[ILUG] Perl guru advice needed

[Kevin Philp]

A quick question for any perl guru's out there.

We have a small programme that downloads data from a website and dumps the 
data in a MySQL database. The website contains a lot of hex coded entities 
such as:

® in place of ® and others

We use a programme that is based on the LWP module and when it downloads the 
text with the get_text it automatically decodes all entities. Unfortunately 
it makes a complete pigs ear of the whole thing so:

using decode_entities: ® = ®

using get_trimmed_text: ® = î

I would prefer not to decode the entities at all but looking at the module 
reference you can't switch it off for get_text.

As anyone any suggestions to get around this or an algorithm to convert the 
corrupt text back to something sensible?

Thanks

Kevin.