# Reports in csv format the current popular packages on the www.npmjs.com
# by collecting all of the links to the popular packages and then retrieving
# the package page to capture the number of downloads for the package within the past month

# Features used in this script:
# -----------------------------
# Using an initial step with a set action to set a reusable variable for the base-url
# Emitting an initial csv header
# Using some advanced jquery selectors to find content in a page
# Capturing an array of hyperlinks and names
# Playing back a request that utilizes the current array value with <%@ .. %>
# Emitting a csv-formatted line for the current package
# Using the iterate action to advance the link and name arrays' current index and to 
#   repeat the current request until the arrays are exhausted.

- actions:
    - set:
        baseUrl: https://www.npmjs.com
    - emit: package,downloads (past month),link
- request: <% baseUrl %>
  response:
    - jquery: "h2:contains(\"Packages People 'npm install' a Lot\") + ul.columnar a.name"
      capture: 
        link: [ "@href" ]
        name: [ text ]
- request: <% baseUrl %><%@ link %>
  response:
    - jquery: "ul>li:contains('downloads in the last month') strong"
      capture:
        last_month: text
    - emit: <%@# name %>,<% last_month %>,<% baseUrl %><%@ link %>
    - iterate: [ link, name ]