Handle error result from unescape_tokens in tokenizer

[lukaszsamson]

I noticed this crash

** (MatchError) no match of right hand side value: {:error, {[line: 1, column: 22], ~c"invalid Unicode escape character, expected \\uHHHH or \\u{H*} where H is a hexadecimal digit. Syntax error after: ", ~c"\\u"}}
    (elixir 1.18.4) src/elixir_tokenizer.erl:929: :elixir_tokenizer.handle_dot/6

All other usages of unescape_tokens properly handle errors

[josevalim]

Do you know which input can trigger this? If not, I can reverse engineer a test. :)

[lukaszsamson]

Do you know which input can trigger this? If not, I can reverse engineer a test. :)

46.y .'9VQj}.{k*J\%yDHqULDXI!g5V"!qJWhV|U%"XJN\xgy(<=&|6^GDk.ui+zOO}\,mZ5QA%T]'BUP@Q}8:-yRNL7C\gG'

[josevalim]

If you can push a test with a minimum string, that would be fantastic, otherwise I will do it later, thanks!

[sabiwara]

I think I got a minimum string:

iex> Code.string_to_quoted "a.'\\u'"
** (MatchError) no match of right hand side value: {:error, {[line: 1, column: 3], ~c"invalid Unicode escape character, expected \\uHHHH or \\u{H*} where H is a hexadecimal digit. Syntax error after: ", ~c"\\u"}}
    (elixir 1.18.4) src/elixir_tokenizer.erl:929: :elixir_tokenizer.handle_dot/6

[josevalim]

💚 💙 💜 💛 ❤️