Pentaho Product: Data Integration
Operating system: Any
The attached transformations display an issue when trying to replace backslash
and double-quote characters using the "replace in string" step.
The data grid step has two fields defined - "value" and "shouldbe".
The value field simulates a string returned from a database that contains backslashes
and double quotes. The "shouldbe" field is how the field should look after the
backslashes and double quotes have been correctly escaped.
For example, the value returned as b"\ when escaped correctly should look like b\"
The "Search" field in the "replace in string" step uses regexp logic to identify the
backslashes and double quotes and should precede them with the escape character (another
backslash), only nothing is changed. This is verified by previewing the "filter rows"
The regexps used have been checked on online regexp utilities (such as http://www.regexr.com) so I believe them to be correct.
Is it possible to investigate this problem in order to:
- have consistent regexp handling
- Avoid having to use scripting steps to handle replacing characters
- improve performance