fix: Preprocess xml content before sending it to ElementTree parser

Author: mkundu1

src/ansys/fluent/core/filereader/case_file.py

@@ -45,12 +45,13 @@
 from pathlib import Path
 from typing import Dict, List
 
-from defusedxml.ElementTree import parse
+import defusedxml.ElementTree as ET
 import numpy as np
 
 from ansys.fluent.core.solver.error_message import allowed_name_error_message
 
 from . import lispy
+from .pre_processor import remove_unsupported_xml_chars
 
 try:
     import h5py
@@ -731,15 +732,17 @@ def _get_processed_string(input_string: bytes) -> str:
 
 
 def _get_case_file_name_from_flprj(flprj_file):
-    tree = parse(flprj_file)
-    root = tree.getroot()
-    folder_name = root.find("Metadata").find("CurrentSimulation").get("value")[5:-1]
-    # If the project file name begins with a digit then the node to find will be prepended
-    # with "_". Rather than making any assumptions that this is a hard rule, or what
-    # the scope of the rule is, simply retry with the name prepended:
-    folder_obj = (
-        root.find(folder_name)
-        if root.find(folder_name) and len(root.find(folder_name)) > 0
-        else root.find("_" + folder_name)
-    )
-    return folder_obj.find("Input").find("Case").find("Target").get("value")
+    with open(flprj_file, "r") as file:
+        content = file.read()
+        content = remove_unsupported_xml_chars(content)
+        root = ET.fromstring(content)
+        folder_name = root.find("Metadata").find("CurrentSimulation").get("value")[5:-1]
+        # If the project file name begins with a digit then the node to find will be prepended
+        # with "_". Rather than making any assumptions that this is a hard rule, or what
+        # the scope of the rule is, simply retry with the name prepended:
+        folder_obj = (
+            root.find(folder_name)
+            if root.find(folder_name) and len(root.find(folder_name)) > 0
+            else root.find("_" + folder_name)
+        )
+        return folder_obj.find("Input").find("Case").find("Target").get("value")

src/ansys/fluent/core/filereader/data_file.py

@@ -39,10 +39,11 @@
 from os.path import dirname
 from pathlib import Path
 
-from defusedxml.ElementTree import parse
+import defusedxml.ElementTree as ET
 import numpy as np
 
 from . import lispy
+from .pre_processor import remove_unsupported_xml_chars
 
 try:
     import h5py
@@ -224,7 +225,15 @@ def get_face_vector_field_data(self, phase_name: str, surface_id: int) -> np.arr
 
 
 def _get_data_file_name_from_flprj(flprj_file):
-    tree = parse(flprj_file)
-    root = tree.getroot()
-    folder_name = root.find("Metadata").find("CurrentSimulation").get("value")[5:-1]
-    return root.find(folder_name).find("Input").find("Case").find("Target").get("value")
+    with open(flprj_file, "r") as file:
+        content = file.read()
+        content = remove_unsupported_xml_chars(content)
+        root = ET.fromstring(content)
+        folder_name = root.find("Metadata").find("CurrentSimulation").get("value")[5:-1]
+        return (
+            root.find(folder_name)
+            .find("Input")
+            .find("Case")
+            .find("Target")
+            .get("value")
+        )

src/ansys/fluent/core/filereader/pre_processor.py

@@ -0,0 +1,13 @@
+"""
+Pre-processor for XML content from Fluent project files.
+"""
+
+import re
+
+
+def remove_unsupported_xml_chars(content: str):
+    """Remove unsupported XML characters from the content."""
+    # Replace double colons in tag names with double underscores
+    content = re.sub(r"</?([^> ]+)::", r"<\1__", content)
+
+    return content

tests/test_casereader.py

@@ -24,6 +24,7 @@
 import pathlib
 import shutil
 
+import defusedxml.ElementTree as ET
 import pytest
 
 from ansys.fluent.core import examples
@@ -35,6 +36,7 @@
     _get_processed_string,
 )
 from ansys.fluent.core.filereader.case_file import CaseFile as CaseReader
+from ansys.fluent.core.filereader.pre_processor import remove_unsupported_xml_chars
 
 
 def call_casereader(
@@ -344,3 +346,25 @@ def test_mesh_reader():
     assert mesh_reader_2d.precision() is None
     assert mesh_reader_3d.precision() is None
     assert case_reader.precision() == 2
+
+
+def test_preprocessor():
+    content = """
+    <Project type="object" class="PFolder">
+    <Metadata type="object" class="ansys::Project::MetadataHoof">
+        <ProjectStoragePolicy::FolderEnabled class="string" value="true"/>
+    </Metadata>
+    </Project>
+    """
+    expected = """
+    <Project type="object" class="PFolder">
+    <Metadata type="object" class="ansys::Project::MetadataHoof">
+        <ProjectStoragePolicy__FolderEnabled class="string" value="true"/>
+    </Metadata>
+    </Project>
+    """
+    with pytest.raises(ET.ParseError):
+        ET.fromstring(content)
+    pre_processed = remove_unsupported_xml_chars(content)
+    assert pre_processed == expected
+    assert ET.fromstring(pre_processed)