RetroSearch Browse

Mon Apr 16 03:50:14 EDT 2001 · http://mail.python.org/pipermail/python-list/2001-April/099606.html

"Ken Peek" <Peek at LVCM.comNOSPAM> wrote in message
news:tdku4kqlqjquf2 at corp.supernews.com...
    [snip]
> Here is my question--  If Python WERE to have a "do...until" construct,
what
> would be the best way to implement the syntactical rules?

It would be best to have one general guarded-loop construct implementing
the Knuthian "N and 1/2 times" loop, rather than two different constructs
each implementing only a special case of it.  Exact syntax sugar could be
endlessly (and fruitlessly) debated, but the general idea might be:

    loop:
        [suite 1]
    while [condition]:
        [suite 2]

The 'loop' keyword and associated [suite 1] would be optional, so that just
by omitting this clause one would get back to today's while-loop.  It is
highly debatable whether [suite 2] could be made optional (together with
the trailing colon on the while clause) -- it might look nice but I think it
would run counter to all of Python's regular syntax (colons are never
optional; where a suite is undesired one writes the pass statement).

In today's Python syntax, this general loop can be expressed with exactly
this Knuthian semantic, BUT the guard clause must be indented the same
as the guarded suites in the loop body -- that would be the main difference.
(Curiously, [suite 2] in today's syntax DOES turn out to be optional!).

Minor differences, strictly sugary ones, are on the spelling of the
clauses --
today, 'loop:' is spelled 'while 1:' and 'while condition:' is spelled
'if not condition: break'.

Alex

Home - News ( United States | United Kingdom | Italy | Germany ) - Football scores

Showing content from http://mail.python.org/pipermail/python-list/2001-April/099606.html below:

do...until wisdom needed...