powerloom / py-ipfs-client Goto Github PK

Describe the bug

The add_bytes() method does not return an expected CID against a bytes object uploaded to IPFS.

The cat() method does not support iterating over bytes streams when attempting to read previously uploaded bytes files to IPFS.

To Reproduce

Steps to reproduce the behavior:

1. Call add_bytes() on a bytes object
2. Call cat() on a CID that contains a bytes file (for eg, a gzip or bz2 file) instead of the typical JSON file

Expected behavior
Should accept and returns bytes stream correctly.

Proposed Solution
Parse the response from the IPFS node appropriately in add_bytes()

Support a bytes mode flag in cat() to utilize the httpx response stream's async iteration helper, aiter_bytes()

Caveats
NA

Additional context
NA

Is your feature request related to a problem?
Presently the library expects multiaddr format addresses to locate the IPFS node. This does not allow the library to be used on 3rd party IPFS API providers like Infura which offer HTTP URL endpoints.

Describe the solution you'd like
Parse the URL appropriately during initialization to consider HTTP URLs.

Describe alternatives you've considered
NA

Additional context
NA