Building upon the swagger petstore, we have some files:
- src/main.js. This has the initialization and routing logic.
- src/controllers/Pets.js. Business logic lives here.
- netlify.toml. A netlify configuration file that specifies the functions output
- webpack.conf.js. A simple webpack configuration file.
To run, copy this folder to a new location. In the root of that folder, install
netlify-lambda:
npm i --save-dev netlify-lambda yml-loader
Then install packages in the src folder:
cd src
npm i
AWS Lambda with node is great for hello world, but typically you have dependencies. That means you either bundle everything up in a single JavaScript or upload a zip file. Since zipping is nasty, we'll use webpack to bundle things up.
Note, if you're not used to webpack, get ready for a hill climb. It brings many oddities into the process.
Build the bundle, which goes in ./lambda/main.js. This includes all the
dependencies in src/node_modules:
export NODE_OPTIONS=--openssl-legacy-provider
netlify-lambda build . -c webpack.conf.js
Run the code and let netlify-lambda watch for changes:
netlify-lambda serve . -c webpack.conf.js
Finally, after running the serve command, open your browser to http://localhost:9000/api/.netlify/functions/api/pets to see the calling of the findAll(args) function. If a limit query param is supplied, it'll be passed in the args
hash.
The library exposes a special route, /swagger.json for usage with swagger-ui. You can hit it locally (note the derived basePath!) or just load it up in swagger-ui using the
online petstore:
http://petstore.swagger.io/?url=http://localhost:9000/main/.netlify/functions/main/swagger.json
The example doesn't use a database (just an in-memory map) so don't expect this to work in production.