[GUI Plugin] FanFicFare

[JimmXinu]

Originally Posted by ownedbycats

Question: What does an "uninitialized error" mean in the browser cache? Thankfully I got it working after clearing cache and reloading the chapters into it.

No idea. That string doesn't appear in the FFF code.

Originally Posted by surpachup

I made an example in which:
- #tw is a custom column which should only contain death or mcd if any
- tags is the default tags column which should only contain alternate canon or modern



I tested it with this https://archiveofourown.org/works/41674563 and #kudos, #fandom, #notes and #ships (old settings) are populated but not tags and #tw

It works better to use [code] around config blocks rather than [quote]. The '#' button in the 'advanced' post editor.

add_to_replace_metadata:
 au=~Alternate Canon|Deviates From Canon|Alternate Universe \- Canon Divergence|[Ff]ix-[Ii]t=>alternate canon&&category=~Harry Potter|Cursed Child
 au=~[Mm]odern [Ee]ra=>modern
 tw=~Minor Character Death=>death
 tw=~Major Character Death=>mcd

You can't use =~ there in replace_metadata, it's only for include/exclude_metadata_pre/post or conditionals in replace_metadata. Yes, it's inconsistent.

include_in_tags:au

include_in_tags does nothing, FFF doesn't have a metadata entry named tags. (As standard--it can be defined for individual sites or by users.)

include_subject_tags:

The Calibre Tags column is treated special for historical reasons. The include_subject_tags setting controls what goes into Tags.

This clears the include_subject_tags setting, telling FFF not to set anything in Calibre's Tags field. (Unless you also have extra_subject_tags set--which is 3/4 deprecated.)

[ownedbycats]

Question: When I created my longtext metadata column a long time ago, I made it HTML longtext, which I recently found out was not good practice.

I created a second longtext column, this one plain-text. What would be the best method of changing all them over though?

Even without ffnet-specific problems, re-downloading metadata on every fic would be annoying. And search & replace also copies over the HTML tags.

[JimmXinu]

Just change the existing column.

Backup first, of course, but I tried and that worked fine for me--assuming you haven't manually changed any. But those would be broken either way.

[ownedbycats]

Oh, I didn't realize it was possible to change the longtext column type after the fact. Seems to be working fine

[surpachup]

Originally Posted by JimmXinu

You can't use =~ there in replace_metadata, it's only for include/exclude_metadata_pre/post or conditionals in replace_metadata. Yes, it's inconsistent.

Thank you for your help, I was able to make it work as I intended my personal.ini has never been this long but I can't go back now

[www.archiveofourown.org]

custom_columns_settings:
bookmarkrec=>#rec
kudos=>#kudos
add_to_extra_valid_entries:,au,tw
include_in_genre: freeformtags,genre,status
include_in_au:genre,freeformtags
include_in_tw:genre,freeformtags

replace_metadata:
 ships=>Draco Malfoy/Hermione Granger=>dramione
 fandoms=>Harry Potter \- J\. K\. Rowling=>Harry Potter
 fandoms=>Harry Potter and the Cursed Child \- Thorne & Rowling=>Cursed Child
 au=>.*[Mm]odern [Ee]ra.*=>modern\,au
 au=>.*Alternate Canon.*|Deviates From Canon|Alternate Universe \- Canon Divergence|.*[Ff]ix-[Ii]t.*=>alternate canon&&fandom=~.*Harry Potter.*|.*Cursed Child.*
 tw=>Minor Character Death=>death
 tw=>Major Character Death=>mcd
 
include_metadata_post:
 au==alternate canon
 au==modern
 tw==death
 tw==mcd

add_to_custom_columns_settings:
 tw=>#tw

include_subject_tags:au
add_to_include_subject_tags:,au
include_in_category:fandoms
include_in_freefromtags:freeformtags
include_in_notes: bookmarksummary

[dailyalice]

Originally Posted by ownedbycats

Oh, I didn't realize it was possible to change the longtext column type after the fact. Seems to be working fine

I was confused a bit at first about the Saved Metadata long text column and HTML versus plain text. And I was also surprised you can just flip that from HTML to Text and back again easily.

The "interpret this column as" option appears to tell calibre how to show you the column data (the data itself is not changed). If you choose HTML, when you look at the data, you see it as if you were looking at a web page (with all that formatting in use). If you choose "Text" it is interpreted as or shown as Text (which would be like viewing the source code of a web page, so you'd see all the tags and whatnot).

I like to sometimes look at Saved Metadata at the bottom of Book Details as a way of comparing what I have in my own columns to what data is available in Saved Metadata. So having it interpreted as HTML means I can easily read it there; whereas if it is interpreted as Text, it shows up as code (which I'm geeky enough to understand, but reading HTML code as code is not nearly as nice to look at or easy to read).

The problem comes if you edit Saved Metadata while you have it interpreted as HTML. If you have it visible on the Edit Metadata screen and edit it while it is interpreted as HTML, the code will get messed up and your Saved Metadata won't be useful anymore. It's because of how the HTML editor works, I guess, changing the code beneath what is visible on your screen. Whereas if you have it interpreted as and visible as text (code) in Edit Metadata, you probably could edit it safely. (Though it seems like a bad idea to edit it at all.)

In that epic calibre/FanFicFare guide I'm working on, I explain this (better, I hope, with screenshots to give examples). I also instruct folks to go to calibre Preferences / Look & feel / Edit Metadata and remove all the FanFicFare special columns (Saved Metadata, Errors, FFF Updated) from their Edit Metadata page, that way they won't mistakenly edit them (or be tempted to edit them). (It's always easy enough to return them if you need to get to them, but it's safest to not have them there.)

If I understand things correctly, if you don't touch Saved Metadata (on purpose or by accident), it should be safe to have it interpreted as HTML. If you ever decide you need to edit it for some reason, it's easy enough to change the column settings to "interpret this column as" to Text before you do anything.

(I wound up fetching data out of an older calibre Library where I'd named my Saved Metadata something different than I did in my current one. I ended up using bulk search & replace to move Saved Metadata from one column to another for some works and then I made darn sure both the columns involved were interpreted as text before I tackled it.)

I think I have this right (please correct me if I'm wrong).

[JimmXinu]

Originally Posted by dailyalice

I think I have this right (please correct me if I'm wrong).

It looks like you understand correctly to me.

FYI, even if you disable columns in Edit Metadata under Preferences / Look & feel / Edit Metadata, you can still edit those columns from the library book list just by double clicking when the columns are shown. So it's definitely 'hiding', not 'protecting'.

I still recommend leaving the Saved Meta column interpreted as Text to be a little bit safer.

Calibre changing the HTML in long texts automagically came along after the FFF Saved Metadata Column feature. I would not have used an HTML format when the feature was implemented if I had known that was coming.

[dailyalice]

Originally Posted by JimmXinu

It looks like you understand correctly to me.

Hooray! (And thank you letting me know.)

FYI, even if you disable columns in Edit Metadata under Preferences / Look & feel / Edit Metadata, you can still edit those columns from the library book list just by double clicking when the columns are shown. So it's definitely 'hiding', not 'protecting'.

Ah, this is good to know. I think the actual first thing I suggest in my guide (after column creation) is to return to the place where you add columns and remove the checkmarks from next to the new special columns so they don't show up in column view (what you call library book list I guess I've been calling "column view").

I still recommend leaving the Saved Meta column interpreted as Text to be a little bit safer.

Calibre changing the HTML in long texts automagically came along after the FFF Saved Metadata Column feature. I would not have used an HTML format when the feature was implemented if I had known that was coming.

That makes sense. And yeah, I guess you can't change it now given how many people have used Saved Metadata and still use it from old books/libraries, etc.

I do mention your recommendation in the guide and explain, so hopefully I won't steer anyone too far afield. I strongly urge that people leave that data alone, but I also find it useful to have folks look at Saved Metadata (in Book Details) so they can see what data FanFicFare is actually pulling down and then decide which bits they might want in their own columns.

I suppose once I get my library settings the way I want, I don't necessarily need to look at Saved Metadata myself very often, it's just when I'm messing around with replace_metadata and so on that i find it helpful.

I've been using "HTML" for my long text columns where I write up my own comments on works, but I think I'm going to switch to plain text or text with Markdown because the HTML editor window in Edit Metadata seems to always throw empty paragraph HTML between my paragraphs, which puts a whole lot of extra space between paragraphs and looks awful when viewed in Book Details. (Or maybe if I stick with HTML I should just write my comments in the "HTML Source" tab rather than "Normal view" since I'm used to typing things in HTML anyway and then presumably I wouldn't be surprised when calibre tries to help by throwing in extra paragraph HTML.)

[greenskye]

What setting would I use to stop FFF from populating the tag column with any values (for all sites). I only want to set tags manually and never pre-populate them. Based on the options it seems like I can only choose to set tags on new books, or if I uncheck that option it will add tags on every update as well. There doesn't seem to be any sort of global disable that I could find.

If there is no global disable, can someone help me disable tags from populating from Ao3? I'm confused by the ini settings and am not sure what to add or remove to stop this field from populating.

[JimmXinu]

Set both include_subject_tags and extra_subject_tags to be empty. Doing so under the [overrides] section will force it to apply to all sites:

[overrides]
include_subject_tags:
extra_subject_tags:

[ownedbycats]

Silly question:

Is it possible to
put a generate_cover_settings under the [website] sections to switch to AO3_smalltext/FFNet_smalltext if the title is above xx characters?

[JimmXinu]

Originally Posted by ownedbycats

Silly question:

Is it possible to
put a generate_cover_settings under the [website] sections to switch to AO3_smalltext/FFNet_smalltext if the title is above xx characters?

Should be possible using the {m,n} regex mechanism.

Overriding [defaults] generate_cover_settings will rely on using add_to_ appropriately--using generate_cover_settings under the [site] sections and add_to_generate_cover_settings under [defaults] so the [site] versions have higher precedence.

[dailyalice]

I'm working on a section of my guide to FanFicFare about replace_metadata (and include/exclude metadata) and I'm wondering if there is a list of available metakeys somewhere. Some of them are obvious, of course; I think the list of available title page entries plus extra_valid_entries for specific archives (like AO3) covers a lot of ground. But I know ships_CHARS is one of them not really covered there and I was wondering if there were any others I can't intuit from what is in plugin-defaults.

I could put together my own list and see if there's anything I'm missing, but wanted to check if there's already a list someplace. (I didn't find one when I searched this thread, but sometimes my searches seem to miss things.)

[JimmXinu]

In addition to entries in extra_valid_entries from defaults.ini or personal.ini, there are two lists in the code of all the valid standard metadata entries in the code:
Lists: https://github.com/JimmXinu/FanFicFa...urable.py#L143
Scalars: https://github.com/JimmXinu/FanFicFa...urable.py#L306
(As a general rule, extra_valid_entries will be list types.)

But I will save you the trouble and list them here:

author
authorHTML*
authorId*
authorUrl*
category
characters
cover_image*
dateCreated+
datePublished+
dateUpdated+
description
extratags
formatext*
formatname*
genre
langcode*
language*
lastupdate
numChapters*
numWords
output_css*
publisher
rating
sectionUrl
series
seriesHTML*
seriesUrl*
ships
site*
siteabbrev*
status
storyId*
storyUrl*
title
titleHTML*
version*
warnings
words_added* # logpage only.

* Entry used internally--should not be changed by the user.
+ Entry contains a date, use _format settings to customize, eg: dateUpdated_format.

Honestly, I would probably just leave out the * and + marked entries from the list for replace_metadata. If somebody really wants to customize one of those, they should make a copy entry and customize it instead.

x_LIST (where x is any list type metadata entry) and ships_CHARS are special cases that allow replace_metadata and in/exclude_metadata_pre/post to be used even though that aren't technically metadata entries.

I don't remember any other special cases.

[dailyalice]

This is really useful info, thanks!

Originally Posted by JimmXinu

Honestly, I would probably just leave out the * and + marked entries from the list for replace_metadata. If somebody really wants to customize one of those, they should make a copy entry and customize it instead.

Yeah, that makes sense. I think I will just list the ones without the * & +.

Though I just realized that I have a line in my personal ini right now under replace metadata to change site "archiveofourown.org" to "Archive of Our Own" thinking it looks nicer under Publisher in my calibre library.

I did similar things with ksarchive and other sites, come to think of it. I take it this is a bad idea? (Hmmm, if I could leave the site data alone and just change it under Publisher, that would make sense. Would I need to make a copy of the site data in a new custom column and then send that to the Publisher column?)

Edited to say: Oh, I could just change publisher instead of site and all would be well, I bet.