fix: Parse character, octal, and hex string escapes

[jiang1997]

Support additional protobuf character escapes, including escaped quotes, plus bounded octal and hexadecimal escapes. Add tests for parsed defaults and numeric escape boundaries.

Unicode escapes and validation of malformed or unknown escapes remain unsupported.

[alexander-fenster]

@jiang1997 Do you know what's the behavior of the protoc parser for these cases? I can probably go figure out but if you have a pointer to their code or docs that would've been handy :)

[jiang1997]

@alexander-fenster Yes. I used the protobuf string literal grammar and protoc's tokenizer as references:

• Spec: https://protobuf.dev/reference/protobuf/proto3-spec/#string-literals
• Implementation: https://github.com/protocolbuffers/protobuf/blob/main/src/google/protobuf/io/tokenizer.cc#L1163-L1235

protoc consumes up to three octal digits and two hexadecimal digits. I also verified that the parsed defaults match protoc's output.

[jiang1997]

Separately, Unicode escapes (\uXXXX and \UXXXXXXXX) are still missing. I think supporting them would be cleaner with a small character-by-character scanner instead of extending the current regex.

Would that make sense for a follow-up PR?