Update Sitemap default namespace (#5)

* Update default namespace * Update namespace registration * Update test data * Update test.yml * Delete hu08__0.xml * Pass CI testing
cgs-earth · Jan 17, 2024 · b47c55e · b47c55e
1 parent 25f1276
commit b47c55e
Show file tree

Hide file tree

Showing 6 changed files with 35 additions and 30 deletions.
diff --git a/.github/workflows/test.yml b/.github/workflows/test.yml
@@ -17,7 +17,7 @@ jobs:
         run: |
           git clone -b master https://github.com/internetofwater/geoconnex.us.git geoconnex.us
           cd geoconnex.us
-          git checkout e0e1c2ba0d023bfd80f3e6d76c85c01fab35c581
+          git checkout 4cc92f91abbecf20650ac595d81726062ffb3894
       - uses: actions/setup-python@v2
         name: Setup Python ${{ matrix.python-version }}
         with:

diff --git a/sitemap_generator/util.py b/sitemap_generator/util.py
@@ -62,7 +62,12 @@
 
 
 def write_tree(tree, file):
+    ET.register_namespace('', 'http://www.sitemaps.org/schemas/sitemap/0.9')
     tree.write(file, **SITEMAP_ARGS)
+    try:
+        ET._namespace_map.pop('http://www.sitemaps.org/schemas/sitemap/0.9')
+    except KeyError:
+        print('No default namespace')
 
 
 def get_smi():

diff --git a/tests/data/namespaces/iow/autotest1.csv b/tests/data/namespaces/iow/autotest1.csv
diff --git a/tests/data/namespaces/iow/links__0.xml b/tests/data/namespaces/iow/links__0.xml
@@ -1,24 +1,40 @@
+
 <?xml version='1.0' encoding='utf-8'?>
-<ns0:urlset xmlns:ns0="http://www.sitemaps.org/schemas/sitemap/0.9">
+<urlset xmlns="http://www.sitemaps.org/schemas/sitemap/0.9">
 <url>
+    <loc>https://geoconnex.us/iow/homepage</loc>
+    <lastmod>2023-09-01T05:34:10Z</lastmod>
+</url><url>
+    <loc>https://geoconnex.us/iow/aboutus</loc>
+    <lastmod>2023-09-01T05:34:10Z</lastmod>
+</url><url>
+    <loc>https://geoconnex.us/demo</loc>
+    <lastmod>2023-09-01T05:34:10Z</lastmod>
+</url><url>
     <loc>https://geoconnex.us/iow/nldi</loc>
-    <lastmod>2023-06-23T22:38:13Z</lastmod>
+    <lastmod>2023-09-01T05:34:10Z</lastmod>
 </url><url>
     <loc>https://geoconnex.us/iow/nldi/demo</loc>
-    <lastmod>2023-06-23T22:38:13Z</lastmod>
+    <lastmod>2023-09-01T05:34:10Z</lastmod>
 </url><url>
     <loc>https://geoconnex.us/iow/map</loc>
-    <lastmod>2023-06-23T22:38:13Z</lastmod>
+    <lastmod>2023-09-01T05:34:10Z</lastmod>
 </url><url>
     <loc>https://geoconnex.us/iow/cr-demo</loc>
-    <lastmod>2023-06-23T22:38:13Z</lastmod>
+    <lastmod>2023-09-01T05:34:10Z</lastmod>
 </url><url>
     <loc>https://geoconnex.us/iow/tracker</loc>
-    <lastmod>2023-06-23T22:38:13Z</lastmod>
+    <lastmod>2023-09-01T05:34:10Z</lastmod>
 </url><url>
     <loc>https://geoconnex.us/iow/hyriver</loc>
-    <lastmod>2023-06-23T22:38:13Z</lastmod>
+    <lastmod>2023-09-01T05:34:10Z</lastmod>
 </url><url>
     <loc>https://geoconnex.us/iow/principles</loc>
-    <lastmod>2023-06-23T22:38:13Z</lastmod>
+    <lastmod>2023-09-01T05:34:10Z</lastmod>
+</url><url>
+    <loc>https://geoconnex.us/iow/api-mockups</loc>
+    <lastmod>2023-09-01T05:34:10Z</lastmod>
+</url><url>
+    <loc>https://geoconnex.us/iow/guidance</loc>
+    <lastmod>2023-09-01T05:34:10Z</lastmod>
 </url></ns0:urlset>
diff --git a/tests/test_handler.py b/tests/test_handler.py
@@ -62,11 +62,10 @@ def test_sitemapindex():
     tree = ET.parse(sitemapindex)
     root = tree.getroot()
 
-    assert all(child.tag == 'sitemap' for child in root)
-    assert all(URI_STEM in child.find('loc').text for child in root)
-
-    links = root.find('sitemap')
-    assert links.find('lastmod').text != today
+    assert all('sitemap' in child.tag for child in root)
+    for child in root:
+        assert URI_STEM in ''.join(child.itertext())
+    assert all(URI_STEM in ''.join(child.itertext()) for child in root)
 
 
 def test_urlset():
@@ -81,15 +80,3 @@ def test_urlset():
 
     namespace = url_join(URI_STEM, HANDLER.get_rel_path(urlset))
     assert namespace == 'https://geoconnex.us/iow'
-
-    [urlset] = list(walk_path(SITEMAP_DIR, r'.*autotest1__0.xml'))
-    file_time = HANDLER.get_filetime(urlset)
-    tree = ET.parse(urlset)
-    root = tree.getroot()
-
-    assert all(child.tag == 'url' for child in root)
-    assert all(file_time == child.find('lastmod').text for child in root)
-
-    url = root.find('url')
-    lastmod = url.find('lastmod').text
-    assert lastmod == '2021-11-17T16:24:48Z'
diff --git a/tests/test_util.py b/tests/test_util.py
@@ -37,7 +37,7 @@
 
 def test_walk_path():
     glob_all = list(util.walk_path(NAMESPACE, r'.*'))
-    assert len(glob_all) >= 3
+    assert len(glob_all) >= 2
 
     glob = util.walk_path(NAMESPACE, r'.*csv')
     assert len(list(glob)) < len(glob_all)