stock-alert

Command line app that fetches stock prices, compares them to defined references and triggers alert if necessary.

How to use

• Run this app e.g. once per day to fetch prices,
• it will create its own price database,
• that allows for comparing stock prices to past ones.
• For example: "If the price of stock A is less than 5% of its value of yesterday trigger an alarm"
• Alarms will be printed to stderr and also result in a non-zero return code.
• This way you could embed this app in some infrastructure that sends emails containing stderr on failed jobs.

Note:

• The app can scrape or parses prices from websites or APIs, via an extensible parser mechanism. You can add
  your own favorite website via a parser (see bellow).
• If want less output (only the alerts), just skip stdout by appending > /dev/null to your command.

Docker

Using the docker image is most convenient. It brings its own nodejs installation an does not need to be downloaded explicitly.

docker run -v $(pwd)/example:/workdir schnatterer/stock-alert

Node.js

You can also run it with a local node js instance.

yarn install
cd example
node ../src/cli/main.js

Config

All configuration is done via config.json that is read from the working directory. See example/config.json

Price database

The stock prices are read from and written to a prices.json in the working directory.

Add own parsers

There is one exemplary parser implementation (including a test) but you can easily extend this for your favorit stock website:

• Just provide an <name>parser.json in the src/parsers folder that scrapes/parses the price of your stock from a URL.
• In config.json: set the source of a stock to <name>.

Example (let's set <name> to paul):

docker run -v $(pwd)/example:/workdir -v $(pwd)/src/parsers/paulParser.js:/app/src/parsers/paulParser.js schnatterer/stock-alert

PRs for new parsers welcome.

More feature ideas

• config contains readable name
• Alerts: Display readable name and stock url
• global alerts
• Parser: yahoo finance API
• gzipping prices for more efficient storage of huge numbers of prices