David Chang API

RESTful API for interacting with me programmatically!

Build & run locally (http://localhost:80)

Note

This approach is the quickest way to get everything up and running. However, services are deployed differently than when deployed into GCP so use this as a quick way to build & test functionality.

Configure, build, and run everything using docker compose

$ ./init_env.sh

• Fill out all the details for your local database (replace examples in brackets):

# Shared environment variables for ./api-app, ./webhook-app, ./reverse-proxy, and ./db
ENVIRONMENT=[development]

# ReadMe.io API key secret
README_API_KEY=[""]					# Get from https://dash.readme.com/project/api-davidchang-dev/ > Developer Dashboard > Set Up API Keys

# Database auth/connection variables
DB_USER=[root]
DB_PASSWORD=[password]
DB_NAME=[db]
DB_CONNECTION_NAME=$DB_CONTAINER_NAME:$MYSQL_CONTAINER_PORT
DB_CONNECTION_STRING="$DB_USER:$DB_PASSWORD@tcp($DB_CONNECTION_NAME)/$DB_NAME?charset=utf8&parseTime=True&loc=Local"

# Docker variables

## Network
NETWORK_NAME=[api-bridge]

## Database (built with MySQL locally and Cloud SQL in GCP)
DB_IMAGE_NAME=[mysql-db-image]
DB_CONTAINER_NAME=[db]
DB_HOST_PORT="3306"					# Must match db/Dockerfile
DB_CONTAINER_PORT="3306"				# Must match db/Dockerfile

## API Provider (built with Go)
API_IMAGE_NAME=[go-app-image]
API_CONTAINER_NAME=[api-app]
API_HOST_PORT="8080"					# Must match api-app/Dockerfile
API_CONTAINER_PORT="8080"				# Must match api-app/Dockerfile

## Webhook Provider (built with NodeJS)
WEBHOOK_IMAGE_NAME=[nodejs-app-image]
WEBHOOK_CONTAINER_NAME=[webhook-app]
WEBHOOK_HOST_PORT="8000"				# Must match webhook-app/Dockerfile
WEBHOOK_CONTAINER_PORT="8000"				# Must match webhook-app/Dockerfile

## Reverse Proxy (built with Nginx locally and Cloud Endpoints in GCP)
REVERSE_PROXY_IMAGE_NAME=[nginx-server-image]
REVERSE_PROXY_CONTAINER_NAME=[reverse-proxy]
REVERSE_PROXY_HOST_PORT="80"				# Must match reverse-proxy/Dockerfile
REVERSE_PROXY_CONTAINER_PORT="80"			# Must match reverse-proxy/Dockerfile

• Propagate environments to application and database

$ ./propagate_env.sh

• Generate the db/init.sql file

$ cd db

$ ./init.sh

• Build

$ docker compose build

• Run

$ docker compose up

• Tear down

$ docker compose down

Build & run in GCP

Note

Deploying testing locally uses MySQL as a database and Nginx as a reverse-proxy. Deploying to GCP will use other services such as Cloud SQL and Cloud Endpoints respectively instead. This will require the need to be able to build and deploy each of these services individually to be compatible with the production environment.

Configure, build, and run each service individually

Set up environment

$ ./init_env.sh

• Fill out all the details for your local database (replace examples in brackets):

# Shared environment variables for ./api-app, ./webhook-app, ./reverse-proxy, and ./db
ENVIRONMENT=[development]

# ReadMe.io API key secret
README_API_KEY=[""]					# Get from https://dash.readme.com/project/api-davidchang-dev/ > Developer Dashboard > Set Up API Keys

# Database auth/connection variables
DB_USER=[root]
DB_PASSWORD=[password]
DB_NAME=[db]
DB_CONNECTION_NAME=$DB_CONTAINER_NAME:$MYSQL_CONTAINER_PORT
DB_CONNECTION_STRING="$DB_USER:$DB_PASSWORD@tcp($DB_CONNECTION_NAME)/$DB_NAME?charset=utf8&parseTime=True&loc=Local"

# Docker variables

## Network
NETWORK_NAME=[api-bridge]

## Database (built with MySQL locally and Cloud SQL in GCP)
DB_IMAGE_NAME=[mysql-db-image]
DB_CONTAINER_NAME=[db]
DB_HOST_PORT="3306"					# Must match db/Dockerfile
DB_CONTAINER_PORT="3306"				# Must match db/Dockerfile

## API Provider (built with Go)
API_IMAGE_NAME=[go-app-image]
API_CONTAINER_NAME=[api-app]
API_HOST_PORT="8080"					# Must match api-app/Dockerfile
API_CONTAINER_PORT="8080"				# Must match api-app/Dockerfile

## Webhook Provider (built with NodeJS)
WEBHOOK_IMAGE_NAME=[nodejs-app-image]
WEBHOOK_CONTAINER_NAME=[webhook-app]
WEBHOOK_HOST_PORT="8000"				# Must match webhook-app/Dockerfile
WEBHOOK_CONTAINER_PORT="8000"				# Must match webhook-app/Dockerfile

## Reverse Proxy (built with Nginx locally and Cloud Endpoints in GCP)
REVERSE_PROXY_IMAGE_NAME=[nginx-server-image]
REVERSE_PROXY_CONTAINER_NAME=[reverse-proxy]
REVERSE_PROXY_HOST_PORT="80"				# Must match reverse-proxy/Dockerfile
REVERSE_PROXY_CONTAINER_PORT="80"			# Must match reverse-proxy/Dockerfile

• Propagate environments to application and database

$ ./propagate_env.sh

Create a shared network for your api-app, webhook-app, reverse-proxy and db

$ docker network create $NETWORK_NAME

Build & run a local MySQL database

• Build & run

$ cd ./db

$ ./init.sh

$ docker build -t $DB_IMAGE_NAME .

$ docker run -d -p $DB_HOST_PORT:$DB_CONTAINER_PORT --name $DB_CONTAINER_NAME --network $NETWORK_NAME -e MYSQL_ROOT_PASSWORD=$DB_PASSWORD -e DB_NAME=$DB_NAME $DB_IMAGE_NAME

• Test (TODO)

Build & run API app code

• Build & run

$ cd ./api-app

$ docker build -t $API_IMAGE_NAME .

$ docker run -p $API_HOST_PORT:$API_CONTAINER_PORT -it --rm --name $API_CONTAINER_NAME --network $NETWORK_NAME $API_IMAGE_NAME

• Test

$ curl --request GET --url localhost:$API_HOST_PORT/resumes --header 'accept: application/json'

• Ok if you get a 200 response

Build & run webhook app code

• Build & run

$ cd ./webhook-app

$ docker build -t $WEBHOOK_IMAGE_NAME .

$ docker run -p $WEBHOOK_HOST_PORT:$WEBHOOK_CONTAINER_PORT -it --rm --name $WEBHOOK_CONTAINER_NAME --network $NETWORK_NAME $WEBHOOK_IMAGE_NAME

• Test (TODO)

Build & run Nginx reverse-proxy

$ cd ./reverse-proxy

$ docker build -t $REVERSE_PROXY_IMAGE_NAME .

$ docker run -p $REVERSE_PROXY_HOST_PORT:$REVERSE_PROXY_CONTAINER_PORT -it --rm --name $REVERSE_PROXY_CONTAINER_NAME --network $NETWORK_NAME $REVERSE_PROXY_IMAGE_NAME

• Test (TODO)

Instructions for deploying Cloud Endpoints

Deploy sample image to Cloud Run

$ gcloud run deploy api.davidchang.dev \
    --image="gcr.io/cloudrun/hello" \
    --allow-unauthenticated \
    --platform managed \
    --project=david-chang-websites

Configure Swagger 2.0 specification (aka OpenAPI 2.0.0) to define entire API surface

$ emacs swaggerspec.yaml

Deploy Cloud Endpoints configuration (take note of CONFIG_ID - ie. 2024-03-10r0)

$ gcloud endpoints services deploy swaggerspec.yaml --project david-chang-websites

Build a new ESPv2 image (download and install gcloud_build_image in /usr/local/bin if needed)

$ gcloud_build_image -s api.davidchang.dev \
    -c 2024-03-10r0 -p david-chang-websites

Deploy the new ESPv2 container

$ gcloud run deploy api.davidchang.dev \
  --image="gcr.io/david-chang-websites/endpoints-runtime-serverless:2.47.0-api.davidchang.dev-2024-03-10r0" \
  --allow-unauthenticated \
  --platform managed \
  --project=david-chang-websites

Test

$ export ENDPOINTS_HOST=dev-davidchang-api-cloud-endpoints-oo452quldq-uc.a.run.app
$ curl --request GET \
   --header "content-type:application/json" \
   "https://${ENDPOINTS_HOST}/resumes"

API Design (https://api.davidchang.dev)

Note

The resumes endpoint retrieves and saves resumés (meta data) from/to a database. Future extension to retrieve/save resumés as binary data from/to object storage

/resumes/latest

• GET - Get the most recent resumé

/resumes

• GET - Get all resumés, returned as JSON
• POST - Add a new resumé from request data sent as JSON

/resumes:id

• GET - Get a resumé by ID, returning resumé data as JSON
• PATCH - Update a resumé by ID
• DELETE - Delete a resumé by ID

Documentation

Full documentation can be found at https://docs.davidchang.dev