[GR-45043] Import Prism

PullRequest: truffleruby/4222

Author: andrykonchin

spec/ruby/language/regexp/back-references_spec.rb

@@ -22,6 +22,15 @@
     $10.should == "0"
   end
 
+  it "returns nil for numbered variable with too large index" do
+    -> {
+      eval(<<~CODE).should == nil
+        "a" =~ /(.)/
+        eval('$4294967296')
+      CODE
+    }.should complain(/warning: ('|`)\$4294967296' is too big for a number variable, always nil/)
+  end
+
   it "will not clobber capture variables across threads" do
     cap1, cap2, cap3 = nil
     "foo" =~ /(o+)/

spec/tags/language/hash_tags.txt

@@ -1,9 +1,9 @@
 fails:The ** operator makes a caller-side copy when calling a method taking a positional Hash
 fails:Hash literal checks duplicated keys on initialization
 fails:Hash literal checks duplicated float keys on initialization
-fails(https://github.com/ruby/prism/issues/2115):Hash literal expands an '**{}' or '**obj' element with the last key/value pair taking precedence
-fails(https://github.com/ruby/prism/issues/2115):Hash literal expands an '**{}' and warns when finding an additional duplicate key afterwards
-fails(https://github.com/ruby/prism/issues/2115):Hash literal merges multiple nested '**obj' in Hash literals
+fails(https://github.com/ruby/prism/issues/2005):Hash literal expands an '**{}' or '**obj' element with the last key/value pair taking precedence
+fails(https://github.com/ruby/prism/issues/2005):Hash literal expands an '**{}' and warns when finding an additional duplicate key afterwards
+fails(https://github.com/ruby/prism/issues/2005):Hash literal merges multiple nested '**obj' in Hash literals
 fails(https://github.com/ruby/prism/issues/2129):Hash literal raises an EncodingError at parse time when Symbol key with invalid bytes
 fails(https://github.com/ruby/prism/issues/2129):Hash literal raises an EncodingError at parse time when Symbol key with invalid bytes and 'key: value' syntax used
 fails:The ** operator makes a copy when calling a method taking a positional Hash

spec/tags/language/heredoc_tags.txt

@@ -1,2 +1 @@
 fails:Heredoc string prints a warning if quoted HEREDOC identifier is ending not on same line
-fails(https://github.com/ruby/prism/issues/2074):Heredoc string allows HEREDOC with <<~'identifier', no interpolation, with backslash

spec/tags/language/regexp_tags.txt

@@ -5,5 +5,3 @@ slow:Source files encoded in UTF-16 LE with a BOM are invalid because they conta
 slow:Source files encoded in UTF-16 BE without a BOM are parsed as empty because they contain a NUL byte before the encoding comment
 slow:Source files encoded in UTF-16 BE with a BOM are invalid because they contain an invalid UTF-8 sequence before the encoding comment
 slow:Source files encoded in UTF-16 LE without a BOM are parsed as empty because they contain a NUL byte before the encoding comment
-fails(https://github.com/ruby/prism/issues/2166):Source files encoded in UTF-16 LE with a BOM are invalid because they contain an invalid UTF-8 sequence before the encoding comment
-fails(https://github.com/ruby/prism/issues/2166):Source files encoded in UTF-16 BE with a BOM are invalid because they contain an invalid UTF-8 sequence before the encoding comment

spec/tags/language/source_encoding_tags.txt

@@ -10,4 +10,6 @@ ast: |
   ObjectLiteralNode
       attributes:
           flags = 1
-          object = RubyRegexp(source = abc \xFF, options = RegexpOptions(kcode: NONE, fixed, kcodeDefault, literal), encoding = ASCII-8BIT)
+          object = RubyRegexp(source = abc \xFF, options = RegexpOptions(kcode: NONE, fixed, kcodeDefault, literal), encoding = ASCII-8BIT)
+          sourceCharIndex = 21
+          sourceLength = 10

spec/tags/truffle/parsing/parsing_tags.txt

@@ -21,12 +21,16 @@ ast: |
           methodName = "=~"
           notEmptyKeywordsProfile = false
           notRuby2KeywordsHashProfile = false
+          sourceCharIndex = 21
+          sourceLength = 10
       children:
           arguments = [
               ReadGlobalVariableNodeGen
                   attributes:
                       flags = 0
                       name = "$_"
+                      sourceCharIndex = -1
+                      sourceLength = 0
                   children:
                       lookupGlobalVariableStorageNode =
                           LookupGlobalVariableStorageNodeGen
@@ -38,4 +42,6 @@ ast: |
               ObjectLiteralNode
                   attributes:
                       flags = 0
-                      object = RubyRegexp(source = abc \xFF, options = RegexpOptions(kcode: NONE, fixed, kcodeDefault, literal), encoding = ASCII-8BIT)
+                      object = RubyRegexp(source = abc \xFF, options = RegexpOptions(kcode: NONE, fixed, kcodeDefault, literal), encoding = ASCII-8BIT)
+                      sourceCharIndex = 21
+                      sourceLength = 10

spec/truffle/parsing/fixtures/regexps/encoding/when_non_ascii_characters.yaml

@@ -7,7 +7,7 @@ ast: |
   StringLiteralNode
       attributes:
           encoding = UTF-8
-          flags = 0
+          flags = 1
           sourceCharIndex = 0
           sourceLength = 11
           tstring = foobar

spec/truffle/parsing/fixtures/regexps/encoding_in_boolean_context/when_non_ascii_characters.yaml

@@ -0,0 +1,16 @@
+subject: "String"
+description: "2 string literals concatenated at parse/translation time / with frozen_string_literal: true"
+notes: >
+  Result string should be frozen and represented as a FrozenStringLiteralNode node
+focused_on_node: "org.truffleruby.language.literal.FrozenStringLiteralNode"
+ruby: |
+  # frozen_string_literal: true
+  "foo" "bar"
+ast: |
+  FrozenStringLiteralNode
+      attributes:
+          definition = "expression" (ASCII-8BIT)
+          flags = 1
+          frozenString = "foobar" (UTF-8)
+          sourceCharIndex = 30
+          sourceLength = 11