Added support for iterable input instead of file

[sebdiem]

I added the support for writing an iterable to the zip archive.

This can be useful when you have a very large file (typically resulting from a large database request). In this case you may want to split your request in several requests (pagination) and write them separately to the archive without creating a huge file in memory or on the disk.

[sebdiem]

Hi, would you have time to consider this pull request please ?

BR

Sébastien

[allanlei]

@sebdiem The changes makes example.py fail to run due to the iterable kwargs taking the replacing the positional arg arcname.

I think it would be a more suitable solution to create a zipstream.ZipFile.write_iter(arcname, iter[, compress_type]) method instead of modify write.

• Keeps the function signature the same as stdlib's zipfile.ZipFile.write(filename[, arcname[, compress_type]])
• arcname would be a required positional argument instead of kwarg.
• Let's example.py work again
• There is a similar method/reference zipfile.ZipFile.writestr(zinfo_or_arcname, bytes[, compress_type])

[sebdiem]

Thanks for your review.
I will update my PR

[sebdiem]

Done!

[allanlei]

Can you update your changes to the README? It's still using write(iterable=iterable)

[sebdiem]

Right! I have just updated the README.
In fact I also corrected it because it was referring to an old api (zipstream.ZipFile(path) does not create an archive with files in path)