LAT Transform Configuration

Data Pipelines are now the preferred method for loading data into the Ocient System. For details, see Load Data﻿.
transform
A transform configuration object.
Required keys:
Transform Configuration source only: transform.topics﻿ 
File sources only: transform.file_groups﻿ 
transform.topics
A collection of topics and their associated configuration. Keys are topic names and values are ﻿topic configuration objects﻿ . Each <topic> set as a key in this object represents an  topic defined in Kafka.
When loading from an s3 or local File Source type, transform.file_groups should be used instead of transform.topics. Each key in the file_group must match a file_group_name defined in the Source section of the pipeline configuration.
Type:
object
Required:
Yes
Default:
﻿
transform.database
A database to be used to fully qualify any table names that are not fully qualified.
For example, if transform.database is set to myDatabase, a table name of the form schema.table will become myDatabase.schema.table.
Type:
string
Required:
No
Default:
null
transform.schema
A schema to be used to fully qualify any table names that are not fully qualified. Specifying this property requires that transform.database﻿  is specified.
For example, if transform.database is set to myDatabase and transform.schema is set to mySchema:
A table name of the form table will become myDatabase.mySchema.table
A table name of the form schema.table will become myDatabase.schema.table
A table name of the form database.schema.table will stay as database.schema.table
Type:
string
Required:
No
Default:
null
Kafka Load Transform Example
JSON
{
    ...
    "transform": {
        "topics": {
            "topic_1": {
                "tables": {
                    ...
                }
            },
            "topic_2": {
                "tables": {
                    ...
                }
            }
        }
    }
}
{
    ...
    "transform": {
        "topics": {
            "topic_1": {
                "tables": {
                    ...
                }
            },
            "topic_2": {
                "tables": {
                    ...
                }
            }
        }
    }
}
﻿
File Based Load Transform Example
Unlike Kafka, File loads define file groups in the source section of the pipeline configuration. The "file_groups" defined in the source and transform sections must match.
JSON
{
    "source": {
        "type": "s3",
        ...
        "file_groups": {
            "file_group_1": {
                ...
            },
            "file_group_2": {
                ...
            },
        }
    }
    "transform": {
        "file_groups": {
            "file_group_1": {
                "tables": {
                    ...
                }
            },
            "file_group_2": {
                "tables": {
                    ...
                }
            }
        }
    }
}
{
    "source": {
        "type": "s3",
        ...
        "file_groups": {
            "file_group_1": {
                ...
            },
            "file_group_2": {
                ...
            },
        }
    }
    "transform": {
        "file_groups": {
            "file_group_1": {
                "tables": {
                    ...
                }
            },
            "file_group_2": {
                "tables": {
                    ...
                }
            }
        }
    }
}
﻿
Topics
Topic configuration objects.
Required keys:
﻿transform.topics.<topic>.tables ﻿ 
For file based loads, topics are replaced by file_groups, but all other settings are equivalent.
transform.topics.<topic>.filter
A record filter to apply at the topic level. See LAT Record Filtering﻿ for details.
Type:
string
Required:
Yes
Default:
﻿
transform.topics.<topic>.tables
A collection of tables and their associated configuration. Keys are table names and values are columns configuration﻿. 
Type:
object
Required:
Yes
Default:
﻿
Tables
Table configuration objects.
Required keys:
﻿transform.topics.<topic>.tables.<table>.columns﻿ 
transform.topics.<topic>.tables.<table>.filter
A record filter to apply at the table level. See LAT Record Filtering﻿ for details.
Type:
string
Required:
Yes
Default:
﻿
transform.topics.<topic>.tables.<table>.columns
A collection of columns and their associated configurations. Keys are table names and values are Columns﻿. 
Columns
Column transformation configurations.
Required keys:
﻿transform.topics.<topic>.tables.<table>.columns.<column> ﻿ 
transform.topics.<topic>.tables.<table>.columns.<column>
A column transformation keyed by a column name. A column’s value is defined as a transformation expression. The expression will query the record and return a value that is loaded into the associated column. The grammar of these expressions uses JMESPath enhanced with some custom  transformations and User Defined Transformations (UDTs).
Type:
string
Required:
Yes
Default:
﻿
Complex Transform Example
JSON
{
    "transform": {
        "database": "adtechdb",
        "schema": "adtech",
        "topics" : {
            "dsp.Auctions" : {
                "tables" : {
                    "auction" : {
                        "columns" : {
                            "auctionid" : "auctionId",
                            "created" : "created",
                            "bidder_nodeid" : "bidder.nodeId",
                            "deals_dealid" : "deals[].dealId"
                        }
                    },
                    "trafficSource" : {
                        "columns" : {
                            "auctionid" : "auctionId",
                            "created" : "created",
                            "sourcesellerid" : "trafficSource.sourceSellerId",
                            "domainname" : "trafficSource.domainName",
                            "contextualdataset_externalentryid" : "trafficSource.siteContextualProfile.contextualDataSet[].entries[].externalEntryId[]",
                            "dimensions_hw" : "join('x', [dimensions.height, dimensions.width])"
                        }
                    },
                    "feedback" : {
                        "columns" : {
                            "auctionid" : "auctionId",
                            "created" : "created",
                            "received" : "received",
                            "campaignid" : "EXPLODE(auctionResponseFeedbacks[].campaignId)",
                            "won" : "EXPLODE(auctionResponseFeedbacks[].won)",
                            "winningbid" : "EXPLODE(auctionResponseFeedbacks[].winningBid)",
                            "sourcelossreasoncode" : "EXPLODE(auctionResponseFeedbacks[].sourceLossReasonCode)"
                        }
                    }
                }
            }
        }
    }
}
{
    "transform": {
        "database": "adtechdb",
        "schema": "adtech",
        "topics" : {
            "dsp.Auctions" : {
                "tables" : {
                    "auction" : {
                        "columns" : {
                            "auctionid" : "auctionId",
                            "created" : "created",
                            "bidder_nodeid" : "bidder.nodeId",
                            "deals_dealid" : "deals[].dealId"
                        }
                    },
                    "trafficSource" : {
                        "columns" : {
                            "auctionid" : "auctionId",
                            "created" : "created",
                            "sourcesellerid" : "trafficSource.sourceSellerId",
                            "domainname" : "trafficSource.domainName",
                            "contextualdataset_externalentryid" : "trafficSource.siteContextualProfile.contextualDataSet[].entries[].externalEntryId[]",
                            "dimensions_hw" : "join('x', [dimensions.height, dimensions.width])"
                        }
                    },
                    "feedback" : {
                        "columns" : {
                            "auctionid" : "auctionId",
                            "created" : "created",
                            "received" : "received",
                            "campaignid" : "EXPLODE(auctionResponseFeedbacks[].campaignId)",
                            "won" : "EXPLODE(auctionResponseFeedbacks[].won)",
                            "winningbid" : "EXPLODE(auctionResponseFeedbacks[].winningBid)",
                            "sourcelossreasoncode" : "EXPLODE(auctionResponseFeedbacks[].sourceLossReasonCode)"
                        }
                    }
                }
            }
        }
    }
}
﻿
Related Links
﻿Ingest Data with Legacy LAT Reference﻿﻿
Type:	object
Required:	Yes
Default:
Type:	string
Required:	No
Default:	null
Type:	string
Required:	No
Default:	null
Type:	string
Required:	Yes
Default:
Type:	object
Required:	Yes
Default:
Type:	string
Required:	Yes
Default:
Type:	string
Required:	Yes
Default:
LAT Extract Configuration
LAT Data Transformation with JMESPath