From 645dc68900cc662ce59e06f448981dc9db3f47cd Mon Sep 17 00:00:00 2001
From: Nemo <commits@captnemo.in>
Date: Tue, 27 Jun 2023 12:23:17 +0530
Subject: [PATCH] [eks] Use web archive to run against 1.19 and 1.20

AWS is no longer publishing the history on GitHub, so we
instead rely on older copies of the website to track the old
release dates. This is done at 2 cutoff points right now for
1.19 and 1.20.

Older releases are available on the web archive, but those
pages didn't include release dates so not helpful.
---
 src/eks.py | 46 ++++++++++++++++++++++++++++------------------
 1 file changed, 28 insertions(+), 18 deletions(-)

diff --git a/src/eks.py b/src/eks.py
index bebecd38..bbf58e49 100644
--- a/src/eks.py
+++ b/src/eks.py
@@ -4,30 +4,40 @@ from bs4 import BeautifulSoup
 from common import endoflife
 from datetime import datetime
 
-URL = "https://docs.aws.amazon.com/eks/latest/userguide/platform-versions.html"
+# Now that AWS no longer publishes docs on GitHub,
+# we use the Web Archive to still get the older versions
+# Keep older pages at top of the list
+URLS = [
+    # 1.19.eks.1
+    "https://web.archive.org/web/20221007150452id_/https://docs.aws.amazon.com/eks/latest/userguide/platform-versions.html",
+    # + 1.20
+    "https://web.archive.org/web/20230521061347id_/https://docs.aws.amazon.com/eks/latest/userguide/platform-versions.html",
+    # + latest
+    "https://docs.aws.amazon.com/eks/latest/userguide/platform-versions.html",
+]
 REGEX = r"^(?P<major>\d+)\.(?P<minor>\d+)\.(?P<patch>\d+)$"
 
 
-def parse_platforms_page():
+def parse_platforms_pages():
     all_versions = {}
     print("::group::eks")
-    response = endoflife.fetch_url(URL)
-    soup = BeautifulSoup(response, features="html5lib")
-    for tr in soup.select("#main-col-body")[0].findAll("tr"):
-        td = tr.find("td")
-        if td and re.match(REGEX, td.text.strip()):
-            data = tr.findAll("td")
-            date = data[-1].text.strip()
-            if len(date) > 0:
-                d = datetime.strptime(date, "%B %d, %Y").strftime("%Y-%m-%d")
-                k8s_version = ".".join(data[0].text.strip().split(".")[:-1])
-                eks_version = data[1].text.strip().replace(".", "-")
-                version = f"{k8s_version}-{eks_version}"
-                all_versions[version] = d
-                print(f"{version}: {d}")
+    for url in URLS:
+        response = endoflife.fetch_url(url)
+        soup = BeautifulSoup(response, features="html5lib")
+        for tr in soup.select("#main-col-body")[0].findAll("tr"):
+            td = tr.find("td")
+            if td and re.match(REGEX, td.text.strip()):
+                data = tr.findAll("td")
+                date = data[-1].text.strip()
+                if len(date) > 0:
+                    d = datetime.strptime(date, "%B %d, %Y").strftime("%Y-%m-%d")
+                    k8s_version = ".".join(data[0].text.strip().split(".")[:-1])
+                    eks_version = data[1].text.strip().replace(".", "-")
+                    version = f"{k8s_version}-{eks_version}"
+                    all_versions[version] = d
+                    print(f"{version}: {d}")
     print("::endgroup::")
     return all_versions
 
-
-versions = parse_platforms_page()
+versions = parse_platforms_pages()
 endoflife.write_releases('eks', versions)