On 01/03/2015 10:50 AM, patrick vrijlandt wrote:
Hello list,
Let me first wish you all the best in 2015!
Today I was trying to test for occurrence of a byte in a set ...
>>> sys.version
'3.4.2 (v3.4.2:ab2c023a9432, Oct 6 2014, 22:15:05) [MSC v.1600 32 bit
(Intel)]'
>>> 'b' in 'abc'
True
>>> b'b' in b'abc'
True
>>> 'b' in set('abc')
True
>>> b'b' in set(b'abc')
False
I was surprised by the last result. What happened?
(Examples simplified; I was planning to manipulate the set)
The surprise is really that the 3rd test is True not that the fourth is
False.
First, as should be expected, a byte string is a sequence of (small)
ints. So b'b' is a (short) byte string and the set set(b'abc') is
composed of three ints. You should not expect your inclusion test to
return True when testing for a bytes-type object in a set of int-type
objects. And that explains your False result in the 4th test.
>>> type(b'abc')
<class 'bytes'>
>>> type(b'abc'[0])
<class 'int'>
But things are different for strings. You might think a string is a
sequence of characters, but Python does not have a character type. In
fact the elements of a string are just 1 char long strings:
>>> type('abc')
<class 'str'>
>>> type('abc'[0])
<class 'str'>
You would not logically expect to find a string 'b' in a set of
characters in, say C++, where the two types are different. But that's
not the Python way. In Python a set of characters set('abc') is really
a set of (short) strings, and the character 'b' is really a (short)
string, so the inclusion test works.
Python's way of returning a 1-byte string when indexing a string
(instead of returning an element of type character) allows this
surprising result.
>>> 'abc'[0]
'a'
>>> 'abc'[0][0]
'a'
>>> 'abc'[0][0][0]
'a'
>>> 'abc'[0][0][0][0]
'a'
...
I've never considered this a problem, but a infinitely indexable object
*is* a bit of an oddity.
Patrick
---
Dit e-mailbericht is gecontroleerd op virussen met Avast
antivirussoftware.
http://www.avast.com
--
Dr. Gary Herron
Department of Computer Science
DigiPen Institute of Technology
(425) 895-4418
--
https://mail.python.org/mailman/listinfo/python-list
