From e1f7f3683417c6d869b53b4186b681ae2bc0135c Mon Sep 17 00:00:00 2001
From: Tomas Kopecek <tkopecek@redhat.com>
Date: Jul 18 2017 17:30:25 +0000
Subject: kojira monitors external repos changes


Fixes: https://pagure.io/koji/issue/512

---

diff --git a/util/kojira b/util/kojira
index 4a21332..a0fbabc 100755
--- a/util/kojira
+++ b/util/kojira
@@ -35,10 +35,12 @@ import fnmatch
 import logging
 import logging.handlers
 import pprint
+import requests
 import signal
 import time
 import threading
 import traceback
+from xml.etree import ElementTree
 
 
 
@@ -291,8 +293,31 @@ class RepoManager(object):
                     self.logger.info('Dropping entry for inactive repo: %s', repo_id)
                     del self.repos[repo_id]
 
+    def checkExternalRepo(self, ts, repodata, tag):
+        """Determine which external repos are current"""
+        url = repodata['url']
+        if url not in self.external_repos:
+            newest_ts = 0
+            for arch in self.tag_arches.get(tag, self.session.getTag(tag)['arches'].split()):
+                arch_url = url.replace('$arch', arch)
+                arch_url = os.path.join(arch_url, 'repodata/repomd.xml')
+                try:
+                    r = requests.get(arch_url)
+                    root = ElementTree.fromstring(r.text)
+                    for child in root.iter('{http://linux.duke.edu/metadata/repo}timestamp'):
+                        newest_ts = max(newest_ts, int(child.text))
+                except Exception:
+                    # inaccessible or without timestamps
+                    # treat repo as unchanged (ts = 0)
+                    pass
+            self.external_repos[url] = newest_ts
+        return ts < self.external_repos[url]
+
     def checkCurrentRepos(self, session=None):
         """Determine which repos are current"""
+        # clean one-cycle timestamp + arch cache
+        self.external_repos = {}
+        self.tag_arches = {}
         if session is None:
             session = self.session
         to_check = []
@@ -320,6 +345,18 @@ class RepoManager(object):
         #session.multicall = True
         for repo in to_check:
             changed = session.tagChangedSinceEvent(repo.event_id, repo.taglist)
+            if not changed:
+                # look to external repos
+                for tag in repo.taglist:
+                    for external_repo in session.getExternalRepoList(tag):
+                        changed = self.checkExternalRepo(repo.event_ts, external_repo, tag)
+                        self.logger.debug("Check external repo %s [%s] for tag %s: %s" % (
+                            external_repo['external_repo_id'], external_repo['url'],
+                            tag, changed))
+                        if changed:
+                            break
+                    if changed:
+                        break
         #for repo, [changed] in zip(to_check, session.multiCall(strict=True)):
             if changed:
                 self.logger.info("Repo %i no longer current", repo.repo_id)