robertwb commented on PR #21762: URL: https://github.com/apache/beam/pull/21762#issuecomment-1150530866
If the columns aren't actually strings then an exception will be raised when trying to encode them at strings. I wasn't able to figure out how to get StringDType out of csv either, and convert_dtypes requires looking at the entire PCollection to figure out the proxy object. I think this is pretty safe because objects should be strings immediately following a csv (or json, or most other common types) read. If it's not, exposing this in a cross-langauge way would be hard anyway. (I suppose one may encounter the bytes type, and nested types (though IIRC pandas doesn't do to well with that anyway). -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
