mcritchlow

Summary

https://developers.google.com/analytics/devguides/reporting/core/v4/rest/v4/reports/batchGet

This gist is for doing some batch query testing against GA. The data we need is:

Page level

1. pageviews, users, newUsers => pageview-query.js

setting pageSize to 100 to get the nextPageToken in the response users = "The total number of users for the requested time period"

mcritchlow

# Ruby script to test out parsing the CIL json data and dump to csv
#
# For now, we're concatenating the property hierarchy with colons
# So CIL: { "Image Files" [ "Mime_type": "application/zip" will be CIL:Image Files:Mime_type "application/zip"
require 'open-uri'
require 'json'
require 'byebug'
require 'csv'

# Hackety hacks, don't talk back

mcritchlow

#!/bin/bash
files_to_download=(
  "https://librarytest.ucsd.edu/dc/object/bb8022051s/_2.mp4/download"
  "https://librarytest.ucsd.edu/dc/object/bb8977691g/_2.mp4/download"
  "https://librarytest.ucsd.edu/dc/object/bb55983863/_1.pdf/download"
  "https://librarytest.ucsd.edu/dc/object/bb38236288/_1.pdf/download"
  "https://librarytest.ucsd.edu/dc/object/bb2015477n/_1.pdf/download"
)
# run tests
# setting jobs to 0 will run as many jobs as possible concurrently

mcritchlow

# This is a crazy idea. but, do we really need a graph-based approach?
# Is this stuff really going to change that often? 
# Could we use something like Single Table Inheritance to make this more sane since so much data is shared?

class Work < ApplicationRecord
  has_many :subjects
  has_namy :agents
  
  delegate :topics, :names, :anatomies, to: :subjects
  delegate :creators, :contributors, :publishers, to: :agents