regexp_parser 2.10.0 → 2.11.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: f6ed5457d89738fa1076cf3875cd2d009973f02857ea68e055ef3ef74a78dc91
-  data.tar.gz: d67eb5f0cb37ad106574b2ae327eefcfc13c9d585cddec6661898f4d8166ebcc
+  metadata.gz: 4cb66cfbf1c78a46f36cb24a7cbc9e04b0bc96aa1285fe81de79cec4bfd1c2c1
+  data.tar.gz: f650a1b30acac1298186dce0818eede9944e3b5117e794801abd0576d7b37b9e
 SHA512:
-  metadata.gz: 6b8adbc3c4707fc4c823456ae1d7547f17568802de03008a17fef18a5f95af08b0e42d48ccdfab25a740603a58ab89c036d70cec94405701201e5a5af51ce392
-  data.tar.gz: 9bea98a42ab64a9b45ddc5564cd077d7eb6d2ddc293844759bb8001aa9fefd8aa26b0e03fff7a286ccde9f7aeacacda9fbb187fe04082749d3c2605e0cece7b9
+  metadata.gz: 1b87b74cafd00c2a8a3fe5a44942a005a4756974363c916c650c18e74df719920537bfaecdf21080aed339d24f1988444940d96fd66dc6af847498c04efbc033
+  data.tar.gz: 40e7f8357bd2ff7485c7d7105d852b6c615eaf8902b787b611953df637e246a53990acdc6e1b7f7ff2dd350edf749bd012352981b908de2bbcbee0bc59714513

data/LICENSE

@@ -1,4 +1,4 @@
-Copyright (c) 2010, 2012-2024,  Ammar Ali
+Copyright (c) 2010, 2012-2025,  Ammar Ali
 
 Permission is hereby granted, free of charge, to any person
 obtaining a copy of this software and associated documentation

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

@@ -18,6 +18,7 @@ module Regexp::Expression
     Codepoint   = Class.new(Base) # e.g. \u000A
 
     CodepointList = Class.new(Base) # e.g. \u{A B}
+    UTF8Hex       = Class.new(Base) # e.g. \xE2\x82\xAC
 
     AbstractMetaControlSequence = Class.new(Base)
     Control                     = Class.new(AbstractMetaControlSequence) # e.g. \cB

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

@@ -15,6 +15,12 @@ module Regexp::Expression::EscapeSequence
   Hex.class_eval         { def codepoint; text[/\h+/].hex end }
   Codepoint.class_eval   { def codepoint; text[/\h+/].hex end }
 
+  UTF8Hex.class_eval do
+    def codepoint
+      text.scan(/\h+/).map(&:hex).pack('C*').force_encoding('utf-8').ord
+    end
+  end
+
   CodepointList.class_eval do
     # Maybe this should be a unique top-level expression class?
     def char

data/lib/regexp_parser/parser.rb

@@ -319,6 +319,7 @@ class Regexp::Parser
     when :codepoint_list; node << EscapeSequence::CodepointList.new(token, active_opts)
     when :hex;            node << EscapeSequence::Hex.new(token, active_opts)
     when :octal;          node << EscapeSequence::Octal.new(token, active_opts)
+    when :utf8_hex;       node << EscapeSequence::UTF8Hex.new(token, active_opts)
 
     when :control
       if token.text =~ /\A(?:\\C-\\M|\\c\\M)/

data/lib/regexp_parser/scanner/scanner.rl

@@ -37,7 +37,8 @@
   octal_sequence        = [0-7]{1,3};
 
   hex_sequence          = 'x' . xdigit{1,2};
-  hex_sequence_err      = 'x' . [^0-9a-fA-F{];
+  hex_sequence_err      = 'x' . [^0-9A-Fa-f];
+  high_hex_sequence     = 'x' . [89A-Fa-f] . xdigit . ( '\\x' . [89A-Fa-f] . xdigit )*;
 
   codepoint_single      = 'u' . xdigit{4};
   codepoint_list        = 'u{' . xdigit{1,6} . (space . xdigit{1,6})* . '}';
@@ -210,7 +211,7 @@
         type = :nonposixclass
       end
 
-      unless self.class.posix_classes.include?(class_name)
+      unless POSIX_CLASSES[class_name]
         raise ValidationError.for(:posix_class, text)
       end
 
@@ -256,9 +257,21 @@
   # escape sequence scanner
   # --------------------------------------------------------------------------
   escape_sequence := |*
-    [1-9] {
+    [1-9] . [0-9]* {
       text = copy(data, ts-1, te)
-      emit(:backref, :number, text)
+
+      # If not enough groups have been opened, there is a fallback to either an
+      # octal or literal interpretation for 2+ digit numerical escapes.
+      digits = text[1..-1]
+      if digits.size == 1 || digits.to_i <= self.capturing_group_count
+        emit(:backref, :number, text)
+      elsif digits =~ /\A[0-7]{2,}\z/
+        emit(:escape, :octal, text)
+      else
+        emit(:escape, :literal, text[0..1])
+        emit(:literal, :literal, text[2..-1])
+      end
+
       fret;
     };
 
@@ -321,6 +334,16 @@
       fret;
     };
 
+    high_hex_sequence > (escaped_alpha, 5) {
+      text = copy(data, ts-1, te)
+      if regexp_encoding == Encoding::BINARY
+        text.split(/(?=\\)/).each { |part| emit(:escape, :hex, part) }
+      else
+        emit(:escape, :utf8_hex, text)
+      end
+      fret;
+    };
+
     hex_sequence > (escaped_alpha, 5) @eof(premature_end_error) {
       emit(:escape, :hex, copy(data, ts-1, te))
       fret;
@@ -514,6 +537,7 @@
     };
 
     group_open @group_opened {
+      self.capturing_group_count += 1
       text = copy(data, ts, te)
       emit(:group, :capture, text)
     };
@@ -662,6 +686,7 @@ class Regexp::Scanner
 
     input = input_object.is_a?(Regexp) ? input_object.source : input_object
     self.free_spacing = free_spacing?(input_object, options)
+    self.regexp_encoding = input_object.encoding if input_object.is_a?(::Regexp)
     self.spacing_stack = [{:free_spacing => free_spacing, :depth => 0}]
 
     data  = input.unpack("c*")
@@ -672,6 +697,7 @@ class Regexp::Scanner
 
     self.set_depth = 0
     self.group_depth = 0
+    self.capturing_group_count = 0
     self.conditional_stack = []
     self.char_pos = 0
 
@@ -711,10 +737,9 @@ class Regexp::Scanner
     File.read("#{__dir__}/scanner/properties/#{name}.csv").scan(/(.+),(.+)/).to_h
   end
 
-  def self.posix_classes
+  POSIX_CLASSES =
     %w[alnum alpha ascii blank cntrl digit graph
-       lower print punct space upper word xdigit]
-  end
+       lower print punct space upper word xdigit].to_h { |c| [c, true] }.freeze
 
   # Emits an array with the details of the scanned pattern
   def emit(type, token, text)
@@ -749,7 +774,9 @@ class Regexp::Scanner
   attr_accessor :block,
                 :collect_tokens, :tokens, :prev_token,
                 :free_spacing, :spacing_stack,
+                :regexp_encoding,
                 :group_depth, :set_depth, :conditional_stack,
+                :capturing_group_count,
                 :char_pos
 
   def free_spacing?(input_object, options)