[GUI Plugin] Extract ISBN

[Vortex]

Yes. These are text PDF's, I can copy and paste them from within the front of the books into the metadata, so wondering why the plugin is failing on them?

[BetterRed]

Originally Posted by Vortex

Yes. These are text PDF's, I can copy and paste them from within the front of the books into the metadata, so wondering why the plugin is failing on them?

If its many books, try converting a couple to TXT and see if the PI can find the ISBN in there.

ISBN extraction from PDFs is 'problematic' for the same reasons as converting them is 'problematic'.

BR

[Vortex]

Originally Posted by BetterRed

ISBN extraction from PDFs is 'problematic' for the same reasons as converting them is 'problematic'.
BR

Very diplomatically put, BR. My language is usually a lot more colourful when describing pdf files.

[toomuchreading]

kiwidude:
Thank you for this plugin. It has been very helpful in fixing up my library. I wanted to point out that I recently found an issue with your plugin and the 'mark book' function of Calibre. When Extract ISBN finds an ISBN that needs updating in a book, and you happen to have books marked, when the update is complete, all your marks are then removed as well. I have had two behaviors happen. I just finished testing with 3 books marked. The first two did not have any ISBN's to update but the third did. I ran your extract ISBN plugin, and after being prompted to update my books, all but the book being update had their marks removed (that book just happened to be the last in the list of three). This proved to me that the plugin was in conflict somehow with the mark book function of Calibre. The previous time is noticed this, all marks (over 200) were removed, and I was left with no marked books after the update function of Extract ISBN.
Your help in resolving this would be appreciated. If you need more information, or help testing, let me know and I will do my best to assist.

[BetterRed]

This has been a problem ever since the Mark books function was made available as a user initiated feature. Prior to that Mark books was an internal function, which was only used by this PI and maybe one or two others - such as Find Duplicates.

BR

[commandante]

Hi,

Some of my epubs are wrongly tagged. The publisher put the catalog number inside the ISBN tag. For example there is in content.opf:
<dc:identifier id="bookId" opf:scheme="ISBN">803957</dc:identifier>

Inside the epub, there is a tag with the correct ISBN inside the toc.ncx file:
<meta name="dtb:uid" content="ISBN:978-3-247-05954-9"/>

However, the ISBN is not recognized by the plugin:

---------------------------------------------------
Scanning first 4, then last 2, then remaining 0 files
Invalid ISBN match: 2012-036907
Invalid ISBN match: 1993-2005
Invalid ISBN match: 1971-1991
Scan time: 2.06 secs
The scan failed to find an isbn in 2.06 secs
Failed to extract ISBN
================================================== =
Scan complete, with 1 failures

Any quick fix for that?

Thanks,

[commandante]

I have tried to investigate the issue I reported. When I edit the book and copy "<meta name="dtb:uid" content="ISBN:978-6-257-03951-9"/>" in the first html page, the plugin doesn't recognize the ISBN. However when I edit the tag to "<meta name="dtb:uid" content="ISBN:9786257039519"/>", it works.

From what I understand the plugin doesn't scan the toc.ncx file.

To me there are two different issues:
(1) The plugin doesn't seem to recognize ISBN written this way: 978-6-257-03951-9
(2) The plugin doesn't seem to scan the toc.ncx file

When I check the source code (I don't have any skill in programming), I understand that it relies on a function called HTMLPreProcessor. I don't know if HTMLPreProcessor is supposed to scan toc.ncx or not. If someones passes by, any information on that would be appreciated!

[JSWolf]

When an eBook has multiple ISBN numbers listed, does the plugin get the correct eISBN or eBook ISBN?

[MetDeFiets]

@JSWolf
How could it know which is the right? It only gets the first one which it finds.

[JSWolf]

Originally Posted by MetDeFiets

@JSWolf
How could it know which is the right? It only gets the first one which it finds.

Which is not always correct. A lot of eBooks have e-book, ebook, or eISBN next to the eBook's ISBN. The plugin could look for those words to find the correct ISBN instead of the wrong ISBN. Sometimes there's even words to say it's a paper copy ISBN such as hardcover or paperback. It's not impossible to find the correct ISBN. Also, if there are multiple ISBN, put up a requestor asking us to choose which one we want to use.

As it stands, the plugin is pretty worthless given that it doesn't even bother to try to find the correct ISBN where there are multiple ones listed.

[mbovenka]

Originally Posted by JSWolf

As it stands, the plugin is pretty worthless given that it doesn't even bother to try to find the correct ISBN where there are multiple ones listed.

Speak for yourself, Jon. I find the plugin very useful. Yes, it does grab the wrong ISBN on occasion (perhaps for one in ten books or so), but overall it helps a in getting the right metadata. You have to check the downloaded stuff, yes, but then you have to do that anyway.

[MetDeFiets]

Originally Posted by JSWolf

As it stands, the plugin is pretty worthless given that it doesn't even bother to try to find the correct ISBN where there are multiple ones listed.

I have really a lot of books, and unlike mbovenka I would even rate it less then 10, maybe about 2% of the books which have ISBN for different books inside. So what, if it grebs the ebook, the printed version or the pocket book ISBN - what do you need the ISBN for? Usually for grebbing more information about the book (Title, Author etc.). And for that it's quite the same WHICH of those ISBN you greb.

For those cases, where the book includes promotion for other books: Yeah, that's bad. But since the book is often enough not titled well in Calibre (or not the right book title) and the Book commercial within other books are not in a certain format (it's not listed in the text in a regular form like "Title of the Book" - ISBN:0815 ; "Title of another Book" - ISBN:4711 etc.) - it would be a kind of crazy to impossible to find a solution for all those different formats.

Anyway, it's a FREE add-on and I'm really thankfull for it. But it's definitely NOT a perfect solution for ANY possible case. When it helps me in 95 % of the books, I save a lot of time. For the restly 5 %, the author would have to spend lots and lots and lots of hours of his free time. I do not expect him to do so for my lazyness.

So if you think the add-on is useless, why don't you uninstall it and write a better solution? Just criticising is easy

[crotchety1]

OK, let me first say that I don't usually write reviews; however, I just found this plugin to calibre and had to say something. If there were an award for "Best Plugin of the Year", this plugin would take it hands down. I have more than 7000 books in my library with more than 1000 with odd file names. This plugin has just helped me update all the meta data in no time at all. Thank you for all your hard work.

[skipster10]

I use this plug-in daily. It saves HUGE amounts of time! Thank you to the author!

I often import books for which the ISBN is in the title field and an ISBN may or may not be found in the contents of the book. It'd be a huge timesaver if the Extract ISBN plug-in had an option to check to see if the contents of the title field are a valid ISBN and, if so, copy or move it to the IDs field. Ideally, an option could be set to favor either the Title field ISBN or the found ISBN, if exists.

I realize that this can be done manually, but using Extract ISBN on a group of books is much more efficient.

Unfortunately, I haven't touched computer code in decades, so self-coding isn't an option.

[theducks]

Bulk Mode Metadata edit Search and Replace might help
(\d+{10,13}) <yes, I know there is no ISBN with 11 or 12 digits>
to capture a numeric pattern

replace in the Identifiers , type ISBN \1