[m4mmon]

Hello,

I have searched a lot but no luck, possibly my queries are not relevant, so here is my question.

I use calibre CLI in order to convert an html file to epub. The options I used are derived from those found in logs when using calibre GUI, so I expect to have the same result.
But I am missing something, the result is my epub has sometimes unresolved links to footnotes and/or missing page breaks. I tried to fine tune options, but that did not change anything.

My html source is a filtered html document from Word.
If I try to directly convert it to epub with ebook-convert and "my" options, I have the problem previously described.

If I first import the html file in calibre, get the resulting zip file then ebook-convert it with the exact same options as above, this time I have the expected result, exactly the same that if I did the conversion with calibre GUI.

So my conclusion is I am missing something with the html input. When importing in calibre there is some processing on it, but I don't know how to replicate it with CLI since I cannot see any log related to the import.

I tried to ebook-convert from html to zip first, but the resulting zip is completely different than if imported into calibre.


Can someone provide any tip/information/link to appropriate documentation section/existing forum thread ?
Thank you.