Don't scan (?<!...> as named group

fixes #85
ammar · Nov 16, 2022 · 81340c3 · 81340c3
1 parent 752c040
commit 81340c3
Show file tree

Hide file tree

Showing 3 changed files with 18 additions and 3 deletions.
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -7,6 +7,13 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ## [Unreleased]
 
+### Fixed
+
+- fixed scanning of two negative lookbehind edge cases
+  * `(?<!x)y>` used to raise a ScannerError
+  * `(?<!x>)y` used to be misinterpreted as a named group
+  * thanks to [Sergio Medina](https://github.com/serch) for the report
+
 ## [2.6.0] - 2022-09-26 - [Janosch Müller](mailto:janosch84@gmail.com)
 
 ### Fixed

diff --git a/lib/regexp_parser/scanner/scanner.rl b/lib/regexp_parser/scanner/scanner.rl
@@ -90,8 +90,8 @@
   group_options         = '?' . ( [^!#'():<=>~]+ . ':'? ) ?;
 
   group_ref             = [gk];
-  group_name_id_ab      = ([^0-9\->] | utf8_multibyte) . ([^>] | utf8_multibyte)*;
-  group_name_id_sq      = ([^0-9\-'] | utf8_multibyte) . ([^'] | utf8_multibyte)*;
+  group_name_id_ab      = ([^!0-9\->] | utf8_multibyte) . ([^>] | utf8_multibyte)*;
+  group_name_id_sq      = ([^0-9\-']  | utf8_multibyte) . ([^'] | utf8_multibyte)*;
   group_number          = '-'? . [1-9] . [0-9]*;
   group_level           = [+\-] . [0-9]+;
 

diff --git a/spec/scanner/groups_spec.rb b/spec/scanner/groups_spec.rb
@@ -6,7 +6,9 @@
   include_examples 'scan', '(abc)',           0 => [:group,     :capture,        '(',          0, 1]
 
   # Named groups
-  # only names that start with a hyphen or digit (ascii or other) are invalid
+  # Names that start with a hyphen or digit (ascii or other) are invalid.
+  # ")" is only allowed as first char of the name.
+  # "!" is allowed anywhere, but ?<!...> is treated as a lookbehind by Ruby.
   include_examples 'scan', '(?<name>abc)',    0 => [:group,     :named_ab,       '(?<name>',   0, 8]
   include_examples 'scan', "(?'name'abc)",    0 => [:group,     :named_sq,       "(?'name'",   0, 8]
   include_examples 'scan', '(?<name_1>abc)',  0 => [:group,     :named_ab,       '(?<name_1>', 0,10]
@@ -19,7 +21,11 @@
   include_examples 'scan', "(?'üüuuüü'abc)",  0 => [:group,     :named_sq,       "(?'üüuuüü'", 0,10]
   include_examples 'scan', "(?<😋1234😋>abc)",  0 => [:group,     :named_ab,       "(?<😋1234😋>", 0,10]
   include_examples 'scan', "(?'😋1234😋'abc)",  0 => [:group,     :named_sq,       "(?'😋1234😋'", 0,10]
+  include_examples 'scan', "(?<)x>y)",        0 => [:group,     :named_ab,       '(?<)x>',     0, 6]
+  include_examples 'scan', "(?')x'y)",        0 => [:group,     :named_sq,       "(?')x'",     0, 6]
+  include_examples 'scan', "(?'!x'y)",        0 => [:group,     :named_sq,       "(?'!x'",     0, 6]
 
+  # Passive groups
   include_examples 'scan', '(?:abc)',         0 => [:group,     :passive,        '(?:',        0, 3]
   include_examples 'scan', '(?:)',            0 => [:group,     :passive,        '(?:',        0, 3]
   include_examples 'scan', '(?::)',           0 => [:group,     :passive,        '(?:',        0, 3]
@@ -33,6 +39,8 @@
   include_examples 'scan', '(?!abc)',         0 => [:assertion, :nlookahead,     '(?!',        0, 3]
   include_examples 'scan', '(?<=abc)',        0 => [:assertion, :lookbehind,     '(?<=',       0, 4]
   include_examples 'scan', '(?<!abc)',        0 => [:assertion, :nlookbehind,    '(?<!',       0, 4]
+  include_examples 'scan', '(?<!x)y>',        0 => [:assertion, :nlookbehind,    '(?<!',       0, 4]
+  include_examples 'scan', '(?<!x>)y',        0 => [:assertion, :nlookbehind,    '(?<!',       0, 4]
 
   # Options
   include_examples 'scan', '(?-mix:abc)',     0 => [:group,     :options,        '(?-mix:',    0, 7]