erroneousBollock hit it with the "COLLATE" option on the query, though for your purposes, a query like this would probably do best:

"SELECT rec,id from memrecs where rec collate latin1_bin rlike '[A-Z]'
+" );
[download]

Then again, MySQL does provide exactly what you need -- the "LOWER" function -- so you can skip the initial query and just do the update:

$dbh->do( "UPDATE memrec set rec=LOWER(rec)" );
[download]

Or, you could just apply the LOWER() function whenever you query the table to get "rec" values (SELECT LOWER(rec) ...) and not worry about whether or not there are capital letters in that column of the database.

A few other points you should know:

if ( $_rec =~ /A .. Z/) {
    $_rec =~ tr/A-Z/a-z/g;
    ...
}
[download]

That's the wrong way to check for upper case letters -- you would want  $_rec =~ /[A-Z]/ instead. Also, the "g" modifier does not apply to the tr/// operator (study the section about tr/// in perlop for details on its modifiers).

But the regex match is unnecessary anyway. The two steps can be simplified to just this:

if ( $_rec =~ tr/A-Z/a-z/ ) {
    # tr/// returns the number of characters converted
    # so we get here only if there had been at least one A-Z
}
[download]