Code Coverage
 
Lines
Functions and Methods
Classes and Traits
Total
100.00%
1 / 1
100.00%
1 / 1
 CRAP
100.00%
1 / 1
TestResult
100.00%
1 / 1
100.00%
1 / 1
 1
100.00%
1 / 1
 __construct
100.00%
1 / 1
100.00%
1 / 1
 1
1<?php
2
3declare(strict_types=1);
4
5namespace Phpdftk\WptHarness;
6
7/**
8 * Per-test result row written to the harness ledger
9 * (`var/wpt/report.json`). One instance per WPT test per harness run.
10 *
11 * Phase 4A.5 serialises a collection of these into the published
12 * dashboard at `docs/site/standards/spec-coverage/wpt-report.md`.
13 */
14final readonly class TestResult
15{
16    /**
17     * @param string $testId          Stable WPT test identifier — the
18     *                                relative path under
19     *                                `vendor-data/wpt/`, minus the
20     *                                file extension. Example:
21     *                                `css/css-transforms/transform-2d-001`.
22     * @param TestStatus $status      Where the test landed in the
23     *                                {@see TestStatus} bucket
24     *                                hierarchy.
25     * @param float $diffScore        Perceptual diff score in
26     *                                `[0.0, 1.0]` — the share of the
27     *                                frame that differs. `0.0` is
28     *                                byte-identical to the reference;
29     *                                `1.0` is "completely
30     *                                different". Reported for sorting
31     *                                and trending; it is NOT the pass
32     *                                criterion, which is an absolute
33     *                                differing-pixel budget (see
34     *                                {@see Scorer}).
35     * @param string|null $reason     Human-readable explanation when
36     *                                the status is `OutOfScope`,
37     *                                `PendingSubstrate`, or
38     *                                `Skipped`. Null when status is
39     *                                `Pass` or `Fail`.
40     * @param string|null $diffArtefactPath Path under `var/wpt/diffs/`
41     *                                where rendered.png + ref.png +
42     *                                diff.png live. Null when no
43     *                                diff was produced (out-of-scope
44     *                                / skipped tests).
45     * @param float $renderMicros     Wall-clock time the renderer
46     *                                spent on this test, in
47     *                                microseconds. Tracked so the
48     *                                dashboard can surface
49     *                                performance regressions
50     *                                alongside conformance ones.
51     * @param bool $bothRendersSolid  True when the test PASSED and
52     *                                both the test render and the
53     *                                reference render came out a
54     *                                single solid colour — WPT's
55     *                                `check_if_solid_color`
56     *                                criterion. Because this harness
57     *                                renders both sides with its own
58     *                                engine, an unimplemented feature
59     *                                is missing from both and the two
60     *                                blank pages match: a pass with
61     *                                no evidence behind it. Reported
62     *                                as `blankPass`; never a verdict.
63     */
64    public function __construct(
65        public string $testId,
66        public TestStatus $status,
67        public float $diffScore,
68        public ?string $reason,
69        public ?string $diffArtefactPath,
70        public float $renderMicros,
71        public bool $bothRendersSolid = false,
72    ) {}
73}