Improvements to Example 6-13 that use regular expressions to enable searching by an email address as opposed to an exact string match on the From: field of a JSONified mbox

MTSW2E Example 6-13 Improvements

import json
import pymongo # pip install pymongo
from bson import json_util # Comes with pymongo
import re

# The basis of our query
FROM = "noreply@coursera.org" # As opposed to a value like "Coursera <noreply@coursera.org>"


client = pymongo.MongoClient()
db = client.enron
mbox = db.mbox

# Get the recipient lists for each message

recipients_per_message = db.mbox.aggregate([
  {"$match" : {"From" : re.compile(r".*{0}.*".format(FROM), re.IGNORECASE)}}, 
  {"$project" : {"From" : 1, "To" : 1} }, 
  {"$group" : {"_id" : "$From", "recipients" : {"$addToSet" : "$To" } } }                    
])['result'][0]['recipients']

# Collapse the lists of recipients into a single list

all_recipients = [recipient
                  for message in recipients_per_message
                      for recipient in message]

# Calculate the number of recipients per sent message and sort

recipients_per_message_totals = \
  sorted([len(recipients) 
   for recipients in recipients_per_message])

# Demonstrate how to use $unwind followed by $group to collapse
# the recipient lists into a single list (with no duplicates
# per the $addToSet operator)
    
unique_recipients = db.mbox.aggregate([
  {"$match" : {"From" : re.compile(r".*{0}.*".format(FROM), re.IGNORECASE)}}, 
  {"$project" : {"From" : 1, "To" : 1} }, 
  {"$unwind" : "$To"}, 
  {"$group" : {"_id" : "From", "recipients" : {"$addToSet" : "$To"}} }
])['result'][0]['recipients']

print all_recipients
print "Num total recipients on all messages:", len(all_recipients)
print "Num recipients for each message:", recipients_per_message_totals
print "Num unique recipients", len(unique_recipients)