On Mon, Mar 17, 2014 at 9:38 PM, Christophe Bal <projet...@gmail.com> wrote: > Here is the translation. ;-) > > Hello, > and what about something like that ? > > a @ b @ c -> (a @ b) @ c > a * b @ c -> (a * b) @ c > a @ b * c -> a @ (b * c) > > Easy to remember: the *-product has priority regarding to the @-product, and > we just do @-product from left to right.
In the terminology we've been using in this thread, this is "weak-left". > An advantage of this is that most parsers do analyze from left to right. > > So I really think that it is a better choice than the weak-right one. We've mostly ignored this option because of assuming that if we want left-associativity, we should go with "same-left" instead of "weak-left". Same-left is: a @ b @ c -> (a @ b) @ c a * b @ c -> (a * b) @ c a @ b * c -> (a @ b) * c i.e., even more left-to-right than weak-left :-) Do you think weak-left is better than same-left? -- Nathaniel J. Smith Postdoctoral researcher - Informatics - University of Edinburgh http://vorpus.org _______________________________________________ NumPy-Discussion mailing list NumPy-Discussion@scipy.org http://mail.scipy.org/mailman/listinfo/numpy-discussion