Fix failure case number in test result summary incorrect issue (sonic…

…-net#10485)

What is the motivation for this PR?
In test summary, failures += failure + error, but in fact, if a testcase failed with both failure and error, the count of failure and error would both +1, so it's unnecessary to calculate in that way

How did you do it?
Let failures += failure

How did you verify/test it?
Run a test with both failure and error, and check the test result xml file

test_reporting/junit_xml_parser.py

@@ -388,7 +388,7 @@ def _extract_test_summary(test_cases):
             # The result field is unique per test case, either error or failure.
             # xfails is the counter for all kinds of xfail results (include success/failure/error/skipped)
             test_result_summary["tests"] += 1
-            test_result_summary["failures"] += case["result"] == "failure" or case["result"] == "error"
+            test_result_summary["failures"] += case["result"] == "failure"
             test_result_summary["skipped"] += case["result"] == "skipped"
             test_result_summary["errors"] += case["error"]
             test_result_summary["time"] += float(case["time"])