## Queue text extraction of all matching documents again (tenant:dms:admin)

`POST /api/dms/v1/extraction/reprocess`

Module: DMS · operationId: `reprocessExtraction` · Tag: Text · Auth: access token (`Authorization: Bearer <accessToken>`)

Re-queues every matching unencrypted document of the tenant with `force` semantics (stored
text is discarded); queued and running extractions are left alone. All filters are optional
and combined; send `{}` for every document. Writes one audit entry.

### Request body

`application/json` (required)

- `category` (string): Document type code (case-insensitive) (max length 64)
- `createdFrom` (string (date-time)): Created at or after
- `createdTo` (string (date-time)): Created before
- `status` (string): Current extraction status (one of `NONE`, `DONE`, `PARTIAL`, `SKIPPED`, `FAILED`)

Example:

```json
{
  "category": "incoming_document",
  "createdFrom": "2026-01-01T00:00:00Z",
  "status": "FAILED"
}
```

### Responses

- **200** Number of queued documents (`application/json`)
- **400** Error (`application/json`)

Fields of the 200 answer:

- `queued` (integer, required): Documents queued for extraction

Example 200 answer:

```json
{
  "queued": 37
}
```

### Example request

```bash
curl -X POST 'https://api.mbcom.cloud/api/dms/v1/extraction/reprocess' \
  -H "Authorization: Bearer $ACCESS_TOKEN" \
  -H 'Content-Type: application/json' \
  --data '{"category":"incoming_document","createdFrom":"2026-01-01T00:00:00Z","status":"FAILED"}'
```
