View Source Query Runners

AyeSQL can be extended to support more databases than PostgreSQL by implementing the behaviour AyeSQL.Runner. This library ships with two runners already:

• AyeSQL.Runner.Ecto, (it's the default and assumes PostgreSQL is the database).
• AyeSQL.Runner.Postgrex.

By default, AyeSQL uses the runner AyeSQL.Runner.Ecto. This runner only has one option which is :repo for the repo module. Additionally, it converts the result to a list of maps.

In order to support your database, you just need to implement the behaviour e.g. let's say we implement a MySQL our queries module would be created using something like the following:

defmodule MyApp.Queries do
  use AyeSQL, runner: MyApp.Runner.MySQL, pool: MyApp.MySQL.Pool

  defqueries("queries.sql")
end

or

import AyeSQL, only: [defqueries: 3]

defqueries(Queries, "queries.sql",
  runner: MyApp.Runner.MySQL,
  pool: MyApp.MySQL.Pool
)

And our runner could be implemented like this:

defmodule MyApp.Runner.MySQL do
  use AyeSQL.Runner

  @impl AyeSQL
  def run(%AyeSQL.Query{statement: stmt, arguments: args}, options) do
    query_options = Keyword.frop(options, [:pool, :into])
    stmt = transform_stmt(stmt)

    with {:ok, result} <- MyXQL.query(pool, stmt, args, query_options) do
      result = AyeSQL.Runner.handle_result(result)
      {:ok, result}
    end
  end

  @spec transform_stmt(AyeSQL.Query.statement()) :: AyeSQL.Query.statement()
  defp transform_stmt(stmt) do
    Regex.replace(~r/\$(\d+)/, stmt, "?")
  end
end