Update sitemap

* Update releases: Newton is unmainted; train, ussuri are relased
* We have redirects to some sites, ignore them
* Remove python2 code
* Use https as starting point

Switch repo also to victoria PTI jobs.

Change-Id: I744e725b32a797b0e208ac7e81afe4b8d3578751

.zuul.yaml

@@ -2,6 +2,6 @@
     templates:
       - check-requirements
       - release-notes-jobs-python3
-      - openstack-python3-ussuri-jobs
+      - openstack-python3-victoria-jobs
       - publish-openstack-docs-pti
       - openstack-lower-constraints-jobs

sitemap/generator/spiders/sitemap_file.py

@@ -12,10 +12,7 @@
 
 import re
 import time
-try:
-    import urlparse
-except ImportError:
-    import urllib.parse as urlparse
+import urllib.parse as urlparse
 
 from scrapy import item
 from scrapy import linkextractors
@@ -34,12 +31,13 @@ class SitemapSpider(spiders.CrawlSpider):
     name = 'sitemap'
 
     MAINT_SERIES = [
-        'newton',
         'ocata',
         'pike',
         'queens',
         'rocky',
         'stein',
+        'train',
+        'ussuri',
     ]
     MAINT_RELEASES_PAT = re.compile('^.*/(' + '|'.join(MAINT_SERIES) + ')/')
     LATEST_PAT = re.compile('^.*/latest/')
@@ -69,7 +67,16 @@ class SitemapSpider(spiders.CrawlSpider):
                     r'/juno/',
                     r'/kilo/',
                     r'/liberty/',
-                    r'/mitaka/'
+                    r'/mitaka/',
+                    r'/newton/',
+                ],
+                deny_domains=[
+                    # docs.o.o redirects to a few sites, filter
+                    # them out
+                    'docs.opendev.org',
+                    'opendev.org',
+                    'releases.openstack.org',
+                    'zuul-ci.org',
                 ]
             ),
             follow=True, callback='parse_item'
@@ -80,7 +87,7 @@ class SitemapSpider(spiders.CrawlSpider):
         super(SitemapSpider, self).__init__(*args, **kwargs)
         self.domain = domain
         self.allowed_domains = [domain]
-        self.start_urls = ['http://%s' % domain]
+        self.start_urls = ['https://%s' % domain]
         for url in urls.split(','):
             if not url:
                 continue
@@ -90,13 +97,17 @@ class SitemapSpider(spiders.CrawlSpider):
         item = SitemapItem()
         item['loc'] = response.url
 
-        path = urlparse.urlsplit(response.url).path
+        components = urlparse.urlsplit(response.url)
+
+        # Filter out any redirected URLs to other domains
+        if self.domain != components.netloc:
+            return
 
-        if self.MAINT_RELEASES_PAT.match(path):
+        if self.MAINT_RELEASES_PAT.match(components.path):
             # weekly changefrequency and highest prio for maintained release
             item['priority'] = '1.0'
             item['changefreq'] = 'weekly'
-        elif self.LATEST_PAT.match(path):
+        elif self.LATEST_PAT.match(components.path):
             # daily changefrequency and normal priority for current files
             item['priority'] = '0.5'
             item['changefreq'] = 'daily'

test/test_sitemap_file.py

@@ -44,7 +44,7 @@ class TestSitemapSpider(unittest.TestCase):
         domain = 'docs.openstack.org'
         self.assertEqual(self.spider.domain, domain)
         self.assertEqual(self.spider.allowed_domains, [domain])
-        self.assertEqual(self.spider.start_urls, ['http://%s' % domain])
+        self.assertEqual(self.spider.start_urls, ['https://%s' % domain])
 
     def test_start_urls_get_appended(self):
         urls = 'new.openstack.org, old.openstack.org'