On Tue, Sep 01, 2015 at 11:26:27AM -0400, Robert Haas wrote: > On Tue, Sep 1, 2015 at 9:58 AM, Andrew Dunstan <[email protected]> wrote: > > On 08/31/2015 11:57 PM, Peter Eisentraut wrote: > >> We now have 80+ Perl files in our tree, and it's growing. Some of those > >> files were originally written for Perl 4, and the coding styles and > >> quality are quite, uh, divergent. So I figured it's time to clean up > >> that code a bit. I ran perlcritic over the tree and cleaned up all the > >> warnings at level 5 (the default, least severe). > > > > I don't object to this. Forcing strict mode is good, and I think I stopped > > using bareword file handles around 17 years ago. > > FWIW, I think perlcritic is both useless and annoying. I've always > used bareword file handles, and I don't really see what the problem > with it is, especially in very short script files.
A bareword file handle is a form of global variable, so the criticism is helpful for codebases large enough to make those a maintenance problem. It's important for any CPAN module, so I can understand perlcritic including it. Plenty of uses in our tree are fine. > And what's wrong > with two-argument form of open, if the path is a constant rather than > possibly-tainted user input? Perl advertises that TMTOWTDI, and then > perlcritic complains about which one you picked, mostly AFAICS for no > particularly compelling reason. So I'm pretty meh about this whole > exercise, especially if we follow it up by cleaning up the lower > levels of warnings which, from what I can see, are unnecessary > pedantry on top of unnecessary pedantry. > > But I suspect I'm in the minority here, so feel free to ignore me. (I > certainly do agree that use strict and use warnings are a good thing > to use everywhere. It's just perlcritic I dislike.) I agree with the rest of your message. -- Sent via pgsql-hackers mailing list ([email protected]) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-hackers
