Merge pull request #17440 from code-dot-org/studioPersonScript

Add CSV support for studio_people script.

bin/oneoff/backfill_data/studio_people

@@ -1,31 +1,78 @@
 #!/usr/bin/env ruby
 
-# This script creates a StudioPerson (if one does not already exist) for every teacher.
+# This script creates a StudioPerson (if one does not already exist) for every teacher. It can be
+# run with or without a CSV file. If there is no file, the script iterates over all users, creating
+# StudioPerson's as necessary. If there is a file, the script creates StudioPersons for the users
+# specified in the file.
+#
+# The CSV for production was created by the following SQL query.
+#
+#   SELECT id
+#   FROM users
+#   WHERE user_type = "teacher" AND studio_person_id IS NULL;
+#
+# USAGE: ./bin/oneoff/backfill_data/studio_people
+#        ./bin/oneoff/backfill_data/studio_people studio_people_data.csv
 
+require 'csv'
 require_relative '../../../dashboard/config/environment'
 
 # The user_id variable is used to track the in-progress user in case of an exception.
 user_id = 0
 
 begin
-  # Unfortunately, this query will not be peformant as the result of no index existing on
-  # users.user_type. As we are batching (via find_each), this should not impact live site traffic.
-  # And since we are willing to allow this script to run for hours, we are fine with this.
-  User.
-    with_deleted.
-    where(studio_person_id: nil).
-    where(user_type: User::TYPE_TEACHER).
-    find_each do |user|
-      user_id = user.id
+  # Depending on whether a filename is specified (via ARGV[0]), we iterate over all users or over
+  # the filename, creating StudioPerson's as necessary.
+  if ARGV[0].nil?
+    # Unfortunately, this query will not be peformant as the result of no index existing on
+    # users.user_type. As we are batching (via find_each), this should not impact users of the site.
+    # It may take a long time for the DB updates to occur, though.
+    User.
+      with_deleted.
+      where(studio_person_id: nil).
+      where(user_type: User::TYPE_TEACHER).
+      find_each do |user|
+        user_id = user.id
 
-      puts "PROCESSING ID: #{user_id}..." unless user_id % 10_000
+        puts "PROCESSING ID: #{user_id}..." unless user_id % 10_000
 
-      user.update!(
-        studio_person: (
-          user.email ? StudioPerson.create!(emails: user.email) : StudioPerson.create!
+        user.update!(
+          studio_person: (
+            user.email ? StudioPerson.create!(emails: user.email) : StudioPerson.create!
+          )
         )
-      )
+      end
+  else
+    filename = ARGV[0]
+
+    CSV.foreach(filename, headers: true) do |row|
+      puts "ROW: #{row}"
+      user_id = row['id']
+
+      puts "PROCESSING ID: #{user_id}..." unless user_id % 10_000
+
+      user = User.with_deleted.find_by_id(user_id)
+
+      # This shouldn't happen... But just in case.
+      unless user
+        puts "MISSING USER: #{user_id}"
+        next
+      end
+      # This may happen if a user is saved (with callbacks) between when the CSV was generated and
+      # the script was run.
+      if user.studio_person_id
+        puts "UNNECESSARY USER: #{user_id}"
+        next
+      end
+      # This may happen if a user changes between when the CSV was generated and the script was run.
+      if user.user_type == 'student'
+        puts "STUDENT USER: #{user_id}"
+        next
+      end
+
+      user.update!(studio_person: StudioPerson.create!(emails: user.email))
     end
+  end
 rescue Exception => e
   puts "EXCEPTION (ID: #{user_id}): #{e.message}"
   raise e