Update Microsoft old font links to use Internet Archive since many no longer work

[ronaldtse]

We don't even need to list out the original microsoft.com paths because they won't work.

These bad links lead to errors like these:

• https://github.com/metanorma/metanorma-docker/actions/runs/23124968517/job/67169032813?pr=231

C:/tools/ruby34/lib/ruby/gems/3.4.0/gems/fontist-2.1.4/lib/fontist/resources/archive_resource.rb:32:in 'Fontist::Resources::ArchiveResource#download_file': Invalid URL: https://gitlab.com/fontmirror/archive/-/raw/master/EUupdate.EXE. Error: #<Down::ConnectionError: An operation was attempted on something that is not a socket. - SSL_connect>. Invalid URL: https://download.microsoft.com/download/a/1/8/a180e21e-9c2b-4b54-9c32-bf7fd7429970/EUupdate.EXE. Error: #<Down::ConnectionError: An operation was attempted on something that is not a socket. - SSL_connect>. (Fontist::Errors::InvalidResourceError)
	from C:/tools/ruby34/lib/ruby/gems/3.4.0/gems/fontist-2.1.4/lib/fontist/resources/archive_resource.rb:20:in 'Fontist::Resources::ArchiveResource#archive'
	from C:/tools/ruby34/lib/ruby/gems/3.4.0/gems/fontist-2.1.4/lib/fontist/resources/archive_resource.rb:16:in 'Fontist::Resources::ArchiveResource#excavate'

As a backup we need to ensure all the links we use are backed by archive.org. I wonder if we can write a script to ensure all the font files we refer to are backed by archive.org?

I've manually made some of them here, but maybe it should be batch automated.

EUupdate.EXE:

• https://web.archive.org/web/20061219192404/https://download.microsoft.com/download/a/1/8/a180e21e-9c2b-4b54-9c32-bf7fd7429970/EUupdate.EXE
• https://web.archive.org/web/20260316043734/https://gitlab.com/fontmirror/archive/-/raw/master/EUupdate.EXE

ExcelViewer.exe:

• https://web.archive.org/web/20260316043732/https://gitlab.com/fontmirror/archive/-/raw/master/ExcelViewer.exe

GBEXTSUP.msi:

• https://web.archive.org/web/20260316043913/https://gitlab.com/fontmirror/archive/-/raw/master/GBEXTSUP.msi

[ronaldtse]

From Grok:

To ensure a webpage is listed (i.e., archived) in the Internet Archive's Wayback Machine using curl, you can follow a two-step process: first, check if it's already archived via their availability API, and if not, submit it for archiving. This can be done entirely from the command line. Note that the Internet Archive may have rate limits or require CAPTCHA for frequent saves, but for occasional use, this works.

Step 1: Check if the Page is Already Archived

Use curl to query the Wayback Machine's availability API. This returns a JSON response indicating whether snapshots exist.

curl -s "https://archive.org/wayback/available?url=https://example.com/your-page"

• Replace https://example.com/your-page with your target URL.
• The -s flag silences progress output for cleaner results.

Example response if archived:

{
  "url": "https://example.com/your-page",
  "archived_snapshots": {
    "closest": {
      "status": "200",
      "available": true,
      "url": "http://web.archive.org/web/20230101000000/https://example.com/your-page",
      "timestamp": "20230101000000"
    }
  }
}

• If "available": true, it's already listed (archived).
• If "available": false or the archived_snapshots object is empty, proceed to Step 2.

Step 2: Submit the Page for Archiving (if Not Already Listed)

If the check shows it's not archived, use curl to request a new snapshot. This is done via a GET request to the save endpoint.

curl -s -I "https://web.archive.org/save/https://example.com/your-page"

• Again, replace https://example.com/your-page with your URL.
• The -I flag fetches only the headers, which include the status and the new archive URL if successful.
• Without -I, it would return the full HTML, but headers are sufficient to confirm.

Look for a response header like Content-Location: /web/20260316125000/https://example.com/your-page (the timestamp will vary). A 200 OK status usually means success, and the page should be queued for archiving (it may take a few minutes to become available).

Automation Tip

You can combine these in a simple Bash script to check and save if needed:

#!/bin/bash
URL="https://example.com/your-page"
CHECK=$(curl -s "https://archive.org/wayback/available?url=$URL" | grep -o '"available": true')

if [ -z "$CHECK" ]; then
  echo "Not archived yet. Submitting..."
  curl -s -I "https://web.archive.org/save/$URL"
  echo "Submitted. Check availability again in a few minutes."
else
  echo "Already archived."
fi

• Save this as archive_check.sh, make it executable (chmod +x archive_check.sh), and run with ./archive_check.sh.

Additional Notes

• Rate Limits: The Internet Archive discourages rapid automated saves. Use sparingly to avoid blocks.
• Errors: If you get a 429 (Too Many Requests) or CAPTCHA prompt, try manually via a browser at web.archive.org/save.
• Verification: After saving, re-run the Step 1 check to confirm it's listed.
• Alternatives: For bulk or programmatic use, consider their official APIs or tools like the Wayback Machine's browser extension, but curl suffices for simple cases.