gistfile1.txt

create table mydata_real (id serial, date date, value text);
create table mydata_real_y2015 (check (date >= '2015-01-01' and date < '2016-01-01')) inherits (mydata_real);
create table mydata_real_y2016 (check (date >= '2016-01-01' and date < '2017-01-01')) inherits (mydata_real);

create function mydata_nope() returns trigger language plpgsql
  as $f$ begin raise exception 'insert on wrong table'; return NULL; end; $f$; 

create trigger mydata_nope before insert on mydata_real execute procedure mydata_nope();

create view mydata as select * from mydata_real;

-- need to copy any defaults from mydata_real to the view
alter view mydata alter column id set default nextval('mydata_real_id_seq');

create function mydata_partition() returns trigger language plpgsql
  as $f$
    begin
      case extract(year from NEW.date)
        when 2015
          then insert into mydata_real_y2015 select NEW.*;
        when 2016
          then insert into mydata_real_y2016 select NEW.*;
        else
          raise exception 'date % out of range', NEW.date;
      end case;
      return NEW;
    end;
$f$;

create trigger mydata_partition instead of insert on mydata
  for each row execute procedure mydata_partition();

insert into mydata(date,value) values ('2015-01-10','foo'),('2015-12-12','bar'),('2016-02-02','baz') returning id;

Note that the "mydata" view also works for ordinary selects, updates and deletes too—there's no need to access "mydata_real" directly.