|
| 1 | +```python Python Ingest v2 |
| 2 | +import os |
| 3 | + |
| 4 | +from unstructured_ingest.v2.pipeline.pipeline import Pipeline |
| 5 | +from unstructured_ingest.v2.interfaces import ProcessorConfig |
| 6 | +from unstructured_ingest.v2.processes.connectors.astradb import ( |
| 7 | + AstraDBAccessConfig, |
| 8 | + AstraDBConnectionConfig, |
| 9 | + AstraDBDownloaderConfig, |
| 10 | + AstraDBIndexerConfig, |
| 11 | +) |
| 12 | +from unstructured_ingest.v2.processes.partitioner import PartitionerConfig |
| 13 | +from unstructured_ingest.v2.processes.connectors.local import LocalUploaderConfig |
| 14 | + |
| 15 | +# Chunking and embedding are optional. |
| 16 | + |
| 17 | +if __name__ == "__main__": |
| 18 | + Pipeline.from_configs( |
| 19 | + context=ProcessorConfig(), |
| 20 | + indexer_config=AstraDBIndexerConfig( |
| 21 | + collection_name=os.getenv("ASTRA_DB_COLLECTION"), |
| 22 | + keyspace=os.getenv("ASTRA_DB_KEYSPACE"), |
| 23 | + ), |
| 24 | + downloader_config=AstraDBDownloaderConfig( |
| 25 | + collection_name=os.getenv("ASTRA_DB_COLLECTION"), |
| 26 | + keyspace=os.getenv("ASTRA_DB_KEYSPACE"), |
| 27 | + ), |
| 28 | + source_connection_config=AstraDBConnectionConfig( |
| 29 | + access_config=AstraDBAccessConfig( |
| 30 | + token=os.getenv("ASTRA_DB_APPLICATION_TOKEN"), |
| 31 | + api_endpoint=os.getenv("ASTRA_DB_API_ENDPOINT"), |
| 32 | + ), |
| 33 | + ), |
| 34 | + partitioner_config=PartitionerConfig( |
| 35 | + partition_by_api=True, |
| 36 | + partition_endpoint=os.getenv("UNSTRUCTURED_API_URL"), |
| 37 | + api_key=os.getenv("UNSTRUCTURED_API_KEY"), |
| 38 | + ), |
| 39 | + uploader_config=LocalUploaderConfig(output_dir=os.getenv("LOCAL_FILE_OUTPUT_DIR")), |
| 40 | + ).run() |
| 41 | +``` |
0 commit comments