feat(linkedin-studio): RE-R2a — item→store capture bridge + publishedAt persistence (schema v1→v2, lossless migrate) [skip-docs]
Closes the research-engine capture loop RE-R1 deferred:
- itemToInput(item, capturedAt): pure envelope→TrendInput bridge in item.ts —
injects capturedAt, carries publishedAt verbatim; no id, no re-validate
- publishedAt persisted: TrendRecord/TrendInput gain it; addTrend conditional-spread,
first-sight kept on re-capture (no back-fill). SCHEMA_VERSION 1→2 with a lossless
forward migrate-on-load: Math.max(onDisk, current) + numeric-typeof coercion
(string/NaN/absent → current; non-array trends coercion preserved verbatim)
- `capture` CLI: stdin raw item|batch → normalize → bridge → addTrend → saveStore once;
tally {added,duplicates,merged,errors} from AddResult; content-invalid → errors[],
exit 2 only on bad stdin; --json summary
- wiring: trend-spotter.md Step 4.5 N×`add` → one normalizing `capture` batch; README
add/capture framing corrected; test-runner Section 16h (capture wiring, unconditional)
+ floors bumped (trends 62→79, ASSERT 87→90)
TDD: 17 new tests (12 genuinely-RED logic-RED + 5 regression guards), tsc clean,
gate 105/0/0. No version bump (additive, v0.5.2 dev).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01VmHCQjJHUyWwxGAVVjNLgp
This commit is contained in:
parent
b4e500fad4
commit
7a158030b6
10 changed files with 465 additions and 31 deletions
|
|
@ -9,15 +9,20 @@
|
|||
* node --import tsx src/cli.ts status [--store <path>] [--json]
|
||||
* echo '<raw item|batch>' | node --import tsx src/cli.ts normalize
|
||||
* echo '<scored candidates>' | node --import tsx src/cli.ts score [--mode kortform|long-form] [--threshold N]
|
||||
* echo '<raw item|batch>' | node --import tsx src/cli.ts capture [--store <path>] [--json]
|
||||
*
|
||||
* The capture agent (research-engine) calls `add` to fold a freshly-polled trend
|
||||
* into the store, and `query`/`list` to reason over accumulated history. The
|
||||
* The capture agent (research-engine) folds freshly-polled trends into the store via
|
||||
* `capture` (the normalizing batch path: stdin → normalizeItem(s) → itemToInput →
|
||||
* addTrend), and reasons over accumulated history via `query`/`list`. `add` is the
|
||||
* MANUAL single-trend path (raw flags, no normalization, publish-date-free). The
|
||||
* polling + relevance-scoring itself lives upstream; this is the deterministic store.
|
||||
*
|
||||
* `normalize` + `score` (RE-R1) are the deterministic research-engine seam: both read
|
||||
* their JSON PAYLOAD FROM STDIN (so they do not overload `--json`, which stays an
|
||||
* output toggle) and print JSON to stdout. `normalize` validates raw items into the
|
||||
* canonical envelope; `score` triages scored candidates (composite/band/threshold).
|
||||
* `normalize` + `score` (RE-R1) and `capture` (RE-R2a) are the deterministic
|
||||
* research-engine seam: all read their JSON PAYLOAD FROM STDIN (so they do not overload
|
||||
* `--json`, which stays an output toggle). `normalize` validates raw items into the
|
||||
* canonical envelope; `score` triages scored candidates (composite/band/threshold);
|
||||
* `capture` normalizes + folds each valid item into the store (persisting `publishedAt`),
|
||||
* reporting content-invalid items in the summary `errors[]`, never via the exit code.
|
||||
*
|
||||
* Exit code: 0 on success, 2 on usage error (incl. unparseable stdin / bad flag).
|
||||
*/
|
||||
|
|
@ -33,7 +38,7 @@ import {
|
|||
queryByTopic,
|
||||
saveStore,
|
||||
} from "./store.js";
|
||||
import { normalizeItem, normalizeItems } from "./item.js";
|
||||
import { normalizeItem, normalizeItems, itemToInput } from "./item.js";
|
||||
import { triage } from "./score.js";
|
||||
import type { ScoreMode } from "./score.js";
|
||||
|
||||
|
|
@ -72,7 +77,8 @@ function usage(msg: string): never {
|
|||
" list [--since <YYYY-MM-DD>] [--limit <n>] [--store <path>] [--json]\n" +
|
||||
" status [--store <path>] [--json]\n" +
|
||||
" normalize < raw-item-or-batch.json\n" +
|
||||
" score [--mode kortform|long-form] [--threshold N] < scored-candidates.json",
|
||||
" score [--mode kortform|long-form] [--threshold N] < scored-candidates.json\n" +
|
||||
" capture [--store <path>] [--json] < raw-item-or-batch.json",
|
||||
);
|
||||
process.exit(2);
|
||||
}
|
||||
|
|
@ -228,6 +234,34 @@ function main(): void {
|
|||
return;
|
||||
}
|
||||
|
||||
if (command === "capture") {
|
||||
const payload = readStdinJson(); // exits 2 on empty/unparseable stdin
|
||||
const raw = Array.isArray(payload) ? payload : [payload];
|
||||
const { items, errors } = normalizeItems(raw);
|
||||
const store = loadStore(storePath);
|
||||
// Tally derived from AddResult {added, merged} (no `duplicates` field): a fold is
|
||||
// `added` (new), else `merged` (existing gained topics), else a plain `duplicate`.
|
||||
let added = 0;
|
||||
let merged = 0;
|
||||
let duplicates = 0;
|
||||
for (const item of items) {
|
||||
const res = addTrend(store, itemToInput(item, today()));
|
||||
if (res.added) added++;
|
||||
else if (res.merged) merged++;
|
||||
else duplicates++;
|
||||
}
|
||||
saveStore(storePath, store);
|
||||
if (asJson) {
|
||||
console.log(JSON.stringify({ added, duplicates, merged, errors }, null, 2));
|
||||
return;
|
||||
}
|
||||
console.log(
|
||||
`Captured into ${storePath}: ${added} added, ${merged} merged, ` +
|
||||
`${duplicates} duplicate, ${errors.length} invalid (${store.trends.length} total)`,
|
||||
);
|
||||
return;
|
||||
}
|
||||
|
||||
usage(command ? `unknown command: ${command}` : "no command given");
|
||||
}
|
||||
|
||||
|
|
|
|||
Loading…
Add table
Add a link
Reference in a new issue