Odd behaviour on importing variable from module

[rs387]

Hi,

I've found the following behaviour on importing a variable from a
module somewhat odd. The behaviour is identical in Python 2.5 and
3.0b2.

In summary, here's what happens. I have a module, oddmodule.py
(below), that defines a variable, OddVariable, by assigning a value A
to it. The file I execute, mainfile.py, imports and re-binds
OddVariable to a value B. I have two test modules which import the
OddVariable in two different ways, one via "import oddmodule" and
another via "from oddmodule import OddVariable".

The weird behaviour is that by importing using the former syntax, I
can see OddVariable bound to B (as expected), but the latter syntax
sees it bound A.

Is this the intended behaviour? Am I misunderstanding what is going on
here?


Source code:

<<>>

print("Importing oddmodule")
OddVariable = ["not", "initialized"]

def OddFunction():
print("OddVariable from oddmodule.OddFunction:", OddVariable)


<<>>

import oddmodule
import testmodule1
import testmodule2

print("Initializing OddVariable")

oddmodule.OddVariable = ["some", "list"]

print()
testmodule2.DoTest()
print()
testmodule1.DoTest()


<<>>

from oddmodule import OddVariable, OddFunction

def DoTest():
print("STARTING testmodule1.DoTest")
print("OddVariable from testmodule1:", OddVariable)
OddFunction()
print("FINISHED testmodule1.DoTest")


<<>>

import oddmodule

def DoTest():
print("STARTING testmodule2.DoTest")
print("OddVariable from testmodule2:", oddmodule.OddVariable)
oddmodule.OddFunction()
print("FINISHED testmodule2.DoTest")



OUTPUT:

Importing oddmodule
Initializing OddVariable

STARTING testmodule2.DoTest
OddVariable from testmodule2: ['some', 'list']
OddVariable from oddmodule.OddFunction: ['some', 'list']
FINISHED testmodule2.DoTest

STARTING testmodule1.DoTest
OddVariable from testmodule1: ['not', 'initialized']  !!! OLD
VALUE !!!
OddVariable from oddmodule.OddFunction: ['some', 'list']
FINISHED testmodule1.DoTest


Many thanks,

Roman
--
http://mail.python.org/mailman/listinfo/python-list

Re: Odd behaviour on importing variable from module

[rs387]

> the construct
>
>  from oddmodule import OddVariable, OddFunction
>
> assigns the *values* of the given module names to new variables in the
> importing module's namespace.  that is, you're binding new names to the
> values the variables happen to have when the from-import statement is
> executed.

Ah right, this does indeed explain what I'm seeing. For some reason I
thought "from module import variable" magically makes "variable" an
"alias", so to speak, for "module.var", which of course it does not.

> Yes. Think of
>
> from module import var
>
> as a shortcut for
>
> import module
> var = module.var
> del module

Good tip, I'll remember this. Thanks very much Fredrik, Peter!

Roman
--
http://mail.python.org/mailman/listinfo/python-list

Stuck connection in Python 3.0b2 http.server

[rs387]

Hi All

I've encountered a weird issue when migrating a web server to Python 3
- the browser would wait forever without showing a page, displaying
"Transferring data" in the status bar. I tracked it down to a
reference cycle in my BaseHTTPRequestHandler descendant - one of the
attributes stored a dict of methods. Removing the cycle made the
problem go away.

In Python 2.5.2 the code works fine either way.

Here's a minimal example which runs in both 2.5 and 3.0 - to see stuck
connections run as-is in 3.0 and navigate to http://localhost:8123; to
fix this comment out "self.dummy = self" (alternatively reset
self.dummy = None at the end of the __init__ method).

Am I doing it wrong, or is this a bug?


try:
import http.server
httpmodule = http.server
except:
import BaseHTTPServer
httpmodule = BaseHTTPServer

class BreakageRequest(httpmodule.BaseHTTPRequestHandler):
def __init__(self, request, client_address, server):
self.dummy = self # COMMENT THIS OUT to see the connection
become unstuck
httpmodule.BaseHTTPRequestHandler.__init__(self, request,
client_address, server)
def do_GET(self):
self.send_response(200)
self.send_header("Content-Type", "application/xml")
self.end_headers()
self.wfile.write("".encode('utf-8'))

srv = httpmodule.HTTPServer(('', 8123), BreakageRequest)
srv.serve_forever()
--
http://mail.python.org/mailman/listinfo/python-list

Re: How to emit Cyrillic and Chinese via unicode from console mode?

[rs387]

On Sep 14, 2:03 am, "Siegfried Heintze" <[EMAIL PROTECTED]> wrote:
> Can someone point me to an example of a little program that emits non-ascii
> Unicode characters (Russian or Chinese perhaps)?

The following doesn't quite work, but I'll post it anyway since it
actually ends up printing the characters. Perhaps someone can point
out what causes the exception at the end?

The important thing is to set the console codepage to 65001, which is
UTF-8. This lets you output utf8-encoded text and see the Unicode
chars displayed.



import sys
import encodings.utf_8
import win32console

sys.stdout = encodings.utf_8.StreamWriter(sys.stdout)

win32console.SetConsoleCP(65001)
win32console.SetConsoleOutputCP(65001)

s = "English: ok\n"
s += u'Russian: \u0420\u0443\u0441\u0441\u043a\u0438\u0439\n'
s += u'Greek: \u03bc\u03b5\u03b3\u03b1\u03bb\u03cd
\u03c4\u03b5\u03c1\u03b7\n'

print s



If redirected to file, all is well, this prints everything properly in
UTF-8. If ran on the console, this also prints everything correctly,
but then throws a mysterious exception:

English: ok
Russian: Русский
Greek: μεγαλύτερη
Traceback (most recent call last):
  File "I:\Temp\utf8console.py", line 18, in 
print s
  File "C:\Progs\Python25\lib\codecs.py", line 304, in write
self.stream.write(data)
IOError: [Errno 0] Error

Any ideas?

Roman

P.S. This really ought to Just Work in this day and age, and do so
without all those 65001/utf8 incantations. Pity that it doesn't. Sigh.
--
http://mail.python.org/mailman/listinfo/python-list

Re: recursion gotcha?

[rs387]

On Sep 14, 9:01 am, cnb <[EMAIL PROTECTED]> wrote:
> def suma(xs, acc=0):
>         if len(xs) == 0:
>                 acc
>         else:
>                 suma(xs[1:], acc+xs[0])
>
> it returns none.

Yep, that's because there is no "return" statement anywhere. Python
doesn't return expressions "by default", like functional languages do,
so where you say "suma(xs[1:], acc+xs[0])" this just calls itself and
returns nothing.

Try this:

def suma(xs, acc=0):
if len(xs) == 0:
return acc
else:
return suma(xs[1:], acc+xs[0])

print suma([1, 2, 3, 4, 5])

(prints 15)

Roman
--
http://mail.python.org/mailman/listinfo/python-list

Re: How to emit Cyrillic and Chinese via unicode from console mode?

[rs387]

On Sep 14, 11:51 am, Gertjan Klein <[EMAIL PROTECTED]> wrote:
> Interesting. On my system (Windows XP) the console codepage does not
> change, and hence the characters don't print properly (I get some of the
> CP437 line drawing characters, for example). I have never been able to
> convince windows to assume/support UTF-8 encoding in the console,
> programatically or otherwise. :(

I found that a useful test is to create a directory whose name
contains chars from various languages, then run "cmd" and see if "dir"
displays them correctly. This worked fine for me in WinXP, even though
my system locale is Russian. Would be interesting to know if you can
get the console to display international chars this way.
--
http://mail.python.org/mailman/listinfo/python-list

Add vs in-place add of str to list

[rs387]

Hi

I'm trying to understand why it is that I can do

>>> a = []
>>> a += 'stuff'
>>> a
['s', 't', 'u', 'f', 'f']

but not

>>> a = []
>>> a = a + 'stuff'
Traceback (most recent call last):
  File "", line 1, in 
TypeError: can only concatenate list (not "str") to list

Can someone explain the logic? Why is the in-place add not a type
error but the usual add is? (This applies to both Python 2.6rc1 and
3.0b2)

Thanks
Roman
--
http://mail.python.org/mailman/listinfo/python-list

Re: Add vs in-place add of str to list

[rs387]

On Oct 2, 8:11 am, Erik Max Francis <[EMAIL PROTECTED]> wrote:
> It's because the `+=` operator is doing the equivalent of calling the
> `extend` method, which treats its argument as a generic sequence, and
> doesn't enforce type.

I see. Do you know whether this is seen as a problem with the language
design?

If so, this could be fixed by changing list.__add__ to call .extend on
the left argument if it's a list... Or, by changing __iadd__ to throw
a TypeError, whichever way around is seen as more correct/desirable.
--
http://mail.python.org/mailman/listinfo/python-list

Re: Add vs in-place add of str to list

[rs387]

On Oct 2, 3:50 pm, Mel <[EMAIL PROTECTED]> wrote:
> rs387 wrote:
> > I see. Do you know whether this is seen as a problem with the language
> > design?
>
> No.

OK, I get it now. I was assuming that the "+" could be implemented in
terms of "+=" as follows:

def add(x,y):
temp = list(x)
temp += y
return temp

But it turns out that this is not what "+" means at all. A bit of a
gotcha...

Roman

--
http://mail.python.org/mailman/listinfo/python-list