Sha256: 883c1a641d802e65df2b2ed851b3ad15a98b1b5eb706fed5e7add3ee1ea6dcd3
Contents?: true
Size: 1.51 KB
Versions: 12
Compression:
Stored size: 1.51 KB
Contents
.well
%h2= current_page_title
%p
There are two approaches to custom filters with joined tables. The first thing that comes to mind is to define
a column of a joined table with
%code :attribute
and
%code :assoc
, and submit a list of all possible values of this column to
%code :custom_filter
, like it is done in the example below in the Priority column:
%p
%code g.column name: 'Priority', attribute: 'name', assoc: :priority, custom_filter: %w(Anecdotic High Low Normal Urgent) do |task|
%p
This works but such a filter produces a query with a
%code WHERE
clause comparing a varchar field with one of the string values submitted to
%code :custom_filter
\.
This is not guaranteed to be as efficient as comparing an indexed integer foreign key with an integer primary key, thus, this approach is highly advised against.
%p
To implement filtering by foreign keys, define the column with the foreign key in
%code :attribute
and submit a hash or a two element array containing the IDs and labels of the joined table to
%code :attribute
\.
This has a negative side effect on sorting - the column will now be sorted according to the numerical value of the foreign key.
It can be dealt with by overiding sorting by
%code :custom_order
in
%code initialize_grid
\:
%p
%code custom_order: {'tasks.status_id' => 'statuses.position', 'tasks.project_id' => 'projects.name'}
= show_code
.row-fluid
.col-md-12
= render 'grid'
Version data entries
12 entries across 12 versions & 2 rubygems