I've updated the decompiler slightly so that it now handles all the sample boks bundled with the Reader. I was not able to get 100% recompilable files yet (it seems the format is changed slightly from Librie), but the produced .lrs should give you some insight into how to make better lrfs for Reader.
clicky (http://projects.mobileread.com/reader/users/igorsk/lrf2lrs_03.zip).

Actually I take that back... after I changed the BookID and removed the readings, I could recompile the "Winning" excerpt, and the footnote links worked! (at least in the Connect software... didn't try the Reader yet)

cool thanks, this will be very useful.

igorsk, you continue to rock.

All of you ... 'unofficial developers' totally rock. :)

I just wanted to say that.

igorsk, you continue to rock.

All of you ... 'unofficial developers' totally rock. :)

I just wanted to say that.


I totally agree with you NatCh! :cool:

Quick way to make the output .lrs compilable with lrs2lrf:
1) in the <BookInfo> section, remove all readings, e.g.:
<Title reading="">Winning (Excerpt)</Title>
2) for <BookID> tag, make first two characters read "FB", e.g.:
<BookID>FBUS12511000A00T</BookID>
3) compile away!

P.S. I didn't try it with all books. Let me know if you run into any issues.

Do you know what the reading attribute is supposed to be for?

EDIT: Nevermind, its for sorting

Does anybody know where to find xylogxml.dll that you need for lrs2lrf?

On the yahoo group I read it comes with BookCreator but I have that installed and it's not there.

I think it's bundled with the lrs2lrf distribution.

I think it's bundled with the lrs2lrf distribution.

Could be but not under that name at least. The problem is that the lrf file created by lrs2lrf is 0 bytes. :(

EDIT: THis is the English translation of the error reported in the log file:

The forced termination which just does not have the file which
checks the consistency
of the file

I guess the .lrs generated by lrf2lrs was not valid. Which book did you try to decompile?

I guess the .lrs generated by lrf2lrs was not valid. Which book did you try to decompile?


Different books that I created with makelrf. Then I converted them with lrf2lrs. But when I try to convert the unchanged file back with lrs2lrf it goes wrong.

Do you have a sample lrs file for me that should work?

I got it to work. Played around with the header and then finally got it working. Thanks igorsk.

I have this error poping up and 0 bytes lrs as a result.
Parsing objects... done.
Traceback (most recent call last):
File "C:\lrf2lrs\lrf2lrs.py", line 1584, in <module>
sys.exit(main(sys.argv[1:]))
File "C:\lrf2lrs\lrf2lrs.py", line 1578, in main
out.write(h.toXml());
File "C:\lrf2lrs\lrf2lrs.py", line 1511, in toXml
xml += self.objects[self.TocObjId].toXml(self.objects)
KeyError: 0
Any idea?

Looks like your file has no TOC specified... where did you get it?
You could fix the problemt by changing that line into:
if self.TocObjId in self.objects: xml += self.objects[self.TocObjId].toXml(self.objects)

Looks like your file has no TOC specified... where did you get it?
You could fix the problemt by changing that line into:
if self.TocObjId in self.objects: xml += self.objects[self.TocObjId].toXml(self.objects)
That was librie output of BookCreater. Thank you, the line change helped.
I can only say wow, for quick response.

New version, fixed decompiling of LRFs with embedded fonts and it should now work on Macs (thanks to grahame!).
http://projects.mobileread.com/reader/users/igorsk/lrf2lrs_04.zip

Is it possible to add support for uncompressed LRF files like those generated by HTML2LRF?

I'm not sure why you consider them "uncompressed" but I will see what can be done for HTML2LRF's files.

A better description would be that the files "aren't compressed", usually standard in LRF/LRX files? lrf2lrs gives a zlib error with the HTML2LRF files and the file sizes compared to ones generated by Book Designer are at least twice the size.

zlib error is caused by corrupted data, looks like the Librie DLLs have some bugs in them. And the larger size is because of inefficient conversion from HTML to LRF, with lots of redundant code.