develooper Front page | perl.perl5.porters | Postings from September 2011

eval "Unicode" vs eval "bytes" (Re: Status of UTF-8 work)

Thread Next
From:
Father Chrysostomos
Date:
September 18, 2011 16:56
Subject:
eval "Unicode" vs eval "bytes" (Re: Status of UTF-8 work)
Message ID:
6708D4AB-B9E3-4D77-8A9F-8BE9F92FF098@cpan.org

On Sep 11, 2011, at 4:36 PM, Brian Fraser wrote:

> The only thing which comes to mind right now, that could be both heinously wrong and take a bit longer to fix, is the commit that dealt with eval having different behaviors for the same string, depending on whenever it had been upgraded from Latin-1 or not.

I’ve just remembered something I needed to bring up: Aristotle pointed out that it is currently possible to put a file in a database, and then retrieve it and run it through eval(). Your change would break that. I think we need an evalbytes function for fallback (as Aristotle suggested), *and* a feature.pm pragma to enable Unicode eval(). (The same pragma could enable both.)

(If we are going to do evalbytes, we might as well make it work with source filters as well.)

I’m forwarding this message to p5p.


Thread Next


nntp.perl.org: Perl Programming lists via nntp and http.
Comments to Ask Bjørn Hansen at ask@perl.org | Group listing | About