Automate creating the IF Archive zip of zips

[dfabulich]

https://if-foundation.slack.com/archives/C1UU4KDEH/p1696301742105999

It’s currently a manual operation that I really need to automate. Easy to make mistakes

[dfabulich]

@mjmusante Can we make sure to prioritize this issue this year? This is the main issue creating hours of manual effort for IF Archive intake. Once the big zip creation is automated and in a standard file layout, we can write our scripts around that.

[dfabulich]

Copying and pasting some notes here from Slack:

@erkyrath wrote:

The easiest format for the archive is "a zip of zips", such that we go into games/competition2026 and type unzip IFComp2026.zip and it comes out looking like https://ifarchive.org/indexes/if-archive/games/competition2025/ and we say "yay all done"

For the names, I observe a de facto policy of "sluggifying" (spaces to underscores, drop non-ascii characters) which I guess includes hand-tweaking if there's a name collision.

I wrote:

I think it would be helpful to have some clear mapping of IFComp entry IDs to sluggified folder name

That doesn’t have to be in the zip, per se; it could live in the JSON in https://ifcomp.org/comp/2025/json

You could imagine:

"id": 3094,
"title": "The Wise-Woman's Dog",
"slug": "Wise-Woman's_Dog"

(We could embed the JSON in the zip, or fetch it from the live site, whatever’s easiest for you)

Which is to say, one of my goals for automating the big zip is to fully automate intake, including metadata

The automatically generated IFDB entries each know their own IFComp entry ID. (That’s how populate_ifdb_ids.pl works today.)

So some deterministic way of saying “folder X = entry ID Y = IFDB TUID Z” would certainly help.