Where is metadata for epubs stored?

[jist]

I did a web- and a forum- search on this, but couldn't find much information on it:

Where is metadata for epubs stored?
On the f.a.q. page I read "Metadata about the books is stored in the file metadata.db"

But each book also is accompanied by an opf file, which also contains what I guess you could call metadata.

And the epub format is also capable of containing metadata. (correct me if I am wrong)

So it looks like there are three locations that can hold metadata about the books.
Is there information or an overview to be found somewhere that explains what is—or can be—stored in the epub file itself (which you won't loose when copying the ebook somewhere), and what is stored in Calibre's metadata.db and the accompanying opf file?

[DNSB]

Originally Posted by jist

Is there information or an overview to be found somewhere that explains what is—or can be—stored in the epub file itself (which you won't loose when copying the ebook somewhere), and what is stored in Calibre's metadata.db and the accompanying opf file?

The database is the master copy, the metadata.opf is a backup copy and can be used when restoring your library. Unless you use Polish, ModifyEpub or a similar tool, the content.opf file internal to the epub is not modified in the library. When you Send to device for Save to disk, the content.opf will be updated to match the metadata in the database.

Hope this is of some help to you.

[jist]

Thanks DNSB.

Some questions though.

I don't understand where you say: "the content.opf file internal to the epub is not modified in the library."

I don't see any 'content.opf' files. Do you mean that they are somehow contained within the epub file and not visible?

If you meant to say 'metadata.opf' instead, I checked, and they do get updated as soon as you make an edit in Calibre.

So, Calibre does not store or read metadata in the epub files themselves.
Hm, I was guessing that was possible and even useful when copying a book to an ereader that can read metadata from an epub.

But then again, I have no clue if or how epubs support metadata.
I am acquainted with how metadata works for mp3 and flac audio files, but I guess it's a completely different ballgame for epubs.

[DNSB]

Originally Posted by jist

Thanks DNSB.

Some questions though.

I don't understand where you say: "the content.opf file internal to the epub is not modified in the library."

I don't see any 'content.opf' files. Do you mean that they are somehow contained within the epub file and not visible?

What I said is what I meant. An epub uses an .opf file for it's metadata. content.opf is commonly used but any filename.opf can be used as the container.xml file contains a pointer to it. Since an epub is a .zip container, you can see the files internal to an epub in several ways. You can open it with a zip utility such as 7Zip and examine the contents or you can use an epub editor such as Sigil or calibre's internal editor to open the epub.

Originally Posted by jist

If you meant to say 'metadata.opf' instead, I checked, and they do get updated as soon as you make an edit in Calibre.

As mentioned, the metadata.opf stored with the epub file is a backup copy of the data in the database and can be used when a database is restored.

Originally Posted by jist

So, Calibre does not store or read metadata in the epub files themselves.
Hm, I was guessing that was possible and even useful when copying a book to an ereader that can read metadata from an epub.

As mentioned in my previous message, calibre will insert the metadata into an epub as part of a Send to device or Save to disk operation. The copy in the library is not modified, the copy sent to the device or written to disk is.

Originally Posted by jist

But then again, I have no clue if or how epubs support metadata.
I am acquainted with how metadata works for mp3 and flac audio files, but I guess it's a completely different ballgame for epubs.

You can find quite a bit of information on the web. Check out Check What’s in an ePUB?: The OPF File for a quick bit of information on the .opf file and it's contents. Then you can move on to Dublin Core metadata, etc.

[jist]

Great, it's becoming clear to me.
Thanks!

[DNSB]

Originally Posted by jist

Great, it's becoming clear to me.
Thanks!

Hmmm... might be time for a sanity roll.

[jbacelar]

Originally Posted by DNSB

When you Send to device for Save to disk, the content.opf will be updated to match the metadata in the database..

This will only happen if in Preferences - Save books to disk, the option to Update the metadata of the saved copies is checked.

[JSWolf]

Metadata for ePubs is stored in the OPF file inside the ePub.

[theducks]

STD does also allow you to export a separate copy of the cover or OPF along with the format.
This is in addition to the books structure (some formats may not contain these in an easy to use form)

[JSWolf]

Originally Posted by theducks

STD does also allow you to export a separate copy of the cover or OPF along with the format.
This is in addition to the books structure (some formats may not contain these in an easy to use form)

You do know what an STD is I hope.

[DNSB]

Originally Posted by JSWolf

You do know what an STD is I hope.

Hmmm.... STD is a TLA for standard wall thickness. Not to mention quite a few other meanings.

[ownedbycats]

Originally Posted by DNSB

Not to mention quite a few other meanings.

https://en.wikipedia.org/wiki/STD

[hobnail]

Originally Posted by ownedbycats

https://en.wikipedia.org/wiki/STD

Hmm, "short term disability"; yeah my brain suffers from those all the time. Sometimes also called a senior moment.

[Freakeao]

Is there a way to find out which EPUBs have content.opf that is out of date (or different) from what is stored in the Calibre database? I'd like to have all my EPUBs with matching metadata compared to calibre, because I use other programs to serve OPDS feeds and sometimes just email someone a EPUB that I have.

Going forward what is the easiest way to keep content.opf updated when I update calibre metadata?

Thanks.

[DNSB]

Originally Posted by Freakeao

Is there a way to find out which EPUBs have content.opf that is out of date (or different) from what is stored in the Calibre database? I'd like to have all my EPUBs with matching metadata compared to calibre, because I use other programs to serve OPDS feeds and sometimes just email someone a EPUB that I have.

Going forward what is the easiest way to keep content.opf updated when I update calibre metadata?

Thanks.

As part of my import to calibre process, I import to an Intake library where I update the metadata and covers, check the epub structures, and fix any errors. Once that process is finished, I run Modify ePub to add the updated covers and metadata to the files and remove some crud before moving them to my Main library.