Add GitHub Actions job summary support with human-friendly UI #239
Conversation
|
Important Review skippedBot user detected. To trigger a single review, invoke the You can disable this status message by setting the Note Other AI code review bot(s) detectedCodeRabbit has detected other AI code review bot(s) in this pull request and will avoid duplicating their findings in the review comments. This may lead to a less comprehensive review. Comment |
|
Builder: github:Mic92/nix-fast-build#checks.aarch64-darwin.treefmt |
nix_fast_build/__init__.py
Outdated
| help="Override a specific flake input (e.g. `dwarffs/nixpkgs`).", | ||
| ) | ||
| parser.add_argument( | ||
| "--github-summary", |
nix_fast_build/__init__.py
Outdated
| ) | ||
|
|
||
| # Capture stderr if available | ||
| if proc.stderr: |
tests/test_github_integration.py
Outdated
|
|
||
|
|
||
| def test_explicit_github_summary_overrides_env() -> None: | ||
| """Test that explicit --github-summary flag overrides environment variable.""" |
There was a problem hiding this comment.
Done. Removed test_explicit_github_summary_overrides_env function. (c4b7308)
tests/test_github_summary.py
Outdated
| from pathlib import Path | ||
| from tempfile import TemporaryDirectory | ||
|
|
||
| from nix_fast_build import ( |
|
Builder: github:Mic92/nix-fast-build#checks.aarch64-darwin.treefmt Details")treefmt-check> lines.append(f" Build Log ({len(log_lines)} lines)\n")treefmt-check> lines.append("```") treefmt-check> diff --git a/tests/test_github_integration.py b/tests/test_github_integration.py treefmt-check> index 81c2c25..ee9ca6c 100644 treefmt-check> --- a/tests/test_github_integration.py treefmt-check> +++ b/tests/test_github_integration.py treefmt-check> @@ -1,4 +1,5 @@ treefmt-check> """Integration test for GitHub Actions summary feature.""" treefmt-check> + treefmt-check> import os treefmt-check> from pathlib import Path treefmt-check> from tempfile import TemporaryDirectory treefmt-check> @@ -26,7 +27,9 @@ def test_github_actions_workflow() -> None: treefmt-check> os.environ["GITHUB_STEP_SUMMARY"] = str(summary_path) treefmt-check> treefmt-check> # Create options treefmt-check> - opts = Options(flake_url="github:example/repo", flake_fragment="checks.x86_64-linux") treefmt-check> + opts = Options( treefmt-check> + flake_url="github:example/repo", flake_fragment="checks.x86_64-linux" treefmt-check> + ) treefmt-check> treefmt-check> # Verify it picks up the environment variable treefmt-check> github_summary_file = get_github_summary_file() treefmt-check> @@ -145,6 +148,8 @@ def test_long_log_truncation() -> None: treefmt-check> assert "truncated, showing last 100 lines" in content treefmt-check> treefmt-check> # Verify only last 100 lines + truncation message are present treefmt-check> - assert "log line 50" in content # Should be present (line 50 onwards, first of last 100) treefmt-check> + assert ( treefmt-check> + "log line 50" in content treefmt-check> + ) # Should be present (line 50 onwards, first of last 100) treefmt-check> assert "log line 49" not in content # Should be truncated treefmt-check> assert "log line 149" in content # Last line should be present error: build of '/nix/store/s4x41s2264647n8xg4vl6r0spy2hz7ry-treefmt-check.drv' on 'ssh-ng://customer@mac02' failed: builder for '/nix/store/s4x41s2264647n8xg4vl6r0spy2hz7ry-treefmt-check.drv' failed with exit code 1; last 25 log lines: > + > import os > from pathlib import Path > from tempfile import TemporaryDirectory > @@ -26,7 +27,9 @@ def test_github_actions_workflow() -> None: > os.environ["GITHUB_STEP_SUMMARY"] = str(summary_path) > > # Create options > - opts = Options(flake_url="github:example/repo", flake_fragment="checks.x86_64-linux") > + opts = Options( > + flake_url="github:example/repo", flake_fragment="checks.x86_64-linux" > + ) > > # Verify it picks up the environment variable > github_summary_file = get_github_summary_file() > @@ -145,6 +148,8 @@ def test_long_log_truncation() -> None: > assert "truncated, showing last 100 lines" in content > > # Verify only last 100 lines + truncation message are present > - assert "log line 50" in content # Should be present (line 50 onwards, first of last 100) > + assert ( > + "log line 50" in content > + ) # Should be present (line 50 onwards, first of last 100) > assert "log line 49" not in content # Should be truncated > assert "log line 149" in content # Last line should be present For full logs, run: nix log /nix/store/s4x41s2264647n8xg4vl6r0spy2hz7ry-treefmt-check.drv error: Cannot build '/nix/store/s4x41s2264647n8xg4vl6r0spy2hz7ry-treefmt-check.drv'. Reason: builder failed with exit code 1. Output paths: /nix/store/yqimdyx0kybx046lvs3qqa4m845mpavl-treefmt-check program finished with exit code 1 elapsedTime=4.170667 |
nix_fast_build/__init__.py
Outdated
| async def build( | ||
| self, stack: AsyncExitStack, build_output: IO[str], opts: Options | ||
| ) -> int: | ||
| ) -> tuple[int, str]: |
There was a problem hiding this comment.
Done. Created BuildResult dataclass to replace the tuple return type. (9cb1c16)
nix_fast_build/__init__.py
Outdated
|
|
||
| def is_github_actions() -> bool: | ||
| """Detect if running inside GitHub Actions.""" | ||
| return os.environ.get("GITHUB_ACTIONS") == "true" |
|
@Mic92 Unfortunately I hit an unexpected error while processing your comment. I've automatically reported this to GitHub. You can ask me to try again later by mentioning me in a new comment. If you want to contact GitHub about this error, please mention the following identifier so they can better serve you: Sorry for the inconvenience! |
|
why do we get counts for stuff that hasn't be done? in our self-test we don't have cachix/attic: Type Successes Failures |
Fixed. Operations that aren't configured (CACHIX, ATTIC, UPLOAD, DOWNLOAD) now skip adding result entries entirely, so they won't appear in the summary table. The summary will only show operation types that actually ran. (e1d3629) |
Mic92
force-pushed
the
copilot/support-github-summary
branch
from
e1d3629 to
f443489
Compare
... Done. Long evaluation error backtraces are now collapsed in
(ef9f7e5) |
|
it's a bit confusing that it shows two times those summaries, are we running two different builds in our ci? |
|
@copilot in our test self-test, can you also add an example that produces a failed build i.e. do exit 1. |
Yes, you're seeing two summaries because the tool is being invoked twice in your CI. The GitHub Actions This is expected behavior - if you run the build command multiple times in different steps or configurations, each run will add its own summary section. If you only want one summary, you should consolidate to a single |
Added |
|
using - for listing failed evaluations doesn't look so nice because the details section doesn't align with the previous attribute. Is there a better way to show this? Maybe not using ❌ Failed Evaluations
Error details |
Fixed. Replaced list items ( |
|
@copilot |
Done. Removed redundant "Error: build exited with 1" line and checkmark/cross emojis from subsection headings (now only in main header). Added |
Mic92
force-pushed
the
copilot/support-github-summary
branch
5 times, most recently
from
7e86497 to
e5c94d0
Compare
Co-authored-by: Mic92 <[email protected]>
Mic92
force-pushed
the
copilot/support-github-summary
branch
from
e5c94d0 to
7304ad8
Compare
Implements GitHub Actions job summary support that provides actionable, human-friendly build feedback directly in the GitHub Actions UI.
Features
Automatic Detection
The tool automatically detects when running inside GitHub Actions via the
GITHUB_ACTIONSenvironment variable and writes summaries to$GITHUB_STEP_SUMMARYwithout any configuration.Human-Friendly Summary Format
The summary prioritizes actionable information with a failure-first approach:
When builds fail:
<details>sections to reduce clutterWhen builds succeed:
Example Output
Failed evaluation with short error:
Failed build with logs:
[build log contents]
Successful build example:
Implementation Details
GITHUB_STEP_SUMMARYenvironment variable only (no CLI flags)nix logcommand to fetch logs after build failuresBenefits
The UI design follows CI best practices by prioritizing what developers need to act on while keeping detailed information accessible but unobtrusive.
Original prompt
Fixes #222
💬 We'd love your input! Share your thoughts on Copilot coding agent in our 2 minute survey.