testing with requirements.txt file

docs/requirements.txt

@@ -1,2 +1,3 @@
 sphinx==7.1.2
-sphinx-rtd-theme==1.3.0rc1
+sphinx-rtd-theme==1.3.0rc1
+tulit

docs/source/parsers.rst

@@ -3,6 +3,11 @@ Parsers
 
 This package contains modules for parsing various types of legal documents. Below are the details for each module.
 
+.. automodule:: parsers.parser
+    :members:
+    :undoc-members:
+    :show-inheritance:
+
 .. automodule:: parsers.formex
     :members:
     :undoc-members:

tests/parsers/test_parser.py

@@ -1,42 +1,8 @@
 import unittest
-from tulit.parsers.parser import validate_xml
 import os 
 
 file_path = os.path.join(os.path.dirname(__file__), '..\\data\\akn\\eu')
 
 
-class TestXMLValidator(unittest.TestCase):
-    def setUp(self):
-        """Set up test paths."""
-
-        # Update these paths to match your actual file locations
-        self.valid_xml = os.path.join(file_path, "32014L0092.akn")
-        #self.invalid_xml = "path/to/invalid.xml"
-        self.xsd_file = os.path.join(os.path.dirname(__file__), "..\\metadata\\schemas\\akomantoso30.xsd")
-
-   # def test_valid_xml(self):
-   #     """Test validation with valid XML file."""
-   #     is_valid, error = validate_xml(self.valid_xml, self.xsd_file)
-   #     self.assertTrue(is_valid)
-   #     self.assertIsNone(error)
-
-#    def test_invalid_xml(self):
-#        """Test validation with invalid XML file."""
-#        is_valid, error = validate_xml_against_xsd(self.invalid_xml, self.xsd_file)
-#        self.assertFalse(is_valid)
-#        self.assertIsNotNone(error)
-
-#    def test_nonexistent_xml_file(self):
-#        """Test validation with non-existent XML file."""
-#        is_valid, error = validate_xml("data/nonexistent.xml", self.xsd_file)
-#        self.assertFalse(is_valid)
-#        self.assertIsNotNone(error)
-
-#    def test_nonexistent_xsd_file(self):
-#        """Test validation with non-existent XSD schema file."""
-#        is_valid, error = validate_xml(self.valid_xml, "nonexistent.xsd")
-#        self.assertFalse(is_valid)
-#        self.assertIsNotNone(error)
-
 if __name__ == '__main__':
     unittest.main(verbosity=2)

tulit/parsers/html.py

@@ -187,7 +187,7 @@ def get_lists(self, parent_id: str, container):
 
     def get_articles(self):
         """
-        Extracts articles from the HTML. Each <div> with an id starting with 'art_' is treated as an article (eId).
+        Extracts articles from the HTML. Each <div> with an id starting with "art" is treated as an article (eId).
         Subsequent subdivisions are processed based on the closest parent with an id.
 
         Returns:

tulit/parsers/parser.py

@@ -65,61 +65,4 @@ def remove_node(self, tree, node):
                         # If no siblings, add the tail text to the parent's text
                         parent.text = (parent.text or '') + tail_text
 
-        return tree
-
-def validate_xml(xml_path: str, xsd_path: str): 
-    """
-    Validate an XML file against an XSD schema.
-    
-    Args:
-        xml_path (str): Path to the XML file to validate
-        xsd_path (str): Path to the XSD schema file
-        
-    Returns:
-        Tuple[bool, Union[str, None]]: A tuple containing:
-            - bool: True if validation successful, False otherwise
-            - Union[str, None]: Error message if validation failed, None if successful
-    """
-    try:
-         # Create XML parser with schema location resolution
-        parser = etree.XMLParser(remove_blank_text=True)
-
-        # Create a custom resolver to handle relative paths
-        class LocalResolver(etree.Resolver):
-            def resolve(self, url, id, context):
-                # Get the directory of the main XSD file
-                schema_dir = os.path.dirname(os.path.abspath(xsd_path))
-                # Construct full path to the imported schema
-                schema_path = os.path.join(schema_dir, os.path.basename(url))
-
-                if os.path.exists(schema_path):
-                    return self.resolve_filename(schema_path, context)
-                return None
-
-        # Add the resolver to the parser
-        parser.resolvers.add(LocalResolver())
-
-        # Parse and validate
-        xmlschema_doc = etree.parse(xsd_path, parser)
-        xmlschema = etree.XMLSchema(xmlschema_doc)
-        xml_doc = etree.parse(xml_path, parser)
-
-        xmlschema.assertValid(xml_doc)
-        return True
-
-    except etree.XMLSyntaxError as e:
-        error_msg = f"XML Syntax Error: {str(e)}"
-        logging.error(error_msg)
-        return False, error_msg
-
-    except etree.DocumentInvalid as e:
-        error_msg = f"Schema Validation Error: {str(e)}"
-        logging.error(error_msg)
-        return False, error_msg
-
-    except Exception as e:
-        error_msg = f"Unexpected Error: {str(e)}"
-        logging.error(error_msg)
-        return False, error_msg
-
-
+        return tree