From 3f881126c2498d21a7af7997e11974ae7f2c6840 Mon Sep 17 00:00:00 2001
From: alpaca-tc <alpaca-tc@alpaca.tc>
Date: Sat, 2 Aug 2025 22:34:38 +0900
Subject: [PATCH] Handle SyntaxError#message with invalid encoding

related: https://bugs.ruby-lang.org/issues/21528

This patch fixes an issue in IRB where it fails to handle a `SyntaxError#message` that has invalid encoding.

The root cause is that prism may generate a `SyntaxError#message` with broken encoding when the source contains multibyte characters. This patch ensures irb can handle such cases gracefully.
---
 lib/irb/ruby-lex.rb                  | 92 ++++++++++++++--------------
 test/irb/command/test_show_source.rb | 24 ++++++++
 2 files changed, 70 insertions(+), 46 deletions(-)

diff --git a/lib/irb/ruby-lex.rb b/lib/irb/ruby-lex.rb
index dd4a8d060..3aef64497 100644
--- a/lib/irb/ruby-lex.rb
+++ b/lib/irb/ruby-lex.rb
@@ -73,6 +73,48 @@ def defined? do
       yield
     ]
 
+    SYNTAX_ERROR_PATTERNS = {
+      # "syntax error, unexpected keyword_end"
+      #
+      #   example:
+      #     if (
+      #     end
+      #
+      #   example:
+      #     end
+      "unexpected keyword_end" => :unrecoverable_error,
+      # "syntax error, unexpected '.'"
+      #
+      #   example:
+      #     .
+      "unexpected '.'" => :unrecoverable_error,
+      # "syntax error, unexpected tREGEXP_BEG, expecting keyword_do or '{' or '('"
+      #
+      #   example:
+      #     method / f /
+      "unexpected tREGEXP_BEG" => :unrecoverable_error,
+      # "unterminated regexp meets end of file"
+      #
+      #   example:
+      #     /
+      #
+      # "unterminated string meets end of file"
+      #
+      #   example:
+      #     '
+      "unterminated string meets end of file" => :recoverable_error,
+      "unterminated regexp meets end of file" => :recoverable_error,
+      # "syntax error, unexpected end-of-input, expecting keyword_end"
+      #
+      #   example:
+      #     if true
+      #       hoge
+      #       if false
+      #         fuga
+      #       end
+      "unexpected end-of-input" => :recoverable_error,
+    }.freeze
+
     class TerminateLineInput < StandardError
       def initialize
         super("Terminate Line Input")
@@ -252,53 +294,11 @@ def check_code_syntax(code, local_variables:)
         # This is for a hash with invalid encoding symbol, {"\xAE": 1}
         :unrecoverable_error
       rescue SyntaxError => e
-        case e.message
-        when /unexpected keyword_end/
-          # "syntax error, unexpected keyword_end"
-          #
-          #   example:
-          #     if (
-          #     end
-          #
-          #   example:
-          #     end
-          return :unrecoverable_error
-        when /unexpected '\.'/
-          # "syntax error, unexpected '.'"
-          #
-          #   example:
-          #     .
-          return :unrecoverable_error
-        when /unexpected tREGEXP_BEG/
-          # "syntax error, unexpected tREGEXP_BEG, expecting keyword_do or '{' or '('"
-          #
-          #   example:
-          #     method / f /
-          return :unrecoverable_error
-        when /unterminated (?:string|regexp) meets end of file/
-          # "unterminated regexp meets end of file"
-          #
-          #   example:
-          #     /
-          #
-          # "unterminated string meets end of file"
-          #
-          #   example:
-          #     '
-          return :recoverable_error
-        when /unexpected end-of-input/
-          # "syntax error, unexpected end-of-input, expecting keyword_end"
-          #
-          #   example:
-          #     if true
-          #       hoge
-          #       if false
-          #         fuga
-          #       end
-          return :recoverable_error
-        else
-          return :other_error
+        SYNTAX_ERROR_PATTERNS.each do |pattern, error|
+          return error if e.message.include?(pattern)
         end
+
+        return :other_error
       ensure
         $VERBOSE = verbose
       end
diff --git a/test/irb/command/test_show_source.rb b/test/irb/command/test_show_source.rb
index 7ef879e81..c9eb845a8 100644
--- a/test/irb/command/test_show_source.rb
+++ b/test/irb/command/test_show_source.rb
@@ -423,5 +423,29 @@ class B
       assert_match(%r[#{@ruby_file.to_path}:7\s+Z = 1], out)
       assert_match(%r[#{@ruby_file.to_path}:8\s+Array = 1], out)
     end
+
+    def test_show_source_with_prism_returns_invalid_utf8_string
+      write_ruby <<~RUBY
+        class A
+          def call
+            if true
+              nil
+            # あああああああああああああああああああああああ
+            # あああああああああああああああああああああああ
+            end
+          end
+        end
+
+        binding.irb
+      RUBY
+
+      out = run_ruby_file do
+        type "inst = A.new"
+        type "show_source inst.call"
+        type "exit"
+      end
+
+      assert_match(/def call/, out)
+    end
   end
 end