[teiid-issues] [JBoss JIRA] (TEIID-2666) Extend OData output formats, currently JSON and XML supported

[Steven Hawkins (JIRA)]

    [ https://issues.jboss.org/browse/TEIID-2666?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12968191#comment-12968191 ] 

Steven Hawkins edited comment on TEIID-2666 at 5/15/14 1:58 PM:
----------------------------------------------------------------

Just wanted to add a comment; while I don't need CSV and can certainly use an ESB layer to handle streaming JSON transformations, it would be really useful to extend or change the JSON formatter without having to patch OData4j.  Here's an example of what we do:

* The JSON initially looks either like so: 
{code}
  {
	"d" : {
	  "results" : [
		{
		  "__metadata" : {
			"uri" : "http://somelongURL/odata/etc",
			"type" : "SchemaName.TableName"
		  },
		  "COLUMN_A":value1,
		  "COLUMN_B":value1
		},
		{
		  "__metadata" : {
			"uri" : "http://somelongURL/odata/etc",
			"type" : "SchemaName.TableName"
		  },
		  "COLUMN_A":value2,
		  "COLUMN_B":value2
		}
	  ]
	}
  }
{code}
* Or like so:
{code}
  {
	"d" : {
	  {
		"__metadata" : {
		  "uri" : "http://somelongURL/odata/etc",
		  "type" : "SchemaName.TableName"
		},
		"COLUMN_A":value1,
		"COLUMN_B":value1
	  }
	}
  }
{code} 
* And we want either this (1st case):
{code}
  {
	 "TableNames" : [
	   {
		 "COLUMN_A":value1,
		 "COLUMN_B":value1
	   },
	   {
		 "COLUMN_A":value2,
		 "COLUMN_B":value2
	   }
	 ]
  }
{code}
* Or this (2nd case):
{code}
  {
	 "TableName" : {
	   "COLUMN_A":value1,
	   "COLUMN_B":value1
	 }
  }
{code}
It's really computationally intensive and expensive to have to reformat the JSON responses coming out of Teiid.

Thanks,
John Muller


was (Author: blue666man):
Just wanted to add a comment; while I don't need CSV and can certainly use an ESB layer to handle streaming JSON transformations, it would be really useful to extend or change the JSON formatter without having to patch OData4j.  Here's an example of what we do:

  #  The JSON initially looks either like so: 
  {
	"d" : {
	  "results" : [
		{
		  "__metadata" : {
			"uri" : "http://somelongURL/odata/etc",
			"type" : "SchemaName.TableName"
		  },
		  "COLUMN_A":value1,
		  "COLUMN_B":value1
		},
		{
		  "__metadata" : {
			"uri" : "http://somelongURL/odata/etc",
			"type" : "SchemaName.TableName"
		  },
		  "COLUMN_A":value2,
		  "COLUMN_B":value2
		}
	  ]
	}
  }
 
  # Or like so:
  {
	"d" : {
	  {
		"__metadata" : {
		  "uri" : "http://somelongURL/odata/etc",
		  "type" : "SchemaName.TableName"
		},
		"COLUMN_A":value1,
		"COLUMN_B":value1
	  }
	}
  }
 
  # And we want either this (1st case):
  {
	 "TableNames" : [
	   {
		 "COLUMN_A":value1,
		 "COLUMN_B":value1
	   },
	   {
		 "COLUMN_A":value2,
		 "COLUMN_B":value2
	   }
	 ]
  }
 
  # Or this (2nd case):
  {
	 "TableName" : {
	   "COLUMN_A":value1,
	   "COLUMN_B":value1
	 }
  }

It's really computationally intensive and expensive to have to reformat the JSON responses coming out of Teiid.

Thanks,
John Muller

> Extend OData output formats, currently JSON and XML supported
> -------------------------------------------------------------
>
>                 Key: TEIID-2666
>                 URL: https://issues.jboss.org/browse/TEIID-2666
>             Project: Teiid
>          Issue Type: Feature Request
>          Components: Server
>    Affects Versions: 8.4
>         Environment: N/A
>            Reporter: Cojan van Ballegooijen
>            Priority: Trivial
>
> We would like to see the available output formats of the OData services be extended with CSV (Command Seperated Values) upon customer request.
> It would be great if we could extend it ourselves using for example XSLT or other transformation mechanism.



--
This message was sent by Atlassian JIRA
(v6.2.3#6260)