from langchain_text_splitters import Language, RecursiveCharacterTextSplitter

PERL is in the Language enum but has no separator implementation

splitter = RecursiveCharacterTextSplitter.from_language(Language.PERL, chunk_size=50, chunk_overlap=0)

Raises: ValueError: Language perl is not implemented yet!

TL;DR

Add a Perl separator list to get_separators_for_language() to support the Perl language.

Guidance

• Verify that the Language enum in base.py includes PERL and that get_separators_for_language() in character.py does not have a handler for it.
• Add a separator list for Perl to get_separators_for_language() with keywords such as \nsub , \nmy , \nour , \nif , \nforeach , \nwhile , \nuse , \npackage .
• Test RecursiveCharacterTextSplitter.from_language(Language.PERL) after implementing the separator list to ensure it no longer raises a ValueError.
• Consider updating the documentation to reflect the addition of Perl support.

Example

def get_separators_for_language(language):
    # ... existing code ...
    if language == Language.PERL:
        return ["\nsub ", "\nmy ", "\nour ", "\nif ", "\nforeach ", "\nwhile ", "\nuse ", "\npackage "]
    # ... existing code ...

Notes

This fix assumes that the provided separator list is sufficient for the Perl language. Additional testing may be necessary to ensure the correctness of the implementation.

Recommendation

Apply workaround: add the Perl separator list to get_separators_for_language() as described above, because the issue is due to a missing implementation for the Perl language.