imgbrd-grabber
imgbrd-grabber copied to clipboard
Adding a metadata.sqlite to Grabber?
As the title says, while using md5 as the common ID between websites tags for each single files?
%website%
would have to give the website's URL (without HTTP and trailing slash) for each board.
%websiteid%
would have to give the website's URL (without HTTP and trailing slash) with the unique image ID for each board.
%websitename%
would have to give the website name as set in the sources' settings for each board.
For example with these (md5:6533cf0248229ac131607dcef46a74dd
):
https://danbooru.donmai.us/posts/4204575
https://gelbooru.com/index.php?page=post&s=view&id=5680029
https://tbib.org/index.php?page=post&s=view&id=8945870
%website%
would give (if the separator was a ;
semicolon):
danbooru.donmai.us;gelbooru.com;tbib.org
%websiteid%
would give (if the separator was a
blank space):
danbooru.donmai.us:4204575 gelbooru.com:5680029 tbib.org:8945870
%websitename%
would give (if the separator was a +
sign):
Danbooru+Gelbooru+The Big ImageBoard (TBIB)
Also combining all the useful tags like %artist%
, %copyright%
, %character%
, %model%
, %general%
, %rating%
, %all%
, etc...
The column of the database could be something like:
-
id
for a unique ID to find a specific item in the database. -
md5
for the specific file to which the tags apply. -
website
for the specific website source from which the following tags come. -
item
for the specific website's unique image ID. -
artist
for the specific website's%artist%
tags. -
copyright
for the specific website's%copyright%
tags. -
character
for the specific website's%character%
tags. -
model
for the specific website's%model%
tags. -
general
for the specific website's%general%
tags. -
rating
for the specific website's%rating%
tags. -
all
for the specific website's%all%
tags. - Etc...
You could achieve something like this using the SQL commands feature. Not quite sure what you're trying to achieve with this though given you need to download the images for this to work, since otherwise Grabber simply doesn't have access to this information.
Aren't you better off using a local booru for this?
The md5s.sqlite
is a similar idea although it's only a MP5 to location mapping on the disk.
I want to be able to update the metadata without having to redownload the files.
So this metadata.sqlite
would be used by Grabber to aggregate everything for updating metadata.
So I guess that I also suggest metadata updating as a feature. 🤔