[GUI Plugin] FanFicFare

[JimmXinu]

Originally Posted by eb21145

Back with another question with a probably very obvious solution: I'm trying to clean up the freeform tags that are populating in my custom AO3 column and I'm running into a problem where if a tag that was not initially excluded when the fic was first downloaded/loaded into Calibre is now excluded AND it's the only tag left in the column, Calibre won't delete the tag and leave the freeformtags column empty when I update the metadata.

custom_columns_settings 'r' (for replace) mode doesn't call set_custom() if there's no value to set. So it never clears a column. It's been that way for at least 10 years, so I'm reluctant to change it now.

(The FFF Custom Columns GUI setting, OTOH, does set empty values, but only knows about the 'standard' FFF metadata entries.)

You might consider clearing that column in mass and then using 'Update from Saved Metadata'.

I might consider adding another 'mode' for custom_columns_settings -- maybe 'R' for really replace. Would anyone else find that useful?

[Jade Aislin]

This ([^/&]+) worked mostly great.

There were a couple issues.

When dealing with more than two names and the '&' sign, it would mess up on configurations I didn't specify. I'm not even sure how I'm getting those.

ie. Moon & Sun & Reader became Moon & Reader& Sun

I'm not sure if there is an easier way to fix this, but I added this code:

 ships=>(?<! )(&amp;)=> \1

Another thing is my code struggles with names containing multiple parentheses, breaking the pairing up. Instead of Thor/Michael turning into Thor (Norse Mythology)/Michael (Supernatural) and Michael (Supernatural)/Thor (Norse Mythology), I get two separate entries. ie Michael (Supernatural) as one pairing and Thor (Norse Mythology) as another.


I don't know how to fix it and I already have two urls affected (https://archiveofourown.org/works/11...pters/27110763), (https://archiveofourown.org/works/57...pters/13190308).

My code:

 category,characters,ships,genre=> \(.*\)=>
 category=>Religion & Lore=>Mythology
 
 characters,ships=>(Þórr \| )?Thor=>Thor (Norse Mythology)&&category=>Norse Mythology
 characters,ships=>Gabriel=>Gabriel (Supernatural)&&category=>Supernatural
 characters,ships=>(?<!\()Lucifer( Morningstar)?=>Lucifer\1 (Supernatural)&&category=>Supernatural
 characters,ships=>Michael=>Michael (Supernatural)&&category=>Supernatural


####################
###        SHIPS         ###
###  EXCEPTIONS  ###
####################

### 4 names (name1/name2/name3/name4)
 ships=>^([^/&]+)(/| &amp; )([^/&]+)(/| &amp; )([^/&]+)(/| &amp; )([^/&]+)$=>\1\2\3\4\5\6\7\,\3\2\5\4\7\6\1\,\5\2\7\4\1\6\3\,\7\2\1\4\3\6\5

### 3 names (name1/name2/name3)
 ships=>^([^/&]+)(/| &amp; )([^/&]+)(/| &amp; )([^/&]+)$=>\1\2\3\4\5\,\3\2\5\4\1\,\5\2\1\4\3

### 2 names (name1/name2)
 ships=>^([^/&]+)(/| &amp; )([^/&]+)=>\3\2\1\,\1\2\3

 ships=>(?<! )(&amp;)=> \1

##Make sure certain names are not first

### 3 names (name1/name2/name3)
 ships=>^(Chloe Decker)(/| &amp; )([^/&]+)(/| &amp; )([^/&]+)$=>\3\2\5\4\1
 ships=>^(Elena Gilbert)(/| &amp; )([^/&]+)(/| &amp; )([^/&]+)$=>\3\2\5\4\1

### 2 names (name1/name2)
 ships=>^(Chloe Decker)(/| &amp; )([^/&]*)$=>\3\2\1  
 ships=>^(Elena Gilbert)(/| &amp; )([^/&]*)$=>\3\2\1  

 ships=>^(Original .+ Character)(/| &amp; )([^/&]*)$=>\3\2\1

[JimmXinu]

Originally Posted by Jade Aislin

This ([^/&]+) worked mostly great.

There were a couple issues.

When dealing with more than two names and the '&' sign, it would mess up on configurations I didn't specify. I'm not even sure how I'm getting those.

ie. Moon & Sun & Reader became Moon & Reader& Sun

I'm not sure if there is an easier way to fix this, but I added this code:

 ships=>(?<! )(&amp;)=> \1

If that works for you. The atomic group feature appears to be new in Python version 3.11 and I confess I haven't used it before.

My instinct would be to exclude the spaces and then add them back, something like this (warning untested code):

### 4 names (name1/name2/name3/name4)
 ships=>^([^/&]+) *(/|&) *([^/&]+) *(/| & ) *([^/&]+) *(/|&) *([^/&]+)$=>\1\2\3\4\5\6\7\,\3\2\5\4\7\6\1\,\5\2\7\4\1\6\3\,\7\2\1\4\3\6\5
## 3 & 2 are similar.  \s is an explicit spaces  (The config parser
## tends to discard trailing spaces.)
 ships=>(&)=> \1\s

That will also match Alice&Bob ships, because AO3 doesn't actually enforce spaces around & as far as I know.

Originally Posted by Jade Aislin

Another thing is my code struggles with names containing multiple parentheses, breaking the pairing up. Instead of Thor/Michael turning into Thor (Norse Mythology)/Michael (Supernatural) and Michael (Supernatural)/Thor (Norse Mythology), I get two separate entries. ie Michael (Supernatural) as one pairing and Thor (Norse Mythology) as another.


I don't know how to fix it and I already have two urls affected (https://archiveofourown.org/works/11...pters/27110763), (https://archiveofourown.org/works/57...pters/13190308).

My code:

Spoiler:

 category,characters,ships,genre=> \(.*\)=>
 category=>Religion & Lore=>Mythology
 
 characters,ships=>(Þórr \| )?Thor=>Thor (Norse Mythology)&&category=>Norse Mythology
 characters,ships=>Gabriel=>Gabriel (Supernatural)&&category=>Supernatural
 characters,ships=>(?<!\()Lucifer( Morningstar)?=>Lucifer\1 (Supernatural)&&category=>Supernatural
 characters,ships=>Michael=>Michael (Supernatural)&&category=>Supernatural


####################
###        SHIPS         ###
###  EXCEPTIONS  ###
####################

### 4 names (name1/name2/name3/name4)
 ships=>^([^/&]+)(/| &amp; )([^/&]+)(/| &amp; )([^/&]+)(/| &amp; )([^/&]+)$=>\1\2\3\4\5\6\7\,\3\2\5\4\7\6\1\,\5\2\7\4\1\6\3\,\7\2\1\4\3\6\5

### 3 names (name1/name2/name3)
 ships=>^([^/&]+)(/| &amp; )([^/&]+)(/| &amp; )([^/&]+)$=>\1\2\3\4\5\,\3\2\5\4\1\,\5\2\1\4\3

### 2 names (name1/name2)
 ships=>^([^/&]+)(/| &amp; )([^/&]+)=>\3\2\1\,\1\2\3

 ships=>(?<! )(&amp;)=> \1

##Make sure certain names are not first

### 3 names (name1/name2/name3)
 ships=>^(Chloe Decker)(/| &amp; )([^/&]+)(/| &amp; )([^/&]+)$=>\3\2\5\4\1
 ships=>^(Elena Gilbert)(/| &amp; )([^/&]+)(/| &amp; )([^/&]+)$=>\3\2\5\4\1

### 2 names (name1/name2)
 ships=>^(Chloe Decker)(/| &amp; )([^/&]*)$=>\3\2\1  
 ships=>^(Elena Gilbert)(/| &amp; )([^/&]*)$=>\3\2\1  

 ships=>^(Original .+ Character)(/| &amp; )([^/&]*)$=>\3\2\1

NGL, I'm confused, too.

If you remove the '(' ')' chars from the lines for Thor and Gabriel , they become "Thor Norse Mythology Norse Mythology" and "Gabriel Supernatural Supernatural". The first line " \(.*\)=>" is masking that.

There's weirdness going on and honestly, I can't be bothered to figure out exactly what.

But I do have a solution that will also address a problem you haven't mentioned yet, at the 'expense' of sorting the ships--which you're then duplicating anyway, so I don't see the harm.

## Turned on so ships_CHARS is available.
sort_ships:true
# use ships_CHARS everywhere you want to affect a person in the ship.

replace_metadata:
 category,characters,ships_CHARS,genre=> \(.*\)=>
 category=>Religion & Lore=>Mythology

# Add ^ and $ around every ship char you want to change--
# otherwise you'll also change Thorn, Michael Landon, etc
 characters,ships_CHARS=>^(Þórr \| )?Thor$=>Thor (Norse Mythology)&&category=>Norse Mythology
 characters,ships_CHARS=>^Gabriel$=>Gabriel (Supernatural)&&category=>Supernatural
 characters,ships_CHARS=>^(?<!\()Lucifer( Morningstar)?$=>Lucifer\1 (Supernatural)&&category=>Supernatural
 characters,ships_CHARS=>^Michael$=>Michael (Supernatural)&&category=>Supernatural

# ...

## needs a trailing '$' also, otherwise it affects the 4 & 3 count lines.
### 2 names (name1/name2)
 ships=>^([^/&]+)(/| &amp; )([^/&]+)$=>\3\2\1\,\1\2\3

I believe you've been here long enough to have seen my rant on the ultimate futility of wrangling AO3 tags, so we can take it as read, yes?

[ownedbycats]

Is there any difference between using ePubMerge on some already-downloaded fanfics and using FFF to download an anthology?

[ownedbycats]

Glitch:

Set Calibre Series URL concatenates the URLs of every series present in an anthology.

[JimmXinu]

Originally Posted by ownedbycats

Is there any difference between using ePubMerge on some already-downloaded fanfics and using FFF to download an anthology?

FFF Anthology, depending on site (ie, AO3) will collect series status and description.

Originally Posted by ownedbycats

Glitch:

Set Calibre Series URL concatenates the URLs of every series present in an anthology.

Example story URL?

[eb21145]

Originally Posted by JimmXinu

custom_columns_settings 'r' (for replace) mode doesn't call set_custom() if there's no value to set. So it never clears a column. It's been that way for at least 10 years, so I'm reluctant to change it now.

Shoot, thought it might be something like that. Well I was planning to do a mass clean out than redownload from the saved metadata anyway.

Thanks!

[ownedbycats]

Originally Posted by JimmXinu

Example story URL?

For example, three stories from a series, the Calibre Series URL comes out as

https://archiveofourown.org/series/4838692,%20https://archiveofourown.org/series/4838692,%20https://archiveofourown.org/series/4838692

and 404s. It adds another for each story in the anthology.

[JimmXinu]

Originally Posted by ownedbycats

For example, three stories from a series, the Calibre Series URL comes out as

https://archiveofourown.org/series/4838692,%20https://archiveofourown.org/series/4838692,%20https://archiveofourown.org/series/4838692

and 404s. It adds another for each story in the anthology.

Please provide a (SFW) story URL that exhibits the problem when asking for help.

EDIT: If you mean you are downloading an FFF anthology for 3 of the 128 books in https://archiveofourown.org/series/4838692, I'm not seeing that happen.

My only guess would be that you are doing replace_metadata on the seriesUrl changing it into a list?

[ownedbycats]

Except for some conditionals, these are only refs to seriesURL in personal.ini:

include_in_series:series,collections
include_in_seriesUrl:seriesUrl

Not sure why I had the second one, tbh. I try removing.

EDIT: Didn't fix it.

[JimmXinu]

I put include_in_series:series,collections in and downloaded https://archiveofourown.org/series/4838692 as an anthology.

I'm not seeing anything unusual happen with seriesUrl, in the titlepage or putting series into custom column.

[ownedbycats]

It's the Calibre Series URL that's affected, the link attached to series column.

[JimmXinu]

New Test Version Posted

2026-04-21
- PI: Correct Series field url link when setanthologyseries (aka "Set 'Series [0]' for New Anthologies")

Originally Posted by ownedbycats

It's the Calibre Series URL that's affected, the link attached to series column.

The standard column Series link on an anthology? Not the identifier:url link? Series field isn't populated for series anthologies. The anthology is the series.

But apparently we have an option for that. After checking "Set 'Series [0]' for New Anthologies?", I do see this happen.

Note: "If checked, the Series column will be set to 'Series Name [0]' when an Anthology for a series is first created." Updating an existing anthology epub will not set either Series or Series link.

Also, FYI, FFF does not set links on series type custom columns.

[slimcat]

I just noticed that when personal.ini is set to EPUB3, anthologies include nav.xhtml files for each work even though the format is EPUB2, and EPUBCheck fails. It's an easy fix: either upgrade book internals to EPUB3, or if you want to keep to EPUB2, manually delete all the nav files. But just wanted to mention it.

While testing this I ended up putting together an Editor Chain that removes leftover package & TOC files, consolidates stylesheets, and upgrades the anthology to EPUB3. It keeps the root toc.ncx for backwards compatibility. (If you don't care about EPUB3 and just want to remove the unused package/TOC files & consolidate stylesheets, just delete steps 3-5).

Attaching it here in case anyone finds it useful, as well as an Action Chain for batch processing.