mirror of
https://github.com/garrytan/gstack.git
synced 2026-08-22 14:07:14 +02:00
Garry TanandClaude Fable 5
820afa26e8
Editing the eval dep-list data no longer forces the full ~$38 / 30-45min suite (measured on 21.9% of recent commits). When touchfiles-data.ts is in the diff, selection now evaluates the BASE version (git show -> mkdtemp -> spawnSync bun child printing the four maps as JSON — sync because e2e-helpers selects at module scope) and JSON-diffs per key: added entries, edited dep lists, and tier flips are selected; keys removed from all maps are reported, never silently dropped; a GLOBAL_TOUCHFILES edit still runs everything. FAIL-CLOSED with named causes: missing-base-ref, git-show-failed, import-failed, shape-mismatch each degrade to run-all and print 'selection: global — touchfiles-data changed (<cause>)' (D9 — silently expensive beats silently wrong, but never silently). eval:select prints 'selected N of M, reason: ...' + removed tests; --base scopes the map-diff too. The temporary conservative GLOBAL entry for touchfiles-data.ts is gone — its changes route through the map-diff. 23 new free tests: pure-core fixtures, selectTests wiring incl. a poison-injection guard, and a temp git repo exercising every fail-closed cause end-to-end. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
94 lines
3.6 KiB
TypeScript
94 lines
3.6 KiB
TypeScript
#!/usr/bin/env bun
|
|
/**
|
|
* Show which E2E and LLM-judge tests would run based on the current git diff.
|
|
*
|
|
* Usage:
|
|
* bun run eval:select # human-readable output
|
|
* bun run eval:select --json # machine-readable JSON
|
|
* bun run eval:select --base main # override base branch
|
|
*/
|
|
|
|
import * as path from 'path';
|
|
import {
|
|
selectTests,
|
|
detectBaseBranch,
|
|
getChangedFiles,
|
|
E2E_TOUCHFILES,
|
|
LLM_JUDGE_TOUCHFILES,
|
|
GLOBAL_TOUCHFILES,
|
|
} from '../test/helpers/touchfiles';
|
|
|
|
const ROOT = path.resolve(import.meta.dir, '..');
|
|
const args = process.argv.slice(2);
|
|
const jsonMode = args.includes('--json');
|
|
const baseIdx = args.indexOf('--base');
|
|
const baseOverride = baseIdx >= 0 ? args[baseIdx + 1] : undefined;
|
|
|
|
// Detect base branch
|
|
const baseBranch = baseOverride || detectBaseBranch(ROOT) || 'main';
|
|
const changedFiles = getChangedFiles(baseBranch, ROOT);
|
|
|
|
if (changedFiles.length === 0) {
|
|
if (jsonMode) {
|
|
console.log(JSON.stringify({ base: baseBranch, changed_files: 0, e2e: 'all', llm_judge: 'all', reason: 'no diff — would run all tests' }));
|
|
} else {
|
|
console.log(`Base: ${baseBranch}`);
|
|
console.log('No changed files detected — all tests would run.');
|
|
}
|
|
process.exit(0);
|
|
}
|
|
|
|
// baseRef/cwd scope the map-diff path (used when touchfiles-data.ts changed)
|
|
// to the same base this script diffed against — including a --base override.
|
|
const selectOpts = { baseRef: baseBranch, cwd: ROOT };
|
|
const e2eSelection = selectTests(changedFiles, E2E_TOUCHFILES, GLOBAL_TOUCHFILES, selectOpts);
|
|
const llmSelection = selectTests(changedFiles, LLM_JUDGE_TOUCHFILES, GLOBAL_TOUCHFILES, selectOpts);
|
|
|
|
if (jsonMode) {
|
|
console.log(JSON.stringify({
|
|
base: baseBranch,
|
|
changed_files: changedFiles,
|
|
e2e: {
|
|
selected: e2eSelection.selected,
|
|
skipped: e2eSelection.skipped,
|
|
reason: e2eSelection.reason,
|
|
removed_tests: e2eSelection.removedTests ?? [],
|
|
count: `${e2eSelection.selected.length}/${Object.keys(E2E_TOUCHFILES).length}`,
|
|
},
|
|
llm_judge: {
|
|
selected: llmSelection.selected,
|
|
skipped: llmSelection.skipped,
|
|
reason: llmSelection.reason,
|
|
count: `${llmSelection.selected.length}/${Object.keys(LLM_JUDGE_TOUCHFILES).length}`,
|
|
},
|
|
}, null, 2));
|
|
} else {
|
|
console.log(`Base: ${baseBranch}`);
|
|
console.log(`Changed files: ${changedFiles.length}`);
|
|
console.log();
|
|
|
|
console.log(`E2E: selected ${e2eSelection.selected.length} of ${Object.keys(E2E_TOUCHFILES).length}, reason: ${e2eSelection.reason}`);
|
|
if (e2eSelection.removedTests && e2eSelection.removedTests.length > 0) {
|
|
console.log(` Removed from maps (reported, not selected): ${e2eSelection.removedTests.join(', ')}`);
|
|
}
|
|
if (e2eSelection.selected.length > 0 && e2eSelection.selected.length < Object.keys(E2E_TOUCHFILES).length) {
|
|
console.log(` Selected: ${e2eSelection.selected.join(', ')}`);
|
|
console.log(` Skipped: ${e2eSelection.skipped.join(', ')}`);
|
|
} else if (e2eSelection.selected.length === 0) {
|
|
console.log(' No E2E tests affected.');
|
|
} else {
|
|
console.log(' All E2E tests selected.');
|
|
}
|
|
console.log();
|
|
|
|
console.log(`LLM-judge: selected ${llmSelection.selected.length} of ${Object.keys(LLM_JUDGE_TOUCHFILES).length}, reason: ${llmSelection.reason}`);
|
|
if (llmSelection.selected.length > 0 && llmSelection.selected.length < Object.keys(LLM_JUDGE_TOUCHFILES).length) {
|
|
console.log(` Selected: ${llmSelection.selected.join(', ')}`);
|
|
console.log(` Skipped: ${llmSelection.skipped.join(', ')}`);
|
|
} else if (llmSelection.selected.length === 0) {
|
|
console.log(' No LLM-judge tests affected.');
|
|
} else {
|
|
console.log(' All LLM-judge tests selected.');
|
|
}
|
|
}
|