fix: pass encoding to partition_csv for CSV loader

[DoTA (dotuananh0712)]

Summary

• Fixed issue Need to pass encoding to partition_csv() in langchain_community/document_loaders/csv_loader.py --> _get_elements #505 where encoding passed to UnstructuredCSVLoader was being ignored
• Now extracts encoding from unstructured_kwargs and passes it directly to partition_csv()
• This allows users to specify custom encodings for CSV files with non-UTF-8 content

Issue

When users pass encoding via unstructured_kwargs, the unstructured library ignores it and defaults to UTF-8, causing UnicodeDecodeError for CSV files with other encodings.

Fix

def _get_elements(self) -> List:
    from unstructured.partition.csv import partition_csv

    input_encoding = self.unstructured_kwargs.get("encoding")
    return partition_csv(
        filename=self.file_path, encoding=input_encoding, **self.unstructured_kwargs
    )

Testing

• Tested with the reproduction case from issue Need to pass encoding to partition_csv() in langchain_community/document_loaders/csv_loader.py --> _get_elements #505

AI assistance used for code review and implementation