Converting a large file to lrf in calibre

[bnajari]

Hi, I just got my PRS505 for graduation a couple of days ago and I've been trying all sorts of ways to fit a medical textbook onto my reader. I have the book in .chm format, and it is 380MB. I used hh.exe to decompile it, zipped it into a zip file, and then used calibre, but I get the following error message:

Detailed traceback:
Traceback (most recent call last):
File "main.py", line 370, in add_books
File "main.py", line 395, in _add_books
File "calibre\gui2\library.pyo", line 123, in add_books
File "calibre\library\database.pyo", line 1295, in add_books
MemoryError

I have also tried ABC Amber Converter, which converts it to a 7MB lit file. When I try using calibre on the lit file, the job just keeps processing and never stops. I've left it for a couple of hours and the hour glass just keeps spinning.

Any advice?

[Deputy-Dawg]

Have you tried doing the conversion from the command line? As in

html2lrf

or

lit2lrf

[bnajari]

Actually, after posting my first message I did try the conversion from the the command line. lit2lrf completed the conversion fine. But when I load the 10MB file onto the reader, the last page is "NaN" and it won't load up either of the pages. I am just stuck on the Begin, End, Bookmarks, TOC, etc screen. There is also no TOC.

When I try html2lrf on the zipped output from hh.exe, I get this:

C:\Program Files\calibre>html2lrf --minimize-memory-usage d:\camp.zip
Processing camp.zip
Traceback (most recent call last):
File "convert_from.py", line 1929, in <module>
File "convert_from.py", line 1923, in main
File "convert_from.py", line 1815, in process_file
File "convert_from.py", line 269, in __init__
File "convert_from.py", line 377, in add_file
File "calibre\ebooks\chardet\__init__.pyo", line 49, in xml_to_unicode
File "calibre\ebooks\chardet\__init__.pyo", line 26, in detect
File "calibre\ebooks\chardet\universaldetector.pyo" , line 115, in feed
File "calibre\ebooks\chardet\charsetgroupprober.pyo ", line 59, in feed
File "calibre\ebooks\chardet\sjisprober.pyo", line 53, in feed
MemoryError

[bnajari]

So just now I tried a smaller .chm book that was only about 30MB, and the same thing happens when I ABC Amber CHM Converter to export a .lit file and the use lit2lrf. It freezes on the 1st screen of the book.

I also did the same hh.exe -decompile, zipped the output file, and run html2lrf. This is what I get with that:

C:\Program Files\calibre>html2lrf d:\SGU.zip
Processing SGU.zip
Traceback (most recent call last):
File "convert_from.py", line 1929, in <module>
File "convert_from.py", line 1923, in main
File "convert_from.py", line 1815, in process_file
File "convert_from.py", line 269, in __init__
File "convert_from.py", line 377, in add_file
File "calibre\ebooks\chardet\__init__.pyo", line 59, in xml_to_unicode
TypeError: decode() argument 1 must be string, not None

[kovidgoyal]

Interesting, send me you small zip file.

[Deputy-Dawg]

Kovid,
I don't think that Mr. bnajari is altogether on the up and up. He did post me the .lit file on one of the share servers. The file contained about 11 pages of binary data that the beginning of the file which choked calibre. Moreover he had done the conversion of the file with a demo version of ABC Amber CHM converter which is limited in several interesting ways but the most interesting limitation would exist in even the production versions in that it creates a single flat file with no directory data and on such a large project the toc would be mandatory I would think.

[kovidgoyal]

Well he sent me the zip file, which I told him how to fix.