ch.brickwork.bsetl.sanitize

Class DbSanitizer

java.lang.Object

ch.brickwork.bsetl.sanitize.DbSanitizer

All Implemented Interfaces:

Preprocessor

public class DbSanitizer
extends Object
implements Preprocessor

A Preprocessor tailored to sanitize values within rows. The preprocessor can be added to the target database, causing inserts of any value covered in the sanitizer configuration to be obfuscated or replaced

Via the configuration, the DBSanitizer applies a series of defined @see ch.brickwork.bsetl.sanitize.ValueSanitizer to do the job

Some ValueSanitizers have a row context allowing for various properties (e.g. name, email) to be populated with different but connected values (e.g. marcel - marcel@camporelli.ch instead of marcel - foo@foo.com; to this end, a mapping function must be provided returning the value of a unique id for the row, like so:

     this.obTarget.addPreprocessor(new DbSanitizer(sanitizerConfig, map  ->
         Objects.toString(map.get("Id")).replaceAll("null", "")
             .concat(Objects.toString(map.get("id")).replaceAll("null", ""))));
   

The id map is per Db, not per table. However, if different tables have different ids, concatenation of any available/required id property will usually work (all "unused" id properties will deliver null).

Constructor Summary

Constructors

Constructor and Description
DbSanitizer(SanitizerConfiguration sanitizerConfiguration) If rows can't be identified; means you will not be able to use contextual value sanitizers; if each row in your table has a unique key (may be composed), it is preferrable to use DbSanitizer(config, function)
DbSanitizer(SanitizerConfiguration sanitizerConfiguration, Function<Map<String,Object>,String> getIdFromMap) If each of your rows can be identified with a unique identifier (may be composed), this is the best constructor to use; if not, use the simple constructor.

Method Summary

Modifier and Type	Method and Description
Set<String>	getUntouchedColumns()
Map<String,Object>	preprocessRow(String tableName, Map<String,Object> inputRow)

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

DbSanitizer

public DbSanitizer(SanitizerConfiguration sanitizerConfiguration)

If rows can't be identified; means you will not be able to use contextual value sanitizers; if each row in your table has a unique key (may be composed), it is preferrable to use DbSanitizer(config, function)

Parameters:

sanitizerConfiguration -

DbSanitizer

public DbSanitizer(SanitizerConfiguration sanitizerConfiguration,
                   Function<Map<String,Object>,String> getIdFromMap)

If each of your rows can be identified with a unique identifier (may be composed), this is the best constructor to use; if not, use the simple constructor. If your table has a column which is unique, the getIdFromMap function is as simple as

map  -> map.get("id")

Parameters:

sanitizerConfiguration -

getIdFromMap - function mapping a row (Map of properties) to a String which is unique across all rows

Method Detail

preprocessRow

public Map<String,Object> preprocessRow(String tableName,
                                        Map<String,Object> inputRow)

Specified by:

preprocessRow in interface Preprocessor

getUntouchedColumns

public Set<String> getUntouchedColumns()