misc: use legacy urls to call up to date urls (#64344)

chrono/utils/publik_urls.py

@@ -23,8 +23,18 @@ def translate_from_publik_url(url):
     if not url:
         return ''
     source_url = urllib.parse.urlparse(url)
+
     if source_url.scheme != 'publik':
-        return url
+        # search in legacy urls
+        legacy_urls_mapping = getattr(settings, 'LEGACY_URLS_MAPPING', None)
+        if not legacy_urls_mapping:
+            return url
+        splitted_url = urllib.parse.urlparse(url)
+        hostname = splitted_url.netloc
+        if hostname not in legacy_urls_mapping:
+            return url
+        return splitted_url._replace(netloc=legacy_urls_mapping[hostname]).geturl()
+
     known_services = getattr(settings, 'KNOWN_SERVICES', None)
     if not known_services:
         return url

tests/settings.py

@@ -29,6 +29,8 @@ KNOWN_SERVICES = {
     },
 }
 
+LEGACY_URLS_MAPPING = {'old.org': 'new.org'}
+
 EXCEPTIONS_SOURCES = {}
 
 SITE_BASE_URL = 'https://example.com'

tests/test_requests.py

@@ -145,3 +145,10 @@ def test_requests_cache():
 
         # check with unicode url
         assert requests.get('http://cache.example.org/éléphant').content == b'hello second world'
+
+
+def test_requests_to_legacy_urls():
+    with mock.patch('chrono.utils.requests_wrapper.RequestsSession.request') as requests_get:
+        requests_get.return_value = mock.Mock(content=b'hello world', status_code=200)
+        assert requests.get('http://old.org/').content == b'hello world'
+        assert requests_get.call_args.args[1] == 'http://new.org/'