afourney
MarkItDown deals with byte streams. You can get the same behavior by doing: ```python import io input_data = b"Test" result = markitdown.convert_stream(io.BytesIO(input_data), file_extension=".html") ``` If it's a string, then perhaps:...
Fantastic find. We should certainly expose this. I'm looking into a mechanism that will allow passing more options to the converters (as well as allowing for more of a plugin...
Thanks for the work on this. It was included in a recent refactor for 0.1.0a1
Arguably, if the xlsx cell is empty, the conversion should be empty. If the excel value is "#N/A" then I believe it should be interpreted as NaN (though it's perhaps...
Not yet, but let me prioritize it for this week. I'm not sold on removing columns, or headers... but swapping NaN for "" make a lot of sense to me
Sounds good. This PR has drifted from when it was submitted. Let me try to migrate the "NaN" -> "" bits to the new main branch, and we can add...
I agree with the need to support reentrance or similar. In the two-agent scenario, we've worked around it in stateless environments (e.g., a http), by repopulating the histories, running without...
Thanks for the report. I will investigate. Does this happen only with the CLI, or also with the library?
Ok, I think I found the problem and fixed it with 0.1.0a6. Please let me know if it works better for you. Once a fix is confirmed, I will close...
This looks very promising. Thank you. I need to check some procedural stuff on my end before I can merge this (re: hosting images rather than just Dockerfiles). Let me...