This is a tiny wrapper around the SQLite C API that's written in RubyMotion. It is intentionally bare so as to limit the surface area of code that interacts with the raw C API (misuse of the C API can cause resource leaks and seg faults). Think of it more as a driver for SQLite than anything else.

Whenever possible, it uses Ruby idioms like blocks and exceptions.

Yes.

db = SQLite3::Database.new(":memory:")
db.execute("CREATE TABLE test (id INTEGER PRIMARY KEY AUTOINCREMENT, name TEXT, age INTEGER)")
db.execute("INSERT INTO test (name, age) VALUES (?, ?)", ["brad", 28])
db.execute("INSERT INTO test (name, age) VALUES (?, ?)", ["sparky", 24])

rows = []
db.execute("SELECT * FROM test") do |row|
  rows << row
end

rows.should == [
  { id: 1, name: "brad", age: 28 },
  { id: 2, name: "sparky", age: 24 }
]

# File is stored in your /project/resources/ folder
db = SQLite3::Database.new(File.join(App.resources_path, "my_db.sqlite"))

rows = []
db.execute("SELECT * FROM test ORDER BY id") do |row|
	rows << row
end

*this code assumes you're using BubbleWrap for access to App.resources_path.

• Pass a filename to the constructor, or ":memory:". The database is held open for the lifetime of the object.
• Use execute to run SQL statements. All SQL statements are first prepared, and parameters can be passed as an Array or a Hash. If a Hash is passed, then the SQLite named parameter syntax is assumed to be in use.
• Use execute_debug to see the SQL statement and paramaters passed in the REPL. You should not use this method in production.
• Use execute_scalar to run SQL statements and return the first column of the first row. This is useful for queries like SELECT COUNT(*) FROM posts.