Chat

Chat server with online toxic message detection.

Running.

• Run ai-service:
  twistd --nodaemon --python chat/server/ai_service.py
  BTW: nodaemon option runs service in the terminal process, you can run in background if you remove that option and add --pidfile=ai --log=ai. ai.pid then contains PID of the service (kill it when you're done).
• Run chat-service:
  twistd --nodaemon --python chat/server/chat_service.py Same as above applies to running in the background.
• Run client: python chat/client.py 8080

Development.

Prerequisites.

1. cd to top level project directory.

2. Create new conda environment from environment.yml file:
   conda env create -n chat_env -f=environment.yml

3. Activate new environment:
   conda activate chat_env

4. Install ai and chat packages in editable mode:
   pip install -e .

Workflow.

Locally:

1. Create new branch from chat_dev branch (naming convention: feature/feature_name):
   git checkout chat_dev
git checkout -b feature/feature_name

2. Make some changes:
   git add file1 file2...
git commit -m "Best commit ever..."

3. Push to your forked repo (-u option to set remote tracking branch):
   git push -u origin feature/feature_name

On GitHub - once you have the feature ready for merge:

4. Create a PR from origin:feature/feature_name to upstream:chat_dev.

5. Once PR is merged, pull changes to your local git repo and push them onto your fork:
   git pull upstream chat_dev
git push origin chat_dev

6. Clean up your working branches (from another branch):
   git branch -d feature/feature_name
git push --delete origin feature/feature_name

Notes:

• Before you create a new branch from chat_dev - do git pull upstream chat_dev.
• If you add new dependency, e.g. scikit-learn - add it to requirements.txt and environment.yml files.