Update the schema

Evolving product requirements are a common challenge in software systems. In this section we will make changes to the Object model, and review the process for persisting those changes to the DB.

Add Phone Support

Add support for a collection of Phone objects to be stored as part of the Customer class.

    List<Phone> phones

    List<Phone> getPhones() { return this.phones }
    void setPhones(List<Phone> phones) { this.phones = phones}

Add appropriate info the customer payload and re-run the save operation. Validate that the saved customer now has a collection of phone objects (should be similar to the Email collection already persisted)

info

Using searches you have already created, search for a record that existed before the phones attribute was added. Notice if any errors or challenges occur from the change.

$elemMatch

To best leverage the new phones structure we need to update our existing customer queries to use the $elemMatch operator. This operator allows you to apply conditions to multiple attributes on the nested phone object that all must evaluate to true before the document is considered to match. For example to search for customers where the phone type is mobile and the value is 913-634-5789 the mongodb query would be

db.customers.find({ phone: {$elemMatch: { type: "mobile", number: "913-634-5789"}}})

To recreate that in Spring-Data it looks like.

    Query query = new Query();
    query.addCriteria(Criteria.where("phones").elemMatch(Criteria.where("type").is("mobile")
         .and("number").is("913-634-5789")));
    mongoTemplate.find(query, Customer.class);

tip

The advanced search approach from the previous lab can also be used here to combine this with the repository pattern.

Because MQL natively supports varied structures in the document model, queries for a specific phone combination are simply treated as not matching when the query experiences a document that doesn't contain the attributes. This flexbility in the query language is a significant difference from SQL and can be a key value when evolving a data model

tip

MongoDB Data Modeling patterns use a Schema Versioning pattern when making structure changes See Maintain Different Schema Versions and Building with Patterns: The Schema Versioning Pattern for details

Email Support

An optional activity is to add in support for an array of email addresses that can be associated with the customer as well.