Reintroduce Regexp mutations

This reverts commit 21d3fef. This was not a clean revert. Note that: - The version of `regexp_parser` was 1.3.0, now it is 1.8.2 to accomodate our current `rubocop` version and because there were some relevant bugfixes implemented between 1.3.x and 1.8.x. We should eventually move to 2.0 but it is currently incompatible with this integration. There are some issues with the frozen Regexp classes getting mutated so we may have to open an issue. - Since "expected exception" support was removed from the specs, I have had to exclude two files entirely. This seems unfortunate as it reduces our overall coverage. - Since unsupported nodes are no longer explicitly tracked, I removed the code that used to handle that for regular expressions. See: #1021 - I had to change the example case for where we are more permissive than `regexp_parser` because `regexp_parser` has decided to become more permissive and try to match Ruby's semantics. It was actually very hard to find a case that failed--I brute-forced 50 million regexp strings that had perfect parity of being accepted and then stumbled onto the single hex escape case by accident. See: ammar/regexp_parser#75 - Changed an access pattern for regexp mutations which became equivalent based on this: https://github.com/ammar/regexp_parser/blame/4ca7cec03b210e3e00473b7b1a7308f963190c1e/lib/regexp_parser/expression/subexpression.rb#L30-L33 - Some other minor conflicts and small spec assertion changes were resolved as well.
mbj · Dec 20, 2020 · 824da89 · 824da89
1 parent fc8c82c
commit 824da89
Show file tree

Hide file tree

Showing 43 changed files with 2,003 additions and 18 deletions.
diff --git a/Changelog.md b/Changelog.md
@@ -1,3 +1,7 @@
+# Unreleased
+
+* Reintroduce regexp mutation support [#1166](https://github.com/mbj/mutant/pull/1166)
+
 # v0.10.20 2020-12-16
 
   [#1159](https://github.com/mbj/mutant/pull/1159)

diff --git a/Gemfile.lock b/Gemfile.lock
@@ -14,6 +14,7 @@ PATH
       mprelude (~> 0.1.0)
       parser (~> 2.7.1)
       procto (~> 0.0.2)
+      regexp_parser (~> 1.8)
       unparser (~> 0.5.4)
       variable (~> 0.0.1)
 
@@ -51,7 +52,7 @@ GEM
       ast (~> 2.4.1)
     procto (0.0.3)
     rainbow (3.0.0)
-    regexp_parser (2.0.0)
+    regexp_parser (1.8.2)
     rexml (3.2.4)
     rspec (3.10.0)
       rspec-core (~> 3.10.0)

diff --git a/lib/mutant.rb b/lib/mutant.rb
@@ -17,6 +17,7 @@
 require 'parser'
 require 'parser/current'
 require 'pathname'
+require 'regexp_parser'
 require 'set'
 require 'singleton'
 require 'stringio'
@@ -52,6 +53,15 @@ module Mutant
 require 'mutant/ast/named_children'
 require 'mutant/ast/node_predicates'
 require 'mutant/ast/find_metaclass_containing'
+require 'mutant/ast/regexp'
+require 'mutant/ast/regexp/transformer'
+require 'mutant/ast/regexp/transformer/direct'
+require 'mutant/ast/regexp/transformer/named_group'
+require 'mutant/ast/regexp/transformer/options_group'
+require 'mutant/ast/regexp/transformer/quantifier'
+require 'mutant/ast/regexp/transformer/recursive'
+require 'mutant/ast/regexp/transformer/root'
+require 'mutant/ast/regexp/transformer/text'
 require 'mutant/ast/meta'
 require 'mutant/ast/meta/send'
 require 'mutant/ast/meta/const'
@@ -74,6 +84,13 @@ module Mutant
 require 'mutant/mutator/util/symbol'
 require 'mutant/mutator/node'
 require 'mutant/mutator/node/generic'
+require 'mutant/mutator/node/regexp'
+require 'mutant/mutator/node/regexp/alternation_meta'
+require 'mutant/mutator/node/regexp/capture_group'
+require 'mutant/mutator/node/regexp/character_type'
+require 'mutant/mutator/node/regexp/end_of_line_anchor'
+require 'mutant/mutator/node/regexp/end_of_string_or_before_end_of_line_anchor'
+require 'mutant/mutator/node/regexp/greedy_zero_or_more'
 require 'mutant/mutator/node/literal'
 require 'mutant/mutator/node/literal/boolean'
 require 'mutant/mutator/node/literal/range'

diff --git a/lib/mutant/ast/regexp.rb b/lib/mutant/ast/regexp.rb
@@ -0,0 +1,43 @@
+# frozen_string_literal: true
+
+module Mutant
+  module AST
+    # Regexp source mapper
+    module Regexp
+      # Parse regex string into expression
+      #
+      # @param regexp [String]
+      #
+      # @return [Regexp::Expression, nil]
+      #
+      # rubocop:disable Lint/SuppressedException
+      def self.parse(regexp)
+        ::Regexp::Parser.parse(regexp)
+      # `regexp_parser` is more strict than MRI
+      # See: https://github.com/ammar/regexp_parser/issues/75
+      rescue ::Regexp::Scanner::PrematureEndError
+      end
+      # rubocop:enable Lint/SuppressedException
+
+      # Convert expression into ast node
+      #
+      # @param expression [Regexp::Expression]
+      #
+      # @return [Parser::AST::Node]
+      def self.to_ast(expression)
+        ast_type = :"regexp_#{expression.token}_#{expression.type}"
+
+        Transformer.lookup(ast_type).to_ast(expression)
+      end
+
+      # Convert node into expression
+      #
+      # @param node [Parser::AST::Node]
+      #
+      # @return [Regexp::Expression]
+      def self.to_expression(node)
+        Transformer.lookup(node.type).to_expression(node)
+      end
+    end # Regexp
+  end # AST
+end # Mutant
diff --git a/lib/mutant/ast/regexp/transformer.rb b/lib/mutant/ast/regexp/transformer.rb
@@ -0,0 +1,150 @@
+# frozen_string_literal: true
+
+module Mutant
+  module AST
+    module Regexp
+      # Regexp bijective mapper
+      #
+      # Transforms parsed regular expression representation from
+      # `Regexp::Expression` instances (provided by `regexp_parser`) into
+      # equivalent representations using `Parser::AST::Node`
+      class Transformer
+        include AbstractType
+
+        REGISTRY = Registry.new
+
+        # Lookup transformer class for regular expression node type
+        #
+        # @param type [Symbol]
+        #
+        # @return [Class<Transformer>]
+        def self.lookup(type)
+          REGISTRY.lookup(type)
+        end
+
+        def self.register(type)
+          REGISTRY.register(type, self)
+        end
+        private_class_method :register
+
+        # Transform expression
+        #
+        # @param expression [Regexp::Expression]
+        #
+        # @return [Parser::AST::Node]
+        def self.to_ast(expression)
+          self::ExpressionToAST.call(expression)
+        end
+
+        # Transform node
+        #
+        # @param node [Parser::AST::Node]
+        #
+        # @return [Regexp::Expression]
+        def self.to_expression(node)
+          self::ASTToExpression.call(node)
+        end
+
+        # Abstract expression transformer
+        class ExpressionToAST
+          PREFIX = :regexp
+
+          include Concord.new(:expression), Procto.call, AST::Sexp, AbstractType, Adamantium
+
+        private
+
+          def ast(*children)
+            s(type, *children)
+          end
+
+          def quantify(node)
+            return node unless expression.quantified?
+
+            Quantifier.to_ast(expression.quantifier).append(node)
+          end
+
+          def children
+            expression.map(&Regexp.public_method(:to_ast))
+          end
+
+          def type
+            :"#{PREFIX}_#{expression.token}_#{expression.type}"
+          end
+        end # ExpressionToAST
+
+        # Abstract node transformer
+        class ASTToExpression
+          include Concord.new(:node), Procto.call, AbstractType, Adamantium
+
+          # Call generic transform method and freeze result
+          #
+          # @return [Regexp::Expression]
+          def call
+            transform.freeze
+          end
+
+        private
+
+          abstract_method :transform
+
+          def subexpressions
+            node.children.map(&Regexp.public_method(:to_expression))
+          end
+        end # ASTToExpression
+
+        # Mixin for node transformers
+        #
+        # Helps construct a mapping from Parser::AST::Node domain to
+        # Regexp::Expression domain
+        module LookupTable
+          Mapping = Class.new.include(Concord::Public.new(:token, :regexp_class))
+
+          # Table mapping ast types to object information for regexp domain
+          class Table
+
+            # Coerce array of mapping information into structured table
+            #
+            # @param [Array(Symbol, Array, Class<Regexp::Expression>)]
+            #
+            # @return [Table]
+            def self.create(*rows)
+              table = rows.map do |ast_type, token, klass|
+                [ast_type, Mapping.new(::Regexp::Token.new(*token), klass)]
+              end.to_h
+
+              new(table)
+            end
+
+            include Concord.new(:table), Adamantium
+
+            # Types defined by the table
+            #
+            # @return [Array<Symbol>]
+            def types
+              table.keys
+            end
+
+            # Lookup mapping information given an ast node type
+            #
+            # @param type [Symbol]
+            #
+            # @return [Mapping]
+            def lookup(type)
+              table.fetch(type)
+            end
+          end # Table
+
+        private
+
+          def expression_token
+            self.class::TABLE.lookup(node.type).token
+          end
+
+          def expression_class
+            self.class::TABLE.lookup(node.type).regexp_class
+          end
+        end # LookupTable
+      end # Transformer
+    end # Regexp
+  end # AST
+end # Mutant
diff --git a/lib/mutant/ast/regexp/transformer/direct.rb b/lib/mutant/ast/regexp/transformer/direct.rb
@@ -0,0 +1,121 @@
+# frozen_string_literal: true
+
+module Mutant
+  module AST
+    module Regexp
+      class Transformer
+        # Transformer for nodes which map directly to other domain
+        #
+        # A node maps "directly" to another domain if the node never
+        # has children or text which needs to be preserved for a mapping
+        #
+        # @example direct mapping
+        #
+        #     input      = /\d/
+        #     expression = Regexp::Parser.parse(input).first
+        #     node       = Transformer::Direct.to_ast(expression)
+        #
+        #     # the digit type always has the same text and no children
+        #     expression.text      # => "\\d"
+        #     expression.terminal? # => true
+        #
+        #     # therefore the `Parser::AST::Node` is always the same
+        #     node # => s(:regexp_digit_type)
+        class Direct < self
+          # Mapper from `Regexp::Expression` to `Parser::AST::Node`
+          class ExpressionToAST < Transformer::ExpressionToAST
+            # Transform expression into node
+            #
+            # @return [Parser::AST::Node]
+            def call
+              quantify(ast)
+            end
+          end # ExpressionToAST
+
+          # Mapper from `Parser::AST::Node` to `Regexp::Expression`
+          class ASTToExpression < Transformer::ASTToExpression
+            include LookupTable
+
+            # rubocop:disable Layout/LineLength
+            TABLE = Table.create(
+              [:regexp_alnum_posixclass,         [:posixclass,    :alnum,            '[:alnum:]'],    ::Regexp::Expression::PosixClass],
+              [:regexp_alpha_posixclass,         [:posixclass,    :alpha,            '[:alpha:]'],    ::Regexp::Expression::PosixClass],
+              [:regexp_alpha_property,           [:property,      :alpha,            '\p{Alpha}'],    ::Regexp::Expression::UnicodeProperty::Alpha],
+              [:regexp_alternation_escape,       [:escape,        :alternation,      '\|'],           ::Regexp::Expression::EscapeSequence::Literal],
+              [:regexp_arabic_property,          [:property,      :arabic,           '\p{Arabic}'],   ::Regexp::Expression::UnicodeProperty::Script],
+              [:regexp_ascii_posixclass,         [:posixclass,    :ascii,            '[:ascii:]'],    ::Regexp::Expression::PosixClass],
+              [:regexp_backspace_escape,         [:escape,        :backspace,        '\b'],           ::Regexp::Expression::EscapeSequence::Backspace],
+              [:regexp_bell_escape,              [:escape,        :bell,             '\a'],           ::Regexp::Expression::EscapeSequence::Literal],
+              [:regexp_blank_posixclass,         [:posixclass,    :blank,            '[:blank:]'],    ::Regexp::Expression::PosixClass],
+              [:regexp_bol_anchor,               [:anchor,        :bol,              '^'],            ::Regexp::Expression::Anchor::BeginningOfLine],
+              [:regexp_bol_escape,               [:escape,        :bol,              '\^'],           ::Regexp::Expression::EscapeSequence::Literal],
+              [:regexp_bos_anchor,               [:anchor,        :bos,              '\\A'],          ::Regexp::Expression::Anchor::BeginningOfString],
+              [:regexp_carriage_escape,          [:escape,        :carriage,         '\r'],           ::Regexp::Expression::EscapeSequence::Literal],
+              [:regexp_cntrl_posixclass,         [:posixclass,    :cntrl,            '[:cntrl:]'],    ::Regexp::Expression::PosixClass],
+              [:regexp_digit_posixclass,         [:posixclass,    :digit,            '[:digit:]'],    ::Regexp::Expression::PosixClass],
+              [:regexp_digit_type,               [:type,          :digit,            '\d'],           ::Regexp::Expression::CharacterType::Digit],
+              [:regexp_dot_escape,               [:escape,        :dot,              '\.'],           ::Regexp::Expression::EscapeSequence::Literal],
+              [:regexp_dot_meta,                 [:meta,          :dot,              '.'],            ::Regexp::Expression::CharacterType::Any],
+              [:regexp_eol_anchor,               [:anchor,        :eol,              '$'],            ::Regexp::Expression::Anchor::EndOfLine],
+              [:regexp_eol_escape,               [:escape,        :eol,              '\$'],           ::Regexp::Expression::EscapeSequence::Literal],
+              [:regexp_eos_anchor,               [:anchor,        :eos,              '\\z'],          ::Regexp::Expression::Anchor::EndOfString],
+              [:regexp_eos_ob_eol_anchor,        [:anchor,        :eos_ob_eol,       '\\Z'],          ::Regexp::Expression::Anchor::EndOfStringOrBeforeEndOfLine],
+              [:regexp_escape_escape,            [:escape,        :escape,           '\e'],           ::Regexp::Expression::EscapeSequence::AsciiEscape],
+              [:regexp_form_feed_escape,         [:escape,        :form_feed,        '\f'],           ::Regexp::Expression::EscapeSequence::FormFeed],
+              [:regexp_graph_posixclass,         [:posixclass,    :graph,            '[:graph:]'],    ::Regexp::Expression::PosixClass],
+              [:regexp_group_close_escape,       [:escape,        :group_close,      '\)'],           ::Regexp::Expression::EscapeSequence::Literal],
+              [:regexp_group_open_escape,        [:escape,        :group_open,       '\('],           ::Regexp::Expression::EscapeSequence::Literal],
+              [:regexp_han_property,             [:property,      :han,              '\p{Han}'],      ::Regexp::Expression::UnicodeProperty::Script],
+              [:regexp_hangul_property,          [:property,      :hangul,           '\p{Hangul}'],   ::Regexp::Expression::UnicodeProperty::Script],
+              [:regexp_hex_type,                 [:type,          :hex,              '\h'],           ::Regexp::Expression::CharacterType::Hex],
+              [:regexp_hiragana_property,        [:property,      :hiragana,         '\p{Hiragana}'], ::Regexp::Expression::UnicodeProperty::Script],
+              [:regexp_interval_close_escape,    [:escape,        :interval_close,   '\}'],           ::Regexp::Expression::EscapeSequence::Literal],
+              [:regexp_interval_open_escape,     [:escape,        :interval_open,    '\{'],           ::Regexp::Expression::EscapeSequence::Literal],
+              [:regexp_katakana_property,        [:property,      :katakana,         '\p{Katakana}'], ::Regexp::Expression::UnicodeProperty::Script],
+              [:regexp_letter_property,          [:property,      :letter,           '\p{L}'],        ::Regexp::Expression::UnicodeProperty::Letter::Any],
+              [:regexp_linebreak_type,           [:type,          :linebreak,        '\R'],           ::Regexp::Expression::CharacterType::Linebreak],
+              [:regexp_lower_posixclass,         [:posixclass,    :lower,            '[:lower:]'],    ::Regexp::Expression::PosixClass],
+              [:regexp_mark_keep,                [:keep,          :mark,             '\K'],           ::Regexp::Expression::Keep::Mark],
+              [:regexp_match_start_anchor,       [:anchor,        :match_start,      '\\G'],          ::Regexp::Expression::Anchor::MatchStart],
+              [:regexp_newline_escape,           [:escape,        :newline,          '\n'],           ::Regexp::Expression::EscapeSequence::Literal],
+              [:regexp_nondigit_type,            [:type,          :nondigit,         '\D'],           ::Regexp::Expression::CharacterType::NonDigit],
+              [:regexp_nonhex_type,              [:type,          :nonhex,           '\H'],           ::Regexp::Expression::CharacterType::NonHex],
+              [:regexp_nonspace_type,            [:type,          :nonspace,         '\S'],           ::Regexp::Expression::CharacterType::NonSpace],
+              [:regexp_nonword_boundary_anchor,  [:anchor,        :nonword_boundary, '\\B'],          ::Regexp::Expression::Anchor::NonWordBoundary],
+              [:regexp_nonword_type,             [:type,          :nonword,          '\W'],           ::Regexp::Expression::CharacterType::NonWord],
+              [:regexp_one_or_more_escape,       [:escape,        :one_or_more,      '\+'],           ::Regexp::Expression::EscapeSequence::Literal],
+              [:regexp_print_nonposixclass,      [:nonposixclass, :print,            '[:^print:]'],   ::Regexp::Expression::PosixClass],
+              [:regexp_print_nonproperty,        [:nonproperty,   :print,            '\P{Print}'],    ::Regexp::Expression::UnicodeProperty::Print],
+              [:regexp_print_posixclass,         [:posixclass,    :print,            '[:print:]'],    ::Regexp::Expression::PosixClass],
+              [:regexp_print_posixclass,         [:posixclass,    :print,            '[:print:]'],    ::Regexp::Expression::PosixClass],
+              [:regexp_print_property,           [:property,      :print,            '\p{Print}'],    ::Regexp::Expression::UnicodeProperty::Print],
+              [:regexp_punct_posixclass,         [:posixclass,    :punct,            '[:punct:]'],    ::Regexp::Expression::PosixClass],
+              [:regexp_set_close_escape,         [:escape,        :set_close,        '\]'],           ::Regexp::Expression::EscapeSequence::Literal],
+              [:regexp_set_open_escape,          [:escape,        :set_open,         '\['],           ::Regexp::Expression::EscapeSequence::Literal],
+              [:regexp_space_posixclass,         [:posixclass,    :space,            '[:space:]'],    ::Regexp::Expression::PosixClass],
+              [:regexp_space_type,               [:type,          :space,            '\s'],           ::Regexp::Expression::CharacterType::Space],
+              [:regexp_upper_posixclass,         [:posixclass,    :upper,            '[:upper:]'],    ::Regexp::Expression::PosixClass],
+              [:regexp_vertical_tab_escape,      [:escape,        :vertical_tab,     '\v'],           ::Regexp::Expression::EscapeSequence::VerticalTab],
+              [:regexp_word_boundary_anchor,     [:anchor,        :word_boundary,    '\b'],           ::Regexp::Expression::Anchor::WordBoundary],
+              [:regexp_word_posixclass,          [:posixclass,    :word,             '[:word:]'],     ::Regexp::Expression::PosixClass],
+              [:regexp_word_type,                [:type,          :word,             '\w'],           ::Regexp::Expression::CharacterType::Word],
+              [:regexp_xdigit_posixclass,        [:posixclass,    :xdigit,           '[:xdigit:]'],   ::Regexp::Expression::PosixClass],
+              [:regexp_xgrapheme_type,           [:type,          :xgrapheme,        '\X'],           ::Regexp::Expression::CharacterType::ExtendedGrapheme],
+              [:regexp_zero_or_more_escape,      [:escape,        :zero_or_more,     '\*'],           ::Regexp::Expression::EscapeSequence::Literal],
+              [:regexp_zero_or_one_escape,       [:escape,        :zero_or_one,      '\?'],           ::Regexp::Expression::EscapeSequence::Literal]
+            )
+          # rubocop:enable Layout/LineLength
+
+          private
+
+            def transform
+              expression_class.new(expression_token)
+            end
+          end # ASTToExpression
+
+          ASTToExpression::TABLE.types.each(&method(:register))
+        end # Direct
+      end # Transformer
+    end # Regexp
+  end # AST
+end # Mutant