cim2pp converter: Using lxml to parse XML files (better performance) (e…

…2nIEE#2400)

* - [FIXED] bug in :code:`cim2pp`: Changed zero prioritized generators with voltage controller to sgens (like PowerFactory does)
- [ADDED] cim2pp: added description fields for each asset and added BusbarSection information to nodes

* changed logging TopologicalNodes warning

* changed logging TopologicalNodes warning

* cim converter: Avoid huge logging output when ignore_erros = True

* fixed some small bugs in the CGMES converter and improved its speed

* cim2pp converter: Using lxml to parse XML files (better performance)

* cim2pp converter: Using lxml to parse XML files (better performance)

* cim2pp converter: Using lxml to parse XML files (better performance)

CHANGELOG.rst

@@ -80,6 +80,7 @@ Change Log
 - [CHANGED] accelerate _integrate_power_elements_connected_with_switch_buses() in get_equivalent()
 - [CHANGED] accelerate distributed slack power flow calculation by using sparse-aware operations in _subnetworks()
 - [ADDED] Discrete shunt controller for local voltage regulation with shunt steps
+- [ADDED] cim2pp converter: Using lxml to parse XML files (better performance)
 
 [2.14.7] - 2024-06-14
 -------------------------------

pandapower/converter/cim/cim_classes.py

@@ -12,8 +12,7 @@
 from typing import Dict, List
 import pandas as pd
 import numpy as np
-import xml.etree.ElementTree
-import xml.etree.cElementTree as xmlET
+from lxml import etree
 from .other_classes import ReportContainer, Report, LogLevel, ReportCode
 from .cim_tools import get_cim_schema
 
@@ -389,7 +388,7 @@ def _parse_element(self, element, parsed=None):
     def _get_df(self, items):
         return pd.DataFrame([self._parse_element(child) for child in iter(items)])
 
-    def _get_cgmes_profile_from_xml(self, root: xml.etree.ElementTree.Element, ignore_errors: bool = False,
+    def _get_cgmes_profile_from_xml(self, root: etree.Element, ignore_errors: bool = False,
                                     default_profile: str = 'unknown') -> str:
         """
         Get the CGMES profile from the XML file.
@@ -486,17 +485,16 @@ def _parse_source_file(self, file: str, output: dict, encoding: str, profile_nam
             temp_dir.cleanup()
             del temp_dir, temp_dir_path
             return
-        with open(file, mode='r', encoding=encoding, errors='ignore') as f:
-            cim_str = f.read()
-        xml_tree = xmlET.fromstring(cim_str)
+        parser = etree.XMLParser(encoding=encoding, resolve_entities=False)
+        xml_tree = etree.parse(file, parser)
         if profile_name is None:
-            prf = self._get_cgmes_profile_from_xml(xml_tree)
+            prf = self._get_cgmes_profile_from_xml(xml_tree.getroot())
         else:
             prf = profile_name
         self.file_names[prf] = file
-        self._parse_xml_tree(xml_tree, prf, output)
+        self._parse_xml_tree(xml_tree.getroot(), prf, output)
 
-    def _parse_xml_tree(self, xml_tree: xmlET, profile_name: str, output: Dict | None = None):
+    def _parse_xml_tree(self, xml_tree: etree.ElementTree, profile_name: str, output: Dict | None = None):
         output = self.cim if output is None else output
         # get all CIM elements to parse
         element_types = pd.Series([ele.tag for ele in list(xml_tree)])
@@ -507,6 +505,8 @@ def _parse_xml_tree(self, xml_tree: xmlET, profile_name: str, output: Dict | Non
         if prf not in ns_dict.keys():
             ns_dict[prf] = dict()
         for _, element_type in element_types.items():
+            if not isinstance(element_type, str):
+                continue
             element_type_c = re.sub('{.*}', '', element_type)
             prf_content[element_type_c] = self._get_df(xml_tree.findall(element_type))
             # rename the columns (remove the namespaces)

pyproject.toml

@@ -42,6 +42,7 @@ dependencies = [
     "tqdm",
     "deepdiff",
     "geojson",
+    "lxml",
     "typing_extensions~=4.9",
 ]
 keywords = [