A data engineer performs the following always-on streaming query to update the downstream table mobiles

spark.readStream.format(‘delta’)
.option(‘readChangeFeed’, ‘true’)
.table(‘customers’)
.select(‘mobile’)
.writeStream
.format(‘delta’)
.option(‘checkpointLocation’, ‘/tmp/cust/_checkpoints/’)
.table(‘mobiles’)

spark.readStream.format(‘delta’)
.option(‘readChangeFeed’, ‘true’)
.table(‘customers’)
.select(‘mobile’)
.writeStream
.format(‘delta’)
.option(‘checkpointLocation’, ‘/tmp/cust/_checkpoints/’)
.table(‘mobiles’)

Assuming that before starting the query, both the tables(mobiles and customers) were empty, what will be the number of records in the mobiles table after the execution of the following statements:

INSERT INTO customers VALUES (‘Tom’, ‘8635875’); 
INSERT INTO customers VALUES (‘Paul’, ‘24984567′); 
INSERT INTO customers VALUES (‘Ben’, ‘79327492’); 
UPDATE customers SET mobile = ‘8474738’ WHERE name = ‘Ben’; 
DELETE FROM customers WHERE mobile = ‘24984567’;

INSERT INTO customers VALUES (‘Tom’, ‘8635875’); 
INSERT INTO customers VALUES (‘Paul’, ‘24984567′); 
INSERT INTO customers VALUES (‘Ben’, ‘79327492’); 
UPDATE customers SET mobile = ‘8474738’ WHERE name = ‘Ben’; 
DELETE FROM customers WHERE mobile = ‘24984567’;