|
Calibre doesn't actually re-order anything during pdf conversion. It does attempt to detect the pdf chapters and insert split points at each chapter. If things are truly being reordered this would indicate some problem with the splitting code, but that would affect lots of formats, not just pdf - this is unlikely to be the root cause.
The most likely alternative theory is that the pdf is formatted in an obscure way that's tripping up pdftohtml - i.e. Garbage In Garbage Out. You could validate this by enabling the debug directory option and see if the html in the 'input' folder is already screwed up.
|