Terry J. Reedy <[EMAIL PROTECTED]> added the comment: I changed this to a doc issue for 2.6/3.0 whenever.
I have two objections to adding "An empty deque evaluates as false". First, it implies (falsely) that it could be otherwise; since deque has no __bool__ method, its __len__ method is used, so that bool(d) == (len(d)!=0). Second, it misses better doc enhancements that might make the statement I just made clearer and easier to find. 1. Ref manual Expressions Boolean Operations says "In the context of Boolean operations, and also when expressions are used by control flow statements, the following values are interpreted as false: False, None, numeric zero of all types, and empty strings and containers (including strings, tuples, lists, dictionaries, sets and frozensets)." For 3.0, I suggest replacing "and empty strings..." with "empty strings and sequences (including strings, bytes, bytearrays, tuples, lists, and Userlists and deques from the collections module), and other empty containers (sets, frozensets, dictionaries, and Userdicts and defaultdicts from the collections module)." Anything else I forgot? Adjust for 2.5/6. The sentence after next "User-defined objects can customize their truth value by providing a __bool__() method." should say '... __bool__ or __len__ method.', with __len__ linked to object.__len__ just as __bool__ is linked to object.__bool__. 2. The LibRef entry for built-in function bool says simply "Convert a value to a Boolean, using the standard truth testing procedure". Extended that with " described in the Language reference in the __bool__ and __len__ entries of the Special methods subsection and in the Boolean operations subsection." ---------- assignee: rhettinger -> georg.brandl components: +Documentation -Extension Modules, Library (Lib) nosy: +georg.brandl, tjreedy versions: +Python 2.6, Python 3.0 -Python 2.7, Python 3.1 _______________________________________ Python tracker <[EMAIL PROTECTED]> <http://bugs.python.org/issue3891> _______________________________________ _______________________________________________ Python-bugs-list mailing list Unsubscribe: http://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
