Re^3: Subclassing a class that uses an internal dispatch table

We already have one. It's called the Perl dispatcher.

    # Sanitize $cmd here if you want. Maybe you go against a list of a
+cceptable
    # commands that's accessible via $self->add_command( 'cmd1' );
    my $method = $self->can( $cmd );
    $self->$method( @args ) if $method;
[download]

Is there something I'm missing? This is very similar to what SQL::Parser does and it's pretty successful.

My criteria for good software:

Does it work?
Can someone else come in, make a change, and be reasonably certain no bugs were introduced?

Comment on Re^3: Subclassing a class that uses an internal dispatch table Download Code

Replies are listed 'Best First'.
Re^4: Subclassing a class that uses an internal dispatch table by BrowserUk (Patriarch) on Oct 30, 2007 at 21:16 UTC
But isn't that breaking one of the cardinal rules? Examine what is said, not who speaks -- Silence betokens consent -- Love the truth but pardon error. "Science is about questioning the status quo. Questioning authority". In the absence of evidence, opinion is indistinguishable from prejudice. "Too many [] have been sedated by an oppressive environment of political correctness and risk aversion."	[reply]
Re^5: Subclassing a class that uses an internal dispatch table by dragonchild (Archbishop) on Oct 30, 2007 at 22:57 UTC
Nope. Think of it as a hash lookup that is a huge set of hashes looked at in the proper order. My criteria for good software: Does it work? Can someone else come in, make a change, and be reasonably certain no bugs were introduced?	[reply]
Re^6: Subclassing a class that uses an internal dispatch table by BrowserUk (Patriarch) on Oct 30, 2007 at 23:28 UTC
Hm. That could be said about any code that relies upon the use of symbolic references. Perl's symbol tables are hashes, and using a symbolic reference is simply looking up the 'name' in a one of those hashes. But still, the primary argument against using symbolic references hold true. If someone sends you a correctly formed cmd message that contains a cmd that coincides with a method name that is not a part of the interface, that private method could be invoked by a third party. I appreciate that you suggested in your comments that received command could be sanatised against a list of those that are a legitimate part of the interface, but the same could be done for any symbolic reference usage. I guess it just reenforces the idea that best practices and even cardinal rules are just guidlines who's purpose should be to suggest caution, not to prohibit absolutely. Upshot: I agree with you that this would be a far better way of implementing this requirement. Examine what is said, not who speaks -- Silence betokens consent -- Love the truth but pardon error. "Science is about questioning the status quo. Questioning authority". In the absence of evidence, opinion is indistinguishable from prejudice. "Too many [] have been sedated by an oppressive environment of political correctness and risk aversion."	[reply]