This project is a deployment of the pywb web archive replay and index server to provide an index query mechanism for datasets provided by Common Crawl

To run locally, please install with pip install -r requirements.txt

Common Crawl stores data on Amazon S3 and the data can be accessed via s3 or https. Access to CC data using s3 api is restricted to authenticated AWS users.

Currently, individual indexes for each crawl can be accessed under: https://data.commoncrawl.org/commoncrawl/cc-index/collections/[CC-MAIN-YYYY-WW] or s3://commoncrawl/cc-index/collections/[CC-MAIN-YYYY-WW]

Most of the index will be served from S3, however, a smaller secondary index must be installed locally for each collection.

This can be done automatically by running: install-collections.sh which will install all available collections locally. It uses the AWS CLI tool to sync the index.

If successful, there should be collections directory with at least one index.

To run, simply run cdx-server to start up the index server, or optionally wayback, to run pywb replay system along with the cdx server.

If you have docker installed in your system, you can run index server with docker itself.

git clone https://github.com/commoncrawl/cc-index-server.git
cd cc-index-server
docker build . -t cc-index
docker run --rm --publish 8080:8080 -ti cc-index

You can use install-collections.sh to download indexes to your system and mount it on docker.

The API endpoints correspond to existing index collections in collections directory.

For example, one currently available index is CC-MAIN-2015-06 and it can be accessed via

http://localhost:8080/CC-MAIN-2015-06-index?url=commoncrawl.org

Refer to CDX Server API for more detailed instructions on the API itself.

The pywb README provides additional information about pywb.

Please see the webarchive-indexing repository for more info on how the index is built.