Go knowledge compiler with scheduler-driven dataset compilation (#17913)

Ports dataset knowledge compilation (wiki/graph/tree/mindmap) to the Go
scheduler with a status contract, aligns wiki storage/retrieval with
Python, sizes prompts by content_length, and resolves embedding batch
size from provider capability.
This commit is contained in:
Zhichang Yu
2026-08-06 15:54:00 +08:00
committed by GitHub
parent addc5acdc0
commit 2e37997ab9
116 changed files with 6678 additions and 1942 deletions

View File

@@ -25,6 +25,13 @@ from utils import wait_for
GO_ONLY_SKIPS = {
"Go route is not implemented": {
# Dataset-level graph/raptor/mindmap indexing via POST/GET/DELETE
# /datasets/:id/index is a Python-era RunIndex/TraceIndex/DeleteIndex
# contract; the Go port schedules dataset compilation through the
# knowledge_compile scheduler instead and does not serve /index.
"test_dataset_index_endpoints",
"test_dataset_index_trace_and_delete_type_contract",
"test_dataset_index_run_with_document_creates_task",
"test_document_download_by_id_invalid_id_contract",
"test_llm_factories_live_auth_contract",
"test_llm_list_live_auth_contract",