[GUI Plugin] FanFicFare

[JimmXinu]

Originally Posted by belton180

When updating multiple fics from sufficient velocity, all failed with the following error: Background Job Failed, see Calibre Jobs log.

After checking each fic one by one, only one fic remained with the error.

The URL is https://forums.sufficientvelocity.co...-quest.136754/ and attached is a copy of the debug log.

That a problem we've seen before in github issue 1298. In fact, it was with the same story.

It's caused by Calibre's image processing not handling SVG files well in background jobs (it's Qt's fault, as usual).

FFF has code to not process SVG files, but it looks specifically for '.svg' and that story has SVG files named like ".../research_sfdp-svg.7211/"

You can avoid it by using no_image_processing_regexp to not do image processing on any file containing 'svg' in the URL:

[https://forums.sufficientvelocity.com/threads/vox-vitae-warhammer-ai-quest.136754]
no_image_processing_regexp:svg

This has come up twice now, but OTOH, it's in the same story. I will do a bit more investigation about whether that setting should be in defaults.ini, or the .svg avoidance code expanded. Or we chalk it up to one weird story...

[HoppHagg]

Hi JimmXinu,

I've attached the requested the log.

Thank you :-)

[HoppHagg]

Originally Posted by HoppHagg

Hi JimmXinu,

I've attached the requested the log.

Thank you :-)

For comparison I've attached the debug log for version 4.53, maybe that's helpful. Thanks :-)

[JimmXinu]

Originally Posted by HoppHagg

Hi JimmXinu,

I've attached the requested the log.

Thank you :-)

This is being caused by embedded fonts in CSS being interpreted by FFF as images. I forgot CSS url() could be used for fonts when I wrote that code.

To determine how FFF should handle this, I first need to know how did you get embedded fonts in the first place?

[HoppHagg]

I've downloaded the font from here "https://papierlos-lesen.de/bookerly-fuer-cybook-kobo-pocketbook/" and installed them via 'windows/control panel/fonts'.

Then I use at first the 'modify epub' plugin and afterwards I use calibre's internal 'convert' function, where I can select to embed the desired font.

Thanks in advance, Marc

[JimmXinu]

Originally Posted by HoppHagg

I've downloaded the font from here "https://papierlos-lesen.de/bookerly-fuer-cybook-kobo-pocketbook/" and installed them via 'windows/control panel/fonts'.

Then I use at first the 'modify epub' plugin and afterwards I use calibre's internal 'convert' function, where I can select to embed the desired font.

Okay. I will fix FFF to not fail entirely on embedded fonts.

But FFF will not keep embedded fonts on update. That's how it worked in v4.53, too.

[HoppHagg]

JimmXinu, thank you very much for your quick replies :-)

Kind regards, Marc

[JimmXinu]

New Test Version Posted

2026-02-05
- Ignore CSS url() when ttf/otf/woff/woff2 font files

For HoppHagg, only matters if user has manually embedded font files. Note that FFF will discard the embedded font files on update, but will not crash now.

[HoppHagg]

Originally Posted by JimmXinu

New Test Version Posted

2026-02-05
- Ignore CSS url() when ttf/otf/woff/woff2 font files

This version works flawlessly, again, many thanks, JimmXinu :-)

Kind regards, Marc

[JimmXinu]

New Test Version Posted

2026-02-05
- Ignore CSS url() when ttf/otf/woff/woff2 font files
- Additional checks for svg images to reject--Calibre only. Related to #1298

Originally Posted by HoppHagg

This version works flawlessly, again, many thanks, JimmXinu :-)

Kind regards, Marc

Regarding belton180's issue with a weirdly named svg URL:

Spoiler:

https://forums.sufficientvelocity.com//attachments/research_sfdp-svg.7211/

I'm not prepared to reject all images with 'svg' in the URL--I can absolutely see someone naming their converted file logo_svg.jpg.

But I did add some code to look at the first 1000 bytes of each image for '<svg ' or 'xmlns="http://www.w3.org/2000/svg".

[steltek]

Originally Posted by texasnightowl

Thanks! I will go add it to Firefox and give it a whirl! Thankfully most of my reading these days is on AO3, but still!

If you are using a Firefox-based browser, there is an easier way to download long fics on Fanfiction.net and get them into Calibre.

Download and install the FicLab extension (it still works on Firefox, but not on Chrome-based browsers due to Google's Manifest v3 stupidity).

https://www.ficlab.com/

As part of Ficlab's download process, it actually caches the pages into the browser cache. It also adds wait times as it downloads to account for Fanfiction's rate limiting to avoid captchas and Cloudflare issues, making it convenient to use to download longer fics. You don't have to manually load chapters as it downloads them all automatically with the implemented rate limiting.

Once the entire fic is downloaded, Ficlab will prompt you to save it as an ePub. If the fic is complete (or abandoned and not likely to be updated), you can just save it as an ePub and then import the resulting ePub into Calibre. The ePub generated is not compatible with FFF, though, so you only want to do this for fics that are complete or which likely won't be updated (unless you want to re-download the entire thing every time it updates).

If it is an ongoing fic (or one that is unfinished that you expect might be updated in the future), hit cancel at the Ficlab save prompt. Then, use the Fanfiction URL for the fic in FFF to retrieve the downloaded fic via the browser cache. Once you have the current state of the fic imported to a compatible ePub via FFF, you can then use FFF to do future updates a chapter or two at a time as the fic is updated over time rather than having to re-download the entire fic again every time it updates.

[ownedbycats]

Is this feasible?

A [metadata]_NUM, used for conditionals, used with greater/less/equal operators.

One problem this would solve is my Star Trek: Original Series/Strange New Worlds problem - I would able to exclude SNW only if there's 2 or more fandoms

[JimmXinu]

I'd need to see more than one obscure use case before I'd be interested in that.

And it would still suffer from recursion blocking if you tried using it on the same entry. So, no, I don't think it would solve your problem.

[ownedbycats]

Ah, I was hoping it could be a way around the recursions

[marshie250]

Has something recently changed in the way anthology/series downloads are processed? I used to be able to delete a chapter (for example, if an author edited it and I wanted the newer version) or an image from the book using the Edit Book window, and then Update Anthology would redownload it. Now, I get the message "There is no item named 'OEBPS/images/ffdl-8.jpg' in the archive" (for example).

I do see these errors in theEdit Book window when I use the Check Book option:

for content.opf and toc.ncx files:


"The file 2/content.opf has its MIME type specified as origrootfile/xml in the OPF file. The recommended MIME type for files with the extension "opf" is application/oebps-package+xml. You should change either the file extension or the MIME type in the OPF.
Change the MIME type for this file in the OPF to application/oebps-package+xml"