[Shivam Sharma (ShivamSharma43)]

Description

Fixes #4260.

With encrypt_type="custom", _stringify_decrypt() raised ValueError: Illegal character sequence in user decrypt function! for any decryption code containing a backslash followed by a digit. The check was there to avoid "invalid group reference" errors when the YAML template was applied with re.sub. convert_async() now renders the template with SeedPrompt.render_template_value() (a Jinja2 sandbox), which inserts decrypt_function as is, so the check only rejected valid code, such as a regex backreference (r"\2 \1") or '\0'. Because it ran in __init__, the converter couldn't be built at all.

• Removed the check and the now unused re import from codechameleon_converter.py.
• New test builds a custom converter whose decrypt function uses re.sub(..., r"\2 \1", ...) plus a SEPARATOR = '\0' string element. It checks that the code is rendered into the prompt unchanged, then runs the rendered code and checks it decrypts the encoded prompt.

Before / after for the repro in #4260:

before: ValueError: Illegal character sequence in user decrypt function! Invalid chars: ['\2', '\1']
after:      return re.sub(r"(\S+) (\S+)", r"\2 \1", text)
        The encrypted_problem you need to solve is: to How down cut tree? a

This change was prepared with the help of Claude Code (an AI coding assistant). I reviewed the change and the tests.

🤖 Generated with Claude Code

[Copilot]

Copilot was unable to review this pull request because the user who requested the review has insufficient remaining quota. Try requesting a review at a lower effort level.