Use correct newlines for heredoc inner lines

Co-authored-by: Jason Kim <jasonkim@github.com>
Co-authored-by: Adam Hess <HParker@github.com>

Author: 3 people

lib/prism/translation/parser/compiler.rb

@@ -2057,23 +2057,31 @@ def visit_heredoc(node)
           node.parts.each do |part|
             pushing =
               if part.is_a?(StringNode) && part.unescaped.include?("\n")
-                unescaped = part.unescaped.lines(chomp: true)
-                escaped = part.content.lines(chomp: true)
+                unescaped = part.unescaped.lines
+                escaped = part.content.lines
 
-                escaped_lengths =
-                  if node.opening.end_with?("'")
-                    escaped.map { |line| line.bytesize + 1 }
-                  else
-                    escaped.chunk_while { |before, after| before.match?(/(?<!\\)\\$/) }.map { |line| line.join.bytesize + line.length }
+                escaped_lengths = []
+                normalized_lengths = []
+
+                if node.opening.end_with?("'")
+                  escaped.each do |line|
+                    escaped_lengths << line.bytesize
+                    normalized_lengths << (line.chomp.bytesize + 1)
                   end
+                else
+                  escaped
+                    .chunk_while { |before, after| before.match?(/(?<!\\)\\\r?\n$/) }
+                    .each do |lines|
+                      escaped_lengths << lines.sum(&:bytesize)
+                      normalized_lengths << lines.sum { |line| line.chomp.bytesize + 1 }
+                    end
+                end
 
                 start_offset = part.location.start_offset
-                end_offset = nil
 
-                unescaped.zip(escaped_lengths).map do |unescaped_line, escaped_length|
-                  end_offset = start_offset + (escaped_length || 0)
-                  inner_part = builder.string_internal(["#{unescaped_line}\n", srange_offsets(start_offset, end_offset)])
-                  start_offset = end_offset
+                unescaped.map.with_index do |unescaped_line, index|
+                  inner_part = builder.string_internal([unescaped_line, srange_offsets(start_offset, start_offset + normalized_lengths.fetch(index, 0))])
+                  start_offset += escaped_lengths.fetch(index, 0)
                   inner_part
                 end
               else

test/prism/ruby/parser_test.rb

@@ -57,14 +57,12 @@ class ParserTest < TestCase
     # skip them for now.
     skip_all = skip_incorrect | [
       "dash_heredocs.txt",
-      "heredocs_with_ignored_newlines.txt",
+      # "heredocs_with_ignored_newlines.txt",
       "regex.txt",
       "regex_char_width.txt",
       "unescaping.txt",
       "seattlerb/bug190.txt",
       "seattlerb/heredoc_nested.txt",
-      "seattlerb/heredoc_with_carriage_return_escapes_windows.txt",
-      "seattlerb/heredoc_with_carriage_return_escapes.txt",
       "seattlerb/heredoc_with_extra_carriage_returns_windows.txt",
       "seattlerb/heredoc_with_only_carriage_returns_windows.txt",
       "seattlerb/heredoc_with_only_carriage_returns.txt",
@@ -116,6 +114,8 @@ class ParserTest < TestCase
       "seattlerb/heredoc_squiggly_visually_blank_lines.txt",
       "seattlerb/heredoc_squiggly.txt",
       "seattlerb/heredoc_unicode.txt",
+      "seattlerb/heredoc_with_carriage_return_escapes_windows.txt",
+      "seattlerb/heredoc_with_carriage_return_escapes.txt",
       "seattlerb/heredoc_with_interpolation_and_carriage_return_escapes_windows.txt",
       "seattlerb/heredoc_with_interpolation_and_carriage_return_escapes.txt",
       "seattlerb/interpolated_symbol_array_line_breaks.txt",