regexp_parser 2.2.0 → 2.3.1

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 332259c898b9b344e10961053bb2b761f4dd5530182a5f6195639dba9cbb99f9
-  data.tar.gz: b537f9bd23db799ee562494633f1e8423501651540a04b634ae07dfe8f3b19c3
+  metadata.gz: 6dedba6b051b22dd917febd957e35e6800b37af958780e331007de1f59b2b466
+  data.tar.gz: 55fa98afa6031a38cac2045f8c592290fad9a4e500597277f3f79be75f1076f3
 SHA512:
-  metadata.gz: 393ecc1cc20189e4a79252e6acf6dab7dd6dc07ba9c47ae7479746eaf8ebe2ccfd1ebcb82fd027edc2c5c938eb490f2f36a93587d2405a54017e0e2727a35a15
-  data.tar.gz: 6c961232ce5f3f409c91d0b66dd23c809e92f47aa6c1f94f2f1929e8eeccfb4bc25fcdf5935fc968d7e0c0ae632992a6d38bc8e982858f2da996a8eac54d3c89
+  metadata.gz: 6b2b463f3a28450527691d90bcfc8901b815bd4a32e88bcdee1f95db4588599993766687a4c7a1b785b450c0f0025039caf07bb93aaf1013fa365e4ed16fc040
+  data.tar.gz: 8aabccda06bb1f20485610076ad679a0d62e8630f992c55be7158c01f3b8b6dd6eb44a206f71b613c0335db13789d8e21d652973ecfb7c262c25a6d54e35a371

data/CHANGELOG.md

@@ -1,10 +1,43 @@
 ## [Unreleased]
 
+## [2.3.1] - 2022-04-24 - [Janosch Müller](mailto:janosch84@gmail.com)
+
+### Fixed
+
+- removed five inexistent unicode properties from `Syntax#features`
+  - these were never supported by Ruby but incorrectly accepted by the parser
+  - thanks to [Markus Schirp](https://github.com/mbj) for the report
+
+## [2.3.0] - 2022-04-08 - [Janosch Müller](mailto:janosch84@gmail.com)
+
+### Added
+
+- improved parsing performance through `Syntax` refactoring
+  - instead of fresh `Syntax` instances, pre-loaded constants are now re-used
+  - this approximately doubles the parsing speed for simple regexps
+- added methods to `Syntax` classes to show relative feature sets
+  - e.g. `Regexp::Syntax::V3_2_0.added_features`
+- support for new unicode properties of Ruby 3.2 / Unicode 14.0
+
+## [2.2.1] - 2022-02-11 - [Janosch Müller](mailto:janosch84@gmail.com)
+
+### Fixed
+
+- fixed Syntax version of absence groups (`(?~...)`)
+  - the lexer accepted them for any Ruby version
+  - now they are only recognized for Ruby >= 2.4.1 in which they were introduced
+- reduced gem size by excluding specs from package
+- removed deprecated `test_files` gemspec setting
+- no longer depend on `yaml`/`psych` (except for Ruby <= 2.4)
+- no longer depend on `set`
+  - `set` was removed from the stdlib and made a standalone gem as of Ruby 3
+  - this made it a hidden/undeclared dependency of `regexp_parser`
+
 ## [2.2.0] - 2021-12-04 - [Janosch Müller](mailto:janosch84@gmail.com)
 
 ### Added
 
-- Added support for 13 new unicode properties introduced in Ruby 3.1.0-dev
+- added support for 13 new unicode properties introduced in Ruby 3.1.0
 
 ## [2.1.1] - 2021-02-23 - [Janosch Müller](mailto:janosch84@gmail.com)
 

data/Gemfile

@@ -5,9 +5,10 @@ gemspec
 group :development, :test do
   gem 'ice_nine', '~> 0.11.2'
   gem 'rake', '~> 13.0'
-  gem 'regexp_property_values', '~> 1.0'
+  gem 'regexp_property_values', '~> 1.3'
   gem 'rspec', '~> 3.10'
   if RUBY_VERSION.to_f >= 2.7
+    gem 'benchmark-ips', '~> 2.1'
     gem 'gouteur'
     gem 'rubocop', '~> 1.7'
   end

data/LICENSE

@@ -1,4 +1,4 @@
-Copyright (c) 2010, 2012-2015,  Ammar Ali
+Copyright (c) 2010, 2012-2022,  Ammar Ali
 
 Permission is hereby granted, free of charge, to any person
 obtaining a copy of this software and associated documentation

data/README.md

@@ -157,31 +157,41 @@ flavor). Syntax classes act as lookup tables, and are layered to create
 flavor variations. Syntax only comes into play in the lexer.
 
 #### Example
-The following instantiates syntax objects for Ruby 2.0, 1.9, 1.8, and
+The following fetches syntax objects for Ruby 2.0, 1.9, 1.8, and
 checks a few of their implementation features.
 
 ```ruby
 require 'regexp_parser'
 
-ruby_20 = Regexp::Syntax.new 'ruby/2.0'
+ruby_20 = Regexp::Syntax.for 'ruby/2.0'
 ruby_20.implements? :quantifier,  :zero_or_one             # => true
 ruby_20.implements? :quantifier,  :zero_or_one_reluctant   # => true
 ruby_20.implements? :quantifier,  :zero_or_one_possessive  # => true
 ruby_20.implements? :conditional, :condition               # => true
 
-ruby_19 = Regexp::Syntax.new 'ruby/1.9'
+ruby_19 = Regexp::Syntax.for 'ruby/1.9'
 ruby_19.implements? :quantifier,  :zero_or_one             # => true
 ruby_19.implements? :quantifier,  :zero_or_one_reluctant   # => true
 ruby_19.implements? :quantifier,  :zero_or_one_possessive  # => true
 ruby_19.implements? :conditional, :condition               # => false
 
-ruby_18 = Regexp::Syntax.new 'ruby/1.8'
+ruby_18 = Regexp::Syntax.for 'ruby/1.8'
 ruby_18.implements? :quantifier,  :zero_or_one             # => true
 ruby_18.implements? :quantifier,  :zero_or_one_reluctant   # => true
 ruby_18.implements? :quantifier,  :zero_or_one_possessive  # => false
 ruby_18.implements? :conditional, :condition               # => false
 ```
 
+Syntax objects can also be queried about their complete and relative feature sets.
+
+```ruby
+require 'regexp_parser'
+
+ruby_20 = Regexp::Syntax.for 'ruby/2.0' # => Regexp::Syntax::V2_0_0
+ruby_20.added_features                  # => { conditional: [...], ... }
+ruby_20.removed_features                # => { property: [:newline], ... }
+ruby_20.features                        # => { anchor: [...], ... }
+```
 
 #### Notes
   * Variations on a token, for example a named group with angle brackets (< and >)
@@ -437,7 +447,7 @@ Projects using regexp_parser.
 
 - [meta_re](https://github.com/ammar/meta_re) is a regular expression preprocessor with alias support.
 
-- [mutant](https://github.com/mbj/mutant) (before v0.9.0) manipulates your regular expressions (amongst others) to see if your tests cover their behavior.
+- [mutant](https://github.com/mbj/mutant) manipulates your regular expressions (amongst others) to see if your tests cover their behavior.
 
 - [rubocop](https://github.com/rubocop-hq/rubocop) is a linter for Ruby that uses regexp_parser to lint Regexps.
 
@@ -470,4 +480,4 @@ Documentation and books used while working on this project.
 
 ---
 ##### Copyright
-_Copyright (c) 2010-2020 Ammar Ali. See LICENSE file for details._
+_Copyright (c) 2010-2022 Ammar Ali. See LICENSE file for details._

data/Rakefile

@@ -5,9 +5,7 @@ require 'rake'
 require 'rake/testtask'
 require 'rspec/core/rake_task'
 
-RAGEL_SOURCE_DIR = File.join(__dir__, 'lib/regexp_parser/scanner')
-RAGEL_OUTPUT_DIR = File.join(__dir__, 'lib/regexp_parser')
-RAGEL_SOURCE_FILES = %w{scanner} # scanner.rl includes property.rl
+Dir['tasks/**/*.rake'].each { |file| load(file) }
 
 Bundler::GemHelper.install_tasks
 
@@ -19,63 +17,7 @@ namespace :test do
   task full: [:'ragel:rb', :spec]
 end
 
-namespace :ragel do
-  desc "Process the ragel source files and output ruby code"
-  task :rb do
-    RAGEL_SOURCE_FILES.each do |source_file|
-      output_file = "#{RAGEL_OUTPUT_DIR}/#{source_file}.rb"
-      # using faster flat table driven FSM, about 25% larger code, but about 30% faster
-      sh "ragel -F1 -R #{RAGEL_SOURCE_DIR}/#{source_file}.rl -o #{output_file}"
-
-      contents = File.read(output_file)
-
-      File.open(output_file, 'r+') do |file|
-        contents = "# -*- warn-indent:false;  -*-\n" + contents
-
-        file.write(contents)
-      end
-    end
-  end
-
-  desc "Delete the ragel generated source file(s)"
-  task :clean do
-    RAGEL_SOURCE_FILES.each do |file|
-      sh "rm -f #{RAGEL_OUTPUT_DIR}/#{file}.rb"
-    end
-  end
-end
-
 # Add ragel task as a prerequisite for building the gem to ensure that the
 # latest scanner code is generated and included in the build.
 desc "Runs ragel:rb before building the gem"
 task :build => ['ragel:rb']
-
-namespace :props do
-  desc 'Write new property value hashes for the properties scanner'
-  task :update do
-    require 'regexp_property_values'
-    RegexpPropertyValues.update
-    dir = File.join(__dir__, 'lib/regexp_parser/scanner/properties')
-
-    require 'psych'
-    write_hash_to_file = ->(hash, path) do
-      File.open(path, 'w') do |f|
-        f.puts '#',
-               "# THIS FILE IS AUTO-GENERATED BY `rake props:update`, DO NOT EDIT",
-               '#',
-               hash.sort.to_h.to_yaml
-      end
-      puts "Wrote #{hash.count} aliases to `#{path}`"
-    end
-
-    long_names_to_tokens = RegexpPropertyValues.all.map do |val|
-      [val.identifier, val.full_name.downcase]
-    end
-    write_hash_to_file.call(long_names_to_tokens, "#{dir}/long.yml")
-
-    short_names_to_tokens = RegexpPropertyValues.alias_hash.map do |k, v|
-      [k.identifier, v.full_name.downcase]
-    end
-    write_hash_to_file.call(short_names_to_tokens, "#{dir}/short.yml")
-  end
-end

data/lib/regexp_parser/expression/classes/escape_sequence.rb

@@ -2,16 +2,21 @@ module Regexp::Expression
   # TODO: unify naming with Token::Escape, on way or the other, in v3.0.0
   module EscapeSequence
     class Base < Regexp::Expression::Base
-      require 'yaml'
-
-      def char
-        # poor man's unescape without using eval
-        YAML.load(%Q(---\n"#{text}"\n))
-      end
-
       def codepoint
         char.ord
       end
+
+      if ''.respond_to?(:undump)
+        def char
+          %("#{text}").undump
+        end
+      else
+        # poor man's unescape without using eval
+        require 'yaml'
+        def char
+          YAML.load(%Q(---\n"#{text}"\n))
+        end
+      end
     end
 
     class Literal < EscapeSequence::Base

data/lib/regexp_parser/expression/methods/strfregexp.rb

@@ -43,7 +43,7 @@ module Regexp::Expression
 
       # Order is important! Fields that use other fields in their
       # definition must appear before the fields they use.
-      part_keys = %w{a m b o i l x s e S y k c q Q z Z t ~t T >}
+      part_keys = %w[a m b o i l x s e S y k c q Q z Z t ~t T >]
       part.keys.each {|k| part[k] = "<?#{k}?>"}
 
       part['>'] = print_level ? ('  ' * (print_level + indent_offset)) : ''

data/lib/regexp_parser/lexer.rb

@@ -18,7 +18,7 @@ class Regexp::Lexer
   end
 
   def lex(input, syntax = "ruby/#{RUBY_VERSION}", options: nil, &block)
-    syntax = Regexp::Syntax.new(syntax)
+    syntax = Regexp::Syntax.for(syntax)
 
     self.tokens = []
     self.nesting = 0