Changing to a different epub ereader should never require manually deobfuscating anything.
The list of obfuscated fonts and obfuscation method is provided by the encrypt.xml file found in the epub's META-INF folder. The data used to obfuscate and deobfuscate depends on obfuscation method and is tied to a specific dc:identifiers in the opf metadata this is sometimes the bookid and another uuid based dc:identifiers. The actually obfuscation algorithm involves a fixed number of bytes (again differs by algorithm) at the top of the font file being xored with data from the opf identifier in a very specific manner. The exact algorithm has been published in the epub spec. But this is not something you can easily manually undo.
So as explained earlier, unzip your epub, throw away the broken font files, replace them with known good matching font files (rename exact file if needed), then remove the entire encrypt.xml from the META-INF folder as keeping it will mistakenly cause the reader to obfuscate the font files on load instead of deobfuscating them.
Properly rezip up the epub.
|