Use Cython directive binding=True to get signatures for cython methods
kwankyu opened this issue · 164 comments
In requests for help for cythonized built-in methods, the signature of the method is not shown, unlike normal python methods. For an example,
sage: a=17
sage: a.quo_rem?
Docstring:
Returns the quotient and the remainder of self divided by other.
Note that the remainder returned is always either zero or of the
same sign as other.
INPUT:
* "other" - the divisor
OUTPUT:
* "q" - the quotient of self/other
* "r" - the remainder of self/other
EXAMPLES:
sage: z = Integer(231)
sage: z.quo_rem(2)
(115, 1)
...
To fix this, we set Cython directive binding=True. Thus we buy
- better support of introspection into cython methods,
- consistent help messages for cython methods like python methods,
- better tracebacks on exceptions,
- better behaving for documentation
for slight performance degradation due to increased overhead cost of calling cython methods.
Related tickets: #19100, #20860, #18192
Depends on #32509
Depends on #33864
CC: @jdemeyer @tscrim @mkoeppe
Component: user interface
Author: Kwankyu Lee, Tobias Diez
Branch/Commit: public/26254 @ 326f19c
Reviewer: Tobias Diez, ...
Issue created by migration from https://trac.sagemath.org/ticket/26254
It seems this file
https://github.com/ipython/ipython/blob/master/IPython/core/oinspect.py
is responsible for this issue. For me, it would take some time to scrutinize what this does though.
A fix is to redefine IPython.core.oinspect.getargspec to use sage.misc.sageinspect.sage_getargspec
Replying to @kwankyu:
A fix is to redefine
IPython.core.oinspect.getargspecto usesage.misc.sageinspect.sage_getargspec
This is already done in sage.repl.ipython_extension.init_inspector. But apparently with no effect, strangely.
It turns out that the problem is with IPython.core.oinspect.inspector._get_def, which calls Python's inspect.signature via IPython.utils.signatures module.
This problem is nothing to do with sage_getargspec.
We may just wait for future sage based on Python 3 with inspect.signature supporting cython.
I don't see the signature in a Python 3 build of Sage, either.
Replying to @jhpalmieri:
I don't see the signature in a Python 3 build of Sage, either.
Right.
I don't remember exactly what I meant in my last comment. Perhaps I expected Cython someday support the new signature module shipped with Python 3. Now I don't have any clear idea what should be done on what side.
Branch: public/26254
Based on this discussion:
and consulting:
I made the last commit. Please checkout and try.
Author: Kwankyu Lee
With your patch I get a bunch of doctest errors, of the kind
sage -t --warn-long 55.8 src/sage/graphs/strongly_regular_db.pyx
**********************************************************************
File "src/sage/graphs/strongly_regular_db.pyx", line 1156, in sage.graphs.strongly_regular_db.is_RSHCD
Failed example:
t = is_RSHCD(64,27,10,12); t
Expected:
[<built-in function SRG_from_RSHCD>, 64, 27, 10, 12]
Got:
[<cyfunction SRG_from_RSHCD at 0x7f8616d09890>, 64, 27, 10, 12]
**********************************************************************
sage -t --warn-long 55.8 src/sage/misc/latex.py
**********************************************************************
File "src/sage/misc/latex.py", line 561, in sage.misc.latex.has_latex_attr
Failed example:
T._latex_()
Expected:
Traceback (most recent call last):
...
TypeError: descriptor '_latex_' of 'sage.matrix.matrix0.Matrix' object needs an argument
Got:
<BLANKLINE>
Traceback (most recent call last):
File "/home/scratch2/dimpase/sage/sage/local/lib/python2.7/site-packages/sage/doctest/forker.py", line 681, in _run
self.compile_and_execute(example, compiler, test.globs)
File "/home/scratch2/dimpase/sage/sage/local/lib/python2.7/site-packages/sage/doctest/forker.py", line 1105, in compile_and_execute
exec(compiled, globs)
File "<doctest sage.misc.latex.has_latex_attr[5]>", line 1, in <module>
T._latex_()
TypeError: unbound method cython_function_or_method object must be called with Matrix_integer_dense instance as first argument (got nothing instead)
--- this is of course not a surpise, but it needs to be fixed on this ticket.
Otherwise, I like it - e.g. notice how much more informative the error messages are.
This comes with the penalty of producing a wrapped object every time a method is accessed on a cython object. I suspect cythonized access avoids that, so it may be that in most scenarios this doesn't come with a performance penalty, but one should check carefully that sage doesn't rely on the situations where it does.
Also, there is a reason why cython.binding==False by default: that's the behaviour built-in methods exhibit: [].insert returns a built-in method insert of list object ... rather than a bound method, and cython by default does the same. If you want more informative tracemacks, wouldn't it be better to solve it in such a way that straight-up CPython (and its C extension classes; of which cython is a special case) also benefit?
Replying to @nbruin:
This comes with the penalty of producing a wrapped object every time a method is accessed on a cython object. I suspect cythonized access avoids that, so it may be that in most scenarios this doesn't come with a performance penalty, but one should check carefully that sage doesn't rely on the situations where it does.
Also, there is a reason why
cython.binding==Falseby default: that's the behaviour built-in methods exhibit:[].insertreturns abuilt-in method insert of list object ...rather than abound method, and cython by default does the same.
[].insert? shows correct signature. So built-in methods can behave well with respect to introspection. Then why cythonized built-in methods do not? How can we make cythonized built-in methos behave well like standard built-in methods of python?
[].insert?shows correct signature. So built-in methods can behave well with respect to introspection. Then why cythonized built-in methods do not? How can we make cythonized built-in methods behave well like standard built-in methods of python?
An answer can be found here:
and
https://docs.python.org/3/howto/clinic.html
Now it seems to me that cython should do a better job in making cythonized built-ins more introspectable.
To summarize the current situation, there are two options:
Option 1: We accept the current patch, which turns on cython directive "binding=True" so that all cythonized methods become bound methods that already support the inspect.signature module well. If we take this path, then there is nothing for us to do except fixing a few doctests.
Option 2: We wait for upstream cython fixes that will make all cythonized built-in methods properly support the inspect.signature module. This is the path that standard built-in methods follow. We don't know when the upstream fix would be available.
Please give your preference and why.
To go with option 1, we need benchmarking results on whether it affects the performance a lot.
Replying to @dimpase:
To go with option 1, we need benchmarking results on whether it affects the performance a lot.
If it affects any bit of the runtime performance in other aspect than introspection, then option 1 should be discarded. I think this should be decided not by experiments but analysis of how python and cython works.
We can configure this in build time, to begin with. It is helpful for debugging - I would not care about a 5% or 15% performance hit, if error messages made more sense.
Replying to @dimpase:
We can configure this in build time, to begin with. It is helpful for debugging - I would not care about a 5% or 15% performance hit, if error messages made more sense.
I would.
This is what Jeroen has been working on for like, literally the last year, perhaps longer :)
Yes, the solution is to use binding=True to enable use of cyfunctions. However, using cyfunctions across the board can introduce a significant performance penalty in many cases, as the Python interpreter has some built-in optimizations for built-in functions that don't work for cyfunctions.
Jeroen has been fighting for a series of PEPs that would overhaul Python's function type hierarchy in such a way that the basic function type can be extended (e.g. as with Cython's cyfunction) while still keeping those optimizations working.
So while this seems like it should be an easy problem to solve, it's completely non-trivial.
Point being, let's not duplicate effort here.
Replying to @embray:
Point being, let's not duplicate effort here.
Thanks for the expert advice.
Ticket retargeted after milestone closed
Moving tickets to milestone sage-9.2 based on a review of last modification date, branch status, and severity.
Replying to @embray:
This is what Jeroen has been working on for like, literally the last year, perhaps longer :)
Jeroen has been fighting for a series of PEPs that would overhaul Python's function type hierarchy in such a way that the basic
functiontype can be extended (e.g. as with Cython'scyfunction) while still keeping those optimizations working.
I searched for these PEPs, and reached to
I am curious if and how theses PEPs would eventually solve the problem of this ticket. I only guess that after the PEPs made into CPython, Cython is updated to use the new CPython features, and then the signature issue in Sage is automatically fixed. Am I right?
It would be interesting to know whether the upcoming Cython 3 (#29863) has improvements in this direction
Replying to @kwankyu:
Replying to @embray:
This is what Jeroen has been working on for like, literally the last year, perhaps longer :)
Jeroen has been fighting for a series of PEPs that would overhaul Python's function type hierarchy in such a way that the basic
functiontype can be extended (e.g. as with Cython'scyfunction) while still keeping those optimizations working.I searched for these PEPs, and reached to
I am curious if and how theses PEPs would eventually solve the problem of this ticket. I only guess that after the PEPs made into CPython, Cython is updated to use the new CPython features, and then the signature issue in Sage is automatically fixed. Am I right?
That's correct--this would allow us to use Cython's own function subclass, which includes support for better signature documentation among other things, without losing any performance.
Setting new milestone based on a cursory review of ticket status, priority, and last modification date.
Setting a new milestone for this ticket based on a cursory review.
binding=True will be the default in Cython 3 (https://cython.readthedocs.io/en/latest/src/userguide/source_files_and_compilation.html#compiler-directives), so eventually we will make this switch anyway; so why not now.
Branch pushed to git repo; I updated commit sha1. This was a forced push. New commits:
f36275a | Turn on cython directive binding |
Rebased on 9.5.beta0
I have set it to "needs review" so that the patchbot runs on it.
Are there performance penalties for doing this, perhaps that Cython 3 is going to address?
Branch pushed to git repo; I updated commit sha1. New commits:
36ec493 | Reformat the comment |
Replying to @jhpalmieri:
Are there performance penalties for doing this...?
How can we see the performance penalty?
Replying to @kwankyu:
Replying to @jhpalmieri:
Are there performance penalties for doing this...?
How can we see the performance penalty?
Try builds with and without and compare some timings?
Replying to @jhpalmieri:
Replying to @kwankyu:
Replying to @jhpalmieri:
Are there performance penalties for doing this...?
How can we see the performance penalty?
Try builds with and without and compare some timings?
I tried a very simple script like: timeit('a=17;a.quo_rem(5); del a'), and find no difference.
I wonder what is a proper way to see the difference...
Replying to @kwankyu:
Replying to @jhpalmieri:
Replying to @kwankyu:
Replying to @jhpalmieri:
Are there performance penalties for doing this...?
How can we see the performance penalty?
Try builds with and without and compare some timings?
I tried a very simple script like:
timeit('a=17;a.quo_rem(5); del a'), and find no difference.I wonder what is a proper way to see the difference...
I ran ./sage -t --long src/sage/matrix/*.pyx a few times:
Develop: average time 83.3 seconds.
This ticket: average time 93.1 seconds.
I also tried ./sage -t --long src/sage/matrix/matrix_gfpn_dense.pyx a few times:
Develop: average time 28.4 seconds
This ticket: average time 37.0 seconds
Some other files in matrix showed very little difference, so maybe the slowdown only occurs in certain types of operations.
From:
https://cython.readthedocs.io/en/latest/src/userguide/source_files_and_compilation.html
... When enabled, functions will bind to an instance when looked up as a class attribute
I don't know what triggers the binding behaviour, but I imagine there may be a code path that runs into this and perhaps ends up not binding anyway (thus creating overhead) or ends up binding in a way that was previously done in a more efficient way (cached perhaps?)
The timings above show the impact can be quite significant: I think too high a penalty to incur in general. Note that the documentation also says:
Changed in version 3.0.0: Default changed from False to True
so figuring out what's causing the slowdown is a prereq to upgrading to 3.0.0 (once that finally is released). If there's a particular scenario where it's bad to have binding, we might just be able to turn it off in those cases.
What's the status here? The performance issues don't seem to be too bad, especially since they apparently only affect certain functions/modules.
Binding=true is also required for #30884 since the decorator library internally uses inspection on the decorated function. So if one wants to decorate cython functions, then they have to be "bound".
Replying to @tobiasdiez:
What's the status here? The performance issues don't seem to be too bad, especially since they apparently only affect certain functions/modules.
I agree with comment:48. The penalty seems significant to me. We need to know how much damage we would get and where, and to see if there is a way to reduce the damage.
Branch pushed to git repo; I updated commit sha1. New commits:
52b5089 | Merge remote-tracking branch 'origin/develop' into public/26254 |
After merging the latest dev branch, this now throws
File "/__w/sagetrac-mirror/sagetrac-mirror/src/sage/all_cmdline.py", line 19, in <module>
from sage.all import *
File "/__w/sagetrac-mirror/sagetrac-mirror/src/sage/all.py", line 135, in <module>
from sage.symbolic.all import *
File "/__w/sagetrac-mirror/sagetrac-mirror/src/sage/symbolic/all.py", line 8, in <module>
import sage.symbolic.expression # initialize pynac before .ring
File "sage/symbolic/expression.pyx", line 1, in init sage.symbolic.expression
# -*- coding: utf-8 -*-
File "sage/symbolic/function.pyx", line 144, in init sage.symbolic.function
from .expression import (
ImportError: cannot import name call_registered_function
when importing sage.all, see eg github build workflow. Any idea on how to fix it?
This method was introduced in #32386 (or one of its dependencies).
Branch pushed to git repo; I updated commit sha1. New commits:
0517de6 | Merge remote-tracking branch 'origin/develop' into public/26254 |
I tried to investigate this is a bit, but didn't get far.
With bindings=true, the expression.cpp file contains the new lines
...
static PyMethodDef __pyx_mdef_4sage_8symbolic_10expression_121call_registered_function = {"call_registered_function", (PyCFunction)(void*)(PyCFunctionWithKeywords)__pyx_pw_4sage_8symbolic_10expression_121call_registered_function, METH_VARARGS|METH_KEYWORDS, __pyx_doc_4sage_8symbolic_10expression_120call_registered_function};
....
static PyMethodDef __pyx_mdef_4sage_8symbolic_10expression_123find_registered_function = {"find_registered_function", (PyCFunction)(void*)(PyCFunctionWithKeywords)__pyx_pw_4sage_8symbolic_10expression_123find_registered_function, METH_VARARGS|METH_KEYWORDS, __pyx_doc_4sage_8symbolic_10expression_122find_registered_function};
....
However, it no longer contains
static PyMethodDef __pyx_methods[] = {
{"is_Expression", (PyCFunction)__pyx_pw_4sage_8symbolic_10expression_79is_Expression, METH_O, __pyx_doc_4sage_8symbolic_10expression_78is_Expression},
{"is_SymbolicEquation", (PyCFunction)__pyx_pw_4sage_8symbolic_10expression_81is_SymbolicEquation, METH_O, __pyx_doc_4sage_8symbolic_10expression_80is_SymbolicEquation},
{"_is_SymbolicVariable", (PyCFunction)__pyx_pw_4sage_8symbolic_10expression_83_is_SymbolicVariable, METH_O, __pyx_doc_4sage_8symbolic_10expression_82_is_SymbolicVariable},
{"_repr_Expression", (PyCFunction)__pyx_pw_4sage_8symbolic_10expression_91_repr_Expression, METH_O, __pyx_doc_4sage_8symbolic_10expression_90_repr_Expression},
{"_latex_Expression", (PyCFunction)__pyx_pw_4sage_8symbolic_10expression_93_latex_Expression, METH_O, __pyx_doc_4sage_8symbolic_10expression_92_latex_Expression},
{"new_Expression", (PyCFunction)(void*)(PyCFunctionWithKeywords)__pyx_pw_4sage_8symbolic_10expression_99new_Expression, METH_VARARGS|METH_KEYWORDS, __pyx_doc_4sage_8symbolic_10expression_98new_Expression},
{"new_Expression_from_pyobject", (PyCFunction)(void*)(PyCFunctionWithKeywords)__pyx_pw_4sage_8symbolic_10expression_101new_Expression_from_pyobject, METH_VARARGS|METH_KEYWORDS, __pyx_doc_4sage_8symbolic_10expression_100new_Expression_from_pyobject},
{"new_Expression_wild", (PyCFunction)(void*)(PyCFunctionWithKeywords)__pyx_pw_4sage_8symbolic_10expression_103new_Expression_wild, METH_VARARGS|METH_KEYWORDS, __pyx_doc_4sage_8symbolic_10expression_102new_Expression_wild},
{"new_Expression_symbol", (PyCFunction)(void*)(PyCFunctionWithKeywords)__pyx_pw_4sage_8symbolic_10expression_105new_Expression_symbol, METH_VARARGS|METH_KEYWORDS, __pyx_doc_4sage_8symbolic_10expression_104new_Expression_symbol},
{"print_order", (PyCFunction)(void*)(PyCFunctionWithKeywords)__pyx_pw_4sage_8symbolic_10expression_107print_order, METH_VARARGS|METH_KEYWORDS, __pyx_doc_4sage_8symbolic_10expression_106print_order},
{"print_sorted", (PyCFunction)__pyx_pw_4sage_8symbolic_10expression_109print_sorted, METH_O, __pyx_doc_4sage_8symbolic_10expression_108print_sorted},
{"math_sorted", (PyCFunction)__pyx_pw_4sage_8symbolic_10expression_111math_sorted, METH_O, __pyx_doc_4sage_8symbolic_10expression_110math_sorted},
{"mixed_order", (PyCFunction)(void*)(PyCFunctionWithKeywords)__pyx_pw_4sage_8symbolic_10expression_113mixed_order, METH_VARARGS|METH_KEYWORDS, __pyx_doc_4sage_8symbolic_10expression_112mixed_order},
{"mixed_sorted", (PyCFunction)__pyx_pw_4sage_8symbolic_10expression_115mixed_sorted, METH_O, __pyx_doc_4sage_8symbolic_10expression_114mixed_sorted},
{"call_registered_function", (PyCFunction)(void*)(PyCFunctionWithKeywords)__pyx_pw_4sage_8symbolic_10expression_121call_registered_function, METH_VARARGS|METH_KEYWORDS, __pyx_doc_4sage_8symbolic_10expression_120call_registered_function},
{"find_registered_function", (PyCFunction)(void*)(PyCFunctionWithKeywords)__pyx_pw_4sage_8symbolic_10expression_123find_registered_function, METH_VARARGS|METH_KEYWORDS, __pyx_doc_4sage_8symbolic_10expression_122find_registered_function},
{"register_or_update_function", (PyCFunction)(void*)(PyCFunctionWithKeywords)__pyx_pw_4sage_8symbolic_10expression_125register_or_update_function, METH_VARARGS|METH_KEYWORDS, __pyx_doc_4sage_8symbolic_10expression_124register_or_update_function},
{"get_sfunction_from_serial", (PyCFunction)__pyx_pw_4sage_8symbolic_10expression_127get_sfunction_from_serial, METH_O, __pyx_doc_4sage_8symbolic_10expression_126get_sfunction_from_serial},
{"get_sfunction_from_hash", (PyCFunction)__pyx_pw_4sage_8symbolic_10expression_129get_sfunction_from_hash, METH_O, __pyx_doc_4sage_8symbolic_10expression_128get_sfunction_from_hash},
{"make_map", (PyCFunction)__pyx_pw_4sage_8symbolic_10expression_131make_map, METH_O, __pyx_doc_4sage_8symbolic_10expression_130make_map},
{0, 0, 0, 0}
};
which are present with bindings=false.
Maybe its related to cython/cython#1658.
After replacing expression.cpp with the one from develop everything seems to work fine.
Does anyone has an idea how to properly fix this? I tried setting bindings=false for expression.pyx, but then the compilation doesn't succeed anymore.
I took a few doctest runs as benchmarks, and could only find a very small negative impact (about 1%) if at all.
bindings=True
hyperfine -i --warmup 3 './sage -t --long src/sage/manifolds/differentiable/pseudo_riemannian.py' --show-output
Time (mean ± σ): 13.183 s ± 0.266 s [User: 14.732 s, System: 1.173 s]
Range (min … max): 12.937 s … 13.730 s 10 runs
hyperfine -i --warmup 3 './sage -t --long src/sage/matrix/matrix0.pyx' --show-output
Time (mean ± σ): 7.719 s ± 0.176 s [User: 7.039 s, System: 0.918 s]
Range (min … max): 7.472 s … 7.993 s 10 runs
develop
hyperfine -i --warmup 3 './sage -t --long src/sage/manifolds/differentiable/pseudo_riemannian.py' --show-output
Time (mean ± σ): 13.096 s ± 0.199 s [User: 14.618 s, System: 1.167 s]
Range (min … max): 12.889 s … 13.524 s 10 runs
hyperfine -i --warmup 3 './sage -t --long src/sage/matrix/matrix0.pyx' --show-output
Time (mean ± σ): 7.675 s ± 0.165 s [User: 6.940 s, System: 0.970 s]
Range (min … max): 7.364 s … 7.841 s 10 runs
(disclaimer: these were run on gitpod, so there is no guarantee that the same resources were available)
Branch pushed to git repo; I updated commit sha1. New commits:
8d9bad0 | Merge branch 'develop' into public/26254 |
Branch pushed to git repo; I updated commit sha1. New commits:
bea228b | Try with localized import |
Branch pushed to git repo; I updated commit sha1. New commits:
803804f | and one more test |
This is now mostly working for me. A few doctests are still failing, mostly around cached functions and latex expressions. For some reason, some of the TypeErrors are now IndexErrors complaining that the index is out of bounds. Any idea where this is coming from?
Attachment: Screenshot 2022-05-08 133226.jpg
I tried to track down the source of the following errors:
File "/__w/sagetrac-mirror/sagetrac-mirror/src/sage/repl/display/pretty_print.py", line 144, in pretty
ok = representation(obj, self, cycle)
File "/__w/sagetrac-mirror/sagetrac-mirror/src/sage/repl/display/fancy_repr.py", line 348, in __call__
ascii_art_repr = ascii_art_repr or o.parent()._repr_option('element_ascii_art')
IndexError: tuple index out of range
Debugging the following test script
import sage.all
from sage.combinat.root_system.cartan_type import CartanType
from sage.repl.rich_output import pretty_print
C = CartanType(["A", 3, 1])
class MyCartanType:
def my_method(self):
return "I am here!"
C._add_abstract_superclass(MyCartanType)
pretty_print(C.__class__)
# <class 'sage.combinat.root_system.type_A_affine.CartanType_with_superclass_with_superclass'>
pretty_print(C.__class__.__bases__)
# This triggers the index error, but should print
# (<class 'sage.combinat.root_system.type_A_affine.CartanType_with_superclass'>,
# <class ...__main__.MyCartanType...>)
leads to the following output in fancy_repr.py, line 348:
For some reason calling parent throws the IndexError. The python debugger doesn't give more insights, and I have no experience with cython debugging to see where the indexerror is thrown. So it would be nice if someone with more cython experience can take this from here.
Replying to @tobiasdiez:
After merging the latest dev branch, this now throws
File "/__w/sagetrac-mirror/sagetrac-mirror/src/sage/all_cmdline.py", line 19, in <module> from sage.all import * File "/__w/sagetrac-mirror/sagetrac-mirror/src/sage/all.py", line 135, in <module> from sage.symbolic.all import * File "/__w/sagetrac-mirror/sagetrac-mirror/src/sage/symbolic/all.py", line 8, in <module> import sage.symbolic.expression # initialize pynac before .ring File "sage/symbolic/expression.pyx", line 1, in init sage.symbolic.expression # -*- coding: utf-8 -*- File "sage/symbolic/function.pyx", line 144, in init sage.symbolic.function from .expression import ( ImportError: cannot import name call_registered_functionwhen importing
sage.all, see eg github build workflow. Any idea on how to fix it?This method was introduced in #32386 (or one of its dependencies).
It seems that cython fails to import the function call_registered_function from sage.symbolic.expression via the src/sage/symbolic/pynac_function_impl.pxi file, where it is defined. I am not sure if we could regard this as a bug of cython as .pxi files are not supposed to contain function definitions. I wonder why it is moved there in the first place. Anyway I think we should ask to cython community about this first before we refactor our code.
Replying to @kwankyu:
Anyway I think we should ask to cython community about this first before we refactor our code.
I filed an issue:
expecting any input from cython experts.
Thanks for opening the issue.
Do you have an idea where the IndexError is coming from?
Branch pushed to git repo; I updated commit sha1. New commits:
a23ffe9 | Tiny edits |
Replying to @kwankyu:
Replying to @kwankyu:
Anyway I think we should ask to cython community about this first before we refactor our code.
I filed an issue:
expecting any input from cython experts.
According to the answer, there was a circular import that somehow was hidden but revealed only with binding=True. Hence I think changing global imports to local imports are the right solution.
Replying to @tobiasdiez:
Do you have an idea where the IndexError is coming from?
Note that with binding=False:
class NewClass:
def parent(self):
return 1
NewClass.parent()
...
TypeError: parent() missing 1 required positional argument: 'self'
%%cython
cdef class NewClass:
def parent(self):
return 1
NewClass.parent()
...
TypeError: unbound method NewClass.parent() needs an argument
With binding=True:
%%cython
cimport cython
@cython.binding(True)
cdef class NewClass:
def parent(self):
return 1
NewClass.parent()
...
IndexError: tuple index out of range
NewClass.parent(2)
1
In the latter, NewClass is of course not an object but a class, hence NewClass.parent() invokes unbound method .parent which is a cython function that thinks it is bound. Hence it tries to get "self" from its dictionary of arguments. As the dictionary is empty, it raises IndexError. I think this is a bug of cython. It should raise TypeError as it should follow Python behavior.
Cython issue:
Replying to @kwankyu:
Cython issue:
They are fixing this bug. I think we should wait for the new Cython instead of "fixing it" on our side.
Replying to @tobiasdiez:
I took a few doctest runs as benchmarks, and could only find a very small negative impact (about 1%) if at all.
Now I think we should regard this 1% as our debt (it was not ours), or as something to pay to get proper introspection to cython functions, which weighs more than the 1%.
Perhaps we may selectively turn on "binding=False" for some cython functions critical for performance.
Replying to @kwankyu:
Replying to @tobiasdiez:
Do you have an idea where the IndexError is coming from?
Note that with
binding=False:class NewClass: def parent(self): return 1 NewClass.parent() ... TypeError: parent() missing 1 required positional argument: 'self'%%cython cdef class NewClass: def parent(self): return 1 NewClass.parent() ... TypeError: unbound method NewClass.parent() needs an argumentWith
binding=True:%%cython cimport cython @cython.binding(True) cdef class NewClass: def parent(self): return 1 NewClass.parent() ... IndexError: tuple index out of range NewClass.parent(2) 1In the latter,
NewClassis of course not an object but a class, henceNewClass.parent()invokes unbound method.parentwhich is a cython function that thinks it is bound. Hence it tries to get "self" from its dictionary of arguments. As the dictionary is empty, it raisesIndexError. I think this is a bug of cython. It should raiseTypeErroras it should follow Python behavior.Cython issue:
Nice! Thanks for doing the analysis and finding a nice minimal example reproducing the issue.
Should be fixed in the new cython release. Updating to it is tracked at #33864.
Seems like the update of cython to 0.29.30 didn't help, the index errors are still present (at least in the github workflow).
Branch pushed to git repo; I updated commit sha1. New commits:
02d6282 | Merge remote-tracking branch 'origin/develop' into public/26254 |
This is mostly working now. There are a few failing doctests related to pickling, and what I believe is connected, module names (see the build & test workflow).
Moreover, a couple of doctests fail in pynac_impl.pxi. I sadly don't have the expertise to fix these (on my own) and would appreciate input from cython experts.
Branch pushed to git repo; I updated commit sha1. This was a forced push. New commits:
f38dc83 | Set cython directive binding=True |
This is working now. While I am not a cython expert, I managed to fix remaining failing doctests. Setting "needs review" to test.
To break circular imports, it is now not possible to import
call_registered_function,
find_registered_function,
register_or_update_function,
get_sfunction_from_hash,
get_sfunction_from_serial
from sage.symbolic.function. They should be imported from sage.symbolic.expression, where they are defined.
But then I could not figure out how to deprecate old imports. Or it would be better to make them importable from sage.symbolic.function as well. But I couldn't figure out how to do this either!
Thanks, nice that its working now.
Changed author from Kwankyu Lee to Kwankyu Lee, Tobias Diez
Reviewer: Tobias Diez, ...
Branch pushed to git repo; I updated commit sha1. New commits:
80375a2 | Put symbolic function related functions back into sage.symbolic.function |
The last commit allows importing
call_registered_function,
find_registered_function,
register_or_update_function,
get_sfunction_from_hash,
get_sfunction_from_serial
from sage.symbolic.function.
I think that Cython has to be rebuilt for this to take effect. (At least, I got some doctest failures when I built by just using the branch, and they went away after doing make distclean and trying again.) Can you modify the branch to trigger that, for example by bumping up the patch level on the Cython version number?
Replying to John Palmieri:
I think that Cython has to be rebuilt for this to take effect.
Perhaps no for Cython itself, but yes all cython files must be rebuilt.
Can you modify the branch to trigger that, for example by bumping up the patch level on the Cython version number?
Okay.
Branch pushed to git repo; I updated commit sha1. New commits:
735fbcf | Bump up patch level for Cython |