SQL Text Filtering

Question

I want to filter out several key phrases out of a dataset. Unfortunately, the only algorithm I've been able to come up with thus far is nested replace statements, such as:

SELECT 
    REPLACE(
      REPLACE(FIELDNAME,'</u>','')
      ,'<u>','')

where FIELDNAME is raw HTML code stored in a table. As you can see, this is hideous. Any better ideas?

Answer 1

I don't think there's a better way in TSQL.

If you've got another environment on top of the SQL layer (e.g asp.net) you might have more luck doing the filtering in that.

Answer 2

Such string manipulations are best handled by CLR scalar valued functions.