From 28865b078363c1023fe272c8caa99b39c32a80a1 Mon Sep 17 00:00:00 2001
From: Edward Betts <edward@4angle.com>
Date: Sun, 13 Oct 2024 16:27:38 +0100
Subject: [PATCH] Detect and handle redirects

Closes: #6
---
 check.py | 36 ++++++++++++++++++++++++++++++++++--
 1 file changed, 34 insertions(+), 2 deletions(-)

diff --git a/check.py b/check.py
index 79d9dc2..efcfa08 100755
--- a/check.py
+++ b/check.py
@@ -9,6 +9,7 @@ import warnings
 from datetime import date
 from email.mime.text import MIMEText
 from email.utils import formatdate, make_msgid
+from urllib.parse import urlparse, urlunparse
 
 import requests
 import yaml
@@ -105,8 +106,25 @@ def get_title(html: str) -> str:
     return m.group(1).strip() if m and m.group(1) else "no title"
 
 
-def check_conference(name: str, url: str) -> tuple[bool, str]:
+def normalize_url(url: str) -> str:
+    """
+    Normalize the URL by parsing and reconstructing to ensure uniformity.
+
+    This handles cases like differing schemes, casing in the domain
+    and trailing slashes.
+    """
+    # Parse the URL into components
+    parsed_url = urlparse(url)
+    # Normalize the domain to lowercase and remove any trailing dot
+    normalized_netloc = parsed_url.netloc.lower().rstrip(".")
+    # Reconstruct the URL with normalized components
+    return urlunparse(parsed_url._replace(netloc=normalized_netloc))
+
+
+def check_conference(name: str, src_url: str, year: int) -> tuple[bool, str]:
     """Check if conference is live."""
+    url = src_url.format(year=year)
+    past_url = src_url.format(year=year - 1)
     try:
         # SotM Baltics has an invalid TLS certificate, but we don't care
         r = s.get(url, verify=False)
@@ -114,6 +132,19 @@ def check_conference(name: str, url: str) -> tuple[bool, str]:
         return (False, "connection refused")
 
     not_here = find_not_here_message(r.text)
+    if (
+        len(r.text) < 2048
+        and 'http-equiv="refresh"' in r.text
+        and str(year) not in r.text
+    ):
+        return (False, "redirect to URL without year")
+
+    if str(year) not in r.url:
+        return (False, "redirect to URL without year")
+
+    if normalize_url(r.url) == normalize_url(past_url):
+        return (False, "redirect to previous year")
+
     return (False, not_here) if not_here else (True, get_title(r.text))
 
 
@@ -143,7 +174,8 @@ def send_mail(subject: str, body: str) -> None:
 def check_conference_web_site(name: str, src_url: str, year: int) -> bool:
     """Check if an individual web site is live."""
     assert "{year}" in src_url
-    live, msg = check_conference(name, url := src_url.format(year=year))
+    live, msg = check_conference(name, src_url, year)
+    url = src_url.format(year=year)
     if live:
         body = f"{name}\n{url}\nWeb page title: {msg}"
         send_mail(f"Conference site live: {name} - {year}", body)