re2 2.9.0 → 2.27.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: be3eb546350524548db3855e8e75c4941805cfd6a27fdec30c39f6bbcdcdca04
-  data.tar.gz: 7b824cce1ca6731957f9bd066b79fc5fc63afb8c7bdefad29c75a7c41ed9ca2f
+  metadata.gz: ac8a9eeb58a93e3f4241d2677a51a15016d543fd55a003e358f1d8442717c517
+  data.tar.gz: d505664cefe09a08d4c14f174791bc3f213cec515897943f12adc2867a91df6f
 SHA512:
-  metadata.gz: c324a2506c302709dd302013cf205630e0b4df5322d1ee3342c7e609cb6a1ed1c405a7ad7f1ee185dc0990d796f979d4d38e22ae35a3dc6c0e08bf984800bc0a
-  data.tar.gz: 7466a5c1d97c8a4b4247dc30e8691cea0c409fde8ae2196858f59ed1d12f6019fc04ae6f9c25b6e35b2f39d8d03e070e2d960fc36327c04038274cc9dae66d9e
+  metadata.gz: 31affec3aeee1a47021d3e96360b52c7dc5065d2abcbd748c35d305e6fffdd3b7a521f92bb6c97aaca67b265bb9df34867f18c5e00edf647bd8e609e093c29aa
+  data.tar.gz: 76654d86eea56b7e66102fc6780d16f94eaba95f0d2c605a46a1eece51c2158414d26edfdcfc3732b2301fe35a6fdbe8860336ce8efc1bd416d0830ee0a64e93

data/Gemfile

@@ -1,5 +1,11 @@
+# frozen_string_literal: true
+
 source "https://rubygems.org"
 
 gemspec
 
 gem "rake", "> 12.3.2"
+
+group :memcheck, optional: true do
+  gem "ruby_memcheck"
+end

data/README.md

@@ -6,8 +6,8 @@ Python".
 
 [![Build Status](https://github.com/mudge/re2/actions/workflows/tests.yml/badge.svg?branch=main)](https://github.com/mudge/re2/actions)
 
-**Current version:** 2.9.0  
-**Bundled RE2 version:** libre2.11 (2024-03-01)  
+**Current version:** 2.27.0  
+**Bundled RE2 version:** libre2.11 (2025-11-05)  
 
 ```ruby
 RE2('h.*o').full_match?("hello")    #=> true
@@ -27,6 +27,8 @@ RE2('(\w+):(\d+)').full_match("ruby:1234")
     * [Submatch extraction](#submatch-extraction)
     * [Scanning text incrementally](#scanning-text-incrementally)
     * [Searching simultaneously](#searching-simultaneously)
+    * [Replacing and extracting](#replacing-and-extracting)
+    * [Escaping](#escaping)
     * [Encoding](#encoding)
 * [Requirements](#requirements)
     * [Native gems](#native-gems)
@@ -165,7 +167,72 @@ m["word"]   #=> "ruby"
 m["number"] #=> "1234"
 ```
 
-They can also be used with Ruby's [pattern matching](https://docs.ruby-lang.org/en/3.2/syntax/pattern_matching_rdoc.html):
+Multiple submatches can be retrieved at the same time by numeric index or name with [`values_at`](https://mudge.name/re2/RE2/MatchData.html#values_at-instance_method):
+
+```ruby
+m = RE2('(?P<word>\w+):(?P<number>\d+):(\d+)').full_match("ruby:1234:5678")
+#=> #<RE2::MatchData "ruby:1234:5678" 1:"ruby" 2:"1234" 3:"5678">
+
+m.values_at("word", :number, 3)
+#=> ["ruby", "1234", "5678"]
+```
+
+All captures can be returned as an array with [`captures`](https://mudge.name/re2/RE2/MatchData.html#captures-instance_method):
+
+```ruby
+m = RE2('(?P<word>\w+):(?P<number>\d+):(\d+)').full_match("ruby:1234:5678")
+#=> #<RE2::MatchData "ruby:1234:5678" 1:"ruby" 2:"1234" 3:"5678">
+
+m.captures #=> ["ruby", "1234", "5678"]
+```
+
+Capturing group names are available on both `RE2::Regexp` and `RE2::MatchData`:
+
+```ruby
+re = RE2('(?P<word>\w+):(?P<number>\d+):(\d+)')
+re.names           #=> ["number", "word"]
+
+m = re.full_match("ruby:1234:5678")
+m.names            #=> ["number", "word"]
+```
+
+Named captures can be returned as a hash with [`named_captures`](https://mudge.name/re2/RE2/MatchData.html#named_captures-instance_method):
+
+```ruby
+m = RE2('(?P<word>\w+):(?P<number>\d+):(\d+)').full_match("ruby:1234:5678")
+#=> #<RE2::MatchData "ruby:1234:5678" 1:"ruby" 2:"1234" 3:"5678">
+
+m.named_captures
+#=> {"number" => "1234", "word" => "ruby"}
+m.named_captures(symbolize_names: true)
+#=> {number: "1234", word: "ruby"}
+```
+
+This is [also available](https://mudge.name/re2/RE2/Regexp.html#named_captures-instance_method) on the original `RE2::Regexp` but will return the corresponding numerical index for each group:
+
+```ruby
+re = RE2('(?P<word>\w+):(?P<number>\d+):(\d+)')
+re.named_captures
+#=> {"number" => 2, "word" => 1}
+```
+
+The strings before and after a match can be returned with [`pre_match`](https://mudge.name/re2/RE2/MatchData.html#pre_match-instance_method) and [`post_match`](https://mudge.name/re2/RE2/MatchData.html#post_match-instance_method):
+
+```ruby
+m = RE2::Regexp.new('(\d+)').partial_match("bob 123 456")
+m.pre_match  #=> "bob "
+m.post_match #=> " 456"
+```
+
+The [`offset`](https://mudge.name/re2/RE2/MatchData.html#offset-instance_method) and [`match_length`](https://mudge.name/re2/RE2/MatchData.html#match_length-instance_method) of a match can be retrieved by index or name:
+
+```ruby
+m = RE2::Regexp.new('(\d+)').partial_match("bob 123 456")
+m.offset(1)       #=> [4, 7]
+m.match_length(1) #=> 3
+```
+
+`RE2::MatchData` objects can also be used with Ruby's [pattern matching](https://docs.ruby-lang.org/en/3.2/syntax/pattern_matching_rdoc.html):
 
 ```ruby
 case RE2('(\w+):(\d+)').full_match("ruby:1234")
@@ -224,17 +291,56 @@ the set. After all patterns have been added, the set can be compiled using
 and then
 [`RE2::Set#match`](https://mudge.name/re2/RE2/Set.html#match-instance_method)
 will return an array containing the indices of all the patterns that matched.
+[`RE2::Set#size`](https://mudge.name/re2/RE2/Set.html#size-instance_method)
+will return the number of patterns in the set.
 
 ```ruby
 set = RE2::Set.new
 set.add("abc")         #=> 0
 set.add("def")         #=> 1
 set.add("ghi")         #=> 2
+set.size               #=> 3
 set.compile            #=> true
 set.match("abcdefghi") #=> [0, 1, 2]
 set.match("ghidefabc") #=> [2, 1, 0]
 ```
 
+### Replacing and extracting
+
+[`RE2.replace`](https://mudge.name/re2/RE2.html#replace-class_method) returns a copy of a given string with the first occurrence of a pattern replaced with a given rewrite string:
+
+```ruby
+RE2.replace("hello there", "hello", "howdy") #=> "howdy there"
+```
+
+The pattern can be given as either a string or an `RE2::Regexp`:
+
+```ruby
+re = RE2('hel+o')
+RE2.replace("hello there", re, "yo") #=> "yo there"
+```
+
+To replace _all_ matches and not just the first, use [`RE2.global_replace`](https://mudge.name/re2/RE2.html#global_replace-class_method):
+
+```ruby
+RE2.global_replace("hallo thare", "a", "e")  #=> "hello there"
+```
+
+To extract matches with a given rewrite string including substitutions, use [`RE2.extract`](https://mudge.name/re2/RE2.html#extract-class_method):
+
+```ruby
+RE2.extract("alice@example.com", '(\w+)@(\w+)', '\2-\1')
+#=> "example-alice"
+```
+
+### Escaping
+
+To escape all potentially meaningful regexp characters in a string, use [`RE2.escape`](https://mudge.name/re2/RE2.html#escape-class_method):
+
+```ruby
+RE2.escape("1.5-2.0?") #=> "1\\.5\\-2\\.0\\?"
+```
+
 ### Encoding
 
 > [!WARNING]
@@ -247,50 +353,49 @@ the right encoding so this is the responsibility of the caller, e.g.
 
 ```ruby
 # By default, RE2 will process patterns and text as UTF-8
-RE2(non_utf8_pattern.encode("UTF-8")).match(non_utf8_text.encode("UTF-8"))
+RE2(non_utf8_pattern.encode("UTF-8")).partial_match(non_utf8_text.encode("UTF-8"))
 
 # If the :utf8 option is false, RE2 will process patterns and text as ISO-8859-1
-RE2(non_latin1_pattern.encode("ISO-8859-1"), utf8: false).match(non_latin1_text.encode("ISO-8859-1"))
+RE2(non_latin1_pattern.encode("ISO-8859-1"), utf8: false).partial_match(non_latin1_text.encode("ISO-8859-1"))
 ```
 
 ## Requirements
 
 This gem requires the following to run:
 
-* [Ruby](https://www.ruby-lang.org/en/) 2.6 to 3.3
+* [Ruby](https://www.ruby-lang.org/en/) 3.1 to 4.0
 
 It supports the following RE2 ABI versions:
 
-* libre2.0 (prior to release 2020-03-02) to libre2.11 (2023-07-01 to 2024-03-01)
+* libre2.0 (prior to release 2020-03-02) to libre2.11 (2023-07-01 to 2025-11-05)
 
 ### Native gems
 
 Where possible, a pre-compiled native gem will be provided for the following platforms:
 
 * Linux
-    * `aarch64-linux` and `arm-linux` (requires [glibc](https://www.gnu.org/software/libc/) 2.29+)
-    * `x86-linux` and `x86_64-linux` (requires [glibc](https://www.gnu.org/software/libc/) 2.17+)
-    * [musl](https://musl.libc.org/)-based systems such as [Alpine](https://alpinelinux.org) are supported as long as a [glibc-compatible library is installed](https://wiki.alpinelinux.org/wiki/Running_glibc_programs)
-* macOS `x86_64-darwin` and `arm64-darwin`
-* Windows `x64-mingw32` and `x64-mingw-ucrt`
+    * `aarch64-linux`, `arm-linux`, and `x86_64-linux` (requires [glibc](https://www.gnu.org/software/libc/) 2.29+, RubyGems 3.3.22+ and Bundler 2.3.21+)
+    * [musl](https://musl.libc.org/)-based systems such as [Alpine](https://alpinelinux.org) are supported with Bundler 2.5.6+
+* macOS 10.14+ `x86_64-darwin` and `arm64-darwin`
+* Windows 2022+ `x64-mingw-ucrt`
 
 ### Verifying the gems
 
 SHA256 checksums are included in the [release notes](https://github.com/mudge/re2/releases) for each version and can be checked with `sha256sum`, e.g.
 
 ```console
-$ gem fetch re2 -v 2.8.0
-Fetching re2-2.8.0-arm64-darwin.gem
-Downloaded re2-2.8.0-arm64-darwin
-$ sha256sum re2-2.8.0-arm64-darwin.gem
-962bad5c8de0757b059eff8922c3de31830a6710c418a83f2eab2b8253b8a0bf  re2-2.8.0-arm64-darwin.gem
+$ gem fetch re2 -v 2.18.0
+Fetching re2-2.18.0-arm64-darwin.gem
+Downloaded re2-2.18.0-arm64-darwin
+$ sha256sum re2-2.18.0-arm64-darwin.gem
+953063f0491420163d3484ed256fe2ff616c777ec66ee20aa5ec1a1a1fc39ff5  re2-2.18.0-arm64-darwin.gem
 ```
 
 [GPG](https://www.gnupg.org/) signatures are attached to each release (the assets ending in `.sig`) and can be verified if you import [our signing key `0x39AC3530070E0F75`](https://mudge.name/39AC3530070E0F75.asc) (or fetch it from a public keyserver, e.g. `gpg --keyserver keyserver.ubuntu.com --recv-key 0x39AC3530070E0F75`):
 
 ```console
-$ gpg --verify re2-2.8.0-arm64-darwin.gem.sig re2-2.8.0-arm64-darwin.gem
-gpg: Signature made Wed 31 Jan 15:02:06 2024 GMT
+$ gpg --verify re2-2.18.0-arm64-darwin.gem.sig re2-2.18.0-arm64-darwin.gem
+gpg: Signature made Sun  3 Aug 11:02:26 2025 BST
 gpg:                using RSA key 702609D9C790F45B577D7BEC39AC3530070E0F75
 gpg: Good signature from "Paul Mucur <mudge@mudge.name>" [unknown]
 gpg:                 aka "Paul Mucur <paul@ghostcassette.com>" [unknown]
@@ -328,9 +433,10 @@ You will need a full compiler toolchain for compiling Ruby C extensions (see
 [Nokogiri's "The Compiler
 Toolchain"](https://nokogiri.org/tutorials/installing_nokogiri.html#appendix-a-the-compiler-toolchain))
 plus the toolchain required for compiling the vendored version of RE2 and its
-dependency [Abseil][] which includes
-[CMake](https://cmake.org) and a compiler with C++14 support such as
-[clang](http://clang.llvm.org/) 3.4 or [gcc](https://gcc.gnu.org/) 5. On
+dependency [Abseil][] which includes [CMake](https://cmake.org), a compiler
+with C++17 support such as [clang](http://clang.llvm.org/) 5 or
+[gcc](https://gcc.gnu.org/) 8 and a recent version of
+[pkg-config](https://www.freedesktop.org/wiki/Software/pkg-config/). On
 Windows, you'll also need pkgconf 2.1.0+ to avoid [`undefined reference`
 errors](https://github.com/pkgconf/pkgconf/issues/322) when attempting to
 compile Abseil.
@@ -375,6 +481,8 @@ Alternatively, you can set the `RE2_USE_SYSTEM_LIBRARIES` environment variable i
   improvements in 2.4.0.
 * Thanks to [Manuel Jacob](https://github.com/manueljacob) for reporting a bug
   when passing strings with null bytes.
+* Thanks to [Maciej Gajewski](https://github.com/konieczkow) for helping
+  confirm issues with GC compaction and mutable strings.
 
 ## Contact
 

data/Rakefile

@@ -1,58 +1,43 @@
 # frozen_string_literal: true
 
 require 'rake/extensiontask'
-require 'rspec/core/rake_task'
 require 'rake_compiler_dock'
-require 'yaml'
+require 'rspec/core/rake_task'
 
 require_relative 'ext/re2/recipes'
 
-CLEAN.include FileList['**/*{.o,.so,.dylib,.bundle}'],
-              FileList['**/extconf.h'],
-              FileList['**/Makefile'],
-              FileList['pkg/']
-
-CLOBBER.include FileList['**/tmp'],
-                FileList['**/*.log'],
-                FileList['doc/**'],
-                FileList['tmp/']
-CLOBBER.add("ports/*").exclude(%r{ports/archives$})
-
-RE2_GEM_SPEC = Gem::Specification.load('re2.gemspec')
-
-task :prepare do
-  puts "Preparing project for gem building..."
-  recipes = load_recipes
-  recipes.each { |recipe| recipe.download }
-end
+re2_gemspec = Gem::Specification.load('re2.gemspec')
+abseil_recipe, re2_recipe = load_recipes
 
-task gem: :prepare
+# Add Abseil and RE2's latest archives to the gem files. (Note these will be
+# removed from the precompiled native gems.)
+abseil_archive = File.join("ports/archives", File.basename(abseil_recipe.files[0][:url]))
+re2_archive = File.join("ports/archives", File.basename(re2_recipe.files[0][:url]))
 
-Gem::PackageTask.new(RE2_GEM_SPEC) do |p|
-  p.need_zip = false
-  p.need_tar = false
-end
+re2_gemspec.files << abseil_archive
+re2_gemspec.files << re2_archive
 
-CROSS_RUBY_VERSIONS = %w[3.3.0 3.2.0 3.1.0 3.0.0 2.7.0 2.6.0].join(':')
-CROSS_RUBY_PLATFORMS = %w[
-  aarch64-linux
-  arm-linux
+cross_platforms = %w[
+  aarch64-linux-gnu
+  aarch64-linux-musl
+  arm-linux-gnu
+  arm-linux-musl
   arm64-darwin
   x64-mingw-ucrt
-  x64-mingw32
-  x86-linux
-  x86-mingw32
   x86_64-darwin
-  x86_64-linux
+  x86_64-linux-gnu
+  x86_64-linux-musl
 ].freeze
 
-ENV['RUBY_CC_VERSION'] = CROSS_RUBY_VERSIONS
+RakeCompilerDock.set_ruby_cc_version("~> 3.1", "~> 4.0")
+
+Gem::PackageTask.new(re2_gemspec).define
 
-Rake::ExtensionTask.new('re2', RE2_GEM_SPEC) do |e|
+Rake::ExtensionTask.new('re2', re2_gemspec) do |e|
   e.cross_compile = true
   e.cross_config_options << '--enable-cross-build'
   e.config_options << '--disable-system-libraries'
-  e.cross_platform = CROSS_RUBY_PLATFORMS
+  e.cross_platform = cross_platforms
   e.cross_compiling do |spec|
     spec.files.reject! { |path| File.fnmatch?('ports/*', path) }
     spec.dependencies.reject! { |dep| dep.name == 'mini_portile2' }
@@ -61,72 +46,49 @@ end
 
 RSpec::Core::RakeTask.new(:spec)
 
-namespace 'gem' do
-  def gem_builder(platform)
-    # use Task#invoke because the pkg/*gem task is defined at runtime
-    Rake::Task["native:#{platform}"].invoke
-    Rake::Task["pkg/#{RE2_GEM_SPEC.full_name}-#{Gem::Platform.new(platform)}.gem"].invoke
+begin
+  require 'ruby_memcheck'
+  require 'ruby_memcheck/rspec/rake_task'
+
+  namespace :spec do
+    RubyMemcheck::RSpec::RakeTask.new(valgrind: :compile)
   end
+rescue LoadError
+  # Only define the spec:valgrind task if ruby_memcheck is installed
+end
 
-  CROSS_RUBY_PLATFORMS.each do |platform|
-    # The Linux x86 image (ghcr.io/rake-compiler/rake-compiler-dock-image:1.3.0-mri-x86_64-linux)
-    # is based on CentOS 7 and has two versions of cmake installed:
-    # a 2.8 version in /usr/bin and a 3.25 in /usr/local/bin. The latter is needed by abseil.
-    cmake =
-      case platform
-      when 'x86_64-linux', 'x86-linux'
-        '/usr/local/bin/cmake'
-      else
-        'cmake'
-      end
-
-    desc "build native gem for #{platform} platform"
+namespace :gem do
+  cross_platforms.each do |platform|
+
+    # Compile each platform's native gem, packaging up the result. Note we add
+    # /usr/local/bin to the PATH as it contains the newest version of CMake in
+    # the rake-compiler-dock images.
+    desc "Compile and build native gem for #{platform} platform"
     task platform do
       RakeCompilerDock.sh <<~SCRIPT, platform: platform, verbose: true
+        wget -O - https://apt.kitware.com/keys/kitware-archive-latest.asc 2>/dev/null | gpg --dearmor - | sudo tee /usr/share/keyrings/kitware-archive-keyring.gpg >/dev/null &&
+        echo 'deb [signed-by=/usr/share/keyrings/kitware-archive-keyring.gpg] https://apt.kitware.com/ubuntu/ focal main' | sudo tee /etc/apt/sources.list.d/kitware.list >/dev/null &&
+        sudo apt-get update &&
+        sudo apt-get install -y cmake=3.22.2-0kitware1ubuntu20.04.1 cmake-data=3.22.2-0kitware1ubuntu20.04.1 &&
         gem install bundler --no-document &&
-        bundle &&
-        bundle exec rake gem:#{platform}:builder CMAKE=#{cmake}
+        bundle install &&
+        bundle exec rake native:#{platform} pkg/#{re2_gemspec.full_name}-#{Gem::Platform.new(platform)}.gem PATH="/usr/local/bin:$PATH"
       SCRIPT
     end
-
-    namespace platform do
-      desc "build native gem for #{platform} platform (guest container)"
-      task 'builder' do
-        gem_builder(platform)
-      end
-    end
   end
-
-  desc 'build all native gems'
-  multitask 'native' => CROSS_RUBY_PLATFORMS
-end
-
-def add_file_to_gem(relative_source_path)
-  dest_path = File.join(gem_build_path, relative_source_path)
-  dest_dir = File.dirname(dest_path)
-
-  mkdir_p dest_dir unless Dir.exist?(dest_dir)
-  rm_f dest_path if File.exist?(dest_path)
-  safe_ln relative_source_path, dest_path
-
-  RE2_GEM_SPEC.files << relative_source_path
 end
 
-def gem_build_path
-  File.join 'pkg', RE2_GEM_SPEC.full_name
+# Set up file tasks for Abseil and RE2's archives so they are automatically
+# downloaded when required by the gem task.
+file abseil_archive do
+  abseil_recipe.download
 end
 
-def add_vendored_libraries
-  dependencies = YAML.load_file(File.join(File.dirname(__FILE__), 'dependencies.yml'))
-  abseil_archive = File.join('ports', 'archives', "#{dependencies['abseil']['version']}.tar.gz")
-  libre2_archive = File.join('ports', 'archives', "re2-#{dependencies['libre2']['version']}.tar.gz")
-
-  add_file_to_gem(abseil_archive)
-  add_file_to_gem(libre2_archive)
+file re2_archive do
+  re2_recipe.download
 end
 
-task gem_build_path do
-  add_vendored_libraries
-end
+task default: :spec
 
-task default: [:compile, :spec]
+CLEAN.add("lib/**/*.{o,so,bundle}", "pkg")
+CLOBBER.add("ports")

data/dependencies.yml

@@ -1,7 +1,7 @@
 ---
 libre2:
-  version: '2024-03-01'
-  sha256: 7b2b3aa8241eac25f674e5b5b2e23d4ac4f0a8891418a2661869f736f03f57f4
+  version: '2025-11-05'
+  sha256: 87f6029d2f6de8aa023654240a03ada90e876ce9a4676e258dd01ea4c26ffd67
 abseil:
-  version: '20240116.1'
-  sha256: 3c743204df78366ad2eaf236d6631d83f6bc928d1705dd0000b872e53b73dc6a
+  version: '20260107.1'
+  sha256: 4314e2a7cbac89cac25a2f2322870f343d81579756ceff7f431803c2c9090195