[Suggestion] EH JSON from API

[w158297]

Would it be possible to add an option to let the JSON data written for E-Hentai downloads be from their official API? https://ehwiki.org/wiki/API
The current implementation sadly leaves out a lot of data, although it provides information about Parent/ visibility which the API doesn't.

Also gallery-dl -j $gallery_page option for EH galleries starts going through all the pages instead of just showing the one for the requested URL

[mikf]

leaves out a lot of data

"archiver_key", "category", "thumb", "uploader" "rating", and "torrentcount" from what I can tell. And the total file size is only an rough estimate. Everything else just has a different name.
"category", "thumb", "uploader" "rating", and "torrentcount" would be easy enough to get from the gallery HTML page as well, only "archiver_key" would be missing.

I wouldn't mind adding an option to (also) get the metadata from the API, but that's 1 extra HTTP request which might not be necessary.

Also gallery-dl -j $gallery_page option for EH galleries starts going through all the pages instead of just showing the one for the requested URL

Use -K or -j --range 0

[w158298]

That would also work, thank you very much!

… from what I can tell

Also if there are torrents it gives you the hash,name, torrent+file size for them. Additionally the upload time is missing seconds, which doesn't matter I guess.

On a side note, about the current gallery-dl JSON: what is the "subcategory" entry referring to?
(and for some reason there are the two entries for language entries)

[hellupline]

now that we are adding more info from the EH-api, its possible to redownload only the metadata for those ?

[mikf]

Extracting a gallery_id wasn't the problem, but all format strings still expected the old metadata names, e.g. gallery_id instead of gid from the API. I've now renamed the old metadata fields to match their API counterparts, so this doesn't happen anymore (61fbbd2) and the metadata option is now a simple boolean value.

@hellupline I recommend --range 0 to skip all images, but this will still cost 1 image limit point per gallery.

[hellupline]

btw, I created this gist to upgrade metadata files to use gid and token and other new fields
gist