Output test results as a GitHub comment

After filtering intermittents, output the results as JSON. Update the
GitHub workflow to aggregate this JSON data into an artifact and use the
aggregated data to generate a GitHub comment with details about the try
run. The idea here is that this comment will make it easier to track
intermittent tests and notice when a change affects a test marked as
intermittent -- either causing it to permanently fail or fixing it.
This commit is contained in:
Martin Robinson 2023-01-31 19:17:18 +01:00
parent 3429e8fe3b
commit d2a66fef0c
3 changed files with 192 additions and 21 deletions

View file

@ -146,7 +146,6 @@ jobs:
# --release --processes $(sysctl -n hw.logicalcpu) --timeout-multiplier 8 \
# --total-chunks ${{ env.max_chunk_id }} --this-chunk ${{ matrix.chunk_id }} \
# --log-raw test-wpt.${{ matrix.chunk_id }}.log \
# --log-servojson wpt-jsonsummary.${{ matrix.chunk_id }}.log \
# --filter-intermittents=filtered-wpt-summary.${{ matrix.chunk_id }}.log
# - name: Archive logs
# uses: actions/upload-artifact@v3
@ -155,7 +154,6 @@ jobs:
# name: wpt${{ matrix.chunk_id }}-logs-macos
# path: |
# test-wpt.${{ matrix.chunk_id }}.log
# wpt-jsonsummary.${{ matrix.chunk_id }}.log
# filtered-wpt-summary.${{ matrix.chunk_id }}.log
build-linux:
@ -218,8 +216,14 @@ jobs:
--release --processes $(nproc) --timeout-multiplier 2 \
--total-chunks ${{ env.max_chunk_id }} --this-chunk ${{ matrix.chunk_id }} \
--log-raw test-wpt.${{ matrix.chunk_id }}.log \
--log-servojson wpt-jsonsummary.${{ matrix.chunk_id }}.log \
--filter-intermittents=filtered-wpt-summary.${{ matrix.chunk_id }}.log
--filter-intermittents=filtered-wpt-results.${{ matrix.chunk_id }}.json
- name: Archive filtered results
uses: actions/upload-artifact@v3
if: ${{ always() }}
with:
name: wpt-filtered-results-linux
path: |
filtered-wpt-results.${{ matrix.chunk_id }}.json
- name: Archive logs
uses: actions/upload-artifact@v3
if: ${{ failure() }}
@ -227,15 +231,33 @@ jobs:
name: wpt-logs-linux
path: |
test-wpt.${{ matrix.chunk_id }}.log
wpt-jsonsummary.${{ matrix.chunk_id }}.log
filtered-wpt-summary.${{ matrix.chunk_id }}.log
filtered-wpt-results.${{ matrix.chunk_id }}.json
report_test_results:
name: Reporting test results
runs-on: ubuntu-latest
if: always()
needs:
- "linux-wpt"
steps:
- uses: actions/checkout@v3
with:
fetch-depth: 2
- uses: actions/download-artifact@v3
with:
name: wpt-filtered-results-linux
path: wpt-filtered-results-linux
- name: Comment on PR with results
run: etc/ci/report_aggregated_expected_results.py wpt-filtered-results-linux/*
env:
GITHUB_CONTEXT: ${{ toJson(github) }}
GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
build_result:
name: homu build finished
runs-on: ubuntu-latest
needs:
- "build-win"
- "build-linux"
- "build-mac"
- "linux-wpt"
# - "mac-wpt"

View file

@ -0,0 +1,144 @@
#!/usr/bin/env python
# Copyright 2023 The Servo Project Developers. See the COPYRIGHT
# file at the top-level directory of this distribution.
#
# Licensed under the Apache License, Version 2.0 <LICENSE-APACHE or
# http://www.apache.org/licenses/LICENSE-2.0> or the MIT license
# <LICENSE-MIT or http://opensource.org/licenses/MIT>, at your
# option. This file may not be copied, modified, or distributed
# except according to those terms.
# This allows using types that are defined later in the file.
from __future__ import annotations
import json
import os
import re
import subprocess
import sys
import textwrap
import xml.etree.ElementTree as ElementTree
from typing import Optional
SUBTEST_RESULT_TRUNCATION = 10
class Item:
def __init__(self, title: str, body: str, children: list[Item]):
self.title = title
self.body = body
self.children = children
@classmethod
def from_result(cls, result: dict, title_key: str = "path", title_prefix: str = "", print_stack=True):
expected = result["expected"]
actual = result["actual"]
title = result[title_key]
if expected != actual:
title = f"{actual} [expected {expected}] {title_prefix}{title}"
else:
title = f"{actual} {title_prefix}{title}"
stack = result["stack"] if result["stack"] and print_stack else ""
body = f"{result['message']}\n{stack}".strip()
subtest_results = result.get("unexpected_subtest_results", [])
children = [
cls.from_result(subtest_result, "subtest", "subtest: ", False)
for subtest_result in subtest_results
]
return cls(title, body, children)
def to_string(self, bullet: str = "", indent: str = ""):
output = f"{indent}{bullet}{self.title}\n"
if self.body:
output += textwrap.indent(f"{self.body}\n", " " * len(indent + bullet))
output += "\n".join([child.to_string("", indent + " ")
for child in self.children])
return output.rstrip()
def to_html(self, level: int = 0) -> ElementTree.Element:
if level == 0:
title = result = ElementTree.Element("div")
elif level == 1:
result = ElementTree.Element("details")
title = ElementTree.SubElement(result, "summary")
else:
result = ElementTree.Element("li")
title = ElementTree.SubElement(result, "span")
title.text = self.title
if self.children:
# Some tests have dozens of failing tests, which overwhelm the
# output. Limit the output for subtests in GitHub comment output.
max_children = len(self.children) if level < 2 else SUBTEST_RESULT_TRUNCATION
if len(self.children) > max_children:
children = self.children[:max_children]
children.append(Item(
f"And {len(self.children) - max_children} more unexpected results...",
"", []))
else:
children = self.children
container = ElementTree.SubElement(result, "div" if not level else "ul")
for child in children:
container.append(child.to_html(level + 1))
return result
def get_results() -> Optional[Item]:
unexpected = []
known_intermittents = []
for filename in sys.argv[1:]:
with open(filename, encoding="utf-8") as file:
data = json.load(file)
unexpected += data["unexpected"]
known_intermittents += data["known_intermittents"]
children = []
if unexpected:
children.append(
Item(f"Tests producing unexpected results ({len(unexpected)})", "",
[Item.from_result(result) for result in unexpected]),
)
if known_intermittents:
children.append(
Item("Unexpected results that are known to be intermittent "
f"({len(known_intermittents)})", "",
[Item.from_result(result) for result in known_intermittents])
)
return Item("Results from try job:", "", children) if children else None
def get_pr_number() -> Optional[str]:
github_context = json.loads(os.environ.get("GITHUB_CONTEXT", "{}"))
if "event" not in github_context:
return None
if "head_commit" not in github_context["event"]:
return None
commit_title = github_context["event"]["head_commit"]["message"]
match = re.match(r"^Auto merge of #(\d+)", commit_title)
return match.group(1) if match else None
def main():
results = get_results()
if not results:
print("Did not find any unexpected results.")
return
print(results.to_string())
pr_number = get_pr_number()
if pr_number:
html_string = ElementTree.tostring(results.to_html(), encoding="unicode")
process = subprocess.Popen(['gh', 'pr', 'comment', pr_number, '-F', '-'], stdin=subprocess.PIPE)
process.communicate(input=html_string.encode("utf-8"))[0]
else:
print("Could not find PR number in environment. Not making GitHub comment.")
if __name__ == "__main__":
main()

View file

@ -2,6 +2,7 @@
# License, v. 2.0. If a copy of the MPL was not distributed with this
# file, You can obtain one at https://mozilla.org/MPL/2.0/.
import dataclasses
import grouping_formatter
import json
import os
@ -218,30 +219,34 @@ def filter_intermittents(
else:
filter = TrackerFilter()
intermittents = []
actually_unexpected = []
known_intermittents: List[UnexpectedResult] = []
unexpected: List[UnexpectedResult] = []
for i, result in enumerate(unexpected_results):
print(f" [{i}/{len(unexpected_results)}]", file=sys.stderr, end="\r")
if filter.is_failure_intermittent(result.path):
intermittents.append(result)
known_intermittents.append(result)
else:
actually_unexpected.append(result)
output = "\n".join([
f"{len(intermittents)} known-intermittent unexpected result",
*[str(result) for result in intermittents],
"",
f"{len(actually_unexpected)} unexpected results that are NOT known-intermittents",
*[str(result) for result in actually_unexpected],
])
unexpected.append(result)
if output_file:
with open(output_file, "w", encoding="utf-8") as file:
file.write(output)
file.write(json.dumps({
"known_intermittents":
[dataclasses.asdict(result) for result in known_intermittents],
"unexpected":
[dataclasses.asdict(result) for result in unexpected],
}))
output = "\n".join([
f"{len(known_intermittents)} known-intermittent unexpected result",
*[str(result) for result in known_intermittents],
"",
f"{len(unexpected)} unexpected results that are NOT known-intermittents",
*[str(result) for result in unexpected],
])
print(output)
print(80 * "=")
return not actually_unexpected
return not unexpected
def main():