Batch Processing Documents with Unstructured JS SDK

This script will process all documents in a directory using the Unstructured Serverless API JavaScript SDK.

Setup

1. Clone the repo git clone git@github.com:Unstructured-IO/js-client-batch.git
2. Install dependencies npm i
3. Set environment variables in .env

UNSTRUCTURED_API_URL=https://api.unstructuredapp.io || <the URL of your choice>
UNSTRUCTURED_API_KEY=<Your Unstructured API Key>
STRATEGY='fast' || 'hi_res'
SPLIT_PAGES='true' || 'false'

Running

Documents to process

By default this script looks to ./sample_data for files to process, or you can specify the path as in .env as:

...
DOCS_PATH=full/path/to/documents

Output

By default this script will output partitioned results to ./output or you can specify the output path in .env as:

...
OUTPUT_PATH=full/path/to/output

Start

To run the script npm run start