Add a split packages sitemap

With very low priority, but this should at least give a few more
cross-linking pages to any crawlers using sitemaps.

Signed-off-by: Dan McGee <dan@archlinux.org>

3 files changed, 33 insertions, 3 deletions

diff --git a/packages/utils.py b/packages/utils.py
index 0df0e38..f8e1f2a 100644
--- a/packages/utils.py
+++ b/packages/utils.py
@@ -2,10 +2,10 @@ from collections import defaultdict
 from operator import itemgetter
 
 from django.db import connection
-from django.db.models import Count, Max
+from django.db.models import Count, Max, F
 from django.contrib.auth.models import User
 
-from main.models import Package, Repo
+from main.models import Package, Arch, Repo
 from main.utils import cache_function, groupby_preserve_order, PackageStandin
 from .models import (PackageGroup, PackageRelation,
         SignoffSpecification, Signoff, DEFAULT_SIGNOFF_SPEC)
@@ -49,6 +49,20 @@ def get_group_info(include_arches=None):
             groups.extend(val.itervalues())
     return sorted(groups, key=itemgetter('name', 'arch'))
 
+def get_split_packages_info():
+    '''Return info on split packages that do not have an actual package name
+    matching the split pkgbase.'''
+    pkgnames = Package.objects.values('pkgname')
+    split_pkgs = Package.objects.exclude(pkgname=F('pkgbase')).exclude(
+            pkgbase__in=pkgnames).values('pkgbase', 'repo', 'arch').annotate(
+            last_update=Max('last_update'))
+    all_arches = Arch.objects.in_bulk(set(s['arch'] for s in split_pkgs))
+    all_repos = Repo.objects.in_bulk(set(s['repo'] for s in split_pkgs))
+    for split in split_pkgs:
+        split['arch'] = all_arches[split['arch']]
+        split['repo'] = all_repos[split['repo']]
+    return split_pkgs
+
 class Difference(object):
     def __init__(self, pkgname, repo, pkg_a, pkg_b):
         self.pkgname = pkgname
diff --git a/sitemaps.py b/sitemaps.py
index 8ac5bc4..7718002 100644
--- a/sitemaps.py
+++ b/sitemaps.py
@@ -3,7 +3,7 @@ from django.core.urlresolvers import reverse
 
 from main.models import Package
 from news.models import News
-from packages.utils import get_group_info
+from packages.utils import get_group_info, get_split_packages_info
 
 class PackagesSitemap(Sitemap):
     changefreq = "weekly"
@@ -41,6 +41,21 @@ class PackageGroupsSitemap(Sitemap):
         return '/groups/%s/%s/' % (obj['arch'], obj['name'])
 
 
+class SplitPackagesSitemap(Sitemap):
+    changefreq = "weekly"
+    priority = "0.3"
+
+    def items(self):
+        return get_split_packages_info()
+
+    def lastmod(self, obj):
+        return obj['last_update']
+
+    def location(self, obj):
+        return '/packages/%s/%s/%s/' % (
+                obj['repo'].name.lower(), obj['arch'], obj['pkgbase'])
+
+
 class NewsSitemap(Sitemap):
     changefreq = "never"
     priority = "0.8"
diff --git a/urls.py b/urls.py
index adbc887..1d06f0f 100644
--- a/urls.py
+++ b/urls.py
@@ -18,6 +18,7 @@ our_sitemaps = {
     'packages':       sitemaps.PackagesSitemap,
     'package-files':  sitemaps.PackageFilesSitemap,
     'package-groups': sitemaps.PackageGroupsSitemap,
+    'split-packages': sitemaps.SplitPackagesSitemap,
 }
 
 admin.autodiscover()