[GUI Plugin] FanFictionDownLoader

[JimmXinu]

Originally Posted by Saioko

Oh, not a big problem, but I wanted to ask. Right now it grabs links to all stories - both by author and in Favorite stories. A while ago (a few months at least) it only got stories by author. Was it a deliberate change or something was changed on fanficton.net and it is now impossible to separate those two blocks of links?

The 'get urls from page' feature is deliberately very general. Until recently, I don't think it worked with default ffnet author pages at all. So while I didn't know it's working like that, I'm also unlikely to do anything about it.

But what you can do is use the mobile site to see just an author's own stories or favorites. Example

[DownWithIcarus]

Originally Posted by JimmXinu

Thanks, Doc.

I've very reluctant to do that. ffnet especially puts only a few reviews on each page, so there are often dozens of pages of reviews for even relatively short stories. Since you already have a tool that gets reviews, I'm even more reluctant.


If you use 'Update EPUB if New Chapters', that should essentially be the default behavior. It only fetches new (additional) chapters, keeping the old text for chapters it already has. (The exception is AO3 when 'do_update_hook' is set to true because it allows new chapters to be inserted in the middle of existing stories.)

A Part I: I do have a 'review' tool, it's true, but I find Calibre FanFiction Downloader so versatile and a very accessible tool, so I thought I would ask. It would be for a 'niche' user, so I can understand not wanting to take it on.

A Part II: Oh, I didn't notice that before. That's a very useful tip, thank you. I have always used html, but I will definitely switch.

Thank you.

[FaceDeer]

Originally Posted by DownWithIcarus

A Part II: Oh, I didn't notice that before. That's a very useful tip, thank you. I have always used html, but I will definitely switch.

As something that might also help you detect cases where an author has gutted their story, a trick I use in my library is that I have separate word-count columns for the word count reported by the site and the word count reported by the "count words" plugin. And then I've got a third column that shows the ratio between those two, which I can sort to find stories where the ratio is suspiciously askew one way or the other. This only works if you're doing the "update epub" option, otherwise the whole story gets overwritten and the two numbers will always match fairly closely (though never exactly, of course).

If you want to get *really* paranoid, you could create a custom column called "old word count" and every time you want to do an update you could first run a regex that copies the current word count value into it. Then, after updating, you'll be able to see what changed. If one of the stories got gutted despite using "update epub", dig up a copy from your library's backup and restore it (everyone should keep backups of their important data. ) Calibre can save search-and-replace operations so it should only take a few extra mouse clicks to do this each time.

[fidofido300]

Hi
When I download
http://storiesonline.net/s/70850/a-land-beyond

story is downloaded ok but it does not display the colors of the word which the author defined indivudually.

any Help will be appriciated
sorry for bad english

[cryzed]

This most likely won't be added. It would require either taking the site's CSS and merging it into the resulting e-book file (which might result in strange layout issues etc.), or creating special cases in the adapter to transform the span-tags to either contain their CSS directly in a style-attribute or turn them into font-tags with a matching color-attribute (even if inaccurate) -- which both is hard to automate without a DOM-aware HTML parser (such as a real browsers) or a proper CSS parsing library as far as I know. Then the question also comes up where to stop: Should we also add custom fonts, and other strange CSS definitions for other sites etc.

But maybe Jimm has a completely different opinion, so wait for his response too.

[davidfor]

As cryzed said, the CSS from the site is needed. It's been a while since I built an epubs using that CSS, but from memory, it worked fine. Though looking at the current version, I would strip everything except the margin settings from the body style in the first line.

[ILB]

If I try to download https://www.fanfiction.net/s/2987160/1/ I get the following message:
File is not a zip file

I tried with restarting calibre but the error message stays the same.

[Firedancer885]

Originally Posted by ILB

If I try to download https://www.fanfiction.net/s/2987160/1/ I get the following message:
File is not a zip file

I tried with restarting calibre but the error message stays the same.

I got that down without any problems. Check your library. Maybe Calibre sees an already existing corrupted file.

[JimmXinu]

Originally Posted by cryzed

This most likely won't be added. It would require either taking the site's CSS and merging it into the resulting e-book file (which might result in strange layout issues etc.)
...

You can incorporate as much (or as little) of that CSS file as you like into your EPUB CSS for storiesonline.net in personal.ini. The easiest way is probably like this:

[storiesonline.net]
## Allow customization of CSS.  Make sure to keep at least one space
## at the start of each line and to escape % to %%.
add_to_output_css:
 .red{color:red;}
 .blue{color:navy;}
 .lblue{color:blue;}
 .green{color:green;}
# ... and so on

Since storiesonline.net appears to provide that CSS file for all stories, I could be persuaded to include a set of the CSS for all storiesonline.net stories in the defaults.ini.

But I'd want somebody who actually reads storiesonline.net stories to determine which parts of the CSS file should be included by default.

[JimmXinu]

Originally Posted by ILB

If I try to download https://www.fanfiction.net/s/2987160/1/ I get the following message:
File is not a zip file

I tried with restarting calibre but the error message stays the same.

Downloads fine for me, too.

That error suggests that your epub has been corrupted.

FFDL used to have a problem with removing epub if you can FFDL's auto-convert turned on and calibre's default convert output also set to epub. But that was fixed a while ago.

Try downloading as 'overwrite always' or deleting the book and downloading it fresh.

[JimmXinu]

Test version with site fhsarchive.com added using eFiction Base adapter.

(eFiction Base adapters are generally less specific gathering metadata.)

[Tanjamuse]

I keep get the following error on the same page from AO3 ...

but I can't post it, because it says there are too many images in the "error message".

[cryzed]

I'm not sure what you mean by that, but if it's readable on your PC try taking a screenshot of it and upload it somewhere.

[Tanjamuse]

I don't know what happened, but it's working fine now.

[watani]

At some point between version updates, it seems that the fix for appending "http:" to the source URLs of FimFiction in-story pictures was dropped. Could the fix please be reapplied?