fix: detect proper xml attributs's namespace

fixes #11

Signed-off-by: Jan Kowalleck <jan.kowalleck@gmail.com>

serializable/__init__.py

@@ -460,10 +460,7 @@ def _from_xml(cls: Type[_T], data: Union[TextIOWrapper, Element],
         _namespaces = dict([node for _, node in
                             SafeElementTree.iterparse(StringIO(SafeElementTree.tostring(data, 'unicode')),
                                                       events=['start-ns'])])
-        if 'ns0' in _namespaces:
-            default_namespace = _namespaces['ns0']
-        else:
-            default_namespace = ''
+        default_namespace = (re.compile(r'^\{(.*?)\}.').search(data.tag) or (None, _namespaces.get('')))[1]
 
     _data: Dict[str, Any] = {}
 
@@ -501,7 +498,9 @@ def _from_xml(cls: Type[_T], data: Union[TextIOWrapper, Element],
 
     # Handle Sub-Elements
     for child_e in data:
-        child_e_tag_name = str(child_e.tag).replace('{' + default_namespace + '}', '')
+        child_e_tag_name = str(child_e.tag)
+        if default_namespace is not None:
+            child_e_tag_name = child_e_tag_name.replace(f'{{{default_namespace}}}', '')
 
         decoded_k = CurrentFormatter.formatter.decode(property_name=child_e_tag_name)
         if decoded_k in klass.ignore_during_deserialization:

tests/base.py

@@ -19,11 +19,12 @@
 
 import json
 import os
-from typing import Any
+from typing import Any, Union
 from unittest import TestCase
 
 import lxml  # type: ignore
 from defusedxml import ElementTree as SafeElementTree  # type: ignore
+from sortedcontainers import SortedSet
 from xmldiff import main  # type: ignore
 from xmldiff.actions import MoveNode  # type: ignore
 
@@ -59,3 +60,28 @@ def assertEqualXml(self, a: str, b: str) -> None:
         diff_results = main.diff_texts(a, b, diff_options={'F': 0.5})
         diff_results = list(filter(lambda o: not isinstance(o, MoveNode), diff_results))
         self.assertEqual(len(diff_results), 0, f'There are XML differences: {diff_results}\n- {a}\n+ {b}')
+
+
+class DeepCompareMixin(object):
+    def assertDeepEqual(self, first: Any, second: Any, msg=None) -> None:
+        """costly compare, but very verbose"""
+        self: Union[TestCase, 'DeepCompareMixin']
+        _omd = self.maxDiff
+        try:
+            self.maxDiff = None
+            dd1 = self.__deepDict(first)
+            dd2 = self.__deepDict(second)
+            self.assertDictEqual(dd1, dd2, msg)
+        finally:
+            self.maxDiff = _omd
+
+    def __deepDict(self, o: Any) -> Any:
+        if isinstance(o, (SortedSet, list, tuple)):
+            return tuple(self.__deepDict(i) for i in o)
+        if isinstance(o, dict):
+            return {k: self.__deepDict(v) for k, v in o}
+        if isinstance(o, set):
+            return tuple(sorted((self.__deepDict(i) for i in o), key=repr))
+        if hasattr(o, '__dict__'):
+            return {k: self.__deepDict(v) for k, v in vars(o).items() if not (k.startswith('__') and k.endswith('__'))}
+        return o

tests/fixtures/the-phoenix-project-defaultNS-v4.xml

@@ -0,0 +1,43 @@
+<?xml version='1.0' encoding='utf-8'?>
+<book xmlns="http://the.phoenix.project/testing/defaultNS" isbn_number="978-1942788294">
+    <id>f3758bf0-0ff7-4366-a5e5-c209d4352b2d</id>
+    <title>The Phoenix Project</title>
+    <edition number="5">5th Anniversary Limited Edition</edition>
+    <publish_date>2018-04-16</publish_date>
+    <author>Gene Kim</author>
+    <author>George Spafford</author>
+    <author>Kevin Behr</author>
+    <type>fiction</type>
+    <publisher>
+        <address>10 Downing Street</address>
+        <name>IT Revolution Press LLC</name>
+    </publisher>
+    <references>
+        <reference ref="my-ref-1"/>
+        <reference ref="my-ref-3">
+            <reference ref="sub-ref-2"/>
+        </reference>
+        <reference ref="my-ref-2">
+            <reference ref="sub-ref-1"/>
+            <reference ref="sub-ref-3"/>
+        </reference>
+    </references>
+    <chapters>
+        <chapter>
+            <number>1</number>
+            <title>Tuesday, September 2</title>
+        </chapter>
+        <chapter>
+            <number>2</number>
+            <title>Tuesday, September 2</title>
+        </chapter>
+        <chapter>
+            <number>3</number>
+            <title>Tuesday, September 2</title>
+        </chapter>
+        <chapter>
+            <number>4</number>
+            <title>Wednesday, September 3</title>
+        </chapter>
+    </chapters>
+</book>

tests/fixtures/the-phoenix-project-defaultNS.xml

@@ -1,2 +1,2 @@
 <?xml version='1.0' encoding='utf-8'?>
-<book xmlns="http://the.phoenix.project/testing/defaultNS" isbn_number="978-1942788294"><id>f3758bf0-0ff7-4366-a5e5-c209d4352b2d</id><title>The Phoenix Project</title><edition number="5">5th Anniversary Limited Edition</edition><publish_date>2018-04-16</publish_date><author>Gene Kim</author><author>George Spafford</author><author>Kevin Behr</author><type>fiction</type><publisher><name>IT Revolution Press LLC</name></publisher><chapters><chapter><number>1</number><title>Tuesday, September 2</title></chapter><chapter><number>2</number><title>Tuesday, September 2</title></chapter><chapter><number>3</number><title>Tuesday, September 2</title></chapter><chapter><number>4</number><title>Wednesday, September 3</title></chapter></chapters></book>
+<book xmlns="http://the.phoenix.project/testing/defaultNS" isbn_number="978-1942788294"><id>f3758bf0-0ff7-4366-a5e5-c209d4352b2d</id><title>The Phoenix Project</title><edition number="5">5th Anniversary Limited Edition</edition><publish_date>2018-04-16</publish_date><author>Karl Ranseier</author><type>fiction</type><publisher><name>IT Revolution Press LLC</name></publisher><chapters><chapter><number>1</number><title>Tuesday, September 2</title></chapter><chapter><number>2</number><title>Tuesday, September 2</title></chapter><chapter><number>3</number><title>Tuesday, September 2</title></chapter><chapter><number>4</number><title>Wednesday, September 3</title></chapter></chapters></book>

tests/model.py

@@ -206,17 +206,17 @@ class Book:
     def __init__(self, title: str, isbn: str, publish_date: date, authors: Iterable[str],
                  publisher: Optional[Publisher] = None, chapters: Optional[Iterable[Chapter]] = None,
                  edition: Optional[BookEdition] = None, type: BookType = BookType.FICTION,
-                 id: Optional[UUID] = None, references: Optional[List[BookReference]] = None) -> None:
+                 id: Optional[UUID] = None, references: Optional[Iterable[BookReference]] = None) -> None:
         self._id = id or uuid4()
         self._title = title
         self._isbn = isbn
         self._edition = edition
         self._publish_date = publish_date
-        self._authors = SortedSet(authors)
+        self._authors = set(authors)
         self._publisher = publisher
         self.chapters = list(chapters or [])
         self._type = type
-        self.references = set(references or {})
+        self.references = set(references or [])
 
     @property  # type: ignore[misc]
     @serializable.xml_sequence(1)

tests/test_xml.py

@@ -19,6 +19,7 @@
 
 import logging
 import os
+from copy import deepcopy
 
 from defusedxml import ElementTree as SafeElementTree
 
@@ -28,15 +29,15 @@
     KebabCasePropertyNameFormatter,
     SnakeCasePropertyNameFormatter,
 )
-from tests.base import FIXTURES_DIRECTORY, BaseTestCase
+from tests.base import FIXTURES_DIRECTORY, BaseTestCase, DeepCompareMixin
 from tests.model import Book, SchemaVersion2, SchemaVersion3, SchemaVersion4, ThePhoenixProject, ThePhoenixProject_v1
 
 logger = logging.getLogger('serializable')
 logger.setLevel(logging.DEBUG)
 formatter = logging.Formatter('%(asctime)s - %(name)s - %(levelname)s - %(message)s')
 
 
-class TestXml(BaseTestCase):
+class TestXml(BaseTestCase, DeepCompareMixin):
 
     # region test_serialize
 
@@ -80,16 +81,16 @@ def test_serializable_with_defaultNS(self) -> None:
         from xml.etree import ElementTree
         xmlns = 'http://the.phoenix.project/testing/defaultNS'
         with open(os.path.join(FIXTURES_DIRECTORY, 'the-phoenix-project-defaultNS.xml')) as expected_xml:
-            self.maxDiff = None
-            self.assertEqual(
-                expected_xml.read(),
-                ElementTree.tostring(
-                    ThePhoenixProject.as_xml(as_string=False, xmlns=xmlns),
-                    method='xml',
-                    encoding='unicode', xml_declaration=True,
-                    default_namespace=xmlns,
-                )
-            )
+            expected = expected_xml.read()
+        data = deepcopy(ThePhoenixProject_v1)
+        data._authors = {'Karl Ranseier', }  # only one item, so order is no issue
+        actual = ElementTree.tostring(
+            data.as_xml(SchemaVersion4, as_string=False, xmlns=xmlns),
+            method='xml',
+            encoding='unicode', xml_declaration=True,
+            default_namespace=xmlns,
+        )
+        self.assertEqual(expected, actual)
 
     # endregion test_serialize
 
@@ -185,4 +186,14 @@ def test_deserialize_tfp_sc1(self) -> None:
             self.assertEqual(ThePhoenixProject_v1.authors, book.authors)
             self.assertEqual(ThePhoenixProject_v1.chapters, book.chapters)
 
+    def test_deserializable_with_defaultNS_from_element(self) -> None:
+        """regression test for https://github.com/madpah/serializable/issues/11"""
+        from xml.etree import ElementTree
+        expected = ThePhoenixProject
+        with open(os.path.join(FIXTURES_DIRECTORY, 'the-phoenix-project-defaultNS-v4.xml')) as fixture_xml:
+            fixture = fixture_xml.read()
+        fixture_element = ElementTree.XML(fixture)
+        actual = Book.from_xml(fixture_element)
+        self.assertDeepEqual(expected, actual)
+
     # region test_deserialize