Work in progress

Signed-off-by: Hristo Marinov <hricky@fedoraproject.org>

ci/monitor-timestamps.py

@@ -0,0 +1,116 @@
+#! /usr/bin/env python3
+
+import re
+import socket
+import urllib.error
+import urllib.request
+from datetime import datetime, timedelta, timezone
+
+TIMEOUT = 10
+UTC = timezone.utc
+
+STALE_THRESHOLD_MINUTES = 100
+
+URLS = {
+    "production": "https://docs.fedoraproject.org/en-US/docs/",
+    "staging": "https://docs.stg.fedoraproject.org/en-US/docs/",
+}
+
+
+def fetch_page(url):
+    """Fetch the webpage and return its content."""
+    try:
+        request = urllib.request.Request(url)
+        with urllib.request.urlopen(request, timeout=TIMEOUT) as response:
+            return response.read().decode('utf-8')
+    except socket.timeout:
+        raise RuntimeError(f"Timeout fetching {url}")
+    except urllib.error.HTTPError as e:
+        raise RuntimeError(f"HTTP error {e.code} from {url}")
+    except urllib.error.URLError as e:
+        raise RuntimeError(f"URL error from {url}: {e.reason}")
+
+
+def extract_build_timestamp(html):
+    """Extract the build timestamp from the footer."""
+    footer_match = re.search(r"<footer[^>]*>.*?</footer>", html, re.DOTALL)
+    if not footer_match:
+        raise ValueError("Could not find footer in HTML")
+    footer_content = footer_match.group(0)
+
+    timestamp_match = re.search(
+        r"Last build:\s*(\d{4}-\d{2}-\d{2} \d{2}:\d{2}:\d{2})\s+UTC",
+        footer_content
+    )
+    if not timestamp_match:
+        raise ValueError("Could not find build timestamp in footer")
+
+    timestamp_str = timestamp_match.group(1)
+    dt = datetime.strptime(timestamp_str, "%Y-%m-%d %H:%M:%S")
+    return dt.replace(tzinfo=UTC)
+
+
+def check_staleness(url):
+    """Check if the build timestamp is stale (older than STALE_THRESHOLD_MINUTES)."""
+    html = fetch_page(url)
+    build_timestamp = extract_build_timestamp(html)
+    current_time = datetime.now(UTC)
+    time_diff = current_time - build_timestamp
+    is_stale = time_diff.total_seconds() > STALE_THRESHOLD_MINUTES * 60
+
+    return {
+        'is_stale': is_stale,
+        'build_timestamp': build_timestamp,
+        'current_time': current_time,
+        'time_diff': time_diff,
+    }
+
+
+def format_time_diff(total_seconds):
+    """Format time difference in human-readable form with proper pluralization."""
+    days = int(total_seconds // 86400)
+    hours = int((total_seconds % 86400) // 3600)
+    minutes = int((total_seconds % 3600) // 60)
+
+    parts = []
+    if days > 0:
+        parts.append(f"{days} day{'s' if days != 1 else ''}")
+    if hours > 0:
+        parts.append(f"{hours} hour{'s' if hours != 1 else ''}")
+    if minutes > 0 and days == 0:
+        parts.append(f"{minutes} minute{'s' if minutes != 1 else ''}")
+
+    return " and ".join(parts) if parts else "0 minutes"
+
+
+def main():
+    results = {}
+    for instance, url in URLS.items():
+        try:
+            staleness_result = check_staleness(url)
+            results[instance] = {
+                'status': "STALE" if staleness_result['is_stale'] else "RECENT",
+                'build_timestamp': staleness_result['build_timestamp'],
+                'current_time': staleness_result['current_time'],
+                'time_diff': staleness_result['time_diff'],
+            }
+        except (RuntimeError, ValueError) as e:
+            results[instance] = {'status': "ERROR", 'error': str(e)}
+
+    print("Build Status:")
+    for instance, result in results.items():
+        if result['status'] != "ERROR":
+            time_str = format_time_diff(result['time_diff'].total_seconds())
+            build_ts = result['build_timestamp'].replace(microsecond=0)
+            build_ts_str = build_ts.strftime('%Y-%m-%d %H:%M:%S UTC')
+            print(f" {instance.capitalize():11} {result['status']:6} {build_ts_str} ({time_str} ago)")
+        else:
+            print(f" {instance.capitalize():11} {result['status']:6} {result['error']}")
+
+    stale_instances = [instance.capitalize() for instance, result in results.items() if result['status'] != "RECENT"]
+
+    return 1 if stale_instances else 0
+
+
+if __name__ == "__main__":
+    raise SystemExit(main())