Repo

Repo is a data mapper and DSL on top of equery and epgpool for writing queries and interacting with PostgreSQL databases in Erlang inspired by Ecto.

Here is an example:

app.config

[
 {epgpool, [
  {database_host, "localhost"},
  {database_name, "mydb"},
  {database_user, "test_user"},
  {database_password, "passwd"}
 ]}
]

m_weather.erl

-module(m_weather).

-export([schema/0]).

schema() ->
    #{ fields => #{
           city => #{type => {varchar, 255}},
           temp_lo => #{type => integer},
           temp_hi => #{type => integer}
           prcp    => #{type => numeric}
       },
       table => <<"weather">>
    }.

repl

1> repo:all(m_weather, [
      q:where(fun([#{city := City}]) -> City =:= <<"Kraków">> end),
      %% or repo_utils:like(#{city => <<"Kraków">>}),
      q:order_by(fun([#{temp_lo := T}]) -> [{T, asc}] end),
      q:limit(10)
   ]).

Custom db type converters

You can implement your own custom dbtype encoders and decoders. All you need is to implement repo_types behaviour and set it via PG_TYPES macro. Example for rebar3:

{overrides, [
    {override, repo, [
        {erl_opts, [
            {d, {'PG_TYPES', my_custom_dbtypes}}
        ]}
    ]}
]}.

Notice

A =:= B or A < B, or etc syntax is available through parse_transform.
It's also avaible in repl. But to use in inside modules you need to include
-include_lib("equer/include/equery.hrl"). or use pg_sql:'=:='(A, B) syntax instead. Parse transform works only inside q callbacks.

More documentation will be later...

TODO

• common tests
• 100% coverge
• spec
• docs
• examples