[GUI Plugin] FanFicFare

[Clem2605]

Hello!

Recently, the option to 'update Antology epub' doesn't work on my end, including in cases where it worked before. The error message tells me 'Book isn't an FanFicFare Antology or contains book(s) without valid Story URLs.'

My anthologies come exclusively from AO3 series and I know there was an issue recently with the changes AO3 made to paginate those bigger than 20 (?) stories, so I'm guessing the problem comes from here especially since the timeline fit. However, the issue comes with all anthologies I downloaded before you adapted the plugin, not only the bigger ones.

When I re-download the stories, it works, so my guess is that the URLs have changed and manually changing them could resolve the problem, but I don't know where those are.
Where do I need to modify the URLs so the FFF recognizes my books?

Thanks!

Edit: Sorry, I was debugging while writing so my post is a bit weirdly written, since I don't have that many anthologies, I hadn't noticed those downloaded recently work without issue.

[JimmXinu]

To the best of recollection, no, series URLs didn't change.

Are you sure you didn't edit/polish/convert/etc the anthology books in question?

Do you have one (SFW only please) you could post for me to look at?

[Clem2605]

Originally Posted by JimmXinu

To the best of recollection, no, series URLs didn't change.

Are you sure you didn't edit/polish/convert/etc the anthology books in question?

Do you have one (SFW only please) you could post for me to look at?

I only edit the metadata of my incomplete fanfictions, and I keep track of the complete ones I've edited, so it seems farfetched, especially since, as far as I can tell, all of my antologies downloaded before june/july are concerned.

I've included one example (it should be sfw), here is the AO3 link, if needed:
https://archiveofourown.org/series/3741151 (removed since I've found the issue)

Edit: Now that I think about it, I do convert from epub to epub when I want to change the default generated cover, I'm gonna try and see if it is the issue

Edit 2: Yes it was! Thank you! But now that we have established that, does that mean that I can't convert the antologies that I may have to update in the future or is there a way to make it work?

[JimmXinu]

Originally Posted by Clem2605

Edit: Now that I think about it, I do convert from epub to epub when I want to change the default generated cover, I'm gonna try and see if it is the issue

Edit 2: Yes it was! Thank you! But now that we have established that, does that mean that I can't convert the antologies that I may have to update in the future or is there a way to make it work?

Correct.

FFF's anthology feature relies on EpubMerge's UnMerge feature, which in turn relies on the epub not changing too much. In this case, not discarding 'unused' metadata files from the individual epubs.

Calibre's Convert, while a great tool, rewrites everything.

But! You can use Calibre's "Polish book" to update the cover with far less processing and impact. My tests show it doesn't break anthologies.

[Clem2605]

Thank you!

I had hoped to find a solution that didn't include re-downloading, but it seems this isn't my lucky day...

Have a good day!

[icecold]

Is there a way to normalize SV/SB urls so that if the thread title changes FFF/Calibre won't think it's a new ebook?

[JimmXinu]

Not currently.

We have special code in the plugin for fanfiction.net and scribblehub.com to search for existing story URLs ignoring the title part, but no other sites.

For ffnet, that code was added when we started keeping the title part of story URLs instead of discarding it. Much more recently, somebody nagged me about it on scribblehub until I added it. Some author was changing their story title every day or something.

You can work around it by choosing the story in your library and updating it from there instead of by URL (or email).


This isn't something I've seen very much on SB/SV. Are other people seeing it happen a lot? For specific authors/subforums/fandoms, or across board?

[icecold]

I know of one quest that changes its title every so often when the main character changes. It happened recently to one thread I follow that added Peggy Sue to the title. I want to say that it happens at least once every three months, though I haven't exactly been keeping track of how often it happens. One author I follow just went off hiatus and changed his two stories' titles to reflect that this week.

Edit: I haven't exactly noticed any particular one forum (SB/SV/QQ) doing thread renaming more often than the other forums.

[booster-t]

Here's a new one for me ... all of a sudden, I'm getting this error when trying to download from FFN. —> Browser Cache Failed to Load with error ''utf-8' codec can't decode byte 0x92 in position 0: invalid start byte'. <—

The 0x92 position changes from time to time. I haven't changed anythng since yesterday other than to clear the cache. I'm on a Mac OSX 14.6.1

Stephen

[JimmXinu]

It's hard to say from just that. Could be a simple problem, or it could be unfixable.

Can you get a debug log? Restart in debug mode under Calibre's Preferences dropdown, try to download, quit Calibre, save the debug log txt file. Zip the file, post it here.

[booster-t]

Originally Posted by JimmXinu

It's hard to say from just that. Could be a simple problem, or it could be unfixable.

Can you get a debug log? Restart in debug mode under Calibre's Preferences dropdown, try to download, quit Calibre, save the debug log txt file. Zip the file, post it here.

OK Jim ... here's the log.

Also, I tried this with an existing story and got a different offset. But, fichub can download the file and if I rename it to what the file is called in Calibre, it seamlessly appears. Just have to manually change the update date :-)

[JimmXinu]

The log says that a cache file was found, but that it failed to read the headers.

Works fine for me on Win10, so I borrowed a Mac and I do see the same error there.

Honestly, that's good news--troubleshooting a problem I can see is many times easier than one I can't. More to follow.

[JimmXinu]

Experimental Version Attached

2024-08-27
- New field in Chrome Simple cache?

Okay, it looks like Chrome 'simple' cache--on Mac at least--now has an extra field in the binary header. No idea what it is, but the fields we use are shifted over 4 bytes.

The attached version reads (and discards) the extra field. Plus a bunch of debug output.

The question is: is it just Mac, or all Chrome derived 'simple' cache using browsers? I know some Windows users end up with simple cache, but it's always 'block' cache on my computer.

So if some other people who use 'simple' can test it, that would help.

If you're running in debug mode, the output log will tell which cache version you're using. Search log for string Using. The options are "Using BlockfileCache", "Using FirefoxCache2", and "Using SimpleCache".

[booster-t]

Hmm ... no luck .. same error. Calibre 7.1.7 and plugin v4.37.2

Here's the log

Stephen

[booster-t]

Hello again Jim

I thought I'd try Firefox to see if that made a difference. When I go about:cache, Firefox returns this file name: ~Library/Caches/Firefox/Profiles/whbmctbq.default-release-1721846696620/cache2

I guess that's a hidden file because when I got to the /Profiles folder, it's empty even though my Mac says it's taking 56mb.

The plugin responds by saying: Directory does not contain a known browser cache type: '/~Library/Caches/Firefox/Profiles/whbmctbq.default-release-1721846696620/cache2'

Sometimes I want to go back in time when things just worked <sigh>

Stephen