Closed lokijuhy closed 3 years ago
.... maybe it's worth mentioning as Aron was having a similar error and didn't know what to do when dealing with parquet files...
This is definitely worth documenting somewhere. I think Notion might be more appropriate. I don't want to put specific instructions in datatc itself that will likely go out of date with future changes to pyarrow/macOS.
Fixes
Documentation: Added Supported File Formats page to documentation. Preview the docs here.
Gives a more detailed error message when a user tries to interact with parquet files when they don't have a parquet engine installed. Previously, the user got an
ImportError
(thrown by pandas). Now, that same error message is prepended with additional text, so the error in full reads:datatc
error message aboutpyarrow
orfastparquet
specifically, because theImportError
comes from pandas, not datatc itself - This whole error comes from the fact that pandas has ato_parquet
method yet does not include a parquet engine in its install requirements. So, I don't want to write an error message that just make things more confusing if pandas's dependencies/requirements change in the future. All together, I'm not sure my addition to pandas's error message adds anything valuable, and in fact I think it sounds "un-pythonic". But I gave it a try. Please let me know your opinion on whether the additional description is worth it, or how to make it sound better.Misc