The error message is pretty straightforward; the replace () has to have the target, the old string and the new string.

You really need to clean up the data instead of trying to kludge it into shape. I am sorry you have only one client; I hope business gets better. Did you mean "clientele" or some other collective or plural noun to show this is a set, and not a single record in the file system? Does "city" actually mean "city_name" or is there another attribute property that you left off (look at the ISO 11179 naming rules)? Also, there is no such thing as a generic description; it has to be of something in particular.

In a correctly normalized table, each distinctive attribute will have its own column. This means that the ZIP Code ought to be in a column declared as "zip_code CHAR(5) NOT NULL CHECK (zip_code LIKE'[0-9][0-9][0-9][0-9][0-9]')" so that you have some data integrity. Likewise, the city name should be in its own column and it needs a two-letter state code. Bit of trivia: the reason The Simpsons cartoon characters live in Springfield is that there are so many Springfields in the United States. This trick was originally used on an old TV show called "Father Knows Best" and became part of TV lore 😎

You will probably want to get a CASS validation tool to make sure the city, state and zip codes are all correct. Clean it up in the front and then put it in the database.