Martin von Loewis loewis at informatik.hu-berlin.de
Mon Sep 25 17:36:07 EDT 2000

• Previous message (by thread): Version mismatches of .PYD extensions crash python...
• Next message (by thread): Lazy coerce function?
• Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

---

Treutwein Guido <Guido.Treutwein at nbg.siemens.de> writes:

> The documentation is a little bit on the short side, specifying, that
> numbers are casted to compatible types.
> 
> Python 1.6 reacts as follows:
> coerce(3,'a') # raises a TypeError, as expected.
> coerce('a', 'b') # returns ('a', 'b')
> coerce('a', []) # raises TypeError
> 
> Is there any simple rule under what circumstances coerce raises an
> exception?

The results of coerce are typically passed to arithmetic
operations. If there is some sensible way to convert the arguments
that the results support arithmethic operators, the conversion is
performed - otherwise you get a type error.

A good candidate operator is '+': if plus 'ought to' work on the
operands, coerce will return objects that can be added. That's why
coercing strings and lists does something nontrivial.

Regards,
Martin

---

• Previous message (by thread): Version mismatches of .PYD extensions crash python...
• Next message (by thread): Lazy coerce function?
• Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

---

More information about the Python-list mailing list