Direct Links to archive.is

[muescha]

Looking through the code, I found the following methods interacting with archive.is - but there are some direct links available - I would like to know your opinion before I try to reinvent the wheel and create a PR to use the direct links:

Available Links - Documentation

• https://archive.ph/faq
• https://wiki.archiveteam.org/index.php/Archive.today

1) Direct Link for Opening a URL

https://${workingHostname}/?run=1&url={url}
vs
https://${workingHostname}/latest/{url}

Benefit:
No need to perform the "is this an archive website"-check and simulate a click on the form:

Userscripts/others/Paywall redirect to Archive.today.user.js

Lines 296 to 316 in a8a1aac

	{
	hostname: 'archive',
	check: (doc, win) => {
	const input = doc.querySelector('#HEADER form input[name="q"]');
	if (!input || !input.value) return false;
	let inputHostname
	try {
	const url = new URL(input.value)
	inputHostname = url.hostname
	} catch (err) {
	console.warn('Invalid URL in input:', input.value)
	return false
	}
	return sites.some(site =>
	site.hostname !== 'archive' &&
	inputHostname.includes(site.hostname) &&
	site.check(doc, win)
	)
	},

2) Direct Link for History

https://${workingHostname}/{url}

No need to manually find the history URL — maybe just keep the check to see if multiple versions are available:

Userscripts/others/Paywall redirect to Archive.today.user.js

Lines 317 to 328 in a8a1aac

	action: (doc, win) => {
	// Redirect to history of this page, if there is also a paywall in this archive
	// Only redirect once for this session
	const key = doc.location.href
	const alreadyRedirected = win.sessionStorage.getItem(key)
	const historyLink = Array.from(doc.querySelectorAll('#HEADER form a'))
	.find(e => e.textContent.includes('history'))
	if (!alreadyRedirected && historyLink) {
	win.sessionStorage.setItem(key, '1')
	historyLink.click()
	}

3) Check for Working Hostname

As mentioned in the doc documentation links , the archive.today domain is used as a load balancer and always redirects to a working TLD.

Or is there another reason behind the "check for working hostname" - maybe it is more easy to always use the archive.today domain?

Userscripts/others/Paywall redirect to Archive.today.user.js

Lines 160 to 175 in a8a1aac

	let workingHostname = null
	for (const hostname of hostnames) {
	try {
	window.setTimeout(() => showSpinner(hostname), 0)
	await checkAvailability(hostname)
	workingHostname = hostname
	break
	} catch (err) {
	if (err && 'message' in err && err.message === 'HOST_UNAVAILABLE') {
	console.debug(`${hostname} is NOT available`)
	} else {
	throw err
	}
	}
	}

[cvzi]

Fine with me if you want to create a PR.

Regarding "3) Check for Working Hostname":
When I added that check, the domains were very unreliable. I don't think that is the case anymore, so it is probably fine to remove the check and stick with archive.today