(X)HTML Metadata

rogue_ronin · 12-31-2009, 05:12 PM

What metadata does Calibre support for (X)HTML input? Is there a setting somewhere? I couldn't find anything in the User Manual, but I may have missed it. My experiment with importing XHTML wasn't too effective.

It seems to be able to output such data well, though.

We're having a discussion of Sigil's support for Dublin Core metadata in the <head> of (X)HTML docs used as input sources here. Near the bottom is a list of possible DCTERMS that cover most useful metadata as they relate to book collections, and some outliers. There is a specific list of what Sigil currently supports here.

Zipped, single-file XHTML is my storage format of choice, as it should convert easily to pretty much anything. I also edit a lot of files into decently marked-up XHTML. It'd be nice to have Calibre recognize the info.

Thanks!

m a r

kovidgoyal · 12-31-2009, 05:37 PM

IIRC the HTML metadata reader is optimized for the output of the ereader2html script as that is the most common use case.

rogue_ronin · 12-31-2009, 06:57 PM

Really?

I just imported 180 or so Doc Savage books from Blackmask, using the "Add books from directories, including sub-directories" option of Calibre. Did a fine job, but the title could only be grabbed from the filename. I deleted them, then wrote a bash script to rename all the files from the folder names (which had the full titles) and re-imported. Then a bulk metadata edit took care of most of the rest.

So, not terrible. And these books don't have the best choices of metadata or particularly stringent values eg:

Code:

<TITLE>THE MAN OF BRONZE</TITLE>
<META NAME="Author" CONTENT="A Doc Savage Adventure by Kenneth Robeson">
<META NAME="Description" CONTENT="Mystery, Suspense, History, Gothic, Literature, Books, Arts">

and the original filename was manbronze.htm. Still, it might be useful, especially in the future as things might get better tagged -- if apps support it.

Sort of a chicken-and-egg thing.

Thanks,

m a r

kovidgoyal · 12-31-2009, 09:24 PM

Hmmm maybe html metadata reading is broken, opena ticket and I'll take a look at it when I have some ntime.

rogue_ronin · 12-31-2009, 10:01 PM

Sure thing.

Thanks,

m a r

EDIT: Done!

12-31-2009, 05:12 PM	#1
rogue_ronin Banned Posts: 475 Karma: 796 Join Date: Sep 2008 Location: Honolulu Device: Nokia 770 (fbreader)	(X)HTML Metadata What metadata does Calibre support for (X)HTML input? Is there a setting somewhere? I couldn't find anything in the User Manual, but I may have missed it. My experiment with importing XHTML wasn't too effective. It seems to be able to output such data well, though. We're having a discussion of Sigil's support for Dublin Core metadata in the <head> of (X)HTML docs used as input sources here. Near the bottom is a list of possible DCTERMS that cover most useful metadata as they relate to book collections, and some outliers. There is a specific list of what Sigil currently supports here. Zipped, single-file XHTML is my storage format of choice, as it should convert easily to pretty much anything. I also edit a lot of files into decently marked-up XHTML. It'd be nice to have Calibre recognize the info. Thanks! m a r

12-31-2009, 06:57 PM	#3
rogue_ronin Banned Posts: 475 Karma: 796 Join Date: Sep 2008 Location: Honolulu Device: Nokia 770 (fbreader)	Really? I just imported 180 or so Doc Savage books from Blackmask, using the "Add books from directories, including sub-directories" option of Calibre. Did a fine job, but the title could only be grabbed from the filename. I deleted them, then wrote a bash script to rename all the files from the folder names (which had the full titles) and re-imported. Then a bulk metadata edit took care of most of the rest. So, not terrible. And these books don't have the best choices of metadata or particularly stringent values eg: Code: <TITLE>THE MAN OF BRONZE</TITLE> <META NAME="Author" CONTENT="A Doc Savage Adventure by Kenneth Robeson"> <META NAME="Description" CONTENT="Mystery, Suspense, History, Gothic, Literature, Books, Arts"> and the original filename was manbronze.htm. Still, it might be useful, especially in the future as things might get better tagged -- if apps support it. Sort of a chicken-and-egg thing. Thanks, m a r

12-31-2009, 10:01 PM	#5
rogue_ronin Banned Posts: 475 Karma: 796 Join Date: Sep 2008 Location: Honolulu Device: Nokia 770 (fbreader)	Sure thing. Thanks, m a r EDIT: Done! Last edited by rogue_ronin; 12-31-2009 at 10:31 PM.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
HTML Metadata	drsassafras	Calibre	10	09-17-2010 02:56 PM
Recognition of author and title from html files/reading metadata from a seperate file	Lethe	Calibre	5	04-03-2010 08:35 AM
"metadata" (toc) in HTML documents	pedz	Calibre	8	03-30-2010 09:23 PM
metadata out of Html	horseman	Calibre	0	08-04-2009 08:34 AM
Wide margins in html to epub; font size mngmt; PDF metadata	dementrio	Calibre	2	08-01-2009 01:33 AM

12-31-2009, 05:37 PM	#2
kovidgoyal creator of calibre Posts: 43,858 Karma: 22666666 Join Date: Oct 2006 Location: Mumbai, India Device: Various	IIRC the HTML metadata reader is optimized for the output of the ereader2html script as that is the most common use case.

12-31-2009, 09:24 PM	#4
kovidgoyal creator of calibre Posts: 43,858 Karma: 22666666 Join Date: Oct 2006 Location: Mumbai, India Device: Various	Hmmm maybe html metadata reading is broken, opena ticket and I'll take a look at it when I have some ntime.