What this category's data looks like
Laboratory hazardous chemical procurement and ledgers cover three types of materials: procurement applications, usage registrations, and inventory ledgers. Sources include supplier qualifications, acceptance receipts, dual-signature usage forms, and monthly inventory sheets. The update cadence varies with procurement batches, usage frequency, and inventory cycles. Fields involve product name, CAS number, specification, quantity, unit, date, handler, approver, signature, and receipt number. When materials enter parsing and training tasks, procurement applications serve as the version starting point, usage registrations serve the role of quantity and date flow, and inventory ledgers serve the role of balance and signature verification. The same record must maintain consistent name, quantity, and date across applications, registrations, and ledgers. After a version change, the old version remains traceable, and references are updated synchronously.
Configuration and implementation checks
This step is executed by parsing and training tasks (server-side tasks), outputting data rows and index status (parsed, indexing, indexed). The version ledger and change records are then completed by retrieval and subsequent processing. After parsing is completed, parsing chunks are immediately persisted as visible data rows and bound to a unique data identifier. The index transitions through three states: parsed, indexing, and indexed. Data pending indexing does not participate in recall. Changing the model only rebuilds data that has already completed indexing.
| Object | Check | Verification action |
|---|---|---|
| Procurement application materials | Whether name, quantity, date, and approval signature are consistent between the application and the ledger, and whether the version number increments with changes | Open the data row to check whether the status is parsed, indexing, or indexed; separately check the identifiers of the file record and the data row; check whether the chunk boundaries corresponding to chunkSize and chunkSplitter cut off the product name and quantity fields |
| Usage registration materials | Whether the usage quantity, usage date, and dual signatures correspond one by one with the inventory ledger, and whether missing signatures or receipts are flagged | Check whether the name, quantity, and date of the usage record and the ledger record match; check whether the processing mode corresponding to trainingType covers this batch of data; confirm whether old-version references are updated synchronously |
| Inventory ledger materials | Whether the balance quantity, inventory date, and handler signature are consistent with the latest version, and whether the old version is traceable | Check whether the data rows covered by the rebuild task are limited to data that has already completed indexing; confirm that data pending indexing is not included in the rebuild chain; check that the tokens and array count of the reference text are recorded separately |
Three common errors
- Symptom: The name, quantity, and date of the same hazardous chemical do not match between the procurement application and the inventory ledger. Cause: The two materials are parsed and persisted separately, and fields are not aligned by the same data identifier.
- Symptom: After a version change, old-version references still point to old content, and the current version is not unique. Cause: Model-change rebuilds only cover data that has already completed indexing, and data pending indexing does not participate in synchronization.
- Symptom: A usage registration lacks a signature or receipt, but the ledger still shows it as used. Cause: Successful parsing does not mean retrieval will definitely hit; when the index status has not reached indexed, the retrieval layer isolates it and does not recall it.
How to verify the setup
Acceptance for this feature: upload one material, observe the status transitions of the data row and the coverage scope of the rebuild task, and confirm that the status transitions among parsed, indexing, and indexed and that the rebuild only covers data that has already completed indexing.
Business test: change one piece of content and see whether the version number and references are updated synchronously.
- Open the data row to check whether the status is parsed, indexing, or indexed, and separately check the identifiers of the file record and the data row.
- Check the data rows covered by the rebuild task, confirm that data pending indexing does not participate in the rebuild, and check the parsing and training records corresponding to
chunkSize,chunkSplitter, andtrainingType.