PDF metadata CreationDate and Save to disk

[paulzag]

Thanks for this wonderful software. I'm playing around with it to manage all my documents. I have a lot of PDF's from FIBA.com

Calibre defaults to the current date when I add a pdf. I expect a pdf with a CreationDate to be imported in Calibre with that date as the Date and Publish date in Calibre's metadata. Am I not getting something?

Secondly it save the metadata back to the file I tried Save to disk in a single directory. But Calibre wrote a new renamed file (as the title changed) but I can't get the book record to link to the new file. Any suggestions?

As an aside does anybody know a utility for Windows or Linux to set the OS file creation date to the PDF metadata CreationDate? Lots that let me overwrite the metadata, but that's not the goal here.

Regards

Paul

[paulzag]

bump. Any ideas?

[Starson17]

Originally Posted by paulzag

Calibre defaults to the current date when I add a pdf. I expect a pdf with a CreationDate to be imported in Calibre with that date as the Date and Publish date in Calibre's metadata. Am I not getting something?

Your expectation is wrong, but you seem to correctly understand how it works. The "date" is not the PDF CreationDate, it's the date the record is created. The publication date can be set manually, or you can fetch metadata from the web to set it.

Secondly it save the metadata back to the file I tried Save to disk in a single directory.

Yes, that's what it's supposed to do.

But Calibre wrote a new renamed file (as the title changed)

This is also what it's supposed to do.

but I can't get the book record to link to the new file. Any suggestions?

It's not supposed to link to the new file. You don't normally need metadata updated in the file kept in the database, as it will update during a Convert, a Save or a Send. If you want it updated anyway, you need to convert to the same format.

[paulzag]

Thanks Starson17 for the reply. I thought it would turn out to be something I wasn't understanding. Using calibre a bit more helps my understanding of the calibre way.

The other reason I want to update the PDF's is I use google desktop search more and more.

I'm putting all my ebooks into calibre, however many PDF's are industry or research specific and not created by information scientists. So the metadata is sadly lacking (many Authors are "Administrator"). However online fetch of metadata about the Official Rules of Basketball 2010 edition won't work. Nor will there be any fetching of metadata for a research report on a small Australian copper exploration company.

The PDF metadata creation date is however the closest thing to a publication date. And I'd like to keep that as the source of the Publication date. Do I need to hack the pdf import function (if this is not the best-practice use-case) or do I file an enhancement request (if what I want makes sense)?

I'd rather use one tool to maintain my collection.

I don't mind calibre storing metadata in it's database, but I ALSO want the files to be updated with whatever metadata suits the format. In this case, if I edit the Author, Title, Tags and Comments how do I best get that metadata back into the PDF file? 0.7.13 now smartly forbids saving files back into the library. Converting PDF->PDF seems a slow and inefficient way of updating the metadata.

This wouldn't be a problem for most of my ebooks as I'll read them on my devices. But these PDF's are likely to stay on my netbook which is my PDF reading device of choice.

I get that calibre is like iTunes for ebooks, but in iTunes I can save id3 tags back to the underlying files. Should the database ever get corrupted, I lose some of my data, but not the core tags.

Am I trying to force calibre to be something it's not?

[Starson17]

Originally Posted by paulzag

The other reason I want to update the PDF's is I use google desktop search more and more.

Calibre tries not to "change" a user's book unless he clearly wants it. That's why the default is to save the book without change, and store metadata only in the db. If you want Google desktop search to access Calibre's copy of your books, you're going to have to update the metadata in the pdf.

Converting PDF->PDF seems a slow and inefficient way of updating the metadata.

PDF sucks. According to Kovid, that's how the 3rd party tools work to change metadata, and it's not something that's in Calibre's control to do differently.

I get that calibre is like iTunes for ebooks, but in iTunes I can save id3 tags back to the underlying files. Should the database ever get corrupted, I lose some of my data, but not the core tags.

You can read this discussion elsewhere here, but the bottom line is that ebooks are not like mp3s and metadata for ebooks is not like id3 tags. It's not possible to save most metadata in most ebook formats. Trying to write multiple types of metadata into different and inconsistent ebook formats is not at all like trying to write id3 tags into the mp3 format specifically designed to receive those tags.

Am I trying to force calibre to be something it's not?

Perhaps. It's up to you to decide. PDF format seems of particular interest to you, and of all ebook formats, it's the worst. It's not even a format, it's a container. Some things inside the container (like Portfolio PDFs and images of pages) aren't readable at all by Calibre.

It's also the one format that Calibre has the least control over when creating it or modifying it to add/change metadata. Calibre relies entirely on 3rd party tools for those functions and if those tools don't work, you have to go elsewhere to get the problem fixed. The main focus of Calibre when it comes to PDF is how the heck to read the $%^&* format cleanly to convert to something more ebook friendly.

[paulzag]

Thanks Starson17

I know what you mean about pdf's. They are devil spawned and represent the smallest percentage of my data. But they cause the greatest pain. I've seriously thought about converting them to something friendlier, but most of them have formating that is tricky.

Calibre tries not to "change" a user's book unless he clearly wants it. That's why the default is to save the book without change, and store metadata only in the db. If you want Google desktop search to access Calibre's copy of your books, you're going to have to update the metadata in the pdf.

Once my file is in calibre I delete the original. I'm okay with calibre's paradigm and this it's the smart way to do it.

...ebooks are not like mp3s and metadata for ebooks is not like id3 tags. It's not possible to save most metadata in most ebook formats. Trying to write multiple types of metadata into different and inconsistent ebook formats is not at all like trying to write id3 tags into the mp3 format specifically designed to receive those tags

Don't start me on id3 "standards". oh well, you've helped me a lot. Thanks

I think I'll end up converting most stuff to epub and leave the fracking pdf's to hang