googlecloudplatform--generative-ai/gemini/sample-apps/quickbot/document-search-using-agent-builder/backend

QuickBotApp document-search-using-agent-builder

Setting up

1. Create virtualenv and install dependencies

Create a virtual environment on the root of the application, activate it and install the requirements

# check if you are already in the env
pip -V

# if not then
python3 -m venv .venv
source .venv/bin/activate
pip3 install -r requirements.txt

IMPORTANT! VS Code may not recognize your env, in that case type "ctrl + shift + P", then select "Python: Select Interpreter" and then select "Enter interpreter path..." and then select your .venv python interpreter, in this case .backend/.venv/bin/python

2. Setup gcloud credentials

gcloud auth list
gcloud config list

gcloud auth login
gcloud config set project <your project id> 
gcloud auth application-default set-quota-project <your project id>

gcloud auth list
gcloud config list

3. Add environment variables

If you have Mac or Windows (or if you are using zsh console on Linux)

. ./local.env

If you have Linux

Open the file .venv/bin/activate and paste the env variables from .local.env after the PATH export, like this:

...

_OLD_VIRTUAL_PATH="$PATH"
PATH="$VIRTUAL_ENV/bin:$PATH"
export PATH

# Quickbot env variables
export ENVIRONMENT=development
export FRONTEND_URL=http://localhost:4200
export BIG_QUERY_DATASET=eren
...

Check that the env variables has been taken into account, running:

env

You should see the new env variables set there

4. Running the set up script

python3 setup.py

5. Run the application

Finally run using uvicorn

uvicorn main:app --reload --port 8080

Code Styling & Commit Guidelines

To maintain code quality and consistency:

• TypeScript (Frontend): We follow Angular Coding Style Guide by leveraging the use of Google's TypeScript Style Guide using gts. This includes a formatter, linter, and automatic code fixer.
• Python (Backend): We adhere to the Google Python Style Guide, using tools like pylint and black for linting and formatting.
• Commit Messages: We suggest following Angular's Commit Message Guidelines to create clear and descriptive commit messages.

Frontend (TypeScript with gts)

1. Initialize gts (if not already done in the project): Navigate to the frontend/ directory and run:

   npx gts init
   

   This will set up gts and create necessary configuration files (like tsconfig.json). Ensure your tsconfig.json (or a related gts config file like .gtsrc) includes an extension for gts defaults, typically:

   {
     "extends": "./node_modules/gts/tsconfig-google.json",
     // ... other configurations
   }
   

2. Check for linting issues:

   npm run lint
   

   (This assumes a lint script is defined in package.json, e.g., "lint": "gts lint")
3. Fix linting issues automatically (where possible):

   npm run fix
   

   (This assumes a fix script is defined in package.json, e.g., "fix": "gts fix")

Backend (Python with pylint and black)

1. Ensure Dependencies are Installed: Add pylint and black to your backend/requirements.txt file:

   pylint
   black
   

   Then install them within your virtual environment:

   pip install pylint black
   # or pip install -r requirements.txt
   

2. Configure pylint: It's recommended to have a .pylintrc file in your backend/ directory to configure pylint rules. You might need to copy a standard one or generate one (pylint --generate-rcfile > .pylintrc).
3. Check for linting issues with pylint: Navigate to the backend/ directory and run:

   pylint .
   

   (Or specify modules/packages: pylint your_module_name)
4. Format code with black: To automatically format all Python files in the current directory and subdirectories:

   python -m black . --line-length=80