PDF File Evaluation

Sample project using a workflow and OpenAI API to evaluate PDF files.

Getting Started

Configure the environment variables. Copy example.env to .env and update the values

Create System Identities

The solution use system identities to deploy cloud resources. The following table lists the system identities and their purpose.

System Identities	Authentication	Authorization	Purpose
env.CICD_CLIENT_NAME	OpenId Connect (OIDC) based Federated Identity Credentials	Subscription Contributor access Microsoft Graph API admin consent Permissions: Directory.ReadWrite.All User.Invite.All User.ReadWrite.All	Deploy cloud resources: connectivity resources Common resources Build Docker Images

# Configure the environment variables. Copy `example.env` to `.env` and update the values
cp example.env .env
# load .env vars
[ ! -f .env ] || export $(grep -v '^#' .env | xargs)
# or this version allows variable substitution and quoted long values
[ -f .env ] && while IFS= read -r line; do [[ $line =~ ^[^#]*= ]] && eval "export $line"; done < .env

# Login to az. Only required once per install.
az login --tenant $AZURE_TENANT_ID --use-device-code
az provider register --namespace Microsoft.ContainerService

# Create Azure CICD system identity
./script/create_cicd_sh.sh
# Adds CICD_CLIENT_ID=$created_clientid to .env

Provision Core Resources

# Provision Azure resources
./script/devops.sh provision --name "$APP_NAME" --location "$AZURE_LOCATION"

Deploy the Application

Build and Deploy python api Container App

Development

You'll need to set up a development environment if you want to develop a new feature or fix issues. The project uses a docker based devcontainer to ensure a consistent development environment.

• Open the project in VSCode and it will prompt you to open the project in a devcontainer. This will have all the required tools installed and configured.

Setup local dev environment

If you want to develop outside of a docker devcontainer you can use the following commands to setup your environment.

# Configure the environment variables. Copy example.env to .env and update the values
cp example.env .env

# load .env vars
# [ ! -f .env ] || export $(grep -v '^#' .env | xargs)
# or this version allows variable substitution and quoted long values
# [ -f .env ] && while IFS= read -r line; do [[ $line =~ ^[^#]*= ]] && eval "export $line"; done < .env

# Create and activate a python virtual environment
# Windows
# virtualenv \path\to\.venv -p path\to\specific_version_python.exe
# C:\Users\!Admin\AppData\Local\Programs\Python\Python312\python.exe -m venv .venv
# .venv\scripts\activate

# Linux
# virtualenv .venv /usr/local/bin/python3.12
# python3.12 -m venv .venv
# python3 -m venv .venv
python3 -m venv .venv
source .venv/bin/activate

# Update pip
python -m pip install --upgrade pip

# Install dependencies
pip install -r requirements_dev.txt

# Configure linting and formatting tools
sudo apt-get update
sudo apt-get install -y shellcheck
pre-commit install

Style Guidelines

This project enforces quite strict PEP8 and PEP257 (Docstring Conventions) compliance on all code submitted.

We use Black for uncompromised code formatting.

Summary of the most relevant points:

• Comments should be full sentences and end with a period.
• Imports should be ordered.
• Constants and the content of lists and dictionaries should be in alphabetical order.
• It is advisable to adjust IDE or editor settings to match those requirements.

Use new style string formatting

Prefer f-strings over % or str.format.

# New
f"{some_value} {some_other_value}"
# Old, wrong
"{} {}".format("New", "style")
"%s %s" % ("Old", "style")

One exception is for logging which uses the percentage formatting. This is to avoid formatting the log message when it is suppressed.

_LOGGER.info("Can't connect to the webservice %s at %s", string1, string2)

Testing

Ideally, all code is checked to verify the following:

All the unit tests pass All code passes the checks from the linting tools To run the linters, run the following commands:

# Use pre-commit scripts to run all linting
pre-commit run --all-files

# Run a specific linter via pre-commit
pre-commit run --all-files codespell

# Run linters outside of pre-commit
codespell .
shellcheck -x ./script/*.sh

References

• Container App Private VNet - https://learn.microsoft.com/en-us/azure/container-apps/networking?tabs=workload-profiles-env%2Cazure-cli