|
+
// ARRAYSIZE(arr) works by inspecting sizeof(arr) (the # of bytes in
|
61
|
+
// the array) and sizeof(*(arr)) (the # of bytes in one array
|
62
|
+
// element). If the former is divisible by the latter, perhaps arr is
|
63
|
+
// indeed an array, in which case the division result is the # of
|
64
|
+
// elements in the array. Otherwise, arr cannot possibly be an array,
|
65
|
+
// and we generate a compiler error to prevent the code from
|
66
|
+
// compiling.
|
67
|
+
//
|
68
|
+
// Since the size of bool is implementation-defined, we need to cast
|
69
|
+
// !(sizeof(a) & sizeof(*(a))) to size_t in order to ensure the final
|
70
|
+
// result has type size_t.
|
71
|
+
//
|
72
|
+
// This macro is not perfect as it wrongfully accepts certain
|
73
|
+
// pointers, namely where the pointer size is divisible by the pointee
|
74
|
+
// size. Since all our code has to go through a 32-bit compiler,
|
75
|
+
// where a pointer is 4 bytes, this means all pointers to a type whose
|
76
|
+
// size is 3 or greater than 4 will be (righteously) rejected.
|
77
|
+
//
|
78
|
+
// Starting with Visual C++ 2005, WinNT.h includes ARRAYSIZE.
|
79
|
+
#define ARRAYSIZE_UNSAFE(a) \
|
80
|
+
((sizeof(a) / sizeof(*(a))) / \
|
81
|
+
static_cast<size_t>(!(sizeof(a) % sizeof(*(a)))))
|
82
|
+
|
83
|
+
// A macro to disallow the evil copy constructor and operator= functions
|
84
|
+
// This should be used in the private: declarations for a class
|
85
|
+
#define DISALLOW_EVIL_CONSTRUCTORS(TypeName) \
|
86
|
+
TypeName(const TypeName&); \
|
87
|
+
void operator=(const TypeName&)
|
88
|
+
|
89
|
+
#endif // BASE_BASICTYPES_H__
|
@@ -0,0 +1,50 @@
|
|
1
|
+
require 'mkmf'
|
2
|
+
|
3
|
+
extension_name = 'uri_parser'
|
4
|
+
|
5
|
+
$CFLAGS << ' -Wno-deprecated -g '
|
6
|
+
|
7
|
+
if RUBY_PLATFORM =~ /linux|darwin/
|
8
|
+
$libs << ' -lstdc++'
|
9
|
+
$libs << ' -licuuc'
|
10
|
+
else
|
11
|
+
abort <<END_BAD_PLATFORM
|
12
|
+
+----------------------------------------------------------------------------+
|
13
|
+
| This gem is for use only on Linux, and Mac OSX |
|
14
|
+
+----------------------------------------------------------------------------+
|
15
|
+
END_BAD_PLATFORM
|
16
|
+
end
|
17
|
+
|
18
|
+
# Check for compiler. Extract first word so ENV['CC'] can be a program name with arguments.
|
19
|
+
cc = (ENV["CC"] or Config::CONFIG["CC"] or "gcc").split(' ').first
|
20
|
+
unless find_executable(cc)
|
21
|
+
failure "No C compiler found in ${ENV['PATH']}. See mkmf.log for details."
|
22
|
+
end
|
23
|
+
RbConfig::MAKEFILE_CONFIG['CC'] = cc
|
24
|
+
|
25
|
+
def failure s
|
26
|
+
Logging::message s
|
27
|
+
message s+"\n"
|
28
|
+
exit(1)
|
29
|
+
end
|
30
|
+
|
31
|
+
def find_library_or_fail(lib,func)
|
32
|
+
unless have_library(lib, func)
|
33
|
+
failure "Cannot find required library %s (have you installed icu?)" % lib
|
34
|
+
end
|
35
|
+
end
|
36
|
+
|
37
|
+
def find_header_or_fail hdr
|
38
|
+
unless have_header(hdr)
|
39
|
+
failure "Cannot find required header %s (have you installed icu?)" % hdr
|
40
|
+
end
|
41
|
+
end
|
42
|
+
|
43
|
+
find_header_or_fail("unicode/ucnv.h")
|
44
|
+
find_header_or_fail("unicode/ucnv_cb.h")
|
45
|
+
find_header_or_fail("unicode/uidna.h")
|
46
|
+
|
47
|
+
dir_config(extension_name)
|
48
|
+
|
49
|
+
create_header
|
50
|
+
create_makefile(extension_name)
|
@@ -0,0 +1,322 @@
|
|
1
|
+
#ifndef BASE_SCOPED_PTR_H
|
2
|
+
#define BASE_SCOPED_PTR_H
|
3
|
+
|
4
|
+
// (C) Copyright Greg Colvin and Beman Dawes 1998, 1999.
|
5
|
+
// Copyright (c) 2001, 2002 Peter Dimov
|
6
|
+
//
|
7
|
+
// Permission to copy, use, modify, sell and distribute this software
|
8
|
+
// is granted provided this copyright notice appears in all copies.
|
9
|
+
// This software is provided "as is" without express or implied
|
10
|
+
// warranty, and with no claim as to its suitability for any purpose.
|
11
|
+
//
|
12
|
+
// See http://www.boost.org/libs/smart_ptr/scoped_ptr.htm for documentation.
|
13
|
+
//
|
14
|
+
|
15
|
+
// scoped_ptr mimics a built-in pointer except that it guarantees deletion
|
16
|
+
// of the object pointed to, either on destruction of the scoped_ptr or via
|
17
|
+
// an explicit reset(). scoped_ptr is a simple solution for simple needs;
|
18
|
+
// use shared_ptr or std::auto_ptr if your needs are more complex.
|
19
|
+
|
20
|
+
// *** NOTE ***
|
21
|
+
// If your scoped_ptr is a class member of class FOO pointing to a
|
22
|
+
// forward declared type BAR (as shown below), then you MUST use a non-inlined
|
23
|
+
// version of the destructor. The destructor of a scoped_ptr (called from
|
24
|
+
// FOO's destructor) must have a complete definition of BAR in order to
|
25
|
+
// destroy it. Example:
|
26
|
+
//
|
27
|
+
// -- foo.h --
|
28
|
+
// class BAR;
|
29
|
+
//
|
30
|
+
// class FOO {
|
31
|
+
// public:
|
32
|
+
// FOO();
|
33
|
+
// ~FOO(); // Required for sources that instantiate class FOO to compile!
|
34
|
+
//
|
35
|
+
// private:
|
36
|
+
// scoped_ptr<BAR> bar_;
|
37
|
+
// };
|
38
|
+
//
|
39
|
+
// -- foo.cc --
|
40
|
+
// #include "foo.h"
|
41
|
+
// FOO::~FOO() {} // Empty, but must be non-inlined to FOO's class definition.
|
42
|
+
|
43
|
+
#include <cstddef> // for std::ptrdiff_t
|
44
|
+
#include <assert.h> // for assert
|
45
|
+
#include <stdlib.h> // for free() decl
|
46
|
+
|
47
|
+
template <typename T>
|
48
|
+
class scoped_ptr {
|
49
|
+
private:
|
50
|
+
|
51
|
+
T* ptr;
|
52
|
+
|
53
|
+
scoped_ptr(scoped_ptr const &);
|
54
|
+
scoped_ptr & operator=(scoped_ptr const &);
|
55
|
+
|
56
|
+
public:
|
57
|
+
|
58
|
+
typedef T element_type;
|
59
|
+
|
60
|
+
explicit scoped_ptr(T* p = 0): ptr(p) {}
|
61
|
+
|
62
|
+
~scoped_ptr() {
|
63
|
+
typedef char type_must_be_complete[sizeof(T)];
|
64
|
+
delete ptr;
|
65
|
+
}
|
66
|
+
|
67
|
+
void reset(T* p = 0) {
|
68
|
+
typedef char type_must_be_complete[sizeof(T)];
|
69
|
+
|
70
|
+
if (ptr != p) {
|
71
|
+
delete ptr;
|
72
|
+
ptr = p;
|
73
|
+
}
|
74
|
+
}
|
75
|
+
|
76
|
+
T& operator*() const {
|
77
|
+
assert(ptr != 0);
|
78
|
+
return *ptr;
|
79
|
+
}
|
80
|
+
|
81
|
+
T* operator->() const {
|
82
|
+
assert(ptr != 0);
|
83
|
+
return ptr;
|
84
|
+
}
|
85
|
+
|
86
|
+
bool operator==(T* p) const {
|
87
|
+
return ptr == p;
|
88
|
+
}
|
89
|
+
|
90
|
+
bool operator!=(T* p) const {
|
91
|
+
return ptr != p;
|
92
|
+
}
|
93
|
+
|
94
|
+
T* get() const {
|
95
|
+
return ptr;
|
96
|
+
}
|
97
|
+
|
98
|
+
void swap(scoped_ptr & b) {
|
99
|
+
T* tmp = b.ptr;
|
100
|
+
b.ptr = ptr;
|
101
|
+
ptr = tmp;
|
102
|
+
}
|
103
|
+
|
104
|
+
T* release() {
|
105
|
+
T* tmp = ptr;
|
106
|
+
ptr = 0;
|
107
|
+
return tmp;
|
108
|
+
}
|
109
|
+
|
110
|
+
private:
|
111
|
+
|
112
|
+
// no reason to use these: each scoped_ptr should have its own object
|
113
|
+
template <typename U> bool operator==(scoped_ptr<U> const& p) const;
|
114
|
+
template <typename U> bool operator!=(scoped_ptr<U> const& p) const;
|
115
|
+
};
|
116
|
+
|
117
|
+
template<typename T> inline
|
118
|
+
void swap(scoped_ptr<T>& a, scoped_ptr<T>& b) {
|
119
|
+
a.swap(b);
|
120
|
+
}
|
121
|
+
|
122
|
+
template<typename T> inline
|
123
|
+
bool operator==(T* p, const scoped_ptr<T>& b) {
|
124
|
+
return p == b.get();
|
125
|
+
}
|
126
|
+
|
127
|
+
template<typename T> inline
|
128
|
+
bool operator!=(T* p, const scoped_ptr<T>& b) {
|
129
|
+
return p != b.get();
|
130
|
+
}
|
131
|
+
|
132
|
+
// scoped_array extends scoped_ptr to arrays. Deletion of the array pointed to
|
133
|
+
// is guaranteed, either on destruction of the scoped_array or via an explicit
|
134
|
+
// reset(). Use shared_array or std::vector if your needs are more complex.
|
135
|
+
|
136
|
+
template<typename T>
|
137
|
+
class scoped_array {
|
138
|
+
private:
|
139
|
+
|
140
|
+
T* ptr;
|
141
|
+
|
142
|
+
scoped_array(scoped_array const &);
|
143
|
+
scoped_array & operator=(scoped_array const &);
|
144
|
+
|
145
|
+
public:
|
146
|
+
|
147
|
+
typedef T element_type;
|
148
|
+
|
149
|
+
explicit scoped_array(T* p = 0) : ptr(p) {}
|
150
|
+
|
151
|
+
~scoped_array() {
|
152
|
+
typedef char type_must_be_complete[sizeof(T)];
|
153
|
+
delete[] ptr;
|
154
|
+
}
|
155
|
+
|
156
|
+
void reset(T* p = 0) {
|
157
|
+
typedef char type_must_be_complete[sizeof(T)];
|
158
|
+
|
159
|
+
if (ptr != p) {
|
160
|
+
delete [] ptr;
|
161
|
+
ptr = p;
|
162
|
+
}
|
163
|
+
}
|
164
|
+
|
165
|
+
T& operator[](std::ptrdiff_t i) const {
|
166
|
+
assert(ptr != 0);
|
167
|
+
assert(i >= 0);
|
168
|
+
return ptr[i];
|
169
|
+
}
|
170
|
+
|
171
|
+
bool operator==(T* p) const {
|
172
|
+
return ptr == p;
|
173
|
+
}
|
174
|
+
|
175
|
+
bool operator!=(T* p) const {
|
176
|
+
return ptr != p;
|
177
|
+
}
|
178
|
+
|
179
|
+
T* get() const {
|
180
|
+
return ptr;
|
181
|
+
}
|
182
|
+
|
183
|
+
void swap(scoped_array & b) {
|
184
|
+
T* tmp = b.ptr;
|
185
|
+
b.ptr = ptr;
|
186
|
+
ptr = tmp;
|
187
|
+
}
|
188
|
+
|
189
|
+
T* release() {
|
190
|
+
T* tmp = ptr;
|
191
|
+
ptr = 0;
|
192
|
+
return tmp;
|
193
|
+
}
|
194
|
+
|
195
|
+
private:
|
196
|
+
|
197
|
+
// no reason to use these: each scoped_array should have its own object
|
198
|
+
template <typename U> bool operator==(scoped_array<U> const& p) const;
|
199
|
+
template <typename U> bool operator!=(scoped_array<U> const& p) const;
|
200
|
+
};
|
201
|
+
|
202
|
+
template<class T> inline
|
203
|
+
void swap(::scoped_array<T>& a, ::scoped_array<T>& b) {
|
204
|
+
a.swap(b);
|
205
|
+
}
|
206
|
+
|
207
|
+
template<typename T> inline
|
208
|
+
bool operator==(T* p, const ::scoped_array<T>& b) {
|
209
|
+
return p == b.get();
|
210
|
+
}
|
211
|
+
|
212
|
+
template<typename T> inline
|
213
|
+
bool operator!=(T* p, const ::scoped_array<T>& b) {
|
214
|
+
return p != b.get();
|
215
|
+
}
|
216
|
+
|
217
|
+
|
218
|
+
// This class wraps the c library function free() in a class that can be
|
219
|
+
// passed as a template argument to scoped_ptr_malloc below.
|
220
|
+
class ScopedPtrMallocFree {
|
221
|
+
public:
|
222
|
+
inline void operator()(void* x) const {
|
223
|
+
free(x);
|
224
|
+
}
|
225
|
+
};
|
226
|
+
|
227
|
+
// scoped_ptr_malloc<> is similar to scoped_ptr<>, but it accepts a
|
228
|
+
// second template argument, the functor used to free the object.
|
229
|
+
|
230
|
+
template<typename T, typename FreeProc = ScopedPtrMallocFree>
|
231
|
+
class scoped_ptr_malloc {
|
232
|
+
private:
|
233
|
+
|
234
|
+
T* ptr;
|
235
|
+
|
236
|
+
scoped_ptr_malloc(scoped_ptr_malloc const &);
|
237
|
+
scoped_ptr_malloc & operator=(scoped_ptr_malloc const &);
|
238
|
+
|
239
|
+
public:
|
240
|
+
|
241
|
+
typedef T element_type;
|
242
|
+
|
243
|
+
explicit scoped_ptr_malloc(T* p = 0): ptr(p) {}
|
244
|
+
|
245
|
+
~scoped_ptr_malloc() {
|
246
|
+
typedef char type_must_be_complete[sizeof(T)];
|
247
|
+
free_((void*) ptr);
|
248
|
+
}
|
249
|
+
|
250
|
+
void reset(T* p = 0) {
|
251
|
+
typedef char type_must_be_complete[sizeof(T)];
|
252
|
+
|
253
|
+
if (ptr != p) {
|
254
|
+
free_((void*) ptr);
|
255
|
+
ptr = p;
|
256
|
+
}
|
257
|
+
}
|
258
|
+
|
259
|
+
T& operator*() const {
|
260
|
+
assert(ptr != 0);
|
261
|
+
return *ptr;
|
262
|
+
}
|
263
|
+
|
264
|
+
T* operator->() const {
|
265
|
+
assert(ptr != 0);
|
266
|
+
return ptr;
|
267
|
+
}
|
268
|
+
|
269
|
+
bool operator==(T* p) const {
|
270
|
+
return ptr == p;
|
271
|
+
}
|
272
|
+
|
273
|
+
bool operator!=(T* p) const {
|
274
|
+
return ptr != p;
|
275
|
+
}
|
276
|
+
|
277
|
+
T* get() const {
|
278
|
+
return ptr;
|
279
|
+
}
|
280
|
+
|
281
|
+
void swap(scoped_ptr_malloc & b) {
|
282
|
+
T* tmp = b.ptr;
|
283
|
+
b.ptr = ptr;
|
284
|
+
ptr = tmp;
|
285
|
+
}
|
286
|
+
|
287
|
+
T* release() {
|
288
|
+
T* tmp = ptr;
|
289
|
+
ptr = 0;
|
290
|
+
return tmp;
|
291
|
+
}
|
292
|
+
|
293
|
+
private:
|
294
|
+
|
295
|
+
// no reason to use these: each scoped_ptr_malloc should have its own object
|
296
|
+
template <typename U, typename GP>
|
297
|
+
bool operator==(scoped_ptr_malloc<U, GP> const& p) const;
|
298
|
+
template <typename U, typename GP>
|
299
|
+
bool operator!=(scoped_ptr_malloc<U, GP> const& p) const;
|
300
|
+
|
301
|
+
static FreeProc const free_;
|
302
|
+
};
|
303
|
+
|
304
|
+
template<typename T, typename FP>
|
305
|
+
FP const scoped_ptr_malloc<T,FP>::free_ = FP();
|
306
|
+
|
307
|
+
template<typename T, typename FP> inline
|
308
|
+
void swap(scoped_ptr_malloc<T,FP>& a, scoped_ptr_malloc<T,FP>& b) {
|
309
|
+
a.swap(b);
|
310
|
+
}
|
311
|
+
|
312
|
+
template<typename T, typename FP> inline
|
313
|
+
bool operator==(T* p, const scoped_ptr_malloc<T,FP>& b) {
|
314
|
+
return p == b.get();
|
315
|
+
}
|
316
|
+
|
317
|
+
template<typename T, typename FP> inline
|
318
|
+
bool operator!=(T* p, const scoped_ptr_malloc<T,FP>& b) {
|
319
|
+
return p != b.get();
|
320
|
+
}
|
321
|
+
|
322
|
+
#endif // #ifndef BASE_SCOPED_PTR_H
|