|
Preprocessing isn't guaranteed to work, it tries to match the most common cases, but there are myriad ways that the chapters could be marked. It also doesn't work for epub, if your source format is epub - try renaming it to .zip instead of .epub and add it to the book record as a compressed html bundle, and convert that format instead.
Also note preprocessing won't build the TOC, it just finds the most common types of chapter headings and wraps h2 tags around them, but the xpath for chapter detection still needs to match the contents of those h2 tags.
If you open a bug with some examples I can see if your case is something that can be accommodated.
|