Source code
Revision control
Copy as Markdown
Other Tools
#!/usr/bin/python3
# This Source Code Form is subject to the terms of the Mozilla Public
# License, v. 2.0. If a copy of the MPL was not distributed with this
"""
Script to parse and print UI test JUnit results from a FullJUnitReport.xml file.
This script processes JUnit test results generated by Flank and Firebase Test Lab.
It reads test results from a specified directory containing a full JUnit report,
parses the results to identify test failures and flaky tests, and prints a formatted
table summarizing these results to the console.
- Parses JUnit XML test result files, including custom 'flaky' attributes.
- Identifies and displays unique test failures and flaky tests.
- Prints the results in a readable table format using BeautifulTable.
- Provides detailed failure messages and test case information.
- Designed for use in Taskcluster following a Firebase Test Lab test execution.
Usage:
python3 parse-junit-results.py --results <path_to_results_directory>
"""
import argparse
import sys
import xml
from pathlib import Path
from beautifultable import BeautifulTable
from junitparser import Attr, Failure, JUnitXml, TestCase, TestSuite
def parse_args(cmdln_args):
parser = argparse.ArgumentParser(
description="Parse and print UI test JUnit results"
)
parser.add_argument(
"--results",
type=Path,
help="Directory containing task artifact results",
required=True,
)
return parser.parse_args(args=cmdln_args)
class test_suite(TestSuite):
flakes = Attr()
class test_case(TestCase):
flaky = Attr()
def parse_print_failure_results(results):
"""
Parses the given JUnit test results and prints a formatted table of failures and flaky tests.
Args:
results (JUnitXml): Parsed JUnit XML results.
Returns:
int: The number of test failures.
The function processes each test suite and each test case within the suite.
If a test case has a result that is an instance of Failure, it is added to the table.
The test case is marked as 'Flaky' if the flaky attribute is set to "true", otherwise it is marked as 'Failure'.
Example of possible JUnit XML (FullJUnitReport.xml):
<testsuites>
<testsuite name="ExampleSuite" tests="2" failures="1" flakes="1" time="0.003">
<testcase classname="example.TestClass" name="testSuccess" flaky="true" time="0.001">
<failure message="Assertion failed">Expected true but was false</failure>
</testcase>
<testcase classname="example.TestClass" name="testFailure" time="0.002">
<failure message="Assertion failed">Expected true but was false</failure>
<failure message="Assertion failed">Expected true but was false</failure>
</testcase>
</testsuite>
</testsuites>
"""
table = BeautifulTable(maxwidth=256)
table.columns.header = ["UI Test", "Outcome", "Details"]
table.columns.alignment = BeautifulTable.ALIGN_LEFT
table.set_style(BeautifulTable.STYLE_GRID)
failure_count = 0
# Dictionary to store the last seen failure details for each test case
last_seen_failures = {}
for suite in results:
cur_suite = test_suite.fromelem(suite)
for case in cur_suite:
cur_case = test_case.fromelem(case)
if cur_case.result:
for entry in case.result:
if isinstance(entry, Failure):
flaky_status = getattr(cur_case, "flaky", "false") == "true"
if flaky_status:
test_id = "%s#%s" % (case.classname, case.name)
details = (
entry.text.replace("\t", " ") if entry.text else ""
)
# Check if the current failure details are different from the last seen ones
if details != last_seen_failures.get(test_id, ""):
table.rows.append(
[
test_id,
"Flaky",
details,
]
)
last_seen_failures[test_id] = details
else:
test_id = "%s#%s" % (case.classname, case.name)
details = (
entry.text.replace("\t", " ") if entry.text else ""
)
# Check if the current failure details are different from the last seen ones
if details != last_seen_failures.get(test_id, ""):
table.rows.append(
[
test_id,
"Failure",
details,
]
)
print(f"TEST-UNEXPECTED-FAIL | {test_id} | {details}")
failure_count += 1
# Update the last seen failure details for this test case
last_seen_failures[test_id] = details
print(table)
return failure_count
def load_results_file(filename):
ret = None
try:
f = open(filename)
try:
ret = JUnitXml.fromfile(f)
except xml.etree.ElementTree.ParseError as e:
print(f"Error parsing {filename} file: {e}")
finally:
f.close()
except OSError as e:
print(e)
return ret
def main():
args = parse_args(sys.argv[1:])
failure_count = 0
junitxml = load_results_file(args.results.joinpath("FullJUnitReport.xml"))
if junitxml:
failure_count = parse_print_failure_results(junitxml)
return failure_count
if __name__ == "__main__":
sys.exit(main())