[DIT-13447] Introduce escape decoding on extractor level

[jholiga]

Overview

Introduces handling specifically for escaped values like \n, \t, etc.

The root issue is that we encode the escape character to maintain the raw value, but never decode it when going to store it in Ditto. It's handled at the extractor level so that we can abide by certain per-language rules (for instance in kotlin a raw string """hello\nworld""", the \n would be intended to show.

Also addresses an issue found during testing for android xml specifically where tags were throwing off some of the parsing

Context

Screenshots

Test Plan

• Unit cover most of the previously failing cases
• A CLI yarn scan works as intended against https://github.com/dittowords/ditto-android-demo
  • If you can add some \n, \t, \\n etc to the code you're scanning that's a bonus, but I think the unit tests cover that logic pretty well.