On 4/14/07, Christian N. [email protected] wrote:
“Robert D.” [email protected] writes:
You see it is not the isolated semantics of x=x, but the context of
other semantics that disturbs me (well disturbs me just a tiny little
bit ;).
So you are proposing all undefined locals return nil too? }}
No I am too young to die, because that means we need python call syntax
for the implicit receiver, maybe you want some parens too ;).
Beware I am writing this in the best intent.
Ok fooled around enough.
The real problem is the implicit declaration of variables of course.
I feel that it should be hidden which means it should take place after
the RHS evaluation, but in practice it takes place before.
Well I can life with that
Cheers
Robert
“Robert D.” [email protected] writes:
Beware I am writing this in the best intent.
Ok fooled around enough.
The real problem is the implicit declaration of variables of course.
I feel that it should be hidden which means it should take place after
the RHS evaluation, but in practice it takes place before.
Well I can life with that
In my five years of using Ruby, this has bitten me maybe twice.
I think I misspelled “initialize” a lot more. 
On 4/15/07, Christian N. [email protected] wrote:
I think I misspelled “initialize” a lot more.
That for sure, that was a great PITN at first!!
Cheers
Robert
–
Christian N. [email protected] http://chneukirchen.org
Robert
On 4/14/07, Christian N. [email protected] wrote:
a does exist at the assignment and initialized as nil.
It’s very philosophical:
$ ruby -e ‘thing = thing; p thing’
nil
That might be to my liking too if there were not
======================
501/1 > ruby -e ‘@x=@x;p @x’
nil
--------------------------->
503/3 > ruby -e ‘p @x’
nil
504/4 > ruby -e ‘x=x;p x’
nil
-------------------------->
505/5 > ruby -e ‘p x’
-e:1: undefined local variable or method `x’ for main:Object (NameError)
You see it is not the isolated semantics of x=x, but the context of
other semantics that disturbs me (well disturbs me just a tiny little
bit ;).
Cheers
Robert
assuming a is already defined
class Object; def yield; yield(self) end end
func(b).yield{|fb| a = fb if fb}
On Apr 14, 10:49 am, Christian N. [email protected]
wrote:
“Robert D.” [email protected] writes:
You see it is not the isolated semantics of x=x, but the context of
other semantics that disturbs me (well disturbs me just a tiny little
bit ;).
So you are proposing all undefined locals return nil too? }}
–
Christian N. [email protected] http://chneukirchen.org
Perhaps Kernel.variable_missing?
Or I’m just crazy…
On 4/15/07, CHubas [email protected] wrote:
Christian N. [email protected] http://chneukirchen.org
Perhaps Kernel.variable_missing?
Or I’m just crazy…
I hope not ;), but I think this will enter in conflict with
method_missing.
Robert
On Apr 15, 3:28 pm, “Robert D.” [email protected] wrote:
So you are proposing all undefined locals return nil too? }}
Perhaps Kernel.variable_missing?
Or I’m just crazy…
I hope not ;), but I think this will enter in conflict with method_missing.
Robert
I think it’s not a preposterous (translation dictionaries are silly)
idea after all, if handled well. Say, method_missing would have more
precedence than variable_missing.
As for now, method missing is worth a try
def method_missing
nil
end
foo
=> nil
x ||= 10
=> 10
x
=> 10
but this is obviously wicked. Specially with typos and lexical
mistakes. I think it’s worse to get bad results from bad valued than
get an error and make things right instead.
Oh, and it doesn’t handle constants.
FOO
=> NameError: uninitialized constant OK
from (irb):32
from :0
On Apr 13, 1:02 am, Jonathan [email protected] wrote:
Is there a cool way to do this without calling the function twice?:
a = func(b) unless func(b).nil? (AKA)
a = func(b) if func(b)
greg wrote:
assuming a is already defined
class Object; def yield; yield(self) end end
func(b).yield{|fb| a = fb if fb}
Shades of forth:
class Object; def if?; yield(self) if self end end
a = 1
“fred wilma”[/fred/].if?{|x| a = x}
p a # ==> “fred”
“fred wilma”[/barney/].if?{|x| a = x}
p a # ==> “fred”
Btw, I think this construct has appeared before, but I can’t seem to
recall what it was called…