Skip to main content

SQLite With Swift Tutorial: Getting Started

2020년 1월 13일About 18 minSwiftSQLiteArticle(s)blogkodeco.comswiftiosxcodesqlite

SQLite With Swift Tutorial: Getting Started 관련

Swift > Article(s)
Article(s)
SQLite > Article(s)
Article(s)
SQLite With Swift Tutorial: Getting Started
In this SQLite with Swift tutorial, you’ll learn to use a SQLite database with Swift projects by creating tables and inserting, updating and deleting rows.

In software development, it doesn’t take long before you need to persist app data. Often, this comes in the form of data structures. But how do you store those structures efficiently?

Fortunately, some great minds have developed solutions for storing structured data in databases and writing language features to access that data. SQLite is available by default on iOS.

In fact, if you’ve used Core Data before, you’ve already used SQLite. Core Data is just a layer on top of SQLite that provides a more convenient API.

In this SQLite with Swift tutorial, you’ll learn how to perform the following database operations:

  • Create and connect to a database.
  • Create a table.
  • Insert, update, and delete rows.
  • Query the database.
  • Handle SQLite errors.

After learning how to perform these fundamental operations, you’ll see how to wrap them in a Swift-like manner. This will let you write abstraction APIs for your apps and avoid working with the more complicated SQLite C APIs.

Finally, you’ll briefly learn about the popular open-source Swift wrapper stephencelis/SQLite. This will give you a basic understanding of how underlying frameworks work within a wrapper.

Note

Databases like SQLite are massive topics to cover, and covering how to use them is out of scope for this tutorial. This tutorial assumes that you have a basic understanding of relational databases and that you’re primarily here to learn to use SQLite with Swift.

Getting Started

Start by using the [Download Materials] button at the top or bottom of this tutorial, then open SQLiteTutorial.xcworkspace from the starter folder. Next, open the [Tutorial] playground from the [Project navigator].

Note

The project is in an Xcode workspace because it uses the SQLite3 dependency as an embedded binary. This binary contains all the functionality for the SQLite code you’ll write in this tutorial.

Long-click the Play button at the bottom and notice that your playground runs manually instead of automatically:

Manual Run drop-down menu
Manual Run drop-down menu

This means it will only execute when you explicitly invoke the execution by clicking the Play button.

You might also see a destroyPart1Database() call at the top of the page. You can safely ignore this since the database file self-destructs each time the playground runs. This ensures all statements execute successfully as you move through this SQLite with Swift tutorial.

Why Should You Choose SQLite?

SQLite isn’t the only way to persist data on iOS. Besides Core Data, there are many other alternatives for data persistence including Realm, Couchbase Lite, Firebase and NSCoding.

Each of these has its own pros and cons — including SQLite itself. There’s no silver bullet for data persistence. As the developer, it’s up to you to determine which option surpasses the others based on your app’s requirements.

SQLite has some advantages:

  • Shipped with iOS, it adds no overhead to your app’s bundle.
  • SQLite released version 1.0 in August 2000, so it’s tried and tested.
  • It’s well-maintained with frequent releases.
  • Open-source.
  • It uses a query language that’s familiar to database developers and administrators.
  • Cross-platform.

The cons of SQLite are subjective and opinionated, so that research is up to you!

Now that you’ve committed to finding out how SQLite works with Swift, it’s time to get to know some of the most basic SQLite APIs.

The C API

This part of the SQLite with Swift tutorial walks you through the most common SQLite APIs. Wrapping the C API in Swift methods is ideal, but work through the C code first. You’ll wrap your C code in the second part of this tutorial.

Opening a Connection

Before doing anything, create a database connection.

Add the following function in your playground:

func openDatabase() -> OpaquePointer? {
  var db: OpaquePointer?
  guard let part1DbPath = part1DbPath else {
    print("part1DbPath is nil.")
    return nil
  }
  if sqlite3_open(part1DbPath, &db) == SQLITE_OK {
    print("Successfully opened connection to database at \(part1DbPath)")
    return db
  } else {
    print("Unable to open database.")
    PlaygroundPage.current.finishExecution()
  }
}

This function calls sqlite3_open(), which opens or creates a new database file. If it’s successful, it returns an OpaquePointer, which is a Swift type for C pointers. When you call this function, you’ll have to capture the returned pointer to interact with the database.

Many of the SQLite functions return an Int32 result code, which is usually a constant in the SQLite library. For example, SQLITE_OK represents the result code 0. Find a list of result codes on the official SQLite site.

Add the following to open the database:

let db = openDatabase()

Click the Play button to run the playground and watch the console output. If the console isn’t open, click the [Show the Debug area] button to the left of the Play button:

Showing the Debug area
Showing the Debug area

If openDatabase() succeeds, you’ll see some output similar to that below:

Successfully opened connection to database at /Users/username/Library/Developer/XCPGDevices/A1EF655D-E318-4CDC-A35A-54349AFE1E62/data/Containers/Data/Application/43A00877-DF24-4E86-8682-22876DEEAFE4/Documents/Part1.sqlite

Creating a Table

Now that you have a connection to a database file, you can create a table. You’ll work with a very simple table to store contacts.

The table consists of two columns: Id, which is an INT and a PRIMARY KEY, and Name, which is a CHAR(255).

The table with ID and Name headers and filled-out rows
The table with ID and Name headers and filled-out rows

Add the following statement, which contains an SQL statement to create the table. Note you’re using Swift’s multi-line syntax:

let createTableString = """
CREATE TABLE Contact(
Id INT PRIMARY KEY NOT NULL,
Name CHAR(255));
"""

Next, add this function:

func createTable() {
  // 1
  var createTableStatement: OpaquePointer?
  // 2
  if sqlite3_prepare_v2(db, createTableString, -1, &createTableStatement, nil) == 
      SQLITE_OK {
    // 3
    if sqlite3_step(createTableStatement) == SQLITE_DONE {
      print("\nContact table created.")
    } else {
      print("\nContact table is not created.")
    }
  } else {
    print("\nCREATE TABLE statement is not prepared.")
  }
  // 4
  sqlite3_finalize(createTableStatement)
}

Going over this, step-by-step:

Next, you check the returned status code to ensure the statement compiled successfully. If so, the process moves to step 3. Otherwise, you print a message noting the statement didn’t compile.

  1. First, you create a pointer to reference in the next step.
  2. sqlite3_prepare_v2() compiles the SQL statement into byte code and returns a status code — an important step before executing arbitrary statements against your database.
  3. sqlite3_step() runs the compiled statement. In this case, you only step once because this statement has a single result. Later in this tutorial, you’ll see when it’s necessary to step multiple times for a single statement.
  4. You must always call sqlite3_finalize() on your compiled statement to delete it and avoid resource leaks. Once a statement finalizes, you should never use it again.

Note

If you want to learn more about what happened in step 2, check out the SQLite documentation about compiling an SQL statement.

Now, add the following function call:

createTable()

Run your playground. You’ll see the following in your console output:

Contact table created.

Now that you have a table, it’s time to add some data to it. You’re going to add a single row with an Id of 1 and Name of Ray.

Inserting Data

Add the following SQL statement:

let insertStatementString = "INSERT INTO Contact (Id, Name) VALUES (?, ?);"

This might look strange if you haven’t had much SQL experience. Why are the values represented by question marks?

Earlier, you used sqlite3_prepare_v2() to compile your statement. The ? syntax tells the compiler that you’ll provide real values when you actually execute the statement.

This lets you compile statements ahead of time, which improves performance since compilation is a costly operation. You can then reuse the compiled statements with different values.

Next, create the following function:

func insert() {
  var insertStatement: OpaquePointer?
  // 1
  if sqlite3_prepare_v2(db, insertStatementString, -1, &insertStatement, nil) == 
      SQLITE_OK {
    let id: Int32 = 1
    let name: NSString = "Ray"
    // 2
    sqlite3_bind_int(insertStatement, 1, id)
    // 3
    sqlite3_bind_text(insertStatement, 2, name.utf8String, -1, nil)
    // 4
    if sqlite3_step(insertStatement) == SQLITE_DONE {
      print("\nSuccessfully inserted row.")
    } else {
      print("\nCould not insert row.")
    }
  } else {
    print("\nINSERT statement is not prepared.")
  }
  // 5
  sqlite3_finalize(insertStatement)
}

Here’s how the function above works:

The first parameter of the function is the statement to bind to, while the second is a non-zero-based index for the position of the ? you’re binding to. The third and final parameter is the value itself. This binding call returns a status code, but for now, you assume that it succeeds.

  1. First, compile the statement and verify that all is well.
  2. Here, you define a value for the ? placeholder. The function’s name — sqlite3_bind_int() — implies you’re binding an Int to the statement.
  3. Perform the same binding process, but this time for a text value. There are two additional parameters on this call. For the purposes of this tutorial, simply pass -1 and nil for them.
  4. Use sqlite3_step() to execute the statement and verify that it finished.
  5. Finalize the statement. If you were going to insert multiple contacts, you’d retain the statement and reuse it with different values.

Note

If you’d like, you can read more about binding parameters on the SQLite website.

Next, call your new function:

insert()

Run your playground and verify that you see the following in your console output:

Successfully inserted row.

Challenge: Multiple Inserts

Challenge time! Your task is to update insert() to insert an array of contacts.

As a hint, you’ll need to reset your compiled statement to its initial state by calling sqlite3_reset() before you execute it again.

Before
func insert() {
  var insertStatement: OpaquePointer?
  // 1
  if sqlite3_prepare_v2(db, insertStatementString, -1, &insertStatement, nil) == SQLITE_OK {
    let id: Int32 = 1
    let name: NSString = "Ray"
    // 2
    sqlite3_bind_int(insertStatement, 1, id)
    // 3
    sqlite3_bind_text(insertStatement, 2, name.utf8String, -1, nil)
    // 4
    if sqlite3_step(insertStatement) == SQLITE_DONE {
      print("\nSuccessfully inserted row.")
    } else {
      print("\nCould not insert row.")
    }
  } else {
    print("\nINSERT statement is not prepared.")
  }
  // 5
  sqlite3_finalize(insertStatement)
}

Great, you’ve now mastered adding data to your database. In the next section of this tutorial, you’ll learn how to access that data when you need it.

Querying Contacts

Now that you’ve inserted some rows, it’d be nice to verify they’re really there.

Add the following to the playground:

let queryStatementString = "SELECT * FROM Contact;"

This query retrieves all records from the contact table. Using a * returns all columns.

Now, add the following function to the query:

func query() {
  var queryStatement: OpaquePointer?
  // 1
  if sqlite3_prepare_v2(db, queryStatementString, -1, &queryStatement, nil) == SQLITE_OK {
    // 2
    if sqlite3_step(queryStatement) == SQLITE_ROW {
      // 3
      let id = sqlite3_column_int(queryStatement, 0)
      // 4
      guard let queryResultCol1 = sqlite3_column_text(queryStatement, 1) else {
        print("Query result is nil")
        return
      }
      let name = String(cString: queryResultCol1)
      // 5
      print("\nQuery Result:")
      print("\(id) | \(name)")
  } else {
      print("\nQuery returned no results.")
  }
  } else {
      // 6
    let errorMessage = String(cString: sqlite3_errmsg(db))
    print("\nQuery is not prepared \(errorMessage)")
  }
  // 7
  sqlite3_finalize(queryStatement)
}

Taking each numbered comment in turn:

  1. You prepare the statement.
  2. Next, you execute the statement. Note that you’re now checking for the status code SQLITE_ROW, which means that you retrieved a row when you stepped through the result.
  3. Now, you read the values from the returned row. Given what you know about the table’s structure and your query, you can access the row’s values column-by-column. The first column is an Int, so you use sqlite3_column_int() and pass in the statement and a zero-based column index. You assign the returned value to the locally-scoped id constant.
  4. Here, you fetch the text value from the Name column. This is a bit messy due to the C API. First, you capture the value as queryResultCol1 so you can convert it to a proper Swift string on the next line.
  5. Print the results.
  6. Print an error, if any. Want to get adventurous? Change the table name in queryStatementString and see what happens.
  7. You finalize the statement. Now, call your new function:
query()

Run your playground. You’ll see the following output in your console:

Query Result:
1 | Ray

WOOt! It looks like your data made it to the database after all, unless you got adventurous.

Challenge: Printing Every Row

Your task is to update query() to print out every contact in the table.

Before
func query() {
  var queryStatement: OpaquePointer?
  // 1
  if sqlite3_prepare_v2(db, queryStatementString, -1, &queryStatement, nil) == 
      SQLITE_OK {
    // 2
    if sqlite3_step(queryStatement) == SQLITE_ROW {
      // 3
      let id = sqlite3_column_int(queryStatement, 0)
      // 4
      guard let queryResultCol1 = sqlite3_column_text(queryStatement, 1) else {
        print("Query result is nil")
        return
      }
      let name = String(cString: queryResultCol1)
      // 5
      print("\nQuery Result:")
      print("\(id) | \(name)")
  } else {
      print("\nQuery returned no results.")
  }
  } else {
      // 6
    let errorMessage = String(cString: sqlite3_errmsg(db))
    print("\nQuery is not prepared \(errorMessage)")
  }
  // 7
  sqlite3_finalize(queryStatement)
}

So now, you have created a database and can add contacts to it. But what if you need to make changes to any of those contacts? You’ll tackle that issue in the next section.

Updating Contacts

Your next step is to update an existing row. Get started by creating the UPDATE statement:

let updateStatementString = "UPDATE Contact SET Name = 'Adam' WHERE Id = 1;"

Here, you’re using real values instead of ? placeholders. You’d usually use the placeholders and bind the statements properly, but this tutorial will skip those steps for brevity.

Next, add the following function to the playground:

func update() {
  var updateStatement: OpaquePointer?
  if sqlite3_prepare_v2(db, updateStatementString, -1, &updateStatement, nil) == 
      SQLITE_OK {
    if sqlite3_step(updateStatement) == SQLITE_DONE {
      print("\nSuccessfully updated row.")
    } else {
      print("\nCould not update row.")
    }
  } else {
    print("\nUPDATE statement is not prepared")
  }
  sqlite3_finalize(updateStatement)
}

This is a similar flow to what you’ve seen before: prepare, step and finalize.

Next, add the following to your playground:

update()
query()

This executes your new function, then calls your previously-defined query() function so that you can see the results:

Successfully updated row.

Query Result:
1 | Adam

Congratulations on updating your first row! How easy was that?

You’re doing great! You can now create a table, add contacts to it and update those contacts. In the next step, you’ll learn how to delete those contacts.

Deleting Contacts

The final step on the path to becoming an SQLite ninja is to delete the row you created. Again, you’ll use the familiar pattern of prepare, step and finalize.

Add the following to the playground:

let deleteStatementString = "DELETE FROM Contact WHERE Id = 1;"

Now, add the following function to execute the statement:

func delete() {
  var deleteStatement: OpaquePointer?
  if sqlite3_prepare_v2(db, deleteStatementString, -1, &deleteStatement, nil) == 
      SQLITE_OK {
    if sqlite3_step(deleteStatement) == SQLITE_DONE {
      print("\nSuccessfully deleted row.")
    } else {
      print("\nCould not delete row.")
    }
  } else {
    print("\nDELETE statement could not be prepared")
  }
  
  sqlite3_finalize(deleteStatement)
}

Are you feeling it now? Prepare, step and finalize.

Execute this new function as below:

delete()
query()

Run your playground and you should see the following output in your console:

Successfully deleted row.
Query returned no results.

Note

If you completed the Multiple Inserts challenge above, your output will look different from that above due to rows still being present in the table.

So you’ve now mastered working with your table, but what about when things inevitably go wrong? In the next section, you’ll learn how to deal with errors.

Handling Errors

Hopefully, you’ve managed to avoid SQLite errors up to this point. But the time will come when you make a call that doesn’t make sense or simply won’t compile.

Handling error message when these things happen saves a lot of development time. It also gives you the opportunity to present meaningful error messages to your users.

To start, you need an error to handle. Add the following statement, which is intentionally malformed:

let malformedQueryString = "SELECT Stuff from Things WHERE Whatever;"

Now, add a function to execute this malformed statement:

func prepareMalformedQuery() {
  var malformedStatement: OpaquePointer?
  // 1
  if sqlite3_prepare_v2(db, malformedQueryString, -1, &malformedStatement, nil) 
      == SQLITE_OK {
    print("\nThis should not have happened.")
  } else {
    // 2
    let errorMessage = String(cString: sqlite3_errmsg(db))
    print("\nQuery is not prepared! \(errorMessage)")
  }
  
  // 3
  sqlite3_finalize(malformedStatement)
}

Here’s how you’re going to force an error:

  1. Prepare the statement, which will fail and will NOT return SQLITE_OK.
  2. Get the error message from the database using sqlite3_errmsg(). This returns a textual description of the most recent error. You then print the error to the console.
  3. Finalize the statement.

Call the function to see the error message:

prepareMalformedQuery()

Run your playground. You’ll see the following output in your console:

Query is not prepared! no such table: Things

Well, that’s actually helpful — you obviously cannot run a SELECT statement on a table that doesn’t exist! Despite the intentional error that you created, you should use this example as a pattern for handling other errors from SQLite. Nice job!

Closing the Database Connection

When you’re done with a database connection, you’re responsible for closing it. But beware — there are a number of things you must do before you can successfully close your database, as described in the SQLite documentation.

Call the close function:

sqlite3_close(db)

Run your playground. You’ll see a status code of 0 on the results sidebar, which represents SQLITE_OK.

Successful close call
Successful close call

Great, your close call succeeded!

You’ve successfully created a database, added a table, added rows to the table, queried rows, updated rows and even deleted a row, all using the SQLite C APIs from Swift. Great job!

In the next section, you’ll see how to wrap some of these calls in Swift.

SQLite With Swift

As a Swift developer, you might feel a little uneasy about the first part of this tutorial. The good news is you can take the power of Swift and wrap those C routines to make things easier for yourself.

swift-weights
swift-weights

For this part of the SQLite with Swift tutorial, click the Making It Swift link at the bottom of the playground:

Where to find the Making It Swift link
Where to find the Making It Swift link

Wrapping Errors

Getting errors from the C API is a bit awkward as a Swift developer. Checking a result code and then calling another function doesn’t make sense in this new world. It would make more sense if functions that fail threw an error. Your next step will be to make that happen in your code.

Add the following:

enum SQLiteError: Error {
  case OpenDatabase(message: String)
  case Prepare(message: String)
  case Step(message: String)
  case Bind(message: String)
}

This is a custom Error enum that covers four of the main operations you’re using that can fail. Note how each case has an associated value to hold the error message.

Wrapping the Database Connection

Another not-so-Swifty aspect of the work you’ve done so far is those blasted OpaquePointer types. Your next step will be to take care of those.

Wrap up the database connection pointer in its own class, as shown below:

class SQLiteDatabase {
  private let dbPointer: OpaquePointer?
  private init(dbPointer: OpaquePointer?) {
    self.dbPointer = dbPointer
  }
  deinit {
    sqlite3_close(dbPointer)
  }
}

This looks much better. When you need a database connection, you create a reference to a more meaningful type of SQLiteDatabase rather than using an OpaquePointer.

You’ll notice the initializer is private. That’s because you don’t want your Swift developers passing in that OpaquePointer. Instead, you let them instantiate this class with a path to the database file.

Add the following static method to SQLiteDatabase:

static func open(path: String) throws -> SQLiteDatabase {
  var db: OpaquePointer?
  // 1
  if sqlite3_open(path, &db) == SQLITE_OK {
    // 2
    return SQLiteDatabase(dbPointer: db)
  } else {
    // 3
    defer {
      if db != nil {
        sqlite3_close(db)
      }
    }
    if let errorPointer = sqlite3_errmsg(db) {
      let message = String(cString: errorPointer)
      throw SQLiteError.OpenDatabase(message: message)
    } else {
      throw SQLiteError
        .OpenDatabase(message: "No error message provided from sqlite.")
    }
  }
}

Here’s what’s happening:

  1. You attempt to open the database at the provided path.
  2. If successful, you return a new instance of SQLiteDatabase.
  3. Otherwise, you defer closing the database if the status code is anything but SQLITE_OK and throw an error.

Now you can create and open a database connection using much cleaner syntax.

Add the following outside the SQLiteDatabase class:

let db: SQLiteDatabase
do {
    db = try SQLiteDatabase.open(path: part2DbPath ?? "")
    print("Successfully opened connection to database.")
} catch SQLiteError.OpenDatabase(_) {
    print("Unable to open database.")
    PlaygroundPage.current.finishExecution()
}

Ah, much more Swift-like. Here, you wrap the attempt to open the database in a do-try-catch block and pass the error message from SQLite to the catch block, thanks to that custom enum you added earlier.

Run your playground and watch the console output. You’ll see the following:

Successfully opened connection to database.

Now, you can use and inspect the db instance as a proper and meaningful type.

Before moving on to writing methods that execute statements, it would be nice if SQLiteDatabase let you easily access SQLite error messages.

Add the following computed property to the SQLiteDatabase class:

fileprivate var errorMessage: String {
  if let errorPointer = sqlite3_errmsg(dbPointer) {
    let errorMessage = String(cString: errorPointer)
    return errorMessage
  } else {
    return "No error message provided from sqlite."
  }
}

Here, you’ve added a computed property, which simply returns the most recent error SQLite knows about. If there is no error, it returns a generic message stating as much.

Wrapping the Prepare Call

Since you make prepare calls so often, it makes sense to wrap them like the other methods. As you move forward and add functionality to the SQLiteDatabase class, you’ll use class extensions.

Add the following extension. This will be useful in your future methods to invoke sqlite3_prepare_v2() on SQL statements:

extension SQLiteDatabase {
 func prepareStatement(sql: String) throws -> OpaquePointer? {
  var statement: OpaquePointer?
  guard sqlite3_prepare_v2(dbPointer, sql, -1, &statement, nil) 
      == SQLITE_OK else {
    throw SQLiteError.Prepare(message: errorMessage)
  }
  return statement
 }
}

Here, you declare that prepareStatement(_:) can throw an error and then use guard to throw that error should sqlite3_prepare_v2() fail. Just like before, you pass the error message from SQLite to the relevant case of your custom enum.

Creating a Contact Struct

Add the following:

struct Contact {
  let id: Int32
  let name: NSString
}

In these examples, you’ll use the same Contact table as before, so it makes sense to define a proper struct to represent a contact. Next, you’ll learn to actually create the table for a Contact

Wrapping the Table Creation

You’ll knock out the same database tasks as before, but this time you’ll use a “Swift-ier” approach.

To create a table, first, create the following protocol for that purpose in your playground:

protocol SQLTable {
  static var createStatement: String { get }
}

Now, extend Contact to conform to this new protocol:

extension Contact: SQLTable {
  static var createStatement: String {
    return """
    CREATE TABLE Contact(
      Id INT PRIMARY KEY NOT NULL,
      Name CHAR(255)
    );
    """
  }
}

This code defines createStatement and adds a CREATE TABLE statement on Contact which is useful for keeping the code grouped together.

Now, you’re able to write the following method that accepts types that conform to SQLTable to create a table:

extension SQLiteDatabase {
  func createTable(table: SQLTable.Type) throws {
    // 1
    let createTableStatement = try prepareStatement(sql: table.createStatement)
    // 2
    defer {
      sqlite3_finalize(createTableStatement)
    }
    // 3
    guard sqlite3_step(createTableStatement) == SQLITE_DONE else {
      throw SQLiteError.Step(message: errorMessage)
    }
    print("\(table) table created.")
  }
}

Here’s a breakdown of what’s happening:

  1. prepareStatement() throws, so you must use try. You don’t use a do-try-catch block because this method itself throws, so any error from prepareStatement() simply passes to the caller of createTable().
  2. With the power of defer, you ensure that your statements are always finalized, regardless of how this method exits its scope.
  3. guard lets you write a more expressive check for the SQLite status codes.

Give your new method a try by adding the following:

do {
  try db.createTable(table: Contact.self)
} catch {
  print(db.errorMessage)
}

Here, you simply attempt to create the Contact and catch an error, if there is one.

Run your playground. The following will appear in your console:

Contact table created.

Fantastic! Isn’t that a much cleaner API to work with?

Wrapping Insertions

Moving along, it’s time to insert a row into the Contact table. Add the following method in your playground:

extension SQLiteDatabase {
  func insertContact(contact: Contact) throws {
    let insertSql = "INSERT INTO Contact (Id, Name) VALUES (?, ?);"
    let insertStatement = try prepareStatement(sql: insertSql)
    defer {
      sqlite3_finalize(insertStatement)
    }
    let name: NSString = contact.name
    guard 
      sqlite3_bind_int(insertStatement, 1, contact.id) == SQLITE_OK  &&
      sqlite3_bind_text(insertStatement, 2, name.utf8String, -1, nil) 
        == SQLITE_OK 
      else {
        throw SQLiteError.Bind(message: errorMessage)
    }
    guard sqlite3_step(insertStatement) == SQLITE_DONE else {
      throw SQLiteError.Step(message: errorMessage)
    }
    print("Successfully inserted row.")
  }
}

Now that you’ve got your SQLegs – see what I did there? – this code shouldn’t be too surprising. Given a Contact instance, you prepare a statement, bind the values, execute and finalize.

Again, using a potent mix of defer, guard and throw allows you to take advantage of modern Swift language features.

Call this new method:

do {
  try db.insertContact(contact: Contact(id: 1, name: "Ray"))
} catch {
  print(db.errorMessage)
}

Run your playground. You’ll see the following in your console:

Successfully inserted row.

Wrapping Reads

To wrap up the last section on creating the Swift wrapper, you’ll handle querying the database.

Add the following method, which will query the database:

extension SQLiteDatabase {
  func contact(id: Int32) -> Contact? {
    let querySql = "SELECT * FROM Contact WHERE Id = ?;"
    guard let queryStatement = try? prepareStatement(sql: querySql) else {
      return nil
    }
    defer {
      sqlite3_finalize(queryStatement)
    }
    guard sqlite3_bind_int(queryStatement, 1, id) == SQLITE_OK else {
      return nil
    }
    guard sqlite3_step(queryStatement) == SQLITE_ROW else {
      return nil
    }
    let id = sqlite3_column_int(queryStatement, 0)
    guard let queryResultCol1 = sqlite3_column_text(queryStatement, 1) else {
      print("Query result is nil.")
      return nil
    }
    let name = String(cString: queryResultCol1) as NSString
    return Contact(id: id, name: name)
  }
}

This method simply takes the ID of a contact and either returns that contact or nil, if there isn’t a contact with that ID. Again, these statements should feel somewhat familiar by now.

Write the code to query the first contact:

if let first = db.contact(id: 1) {
  print("\(first.id) \(first.name)")
}

Run your playground. You’ll see the following output in the console:

1 Ray

By now, you’ve probably identified some calls you could create in a generic fashion and apply to entirely different tables. The point of the above exercise is to show how you can use Swift to wrap low-level C APIs. This is no simple task for SQLite; there are a ton of intricacies to SQLite that this tutorial didn’t cover.

You might be thinking, “Hasn’t someone already created a wrapper for this?” That’s what the final section is all about.

Introducing SQLite.swift

Stephen Celis has graciously written a fully-featured Swift wrapper for SQLite named SQLite.swift. Check it out if you decide that SQLite fits the bill for data storage in your app.

SQLite.swift provides an expressive way to represent tables. This lets you get started with SQLite without worrying about its underlying details and idiosyncrasies.

You may even consider wrapping SQLite.swift itself to create a high-level API for your app’s domain model.

Check out the [README.md for SQLite.swift and decide for yourself if it has a place in your personal code toolbox.

Where to Go From Here?

Download the completed project by using the [Download Materials] button at the top or bottom of this tutorial. There, you can see other common tasks that this tutorial didn’t have time to cover, including the implemented updates, deletes and multiple row handling.

Debugging is another important part of using SQLite with Swift. In many cases, you’ll need some kind of database browser to see what’s going on under the hood. There are many different apps that can help with this. A quick internet search will show which ones are best for you.

You can also access your SQLite databases directly from your Terminal by typing sqlite3 file.db. From there, use the .help command to see a list of commands, or simply start executing SQL statements directly at the prompt. Find more information on the command-line SQLite client on the main SQLite site.

I hope you enjoyed this whirlwind introduction to working with SQLite and Swift! If you have any questions or comments, please join the discussion below.