While writing a mobi file parser, I tried to implement the reverse engineered specifications of the mobi format from http://wiki.mobileread.com/wiki/MOBI
, but I struggled to understand how to correctly parse the index defined by the INDX, IDXT and TAGX sections.
In order to solve this problem, I inspected a mobi file generated by kindlegen
, which (unfortunately) only increased my confusion. The order of the sections defining only a single entry was something like "INDX - TAGX - IDXT - INDX - IDXT". (followed by the RESC section; may be irrelevant)
I would appreciate any explanations which would help me to understand the structures of these indices very much!