mirror of
https://github.com/Unstructured-IO/unstructured.git
synced 2025-09-01 12:53:58 +00:00
6c78677ebb
Thanks to @erichare we now have an AstraDB source connector. updating constant names to be more aligned with AstraDB
23 lines
676 B
Bash
23 lines
676 B
Bash
#!/usr/bin/env bash
|
||
|
||
# Processes the Unstructured-IO/unstructured repository
|
||
# through Unstructured's library in 2 processes.
|
||
|
||
# Structured outputs are stored in astra-ingest-output/
|
||
|
||
# NOTE, this script is not ready-to-run!
|
||
# You must enter a token, endpoint and collection name
|
||
# before running.
|
||
|
||
SCRIPT_DIR=$(cd -- "$(dirname -- "${BASH_SOURCE[0]}")" &>/dev/null && pwd)
|
||
cd "$SCRIPT_DIR"/../../.. || exit 1
|
||
|
||
PYTHONPATH=. ./unstructured/ingest/main.py \
|
||
astra \
|
||
--token "<AstraDB Application Token>" \
|
||
--api-endpoint "<AstraDB Api Endpoint>" \
|
||
--collection-name "<AstraDB Collection Name>" \
|
||
--num-processes "2" \
|
||
--output-dir astra-ingest-output \
|
||
--verbose
|