Quote:
Originally Posted by Ken Maltby
Hmm... I thought TOC (Table of Contents) for epubs were constructed within the epub.
The reading software (Adobe Reader Mobile) would be extracting, using and presenting the TOC already in the epub. Just extracting the data contained in the TOC, for the database, when first accessing the file should be instantaneous by human perceptions.
I completely miss whatever logic there would be for having data from each "split" in the
database, at all. They are accessed sequentially by the reading software and are indexed for random page access, already. I doubt very much that Adobe made a special reader program that uses Kobo's database, for anything.
|
Kobo is using the TOC (from the NCX file) in the epub. But, it is read once when the books is first processed, loaded into the database and used from there when reading. I have played with the different pieces to prove this to my satisfaction.
The split files rows are only for kepubs. These don't go through the Adobe code, but through the ACCESS reader. I haven't spent as much time looking at these and haven't really seen the reason for them. Especially as they also record the actual TOC entries.
I am sure that ARM has no interaction with the database. I assume that it talks to the Kobo software which stores this.