osu-ctl-scraper 1.0.0

checksums.yaml

@@ -0,0 +1,7 @@
+---
+SHA1:
+  metadata.gz: 0f61f5674ffa411e662d676ca6a1a5c68c3c35b8
+  data.tar.gz: 08336a7892b67f47adf6ad823005f23b057ff9b6
+SHA512:
+  metadata.gz: b9283653ac17768853f815ecad02e26011df72bf67b099e70b5469c60a4873e73ca717457ba568948b4f473ac646905521d1a5c48298831c63494c8ea144ecb1
+  data.tar.gz: d427cdc637e0eecf8072ba812b7d1d08b44b8614f09e25537c415b39b94c634a74e3027ab77fd47d9281ad797ad65763d85d987c6d09a0f583c86f7adc964ad8

data/bin/osu-ctl-scraper

@@ -0,0 +1,9 @@
+#!/usr/bin/env ruby
+
+require "csv"
+require "osu-ctl-scraper"
+
+books = OsuCtlScraper::Book.where("BA", 2015, :spring)
+books.each do |book|
+  $stdout.puts(book.values.to_csv)
+end

data/lib/osu-ctl-scraper.rb

@@ -0,0 +1,10 @@
+require "net/http"
+require "nokogiri"
+require "json"
+require "open-uri"
+
+module OsuCtlScraper
+end
+
+require "osu-ctl-scraper/department"
+require "osu-ctl-scraper/book"

data/lib/osu-ctl-scraper/book.rb

@@ -0,0 +1,88 @@
+module OsuCtlScraper
+  class Book
+    # Returns an array of Books for the given options
+    #
+    # @param  [Hash] options
+    # @return [Array<Hash>]
+    def self.where(subject_code, year, term)
+      res = get_html(subject_code, year, term)
+      html = format_response(res.body)
+      process_html(subject_code, html)
+    end
+
+    # @param  [String] subject_code
+    # @param  [Integer] year
+    # @param  [Symbol] term Must be one of: "winter", "spring", "summer", or "fall"
+    # @return [String]
+    def self.get_html(subject_code, year, term)
+      url = URI.parse("http://osubeaverstore.com/Faculty/GetTextbooks/")
+      params = form_params(subject_code, year, term)
+      res = Net::HTTP.post_form(url, params)
+    end
+
+    # @param  [String] subject_code
+    # @param  [Integer] year
+    # @param  [Symbol] term Must be one of: "winter", "spring", "summer", or "fall"
+    # @return [Hash]
+    def self.form_params(subject_code, year, term)
+      {
+        termcode: termcode(year, term),
+        dept: subject_code
+      }
+    end
+
+    # Formats the year and term into the expected server request format
+    #
+    # @param  [Integer] year
+    # @param  [Symbol] term Must be one of: "winter", "spring", "summer", or "fall"
+    # @return [String]
+    def self.termcode(year, term)
+      terms = { winter: "A", spring: "B", summer: "C", fall: "D" }
+      "#{year}#{terms[term]}"
+    end
+
+    # Marshals the server response into a format understandable by Nokogiri
+    #
+    # @param  [String] body
+    # @return [String]
+    def self.format_response(body)
+      body.
+        gsub('\u003c', '<').
+        gsub('\u003e', '>').
+        gsub('\\"', '"')
+    end
+
+    # @param  [String] subject_code
+    # @param  [String] html
+    # @return [Array<Hash>]
+    def self.process_html(subject_code, html)
+      books = []
+      ng = Nokogiri::HTML(html)
+      ng.css("tr:not(:first-child)").each do |row|
+        book = process_row(row)
+        book[:subject_code] = subject_code
+        books << book
+      end
+      books
+    end
+
+    # @param  [Nokogiri::XML::Element] row
+    # @return [Hash]
+    def self.process_row(row)
+      {
+        course:      row.css("td:nth-child(1)").text.strip,
+        section:     row.css("td:nth-child(2)").text.strip,
+        instructor:  row.css("td:nth-child(3)").text.strip,
+        title:       row.css("td:nth-child(4)").text.strip,
+        edition:     row.css("td:nth-child(5)").text.strip,
+        author:      row.css("td:nth-child(6)").text.strip,
+        isbn:        row.css("td:nth-child(7)").text.strip,
+        publisher:   row.css("td:nth-child(8)").text.strip,
+        requirement: row.css("td:nth-child(9)").text.strip,
+        sku:         row.css("td:nth-child(10)").text.strip,
+        comments:    row.css("td:nth-child(11)").text.strip,
+        req_date:    row.css("td:nth-child(12)").text.strip
+      }
+    end
+  end
+end

data/lib/osu-ctl-scraper/department.rb

@@ -0,0 +1,40 @@
+module OsuCtlScraper
+  class Department
+    # @return [Array<Hash>]
+    def self.all
+      html = get_html
+      process_html(html)
+    end
+
+    # @return [String] html
+    def self.get_html
+      open("http://osubeaverstore.com/faculty/textbooks/").read
+    end
+
+    # @param  [String] html
+    # @return [Array<Hash>]
+    def self.process_html(html)
+      departments = []
+      ng = Nokogiri.HTML(html)
+      ng.css("select[id='Dept'] option:not(:first-child)").each do |option|
+        departments << process_option(option)
+      end
+      departments
+    end
+
+    # @param  [String] option
+    # @return [Hash]
+    def self.process_option(option)
+      {
+        subject_code: option["value"],
+        title: process_title(option.text)
+      }
+    end
+
+    # @param  [String] title
+    # @return [String]
+    def self.process_title(title)
+      title.split(':')[1].strip
+    end
+  end
+end

metadata

@@ -0,0 +1,63 @@
+--- !ruby/object:Gem::Specification
+name: osu-ctl-scraper
+version: !ruby/object:Gem::Version
+  version: 1.0.0
+platform: ruby
+authors:
+- Jonah George
+autorequire: 
+bindir: bin
+cert_chain: []
+date: 2015-08-25 00:00:00.000000000 Z
+dependencies:
+- !ruby/object:Gem::Dependency
+  name: nokogiri
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - '='
+      - !ruby/object:Gem::Version
+        version: 1.6.6.2
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - '='
+      - !ruby/object:Gem::Version
+        version: 1.6.6.2
+description: A script to gather course data from Oregon State's Current Textbook List.
+email: jonah.george@me.com
+executables:
+- osu-ctl-scraper
+extensions: []
+extra_rdoc_files: []
+files:
+- bin/osu-ctl-scraper
+- lib/osu-ctl-scraper.rb
+- lib/osu-ctl-scraper/book.rb
+- lib/osu-ctl-scraper/department.rb
+homepage: https://github.com/jonahgeorge/osu-ctl-scraper
+licenses:
+- MIT
+metadata: {}
+post_install_message: 
+rdoc_options: []
+require_paths:
+- lib
+required_ruby_version: !ruby/object:Gem::Requirement
+  requirements:
+  - - ">="
+    - !ruby/object:Gem::Version
+      version: '0'
+required_rubygems_version: !ruby/object:Gem::Requirement
+  requirements:
+  - - ">="
+    - !ruby/object:Gem::Version
+      version: '0'
+requirements: []
+rubyforge_project: 
+rubygems_version: 2.4.8
+signing_key: 
+specification_version: 4
+summary: OSU Current Textbook List Scraper
+test_files: []
+has_rdoc: