asynchronous support?

[fuad00]

Any plans on this?

[gagb]

There is no current plan yet, but PRs welcome! @afourney

[0xRaduan]

Would be great, I can try to take a stab on this tomorrow.

[0xRaduan]

Just an update from my end: I checked the current implementation, and while there are some benefits of async support, many libraries that are used under the hood, are not supporting async:

• markdownify
• mammoth
• pandas
• pdfminer
• python-pptx

At this point, if all underlying calls will be sync, I am not sure, how much benefit it would give to provide async APIs on markitdown, outside of a thin wrapper. Otherwise, we need a full rewrite + using a full new set of libraries.

If people want to use MarkItDown in async manner, I think they should just create a thin wrapper for markitdown.

I did a tiny PR on how I would do that: #32;