mirror of
https://github.com/Fission-AI/OpenSpec.git
synced 2026-10-02 05:24:34 +08:00
skill-templates-parity.test.ts pins a SHA-256 per workflow template so an
unintended template edit fails loudly. The cost lands on every intended
edit: the pinned hashes go stale, and because all 37 live in two maps in
one file, two branches editing different templates collide there on rebase.
Resolving that means hand-editing 64-character hashes, which is where
transcription mistakes come from - and the test proves a hash matches its
source, never that the source is right, so a bad value regenerated over a
bad merge passes CI in silence.
Recompute every pinned hash from the built dist/ and rewrite the map in
place, reporting which entries moved. The skill-directory mapping comes
from getSkillTemplates(), the same helper the skills.sh generator uses, so
adding a workflow needs no second list here; function labels resolve
dynamically against the module exports, so there is no hard-coded list at
all.
"Nothing to update" has to mean it, so four things abort the run without
writing:
- dist/ missing or older than src/, which would pin hashes from a stale
build that the parity test - which reads src/ - then rejects
- a pinned label with no matching export, from a renamed or deleted
template
- a pinned hash whose line the patterns do not recognise, counted by
comparing 64-hex literals found against literals rewritten; the count
uses a deliberately broader pattern so it is a real cross-check rather
than a restatement of the same patterns
- a skill the registry deploys that nothing pins, compared in the other
direction: pins-to-registry only sees pins that already exist
That last direction closes a hole that predates this script. A workflow
added to getSkillTemplates() but never pinned was invisible to the parity
test too, which compares only the entries it already lists - so it shipped
with no golden hash while everything reported success. skill-templates-
parity.test.ts now pins the registry itself, so CI catches it whether or
not anyone runs this script.
The rewriting lives in parity-hash-shared.mjs, following the split between
generate-skillssh.mjs and skillssh-shared.mjs, so those guards can be
exercised against fabricated input. Running the script for real from a test
would rewrite the repository's own parity test file mid-suite. Each case in
parity-hash-shared.test.ts was mutation-checked: removing the guard it
covers makes it fail.
The script cannot silently emit a wrong hash: the parity test recomputes
the same values independently and compares, so a drift between the two
copies of stableStringify fails the test. The test stays the authority.
Dev tooling only. scripts/ is not published (package.json files ships just
scripts/postinstall.js), no src/ is touched, and no runtime behaviour
changes - hence no changeset.
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
131 lines
5.4 KiB
JavaScript
131 lines
5.4 KiB
JavaScript
/**
|
|
* Shared helpers for the parity-hash regeneration script and its tests.
|
|
*
|
|
* The rewriting lives here, separate from `regen-parity-hashes.mjs`, so its
|
|
* guards can be exercised against fabricated input instead of the repository's
|
|
* own parity test file. A test that ran the script for real would rewrite
|
|
* `test/core/templates/skill-templates-parity.test.ts` on disk mid-suite.
|
|
*/
|
|
|
|
/**
|
|
* Pinned-hash line patterns.
|
|
*
|
|
* The trailing comma is optional: the last entry of a map may legally omit it,
|
|
* and requiring it silently skipped such a pin.
|
|
*
|
|
* CRLF needs no special handling. `test/**` carries no `text eol=lf` attribute,
|
|
* so a Windows checkout delivers CRLF, but JavaScript treats `\r` as a line
|
|
* terminator under /m - `$` matches before it, so the carriage return is never
|
|
* consumed and survives the rewrite. (Python's `re.M` does not, which is worth
|
|
* knowing before porting these patterns anywhere.)
|
|
*/
|
|
const FUNCTION_PIN = /^(\s+)(get[A-Za-z0-9]+): '([0-9a-f]{64})'(,?)$/gm;
|
|
const CONTENT_PIN = /^(\s+)'(openspec-[a-z0-9-]+)': '([0-9a-f]{64})'(,?)$/gm;
|
|
|
|
/** Any 64-hex literal, however it is written. */
|
|
const HEX_LITERAL = /'[0-9a-f]{64}'/g;
|
|
|
|
/**
|
|
* Rewrite every pinned hash in the parity test's source.
|
|
*
|
|
* `resolveFunctionHash(name)` and `resolveContentHash(dirName)` return the hash
|
|
* a pin should now hold, or `undefined` when the label no longer corresponds to
|
|
* anything - a renamed or deleted template, which is an error rather than a
|
|
* line to leave alone.
|
|
*
|
|
* Throws without returning a partial rewrite when the number of 64-hex literals
|
|
* found does not match the number rewritten. That count uses a deliberately
|
|
* broader pattern than the two above, so it is a real cross-check: were it
|
|
* derived from the same patterns, a line they miss would go missing from both
|
|
* sides and prove nothing.
|
|
*
|
|
* @param {string} source - contents of the parity test file
|
|
* @param {{
|
|
* resolveFunctionHash: (name: string) => string | undefined,
|
|
* resolveContentHash: (dirName: string) => string | undefined,
|
|
* knownContentKeys?: Iterable<string>,
|
|
* sourceLabel?: string,
|
|
* }} resolvers
|
|
* @returns {{ source: string, moved: string[] }} rewritten source and the
|
|
* labels whose hash changed
|
|
*/
|
|
export function rewriteParityHashes(
|
|
source,
|
|
{ resolveFunctionHash, resolveContentHash, knownContentKeys = [], sourceLabel = 'the parity test' }
|
|
) {
|
|
const moved = [];
|
|
const seenContentKeys = new Set();
|
|
let seen = 0;
|
|
|
|
const totalHexLiterals = (source.match(HEX_LITERAL) ?? []).length;
|
|
|
|
let rewritten = source.replace(FUNCTION_PIN, (_match, indent, name, previous, comma) => {
|
|
const next = resolveFunctionHash(name);
|
|
if (next === undefined) {
|
|
throw new Error(`${name} is pinned in the parity test but not exported from skill-templates.js`);
|
|
}
|
|
if (next !== previous) moved.push(name);
|
|
seen += 1;
|
|
return `${indent}${name}: '${next}'${comma}`;
|
|
});
|
|
|
|
rewritten = rewritten.replace(CONTENT_PIN, (_match, indent, dirName, previous, comma) => {
|
|
const next = resolveContentHash(dirName);
|
|
if (next === undefined) {
|
|
throw new Error(`'${dirName}' is pinned in the parity test but not returned by getSkillTemplates()`);
|
|
}
|
|
if (next !== previous) moved.push(dirName);
|
|
seenContentKeys.add(dirName);
|
|
seen += 1;
|
|
return `${indent}'${dirName}': '${next}'${comma}`;
|
|
});
|
|
|
|
if (seen !== totalHexLiterals) {
|
|
throw new Error(
|
|
`Rewrote ${seen} of ${totalHexLiterals} 64-hex literals in ${sourceLabel}.\n` +
|
|
'Every one is assumed to be a pinned hash, so the two counts must agree. Either:\n' +
|
|
' - a pinned hash is formatted in a way the patterns here do not match, and ' +
|
|
'would have been left stale without warning: widen them; or\n' +
|
|
' - the file gained a 64-hex literal that is not a pin: narrow the count above ' +
|
|
'so it stops being mistaken for one.'
|
|
);
|
|
}
|
|
|
|
// The checks above only see pins that exist. A workflow added to the registry
|
|
// but never pinned is invisible to them AND to the parity test, which compares
|
|
// only the entries it already lists - so it would ship with no golden hash at
|
|
// all while this reported success. Compare the other direction too.
|
|
const unpinned = [...knownContentKeys].filter((key) => !seenContentKeys.has(key));
|
|
if (unpinned.length > 0) {
|
|
throw new Error(
|
|
`getSkillTemplates() returns ${unpinned.length} skill(s) with no pinned hash in ${sourceLabel}:\n` +
|
|
unpinned.map((key) => ` ${key}`).join('\n') +
|
|
'\nAdd each to EXPECTED_GENERATED_SKILL_CONTENT_HASHES and GENERATED_SKILL_FACTORIES.\n' +
|
|
'Until then the skill ships with no parity coverage, so this run would have ' +
|
|
'reported success while leaving it unguarded.'
|
|
);
|
|
}
|
|
|
|
return { source: rewritten, moved };
|
|
}
|
|
|
|
/**
|
|
* Stable, key-sorted serialisation used to hash a template's payload.
|
|
*
|
|
* Must stay byte-compatible with the copy in
|
|
* `test/core/templates/skill-templates-parity.test.ts`. A divergence cannot pass
|
|
* unnoticed: that test recomputes the hashes independently and compares.
|
|
*/
|
|
export function stableStringify(value) {
|
|
if (Array.isArray(value)) {
|
|
return `[${value.map(stableStringify).join(',')}]`;
|
|
}
|
|
if (value && typeof value === 'object') {
|
|
const entries = Object.entries(value)
|
|
.sort(([left], [right]) => left.localeCompare(right))
|
|
.map(([key, item]) => `${JSON.stringify(key)}:${stableStringify(item)}`);
|
|
return `{${entries.join(',')}}`;
|
|
}
|
|
return JSON.stringify(value);
|
|
}
|