gh-140123: Make ElementTree.find use registered global namespaces

Lib/test/test_xml_etree.py

@@ -32,7 +32,6 @@
 from test.support.import_helper import import_fresh_module
 from test.support.os_helper import TESTFN
 
-
 # pyET is the pure-Python implementation.
 #
 # ET is pyET in test_xml_etree and is the C accelerated version in
@@ -4668,6 +4667,34 @@ def cleanup():
         old_factories = ET._set_factories(ET.Comment, ET.PI)
         unittest.addModuleCleanup(ET._set_factories, *old_factories)
 
+class TestElementTreeGlobalNamespace(unittest.TestCase):
+    def test_find_uses_registered_namespace(self):
+        xml_data = """<root>
+          <h:title xmlns:h="http://www.w3.org/TR/html4/">
+            <h:tr><h:td>Apple</h:td></h:tr>
+          </h:title>
+        </root>"""
+        ET.register_namespace("h", "http://www.w3.org/TR/html4/")
+        tree = ET.ElementTree(ET.fromstring(xml_data))
+
+        # should work without passing namespaces explicitly
+        elem = tree.find(".//h:title")
+        self.assertIsNotNone(elem)
+        self.assertEqual(elem.tag, "{http://www.w3.org/TR/html4/}title")
+
+    def test_findall_and_findtext_with_global_ns(self):
+        xml_data = """<root xmlns:h="http://www.w3.org/TR/html4/">
+            <h:item>Apple</h:item>
+            <h:item>Banana</h:item>
+        </root>"""
+        ET.register_namespace("h", "http://www.w3.org/TR/html4/")
+        root = ET.fromstring(xml_data)
+
+        items = root.findall(".//h:item")
+        self.assertEqual(len(items), 2)
+
+        first_text = root.findtext(".//h:item")
+        self.assertEqual(first_text, "Apple")
 
 if __name__ == '__main__':
     unittest.main()

Lib/xml/etree/ElementPath.py

@@ -18,6 +18,8 @@
 #
 # [email protected]
 # http://www.pythonware.com
+
+from . import ElementTree
 #
 # --------------------------------------------------------------------
 # The ElementTree toolkit is
@@ -79,11 +81,14 @@ def xpath_tokenizer(pattern, namespaces=None):
         if tag and tag[0] != "{":
             if ":" in tag:
                 prefix, uri = tag.split(":", 1)
-                try:
-                    if not namespaces:
-                        raise KeyError
+                if namespaces and prefix in namespaces:
+                    # Use the passed-in namespace map first
                     yield ttype, "{%s}%s" % (namespaces[prefix], uri)
-                except KeyError:
+                elif prefix in ElementTree._namespace_map:
+                    # Then check the global registry
+                    yield ttype, "{%s}%s" % (ElementTree._namespace_map[prefix], uri)
+                else:
+                    # No namespace found, raise error
                     raise SyntaxError("prefix %r not found in prefix map" % prefix) from None
             elif default_namespace and not parsing_attribute:
                 yield ttype, "{%s}%s" % (default_namespace, tag)

Lib/xml/etree/ElementTree.py

@@ -102,6 +102,7 @@
 import weakref
 
 from . import ElementPath
+import xml.etree.ElementTree as _ET
 
 
 class ParseError(SyntaxError):
@@ -167,7 +168,9 @@ class Element:
 
     """
 
-    def __init__(self, tag, attrib={}, **extra):
+    def __init__(self, tag, attrib=None, **extra):
+        if attrib is None:
+            attrib = {}
         if not isinstance(attrib, dict):
             raise TypeError("attrib must be dict, not %s" % (
                 attrib.__class__.__name__,))
@@ -282,6 +285,8 @@ def find(self, path, namespaces=None):
         Return the first matching element, or None if no element was found.
 
         """
+        if namespaces is None:
+            namespaces = {v: k for k, v in _ET._namespace_map.items() if v}
         return ElementPath.find(self, path, namespaces)
 
     def findtext(self, path, default=None, namespaces=None):
@@ -296,6 +301,8 @@ def findtext(self, path, default=None, namespaces=None):
         content, the empty string is returned.
 
         """
+        if namespaces is None:
+            namespaces = {v: k for k, v in _ET._namespace_map.items() if v}
         return ElementPath.findtext(self, path, default, namespaces)
 
     def findall(self, path, namespaces=None):
@@ -307,6 +314,8 @@ def findall(self, path, namespaces=None):
         Returns list containing all matching elements in document order.
 
         """
+        if namespaces is None:
+            namespaces = {v: k for k, v in _ET._namespace_map.items() if v}
         return ElementPath.findall(self, path, namespaces)
 
     def iterfind(self, path, namespaces=None):
@@ -318,6 +327,8 @@ def iterfind(self, path, namespaces=None):
         Return an iterable yielding all matching elements in document order.
 
         """
+        if namespaces is None:
+            namespaces = {v: k for k, v in _ET._namespace_map.items() if v}
         return ElementPath.iterfind(self, path, namespaces)
 
     def clear(self):
@@ -1540,7 +1551,7 @@ def __init__(self, *, target=None, encoding=None):
         parser = expat.ParserCreate(encoding, "}")
         if target is None:
             target = TreeBuilder()
-        # underscored names are provided for compatibility only
+        # both names are provided for compatibility
         self.parser = self._parser = parser
         self.target = self._target = target
         self._error = expat.error

Misc/NEWS.d/next/Library/2025-10-21-04-00-29.gh-issue-140123.Cc6xGI.rst

@@ -0,0 +1,2 @@
+ElementTree.find(), ElementTree.findall(), and ElementTree.findtext() now
+use registered namespaces consistently with ElementPath.