PEP-xxx: Unification of for statement and list-comp syntax

[Edward Elliott]

George Sakkis wrote:

> Em Dom, 2006-05-21 às 17:11 +0200, Heiko Wundram escreveu:
>> for node in tree if node.haschildren():
>> <do something with node>
>> 
>> as syntactic sugar for:
>> 
>> for node in tree:
>> if not node.haschildren():
>> continue
>> <do something with node>
[snip]
>
> 2) "There should be one and preferably only one way to do it."

You mean like this:

s = "foo" + "bar"
s = 'foo' + 'bar'
s = 'foo' 'bar'
s = '%s%s' % ('foo', 'bar')

This one and only one way stuff is overrated.  I don't care how many ways
there are as long as:
1. at least one way is intuitive
2. every way is easily comprehendible (readable, no side effects, etc)

Now 'one way to do it' is a good mantra for keeping language complexity down
and achieving the second goal.  But it needs to be flexible.  The proposed
syntax is short, intuitive, and consistent with comprehension/generator
syntax.  I'd say it's entirely in keeping with a number of rules which
trump 'one way':

Beautiful is better than ugly.
Flat is better than nested.
Readability counts.
Special cases aren't special enough to break the rules. (proposal eliminates
the current special case for comprehensions/generators)

-- 
Edward Elliott
UC Berkeley School of Law (Boalt Hall)
complangpython at eddeye dot net