mark-python-client 1.0.0

A script to connect to the MARK server via python

mark-python-client

Description

A client implemented in Python script to connect to the MARK framework and access the MONGO-DB

Project overview:

1. mark_client - source folder of the project

   1.1. mark_python_client.py - script responsible for connecting to the MARK server

   1.2. conftest.py - empty script used by pytest unit test library to point the global PATH variable to the modules and scripts used in the unit tests

2. tests - test folder of the project

   2.1. mark_python_client_test.py - script responsible for testing the mark_client's ability to create a MARK client, set the server url, adding new data to the DB via the server and fetching data from the DB via the server

3. .gitlab-ci.yml - file responsible for the DOCKER tests ran on the project

4. requirements.txt - file responsible for listing all dependencies of the project

Installation

To install all dependencies run :

$pip install -r requirements.txt

Usage

To use the mark_python_client.py script just add it to your python code and you can invoke the different methods defined in the script

from mark_client import mark_python_client as mark_client

#initialize the MARK Client
client = mark_client.MarkClient()
#set the url of the MARK server
client.set_server_url("http://server.ip.address:ip.port")
#add evidence to the database via the MARK server
client.add_evidence(evidence_to_be_added)
#fetch the ranked list of the k-most suspicious entries
ranked_list = client.get_evidence()
#fetch evidences with a specific label and subject
evidences = client.find_evidence(label, subject)

Testing

There are two ways to test the project:

• pytest

By installing the dependencies in requirements.txt, you will also install the pytest module for unit testing. To run pytest just run it from bash in the folder of the project:

$pytest

• running gitlab-ci

The project can be tested using the gitlab-runner tool, distributed by gitlab. First docker and gitlab-runner have to be set up following this tutorial:

Setting up gitlab-runner locally

Create a new test or run the test available in the gitlab-ci file:

test:integration:
  stage: mark-integration
  image: ubuntu:18.04
  script:
    # install what we need
    - apt-get update
    - apt-get install -y mongodb openjdk-8-jdk python2.7 python-pip wget unzip
    - pip install -r requirements.txt
    - service mongodb start
    # Download and unzip mark server
    - wget https://download.cylab.be/mark/server-$MARK_VERSION-standalone.zip
    - unzip server*
    # Launch server and let it run...
    - bash server*/run.sh &
    - sleep 30
    - pytest
    # Kill server and check if db entries are correct
    - kill %1

Once gitlab-runner is set up it can easily be called using:

$gitlab-runner exec docker {name_of_your_test_in_gitlab-ci_file}

When the gitlab-runner is ran it will automatically start a docker container with the image specified in the test that is to be run and test the project in it by downloading the required libraries from the requirements.txt file and also install other needed packages via apt-get.