From 12c0c32f127beab2acb4f04f003a7cdd518524f0 Mon Sep 17 00:00:00 2001
From: Steve McClure <steve.mcclure@mongodb.com>
Date: Wed, 24 Sep 2025 15:56:13 -0400
Subject: [PATCH] SERVER-111373 parse test.xml more robustly with better
 diagnostics (#41734)

GitOrigin-RevId: a3cb25fc5dedbc89e817cc4a431d9a6941465a56
---
 buildscripts/BUILD.bazel                 | 14 +++++++++++++
 buildscripts/create_bazel_test_report.py |  5 +++--
 buildscripts/gather_failed_unittests.py  |  5 +++--
 buildscripts/parse_test_xml.py           | 25 ++++++++++++++++++++++++
 4 files changed, 45 insertions(+), 4 deletions(-)
 create mode 100644 buildscripts/parse_test_xml.py

diff --git a/buildscripts/BUILD.bazel b/buildscripts/BUILD.bazel
index f510714dd73..53ca7953ba8 100644
--- a/buildscripts/BUILD.bazel
+++ b/buildscripts/BUILD.bazel
@@ -217,11 +217,25 @@ sh_binary(
     srcs = ["mount_drives.sh"],
 )
 
+py_library(
+    name = "parse_test_xml",
+    srcs = [
+        "parse_test_xml.py",
+    ],
+    deps = [
+        dependency(
+            "pyyaml",
+            group = "core",
+        ),
+    ],
+)
+
 py_binary(
     name = "gather_failed_unittests",
     srcs = ["gather_failed_unittests.py"],
     visibility = ["//visibility:public"],
     deps = [
+        ":parse_test_xml",
         "//buildscripts/util",
         dependency(
             "typer",
diff --git a/buildscripts/create_bazel_test_report.py b/buildscripts/create_bazel_test_report.py
index 54b078de529..19299e7d2e9 100644
--- a/buildscripts/create_bazel_test_report.py
+++ b/buildscripts/create_bazel_test_report.py
@@ -1,6 +1,5 @@
 import json
 import time
-import xml.etree.ElementTree as ET
 from glob import glob
 from typing import List
 
@@ -8,6 +7,8 @@ import typer
 from typing_extensions import TypedDict
 from util.expansions import get_expansion
 
+from buildscripts.parse_test_xml import parse_test_xml
+
 
 class Result(TypedDict):
     """A single test result"""
@@ -36,7 +37,7 @@ def main(testlog_dir: str):
 
     report = Report({"results": []})
     for test_xml in glob(f"{testlog_dir}/**/test.xml", recursive=True):
-        testsuite = ET.parse(test_xml).getroot().find("testsuite")
+        testsuite = parse_test_xml(test_xml).find("testsuite")
         testcase = testsuite.find("testcase")
 
         # Replace part of the name added by the remote test wrapper script
diff --git a/buildscripts/gather_failed_unittests.py b/buildscripts/gather_failed_unittests.py
index db25c3a4adb..120d93576a3 100644
--- a/buildscripts/gather_failed_unittests.py
+++ b/buildscripts/gather_failed_unittests.py
@@ -1,19 +1,20 @@
 import os
 import shutil
 import subprocess
-import xml.etree.ElementTree as ET
 from glob import glob
 from pathlib import Path
 from typing import List
 
 import typer
 
+from buildscripts.parse_test_xml import parse_test_xml
+
 
 def _collect_test_results(testlog_dir: str) -> List[str]:
     failed_tests = []
     successful_tests = []
     for test_xml in glob(f"{testlog_dir}/**/test.xml", recursive=True):
-        testsuite = ET.parse(test_xml).getroot().find("testsuite")
+        testsuite = parse_test_xml(test_xml).find("testsuite")
         testcase = testsuite.find("testcase")
         test_file = testcase.attrib["name"]
 
diff --git a/buildscripts/parse_test_xml.py b/buildscripts/parse_test_xml.py
new file mode 100644
index 00000000000..1243ee27491
--- /dev/null
+++ b/buildscripts/parse_test_xml.py
@@ -0,0 +1,25 @@
+import os
+import xml.etree.ElementTree as ET
+
+
+def parse_test_xml(xml_file: str) -> ET.ElementTree:
+    """Parse the test.xml file and return the ElementTree object. Throws with diagnostics if unparsable."""
+
+    # Check if file exists and has content
+    if not os.path.exists(xml_file):
+        raise Exception(f"Failed to parse {xml_file}: file does not exist")
+
+    with open(xml_file, "r", encoding="utf-8") as f:
+        content = f.read().strip()
+
+    if not content:
+        raise Exception(f"Failed to parse {xml_file}: file is empty")
+
+    try:
+        root = ET.fromstring(content)
+        return ET.ElementTree(root)
+
+    except (ET.ParseError, UnicodeDecodeError) as e:
+        print(f"Failed to parse {xml_file}: {e}")
+        print(content)
+        raise