[Booki-dev] epub <-> booki-zip round trip

[Douglas Bagnall]

There's bugs, but it generally works.

You make the booki-zip from here:

http://objavi.halo.gen.nz/espri.cgi

Where the thing you put in is the internet archive ID.  For example:

http://objavi.halo.gen.nz/espri.cgi?book=oldfrenchfairyta00sgrich

The http library doesn't like the redirects on some of the books, in
which case you'll see a 500 error.  I don't know exactly what the issue
is, but it seems fixable.

Then you go here:

http://objavi.halo.gen.nz/epubjavi.cgi

The archive ones are at the bottom of the drop-down as it sorts upper
case first.

There is some problem with the ncx file, so the table of contents
doesn't work.  The file looks right to me, just not to fbreader, so I
don't know what is wrong but I expect it is fixable.

Some of the metadata is a bit messed up (e.g. it sticks in a FLOSS
Manuals authorship claim), but most of it seems intact.

With fbreader the books look visually the same after the cycle, but I
don't think fbreader uses the stylesheets so there might be issues there.

I'll be writing some tests today to ensure that the journey is in fact
round.



Douglas