#apps script

Coherently flowing prose often fail me when shifting to a technical discussion, so to break down the work:

this was all in JavaScript. I obviously love Kotlin more these days, but it felt painless/rather accessible being in JS

the primary 'key' passed around for the entries (when participents submitted updates and final entries) tended to just be the fic URL (plus gifter/giftee names) - a good UUID but carries no meta info

almost (tho not all) of the entries were hosted on AO3 which made programatic scraping possible

we definitely needed the author & fic name... which required visiting the page. And once you're on the page, why not grab more details? In the end we wound up scraping: title, author, last update, word count, fandom, rating. And while there's more one could collect, we lacked the space to visually display it along with the results. Remember-- the point is the typesetting!

Apps Script let you make URL requests [getContent]

I used the Cheerio library to process the page's contents and let me comb through it using jQuery. Link goes to GitHub page, which has the Script ID you'll need to import it in the Libraries section.

I found the Google Sheets APIs easy enough to read/work with -- just lots of arrays, really. Also, the quotas that you get seem generous to me (for a free product)

AO3 specific note : had to append "?view_adult=true" to the URLs if I wanted to bypass that check screen, but I also had to handle locked fic, which keep you at a reduced info lock screen. Be sure to test E rated & locked fic when drafting your script (Incognito Window is your friend)

Here is the Gist of the pictured script-- it was a 'round 2' edit where I went back and pulled Rating & Locked status : [LINK]

The above refers mostly to the Sheets side of things. The other most magical feature was the population of Slides w/ sheet data!

Notes as follows:

images are my current weakness-- had to go in later and adjust/layout/upload images manually. SUPER LAME. Next time I'm going to dig into pulling images from the drive...

I was able to embed links programmatically into the slides but it was VERY HACKY code. The generation of the PDF links was still a manual task, and while I imagine I could script that, sounds hairy (I want to avoid all things Auth)

The {{name_pattern}} was/is only a stylistic choice. Doing a find & replace on something so distinct is safest tho.

I did not use Alt Text on my things but I think going forward, it would be how I'd ID/address the non-string replacement actions (finding the right shape to anchor a URL link on was finicky the way I did it)

Here is my (very messy!) script for populating the slides: [LINK]

For next time:

using a script to pull tags/fandoms/pairings will make the gifter/giftee pairing process sooooo much smoother (so far it's been very manual!! Bless @aetherseer's tenacity!)