An API that allows fast searches for anagrams.
git clone git@github.com:giarcjc/anagram-search.git
Pre-requisities:
- Local install of Docker
- Local install of Docker Compose
npm run docker
This command concatenates two scripts, npm run docker-build which builds the docker image, and npm run dc-up which uses the docker-compose file to spin up the Anagram-Search api with a linked Redis DB in it's own container.
The first time you run the docker command, the project will take a bit of time to build the image. However after the initial image is built it should be cached, so you can simply run npm run dc-up to spin the project up quickly.
Once the project is up and running you should see a log file resembling
anagram-search | {"name":"Anagram-Search-Api","hostname":"1639c1de5348","pid":46,"level":30,"msg":"Express server bunyan listening on port 3000","time":"2018-09-07T04:29:36.715Z","v":0}
which lets you know the api is ready for your queries.
In order to verify that anagrams being returned for a given word are actual English words, we utilize a corpus. The corpus is stored in Redis, which means we need to ingest a dictionary file (assumed to be dictionary.txt.gz) into the data store.
The corpus is created at startup by importing the dictionary into Redis. You should see a log message resembling
{"name":"Anagram-Search-Api","hostname":"e14e39f76173","pid":46,"level":30,"msg":"Finished Importing Dictionary File.","time":"2018-09-12T02:44:27.252Z","v":0}
Which lets you know that the dictionary has successfully been imported, thus creating the corpus.
If you accidentally (or intentionally) drop the database via the DELETE /words.json route, the db can be re-seeded with the dictionary file by running:
npm run import
to run the import command which will unzip and import the dictionary.txt.gz located at the root level of the project.
Alternately you can seed the DB by sending a POST request to the /seed endpoint.
curl -d {} localhost:3000/seed
As you can see no payload body is necessary when sending a POST request to /seed.
Ex:
curl -i http://localhost:3000/anagrams/read.json
You can call the api via curl, postman, httpie, or whatever means you prefer, but requests must have a content type of application/json. Currently the app spins up on localhost port 3000. There are 2 resources: anagrams and words. (See Documentation for full details of api endpoints).
npm run docs
The API documentation can be viewed by opening the file at docs/api/index.html with the browser of your choice. I included a convenience command which should launch it in Chrome if you're on a Mac (other OS/browser combos should just open the file directly).
The API documentation uses API Blueprint format and was generated via cURL Trace Parser and Aglio.
Pre-requisities:
If you wish to run the project locally without using docker, you need to have node and redis installed locally. The project was developed using Node v10.8.0 although any version of Node starting from v8.11.0 (LTS) should work ok.
First pull dependencies:
npm i ` or `npm install
In a different bash shell, start the redis server
redis-server
Then back in the project dir, start the api server:
npm run start
npm run test
Runs unit tests written with Mocha Chai and Sinon. It's better to run these without redis or the app running because otherwise you'll get conflicts between port numbers being in use.
npm run test-ruby
This runs a suite of ruby smoke tests that were used to guide the development of the api.
npm run build
This project was developed using Typescript so if you modify any of the source code you'll need to compile the ts files to js. If you're doing active development you can run npm run build-dev to watch for changes. Similarly you can run npm run start-dev to start the app locally with supervisor (requires local install of supervisor).