0 votes
Getting some unusual behaviour from a "Transform String" shaker. When applied to a string column, it does not successfully remove leading/trailing whitespace. When encoding the text to UTF-8 (via the shaker), I can see that there is a non-breaking whitespace encoded there.

Is this behaviour intended? What can we do to go around this?
asked by

1 Answer

0 votes
Best answer

This is the expected behavior as the Trim processor uses the classic definition of whitespaces, not the extended Unicode definition (which includes the dozens of ways there exist to represent spaces).

To remove it, you have to use a Find/Replace with a regular expression

Replace:

\p{Zs}*$

By nothing

 

This means "any number of characters of the Zs Unicode character class immediately followed by the end of the string ($)". To learn more about Unicode character classes, see https://en.wikipedia.org/wiki/Unicode_character_property 

answered by
944 questions
978 answers
989 comments
2,153 users

┬ęDataiku 2012-2018 - Privacy Policy