Streamlit Sharing - Fileupload, where does it go?

[Joe_Pawlowski]

So, I successfully shared my app via Streamlit Sharing. Even the upload file part works but, my quesion is, does the file persist anywhere (i.e., will it be there in a few days somewhere like in Github ( i didn’t see it)) or, does it reside in memory and clear when the browser/session is terminated?

thanx

Great work team.

[randyzwitch]

Hi @Joe_Pawlowski -

When you upload a file using file_uploader, the data is written to RAM. It will persist in memory for a while (depending on the cache settings), but it will never persist on disk unless you choose to do that yourself.

How to process .apk file uploaded using fileuploader?

Hi @allenpbiju, welcome to the Streamlit community! st.file_uploader returns a BytesIO object, which is a buffer, which isn’t one of the types that function accepts (as the error message indicates). To get the bytes from a BytesIO object, you need to read the buffer: a = APK(uploaded_apk.read())

Best,
Randy

[Joe_Pawlowski]

Oh so, for streamlet sharing I have the option to have the uploaded file persist in my GitHub directory? Thanks for your reply

[PablocFonseca]

I don’t think so, you’ll probably have to persist it somewhere else… s3 for instance…

[Joe_Pawlowski]

Thanks for your reply. S3?

[PablocFonseca]

Amazon S3, or other file storage service that you can access programatically.

Pandas should be capable of writing to s3 storages easily

Once you read your uploaded data into a DataFrame you would call something like:

df.to_csv('s3://path/to/your/file')

To save it.

If you need to save files and not data frames. There is a Lib to interface s3

[Joe_Pawlowski]

Ahhhhh. Sneaky!

Seems like that would work with any cloud path.

Awesome. Thank you

[randyzwitch]

Right, there’s no default Streamlit option, but there is the default Python file writing capabilities, pandas, JSON, etc