This chapter introduced the BSON types and type-handling behavior that underpin reliable MongoDB applications.

Before moving on, make sure you can explain the following concepts:

1. Why MongoDB uses BSON rather than plain JSON for stored documents.
2. How BSON strings differ from numbers, Booleans, and dates.
3. Why ObjectId values must be converted when received as hexadecimal strings.
4. The differences between Int32, Int64, Double, and Decimal128.
5. Why Decimal128 or integer minor units may be preferable to Double for monetary values.
6. How to distinguish Boolean false from a string containing false.
7. How explicit null differs from a missing field.
8. How arrays are queried and why unbounded arrays require caution.
9. How embedded documents are queried using dot notation.
10. How BSON Date differs from BSON Timestamp.
11. How regular expressions match strings and how indexes may affect their performance.
12. How binary data and UUID representations should be handled consistently.
13. How $type and $isNumber help audit data.
14. How conversion operators can normalize historical data.
15. How schema validation can prevent future type inconsistencies.

Final assessment

Create a collection called transactions with documents containing:

• A transaction identifier.
• A customer identifier.
• A Decimal128 amount.
• A currency code.
• A Boolean success indicator.
• A BSON Date creation time.
• An array of tags.
• An optional nested metadata document.

Then write queries to:

• Find transactions with a specified currency.
• Find successful transactions.
• Filter transactions by a date range.
• Find documents where the optional metadata field is missing.
• Identify documents whose amount is not stored as Decimal128.
• Query a nested metadata field.
• Convert a legacy numeric string to Decimal128 in a read-only aggregation.

Finally, explain how you would migrate inconsistent historical data and prevent invalid types from being inserted again.

Next chapter

The next chapter will cover MongoDB CRUD operations in depth: inserting one or many documents, finding documents, filtering and projecting results, updating selected fields, using update operators, deleting documents safely, interpreting write results, handling upserts, and understanding atomicity at the document level.