PostgreSQL regex Problem

[MrCBofBCinTX]

I'm using a dbi script on a database.
Someone entered some entries with extra info that I want to remove.

Found section on PostgreSQL's regex functions

I found that the following SQL works great, one by one on the entries.

SQL:
update products set product_description = regexp_replace((select product_description from products where product_description like 'Millstead%' AND product_id=19
), 'Millstead - ', '') where product_id=19;

my perl:
my $SQL = "update $updatetable set $fi3eld = regexp_replace((select $fi3eld from $updatetable where $fi3eld like \'\%$removeExp%\' AND $id_field = $id3), \'$removeExp\', '') where $id_field = $id3;";

With record abcdef,
Everything works fine for removing a complete string anywhere in record. cd->ok abc->ok def->ok, etc.

However, if I try to use ae as match, the entire string is removed. Same if I use ce as match.

I've reread the page a couple of times for this in PostgreSQL manual, but I don't understand what I need to do different.

 

[MrCBofBCinTX]

Well, I figured out part of the answer.
Turns out that works a little more like perl than manual makes clear.
using .* seems to work, so I can clear out something in the middle ok, but that still leaves me with having to review for every record change, since some fields have values like:

Millstead Dimensional Lumber
or
Millstead 2 In. x 4In. x8 Ft. Pressure Treated Lumber
etc

One record has something in between Dimensional and Lumber that I want to keep.

So Mill.*umber can't be used except as pick and choose each time.

A lot of the stuff I want to remove is every to list, but varies in details.