re2 2.0.0 → 2.1.0
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +4 -4
- data/LICENSE-DEPENDENCIES.txt +237 -0
- data/LICENSE.txt +1 -1
- data/README.md +47 -16
- data/ext/re2/re2.cc +74 -23
- data/lib/re2/version.rb +1 -1
- data/re2.gemspec +5 -4
- data/spec/kernel_spec.rb +3 -3
- data/spec/re2/match_data_spec.rb +24 -0
- data/spec/re2/regexp_spec.rb +6 -0
- data/spec/re2/scanner_spec.rb +76 -22
- data/spec/re2/set_spec.rb +41 -1
- data/spec/re2/string_spec.rb +7 -3
- data/spec/re2_spec.rb +104 -10
- data/spec/spec_helper.rb +10 -0
- metadata +6 -4
checksums.yaml
CHANGED
|
@@ -1,7 +1,7 @@
|
|
|
1
1
|
---
|
|
2
2
|
SHA256:
|
|
3
|
-
metadata.gz:
|
|
4
|
-
data.tar.gz:
|
|
3
|
+
metadata.gz: 37e251e652dcc758e2d18f0734cf6fe9af62fa23b3ed127175c27cc439e80451
|
|
4
|
+
data.tar.gz: 536b7cf0e815c480638a381dfca47677149d0c8d2f711105db34439e0c2d9e09
|
|
5
5
|
SHA512:
|
|
6
|
-
metadata.gz:
|
|
7
|
-
data.tar.gz:
|
|
6
|
+
metadata.gz: 711ba1845bbc96f4a2120ec595f77316d08803bcddb995758871232811b23fa5a2c7cda1949d5fee2f86279ecba694001c530efe7bc29767aa7b5371fe4b5613
|
|
7
|
+
data.tar.gz: f13f0c6d838775667d3d0ed507397e6dadad55c352e8968d7af8295c5dd7e87f6f987c813042c7edee453ded849c9dbd9cc26d440966efe48eda698d29db616b
|
|
@@ -0,0 +1,237 @@
|
|
|
1
|
+
# Vendored Dependency Licenses
|
|
2
|
+
|
|
3
|
+
The library re2 (which lives at https://github.com/mudge/re2) may include the source code for RE2 (which lives at https://github.com/google/re2) and Abseil (which lives at https://abseil.io).
|
|
4
|
+
|
|
5
|
+
The license terms shipped with RE2 are included here for your convenience:
|
|
6
|
+
|
|
7
|
+
// Copyright (c) 2009 The RE2 Authors. All rights reserved.
|
|
8
|
+
//
|
|
9
|
+
// Redistribution and use in source and binary forms, with or without
|
|
10
|
+
// modification, are permitted provided that the following conditions are
|
|
11
|
+
// met:
|
|
12
|
+
//
|
|
13
|
+
// * Redistributions of source code must retain the above copyright
|
|
14
|
+
// notice, this list of conditions and the following disclaimer.
|
|
15
|
+
// * Redistributions in binary form must reproduce the above
|
|
16
|
+
// copyright notice, this list of conditions and the following disclaimer
|
|
17
|
+
// in the documentation and/or other materials provided with the
|
|
18
|
+
// distribution.
|
|
19
|
+
// * Neither the name of Google Inc. nor the names of its
|
|
20
|
+
// contributors may be used to endorse or promote products derived from
|
|
21
|
+
// this software without specific prior written permission.
|
|
22
|
+
//
|
|
23
|
+
// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
|
|
24
|
+
// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
|
|
25
|
+
// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
|
|
26
|
+
// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
|
|
27
|
+
// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
|
|
28
|
+
// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
|
|
29
|
+
// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
|
|
30
|
+
// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
|
|
31
|
+
// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
|
|
32
|
+
// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
|
|
33
|
+
// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
|
|
34
|
+
|
|
35
|
+
The license terms shipped with Abseil are included here for your convenience:
|
|
36
|
+
|
|
37
|
+
Apache License
|
|
38
|
+
Version 2.0, January 2004
|
|
39
|
+
https://www.apache.org/licenses/
|
|
40
|
+
|
|
41
|
+
TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
|
|
42
|
+
|
|
43
|
+
1. Definitions.
|
|
44
|
+
|
|
45
|
+
"License" shall mean the terms and conditions for use, reproduction,
|
|
46
|
+
and distribution as defined by Sections 1 through 9 of this document.
|
|
47
|
+
|
|
48
|
+
"Licensor" shall mean the copyright owner or entity authorized by
|
|
49
|
+
the copyright owner that is granting the License.
|
|
50
|
+
|
|
51
|
+
"Legal Entity" shall mean the union of the acting entity and all
|
|
52
|
+
other entities that control, are controlled by, or are under common
|
|
53
|
+
control with that entity. For the purposes of this definition,
|
|
54
|
+
"control" means (i) the power, direct or indirect, to cause the
|
|
55
|
+
direction or management of such entity, whether by contract or
|
|
56
|
+
otherwise, or (ii) ownership of fifty percent (50%) or more of the
|
|
57
|
+
outstanding shares, or (iii) beneficial ownership of such entity.
|
|
58
|
+
|
|
59
|
+
"You" (or "Your") shall mean an individual or Legal Entity
|
|
60
|
+
exercising permissions granted by this License.
|
|
61
|
+
|
|
62
|
+
"Source" form shall mean the preferred form for making modifications,
|
|
63
|
+
including but not limited to software source code, documentation
|
|
64
|
+
source, and configuration files.
|
|
65
|
+
|
|
66
|
+
"Object" form shall mean any form resulting from mechanical
|
|
67
|
+
transformation or translation of a Source form, including but
|
|
68
|
+
not limited to compiled object code, generated documentation,
|
|
69
|
+
and conversions to other media types.
|
|
70
|
+
|
|
71
|
+
"Work" shall mean the work of authorship, whether in Source or
|
|
72
|
+
Object form, made available under the License, as indicated by a
|
|
73
|
+
copyright notice that is included in or attached to the work
|
|
74
|
+
(an example is provided in the Appendix below).
|
|
75
|
+
|
|
76
|
+
"Derivative Works" shall mean any work, whether in Source or Object
|
|
77
|
+
form, that is based on (or derived from) the Work and for which the
|
|
78
|
+
editorial revisions, annotations, elaborations, or other modifications
|
|
79
|
+
represent, as a whole, an original work of authorship. For the purposes
|
|
80
|
+
of this License, Derivative Works shall not include works that remain
|
|
81
|
+
separable from, or merely link (or bind by name) to the interfaces of,
|
|
82
|
+
the Work and Derivative Works thereof.
|
|
83
|
+
|
|
84
|
+
"Contribution" shall mean any work of authorship, including
|
|
85
|
+
the original version of the Work and any modifications or additions
|
|
86
|
+
to that Work or Derivative Works thereof, that is intentionally
|
|
87
|
+
submitted to Licensor for inclusion in the Work by the copyright owner
|
|
88
|
+
or by an individual or Legal Entity authorized to submit on behalf of
|
|
89
|
+
the copyright owner. For the purposes of this definition, "submitted"
|
|
90
|
+
means any form of electronic, verbal, or written communication sent
|
|
91
|
+
to the Licensor or its representatives, including but not limited to
|
|
92
|
+
communication on electronic mailing lists, source code control systems,
|
|
93
|
+
and issue tracking systems that are managed by, or on behalf of, the
|
|
94
|
+
Licensor for the purpose of discussing and improving the Work, but
|
|
95
|
+
excluding communication that is conspicuously marked or otherwise
|
|
96
|
+
designated in writing by the copyright owner as "Not a Contribution."
|
|
97
|
+
|
|
98
|
+
"Contributor" shall mean Licensor and any individual or Legal Entity
|
|
99
|
+
on behalf of whom a Contribution has been received by Licensor and
|
|
100
|
+
subsequently incorporated within the Work.
|
|
101
|
+
|
|
102
|
+
2. Grant of Copyright License. Subject to the terms and conditions of
|
|
103
|
+
this License, each Contributor hereby grants to You a perpetual,
|
|
104
|
+
worldwide, non-exclusive, no-charge, royalty-free, irrevocable
|
|
105
|
+
copyright license to reproduce, prepare Derivative Works of,
|
|
106
|
+
publicly display, publicly perform, sublicense, and distribute the
|
|
107
|
+
Work and such Derivative Works in Source or Object form.
|
|
108
|
+
|
|
109
|
+
3. Grant of Patent License. Subject to the terms and conditions of
|
|
110
|
+
this License, each Contributor hereby grants to You a perpetual,
|
|
111
|
+
worldwide, non-exclusive, no-charge, royalty-free, irrevocable
|
|
112
|
+
(except as stated in this section) patent license to make, have made,
|
|
113
|
+
use, offer to sell, sell, import, and otherwise transfer the Work,
|
|
114
|
+
where such license applies only to those patent claims licensable
|
|
115
|
+
by such Contributor that are necessarily infringed by their
|
|
116
|
+
Contribution(s) alone or by combination of their Contribution(s)
|
|
117
|
+
with the Work to which such Contribution(s) was submitted. If You
|
|
118
|
+
institute patent litigation against any entity (including a
|
|
119
|
+
cross-claim or counterclaim in a lawsuit) alleging that the Work
|
|
120
|
+
or a Contribution incorporated within the Work constitutes direct
|
|
121
|
+
or contributory patent infringement, then any patent licenses
|
|
122
|
+
granted to You under this License for that Work shall terminate
|
|
123
|
+
as of the date such litigation is filed.
|
|
124
|
+
|
|
125
|
+
4. Redistribution. You may reproduce and distribute copies of the
|
|
126
|
+
Work or Derivative Works thereof in any medium, with or without
|
|
127
|
+
modifications, and in Source or Object form, provided that You
|
|
128
|
+
meet the following conditions:
|
|
129
|
+
|
|
130
|
+
(a) You must give any other recipients of the Work or
|
|
131
|
+
Derivative Works a copy of this License; and
|
|
132
|
+
|
|
133
|
+
(b) You must cause any modified files to carry prominent notices
|
|
134
|
+
stating that You changed the files; and
|
|
135
|
+
|
|
136
|
+
(c) You must retain, in the Source form of any Derivative Works
|
|
137
|
+
that You distribute, all copyright, patent, trademark, and
|
|
138
|
+
attribution notices from the Source form of the Work,
|
|
139
|
+
excluding those notices that do not pertain to any part of
|
|
140
|
+
the Derivative Works; and
|
|
141
|
+
|
|
142
|
+
(d) If the Work includes a "NOTICE" text file as part of its
|
|
143
|
+
distribution, then any Derivative Works that You distribute must
|
|
144
|
+
include a readable copy of the attribution notices contained
|
|
145
|
+
within such NOTICE file, excluding those notices that do not
|
|
146
|
+
pertain to any part of the Derivative Works, in at least one
|
|
147
|
+
of the following places: within a NOTICE text file distributed
|
|
148
|
+
as part of the Derivative Works; within the Source form or
|
|
149
|
+
documentation, if provided along with the Derivative Works; or,
|
|
150
|
+
within a display generated by the Derivative Works, if and
|
|
151
|
+
wherever such third-party notices normally appear. The contents
|
|
152
|
+
of the NOTICE file are for informational purposes only and
|
|
153
|
+
do not modify the License. You may add Your own attribution
|
|
154
|
+
notices within Derivative Works that You distribute, alongside
|
|
155
|
+
or as an addendum to the NOTICE text from the Work, provided
|
|
156
|
+
that such additional attribution notices cannot be construed
|
|
157
|
+
as modifying the License.
|
|
158
|
+
|
|
159
|
+
You may add Your own copyright statement to Your modifications and
|
|
160
|
+
may provide additional or different license terms and conditions
|
|
161
|
+
for use, reproduction, or distribution of Your modifications, or
|
|
162
|
+
for any such Derivative Works as a whole, provided Your use,
|
|
163
|
+
reproduction, and distribution of the Work otherwise complies with
|
|
164
|
+
the conditions stated in this License.
|
|
165
|
+
|
|
166
|
+
5. Submission of Contributions. Unless You explicitly state otherwise,
|
|
167
|
+
any Contribution intentionally submitted for inclusion in the Work
|
|
168
|
+
by You to the Licensor shall be under the terms and conditions of
|
|
169
|
+
this License, without any additional terms or conditions.
|
|
170
|
+
Notwithstanding the above, nothing herein shall supersede or modify
|
|
171
|
+
the terms of any separate license agreement you may have executed
|
|
172
|
+
with Licensor regarding such Contributions.
|
|
173
|
+
|
|
174
|
+
6. Trademarks. This License does not grant permission to use the trade
|
|
175
|
+
names, trademarks, service marks, or product names of the Licensor,
|
|
176
|
+
except as required for reasonable and customary use in describing the
|
|
177
|
+
origin of the Work and reproducing the content of the NOTICE file.
|
|
178
|
+
|
|
179
|
+
7. Disclaimer of Warranty. Unless required by applicable law or
|
|
180
|
+
agreed to in writing, Licensor provides the Work (and each
|
|
181
|
+
Contributor provides its Contributions) on an "AS IS" BASIS,
|
|
182
|
+
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
|
|
183
|
+
implied, including, without limitation, any warranties or conditions
|
|
184
|
+
of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
|
|
185
|
+
PARTICULAR PURPOSE. You are solely responsible for determining the
|
|
186
|
+
appropriateness of using or redistributing the Work and assume any
|
|
187
|
+
risks associated with Your exercise of permissions under this License.
|
|
188
|
+
|
|
189
|
+
8. Limitation of Liability. In no event and under no legal theory,
|
|
190
|
+
whether in tort (including negligence), contract, or otherwise,
|
|
191
|
+
unless required by applicable law (such as deliberate and grossly
|
|
192
|
+
negligent acts) or agreed to in writing, shall any Contributor be
|
|
193
|
+
liable to You for damages, including any direct, indirect, special,
|
|
194
|
+
incidental, or consequential damages of any character arising as a
|
|
195
|
+
result of this License or out of the use or inability to use the
|
|
196
|
+
Work (including but not limited to damages for loss of goodwill,
|
|
197
|
+
work stoppage, computer failure or malfunction, or any and all
|
|
198
|
+
other commercial damages or losses), even if such Contributor
|
|
199
|
+
has been advised of the possibility of such damages.
|
|
200
|
+
|
|
201
|
+
9. Accepting Warranty or Additional Liability. While redistributing
|
|
202
|
+
the Work or Derivative Works thereof, You may choose to offer,
|
|
203
|
+
and charge a fee for, acceptance of support, warranty, indemnity,
|
|
204
|
+
or other liability obligations and/or rights consistent with this
|
|
205
|
+
License. However, in accepting such obligations, You may act only
|
|
206
|
+
on Your own behalf and on Your sole responsibility, not on behalf
|
|
207
|
+
of any other Contributor, and only if You agree to indemnify,
|
|
208
|
+
defend, and hold each Contributor harmless for any liability
|
|
209
|
+
incurred by, or claims asserted against, such Contributor by reason
|
|
210
|
+
of your accepting any such warranty or additional liability.
|
|
211
|
+
|
|
212
|
+
END OF TERMS AND CONDITIONS
|
|
213
|
+
|
|
214
|
+
APPENDIX: How to apply the Apache License to your work.
|
|
215
|
+
|
|
216
|
+
To apply the Apache License to your work, attach the following
|
|
217
|
+
boilerplate notice, with the fields enclosed by brackets "[]"
|
|
218
|
+
replaced with your own identifying information. (Don't include
|
|
219
|
+
the brackets!) The text should be enclosed in the appropriate
|
|
220
|
+
comment syntax for the file format. We also recommend that a
|
|
221
|
+
file or class name and description of purpose be included on the
|
|
222
|
+
same "printed page" as the copyright notice for easier
|
|
223
|
+
identification within third-party archives.
|
|
224
|
+
|
|
225
|
+
Copyright [yyyy] [name of copyright owner]
|
|
226
|
+
|
|
227
|
+
Licensed under the Apache License, Version 2.0 (the "License");
|
|
228
|
+
you may not use this file except in compliance with the License.
|
|
229
|
+
You may obtain a copy of the License at
|
|
230
|
+
|
|
231
|
+
https://www.apache.org/licenses/LICENSE-2.0
|
|
232
|
+
|
|
233
|
+
Unless required by applicable law or agreed to in writing, software
|
|
234
|
+
distributed under the License is distributed on an "AS IS" BASIS,
|
|
235
|
+
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
|
236
|
+
See the License for the specific language governing permissions and
|
|
237
|
+
limitations under the License.
|
data/LICENSE.txt
CHANGED
data/README.md
CHANGED
|
@@ -1,18 +1,19 @@
|
|
|
1
1
|
re2 [](https://github.com/mudge/re2/actions)
|
|
2
2
|
===
|
|
3
3
|
|
|
4
|
-
|
|
5
|
-
|
|
4
|
+
Ruby bindings to [RE2][], a "fast, safe, thread-friendly alternative to
|
|
5
|
+
backtracking regular expression engines like those used in PCRE, Perl, and
|
|
6
|
+
Python".
|
|
6
7
|
|
|
7
|
-
**Current version:** 2.
|
|
8
|
+
**Current version:** 2.1.0
|
|
8
9
|
**Supported Ruby versions:** 2.6, 2.7, 3.0, 3.1, 3.2
|
|
9
|
-
**Bundled
|
|
10
|
-
**Supported
|
|
10
|
+
**Bundled RE2 version:** libre2.11 (2023-09-01)
|
|
11
|
+
**Supported RE2 versions:** libre2.0 (< 2020-03-02), libre2.1 (2020-03-02), libre2.6 (2020-03-03), libre2.7 (2020-05-01), libre2.8 (2020-07-06), libre2.9 (2020-11-01), libre2.10 (2022-12-01), libre2.11 (2023-07-01)
|
|
11
12
|
|
|
12
13
|
Installation
|
|
13
14
|
------------
|
|
14
15
|
|
|
15
|
-
The gem comes bundled with a version of [
|
|
16
|
+
The gem comes bundled with a version of [RE2][] and will compile itself (and
|
|
16
17
|
any dependencies) on install. As compilation can take a while, precompiled
|
|
17
18
|
native gems are available for Linux, Windows and macOS.
|
|
18
19
|
|
|
@@ -27,10 +28,10 @@ on these platforms:
|
|
|
27
28
|
- `x86_64-darwin`
|
|
28
29
|
- `x86_64-linux` (requires: glibc >= 2.17)
|
|
29
30
|
|
|
30
|
-
If you wish to opt out of using the bundled libraries, you will need
|
|
31
|
+
If you wish to opt out of using the bundled libraries, you will need RE2
|
|
31
32
|
installed as well as a C++ compiler such as [gcc][] (on Debian and Ubuntu, this
|
|
32
|
-
is provided by the [build-essential][] package). If you are using
|
|
33
|
-
recommend installing
|
|
33
|
+
is provided by the [build-essential][] package). If you are using macOS, I
|
|
34
|
+
recommend installing RE2 with [Homebrew][] by running the following:
|
|
34
35
|
|
|
35
36
|
$ brew install re2
|
|
36
37
|
|
|
@@ -38,7 +39,7 @@ If you are using Debian, you can install the [libre2-dev][] package like so:
|
|
|
38
39
|
|
|
39
40
|
$ sudo apt-get install libre2-dev
|
|
40
41
|
|
|
41
|
-
Recent versions of
|
|
42
|
+
Recent versions of RE2 require a compiler with C++14 support such as
|
|
42
43
|
[clang](http://clang.llvm.org/) 3.4 or [gcc](https://gcc.gnu.org/) 5.
|
|
43
44
|
|
|
44
45
|
If you are using a packaged Ruby distribution, make sure you also have the
|
|
@@ -47,7 +48,7 @@ on Debian and Ubuntu.
|
|
|
47
48
|
|
|
48
49
|
You can then install the library via RubyGems with `gem install re2 --platform=ruby --
|
|
49
50
|
--enable-system-libraries` or `gem install re2 --platform=ruby -- --enable-system-libraries
|
|
50
|
-
--with-re2-dir=/path/to/re2/prefix` if
|
|
51
|
+
--with-re2-dir=/path/to/re2/prefix` if RE2 is not installed in any of the
|
|
51
52
|
following default locations:
|
|
52
53
|
|
|
53
54
|
* `/usr/local`
|
|
@@ -66,7 +67,7 @@ Documentation
|
|
|
66
67
|
Full documentation automatically generated from the latest version is
|
|
67
68
|
available at <http://mudge.name/re2/>.
|
|
68
69
|
|
|
69
|
-
Note that
|
|
70
|
+
Note that RE2's regular expression syntax differs from PCRE and Ruby's
|
|
70
71
|
built-in [`Regexp`][Regexp] library, see the [official syntax page][] for more
|
|
71
72
|
details.
|
|
72
73
|
|
|
@@ -192,6 +193,24 @@ end
|
|
|
192
193
|
# My name is Alice and I am 42 years old
|
|
193
194
|
```
|
|
194
195
|
|
|
196
|
+
Encoding
|
|
197
|
+
--------
|
|
198
|
+
|
|
199
|
+
Note RE2 only supports UTF-8 and ISO-8859-1 encoding so strings will be
|
|
200
|
+
returned in UTF-8 by default or ISO-8859-1 if the `:utf8` option for the
|
|
201
|
+
`RE2::Regexp` is set to false (any other encoding's behaviour is undefined).
|
|
202
|
+
|
|
203
|
+
For backward compatibility: re2 won't automatically convert string inputs to
|
|
204
|
+
the right encoding so this is the responsibility of the caller, e.g.
|
|
205
|
+
|
|
206
|
+
```ruby
|
|
207
|
+
# By default, RE2 will process patterns and text as UTF-8
|
|
208
|
+
RE2(non_utf8_pattern.encode("UTF-8")).match(non_utf8_text.encode("UTF-8"))
|
|
209
|
+
|
|
210
|
+
# If the :utf8 option is false, RE2 will process patterns and text as ISO-8859-1
|
|
211
|
+
RE2(non_latin1_pattern.encode("ISO-8859-1"), :utf8 => false).match(non_latin1_text.encode("ISO-8859-1"))
|
|
212
|
+
```
|
|
213
|
+
|
|
195
214
|
Features
|
|
196
215
|
--------
|
|
197
216
|
|
|
@@ -238,9 +257,9 @@ Contributions
|
|
|
238
257
|
* Thanks to [Jason Woods](https://github.com/driskell) who contributed the
|
|
239
258
|
original implementations of `RE2::MatchData#begin` and `RE2::MatchData#end`;
|
|
240
259
|
* Thanks to [Stefano Rivera](https://github.com/stefanor) who first contributed C++11 support;
|
|
241
|
-
* Thanks to [Stan Hu](https://github.com/stanhu) for reporting a bug with empty patterns and `RE2::Regexp#scan`, contributing support for libre2.11 (2023-07-01) and for vendoring
|
|
260
|
+
* Thanks to [Stan Hu](https://github.com/stanhu) for reporting a bug with empty patterns and `RE2::Regexp#scan`, contributing support for libre2.11 (2023-07-01) and for vendoring RE2 and abseil and compiling native gems in 2.0;
|
|
242
261
|
* Thanks to [Sebastian Reitenbach](https://github.com/buzzdeee) for reporting
|
|
243
|
-
the deprecation and removal of the `utf8` encoding option in
|
|
262
|
+
the deprecation and removal of the `utf8` encoding option in RE2;
|
|
244
263
|
* Thanks to [Sergio Medina](https://github.com/serch) for reporting a bug when
|
|
245
264
|
using `RE2::Scanner#scan` with an invalid regular expression;
|
|
246
265
|
* Thanks to [Pritam Baral](https://github.com/pritambaral) for contributed the
|
|
@@ -251,7 +270,19 @@ Contact
|
|
|
251
270
|
|
|
252
271
|
All issues and suggestions should go to [GitHub Issues](https://github.com/mudge/re2/issues).
|
|
253
272
|
|
|
254
|
-
|
|
273
|
+
License
|
|
274
|
+
-------
|
|
275
|
+
|
|
276
|
+
This library is licensed under the BSD 3-Clause License, see `LICENSE.txt`.
|
|
277
|
+
|
|
278
|
+
Dependencies
|
|
279
|
+
------------
|
|
280
|
+
|
|
281
|
+
The source code of [RE2][] is distributed in the `ruby` platform gem. This code is licensed under the BSD 3-Clause License, see `LICENSE-DEPENDENCIES.txt`.
|
|
282
|
+
|
|
283
|
+
The source code of [Abseil][] is distributed in the `ruby` platform gem. This code is licensed under the Apache License 2.0, see `LICENSE-DEPENDENCIES.txt`.
|
|
284
|
+
|
|
285
|
+
[RE2]: https://github.com/google/re2
|
|
255
286
|
[gcc]: http://gcc.gnu.org/
|
|
256
287
|
[ruby-dev]: http://packages.debian.org/ruby-dev
|
|
257
288
|
[build-essential]: http://packages.debian.org/build-essential
|
|
@@ -260,4 +291,4 @@ All issues and suggestions should go to [GitHub Issues](https://github.com/mudge
|
|
|
260
291
|
[Homebrew]: http://mxcl.github.com/homebrew
|
|
261
292
|
[libre2-dev]: http://packages.debian.org/search?keywords=libre2-dev
|
|
262
293
|
[official syntax page]: https://github.com/google/re2/wiki/Syntax
|
|
263
|
-
|
|
294
|
+
[Abseil]: https://abseil.io
|
data/ext/re2/re2.cc
CHANGED
|
@@ -39,18 +39,9 @@ using std::vector;
|
|
|
39
39
|
rb_enc_associate_index(_string, _enc); \
|
|
40
40
|
_string; \
|
|
41
41
|
})
|
|
42
|
-
#define ENCODED_STR_NEW2(str, length, str2) \
|
|
43
|
-
({ \
|
|
44
|
-
VALUE _string = rb_str_new(str, length); \
|
|
45
|
-
int _enc = rb_enc_get_index(str2); \
|
|
46
|
-
rb_enc_associate_index(_string, _enc); \
|
|
47
|
-
_string; \
|
|
48
|
-
})
|
|
49
42
|
#else
|
|
50
43
|
#define ENCODED_STR_NEW(str, length, encoding) \
|
|
51
44
|
rb_str_new((const char *)str, (long)length)
|
|
52
|
-
#define ENCODED_STR_NEW2(str, length, str2) \
|
|
53
|
-
rb_str_new((const char *)str, (long)length)
|
|
54
45
|
#endif
|
|
55
46
|
|
|
56
47
|
#ifdef HAVE_RB_STR_SUBLEN
|
|
@@ -284,6 +275,10 @@ static VALUE re2_scanner_rewind(VALUE self) {
|
|
|
284
275
|
* Scan the given text incrementally for matches, returning an array of
|
|
285
276
|
* matches on each subsequent call. Returns nil if no matches are found.
|
|
286
277
|
*
|
|
278
|
+
* Note RE2 only supports UTF-8 and ISO-8859-1 encoding so strings will be
|
|
279
|
+
* returned in UTF-8 by default or ISO-8859-1 if the :utf8 option for the
|
|
280
|
+
* RE2::Regexp is set to false (any other encoding's behaviour is undefined).
|
|
281
|
+
*
|
|
287
282
|
* @return [Array<String>] the matches.
|
|
288
283
|
* @example
|
|
289
284
|
* s = RE2::Regexp.new('(\w+)').scan("Foo bar baz")
|
|
@@ -312,7 +307,6 @@ static VALUE re2_scanner_scan(VALUE self) {
|
|
|
312
307
|
original_input_size = c->input->size();
|
|
313
308
|
|
|
314
309
|
for (i = 0; i < c->number_of_capturing_groups; i++) {
|
|
315
|
-
matches[i] = "";
|
|
316
310
|
argv[i] = &matches[i];
|
|
317
311
|
args[i] = &argv[i];
|
|
318
312
|
}
|
|
@@ -504,6 +498,10 @@ static VALUE re2_regexp_allocate(VALUE klass) {
|
|
|
504
498
|
/*
|
|
505
499
|
* Returns the array of matches.
|
|
506
500
|
*
|
|
501
|
+
* Note RE2 only supports UTF-8 and ISO-8859-1 encoding so strings will be
|
|
502
|
+
* returned in UTF-8 by default or ISO-8859-1 if the :utf8 option for the
|
|
503
|
+
* RE2::Regexp is set to false (any other encoding's behaviour is undefined).
|
|
504
|
+
*
|
|
507
505
|
* @return [Array<String, nil>] the array of matches
|
|
508
506
|
* @example
|
|
509
507
|
* m = RE2::Regexp.new('(\d+)').match("bob 123")
|
|
@@ -579,6 +577,10 @@ static VALUE re2_matchdata_named_match(const char* name, VALUE self) {
|
|
|
579
577
|
/*
|
|
580
578
|
* Retrieve zero, one or more matches by index or name.
|
|
581
579
|
*
|
|
580
|
+
* Note RE2 only supports UTF-8 and ISO-8859-1 encoding so strings will be
|
|
581
|
+
* returned in UTF-8 by default or ISO-8859-1 if the :utf8 option for the
|
|
582
|
+
* RE2::Regexp is set to false (any other encoding's behaviour is undefined).
|
|
583
|
+
*
|
|
582
584
|
* @return [Array<String, nil>, String, Boolean]
|
|
583
585
|
*
|
|
584
586
|
* @overload [](index)
|
|
@@ -646,6 +648,10 @@ static VALUE re2_matchdata_to_s(VALUE self) {
|
|
|
646
648
|
/*
|
|
647
649
|
* Returns a printable version of the match.
|
|
648
650
|
*
|
|
651
|
+
* Note RE2 only supports UTF-8 and ISO-8859-1 encoding so strings will be
|
|
652
|
+
* returned in UTF-8 by default or ISO-8859-1 if the :utf8 option for the
|
|
653
|
+
* RE2::Regexp is set to false (any other encoding's behaviour is undefined).
|
|
654
|
+
*
|
|
649
655
|
* @return [String] a printable version of the match
|
|
650
656
|
* @example
|
|
651
657
|
* m = RE2::Regexp.new('(\d+)').match("bob 123")
|
|
@@ -690,6 +696,10 @@ static VALUE re2_matchdata_inspect(VALUE self) {
|
|
|
690
696
|
/*
|
|
691
697
|
* Returns the array of submatches for pattern matching.
|
|
692
698
|
*
|
|
699
|
+
* Note RE2 only supports UTF-8 and ISO-8859-1 encoding so strings will be
|
|
700
|
+
* returned in UTF-8 by default or ISO-8859-1 if the :utf8 option for the
|
|
701
|
+
* RE2::Regexp is set to false (any other encoding's behaviour is undefined).
|
|
702
|
+
*
|
|
693
703
|
* @return [Array<String, nil>] the array of submatches
|
|
694
704
|
* @example
|
|
695
705
|
* m = RE2::Regexp.new('(\d+)').match("bob 123")
|
|
@@ -735,6 +745,10 @@ static VALUE re2_matchdata_deconstruct(VALUE self) {
|
|
|
735
745
|
* more keys than there are capturing groups. Given keys will populate the hash in
|
|
736
746
|
* order but an invalid name will cause the hash to be immediately returned.
|
|
737
747
|
*
|
|
748
|
+
* Note RE2 only supports UTF-8 and ISO-8859-1 encoding so strings will be
|
|
749
|
+
* returned in UTF-8 by default or ISO-8859-1 if the :utf8 option for the
|
|
750
|
+
* RE2::Regexp is set to false (any other encoding's behaviour is undefined).
|
|
751
|
+
*
|
|
738
752
|
* @return [Hash] a hash of capturing group names to submatches
|
|
739
753
|
* @param [Array<Symbol>, nil] keys an array of Symbol capturing group names or nil to return all names
|
|
740
754
|
* @example
|
|
@@ -866,6 +880,10 @@ static VALUE re2_regexp_initialize(int argc, VALUE *argv, VALUE self) {
|
|
|
866
880
|
/*
|
|
867
881
|
* Returns a printable version of the regular expression +re2+.
|
|
868
882
|
*
|
|
883
|
+
* Note RE2 only supports UTF-8 and ISO-8859-1 encoding so strings will be
|
|
884
|
+
* returned in UTF-8 by default or ISO-8859-1 if the :utf8 option for the
|
|
885
|
+
* RE2::Regexp is set to false (any other encoding's behaviour is undefined).
|
|
886
|
+
*
|
|
869
887
|
* @return [String] a printable version of the regular expression
|
|
870
888
|
* @example
|
|
871
889
|
* re2 = RE2::Regexp.new("woo?")
|
|
@@ -889,6 +907,10 @@ static VALUE re2_regexp_inspect(VALUE self) {
|
|
|
889
907
|
/*
|
|
890
908
|
* Returns a string version of the regular expression +re2+.
|
|
891
909
|
*
|
|
910
|
+
* Note RE2 only supports UTF-8 and ISO-8859-1 encoding so strings will be
|
|
911
|
+
* returned in UTF-8 by default or ISO-8859-1 if the :utf8 option for the
|
|
912
|
+
* RE2::Regexp is set to false (any other encoding's behaviour is undefined).
|
|
913
|
+
*
|
|
892
914
|
* @return [String] a string version of the regular expression
|
|
893
915
|
* @example
|
|
894
916
|
* re2 = RE2::Regexp.new("woo?")
|
|
@@ -1116,6 +1138,10 @@ static VALUE re2_regexp_error(VALUE self) {
|
|
|
1116
1138
|
* If the RE2 could not be created properly, returns
|
|
1117
1139
|
* the offending portion of the regexp otherwise returns nil.
|
|
1118
1140
|
*
|
|
1141
|
+
* Note RE2 only supports UTF-8 and ISO-8859-1 encoding so strings will be
|
|
1142
|
+
* returned in UTF-8 by default or ISO-8859-1 if the :utf8 option for the
|
|
1143
|
+
* RE2::Regexp is set to false (any other encoding's behaviour is undefined).
|
|
1144
|
+
*
|
|
1119
1145
|
* @return [String, nil] the offending portion of the regexp or nil
|
|
1120
1146
|
*/
|
|
1121
1147
|
static VALUE re2_regexp_error_arg(VALUE self) {
|
|
@@ -1212,6 +1238,10 @@ static VALUE re2_regexp_number_of_capturing_groups(VALUE self) {
|
|
|
1212
1238
|
/*
|
|
1213
1239
|
* Returns a hash of names to capturing indices of groups.
|
|
1214
1240
|
*
|
|
1241
|
+
* Note RE2 only supports UTF-8 and ISO-8859-1 encoding so strings will be
|
|
1242
|
+
* returned in UTF-8 by default or ISO-8859-1 if the :utf8 option for the
|
|
1243
|
+
* RE2::Regexp is set to false (any other encoding's behaviour is undefined).
|
|
1244
|
+
*
|
|
1215
1245
|
* @return [Hash] a hash of names to capturing indices
|
|
1216
1246
|
*/
|
|
1217
1247
|
static VALUE re2_regexp_named_capturing_groups(VALUE self) {
|
|
@@ -1242,16 +1272,23 @@ static VALUE re2_regexp_named_capturing_groups(VALUE self) {
|
|
|
1242
1272
|
* @return [Boolean, RE2::MatchData]
|
|
1243
1273
|
*
|
|
1244
1274
|
* @overload match(text)
|
|
1245
|
-
* Returns an {RE2::MatchData} containing the matching
|
|
1246
|
-
*
|
|
1247
|
-
*
|
|
1275
|
+
* Returns an {RE2::MatchData} containing the matching pattern and all
|
|
1276
|
+
* subpatterns resulting from looking for the regexp in +text+ if the pattern
|
|
1277
|
+
* contains capturing groups.
|
|
1278
|
+
*
|
|
1279
|
+
* Returns either true or false indicating whether a successful match was
|
|
1280
|
+
* made if the pattern contains no capturing groups.
|
|
1248
1281
|
*
|
|
1249
1282
|
* @param [String] text the text to search
|
|
1250
|
-
* @return [RE2::MatchData] the
|
|
1283
|
+
* @return [RE2::MatchData] if the pattern contains capturing groups
|
|
1284
|
+
* @return [Boolean] if the pattern does not contain capturing groups
|
|
1251
1285
|
* @raise [NoMemoryError] if there was not enough memory to allocate the matches
|
|
1252
|
-
* @example
|
|
1286
|
+
* @example Matching with capturing groups
|
|
1253
1287
|
* r = RE2::Regexp.new('w(o)(o)')
|
|
1254
1288
|
* r.match('woo') #=> #<RE2::MatchData "woo" 1:"o" 2:"o">
|
|
1289
|
+
* @example Matching without capturing groups
|
|
1290
|
+
* r = RE2::Regexp.new('woo')
|
|
1291
|
+
* r.match('woo') #=> true
|
|
1255
1292
|
*
|
|
1256
1293
|
* @overload match(text, 0)
|
|
1257
1294
|
* Returns either true or false indicating whether a
|
|
@@ -1390,6 +1427,10 @@ static VALUE re2_regexp_scan(VALUE self, VALUE text) {
|
|
|
1390
1427
|
* Returns a copy of +str+ with the first occurrence +pattern+
|
|
1391
1428
|
* replaced with +rewrite+.
|
|
1392
1429
|
*
|
|
1430
|
+
* Note RE2 only supports UTF-8 and ISO-8859-1 encoding so strings will be
|
|
1431
|
+
* returned in UTF-8 by default or ISO-8859-1 if the :utf8 option for the
|
|
1432
|
+
* RE2::Regexp is set to false (any other encoding's behaviour is undefined).
|
|
1433
|
+
*
|
|
1393
1434
|
* @param [String] str the string to modify
|
|
1394
1435
|
* @param [String, RE2::Regexp] pattern a regexp matching text to be replaced
|
|
1395
1436
|
* @param [String] rewrite the string to replace with
|
|
@@ -1404,7 +1445,9 @@ static VALUE re2_Replace(VALUE self, VALUE str, VALUE pattern,
|
|
|
1404
1445
|
UNUSED(self);
|
|
1405
1446
|
re2_pattern *p;
|
|
1406
1447
|
|
|
1407
|
-
/*
|
|
1448
|
+
/* Take a copy of str so it can be modified in-place by
|
|
1449
|
+
* RE2::Replace.
|
|
1450
|
+
*/
|
|
1408
1451
|
string str_as_string(StringValuePtr(str));
|
|
1409
1452
|
|
|
1410
1453
|
/* Do the replacement. */
|
|
@@ -1418,8 +1461,8 @@ static VALUE re2_Replace(VALUE self, VALUE str, VALUE pattern,
|
|
|
1418
1461
|
RE2::Replace(&str_as_string, StringValuePtr(pattern),
|
|
1419
1462
|
StringValuePtr(rewrite));
|
|
1420
1463
|
|
|
1421
|
-
return
|
|
1422
|
-
|
|
1464
|
+
return ENCODED_STR_NEW(str_as_string.data(), str_as_string.size(),
|
|
1465
|
+
"UTF-8");
|
|
1423
1466
|
}
|
|
1424
1467
|
|
|
1425
1468
|
}
|
|
@@ -1427,6 +1470,10 @@ static VALUE re2_Replace(VALUE self, VALUE str, VALUE pattern,
|
|
|
1427
1470
|
/*
|
|
1428
1471
|
* Return a copy of +str+ with +pattern+ replaced by +rewrite+.
|
|
1429
1472
|
*
|
|
1473
|
+
* Note RE2 only supports UTF-8 and ISO-8859-1 encoding so strings will be
|
|
1474
|
+
* returned in UTF-8 by default or ISO-8859-1 if the :utf8 option for the
|
|
1475
|
+
* RE2::Regexp is set to false (any other encoding's behaviour is undefined).
|
|
1476
|
+
*
|
|
1430
1477
|
* @param [String] str the string to modify
|
|
1431
1478
|
* @param [String, RE2::Regexp] pattern a regexp matching text to be replaced
|
|
1432
1479
|
* @param [String] rewrite the string to replace with
|
|
@@ -1440,7 +1487,9 @@ static VALUE re2_GlobalReplace(VALUE self, VALUE str, VALUE pattern,
|
|
|
1440
1487
|
VALUE rewrite) {
|
|
1441
1488
|
UNUSED(self);
|
|
1442
1489
|
|
|
1443
|
-
/*
|
|
1490
|
+
/* Take a copy of str so it can be modified in-place by
|
|
1491
|
+
* RE2::GlobalReplace.
|
|
1492
|
+
*/
|
|
1444
1493
|
re2_pattern *p;
|
|
1445
1494
|
string str_as_string(StringValuePtr(str));
|
|
1446
1495
|
|
|
@@ -1455,8 +1504,8 @@ static VALUE re2_GlobalReplace(VALUE self, VALUE str, VALUE pattern,
|
|
|
1455
1504
|
RE2::GlobalReplace(&str_as_string, StringValuePtr(pattern),
|
|
1456
1505
|
StringValuePtr(rewrite));
|
|
1457
1506
|
|
|
1458
|
-
return
|
|
1459
|
-
|
|
1507
|
+
return ENCODED_STR_NEW(str_as_string.data(), str_as_string.size(),
|
|
1508
|
+
"UTF-8");
|
|
1460
1509
|
}
|
|
1461
1510
|
}
|
|
1462
1511
|
|
|
@@ -1579,11 +1628,12 @@ static VALUE re2_set_initialize(int argc, VALUE *argv, VALUE self) {
|
|
|
1579
1628
|
* set.add("def") #=> 1
|
|
1580
1629
|
*/
|
|
1581
1630
|
static VALUE re2_set_add(VALUE self, VALUE pattern) {
|
|
1582
|
-
|
|
1631
|
+
StringValue(pattern);
|
|
1583
1632
|
re2::StringPiece regex(RSTRING_PTR(pattern), RSTRING_LEN(pattern));
|
|
1584
1633
|
std::string err;
|
|
1585
1634
|
re2_set *s;
|
|
1586
1635
|
Data_Get_Struct(self, re2_set, s);
|
|
1636
|
+
|
|
1587
1637
|
int index = s->set->Add(regex, &err);
|
|
1588
1638
|
if (index < 0) {
|
|
1589
1639
|
rb_raise(rb_eArgError, "str rejected by RE2::Set->Add(): %s", err.c_str());
|
|
@@ -1669,7 +1719,8 @@ static VALUE re2_set_match(int argc, VALUE *argv, VALUE self) {
|
|
|
1669
1719
|
VALUE str, options, exception_option;
|
|
1670
1720
|
bool raise_exception = true;
|
|
1671
1721
|
rb_scan_args(argc, argv, "11", &str, &options);
|
|
1672
|
-
|
|
1722
|
+
|
|
1723
|
+
StringValue(str);
|
|
1673
1724
|
re2::StringPiece data(RSTRING_PTR(str), RSTRING_LEN(str));
|
|
1674
1725
|
std::vector<int> v;
|
|
1675
1726
|
re2_set *s;
|
data/lib/re2/version.rb
CHANGED
data/re2.gemspec
CHANGED
|
@@ -2,8 +2,8 @@ require_relative 'lib/re2/version'
|
|
|
2
2
|
|
|
3
3
|
Gem::Specification.new do |s|
|
|
4
4
|
s.name = "re2"
|
|
5
|
-
s.summary = "Ruby bindings to
|
|
6
|
-
s.description = 'Ruby bindings to
|
|
5
|
+
s.summary = "Ruby bindings to RE2."
|
|
6
|
+
s.description = 'Ruby bindings to RE2, "a fast, safe, thread-friendly alternative to backtracking regular expression engines like those used in PCRE, Perl, and Python".'
|
|
7
7
|
s.version = RE2::VERSION
|
|
8
8
|
s.authors = ["Paul Mucur", "Stan Hu"]
|
|
9
9
|
s.homepage = "https://github.com/mudge/re2"
|
|
@@ -22,6 +22,7 @@ Gem::Specification.new do |s|
|
|
|
22
22
|
"lib/re2/string.rb",
|
|
23
23
|
"lib/re2/version.rb",
|
|
24
24
|
"LICENSE.txt",
|
|
25
|
+
"LICENSE-DEPENDENCIES.txt",
|
|
25
26
|
"README.md",
|
|
26
27
|
"Rakefile",
|
|
27
28
|
"re2.gemspec"
|
|
@@ -36,8 +37,8 @@ Gem::Specification.new do |s|
|
|
|
36
37
|
"spec/re2/set_spec.rb",
|
|
37
38
|
"spec/re2/scanner_spec.rb"
|
|
38
39
|
]
|
|
39
|
-
s.add_development_dependency
|
|
40
|
-
s.add_development_dependency
|
|
40
|
+
s.add_development_dependency("rake-compiler", "~> 1.2.1")
|
|
41
|
+
s.add_development_dependency("rake-compiler-dock", "~> 1.3.0")
|
|
41
42
|
s.add_development_dependency("rspec", "~> 3.2")
|
|
42
43
|
s.add_runtime_dependency("mini_portile2", "~> 2.8.4") # keep version in sync with extconf.rb
|
|
43
44
|
end
|
data/spec/kernel_spec.rb
CHANGED
|
@@ -1,13 +1,13 @@
|
|
|
1
1
|
RSpec.describe Kernel do
|
|
2
|
-
describe "
|
|
2
|
+
describe ".RE2" do
|
|
3
3
|
it "returns an RE2::Regexp instance given a pattern" do
|
|
4
4
|
expect(RE2('w(o)(o)')).to be_a(RE2::Regexp)
|
|
5
5
|
end
|
|
6
6
|
|
|
7
7
|
it "returns an RE2::Regexp instance given a pattern and options" do
|
|
8
8
|
re = RE2('w(o)(o)', :case_sensitive => false)
|
|
9
|
-
|
|
10
|
-
expect(re).
|
|
9
|
+
|
|
10
|
+
expect(re).not_to be_case_sensitive
|
|
11
11
|
end
|
|
12
12
|
end
|
|
13
13
|
end
|
data/spec/re2/match_data_spec.rb
CHANGED
|
@@ -10,6 +10,18 @@ RSpec.describe RE2::MatchData do
|
|
|
10
10
|
a = RE2::Regexp.new('(\d?)(a)(b)').match('ab').to_a
|
|
11
11
|
expect(a).to eq(["ab", nil, "a", "b"])
|
|
12
12
|
end
|
|
13
|
+
|
|
14
|
+
it "returns UTF-8 strings if the pattern is UTF-8" do
|
|
15
|
+
a = RE2::Regexp.new('w(o)(o)').match('woo').to_a
|
|
16
|
+
|
|
17
|
+
expect(a.map(&:encoding)).to all eq(Encoding::UTF_8)
|
|
18
|
+
end
|
|
19
|
+
|
|
20
|
+
it "returns ISO-8859-1 strings if the pattern is not UTF-8" do
|
|
21
|
+
a = RE2::Regexp.new('w(o)(o)', :utf8 => false).match('woo').to_a
|
|
22
|
+
|
|
23
|
+
expect(a.map(&:encoding)).to all eq(Encoding::ISO_8859_1)
|
|
24
|
+
end
|
|
13
25
|
end
|
|
14
26
|
|
|
15
27
|
describe "#[]" do
|
|
@@ -19,6 +31,18 @@ RSpec.describe RE2::MatchData do
|
|
|
19
31
|
expect(md[2]).to eq("23")
|
|
20
32
|
end
|
|
21
33
|
|
|
34
|
+
it "returns a UTF-8 string by numerical index if the pattern is UTF-8" do
|
|
35
|
+
md = RE2::Regexp.new('(\d)(\d{2})').match("123")
|
|
36
|
+
|
|
37
|
+
expect(md[1].encoding).to eq(Encoding::UTF_8)
|
|
38
|
+
end
|
|
39
|
+
|
|
40
|
+
it "returns a ISO-8859-1 string by numerical index if the pattern is not UTF-8" do
|
|
41
|
+
md = RE2::Regexp.new('(\d)(\d{2})', :utf8 => false).match("123")
|
|
42
|
+
|
|
43
|
+
expect(md[1].encoding).to eq(Encoding::ISO_8859_1)
|
|
44
|
+
end
|
|
45
|
+
|
|
22
46
|
it "has the whole match as the 0th item" do
|
|
23
47
|
md = RE2::Regexp.new('(\d)(\d{2})').match("123")
|
|
24
48
|
expect(md[0]).to eq("123")
|
data/spec/re2/regexp_spec.rb
CHANGED
|
@@ -287,6 +287,12 @@ RSpec.describe RE2::Regexp do
|
|
|
287
287
|
expect(re.match("My age is 99", 0)).to eq(false)
|
|
288
288
|
end
|
|
289
289
|
|
|
290
|
+
it "returns only true or false if the pattern has no capturing groups" do
|
|
291
|
+
re = RE2::Regexp.new('My name is')
|
|
292
|
+
|
|
293
|
+
expect(re.match('My name is Robert Paulson')).to eq(true)
|
|
294
|
+
end
|
|
295
|
+
|
|
290
296
|
it "raises an exception when given nil" do
|
|
291
297
|
expect { re.match(nil) }.to raise_error(TypeError)
|
|
292
298
|
end
|
data/spec/re2/scanner_spec.rb
CHANGED
|
@@ -21,9 +21,10 @@ RSpec.describe RE2::Scanner do
|
|
|
21
21
|
end
|
|
22
22
|
|
|
23
23
|
describe "#scan" do
|
|
24
|
-
it "returns the next array of matches" do
|
|
24
|
+
it "returns the next array of matches", :aggregate_failures do
|
|
25
25
|
r = RE2::Regexp.new('(\w+)')
|
|
26
26
|
scanner = r.scan("It is a truth universally acknowledged")
|
|
27
|
+
|
|
27
28
|
expect(scanner.scan).to eq(["It"])
|
|
28
29
|
expect(scanner.scan).to eq(["is"])
|
|
29
30
|
expect(scanner.scan).to eq(["a"])
|
|
@@ -33,41 +34,73 @@ RSpec.describe RE2::Scanner do
|
|
|
33
34
|
expect(scanner.scan).to be_nil
|
|
34
35
|
end
|
|
35
36
|
|
|
37
|
+
it "returns UTF-8 matches if the pattern is UTF-8" do
|
|
38
|
+
r = RE2::Regexp.new('(\w+)')
|
|
39
|
+
scanner = r.scan("It")
|
|
40
|
+
matches = scanner.scan
|
|
41
|
+
|
|
42
|
+
expect(matches.first.encoding).to eq(Encoding::UTF_8)
|
|
43
|
+
end
|
|
44
|
+
|
|
45
|
+
it "returns ISO-8859-1 matches if the pattern is not UTF-8" do
|
|
46
|
+
r = RE2::Regexp.new('(\w+)', :utf8 => false)
|
|
47
|
+
scanner = r.scan("It")
|
|
48
|
+
matches = scanner.scan
|
|
49
|
+
|
|
50
|
+
expect(matches.first.encoding).to eq(Encoding::ISO_8859_1)
|
|
51
|
+
end
|
|
52
|
+
|
|
53
|
+
it "returns multiple capturing groups at a time", :aggregate_failures do
|
|
54
|
+
r = RE2::Regexp.new('(\w+) (\w+)')
|
|
55
|
+
scanner = r.scan("It is a truth universally acknowledged")
|
|
56
|
+
|
|
57
|
+
expect(scanner.scan).to eq(["It", "is"])
|
|
58
|
+
expect(scanner.scan).to eq(["a", "truth"])
|
|
59
|
+
expect(scanner.scan).to eq(["universally", "acknowledged"])
|
|
60
|
+
expect(scanner.scan).to be_nil
|
|
61
|
+
end
|
|
62
|
+
|
|
36
63
|
it "returns an empty array if there are no capturing groups" do
|
|
37
64
|
r = RE2::Regexp.new('\w+')
|
|
38
65
|
scanner = r.scan("Foo bar")
|
|
66
|
+
|
|
39
67
|
expect(scanner.scan).to eq([])
|
|
40
68
|
end
|
|
41
69
|
|
|
42
70
|
it "returns nil if there is no match" do
|
|
43
71
|
r = RE2::Regexp.new('\d+')
|
|
44
72
|
scanner = r.scan("Foo bar")
|
|
73
|
+
|
|
45
74
|
expect(scanner.scan).to be_nil
|
|
46
75
|
end
|
|
47
76
|
|
|
48
77
|
it "returns nil if the regexp is invalid" do
|
|
49
78
|
r = RE2::Regexp.new('???', :log_errors => false)
|
|
50
79
|
scanner = r.scan("Foo bar")
|
|
80
|
+
|
|
51
81
|
expect(scanner.scan).to be_nil
|
|
52
82
|
end
|
|
53
83
|
|
|
54
|
-
it "returns an empty array if the input is empty" do
|
|
84
|
+
it "returns an empty array if the input is empty", :aggregate_failures do
|
|
55
85
|
r = RE2::Regexp.new("")
|
|
56
86
|
scanner = r.scan("")
|
|
87
|
+
|
|
57
88
|
expect(scanner.scan).to eq([])
|
|
58
89
|
expect(scanner.scan).to be_nil
|
|
59
90
|
end
|
|
60
91
|
|
|
61
|
-
it "returns an array of nil with an empty input and capture" do
|
|
92
|
+
it "returns an array of nil with an empty input and capture", :aggregate_failures do
|
|
62
93
|
r = RE2::Regexp.new("()")
|
|
63
94
|
scanner = r.scan("")
|
|
95
|
+
|
|
64
96
|
expect(scanner.scan).to eq([nil])
|
|
65
97
|
expect(scanner.scan).to be_nil
|
|
66
98
|
end
|
|
67
99
|
|
|
68
|
-
it "returns an empty array for every match if the pattern is empty" do
|
|
100
|
+
it "returns an empty array for every match if the pattern is empty", :aggregate_failures do
|
|
69
101
|
r = RE2::Regexp.new("")
|
|
70
102
|
scanner = r.scan("Foo")
|
|
103
|
+
|
|
71
104
|
expect(scanner.scan).to eq([])
|
|
72
105
|
expect(scanner.scan).to eq([])
|
|
73
106
|
expect(scanner.scan).to eq([])
|
|
@@ -75,9 +108,10 @@ RSpec.describe RE2::Scanner do
|
|
|
75
108
|
expect(scanner.scan).to be_nil
|
|
76
109
|
end
|
|
77
110
|
|
|
78
|
-
it "returns an array of nil if the pattern is an empty capturing group" do
|
|
111
|
+
it "returns an array of nil if the pattern is an empty capturing group", :aggregate_failures do
|
|
79
112
|
r = RE2::Regexp.new("()")
|
|
80
113
|
scanner = r.scan("Foo")
|
|
114
|
+
|
|
81
115
|
expect(scanner.scan).to eq([nil])
|
|
82
116
|
expect(scanner.scan).to eq([nil])
|
|
83
117
|
expect(scanner.scan).to eq([nil])
|
|
@@ -85,9 +119,10 @@ RSpec.describe RE2::Scanner do
|
|
|
85
119
|
expect(scanner.scan).to be_nil
|
|
86
120
|
end
|
|
87
121
|
|
|
88
|
-
it "returns array of nils with multiple empty capturing groups" do
|
|
122
|
+
it "returns array of nils with multiple empty capturing groups", :aggregate_failures do
|
|
89
123
|
r = RE2::Regexp.new("()()()")
|
|
90
124
|
scanner = r.scan("Foo")
|
|
125
|
+
|
|
91
126
|
expect(scanner.scan).to eq([nil, nil, nil])
|
|
92
127
|
expect(scanner.scan).to eq([nil, nil, nil])
|
|
93
128
|
expect(scanner.scan).to eq([nil, nil, nil])
|
|
@@ -95,17 +130,34 @@ RSpec.describe RE2::Scanner do
|
|
|
95
130
|
expect(scanner.scan).to be_nil
|
|
96
131
|
end
|
|
97
132
|
|
|
98
|
-
it "supports empty groups with multibyte characters" do
|
|
133
|
+
it "supports empty groups with multibyte characters", :aggregate_failures do
|
|
99
134
|
r = RE2::Regexp.new("()€")
|
|
100
135
|
scanner = r.scan("€")
|
|
136
|
+
|
|
101
137
|
expect(scanner.scan).to eq([nil])
|
|
102
138
|
expect(scanner.scan).to be_nil
|
|
103
139
|
end
|
|
140
|
+
|
|
141
|
+
it "raises a Type Error if given input that can't be coerced to a String" do
|
|
142
|
+
r = RE2::Regexp.new('(\w+)')
|
|
143
|
+
|
|
144
|
+
expect { r.scan(0) }.to raise_error(TypeError)
|
|
145
|
+
end
|
|
146
|
+
|
|
147
|
+
it "accepts input that can be coerced to a String", :aggregate_failures do
|
|
148
|
+
r = RE2::Regexp.new('(\w+)')
|
|
149
|
+
scanner = r.scan(StringLike.new("Hello world"))
|
|
150
|
+
|
|
151
|
+
expect(scanner.scan).to eq(["Hello"])
|
|
152
|
+
expect(scanner.scan).to eq(["world"])
|
|
153
|
+
expect(scanner.scan).to be_nil
|
|
154
|
+
end
|
|
104
155
|
end
|
|
105
156
|
|
|
106
157
|
it "is enumerable" do
|
|
107
158
|
r = RE2::Regexp.new('(\d)')
|
|
108
159
|
scanner = r.scan("There are 1 some 2 numbers 3")
|
|
160
|
+
|
|
109
161
|
expect(scanner).to be_a(Enumerable)
|
|
110
162
|
end
|
|
111
163
|
|
|
@@ -113,12 +165,8 @@ RSpec.describe RE2::Scanner do
|
|
|
113
165
|
it "yields each match" do
|
|
114
166
|
r = RE2::Regexp.new('(\d)')
|
|
115
167
|
scanner = r.scan("There are 1 some 2 numbers 3")
|
|
116
|
-
matches = []
|
|
117
|
-
scanner.each do |match|
|
|
118
|
-
matches << match
|
|
119
|
-
end
|
|
120
168
|
|
|
121
|
-
expect(
|
|
169
|
+
expect { |b| scanner.each(&b) }.to yield_successive_args(["1"], ["2"], ["3"])
|
|
122
170
|
end
|
|
123
171
|
|
|
124
172
|
it "returns an enumerator when not given a block" do
|
|
@@ -135,22 +183,28 @@ RSpec.describe RE2::Scanner do
|
|
|
135
183
|
end
|
|
136
184
|
|
|
137
185
|
describe "#rewind" do
|
|
138
|
-
it "resets any consumption" do
|
|
186
|
+
it "resets any consumption", :aggregate_failures do
|
|
139
187
|
r = RE2::Regexp.new('(\d)')
|
|
140
188
|
scanner = r.scan("There are 1 some 2 numbers 3")
|
|
189
|
+
|
|
141
190
|
expect(scanner.to_enum.first).to eq(["1"])
|
|
142
191
|
expect(scanner.to_enum.first).to eq(["2"])
|
|
192
|
+
|
|
143
193
|
scanner.rewind
|
|
194
|
+
|
|
144
195
|
expect(scanner.to_enum.first).to eq(["1"])
|
|
145
196
|
end
|
|
146
197
|
|
|
147
|
-
it "resets the eof? check" do
|
|
198
|
+
it "resets the eof? check", :aggregate_failures do
|
|
148
199
|
r = RE2::Regexp.new('(\d)')
|
|
149
200
|
scanner = r.scan("1")
|
|
150
201
|
scanner.scan
|
|
151
|
-
|
|
202
|
+
|
|
203
|
+
expect(scanner).to be_eof
|
|
204
|
+
|
|
152
205
|
scanner.rewind
|
|
153
|
-
|
|
206
|
+
|
|
207
|
+
expect(scanner).not_to be_eof
|
|
154
208
|
end
|
|
155
209
|
end
|
|
156
210
|
|
|
@@ -159,7 +213,7 @@ RSpec.describe RE2::Scanner do
|
|
|
159
213
|
r = RE2::Regexp.new('(\d)')
|
|
160
214
|
scanner = r.scan("1 2 3")
|
|
161
215
|
|
|
162
|
-
expect(scanner
|
|
216
|
+
expect(scanner).not_to be_eof
|
|
163
217
|
end
|
|
164
218
|
|
|
165
219
|
it "returns true if the input has been consumed" do
|
|
@@ -167,7 +221,7 @@ RSpec.describe RE2::Scanner do
|
|
|
167
221
|
scanner = r.scan("1")
|
|
168
222
|
scanner.scan
|
|
169
223
|
|
|
170
|
-
expect(scanner
|
|
224
|
+
expect(scanner).to be_eof
|
|
171
225
|
end
|
|
172
226
|
|
|
173
227
|
it "returns false if no match is made" do
|
|
@@ -175,14 +229,14 @@ RSpec.describe RE2::Scanner do
|
|
|
175
229
|
scanner = r.scan("a")
|
|
176
230
|
scanner.scan
|
|
177
231
|
|
|
178
|
-
expect(scanner
|
|
232
|
+
expect(scanner).not_to be_eof
|
|
179
233
|
end
|
|
180
234
|
|
|
181
235
|
it "returns false with an empty input that has not been scanned" do
|
|
182
236
|
r = RE2::Regexp.new("")
|
|
183
237
|
scanner = r.scan("")
|
|
184
238
|
|
|
185
|
-
expect(scanner
|
|
239
|
+
expect(scanner).not_to be_eof
|
|
186
240
|
end
|
|
187
241
|
|
|
188
242
|
it "returns false with an empty input that has not been matched" do
|
|
@@ -190,7 +244,7 @@ RSpec.describe RE2::Scanner do
|
|
|
190
244
|
scanner = r.scan("")
|
|
191
245
|
scanner.scan
|
|
192
246
|
|
|
193
|
-
expect(scanner
|
|
247
|
+
expect(scanner).not_to be_eof
|
|
194
248
|
end
|
|
195
249
|
|
|
196
250
|
it "returns true with an empty input that has been matched" do
|
|
@@ -198,7 +252,7 @@ RSpec.describe RE2::Scanner do
|
|
|
198
252
|
scanner = r.scan("")
|
|
199
253
|
scanner.scan
|
|
200
254
|
|
|
201
|
-
expect(scanner
|
|
255
|
+
expect(scanner).to be_eof
|
|
202
256
|
end
|
|
203
257
|
end
|
|
204
258
|
end
|
data/spec/re2/set_spec.rb
CHANGED
|
@@ -67,11 +67,17 @@ RSpec.describe RE2::Set do
|
|
|
67
67
|
end
|
|
68
68
|
end
|
|
69
69
|
|
|
70
|
-
it "raises an error if given a
|
|
70
|
+
it "raises an error if given a pattern that can't be coerced to a String" do
|
|
71
71
|
set = RE2::Set.new(:unanchored, :log_errors => false)
|
|
72
72
|
|
|
73
73
|
expect { set.add(0) }.to raise_error(TypeError)
|
|
74
74
|
end
|
|
75
|
+
|
|
76
|
+
it "accepts a pattern that can be coerced to a String" do
|
|
77
|
+
set = RE2::Set.new
|
|
78
|
+
|
|
79
|
+
expect(set.add(StringLike.new("abc"))).to eq(0)
|
|
80
|
+
end
|
|
75
81
|
end
|
|
76
82
|
|
|
77
83
|
describe "#compile" do
|
|
@@ -96,6 +102,24 @@ RSpec.describe RE2::Set do
|
|
|
96
102
|
expect(set.match("abcdefghi", :exception => false)).to eq([0, 1, 2])
|
|
97
103
|
end
|
|
98
104
|
|
|
105
|
+
it "returns an empty array if there is no match" do
|
|
106
|
+
set = RE2::Set.new
|
|
107
|
+
set.add("abc")
|
|
108
|
+
set.compile
|
|
109
|
+
|
|
110
|
+
expect(set.match("def", :exception => false)).to be_empty
|
|
111
|
+
end
|
|
112
|
+
|
|
113
|
+
it "returns an empty array if there is no match when :exception is true" do
|
|
114
|
+
skip "Underlying RE2::Set::Match does not output error information" unless RE2::Set.match_raises_errors?
|
|
115
|
+
|
|
116
|
+
set = RE2::Set.new
|
|
117
|
+
set.add("abc")
|
|
118
|
+
set.compile
|
|
119
|
+
|
|
120
|
+
expect(set.match("def")).to be_empty
|
|
121
|
+
end
|
|
122
|
+
|
|
99
123
|
it "raises an error if called before #compile by default" do
|
|
100
124
|
skip "Underlying RE2::Set::Match does not output error information" unless RE2::Set.match_raises_errors?
|
|
101
125
|
|
|
@@ -139,6 +163,22 @@ RSpec.describe RE2::Set do
|
|
|
139
163
|
|
|
140
164
|
expect { set.match("", 0) }.to raise_error(TypeError)
|
|
141
165
|
end
|
|
166
|
+
|
|
167
|
+
it "raises a Type Error if given input that can't be coerced to a String" do
|
|
168
|
+
set = RE2::Set.new
|
|
169
|
+
set.add("abc")
|
|
170
|
+
set.compile
|
|
171
|
+
|
|
172
|
+
expect { set.match(0, :exception => false) }.to raise_error(TypeError)
|
|
173
|
+
end
|
|
174
|
+
|
|
175
|
+
it "accepts input if it can be coerced to a String" do
|
|
176
|
+
set = RE2::Set.new
|
|
177
|
+
set.add("abc")
|
|
178
|
+
set.compile
|
|
179
|
+
|
|
180
|
+
expect(set.match(StringLike.new("abcdef"), :exception => false)).to contain_exactly(0)
|
|
181
|
+
end
|
|
142
182
|
end
|
|
143
183
|
|
|
144
184
|
def silence_stderr
|
data/spec/re2/string_spec.rb
CHANGED
|
@@ -12,7 +12,8 @@ RSpec.describe RE2::String do
|
|
|
12
12
|
|
|
13
13
|
it "doesn't perform an in-place replacement" do
|
|
14
14
|
string = "My name is Robert Paulson"
|
|
15
|
-
|
|
15
|
+
|
|
16
|
+
expect(string.re2_sub('Robert', 'Crobert')).not_to equal(string)
|
|
16
17
|
end
|
|
17
18
|
end
|
|
18
19
|
|
|
@@ -23,13 +24,15 @@ RSpec.describe RE2::String do
|
|
|
23
24
|
|
|
24
25
|
it "doesn't perform an in-place replacement" do
|
|
25
26
|
string = "My name is Robert Paulson"
|
|
26
|
-
|
|
27
|
+
|
|
28
|
+
expect(string.re2_gsub('a', 'e')).not_to equal(string)
|
|
27
29
|
end
|
|
28
30
|
end
|
|
29
31
|
|
|
30
32
|
describe "#re2_match" do
|
|
31
|
-
it "delegates to RE2::Regexp#match to perform matches" do
|
|
33
|
+
it "delegates to RE2::Regexp#match to perform matches", :aggregate_failures do
|
|
32
34
|
md = "My name is Robert Paulson".re2_match('My name is (\S+) (\S+)')
|
|
35
|
+
|
|
33
36
|
expect(md).to be_a(RE2::MatchData)
|
|
34
37
|
expect(md[0]).to eq("My name is Robert Paulson")
|
|
35
38
|
expect(md[1]).to eq("Robert")
|
|
@@ -38,6 +41,7 @@ RSpec.describe RE2::String do
|
|
|
38
41
|
|
|
39
42
|
it "supports limiting the number of matches" do
|
|
40
43
|
md = "My name is Robert Paulson".re2_match('My name is (\S+) (\S+)', 0)
|
|
44
|
+
|
|
41
45
|
expect(md).to eq(true)
|
|
42
46
|
end
|
|
43
47
|
end
|
data/spec/re2_spec.rb
CHANGED
|
@@ -1,5 +1,5 @@
|
|
|
1
1
|
RSpec.describe RE2 do
|
|
2
|
-
describe "
|
|
2
|
+
describe ".Replace" do
|
|
3
3
|
it "only replaces the first occurrence of the pattern" do
|
|
4
4
|
expect(RE2.Replace("woo", "o", "a")).to eq("wao")
|
|
5
5
|
end
|
|
@@ -15,29 +15,68 @@ RSpec.describe RE2 do
|
|
|
15
15
|
it "does not perform replacements in-place" do
|
|
16
16
|
name = "Robert"
|
|
17
17
|
replacement = RE2.Replace(name, "R", "Cr")
|
|
18
|
-
|
|
18
|
+
|
|
19
|
+
expect(name).not_to equal(replacement)
|
|
19
20
|
end
|
|
20
21
|
|
|
21
22
|
it "supports passing an RE2::Regexp as the pattern" do
|
|
22
23
|
re = RE2::Regexp.new('wo{2}')
|
|
24
|
+
|
|
23
25
|
expect(RE2.Replace("woo", re, "miaow")).to eq("miaow")
|
|
24
26
|
end
|
|
25
27
|
|
|
26
28
|
it "respects any passed RE2::Regexp's flags" do
|
|
27
29
|
re = RE2::Regexp.new('gOOD MORNING', :case_sensitive => false)
|
|
30
|
+
|
|
28
31
|
expect(RE2.Replace("Good morning", re, "hi")).to eq("hi")
|
|
29
32
|
end
|
|
30
33
|
|
|
31
|
-
|
|
32
|
-
|
|
33
|
-
|
|
34
|
-
|
|
35
|
-
|
|
36
|
-
|
|
34
|
+
it "supports passing something that can be coerced to a String as input" do
|
|
35
|
+
expect(RE2.Replace(StringLike.new("woo"), "oo", "ah")).to eq("wah")
|
|
36
|
+
end
|
|
37
|
+
|
|
38
|
+
it "supports passing something that can be coerced to a String as a pattern" do
|
|
39
|
+
expect(RE2.Replace("woo", StringLike.new("oo"), "ah")).to eq("wah")
|
|
40
|
+
end
|
|
41
|
+
|
|
42
|
+
it "supports passing something that can be coerced to a String as a replacement" do
|
|
43
|
+
expect(RE2.Replace("woo", "oo", StringLike.new("ah"))).to eq("wah")
|
|
44
|
+
end
|
|
45
|
+
|
|
46
|
+
it "returns UTF-8 strings if the pattern is UTF-8" do
|
|
47
|
+
original = "Foo".encode("ISO-8859-1")
|
|
48
|
+
replacement = RE2.Replace(original, "oo", "ah")
|
|
49
|
+
|
|
50
|
+
expect(replacement.encoding).to eq(Encoding::UTF_8)
|
|
51
|
+
end
|
|
52
|
+
|
|
53
|
+
it "returns ISO-8859-1 strings if the pattern is not UTF-8" do
|
|
54
|
+
original = "Foo"
|
|
55
|
+
replacement = RE2.Replace(original, RE2("oo", :utf8 => false), "ah")
|
|
56
|
+
|
|
57
|
+
expect(replacement.encoding).to eq(Encoding::ISO_8859_1)
|
|
58
|
+
end
|
|
59
|
+
|
|
60
|
+
it "returns UTF-8 strings when given a String pattern" do
|
|
61
|
+
replacement = RE2.Replace("Foo", "oo".encode("ISO-8859-1"), "ah")
|
|
62
|
+
|
|
63
|
+
expect(replacement.encoding).to eq(Encoding::UTF_8)
|
|
64
|
+
end
|
|
65
|
+
|
|
66
|
+
it "raises a Type Error for input that can't be converted to String" do
|
|
67
|
+
expect { RE2.Replace(0, "oo", "ah") }.to raise_error(TypeError)
|
|
68
|
+
end
|
|
69
|
+
|
|
70
|
+
it "raises a Type Error for a non-RE2::Regexp pattern that can't be converted to String" do
|
|
71
|
+
expect { RE2.Replace("woo", 0, "ah") }.to raise_error(TypeError)
|
|
72
|
+
end
|
|
73
|
+
|
|
74
|
+
it "raises a Type Error for a replacement that can't be converted to String" do
|
|
75
|
+
expect { RE2.Replace("woo", "oo", 0) }.to raise_error(TypeError)
|
|
37
76
|
end
|
|
38
77
|
end
|
|
39
78
|
|
|
40
|
-
describe "
|
|
79
|
+
describe ".GlobalReplace" do
|
|
41
80
|
it "replaces every occurrence of a pattern" do
|
|
42
81
|
expect(RE2.GlobalReplace("woo", "o", "a")).to eq("waa")
|
|
43
82
|
end
|
|
@@ -53,23 +92,78 @@ RSpec.describe RE2 do
|
|
|
53
92
|
it "does not perform replacement in-place" do
|
|
54
93
|
name = "Robert"
|
|
55
94
|
replacement = RE2.GlobalReplace(name, "(?i)R", "w")
|
|
56
|
-
|
|
95
|
+
|
|
96
|
+
expect(name).not_to equal(replacement)
|
|
57
97
|
end
|
|
58
98
|
|
|
59
99
|
it "supports passing an RE2::Regexp as the pattern" do
|
|
60
100
|
re = RE2::Regexp.new('wo{2,}')
|
|
101
|
+
|
|
61
102
|
expect(RE2.GlobalReplace("woowooo", re, "miaow")).to eq("miaowmiaow")
|
|
62
103
|
end
|
|
63
104
|
|
|
64
105
|
it "respects any passed RE2::Regexp's flags" do
|
|
65
106
|
re = RE2::Regexp.new('gOOD MORNING', :case_sensitive => false)
|
|
107
|
+
|
|
66
108
|
expect(RE2.GlobalReplace("Good morning Good morning", re, "hi")).to eq("hi hi")
|
|
67
109
|
end
|
|
110
|
+
|
|
111
|
+
it "supports passing something that can be coerced to a String as input" do
|
|
112
|
+
expect(RE2.GlobalReplace(StringLike.new("woo"), "o", "a")).to eq("waa")
|
|
113
|
+
end
|
|
114
|
+
|
|
115
|
+
it "supports passing something that can be coerced to a String as a pattern" do
|
|
116
|
+
expect(RE2.GlobalReplace("woo", StringLike.new("o"), "a")).to eq("waa")
|
|
117
|
+
end
|
|
118
|
+
|
|
119
|
+
it "supports passing something that can be coerced to a String as a replacement" do
|
|
120
|
+
expect(RE2.GlobalReplace("woo", "o", StringLike.new("a"))).to eq("waa")
|
|
121
|
+
end
|
|
122
|
+
|
|
123
|
+
it "returns UTF-8 strings if the pattern is UTF-8" do
|
|
124
|
+
original = "Foo".encode("ISO-8859-1")
|
|
125
|
+
replacement = RE2.GlobalReplace(original, "oo", "ah")
|
|
126
|
+
|
|
127
|
+
expect(replacement.encoding).to eq(Encoding::UTF_8)
|
|
128
|
+
end
|
|
129
|
+
|
|
130
|
+
it "returns ISO-8859-1 strings if the pattern is not UTF-8" do
|
|
131
|
+
original = "Foo"
|
|
132
|
+
replacement = RE2.GlobalReplace(original, RE2("oo", :utf8 => false), "ah")
|
|
133
|
+
|
|
134
|
+
expect(replacement.encoding).to eq(Encoding::ISO_8859_1)
|
|
135
|
+
end
|
|
136
|
+
|
|
137
|
+
it "returns UTF-8 strings when given a String pattern" do
|
|
138
|
+
replacement = RE2.GlobalReplace("Foo", "oo".encode("ISO-8859-1"), "ah")
|
|
139
|
+
|
|
140
|
+
expect(replacement.encoding).to eq(Encoding::UTF_8)
|
|
141
|
+
end
|
|
142
|
+
|
|
143
|
+
it "raises a Type Error for input that can't be converted to String" do
|
|
144
|
+
expect { RE2.GlobalReplace(0, "o", "a") }.to raise_error(TypeError)
|
|
145
|
+
end
|
|
146
|
+
|
|
147
|
+
it "raises a Type Error for a non-RE2::Regexp pattern that can't be converted to String" do
|
|
148
|
+
expect { RE2.GlobalReplace("woo", 0, "a") }.to raise_error(TypeError)
|
|
149
|
+
end
|
|
150
|
+
|
|
151
|
+
it "raises a Type Error for a replacement that can't be converted to String" do
|
|
152
|
+
expect { RE2.GlobalReplace("woo", "o", 0) }.to raise_error(TypeError)
|
|
153
|
+
end
|
|
68
154
|
end
|
|
69
155
|
|
|
70
156
|
describe "#QuoteMeta" do
|
|
71
157
|
it "escapes a string so it can be used as a regular expression" do
|
|
72
158
|
expect(RE2.QuoteMeta("1.5-2.0?")).to eq('1\.5\-2\.0\?')
|
|
73
159
|
end
|
|
160
|
+
|
|
161
|
+
it "raises a Type Error for input that can't be converted to String" do
|
|
162
|
+
expect { RE2.QuoteMeta(0) }.to raise_error(TypeError)
|
|
163
|
+
end
|
|
164
|
+
|
|
165
|
+
it "supports passing something that can be coerced to a String as input" do
|
|
166
|
+
expect(RE2.QuoteMeta(StringLike.new("1.5"))).to eq('1\.5')
|
|
167
|
+
end
|
|
74
168
|
end
|
|
75
169
|
end
|
data/spec/spec_helper.rb
CHANGED
|
@@ -1,5 +1,15 @@
|
|
|
1
1
|
require "re2"
|
|
2
2
|
|
|
3
|
+
# To test passing objects that can be coerced to a String.
|
|
4
|
+
class StringLike
|
|
5
|
+
attr_reader :str
|
|
6
|
+
alias_method :to_str, :str
|
|
7
|
+
|
|
8
|
+
def initialize(str)
|
|
9
|
+
@str = str
|
|
10
|
+
end
|
|
11
|
+
end
|
|
12
|
+
|
|
3
13
|
RSpec.configure do |config|
|
|
4
14
|
config.expect_with :rspec do |expectations|
|
|
5
15
|
expectations.include_chain_clauses_in_custom_matcher_descriptions = true
|
metadata
CHANGED
|
@@ -1,7 +1,7 @@
|
|
|
1
1
|
--- !ruby/object:Gem::Specification
|
|
2
2
|
name: re2
|
|
3
3
|
version: !ruby/object:Gem::Version
|
|
4
|
-
version: 2.
|
|
4
|
+
version: 2.1.0
|
|
5
5
|
platform: ruby
|
|
6
6
|
authors:
|
|
7
7
|
- Paul Mucur
|
|
@@ -9,7 +9,7 @@ authors:
|
|
|
9
9
|
autorequire:
|
|
10
10
|
bindir: bin
|
|
11
11
|
cert_chain: []
|
|
12
|
-
date: 2023-09-
|
|
12
|
+
date: 2023-09-16 00:00:00.000000000 Z
|
|
13
13
|
dependencies:
|
|
14
14
|
- !ruby/object:Gem::Dependency
|
|
15
15
|
name: rake-compiler
|
|
@@ -67,7 +67,8 @@ dependencies:
|
|
|
67
67
|
- - "~>"
|
|
68
68
|
- !ruby/object:Gem::Version
|
|
69
69
|
version: 2.8.4
|
|
70
|
-
description: Ruby bindings to
|
|
70
|
+
description: Ruby bindings to RE2, "a fast, safe, thread-friendly alternative to backtracking
|
|
71
|
+
regular expression engines like those used in PCRE, Perl, and Python".
|
|
71
72
|
email:
|
|
72
73
|
executables: []
|
|
73
74
|
extensions:
|
|
@@ -76,6 +77,7 @@ extra_rdoc_files: []
|
|
|
76
77
|
files:
|
|
77
78
|
- ".rspec"
|
|
78
79
|
- Gemfile
|
|
80
|
+
- LICENSE-DEPENDENCIES.txt
|
|
79
81
|
- LICENSE.txt
|
|
80
82
|
- README.md
|
|
81
83
|
- Rakefile
|
|
@@ -120,7 +122,7 @@ requirements: []
|
|
|
120
122
|
rubygems_version: 3.4.10
|
|
121
123
|
signing_key:
|
|
122
124
|
specification_version: 4
|
|
123
|
-
summary: Ruby bindings to
|
|
125
|
+
summary: Ruby bindings to RE2.
|
|
124
126
|
test_files:
|
|
125
127
|
- spec/spec_helper.rb
|
|
126
128
|
- spec/re2_spec.rb
|