This should only be used where compatibility with StringKeyFormat is required, for all other
uses, please see the more flexible InlineStringCodec.
Provides an codec for Pegasus data that is compatible with StringKeyFormat.
While this codec encodes data in the same format as StringKeyFormat, it is
designed to be used with Courier, not with Play JSON Formats.
Limitations:
- The pegasus bytes type is not supported, consider base64 encoding to a string instead
- The pegasus union type is not supported, please migrate to InlineStringCodec if needed
- The pegasus map type is not supported, please migrate to InlineStringCodec if needed
- Records containing optional fields are not allowed
This codec is "schema aided", meaning that the correct Pegasus schema is required to serialize
or deserialize data, even to the raw DataMap and DataList types. This is because
the order and names of record fields, defined in the schema, must be used by the codec to
correctly serialize/deserialize to StringKeyFormat tuples.
The important type relations are:
StringKeyFormat type | Pegasus raw type | Pegasus schema type | Scala type
---------------------|------------------|---------------------|----------------------------------
Tuple | DataMap | record | case class <TypeName>
Seq | DataList | array | <ItemName>Array extends IndexedSeq[T]
StringKeyFormat tuples are positionally ordered. The order of the tuple values must match the
order of fields in a pegasus record. E.g. the pegasus record:
Intended for Legacy support only!
This should only be used where compatibility with
StringKeyFormatis required, for all other uses, please see the more flexibleInlineStringCodec.Provides an codec for Pegasus data that is compatible with
StringKeyFormat.While this codec encodes data in the same format as
StringKeyFormat, it is designed to be used with Courier, not with Play JSON Formats.Limitations: - The pegasus bytes type is not supported, consider base64 encoding to a string instead - The pegasus union type is not supported, please migrate to
InlineStringCodecif needed - The pegasus map type is not supported, please migrate toInlineStringCodecif needed - Records containing optional fields are not allowedThis codec is "schema aided", meaning that the correct Pegasus schema is required to serialize or deserialize data, even to the raw
DataMapandDataListtypes. This is because the order and names of record fields, defined in the schema, must be used by the codec to correctly serialize/deserialize toStringKeyFormattuples.The important type relations are:
StringKeyFormat type | Pegasus raw type | Pegasus schema type | Scala type ---------------------|------------------|---------------------|---------------------------------- Tuple | DataMap | record | case class <TypeName> Seq | DataList | array | <ItemName>Array extends IndexedSeq[T]
The "Tuple" encoding --------------------
- Example:
[email protected]- JSON Equivalent:{ "email": "[email protected]", "message": "Greetings", "recipient": "John"}- Reserved chars:!~- Escape char:!StringKeyFormat tuples are positionally ordered. The order of the tuple values must match the order of fields in a pegasus record. E.g. the pegasus record:
{ "name": "...", "type": "record", "fields": [ { "name": "message", "type": "..." }, { "name": "recipient", "type": "..." }, { "name": "email", "type": "..." } ] }Would be required for the above tuple example.
An empty string is parsed to a tuple of size 1 containing a single empty string.
The "Seq" encoding ------------------
- Example:
one~two~three- JSON Equivalent:[ "one", "two", "three" ]- Reserved chars:!,- Escape char:!An empty string is parsed to an empty Seq.