LookupJoin

Short Description

LookupJoin is a general purpose joiner. It merges potentially unsorted records from one data source incoming through the single input port with another data source from a lookup table based on a common key.

COMPONENT	SAME INPUT METADATA	SORTED INPUTS	SLAVE INPUTS	OUTPUTS	OUTPUT FOR DRIVER WITHOUT SLAVE	OUTPUT FOR SLAVES WITHOUT DRIVER	JOINING BASED ON EQUALITY	AUTO-PROPAGATED METADATA
LookupJoin	x	x	1 (virtual)	1-2	✓	x	✓	✓

Ports

The joined data is then sent to the first output port.
The second output port can optionally be used to capture unmatched master records.

PORT TYPE	NUMBER	REQUIRED	DESCRIPTION	METADATA
Input	0	✓	Master input port	Any
	1 (virtual)	✓	Slave input port	Any
Output	0	✓	Output port for the joined data	Any
	1	x	Optional port for master data records without slave matches. (Only if the Join type attribute is set to `Inner join`.) This applies only to LookupJoin and DBJoin.	Input 0

Metadata

LookupJoin propagates metadata from the first input port to the second output port and from the second output port to the first input port. The propagation does not change the priority of metadata.
LookupJoin has no metadata template.
Either data source (input port and lookup table) may potentially have a different metadata structure.

LookupJoin Attributes

ATTRIBUTE	REQ	DESCRIPTION	POSSIBLE VALUES
BASIC
Join key	yes	A key according to which incoming data flows are joined. See Join Key below.
Left outer join		If set to `true`, also driver records without corresponding slave are parsed. Otherwise, `inner join` is performed.	false (default) \| true
Lookup table	yes	ID of the lookup table to be used as the resource of slave records. Number of lookup key fields and their data types must be the same as those of Join key. These fields values are compared and matched records are joined.
Transform	[bl	Transformation in CTL or Java defined in the graph.
Transform URL	[bl	External file defining the transformation in CTL or Java.
Transform class	[bl	External transformation class.
Transform source charset		Encoding of the external file defining the transformation. The default encoding depends on DEFAULT_SOURCE_CODE_CHARSET in defaultProperties.	E.g. UTF-8
ADVANCED
Clear lookup table after finishing		When set to `true`, memory caches of the lookup table will be emptied at the end of the execution of this component. This has different effects on different lookup table types. Simple lookup table and Range lookup table will contain 0 entries after this operation. For the other lookup table types, this will only erase cached data and therefore make more memory available, but the lookup table will still contain the same entries.	false (default) \| true
DEPRECATED
Error actions		Definition of the action that should be performed when the specified transformation returns an Error code. See Return Values of Transformations.
Error log		URL of the file to which error messages for specified Error actions should be written. If not set, they are written to Console.

[1] One of these must be set. These transformation attributes must be specified. Any of these transformation attributes must use a common CTL template for Joiners or implement a RecordTransform interface.

Details

LookupJoin is a general purpose joiner used in most common situations. It does not require the input to be sorted and is very fast as it is processed in memory.

The data attached to the first input port is called the master, the second data source is called the slave. Its data is considered as if it were coming through the second (virtual) input port. Each master record is matched to the slave record on one or more fields known as the join key. The output is produced by applying a transformation which maps joined inputs to the output.

Slave data is pulled out from a lookup table, so depending on the lookup table the data can be stored in the memory. This also depends on the lookup table type, e.g. Database lookup stores only the values which have already been queried. Master data is not stored in the memory.

Join Key
Join key is a sequence of field names from the input metadata separated by a semicolon, colon or pipe. You can define the key in the Edit key wizard.

A counterpart of this Join key of the input metadata is the key of lookup table in lookup tables. It is specified in the lookup table itself.

Example 45. Join Key for LookupJoin

$first_name;$last_name

This is the master part of fields that should serve to join master records with slave records.

Lookup key should look like this:

$fname;$lname

Corresponding fields will be compared and matching values will serve to join master and slave records.

Key Duplicates in Lookup Table
If the lookup table allows key duplicates, more output records can be created from a single input record.

CTL Interface
All Joiners share the same transformation template which can be found in CTL Templates for Joiners.

Java Interfaces
If you define your transformation in Java, it must implement the following interface that is common for all Joiners: Java Interfaces for Joiners.

Examples

Enrichment of Records Using Data from Lookup Table
Given a list of number of customers for particular year per country with metadata fields CountryCode, Customers and Year.

CZ |341 |2013
CZE|417 |2014
DE |2573|2013
DE |3791|2014
SK |457 |2014
...

Replace the country code by country name. The list of country codes and corresponding country names is available from lookup table CountryCodeLookup.

CZ |Czech Republic
DE |Germany
SK |Slovak Republic
...

Solution
Use the attributes Join Key, Lookup Table and Transform.

ATTRIBUTE	VALUE
Join Key	CountryCode
Lookup Table	CountryCodeLookup
Transform	See the code below

function integer transform() {
    $out.0.Customers = $in.0.Customers;
    $out.0.Country = $in.1.CountryName;
    $out.0.Year = $in.0.Year;

    return ALL;
}

Values found in the lookup table are mapped in the same way as if they came from the second input port.
The result records are

Czech Republic |341 |2013
Germany        |2573|2013
Germany        |3791|2014
Slovak Republic|457 |2014

The country code CZE has not been found in the lookup table, so it has been sent unchanged to the second output port if an edge is connected.

Matching Ranges with Range Lookup Table
This example shows usage of https://docs.primeur.com/docs/types-of-lookup-tables#range-lookup-table table in LookupJoin.

Records on the first data stream contains groups of accounts. Each group of accounts is defined by the lowest and highest account numbers, e.g. 12300 and 12399.

Data on the second data stream contains account numbers. Match the accounts with groups.

Solution
Load the records containing account ranges with LookupTableReaderWriter into Range Lookup Table.

In the next phase, use LookupJoin to match the records from the second data stream.

In LookupJoin set Join Key, Lookup Table, and Transform.

ATTRIBUTE	VALUE
Join Key	accountNumber
Lookup Table	RangeLookupTable0
Transform	Map the fields that are necessary

Note
Matching account into the ranges depends on the data type. If the account ranges and account number are specified as a whole number (integer/long), the records are compared as numbers. If the account ranges and account number are specified as a string, the records are compared as strings.

If account numbers are integers (or longs) and the range is from 10 to 50, account 200 is out of the range.

If account numbers are strings and the range is from 10 to 50, account 200 is within the range.

Best Practices

If the transformation is specified in an external file (with Transform URL), we recommend users to explicitly specify Transform source charset.