Recent changes to feature-requests

Suppressions on line/code/comment level

kxroberto — Fri, 25 Nov 2011 11:49:20 -0000

In practice I don't manage to get all warnings quieted/confirmed adequately and comfortably via .pycheckrc and __pychecker__ .
A mechanism is necessary to suppress locally e.g. via special comments.

Check for valid type in super calls

Thomas Heller — Fri, 02 Dec 2005 09:56:16 -0000

It would be nice if pychecker could check for the
correct type in super() calls. I common error I make
when copy and pasting is this:

class X(object):

def __init__(self):
super(Y, self).__init__()

Should warn about "%d %d %d" % (1, 2)

Andrew Bennetts — Wed, 31 Aug 2005 11:59:43 -0000

Passing the wrong number of args to string printf-style
formatting should be detected. e.g. this code should
be caught be pychecker:

def f():
"%d %d %d" % (1, 2)

This is a frustrating bug that's all too easy to do.

assert is a keyword, not a function

Toyotomi Hideyoshi — Fri, 29 Jul 2005 00:47:54 -0000

It would be nice if pychecker could warn when assert
is called with a tuple, ie. looking like a function call.

assert(exp,"message")

assert is particularly dangerous when used like this.
print will do something, assert will do nothing, ever.

Warning on other statements could still be good though.

implicit literal string concatenation

Toyotomi Hideyoshi — Sun, 17 Jul 2005 14:34:18 -0000

I found a very annoying bug in some of my code due to
a missing comma in a list defined as follows:

>>> [
... 'dir1/','dir2/'
... 'dir3/'
... ]
['dir1/', 'dir2/dir3/']

Note the missing comma are 'dir2/'

>>> 'foo/''bar'
'foo/bar'

>>> 'foo''bar'
'foobar'

It would be nice if pychecker detected one of these
cases.

identifiers list by function/method

Marcos Dione — Fri, 01 Apr 2005 14:22:46 -0000

one report pychecker could do is to list the identifiers that
appear on each method or function, may be sorted alfabetically,
so it's easier to detect typos, specially on assigments. e.g.:

def main (*unused):
factor1= 73
factor2= 736

resutl= factor1*factor2

print result

the report should look like:

main:
factor1
factor2
result
resutl
unused

it's easier to find out that resutl is the typo and not result, as it
is reported now by pychecker:

pycheck.py:7: No global (result) found

warn on dict.has_key[value]

Adrian Likins — Thu, 27 Jan 2005 20:01:51 -0000

At least for me, a pretty common typo is something
like:

dict = {}
dict['blippy'] = 'baz'

if dict.has_key['blippy']:
whatever

aka, using [] for has_key instead of proper ()

Warn about octal literals.

Simon Brunning — Fri, 14 Jan 2005 12:02:23 -0000

It would be useful if PyChecker warned you when you
specify an octal literal and where the value would
differ from what you might expect if you didn't
*realise* that you were specifying an octal literal.

x = 04 # This doesn't need a warning: 04 == 4
#x = 09 # This doesn't need a warning: it will fail to
compile
x= 012 # This *does* need a warning: 012 == 10

Can pychecker be faster? - Whitelist?

Fabio Zadrozny — Tue, 26 Oct 2004 10:36:18 -0000

Hi, I'm trying to integrate pychecker in pydev (python
plugin for eclipse), and I would like to know if there
is any way to check only a single file.

It could be some sort of a white list or something like
that, as I would be analyzing the file automatically
each time the file is saved, and pychecker gets really
slow when I do that. Some files seem to take forever to
analyze... I have some files that can take up to 5
minutes to be analyzed, as they import many things...
(Having a black list is also not useful, especially
when it is analyzed only after the results are gotten...)

Resuming, I would like to analyze everything in the
first time, after that, I would be working only with
deltas (so it could be much faster having a white list
so that only the delta is analyzed - normally only 1 or
2 files).

Can pychecker be faster? - Whitelist?

Fabio Zadrozny — Mon, 25 Oct 2004 13:08:14 -0000

Hi, I'm trying to integrate pychecker in pydev (python
plugin for eclipse), and I would like to know if there
is any way to check only a single file.