@twinsearth/w2m-dsh-plugin 2.0.3 → 2.0.5

package/src/plugin/dataflow.mjs

@@ -1,72 +0,0 @@
-/**
- * Data flow across the fleet: map results that are too big to return inline travel as blobs, and the reduce step names
- * exactly which blob it read.
- *
- * ## Why a plan instead of a pipe
- *
- * Machines do not send data to each other directly: the relay is the only channel every machine can reach, and it already
- * stores content-addressed blobs. So the shape is map -> PUT blob -> reduce by sha256. The decisions worth testing are the
- * boundaries: when an output is inline, when it becomes a blob, and what happens when a blob a reduce depends on is not
- * there. A pipeline that silently drops a shard is worse than one that fails.
- */
-
-/** Stage stdout at or below this travels inline in the result envelope. */
-export const MAX_INLINE_BYTES = 64 * 1024;
-
-/** Past this a stage output is not a result any more: it is a data problem the operator has to see. */
-export const HARD_CAP_BYTES = 16 * 1024 * 1024;
-
-/**
- * Decide how one machine's stage output travels.
- *
- * @param {{machineId?: string, stageIndex?: number, stdoutBytes?: number}} input
- * @returns {{ok: true, mode: 'inline'|'blob', stdoutBytes: number} | {ok: false, code: string, detail: object}}
- */
-export function planStageOutput(input = {}) {
-  const machineId = typeof input.machineId === 'string' && input.machineId !== '' ? input.machineId : null;
-  if (machineId === null) return { ok: false, code: 'STAGE_MACHINE_MISSING', detail: {} };
-  const stdoutBytes = Number.isFinite(input.stdoutBytes) ? Math.max(0, Math.floor(input.stdoutBytes)) : 0;
-  if (stdoutBytes > HARD_CAP_BYTES) {
-    return { ok: false, code: 'STAGE_OUTPUT_TOO_LARGE', detail: { machineId, stdoutBytes, hardCapBytes: HARD_CAP_BYTES } };
-  }
-  return { ok: true, mode: stdoutBytes <= MAX_INLINE_BYTES ? 'inline' : 'blob', stdoutBytes };
-}
-
-const SHA256 = /^[0-9a-f]{64}$/;
-
-/**
- * Build the reduce input list from the map results.
- *
- * Every map machine must contribute exactly one entry: inline text is carried as-is, a blob result must name a sha256.
- * A missing or malformed hash is refused by name -- the reduce step must never run on a partial fleet and call the answer
- * complete.
- *
- * @param {{results?: Array<{machine_id?: string, mode?: string, stdout?: string, blob_sha256?: string}>,
- *          reducerMachineId?: string}} input
- * @returns {{ok: true, reducerMachineId: string, inputs: Array<{machine_id: string, mode: 'inline'|'blob', stdout?: string, sha256?: string}>} |
- *           {ok: false, code: string, detail: object}}
- */
-export function buildReducePlan(input = {}) {
-  const reducer = typeof input.reducerMachineId === 'string' && input.reducerMachineId !== '' ? input.reducerMachineId : null;
-  if (reducer === null) return { ok: false, code: 'REDUCER_MISSING', detail: {} };
-  const results = Array.isArray(input.results) ? input.results : null;
-  if (results === null || results.length === 0) return { ok: false, code: 'NO_MAP_RESULTS', detail: {} };
-
-  const seen = new Set();
-  const inputs = [];
-  for (const result of results) {
-    const machineId = typeof result?.machine_id === 'string' && result.machine_id !== '' ? result.machine_id : null;
-    if (machineId === null) return { ok: false, code: 'MAP_MACHINE_MISSING', detail: {} };
-    if (seen.has(machineId)) return { ok: false, code: 'MAP_MACHINE_DUPLICATE', detail: { machineId } };
-    seen.add(machineId);
-    const mode = result?.mode === 'blob' ? 'blob' : 'inline';
-    if (mode === 'blob') {
-      const sha = typeof result?.blob_sha256 === 'string' ? result.blob_sha256.toLowerCase() : '';
-      if (!SHA256.test(sha)) return { ok: false, code: 'BLOB_MISSING', detail: { machineId, sha256: result?.blob_sha256 ?? null } };
-      inputs.push({ machine_id: machineId, mode, sha256: sha });
-    } else {
-      inputs.push({ machine_id: machineId, mode, stdout: typeof result?.stdout === 'string' ? result.stdout : '' });
-    }
-  }
-  return { ok: true, reducerMachineId: reducer, inputs };
-}