This script pulls data from politikus.sinarproject.org and cache it in networkx to enable offline processing. The cache can then be saved into a Neo4j database for further processing and visualization.

The project depends on the following tools / python package in order to build and install properly.

1. Python 3.9 and up
2. While the development work targets Neo4j 4.1, earlier version should work.
3. Poetry - follow the installation instruction found here.
4. Python wheel - you can install via pip

   pip3 install wheel
   

5. In order to generate graph, python would need to be compiled to work with tk-dev package on Ubuntu.

1. Clone this project

   git clone https://github.com/Sinar/popit_relationship
   cd popit_relationship
   

2. Install and build the project

   poetry build
   

Install the built project with pip (filename of the .whl file may vary). Please ensure your PATH is configured properly.

pip3 install ./dist/popit_relationship-0.1.0-py3-none-any.whl

If you are reinstalling after pulling the latest changes, add a --force-reinstall flag

pip3 install --force-reinstall ./dist/popit_relationship-0.1.0-py3-none-any.whl

Most of the configuration is saved within .env file, please refer to the .env.example for example. Besides NEO4J_AUTH and NEO4J_URI, the script should work with the default settings.

• NEO4J_AUTH stores the username and passsword pair separated by a backslash character /, e.g. neo4j/s0meCompl!catedPassword
• NEO4J_URI stores the URI to the neo4j database, e.g. bolt:hostname:7687
• ENDPOINT_API stores the ENDPOINT API URI, currently defaulted to https://politikus.sinarproject.org/@search, the script should work with other similar APIs
• CRAWL_INTERVAL stores the time to wait between every API call (defaulted to 1 second)
• CACHE_PATH stores the path to the cache file (defaulted to ./primport-cache.gpickle)
• GRAPHML_PATH stores the path used by primport export graphml (defaulted to ./primport-cache.graphml)

The configuration environment variables can be overwritten while executing the script (please refer to the usage examples below).

After following the installation guide, if the python environment is properly configured, a script named primport should be made available. Sub-commands can then be issued for different tasks.

Configuration options can be overriden as environment variables, e.g. when running primport in Bash

NEO4J_AUTH=neo4j/someOtherPassword primport reset db

• primport reset cache resets the cache file
• primport reset db clears the Neo4j database

• primport sync person fetches the Person API
• primport sync org fetches the Organization API
• primport sync post fetches the Post API
• primport sync membership fetches the Membership API
• primport sync relationship fetches the Relationship API
• primport sync ownership fetches the Ownership Control Statement API
• primport sync all fetches all of the above
• primport visualize $node1 [$node2 $node3 ...] generates a graph from cache including $node1 ($node2, $node3 etc are optional).
  • Each $node is a URI to an entity, for instance https://politikus.sinarproject.org/organizations/government-linked-companies/1mdb-real-estate-sdn-bhd
  • The maximum depth can be overwritten by passing --depth flag, eg. --depth=1 (value is defaulted to 3).

• primport save saves the cached data to the Neo4j database to allow further work.

• primport export graphml writes the cached graph to GRAPHML_PATH
• primport export graphml ./primport-cache.graphml writes the cached graph to the specified GraphML file using NetworkX's built-in GraphML exporter

• The script can be executed normally as follows

  git clone https://github.com/Sinar/popit_relationship
  cd popit_relationship
  poetry install
  poetry run python src/popit_relationship/primport.py reset db
  

  (Just replace primport with poetry run python src/popit_relationship/primport.py)

Test is done through PyTest

poetry run pytest