odoo
/
odoo
2
0
Fork 0
Code Pull Requests Releases Activity

[IMP] document transformation of read_group fields, add an 'aggregates' property to handle aggregate fields of read_group in a cleaner manner 

bzr revid: xmo@openerp.com-20110516080539-4ltj6um1htwdzndb
This commit is contained in:
Xavier Morel 2011-05-16 10:05:39 +02:00
parent 24275a1de5
commit b78b3d4ead
1 changed files with 56 additions and 17 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

73
addons/base/static/src/js/data.js
View File

@ -33,6 +33,55 @@ openerp.base.DataGroup = openerp.base.Controller.extend( /** @lends openerp.bas
this.groups = null;
},
/**
* The format returned by ``read_group`` is absolutely dreadful:
*
* * A ``__context`` key provides future grouping levels
* * A ``__domain`` key provides the domain for the next search
* * The current grouping value is provided through the name of the
* current grouping name e.g. if currently grouping on ``user_id``, then
* the ``user_id`` value for this group will be provided through the
* ``user_id`` key.
* * Similarly, the number of items in the group (not necessarily direct)
* is provided via ``${current_field}_count``
* * Other aggregate fields are just dumped there
*
* This function slightly improves the grouping records by:
*
* * Adding a ``grouped_on`` property providing the current grouping field
* * Adding a ``value`` and a ``length`` properties which replace the
* ``$current_field`` and ``${current_field}_count`` ones
* * Moving aggregate values into an ``aggregates`` property object
*
* Context and domain keys remain as-is, they should not be used externally
* but in case they're needed...
*
* @param {Object} group ``read_group`` record
*/
transform_group: function (group) {
var field_name = this.group_by[0];
var aggregates = {};
_(group).each(function (value, key) {
if (key.indexOf('__') === 0
|| key === field_name
|| key === field_name + '_count') {
return;
}
aggregates[key] = value;
});
return {
__context: group.__context,
__domain: group.__domain,
grouped_on: field_name,
length: group[field_name + '_count'],
value: group[field_name],
aggregates: aggregates
};
},
fetch: function () {
// internal method
var d = new $.Deferred();
@ -47,21 +96,10 @@ openerp.base.DataGroup = openerp.base.Controller.extend( /** @lends openerp.bas
domain: this.domain,
group_by_fields: this.group_by
}, function () { }).then(function (response) {
self.groups = response.result;
// read_group results are annoying: they use the name of the
// field grouped on to hold the value and the count, so no
// generic access to those values is possible.
// Alias them to `value` and `length`.
d.resolveWith(self, [_(response.result).map(function (group) {
console.log(group);
var field_name = self.group_by[0];
return _.extend({}, group, {
// provide field used for grouping
grouped_on: field_name,
length: group[field_name + '_count'],
value: group[field_name]
});
})]);
var data_groups = _(response.result).map(
_.bind(self.transform_group, self));
self.groups = data_groups;
d.resolveWith(self, [data_groups]);
}, function () {
d.rejectWith.apply(d, self, [arguments]);
});
@ -116,8 +154,7 @@ openerp.base.DataGroup = openerp.base.Controller.extend( /** @lends openerp.bas
* The argument to the callback is the list of elements fetched, the
* context (``this``) is the datagroup itself.
*
* The items of a list are the standard objects returned by ``read_group``
* with three properties added:
* The items of a list have the following properties:
*
* ``length``
* the number of records contained in the group (and all of its
@ -132,6 +169,8 @@ openerp.base.DataGroup = openerp.base.Controller.extend( /** @lends openerp.bas
* ``value``
* the value which led to this group (this is the value all contained
* records have for the current ``grouped_on`` field name).
* ``aggregates``
* a mapping of other aggregation fields provided by ``read_group``
*
* @returns {$.Deferred}
*/