Skip to main content
While we recommend that you allow Bracket to build a table for you, here is a guide for using a table that you’ve created yourself

​
General requirements

  1. A unique identifier (we recommend snake case field names), and our default table generation creates a field called bracket_pkey.
  2. To be able to insert into postgres, you must have a default value. We recommend using the builtin function gen_random_uuid().
  3. In order to leverage the a more efficient polling mechanism where only recently updated/inserted records are polled, you must include a field that tracks when records were modified. When Bracket builds a base for you, we create a field called bracket_last_modified and add a trigger/function that updates that field’s value whenever the record is updated.
Here’s some default SQL that can help you get started: 
CREATE TABLE IF NOT EXISTS schema_name.table_name (
    bracket_pkey UUID DEFAULT gen_random_uuid() primary key,
    bracket_last_modified TIMESTAMP DEFAULT NOW(),
    created_date TIMESTAMP DEFAULT NOW(),
    field_1 TEXT
);

create function bracket_update_timestamp() returns trigger
    language plpgsql
as
$$
      BEGIN
        NEW.bracket_last_modified = NOW();
        RETURN NEW;
      END;
      $$;
alter function bracket_update_timestamp() owner to postgres;


CREATE TRIGGER "update_bracket_last_modified_for_table_name"
BEFORE UPDATE ON schema_name.table_name
FOR EACH ROW EXECUTE FUNCTION bracket_update_timestamp();

​
Event driven requirements

Please see the requirements for the streaming sync methods here