This question has been asked several times in the very recent past. See these previous discussions.

• PERL UNIX and strangeness and char conversion
• Substitute and converting to UTF8
• convert tags to punctuation

There seems little point in going through it all again.

I just posted this 2d ago:

" I ran across this string in a document: \xE2\x80\x94 I did tr/\x94/"/, and got no result. (Should have replaced \x94). What is the above code/string and how do I get rid of it? "

Is there a document that tells how to fix/convert this?

I just posted this 2d ago

There is a reply to that node with links to similar threads. And in one of those threads I wrote this: Could you post a representative sample of your input file, following the advice given here and here? In particular, show us a representative sample of your input using hexdump -C file or od -tx1c file, pasted here inside <code> tags.

If you don't show us your actual data, we will be left guessing, and you won't get any good answers. Please help us to help you. See also How do I post a question effectively?, I know what I mean. Why don't you?, and Short, Self-Contained, Correct Example.

I think you might be looking for the solution others provided me just recently for almost the exact same issue. In my case, the trick turned out to be a two-step process. You will find the details to the solution here:

Perl's encoding versus UTF8 octets

The crux of it was this:

$mytext =~ s!\\x(..)!chr(hex($1))!ge;

my $newcode = decode('utf8', $mytext);
[download]

Note the substitution followed by the decode.

Blessings,

~Polyglot~