✨ ALTO: Count alto:Tags

qurator/modstool/alto4pandas.py

@@ -86,7 +86,8 @@ def alto_to_dict(alto, raise_errors=True):
         elif localname == 'Styles':
             pass
         elif localname == 'Tags':
-            pass
+            value[localname] = {}
+            value[localname].update(TagGroup(tag, group).subelement_counts())
         else:
             if raise_errors:
                 print(value)

qurator/modstool/tests/test_alto.py

@@ -37,3 +37,21 @@ def test_Page_counts():
     assert d['Layout_Page_TextBlock-count'] == 1
     assert d['Layout_Page_TextLine-count'] == 3
     assert d['Layout_Page_String-count'] == 6
+
+def test_Tags_counts():
+    d = dict_fromstring("""
+    <alto xmlns="http://www.loc.gov/standards/alto/ns-v2#">
+      <Tags>
+        <NamedEntityTag ID="PER0" LABEL="Pentlings"/>
+        <NamedEntityTag ID="LOC1" LABEL="Pentling"/>
+        <NamedEntityTag ID="LOC2" LABEL="Hamm"/>
+        <NamedEntityTag ID="PER4" LABEL="Hofes Pentling"/>
+        <NamedEntityTag ID="LOC5" LABEL="Hofs Pentling"/>
+        <NamedEntityTag ID="LOC7" LABEL="Hilbeck"/>
+        <NamedEntityTag ID="PER8" LABEL="Hoff"/>
+        <NamedEntityTag ID="PER9" LABEL="L i b e r"/>
+        <NamedEntityTag ID="PER10" LABEL="Jhesu Christi"/>
+      </Tags>
+    </alto>
+    """)
+    assert d['Tags_NamedEntityTag-count'] == 9