__imul__ broken for 'objects' in 2.2.1
Alexander Schmolck
a.schmolck at gmx.net
Mon Jun 10 09:47:06 EDT 2002
More information about the Python-list mailing list
Mon Jun 10 09:47:06 EDT 2002
- Previous message (by thread): __imul__ broken for 'objects' in 2.2.1
- Next message (by thread): __imul__ broken for 'objects' in 2.2.1
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Christos Georgiou <DLNXPEGFQVEB at spammotel.com> writes: > On 07 Jun 2002 18:17:45 +0100, rumours say that Alexander Schmolck > <a.schmolck at gmx.net> might have written: > > >class Breaky(object): > > def __imul__(self, other): > > print "imuling" > > return self > >sq = Breaky() > >sq *=1. > > > >gives: > > > >Traceback (most recent call last):[...] line 10, in ? > > sq *=1. > >TypeError: can't multiply sequence to non-int > > Just in case: it is indeed your intention to multiply the object with a > float, right? > Indeed it is and this nice little bug basically means that the matrix class I've been working on for quite some time is now somewhat screwed (I need a new-style class *and* inplace operations). > PS It seems that a Breaky instance is not (incorrectly) considered a > sequence in 2.3a0 (20020603). Nice to hear, but only a limited consolation since it quite some time will pass before 2.3 comes out and is sufficiently widely adopted. I really wonder where this bizzarre behavior comes from -- it is really none of the interpreter's business to take guesses whether my class is a sequence or not -- finding out if there is a problem with the supplied arguments should be left to the discretion of the method itself. alex
- Previous message (by thread): __imul__ broken for 'objects' in 2.2.1
- Next message (by thread): __imul__ broken for 'objects' in 2.2.1
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
More information about the Python-list mailing list