[issue9136] RuntimeError when profiling Decimal

Mark Dickinson added the comment: Fixed in revisions r82654 through r82657. Thanks everyone for the input. -- resolution: -> fixed stage: commit review -> committed/rejected status: open -> closed ___ Python tracker

[issue9136] RuntimeError when profiling Decimal

Stefan Krah added the comment: Alexander, for me the ternary expressions have the advantage that they take up less vertical space and you have to keep less state in mind. Mark, I'm sidestepping the dict syntax question and reassign to you :) -- assignee: skrah -> mark.dickinson _

[issue9136] RuntimeError when profiling Decimal

Mark Dickinson added the comment: > I don't quite understand the point of catching NameError here So that the initialization of DefaultContext itself doesn't fail. -- ___ Python tracker ___

[issue9136] RuntimeError when profiling Decimal

Alexander Belopolsky added the comment: +try: +dc = DefaultContext +except NameError: +pass + +self.prec = dc.prec if prec is None else prec I don't quite understand the point of catching NameError here, but it looks like if it is caught, it will

[issue9136] RuntimeError when profiling Decimal

Changes by Stefan Krah : Added file: http://bugs.python.org/file17837/issue9136-trunk-3.patch ___ Python tracker ___ ___ Python-bugs-list maili

[issue9136] RuntimeError when profiling Decimal

Stefan Krah added the comment: In the new patches I reinstated the ternary construct. I think it would be nice to use dict comprehensions in py3k. People read the stdlib for guidance, and it would help if ultimately only the most concise idioms remained in py3k. If I'm not mistaken, quite often

[issue9136] RuntimeError when profiling Decimal

Changes by Stefan Krah : Removed file: http://bugs.python.org/file17831/unnamed ___ Python tracker ___ ___ Python-bugs-list mailing list Unsubs

[issue9136] RuntimeError when profiling Decimal

Mark Dickinson added the comment: 2.5 compatibility is fine for the 2.x version of decimal.py. For the 3.x version of decimal.py, I don't see what it buys us. -- ___ Python tracker

[issue9136] RuntimeError when profiling Decimal

Raymond Hettinger added the comment: > I also don't think that the 2.x-to-3.x maintenance issue > is that big a deal any more; it would be surprising if > the 2.x version of Decimal sees anything more than minor > changes (doc fixes, etc.) from this point on. So perhaps > the 3.x version o

[issue9136] RuntimeError when profiling Decimal

Yuv Gre added the comment: Now that we're on the subject of "from_float", I just recalled this slight issue: Python 3.1.2 (r312:79149, Mar 21 2010, 00:41:52) [MSC v.1500 32 bit (Intel)] on win32 Type "help", "copyright", "credits" or "license" for more information. >>> >>> import decimal >>> x

[issue9136] RuntimeError when profiling Decimal

Mark Dickinson added the comment: > ISTM, 2.5 compatible is probably good enough. Okay; that's fine with me. Now we can finally turn that "from_float = classmethod(from_float)" into a proper "@classmethod" decorator. :) I also don't think that the 2.x-to-3.x maintenance issue is that big a

[issue9136] RuntimeError when profiling Decimal

Raymond Hettinger added the comment: I'm not sure that I care about 2.3 compatibility anymore (I was the one who made the rule about 2.3 compatibility and made sure that it was just a preference, not an absolute rule -- as time goes on, it is of less and less value). ISTM, 2.5 compatible is

[issue9136] RuntimeError when profiling Decimal

Mark Dickinson added the comment: [Raymond] > self.Emin = DefaultContext.Emin if Emin is None else Emin I agree that looks better. But we can't spell it that way in 2.x (since conditional expressions aren't 2.3 compatible), and I'd prefer to keep the 2.x and 3.x versions reasonably close, at

[issue9136] RuntimeError when profiling Decimal

Raymond Hettinger added the comment: Stylistically, it would be nice to eliminate the local variable reassignment entirely: self.Emin = DefaultContext.Emin if Emin is None else Emin self.Emax = DefaultContext.Emax if Emax is None else Emax self._ignored_flags = [] if _ignored_flags is None els

[issue9136] RuntimeError when profiling Decimal

Mark Dickinson added the comment: Looks good to me. Please go ahead and apply it to the 3.2 and 3.1 branches, if you want. (Or just assign back to me if you prefer.) It should also be applied to 2.7 and 2.6, eventually, but we should probably wait until after the 2.7 release this weekend fo

[issue9136] RuntimeError when profiling Decimal

Stefan Krah added the comment: Mark, good point about 2.3 compatibility. The unit tests diverge quite a bit though between 2.5, 2.6 and 2.7. I like {s: 0 for s in _signals} slightly better here (but I like dict/list comprehensions in general). So, the new patch still sets the flags/traps sect

[issue9136] RuntimeError when profiling Decimal

Mark Dickinson added the comment: Stefan, one other thought: please also bear in mind that we're restricted to Python 2.3 syntax in the 2.x version of decimal, so any post-Python 2.3-isms will have to be rewritten when the patch is backported. (E.g., use of conditional expressions.) While I

[issue9136] RuntimeError when profiling Decimal

Mark Dickinson added the comment: Nice shade of blue! Just a couple of red spots that I'd prefer repainted, namely: (1) please put bodies of 'try:' and 'except:' on separate lines, and (2) please use 'except NameError' instead of a bare except. And a couple of points that I think also applie

[issue9136] RuntimeError when profiling Decimal

Stefan Krah added the comment: Mark, the patch looks good. I don't find it ugly, but anyway, here's a bike shed version. :) The main point is that I'd like the flags and traps sections to be set apart visually while compressing the boilerplate. -- Added file: http://bugs.python.org/fil

[issue9136] RuntimeError when profiling Decimal

Mark Dickinson added the comment: Here's a patch. It's a little ugly, but I don't see a better way that avoids locals(). Using **kwargs is an option, but would mean abandoning the current nice feature that unexpected keyword arguments raise TypeError (or alternatively, implementing that che

[issue9136] RuntimeError when profiling Decimal

Mark Dickinson added the comment: > it looks like 'locals()' is somewhat magical ... and Objects/frameobject.c is the place to go for a full understanding. PyFrame_FastToLocals is the 'updating' function that updates the locals dict from the 'fast locals' object. -- ___

[issue9136] RuntimeError when profiling Decimal

Mark Dickinson added the comment: Ah, it looks like 'locals()' is somewhat magical. Its docstring says: "Update and return a dictionary containing the current scope's local variables." So I think this explains your (Stefan's) results: in either case, you evaluate locals() (as the target of

[issue9136] RuntimeError when profiling Decimal

Stefan Krah added the comment: The tracker doesn't handle code when posted by mail. Here's the code again: >>> for name, val in locals().items(): ... print(locals()) ... {'name': '__builtins__', 'val': , '__builtins__': , '__package__': None, '__name__': '__main__', '__doc__': None} Trac

[issue9136] RuntimeError when profiling Decimal

Stefan Krah added the comment: Mark Dickinson wrote: > > Mark Dickinson added the comment: > > class Context(object): > def __init__(self): > for name, val in locals().items(): > setattr(self, name, val) > > > Isn't this dodgy anyway, since 'name' and 'val' end up g

[issue9136] RuntimeError when profiling Decimal

Stefan Krah added the comment: Still clueless about profile.py, but I think it creates a parallel stack, and overwriting of locals legitimately occurs in that stack, producing the Exception. So it would appear that by design one simply cannot iterate over locals when using the profile module.

[issue9136] RuntimeError when profiling Decimal

Mark Dickinson added the comment: class Context(object): def __init__(self): for name, val in locals().items(): setattr(self, name, val) Isn't this dodgy anyway, since 'name' and 'val' end up going into locals()? I wonder why the RuntimeError *isn't* raised for a norm

[issue9136] RuntimeError when profiling Decimal

Mark Dickinson added the comment: Removing the decimal module from the equation, the following is enough to trigger this for me. Stefan's suggestion about the profile module writing to locals sounds right on target. import profile class Context(object): def __init__(self): for

[issue9136] RuntimeError when profiling Decimal

Stefan Krah added the comment: Mark Dickinson wrote: > I can't reproduce this on Linux. However, I've seen an informal report of > something like this happening before. It looks a lot like something > threading related, but I don't see any threads in what you're doing. (Though > I don't k

[issue9136] RuntimeError when profiling Decimal

Mark Dickinson added the comment: Okay, I can reproduce by adding a 'time.sleep(0.01)' delay into the body of the 'for name, val in locals().items():' loop. I then get (with py3k): dicki...@alberti:~/Source/py3k> ./python Python 3.2a0 (py3k:82413M, Jul 1 2010, 10:21:02) [GCC 4.2.1 (SUSE Linu

[issue9136] RuntimeError when profiling Decimal

Mark Dickinson added the comment: BTW, is the behaviour consistent, or does it only occur on some runs? -- ___ Python tracker ___ ___

[issue9136] RuntimeError when profiling Decimal

Mark Dickinson added the comment: I can't reproduce this on Linux. However, I've seen an informal report of something like this happening before. It looks a lot like something threading related, but I don't see any threads in what you're doing. (Though I don't know the details of how the p

[issue9136] RuntimeError when profiling Decimal

Changes by Stefan Krah : -- nosy: +skrah ___ Python tracker ___ ___ Python-bugs-list mailing list Unsubscribe: http://mail.python.org/

[issue9136] RuntimeError when profiling Decimal

Changes by Ezio Melotti : -- nosy: +ezio.melotti, mark.dickinson stage: -> unit test needed ___ Python tracker ___ ___ Python-bugs-lis

[issue9136] RuntimeError when profiling Decimal

New submission from Yuv Gre : I'm using Python 3.1.2 (r312:79149, Mar 21 2010, 00:41:52) [MSC v.1500 32 bit (Intel)] on win32. Running the following code: import profile import math import decimal def show_bug(): x = decimal.Decimal.from_float(math.e)**(-2**31