Contracts.py

Create a faked but coherent Redisearch database with options contracts. Here is an example:

Usage

1. Create a virtual env, activate and install the requirements.

   git clone https://github.com/bjbredis/contracts.git
   virtualenv -p python3 contracts
   cd contracts
   . ./bin/activate
   pip install -r requirements.txt
   

2. Create a Redisearch DB with the following schema:

   ft.create contracts schema product TAG SORTABLE expiry NUMERIC SORTABLE delivery_class TAG SORTABLE type TAG SORTABLE details TEXT NOINDEX price NUMERIC SORTABLE qty NUMERIC SORTABLE value NUMERIC SORTABLE market TAG SORTABLE delivery_component TEXT NOSTEM

3. Set your environment variables:

   export REDIS_HOSTNAME=localhost; export REDIS_PORT=14000; export INDEX_NAME=contracts; export COUNT=1000;

4. Run the script with some form of multiplicity ( note: each process will generate at ~1-1.5k records/sec). Example: (10x COUNT)

   export END=10; for i in $(seq 1 $END); do python contracts.py \&; done

Sample queries

Find all contracts of a given product and delivery class:

FT.SEARCH contracts * @delivery_class:{TSLA-DC} @product:{TSLA-P2}

Find all delivery components of a given delivery class:

FT.SEARCH contracts * @delivery_class:{TSLA-DC} return 1 delivery_component

Find outstanding value of all contracts of each delivery class, sort by value descending:

FT.AGGREGATE contracts * GROUPBY 1 @delivery_class REDUCE sum 1 @value as val SORTBY 2 @val desc

Find outstanding aggregate value of all contracts expiring Monday June 1 2020 for each delivery class, descending by value:

FT.AGGREGATE contracts "@expiry:[1590978200 1590978200]" GROUPBY 1 "@delivery_class" REDUCE sum 1 @value as val SORTBY 2 @val desc