Merge branch 'development' into feature/docs_and_annotation

# Conflicts:
#	ebmlite/core.py
#	ebmlite/util.py

.github/workflows/unit-tests.yml

@@ -25,7 +25,7 @@ jobs:
 
     steps:
 
-      - uses: actions/setup-python@v3
+      - uses: actions/setup-python@v4
         with:
           python-version: ${{ matrix.python-version }}
       - run: python -m pip install --upgrade pip

ebmlite/core.py

@@ -603,14 +603,8 @@ def _isValidChild(cls, elId: int) -> bool:
         if not cls.children:
             return False
 
-        if cls._childIds:
-            return elId in cls._childIds
-        else:
-            # The set of valid child IDs hasn't been created yet.
-            cls._childIds = set(cls.children)
-            if cls.schema is not None:
-                cls._childIds.update(cls.schema.globals)
-            return elId in cls._childIds
+        return elId in cls.children or elId in cls.schema.globals
+
 
     @property
     def size(self) -> int:
@@ -1124,7 +1118,7 @@ def __init__(self,
         self.elementInfo = {}  # Raw element schema attributes, keyed by ID
 
         self.globals = {}   # Elements valid for any parent, by ID
-        self.children = {}  # Valid root elements, by ID
+        self.children = set()  # Valid root elements, by ID
 
         # Parse, using the correct method for the schema format.
         schema = ET.parse(source)
@@ -1215,7 +1209,7 @@ def _parseSchema(self, el, parent=None):
             for chEl in el:
                 self._parseSchema(chEl, cls)
 
-    def addElement(self, 
+    def addElement(self,
                    eid: int, 
                    ename: str, 
                    baseClass, 
@@ -1320,7 +1314,7 @@ def _getInt(d, k, default):
                           {'id': eid, 'name': ename, 'schema': self,
                            'mandatory': mandatory, 'multiple': multiple,
                            'precache': precache, 'length': length,
-                           'children': dict(), '__doc__': docs,
+                           'children': set(), '__doc__': docs,
                            '__slots__': baseClass.__slots__})
 
             self.elements[eid] = eclass
@@ -1332,8 +1326,8 @@ def _getInt(d, k, default):
 
         parent = parent or self
         if parent.children is None:
-            parent.children = {}
-        parent.children[eid] = eclass
+            parent.children = set()
+        parent.children.add(eid)
 
         return eclass
 

ebmlite/util.py

@@ -15,7 +15,7 @@
 __credits__ = "David Randall Stokes, Connor Flanigan, Becker Awqatty, Derek Witt"
 
 __all__ = ['createID', 'validateID', 'toXml', 'xml2ebml', 'loadXml', 'pprint',
-           'printSchemata']
+           'printSchemata', 'flatiter']
 
 import ast
 from io import BytesIO
@@ -505,3 +505,30 @@ def printSchemata(paths: Optional[List[str]] = None,
     finally:
         if newfile:
             out.close()
+
+
+#===============================================================================
+#
+#===============================================================================
+
+
+def flatiter(element, depth=None):
+    """ Recursively crawl an EBML document or element, depth-first,
+        yielding all elements (or elements down to a given depth).
+
+        :param element: The EBML `Document` or `Element` to iterate.
+        :param depth: The maximum recursion depth. `None` or a value less
+            than zero will fully recurse without limit.
+    """
+    depth = -1 if depth is None else depth
+
+    def _flatiter(el, d, first):
+        if not first:
+            yield el
+        if abs(d) > 0 and isinstance(el, core.MasterElement):
+            for ch in el:
+                for grandchild in _flatiter(ch, d-1, False):
+                    yield grandchild
+
+    for child in _flatiter(element, depth, True):
+        yield child

tests/test_general.py

@@ -4,6 +4,7 @@
 @author: dstokes
 """
 
+from itertools import zip_longest
 import os.path
 import unittest
 from xml.dom.minidom import parseString
@@ -66,18 +67,11 @@ def testMkv(self):
             xmlDoc2 = util.loadXml(xmlFile2, schema)
 
             # Compare each element from the XML
-            xmlEls1 = [xmlDoc1]
-            xmlEls2 = [xmlDoc2]
-            while len(xmlEls1) > 0:
-                self.assertEqual(xmlEls1[0], xmlEls2[0], 'Element '
-                                 + repr(xmlEls1[0])
-                                 + ' was not converted properly')
-                for x in list(xmlEls1.pop(0).children.values()):
-                    if issubclass(x, core.Element):
-                        xmlEls1.append(x)
-                for x in list(xmlEls2.pop(0).children.values()):
-                    if issubclass(x, core.Element):
-                        xmlEls2.append(x)
+            for el1, el2 in zip_longest(util.flatiter(xmlDoc1),
+                                        util.flatiter(xmlDoc2),
+                                        fillvalue=None):
+                self.assertEqual(el1, el2,
+                                 'Element {!r} was not converted properly'.format(el1))
 
 
     def testIde(self):
@@ -121,19 +115,11 @@ def testIde(self):
             xmlDoc2 = util.loadXml(xmlFile2, schema)
 
             # Compare each element from the XML
-            xmlEls1 = [xmlDoc1]
-            xmlEls2 = [xmlDoc2]
-            while len(xmlEls1) > 0:
-                self.assertEqual(xmlEls1[0], xmlEls2[0], 'Element '
-                                 + repr(xmlEls1[0])
-                                 + ' was not converted properly')
-                for x in list(xmlEls1.pop(0).children.values()):
-                    if issubclass(x, core.Element):
-                        xmlEls1.append(x)
-                for x in list(xmlEls2.pop(0).children.values()):
-                    if issubclass(x, core.Element):
-                        xmlEls2.append(x)
-
+            for el1, el2 in zip_longest(util.flatiter(xmlDoc1),
+                                        util.flatiter(xmlDoc2),
+                                        fillvalue=None):
+                self.assertEqual(el1, el2,
+                                 'Element {!r} was not converted properly'.format(el1))
 
 
     def testPPrint(self):
@@ -298,18 +284,11 @@ def testMkv(self):
             xmlDoc2 = util.loadXml(xmlFile2, schema)
 
             # Compare each element from the XML
-            xmlEls1 = [xmlDoc1]
-            xmlEls2 = [xmlDoc2]
-            while len(xmlEls1) > 0:
-                self.assertEqual(xmlEls1[0], xmlEls2[0], 'Element '
-                                 + repr(xmlEls1[0])
-                                 + ' was not converted properly')
-                for x in list(xmlEls1.pop(0).children.values()):
-                    if issubclass(x, core.Element):
-                        xmlEls1.append(x)
-                for x in list(xmlEls2.pop(0).children.values()):
-                    if issubclass(x, core.Element):
-                        xmlEls2.append(x)
+            for el1, el2 in zip_longest(util.flatiter(xmlDoc1),
+                                        util.flatiter(xmlDoc2),
+                                        fillvalue=None):
+                self.assertEqual(el1, el2,
+                                 'Element {!r} was not converted properly'.format(el1))
 
 
 if __name__ == "__main__":

tests/test_tools.py

@@ -12,6 +12,13 @@
 
 @pytest.mark.script_launch_mode('subprocess')
 def test_ebml2xml(script_runner):
+
+    # This test can only run if the library has been installed,
+    # e.g., in a GitHub action. Bail if not.
+    # TODO: This is a hack and should be redone.
+    if os.getenv("GITHUB_ACTIONS") != "true":
+        return
+
     path_base = os.path.join(".", "tests", "video-4{ext}")
     path_in = path_base.format(ext=".ebml")
     path_out = path_base.format(ext=".ebml.xml")
@@ -56,6 +63,13 @@ def assert_elements_are_equiv(e1, e2):
 
 @pytest.mark.script_launch_mode('subprocess')
 def test_xml2ebml(script_runner):
+
+    # This test can only run if the library has been installed,
+    # e.g., in a GitHub action. Bail if not.
+    # TODO: This is a hack and should be redone.
+    if os.getenv("GITHUB_ACTIONS") != "true":
+        return
+
     path_base = os.path.join(".", "tests", "video-4{ext}")
     path_in = path_base.format(ext=".xml")
     path_out = path_base.format(ext=".xml.ebml")
@@ -84,6 +98,13 @@ def test_xml2ebml(script_runner):
 
 @pytest.mark.script_launch_mode('subprocess')
 def test_view(script_runner):
+
+    # This test can only run if the library has been installed,
+    # e.g., in a GitHub action. Bail if not.
+    # TODO: This is a hack and should be redone.
+    if os.getenv("GITHUB_ACTIONS") != "true":
+        return
+
     path_base = os.path.join(".", "tests", "video-4{ext}")
     path_in = path_base.format(ext=".ebml")
     path_out = path_base.format(ext=".xml.txt")
@@ -112,6 +133,13 @@ def test_view(script_runner):
 
 @pytest.mark.script_launch_mode('subprocess')
 def test_list_schemata(script_runner):
+
+    # This test can only run if the library has been installed,
+    # e.g., in a GitHub action. Bail if not.
+    # TODO: This is a hack and should be redone.
+    if os.getenv("GITHUB_ACTIONS") != "true":
+        return
+
     core.SCHEMA_PATH = [os.path.dirname(schemata.__file__)]
     path_out = os.path.join(".", "tests", "list-schemata.txt")