I needed to extract information from some pages; I wanted to try out functional programming in javascript; I built this.

You can read more about it here.

As described in the article above, this project is mostly about plumbing, i.e. controlling how data is routed between components. These libraries do most of the heavy lifting:

• request: we'll use to perform a GET request to a URL and get the HTML back
• Cheerio: let's us use jQuery-like selectors to select and extract text from HTML elements
• Ramda: the FP utility library
• Sanctuary: we'll just use it for the Maybe and Either implementations. It would have been possible to use it as an FP utility library as well, instead of Ramda.
• Fluture: for the implementation of Future

Currently, there's no library on npm, probably never will be.

Interesting stuff is in the scraper.js file. You should probably leave the scrapeUrl function alone, and just build your own decoder functions (those will depend on the web pages you want to scrape. Look at decodeMovie and decodeActorMovieUrls for examples).

Each decoder function is then passed as an argument to scrapeUrl which will return a Future. You can keep adding stuff to your scraping pipeline by mapping and chaining on that Future until you're ready to fork it.

If anything of this is new to you, I really recommend reading the article above, in which I describe in excruciating detail how I got to this solution.

There are some tests, specifically for the utility functions that define the API for cheerio and request. To run them:

npm test