@ontos-ai/knowhere-sdk 2.4.2 → 2.5.0

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
package/README.md CHANGED
@@ -456,6 +456,18 @@ const knowledge = client.knowledge.withParsedStorage({
456
456
  });
457
457
  ```
458
458
 
459
+ Remote `syncParsedDocument({ documentId })` and `syncParsedDocument({ jobId })`
460
+ stream ZIP entries without loading a full `ParseResult`. Add the optional
461
+ `ParsedDocumentStorage.writeObjectStream` sink for memory usage independent of
462
+ large asset sizes. Each invocation defaults to 100 entries or a 120-second budget
463
+ checked between entries; re-enqueue while `completed` is false. Signed URLs are
464
+ refreshed on every retry, and persisted checkpoints pin the revision and archive.
465
+ `retainResultArchive: true` optionally streams the original ZIP to `result.zip`
466
+ after committing the expanded snapshot. Servers without HTTP Range support need
467
+ temporary disk space for ZIP integrity validation. See the
468
+ [streaming snapshot guide](docs/streamed-parsed-document-sync.md) for the adapter
469
+ contract, retry semantics, and memory validation harness.
470
+
459
471
  If a search result only has a published `documentId`, read-oriented helpers can
460
472
  accept that remote identifier directly. `readChunks` supports display paging;
461
473
  remote fallback requests asset URLs from the Knowhere API, and storage hits use