Files
ragflow/internal/ingestion/task/dataflow_chunk_assets.go
Jack 0dd0ac06f8 Feature: task executor migration to go (#16549)
### Summary

Feature: Integrate parser
2026-07-08 19:08:11 +08:00

15 lines
554 B
Go

package task
// PrepareDataflowChunkAssets applies the minimal pre-index cleanup needed by
// the Go dataflow path so real pipeline output can be stored safely.
func PrepareDataflowChunkAssets(chunks []map[string]any) error {
for _, ck := range chunks {
delete(ck, "_pdf_positions")
// FIXME: production parity with Python _prepare_docs_and_upload should
// upload raw image data and set img_id. During the current bring-up phase
// we drop raw image payloads so the main dataflow -> ES path can proceed.
delete(ck, "image")
}
return nil
}