Skip to content

02_neon_configuration.md

Latest commit

 

History

History
78 lines (73 loc) · 3.22 KB

02_neon_configuration.md

File metadata and controls

78 lines (73 loc) · 3.22 KB

Define structure in neon file

TODO convert to video example

  • Create neon file for example in app/ProductModule/Config/Product.neon
  • Import created file to your application config neon. Usually located inapp/config/config.neon.
includes:
	- ../ProductModule/Config/Product.neon
  • Now you can configure your entity. Each entity has own ElasticSearch type and index with same name. You have to specify only index, type has same name and is only one in index, it is defined in library so you don't have to worry about it because of type deprecation. More details here https://www.elastic.co/guide/en/elasticsearch/reference/master/removal-of-types.html
  • Entity definition is in neon under namespace elasticSearch.entities.EntityName continuing our example in file app/ProductModule/Config/Product.neon:
spameriElasticSearch:
	entities:
		Product:
			index: shop_product
  • You can specify if index has strictly defined mapping. Documentation https://www.elastic.co/guide/en/elasticsearch/reference/current/dynamic.html .
  • In ElasticSearch all data indexed has to have specified mapping if not provided then ElasticSearch tries to guess type from data provided. This is very handy but not always needed and without mistakes. So you can set index mappings as strict. This means newly introduced fields not specified in mapping will throw error when trying to index. Now you can manage all fields introduced and specify their type. But if your application can add fields as needed you need to remember this strict limitation or just do not enable it.
spameriElasticSearch:
	entities:
		Product:
			dynamic: strict
  • Now to specify entity mapping. Each object or encapsulation of sub fields stars with properties: then property name and under it you can specify type and analyzer.
spameriElasticSearch:
	entities:
		Product:
			properties: 
				name:
					type: text
					analyzer: czechDictionary
spameriElasticSearch:
	entities:
		Product:
			properties: 
				name:
					type: text
					analyzer: czechDictionary
				
				details:
					properties:
						productTag:
							type: text

						accessories:
							properties:
								name:
									type: text

								category:
									type: text
  • ElasticSearch does not have array field type instead you can write type of array value. This means all fields can be array of values. In this example product has field productTag which has array of text values.
  • Object type of field is not specified directly, you can just pile fields under it. In this case it is represented by details or accessories field.
  • Accessories can have multiple values of accesory which is object with two fields. More details in entity creation.

Tips

  • Do not use field named id in mapping. Use databaseId or externalId depending from your data originates.
  • Do not use same field name in entity structure.