Skip to main content
Version: 2024.3

Filter Classification Store

With elasticsearch it is possible to index all attributes of Classification Store data without defining an attribute for each single classification store key.

To do so, follow these steps

1) Index Definition

Pimcore ships with a special interpreter to load all classification store attributes to the index. To use this interpreter add following attribute configuration into your index definition:

some_field_name:
fieldname: 'my_classification_store_field_name'
interpreter: Pimcore\Bundle\EcommerceFrameworkBundle\IndexService\Interpreter\DefaultClassificationStore
filtergroup: classificationstore
options:
mapping:
type: 'object'
properties:
keys:
type: 'long'
values:
type: 'object'

This results in a data structure in your index similar to this:

"some_field_name" : {
"values" : {
"1" : [
"one value"
],
"2" : [
"another value"
],
"3" : [
"123",
"99"
],
"5" : [
"AT",
"DZ"
],
"4" : [
""
]
},
"keys" : [
1,
2,
3,
5,
4
]

This contains all the classification store IDs as well as all the data.

2) Filter Type

To interpret the indexed data correctly, a special filter type is needed. As this filter type works only in combination with elasticsearch, it is not activated by default. Following steps are necessary to activate it:

  • make sure the field collection FilteSelectClsStoreAttributes is installed and is allowed in the filter list of the filter definition class.
  • add following filter type mapping (see here for details):
FilteSelectClsStoreAttributes:
filter_type_id: Pimcore\Bundle\EcommerceFrameworkBundle\FilterService\FilterType\ElasticSearch\SelectClassificationStoreAttributes
template: 'product/filters/nested_attributes.html.twig'

3) Template

The template for the filter has to render not only one filter, but the whole list of possible filters for the classification store attributes. For a sample template see our demo.