Initial commit

api/tasks/recover_document_indexing_task.py

@@ -0,0 +1,51 @@
+import datetime
+import logging
+import time
+
+import click
+from celery import shared_task
+from werkzeug.exceptions import NotFound
+
+from core.indexing_runner import IndexingRunner, DocumentIsPausedException
+from extensions.ext_database import db
+from models.dataset import Document
+
+
+@shared_task
+def recover_document_indexing_task(dataset_id: str, document_id: str):
+    """
+    Async recover document
+    :param dataset_id:
+    :param document_id:
+
+    Usage: recover_document_indexing_task.delay(dataset_id, document_id)
+    """
+    logging.info(click.style('Recover document: {}'.format(document_id), fg='green'))
+    start_at = time.perf_counter()
+
+    document = db.session.query(Document).filter(
+        Document.id == document_id,
+        Document.dataset_id == dataset_id
+    ).first()
+
+    if not document:
+        raise NotFound('Document not found')
+
+    try:
+        indexing_runner = IndexingRunner()
+        if document.indexing_status in ["waiting", "parsing", "cleaning"]:
+            indexing_runner.run(document)
+        elif document.indexing_status == "splitting":
+            indexing_runner.run_in_splitting_status(document)
+        elif document.indexing_status == "indexing":
+            indexing_runner.run_in_indexing_status(document)
+        end_at = time.perf_counter()
+        logging.info(click.style('Processed document: {} latency: {}'.format(document.id, end_at - start_at), fg='green'))
+    except DocumentIsPausedException:
+        logging.info(click.style('Document paused, document id: {}'.format(document.id), fg='yellow'))
+    except Exception as e:
+        logging.exception("consume document failed")
+        document.indexing_status = 'error'
+        document.error = str(e)
+        document.stopped_at = datetime.datetime.utcnow()
+        db.session.commit()