Complex evaluation bug

[of]

a = 1+3j
complex(str(a))

Why does this not work ? It should

[Brian Blazer]

I am not exactly sure what is going on, but I get the error:

ValueError: complex() arg is a malformed string

I think that it might be because the value of 'j' is not defined.

But I am a newbie so I could very well be wrong.

Brian Blazer
br***@brianandkate.com


On May 18, 2006, at 11:36 AM, of wrote:

a = 1+3j
complex(str(a))

Why does this not work ? It should
--
http://mail.python.org/mailman/listinfo/python-list

[Gary Herron]

of wrote:

a = 1+3j
complex(str(a))

Why does this not work ? It should

Says who?

By normal conventions in Python, "str" attempts only to make a "nice"
human readable representation. The function "repr" is usually expected
to provide output that can be parsed back into the original object.
(Although for the numeric complex type the two produce identical results.)

Further, constructors are rarely expected to parse a string
representation to return an object. The function "eval" is usually
expected to provide that functionality.

So, putting them together, you could expect
eval(repr(a))
to reproduce a, and in fact it does so.

Gary Herron

[wittempj]

py> c = complex(1, 3)
py> print c
(1+3j)
py> d = complex('1+3j')
py> print d
(1+3j)
py> str(1+3j)
'(1+3j)'
complex takes two numbers, or a string representing a complex number.
the string you supply isn't a representation of valid complex number.

[Mel Wilson]

of wrote:

a = 1+3j
complex(str(a))

Why does this not work ? It should

It would be nice.
Looks like str(1+3j) is returning an expression in string
form. Maybe there is no actual complex literal.

eval (str(1+3j)) works.

Python 2.4.2 (#1, Jan 23 2006, 21:24:54) [GCC 3.3.4] on linux2
Type "help", "copyright", "credits" or "license" for more
information.

1+3j (1+3j) str(1+3j) '(1+3j)' complex (str(1+3j)) Traceback (most recent call last):
File "<stdin>", line 1, in ?
ValueError: complex() arg is a malformed string eval (str(1+3j)) (1+3j) eval(str(1+3j))**2

(-8+6j)

[Paul McGuire]

"Mel Wilson" <mw********@sympatico.ca> wrote in message
news:Yk*******************@news20.bellglobal.com.. .

of wrote:

a = 1+3j
complex(str(a))

Why does this not work ? It should

It would be nice.
Looks like str(1+3j) is returning an expression in string
form. Maybe there is no actual complex literal.

The problem is that str(1+3j) emits a string enclosed in parens. Stripping
them off makes the string acceptable to complex's constructor, without
invoking eval.

a = 1+3j
str(a) '(1+3j)' complex(str(a)) Traceback (most recent call last):
File "<stdin>", line 1, in ?
ValueError: complex() arg is a malformed string complex(str(a)[1:-1]) (1+3j) complex(str(a).strip("()"))

(1+3j)
-- Paul

[Christophe]

Gary Herron a écrit :

of wrote:

a = 1+3j
complex(str(a))

Why does this not work ? It should

Says who?
By normal conventions in Python, "str" attempts only to make a "nice"
human readable representation. The function "repr" is usually expected
to provide output that can be parsed back into the original object.
(Although for the numeric complex type the two produce identical results.)

Further, constructors are rarely expected to parse a string
representation to return an object. The function "eval" is usually
expected to provide that functionality.

So, putting them together, you could expect
eval(repr(a))
to reproduce a, and in fact it does so.

Says who ?

Python 2.4.2 (#67, Sep 28 2005, 12:41:11) [MSC v.1310 32 bit (Intel)] on
win32
Type "help", "copyright", "credits" or "license" for more information.

repr(1+3j) '(1+3j)' complex(repr(1+3j)) Traceback (most recent call last):
File "<stdin>", line 1, in ?
ValueError: complex() arg is a malformed string

[Fredrik Lundh]

Christophe wrote:

So, putting them together, you could expect
eval(repr(a))
to reproduce a, and in fact it does so.

Says who ?

Python 2.4.2 (#67, Sep 28 2005, 12:41:11) [MSC v.1310 32 bit (Intel)] on
win32
Type "help", "copyright", "credits" or "license" for more information.

>>> repr(1+3j) '(1+3j)' >>> complex(repr(1+3j)) Traceback (most recent call last):
File "<stdin>", line 1, in ?
ValueError: complex() arg is a malformed string >>>

in what language is "eval" spelled "complex" ?

</F>

[Christophe]

Fredrik Lundh a écrit :

Christophe wrote:

So, putting them together, you could expect
eval(repr(a))
to reproduce a, and in fact it does so.

Says who ?

Python 2.4.2 (#67, Sep 28 2005, 12:41:11) [MSC v.1310 32 bit (Intel)]
on win32
Type "help", "copyright", "credits" or "license" for more information.

>>> repr(1+3j)

'(1+3j)'

>>> complex(repr(1+3j))

Traceback (most recent call last):
File "<stdin>", line 1, in ?
ValueError: complex() arg is a malformed string

>>>

in what language is "eval" spelled "complex" ?

</F>

Oups, I was too fast to read what was written. I though you only changed
one of the terms ( str -> repr ).

You'll note that repr and str produce the exact same result for complex.
And also, I'm not sure using eval anywhere is a good idea so it probably
doesn't help for what the OP wants to do.

[Paul McGuire]

"Christophe" <ch*************@free.fr> wrote in message
news:44**********************@news.free.fr...
<snip>

Oups, I was too fast to read what was written. I though you only changed
one of the terms ( str -> repr ).

You'll note that repr and str produce the exact same result for complex.
And also, I'm not sure using eval anywhere is a good idea so it probably
doesn't help for what the OP wants to do.

An eval-less approach - the problem is the enclosing parens.

a = 1+3j
str(a) '(1+3j)' complex(str(a)) Traceback (most recent call last):
File "<stdin>", line 1, in ?
ValueError: complex() arg is a malformed string complex(str(a)[1:-1]) (1+3j) complex(str(a).strip("()"))

(1+3j)

-- Paul

[Heiko Wundram]

Am Freitag 19 Mai 2006 18:03 schrieb Paul McGuire:

An eval-less approach - the problem is the enclosing parens.
<snip>

I've just submitted two patches to the Python bugtracker at:

http://sourceforge.net/tracker/index...70&atid=305470

which either change the repr() format (removing the parentheses), which I find
doubtful, because it's not backwards-compatible, or alter the constructor to
accept the repr() format for complex numbers (a bracketed number).

Feel free to comment.

--- Heiko.

[of]

Heiko Wundram wrote:

Am Freitag 19 Mai 2006 18:03 schrieb Paul McGuire:

An eval-less approach - the problem is the enclosing parens.
<snip>

I've just submitted two patches to the Python bugtracker at:

http://sourceforge.net/tracker/index...70&atid=305470

which either change the repr() format (removing the parentheses), which I find
doubtful, because it's not backwards-compatible, or alter the constructor to
accept the repr() format for complex numbers (a bracketed number).

Feel free to comment.

--- Heiko.

thanks