Module elastic_types::document

Base requirements for indexable document mappings.

Structures that can be indexed in Elasticsearch should implement DocumentType. The DocumentType is composed of typical mapping metadata, as well as the mapping for each of its properties.

Documents can be mapped as indexable types, or as an object field on another type.

Examples

Define your Elasticsearch types using Plain Old Rust Structures.

Derive Mapping

Mapping can be generated by deriving ElasticType on a struct:

#[derive(Serialize, ElasticType)]
pub struct MyType {
    pub my_date: Date<DefaultDateFormat>,
    pub my_string: String,
    pub my_num: i32
}

This will produce the following field mapping:

{
    "type": "nested",
    "properties": {
        "my_date": {
            "type": "date",
            "format": "basic_date_time"
        },
        "my_string": {
            "type": "text",
            "fields": {
                "keyword":{
                    "type":"keyword",
                    "ignore_above":256
                }
            }
        },
        "my_num": {
            "type": "integer"
        }
    }
}

It's also possible to adjust the mapping using the #[elastic] attribute.

Override Default Mapping Properties

You can override the mapping meta properties for an object by providing your own mapping type with #[elastic(mapping="{TypeName}")]:

#[derive(Serialize, ElasticType)]
#[elastic(mapping="MyTypeMapping")]
pub struct MyType {
    pub my_date: Date<DefaultDateFormat>,
    pub my_string: String,
    pub my_num: i32
}

#[derive(Default)]
struct MyTypeMapping;
impl DocumentMapping for MyTypeMapping {
    //Give your own name to a type
    fn name() -> &'static str { "my_type" }

    fn data_type() -> &'static str { OBJECT_DATATYPE }
}

This will produce the following field mapping:

{
    "type": "object",
    "properties": {
        "my_date": {
            "type": "date",
            "format": "basic_date_time"
        },
        "my_string": {
            "type": "text",
            "fields": {
                "keyword":{
                    "type":"keyword",
                    "ignore_above":256
                }
            }
        },
        "my_num": {
            "type": "integer"
        }
    }
}

Ignore or Rename Fields

You can then serialise type mappings with #[serde] attributes:

#[derive(ElasticType, Serialize)]
pub struct MyType {
    #[serde(rename="my_renamed_date")]
    pub my_date: Date<DefaultDateFormat>,
    #[serde(skip_serializing)]
    pub ignored: String,
    pub my_num: i32
}

NOTE: Fields with a #[serde(skip_deserializing)] attribute will still be mapped, because they can still be indexed in Elasticsearch.

Limitations

Automatically deriving mapping has the following limitations:

• Generics aren't supported by auto deriving. So you can't #[derive(ElasticType)] on MyType<T>.
• Mapping types can't be shared. This is because they need to map the type fields, so are specific to that type. So you can't share MyTypeMapping between MyType and MyOtherType.

All of the above limitations can be worked around by implementing the mapping manually.

Remember that Elasticsearch will automatically update mappings based on the objects it sees though, so if your 'un-mapped' field is serialised, then an inferred mapping will be added for it.

Manually Implement Mapping

You can build object mappings by manually implementing the DocumentMapping and PropertiesMapping traits:

#[derive(Serialize)]
pub struct MyType {
    pub my_date: Date<DefaultDateFormat>,
    pub my_string: String,
    pub my_num: i32
}

//Implement DocumentType for your type. This binds it to the mapping
impl DocumentType<MyTypeMapping> for MyType { }

//Define the type mapping for our type
#[derive(Default)]
pub struct MyTypeMapping;
impl DocumentMapping for MyTypeMapping {
    fn name() -> &'static str { "my_type" }
}
impl PropertiesMapping for MyTypeMapping {
    fn props_len() -> usize { 3 }

    fn serialize_props<S>(state: &mut S) -> Result<(), S::Error>
    where S: serde::ser::SerializeStruct {
        try!(field_ser(state, "my_date", Date::<DefaultDateFormat>::mapping()));
        try!(field_ser(state, "my_string", String::mapping()));
        try!(field_ser(state, "my_num", i32::mapping()));

        Ok(())
    }
}

Links

• Field Types
• Document Types

Structs

Document

A wrapper type for serialising user types.

Enums

Dynamic

The dynamic setting may be set at the mapping type level, and on each inner object. Inner objects inherit the setting from their parent object or from the mapping type.

Constants

DYNAMIC_DATATYPE	Elasticsearch datatype name.
NESTED_DATATYPE	Elasticsearch datatype name.
OBJECT_DATATYPE	Elasticsearch datatype name.

Traits

DocumentMapping	The base requirements for mapping an object type.
DocumentType	The additional fields available to an indexable Elasticsearch type.
PropertiesMapping	Serialisation for the mapping of object properties.

Functions

doc_ser	Serialise a document mapping using the given serialiser.
field_ser	Serialise a field mapping using the given serialiser.