llama_index icon indicating copy to clipboard operation
llama_index copied to clipboard

Adds an AsanaReader for loading tasks from a workspace

Open daveey opened this issue 2 years ago • 3 comments

daveey avatar Feb 14 '23 03:02 daveey

Hi @daveey, thanks for the reader. The only comment I have is. We can make the asana_token optional and get it from environment variable, if it is not provided. It is up to you.

ahmetkca avatar Feb 14 '23 04:02 ahmetkca

Awesome! Would you be able to port this over to https://github.com/emptycrown/llama-hub? We're porting over all the readers to a dedicated library over there.

Incidentally, we're drawing prizes for all new creators of loaders before 2/24 so this would be a great fit.

emptycrown avatar Feb 14 '23 05:02 emptycrown

@daveey thanks for this PR!

If you have questions on migrating to llamahub, no problem. ping us and we're more than happy to help out

jerryjliu avatar Feb 14 '23 08:02 jerryjliu

I added a PR to llamahub, but was unable to test it:

from gpt_index import download_loader AsanaReader = download_loader('AsanaReader') Traceback (most recent call last): File "", line 1, in File "/Users/daveey/code/gpt_index/gpt_index/readers/download.py", line 56, in download_loader raise ValueError("Loader class name not found in library") ValueError: Loader class name not found in library

I assume this is because the PR has to first be merged?

daveey avatar Feb 16 '23 03:02 daveey

Yeah, has to be merged first. Typically it's enough to just test the class itself in a notebook. Taking a look now, thanks!

emptycrown avatar Feb 16 '23 03:02 emptycrown

Hi @daveey, thanks for the reader. The only comment I have is. We can make the asana_token optional and get it from environment variable, if it is not provided. It is up to you.

I can do that if you feel strongly, but personally I prefer that users explicitly specify auth tokens rather than relying on ENV.

daveey avatar Feb 16 '23 03:02 daveey

Closing in favor of https://github.com/emptycrown/llama-hub/pull/38

daveey avatar Feb 16 '23 03:02 daveey

I added a PR to llamahub, but was unable to test it:

from gpt_index import download_loader

AsanaReader = download_loader('AsanaReader')

Traceback (most recent call last):

File "", line 1, in

File "/Users/daveey/code/gpt_index/gpt_index/readers/download.py", line 56, in download_loader

raise ValueError("Loader class name not found in library")

ValueError: Loader class name not found in library

I assume this is because the PR has to first be merged?

You can actually point to your fork and temporarily test before merging.

ahmetkca avatar Feb 16 '23 17:02 ahmetkca