Major Many OData primitive types are JSON Schema primitive types combined with a set of restrictions, e.g. Edm.Int64 is
- "type":"integer",
- "minimum":-9223372036854775808,
- "maximum":9223372036854775807
or Edm.Guid is - "type":"string",
- "pattern":"^[0-9a-fA-F]
{8}
-[0-9a-fA-F]
{4}-[0-9a-fA-F]{4}-[0-9a-fA-F]
{4}-[0-9a-fA-F]
{12}$"
There are basically three options:
1) place these in definitions in edm.json and reference them there
2) inline these definitions at each occurrence of the OData primitive type
3) just combine the JSON Schema primitive type with an OData-specific format, e.g.
- type: string
- format: guid
or - type: integer
- format: int64
The third approach is the most readable and in sync with what Swagger does, see http://swagger.io/specification, section Data Types, and we can try to get these formats into the next version of JSON Schema, see https://github.com/json-schema/json-schema/wiki/%22format%22-suggestions
| Field | Original Value | New Value |
|---|---|---|
| Proposal | Define specific formats for OData primitive types | |
| Description |
Many OData primitive types are JSON Schema primitive types combined with a set of restrictions, e.g. Edm.Int64 is - "type":"integer", - "minimum":-9223372036854775808, - "maximum":9223372036854775807 or Edm.Guid is - "type":"string", - "pattern":"^[0-9a-fA-F]{8}-[0-9a-fA-F]{4}-[0-9a-fA-F]{4}-[0-9a-fA-F]{4}-[0-9a-fA-F]{12}$" There are basically three options: - place these in definitions in edm.json and reference them there - inline these definitions at each occurrence of the OData primitive type - just combine the JSON Schema primitive type with an OData-specific format, e.g. - type: string - format: guid or - type: integer - format: int64 The third approach is the most readable and in sync with what Swagger does, see http://swagger.io/specification, section Data Types |
Many OData primitive types are JSON Schema primitive types combined with a set of restrictions, e.g. Edm.Int64 is - "type":"integer", - "minimum":-9223372036854775808, - "maximum":9223372036854775807 or Edm.Guid is - "type":"string", - "pattern":"^[0-9a-fA-F]{8}-[0-9a-fA-F]{4}-[0-9a-fA-F]{4}-[0-9a-fA-F]{4}-[0-9a-fA-F]{12}$" There are basically three options: - place these in definitions in edm.json and reference them there - inline these definitions at each occurrence of the OData primitive type - just combine the JSON Schema primitive type with an OData-specific format, e.g. - type: string - format: guid or - type: integer - format: int64 The third approach is the most readable and in sync with what Swagger does, see http://swagger.io/specification, section Data Types, and we can try to get these formats into the next version of JSON Schema, see https://github.com/json-schema/json-schema/wiki/%22format%22-suggestions |
| Proposal | Define specific formats for OData primitive types |
Define specific formats for OData primitive types: - uuid - date - time - duration - byte, int8, int16, int32, int64 - double, single - base64url |
| Description |
Many OData primitive types are JSON Schema primitive types combined with a set of restrictions, e.g. Edm.Int64 is - "type":"integer", - "minimum":-9223372036854775808, - "maximum":9223372036854775807 or Edm.Guid is - "type":"string", - "pattern":"^[0-9a-fA-F]{8}-[0-9a-fA-F]{4}-[0-9a-fA-F]{4}-[0-9a-fA-F]{4}-[0-9a-fA-F]{12}$" There are basically three options: - place these in definitions in edm.json and reference them there - inline these definitions at each occurrence of the OData primitive type - just combine the JSON Schema primitive type with an OData-specific format, e.g. - type: string - format: guid or - type: integer - format: int64 The third approach is the most readable and in sync with what Swagger does, see http://swagger.io/specification, section Data Types, and we can try to get these formats into the next version of JSON Schema, see https://github.com/json-schema/json-schema/wiki/%22format%22-suggestions |
Many OData primitive types are JSON Schema primitive types combined with a set of restrictions, e.g. Edm.Int64 is - "type":"integer", - "minimum":-9223372036854775808, - "maximum":9223372036854775807 or Edm.Guid is - "type":"string", - "pattern":"^[0-9a-fA-F]{8}-[0-9a-fA-F]{4}-[0-9a-fA-F]{4}-[0-9a-fA-F]{4}-[0-9a-fA-F]{12}$" There are basically three options: 1) place these in definitions in edm.json and reference them there 2) inline these definitions at each occurrence of the OData primitive type 3) just combine the JSON Schema primitive type with an OData-specific format, e.g. - type: string - format: guid or - type: integer - format: int64 The third approach is the most readable and in sync with what Swagger does, see http://swagger.io/specification, section Data Types, and we can try to get these formats into the next version of JSON Schema, see https://github.com/json-schema/json-schema/wiki/%22format%22-suggestions |
| Issue Type | Bug [ 1 ] | Improvement [ 4 ] |
| Environment | Proposed |
| Status | New [ 10000 ] | Open [ 1 ] |
| Resolution | Fixed [ 1 ] | |
| Status | Open [ 1 ] | Resolved [ 5 ] |
Questions:
-Should we name the formats with an edm prefix (i.e., Edm.Guid) or not?
==>Since Swagger already defined guid, int64, etc., we should just add what is missing (w/o extensions)
-If we specify a format, do we (for some types) also specify pattern (i.e., guid/uuid, time, datetimeoffset)?
==>open new issue
| Proposal |
Define specific formats for OData primitive types: - uuid - date - time - duration - byte, int8, int16, int32, int64 - double, single - base64url |
Define specific formats for OData primitive types: - uuid - date - time - duration - uint8, int8, int16, int32, int64 - double, single - base64url |
| Environment | Proposed | Applied |
| Resolution | https://www.oasis-open.org/committees/download.php/56552/odata-json-csdl-v4.0-wd01-2015-09-24.docx | |
| Status | Resolved [ 5 ] | Applied [ 10002 ] |
Approved 2015-10-1
| Status | Applied [ 10002 ] | Closed [ 6 ] |
Resolved as proposed 2015-9-17