Hi there, reposting on this channel since llamacloud doesn't seem super active.
I recently started using LlamaParse via the python SDK. However, disappointed to see the parser is not compatible with the abstract file system for reading externally hosted files (e.g. S3) when passing it as the file_extractor to the SimpleDirectoryReader.
I get the following error when trying to load data using the SimpleDirectoryReader with the LlamaParse parser as the file_extractor.
Failed to load file <file-name>.pdf with error: LlamaParse.load_data() got an unexpected keyword argument 'fs'. Skipping...
Anyone have any workarounds for this or any timelines for when this might be added? I guess I can download the file to my server and parse it as a local file but obviously reading direct from S3 would be much cleaner
https://github.com/run-llama/llama_parse/pull/117I noticed this pull request which has been open for some time and possibly isn't fully functional.
