View Single Post
Old 02-21-2011, 07:40 PM   #7
Junior Member
theomeier began at the beginning.
Posts: 8
Karma: 10
Join Date: Sep 2008
Device: Sony PRS-950
Debug or error

The error happens with every book. See the debug protocol enclosed:

calibre Debug log
calibre 0.7.46
('Windows', 'Vista', '6.0.6002')
Python 2.7.0
Windows: ('Vista', '6.0.6002', 'SP2', 'Multiprocessor Free')
Starting up...
Started up in 118.303000212 covers: HTTP Error 404: Not Found covers: cover for 9780312576554 found covers: HTTP Error 404: Not Found covers: cover for 9780982952306 found covers: cover for 9780451229571 found covers: cover for 9780451229571 found

I have the pdf ISBN extraction sript from MEKK installed:

Originally Posted by bazbar View Post
For anyone who cares this is how I finally got the many thousands of ISBNs I had in my unorganized ebooks into Calibre on Windows. This is a giant pain in the ass and I might have missed some steps that I went through but at least this might save you a little time. It is better than getting a repetitive stress injury by doing it all by hand (I know, I did about 2000 books by hand before I gave up on that nonsense).

Again, if anyone cares, I feel like ISBN recognition should be a core attribute of any software that is trying to organize ebooks - it is the critical piece of information that allows for proper organization and recognizing it should be as automated as possible. If you don't want to make it happen automatically at least make it an option or somehow build this Python script I'm running into the system so we can choose to run it on our libraries without all of the nonsense below? Anyway, my $.02. Goodluck to anyone else trying.

Download pdftotext from
Extract the contents, or pdftotext.exe alone, to a directory. Add the directory to your path.

1) download and install Python 2.6. do not use a newer version since you will need binary distributions of packages that don't exist for newer versions.

2) install setuptools that works with Python 2.6

3) install binary of lxml that works with Python 2.6

3a) if the binary above failed with an error about a missing dll then download that dll from the web and put it in the same directory as the exe (awesome, right? I know. I love downloading random dlls because they don't include the visual c distribution stuff)

4) set enviornment variables - make sure the following are all in your path:
C:\Program Files\Calibre2\;C:\Python26\Scripts;

5) open command prompt and try 'easy_install mekk.calibre'

5a) if step 5 fails with some kind of error then you probably need to install a compiler - I used MinGW:
When that runs if you have options make sure you include gcc and MAKE (I think I selected them all but I don't think you need to)

5b) tell easy_install to use the MinGW compiler
This is done by creating a file called "distutils.cfg" at C:\Python26\Lib\distutils that contains following:
add MinGW to the path: c:\MinGW\bin;

5c) close command prompt, reopen, retry easy_install mekk.calibre

6) edit .calibre-utils file (if it exists for you) - it should be in your C:\Documents and Settings\<user> folder or similar based on operating system. If you didn't install one of the commands listed then set it to nothing like so:
catdoc =
archmage =
djvutxt =
calibredb = calibredb
pdftotext = pdftotext

Make sure to keep the calibredb line (Calibre should already be in the path). I only have pdftotext working in mine. If it doesn't exist then run the program once, it will likely error out eventually when it finds a missing command, then track down the config file and edit it.

7) run calibre_guess_and_add_isbn (or any of the other options in mekk's script)

automatic ISBN recognition win32 windows calibre
theomeier is offline   Reply With Quote