Match % strings in parser

Author: kddnewton

lib/prism/translation/parser/compiler.rb

@@ -1689,6 +1689,8 @@ def visit_string_node(node)
             builder.string_compose(token(node.opening_loc), children, closing)
           elsif node.opening == "?"
             builder.character([node.unescaped, srange(node.location)])
+          elsif node.opening&.start_with?("%") && node.unescaped.empty?
+            builder.string_compose(token(node.opening_loc), [], token(node.closing_loc))
           else
             content_lines = node.content.lines
             unescaped_lines = node.unescaped.lines

test/prism/ruby/parser_test.rb

@@ -68,13 +68,11 @@ class ParserTest < TestCase
       "seattlerb/heredoc_with_only_carriage_returns.txt",
       "seattlerb/masgn_double_paren.txt",
       "seattlerb/parse_line_heredoc_hardnewline.txt",
-      "seattlerb/pct_nl.txt",
       "seattlerb/pctW_lineno.txt",
       "seattlerb/regexp_esc_C_slash.txt",
       "seattlerb/TestRubyParserShared.txt",
       "unparser/corpus/literal/assignment.txt",
       "unparser/corpus/literal/block.txt",
-      "unparser/corpus/literal/def.txt",
       "unparser/corpus/literal/dstr.txt",
       "unparser/corpus/literal/literal.txt",
       "unparser/corpus/literal/pattern.txt",