TOC not identifying all chapters

[shilpa]

Hi,

I have just bought a sony prs-505 and have downloaded the calibre software. My books are all in pdf and i have converted some of them to epub. Because i am new to the software i kept the settings as it is in the conversion part. I am quite happy with it. Apart from the chapters. when i do a conversion it picks up chapters correctly upto a certain number then it does not. for example in a book with 51 chapters i will get chapters 1 - 20(so that is fine) but from 20 onwards i get chapter 30, chapter 40 chapter 50. It does this for most of my books. So it misses chapters between 30 - 40 and 40 - 50 etc. Does anybody now why this is happening? is there a setting that i need to change?

[Catire]

Try modifying "table of contents" and/or the "the look and feel". For that you'll need to read the Xpath tutorial



If that doesn't help you let me know
If the book is not in copyright you might find it here or on another site already formatted for your Sony. If its not in copyright and you can't find it post it here and I'll test it.


Btw Welcome to Mobileread.

[shilpa]

ok thanks catire. I will try that.

[ldolse]

PDF conversion is error prone, and there is a point where a regex attempts to identify the chapters, marking them up so that the xpath will later catch them. Editing the xpath from the default won't help with pdf.

You could open a bug on this and attach your pdf, I can look into enhancing the regex if possible. That said, this will continue to happen with other pdfs as the conversion can cause numerous variations in the output text.

Unfortunately, the best way to fix these in general is by hand. Now that Sigil is out it will make it trivial to scroll through the book and mark the missed chapters so that they're added to the TOC. Note if you're using Calibre .6.x you'll need to wait a couple days til a related crash fix is released for Sigil - watch for version 0.1.1.

[shilpa]

I will wait for the crash fix of the sigil comes out.

Also i was wondering how do you enhance the regex of the pdf file? what file do you open and what is that you edit?

[ldolse]

It's in preprocess.py You'd need to download a local version, edit it, and then update your local install.

On OS X I'd use the following command from the directory containing the source file to update my calibre installation (fixed according to Kovid's input):

calibre-debug --update module calibre.ebooks.preprocess ./preprocess.py

The directories vary a bit by platform, everything else is the same.

These are the regexes doing the chapter detection there:

                  # Detect Chapters to match default XPATH in GUI
                  (re.compile(r'(?=<(/?br|p))(<(/?br|p)[^>]*)?>\s*(?P<chap>(<i><b>|<i>|<b>)?(Chapter|Epilogue|Prologue|Book|Part)\s*(\d+|\w+)?(</i></b>|</i>|</b>)?)(</?p[^>]*>|<br[^>]*>)\n?((?=(<i>)?\s*\w+(\s+\w+)?(</i>)?(<br[^>]*>|</?p[^>]*>))((?P<title>(<i>)?\s*\w+(\s+\w+)?(</i>)?)(<br[^>]*>|</?p[^>]*>)))?', re.IGNORECASE), chap_head),
                  (re.compile(r'(?=<(/?br|p))(<(/?br|p)[^>]*)?>\s*(?P<chap>([A-Z \'"!]{5,})\s*(\d+|\w+)?)(</?p[^>]*>|<br[^>]*>)\n?((?=(<i>)?\s*\w+(\s+\w+)?(</i>)?(<br[^>]*>|</?p[^>]*>))((?P<title>.*)(<br[^>]*>|</?p[^>]*>)))?'), chap_head),

[kovidgoyal]

Actually in 0.6 the correct syntax is

calibre-debug --update module calibre.ebooks.preprocess ./preprocess.py

IIRS the old syntax will no longer work, though I'm not sure.