[Old Thread] Auto Extract ISBN-Feature request

[JvdW]

Originally Posted by chrisix

@JvdW

yes, please

OK, here we go:
Download from http://sourceforge.net/projects/unxutils/files/
the following package:
UnxUtils.zip

Unpack the UnxUtils.zip where you want it (ex: c:\unixutils), note where it ends up since we need the path to the bin folder and the path to usr\local\wbin

Download pdftotext from foolabs.com:
ftp://ftp.foolabs.com/pub/xpdf/xpdf-3.02pl4-win32.zip
Extract the contents, or pdftotext.exe alone, to the usr\local\wbin folder

Open a command prompt
type: set PATH=c:\unixutils\bin;c:\unixutils\usr\local\wbin; %PATH%
type: sh
(this should give you a unix like shell which will run the batch file)
type: ls
(this should give you a listing of files in your current directory)
type: exit
You're back at the command prompt of Windows.
Copy the attached isbn.bat into the wbin folder and rename it to isbn.zsh, edit the first line to reflect your installation of UnxUtils.
(For unknow reasons to me I can't upload isbn.zsh )
Copy your PDF's, if you're not trusting me that they won't be touched , into a temporary folder, no sub folders allowed, for example C:\PDF.
In the command prompt type:
cd c:
cd \pdf
type: sh
type: isbn.zsh

The script will try to find the following order of ISBN's:
ISBN-13
ISBN-10
ISBN:
ISBN
If found it will rename the file and move it into the done folder. Locations can be changed in the script if needed but first leave the defaults since it will make finding problems much easier.
After that make sure that Calibre is set to read the isbn number from the filename and start importing. Preferences->Add/Save->Adding Books>Regular expression: (?P<isbn>[0-9].+$)
Further I have 'Read metadata only from file name' not checked.

That should do the trick, I hope ;-)

Regards,

Joop

[chrisix]

Thanks a lot for your great script and help!

Can't run unixutils it in Win7 X64 (abnormal program termination)
So I did it in WinXP in a VMware.

I have a very high ISBN output which is great because it is a folder of files which failed all in e-library, but the script often is stopping without any reason and without any output so I interrupt it and start it again, may be it has something to do with the VMware Shared Folder where the files are in?

[JvdW]

Originally Posted by chrisix

Thanks a lot for your great script and help!

Can't run unixutils it in Win7 X64 (abnormal program termination)
So I did it in WinXP in a VMware.

Sorry I indeed didn't state that it wouldn't work in 64bit Windows.

I have a very high ISBN output which is great because it is a folder of files which failed all in e-library, but the script often is stopping without any reason and without any output so I interrupt it and start it again, may be it has something to do with the VMware Shared Folder where the files are in?

Doesn't it atleast output the name of the last pdf it is going to work on?
If it has something todo with the shared folder I would suspect that every file failed.

Can you show some of the output of the script?

Regards,

Joop

[chrisix]

Originally Posted by JvdW

Doesn't it atleast output the name of the last pdf it is going to work on?
If it has something todo with the shared folder I would suspect that every file failed.
Can you show some of the output of the script?

As I told, it stops without any error, and the pdf at which one it stops is random. I think it has something to do with the unixutlis, your script is clean and I don't see any reason for stopping.

When I have a lot of time I will try to rewrite your script with windows powershell if it is possible

[JvdW]

Originally Posted by chrisix

As I told, it stops without any error, and the pdf at which one it stops is random. I think it has something to do with the unixutlis, your script is clean and I don't see any reason for stopping.

If you look at the script and run the commands just one by one does it then work?
Its really wierd because if the pdf doesn't contain an ISBN number the script will print the name of the PDF and then process the next one, printing the name and move on. One thing that might cause this is a pdf file which isn't strictly conforming to the specs.
We have an application which import a pdf which is first split into single pages and the application which generates this multipage pdf generates an invalid pdf. (MS reporting service)

When I have a lot of time I will try to rewrite your script with windows powershell if it is possible

That thought crossed my mind too but I'm a little more proficient with unix commands then with powershell.

Regards,

Joop

[chrisix]

I already made a working PowerShell Script....
Just needs Windows PowerShell and the pdftotext.exe

The next thing I will figure out is to find ISBN-13, ISBN-10, ISBN in one loop to make it a bit faster.

[UnraisedArc]

I'm surprised this thread is still alive. Unfortunately, my ebook collection has grown to the thousands, and I still have many that have sparse metadata. Good luck guys and keep up the good work!

[chrisix]

If someone is interested in a PDF > ISBN Extractor PowerShell Script, let me know.

[chrisix]

Does anybody know if it is possible to limit the lines/pages output txt with ebook-convert?
I can't find a parameter for it.
I will try to make ISBN detection for pdf, epub and may be more and use only calibre own commands.

[Starson17]

Originally Posted by chrisix

Does anybody know if it is possible to limit the lines/pages output txt with ebook-convert?

I'm pretty certain the answer is no, assuming you are asking if you can insert page breaks into the txt output at regular intervals. Calibre's txt output is for e-readers where there is no concept of a page. The text is continuous.

If you are asking about limiting articles and feeds during news fetching (my primary use for ebook-convert) the answer is yes, but not by page or line, only by article, feed, date, etc.

[JvdW]

Originally Posted by chrisix

If someone is interested in a PDF > ISBN Extractor PowerShell Script, let me know.

Would be nice to look at how you've done it. Might even use it instead of my own script



Joop

[chrisix]

[chrisix]

Originally Posted by JvdW

Would be nice to look at how you've done it. Might even use it instead of my own script

Just made a compare between yours and mine and yours is doing a much better job, so let me tweak it a bit

[chrisix]

Here it is, put pdftotext.exe in same path.

I still have a problem with some situations, but may be you find a better much cleaner way, I already get crazy.....

0.2 > small changes.
0.3 > better results.

[chrisix]

8 downloads no comment?

is it usable for anybody?

chris