44 KiB
| layout | title | tags | description |
|---|---|---|---|
| post | Finding and Replacing Objects in Python | Python | More reflection than you cared to ask for |
Today, we’re going to demonstrate a fairly evil thing in Python, which I call object replacement.
Say you have some program that’s been running for a while, and a particular
object has made its way throughout your code. It lives inside lists, class
attributes, maybe even inside some closures. You want to completely replace
this object with another one; that is to say, you want to find all references
to object A and replace them with object B, enabling A to be garbage
collected. This has some interesting implications for special object types. If
you have methods that are bound to A, you want to rebind them to B. If A
is a class, you want all instances of A to become instances of B. And so
on.
But why on Earth would you want to do that? you ask. I’ll focus on a concrete use case in a future post, but for now, I imagine this could be useful in some kind of advanted unit testing situation with mock objects. Still, it’s fairly insane, so let’s leave it primarily as an intellectual exercise.
This article is written for CPython 2.7.[1]
Review
First, a recap on terminology here. You can skip this section if you know Python well.
In Python, names are what most languages call “variables”. They reference objects. So when we do:
{% highlight python %}
a = [1, 2, 3, 4]
{% endhighlight %}
...we are creating a list object with four integers, and binding it to the name
a. In graph form:[2]
[1, 2, 3, 4]a
In each of the following examples, we are creating new references to the
list object, but we are never duplicating it. Each reference points to the same
memory address (which you can get using id(a)).
{% highlight python %}
b = a
{% endhighlight %}
{% highlight python %}
c = SomeContainerClass() c.data = a
{% endhighlight %}
{% highlight python %}
def wrapper(L): def inner(): return L.pop() return inner
d = wrapper(a)
{% endhighlight %}
d[1, 2, 3, 4]abc.dataL
Note that these references are all equal. a is no more valid a name for the
list than b, c.data, or L (or d.func_closure[0].cell_contents to the
outside world). As a result, if you delete one of these references—explicitly
with del a, or implicitly if a name goes out of scope—then the other
references are still around, and object continues to exist. If all of an
object’s references disappear, then Python’s garbage collector should eliminate
it.
Dead ends
My first thought when approaching this problem was to physically write over the
memory where our target object is stored. This can be done using
ctypes.memmove()
from the Python standard library:
{% highlight pycon %}
class A(object): pass ... class B(object): pass ... obj = A() print obj <__main__.A object at 0x10e3e1190> import ctypes ctypes.memmove(id(A), id(B), object.__sizeof__(A)) 140576340136752 print obj <__main__.B object at 0x10e3e1190>
{% endhighlight %}
What we are doing here is overwriting the fields of the A instance of the
PyClassObject C struct
with fields from the B struct instance. As a result, they now share various
properties, such as their attribute dictionaries
(__dict__).
So, we can do things like this:
{% highlight pycon %}
B.foo = 123 obj.foo 123
{% endhighlight %}
However, there are clear issues. What we’ve done is create a
shallow copy.
Therefore, A and B are still distinct objects, so certain changes made to
one will not be replicated to the other:
{% highlight pycon %}
A is B False B.name = “C” A.name ‘B’
{% endhighlight %}
Also, this won’t work if A and B are different sizes, since we will be
either reading from or writing to memory that we don’t necessarily own:
{% highlight pycon %}
A = () B = [] print A.sizeof(), B.sizeof() 24 40 import ctypes ctypes.memmove(id(A), id(B), A.sizeof()) 4321271888 Python(33575,0x7fff76925300) malloc: *** error for object 0x6f: pointer being freed was not allocated *** set a breakpoint in malloc_error_break to debug Abort trap: 6
{% endhighlight %}
Oh, and there’s a bit of a problem when we deallocate these objects, too...
{% highlight pycon %}
A = [] B = range(8) import ctypes ctypes.memmove(id(A), id(B), A.sizeof()) 4514685728 print A [0, 1, 2, 3, 4, 5, 6, 7] del A del B Segmentation fault: 11
{% endhighlight %}
Fishing for references with Guppy
A more appropriate solution is finding all of the references to the old object, and then updating them to point to the new object, rather than replacing the old object directly.
But how do we track references? Fortunately, there’s a library called
Guppy that allows us to do this. Often used
for diagnosing memory leaks, we can take advantage of its robust object
tracking features here. Install it with pip
(pip install guppy).
I’ve always found Guppy hard to use (as many debuggers are, though justified by the complexity of the task involved), so we’ll begin with a feature demo before delving into the actual problem.
Feature demonstration
Guppy’s interface is deceptively simple. We begin by calling
guppy.hpy(),
to expose the Heapy interface, which is the component of Guppy with the
features we want:
{% highlight pycon %}
import guppy hp = guppy.hpy() hp Top level interface to Heapy. Use eg: hp.doc for more info on hp.
{% endhighlight %}
Calling
hp.heap()
shows us a table of the objects known to Guppy, grouped together
(mathematically speaking,
partitioned) by
type[3] and sorted by how much space
they take up in memory:
{% highlight pycon %}
heap = hp.heap() heap Partition of a set of 45761 objects. Total size = 4699200 bytes. Index Count % Size % Cumulative % Kind (class / dict of class) 0 15547 34 1494736 32 1494736 32 str 1 8356 18 770272 16 2265008 48 tuple 2 346 1 452080 10 2717088 58 dict (no owner) 3 13685 30 328440 7 3045528 65 int 4 71 0 221096 5 3266624 70 dict of module 5 1652 4 211456 4 3478080 74 types.CodeType 6 199 0 210856 4 3688936 79 dict of type 7 1614 4 193680 4 3882616 83 function 8 199 0 177008 4 4059624 86 type 9 124 0 135328 3 4194952 89 dict of class <91 more rows. Type e.g. ‘_.more’ to view.>
{% endhighlight %}
This object (called an
IdentitySet)
looks bizarre, but it can be treated roughly like a list. If we want to take a
look at strings, we can do heap[0]:
{% highlight pycon %}
heap[0] Partition of a set of 22606 objects. Total size = 2049896 bytes. Index Count % Size % Cumulative % Kind (class / dict of class) 0 22606 100 2049896 100 2049896 100 str
{% endhighlight %}
This isn’t very useful, though. What we really want to do is re-partition this subset using another relationship. There are a number of options, such as:
{% highlight pycon %}
heap[0].byid # Group by object ID; each subset therefore has one element Set of 22606 objects. Total size = 2049896 bytes. Index Size % Cumulative % Representation (limited) 0 7480 0.4 7480 0.4 ‘The class Bi... copy of S.\n’ 1 4872 0.2 12352 0.6 “Support for ... ‘error’.\n\n” 2 4760 0.2 17112 0.8 ‘Heap queues...at Art! :-)\n’ 3 4760 0.2 21872 1.1 ‘Heap queues...at Art! :-)\n’ 4 3896 0.2 25768 1.3 ‘This module ...ng function\n’ 5 3824 0.2 29592 1.4 ‘The type of ...call order.\n’ 6 3088 0.2 32680 1.6 ‘t\x00\x00|\x...x00|\x02\x00S’ 7 2992 0.1 35672 1.7 ‘HeapView(roo... size, etc.\n’ 8 2808 0.1 38480 1.9 ‘Directory tr...ories\n\n ' 9 2640 0.1 41120 2.0 ‘The class No... otherwise.\n’ <22596 more rows. Type e.g. ‘_.more’ to view.>
{% endhighlight %}
{% highlight pycon %}
heap[0].byrcs # Group by what types of objects reference the strings Partition of a set of 22606 objects. Total size = 2049896 bytes. Index Count % Size % Cumulative % Referrers by Kind (class / dict of class) 0 6146 27 610752 30 610752 30 types.CodeType 1 5304 23 563984 28 1174736 57 tuple 2 4104 18 237536 12 1412272 69 dict (no owner) 3 1959 9 139880 7 1552152 76 list 4 564 2 136080 7 1688232 82 function, tuple 5 809 4 97896 5 1786128 87 dict of module 6 346 2 71760 4 1857888 91 dict of type 7 365 2 19408 1 1877296 92 dict of module, tuple 8 192 1 16176 1 1893472 92 dict (no owner), list 9 232 1 11784 1 1905256 93 dict of class, function, tuple, types.CodeType <229 more rows. Type e.g. ‘_.more’ to view.>
{% endhighlight %}
{% highlight pycon %}
heap[0].byvia # Group by how the strings are related to their referrers Partition of a set of 22606 objects. Total size = 2049896 bytes. Index Count % Size % Cumulative % Referred Via: 0 2656 12 420456 21 420456 21 ‘[0]’ 1 2095 9 259008 13 679464 33 ‘.co_code’ 2 2095 9 249912 12 929376 45 ‘.co_filename’ 3 564 2 136080 7 1065456 52 ‘.func_doc’, ‘[0]’ 4 243 1 103528 5 1168984 57 “['doc']” 5 1930 9 100584 5 1269568 62 ‘.co_lnotab’ 6 502 2 31128 2 1300696 63 ‘[1]’ 7 306 1 16272 1 1316968 64 ‘[2]’ 8 242 1 12960 1 1329928 65 ‘[3]’ 9 184 1 9872 0 1339800 65 ‘[4]’ <7323 more rows. Type e.g. ‘_.more’ to view.>
{% endhighlight %}
From this, we can see that the plurality of memory devoted to strings is taken
up by those referenced by code objects (types.CodeType represents
Python code—accessible from a non-C-defined function through
func.func_code—and contains things like the names of its local variables and
the actual sequence of opcodes that make it up).
For fun, let’s pick a random string.
{% highlight pycon %}
import random obj = heap[0].byid[random.randrange(0, heap[0].count)] obj Set of 1 object. Total size = 176 bytes. Index Size % Cumulative % Representation (limited) 0 176 100.0 176 100.0 ‘Define names...not listed.\n’
{% endhighlight %}
Interesting. Since this heap subset contains only one element, we can use
.theone
to get the actual object represented here:
{% highlight pycon %}
obj.theone ‘Define names for all type symbols known in the standard interpreter.\n\nTypes that are part of optional modules (e.g. array) are not listed.\n’
{% endhighlight %}
Looks like the docstring for the
types module. We can confirm
by using
.referrers
to get the set of objects that refer to objects in the given set:
{% highlight pycon %}
obj.referrers Partition of a set of 1 object. Total size = 3352 bytes. Index Count % Size % Cumulative % Kind (class / dict of class) 0 1 100 3352 100 3352 100 dict of module
{% endhighlight %}
This is types.__dict__ (since the docstring we got is actually stored as
types.__dict__["__doc__"]), so if we use .referrers again:
{% highlight pycon %}
obj.referrers.referrers Partition of a set of 1 object. Total size = 56 bytes. Index Count % Size % Cumulative % Kind (class / dict of class) 0 1 100 56 100 56 100 module obj.referrers.referrers.theone <module ‘types’ from ‘/usr/local/Cellar/python/2.7.8_2/Frameworks/Python.framework/Versions/2.7/lib/python2.7/types.pyc'> import types types.doc is obj.theone True
{% endhighlight %}
But why did we find an object in the types module if we never imported it?
Well, let’s see. We can use
hp.iso()
to get the Heapy set consisting of a single given object:
{% highlight pycon %}
hp.iso(types) Partition of a set of 1 object. Total size = 56 bytes. Index Count % Size % Cumulative % Kind (class / dict of class) 0 1 100 56 100 56 100 module
{% endhighlight %}
Using a similar procedure as before, we see that types is imported by the
traceback module:
{% highlight pycon %}
hp.iso(types).referrers Partition of a set of 10 objects. Total size = 25632 bytes. Index Count % Size % Cumulative % Kind (class / dict of class) 0 2 20 13616 53 13616 53 dict (no owner) 1 5 50 9848 38 23464 92 dict of module 2 1 10 1048 4 24512 96 dict of guppy.etc.Glue.Interface 3 1 10 1048 4 25560 100 dict of guppy.etc.Glue.Share 4 1 10 72 0 25632 100 tuple hp.iso(types).referrers[1].byid Set of 5 objects. Total size = 9848 bytes. Index Size % Cumulative % Owner Name 0 3352 34.0 3352 34.0 traceback 1 3352 34.0 6704 68.1 warnings 2 1048 10.6 7752 78.7 main 3 1048 10.6 8800 89.4 abc 4 1048 10.6 9848 100.0 guppy.etc.Glue
{% endhighlight %}
...and that is imported by
site:
{% highlight pycon %}
import traceback hp.iso(traceback).referrers Partition of a set of 3 objects. Total size = 15992 bytes. Index Count % Size % Cumulative % Kind (class / dict of class) 0 1 33 12568 79 12568 79 dict (no owner) 1 1 33 3352 21 15920 100 dict of module 2 1 33 72 0 15992 100 tuple hp.iso(traceback).referrers[1].byid Set of 1 object. Total size = 3352 bytes. Index Size % Cumulative % Owner Name 0 3352 100.0 3352 100.0 site
{% endhighlight %}
Since site is imported by Python on startup, we’ve figured out why objects
from types exist, even though we’ve never used them.
We’ve learned something important, too. When objects are stored as ordinary
attributes of a parent object (like types.__doc__, traceback.types, and
site.traceback from above), they are not referenced directly by the parent
object, but by that object’s __dict__ attribute. Therefore, if we want to
replace A with B and A is an attribute of C, we (probably) don’t need
to know anything special about C—just how to modify dictionaries.
A good Guppy/Heapy tutorial, while a bit old and incomplete, can be found on Andrey Smirnov’s website.
Examining paths
Let’s set up an example replacement using class instances:
{% highlight python %}
class A(object): pass
class B(object): pass
a = A() b = B()
{% endhighlight %}
Suppose we want to replace a with b. From the demo above, we know that we
can get the Heapy set of a single object using hp.iso(). We also know we can
use .referrers to get the set of objects that reference the given object:
{% highlight pycon %}
import guppy hp = guppy.hpy() print hp.iso(a).referrers Partition of a set of 1 object. Total size = 1048 bytes. Index Count % Size % Cumulative % Kind (class / dict of class) 0 1 100 1048 100 1048 100 dict of module
{% endhighlight %}
a is only referenced by one object, which makes sense, since we’ve only used
it in one place—as a local variable—meaning hp.iso(a).referrers.theone must
be locals():
{% highlight pycon %}
hp.iso(a).referrers.theone is locals() True
{% endhighlight %}
However, there is a more useful feature available to us:
.pathsin.
This also returns references to the given object, but instead of a Heapy set,
it is a list of Path objects. These are more useful since they tell us not
only what objects are related to the given object, but how they are
related.
{% highlight pycon %}
print hp.iso(a).pathsin 0: Src[‘a’]
{% endhighlight %}
This looks very ambiguous. However, we find that we can extract the source of
the reference using .src:
{% highlight pycon %}
path = hp.iso(a).pathsin[0] print path.src Partition of a set of 1 object. Total size = 1048 bytes. Index Count % Size % Cumulative % Kind (class / dict of class) 0 1 100 1048 100 1048 100 dict of module path.src.theone is locals() True
{% endhighlight %}
...and, we can examine the type of relation by looking at .path[1] (the
actual reason for this isn’t worth getting into, due to Guppy’s lack of
documentation on the subject):
{% highlight pycon %}
relation = path.path[1] relation <guppy.heapy.Path.Based_R_INDEXVAL object at 0x100f38230>
{% endhighlight %}
We notice that relation is a Based_R_INDEXVAL object. Sounds bizarre, but
this tells us that a is a particular indexed value of path.src. What index?
We can get this using relation.r:
{% highlight pycon %}
rel = relation.r print rel a
{% endhighlight %}
Ah ha! So now we know that a is equal to the reference source (i.e.,
path.src.theone) indexed by rel:
{% highlight pycon %}
path.src.theone[rel] is a True
{% endhighlight %}
But path.src.theone is just a dictionary, meaning we know how to modify it
very easily:[4]
{% highlight pycon %}
path.src.theone[rel] = b a <__main__.B object at 0x100dae090> a is b True
{% endhighlight %}
Bingo. We’ve successfully replaced a with b, using a general method that
should work for any case where a is in a dictionary-like object.
Handling different reference types
We’ll continue by wrapping this code up in a nice function, which we will expand as we go:
{% highlight python %}
import guppy from guppy.heapy import Path
hp = guppy.hpy()
def replace(old, new): for path in hp.iso(old).pathsin: relation = path.path[1] if isinstance(relation, Path.R_INDEXVAL): path.src.theone[relation.r] = new
{% endhighlight %}
Dictionaries, lists, and tuples
As noted above, this is versatile to handle many dictionary-like situations,
including __dict__, which means we already know how to replace object
attributes:
{% highlight pycon %}
a, b = A(), B()
class X(object): ... pass ... X.cattr = a x = X() x.iattr = a d1 = {1: a} d2 = [{1: {0: (“foo”, “bar”, {“a”: a, “b”: b})}}]
replace(a, b)
print a <__main__.B object at 0x1042b9910> print X.cattr <__main__.B object at 0x1042b9910> print x.iattr <__main__.B object at 0x1042b9910> print d1[1] <__main__.B object at 0x1042b9910> print d2[0][1][0][2][“a”] <__main__.B object at 0x1042b9910>
{% endhighlight %}
Lists can be handled exactly the same as dictionaries, although the keys in
this case (i.e., relation.r) will always be integers.
{% highlight pycon %}
a, b = A(), B() L = [0, 1, 2, a, b] print L [0, 1, 2, <__main__.A object at 0x104598950>, <__main__.B object at 0x104598910>] replace(a, b) print L [0, 1, 2, <__main__.B object at 0x104598910>, <__main__.B object at 0x104598910>]
{% endhighlight %}
Tuples are interesting. We can’t modify them directly because they’re immutable, but we can create a new tuple with the new value, and then replace that tuple just like we replaced our original object:
{% highlight python %}
# Meanwhile, in replace()...
if isinstance(relation, Path.R_INDEXVAL):
source = path.src.theone
if isinstance(source, tuple):
temp = list(source)
temp[relation.r] = new
replace(source, tuple(temp))
else:
source[relation.r] = new
{% endhighlight %}
As a result:
{% highlight pycon %}
a, b = A(), B() t1 = (0, 1, 2, a) t2 = (0, (1, (2, (3, (4, (5, (a,))))))) replace(a, b) print t1 (0, 1, 2, <__main__.B object at 0x104598e50>) print t2 (0, (1, (2, (3, (4, (5, (<__main__.B object at 0x104598e50>,)))))))
{% endhighlight %}
Bound methods
Here’s a fun one. Let’s upgrade our definitions of A and B:
{% highlight python %}
class A(object): def func(self): return self
class B(object): pass
{% endhighlight %}
After replacing a with b, a.func no longer exists, as we’d expect:
{% highlight pycon %}
a, b = A(), B() a.func() <__main__.A object at 0x10c4a5b10> replace(a, b) a.func() Traceback (most recent call last): File “”, line 1, in AttributeError: ‘B’ object has no attribute ‘func’
{% endhighlight %}
But what if we save a reference to a.func before the replacement?
{% highlight pycon %}
a, b = A(), B() f = a.func replace(a, b) f() <__main__.A object at 0x10c4b6090>
{% endhighlight %}
Hmm. So f has kept a reference to a somehow, but not in a dictionary-like
object. So where is it?
Well, we can reveal it with the attribute f.__self__:
{% highlight pycon %}
f.self <__main__.A object at 0x10c4b6090>
{% endhighlight %}
Unfortunately, this attribute is magical and we can’t write to it directly:
{% highlight pycon %}
f.self = b Traceback (most recent call last): File “”, line 1, in TypeError: readonly attribute
{% endhighlight %}
Python clearly doesn’t want us to re-bind bound methods, and a reasonable
person would give up here, but we still have a few tricks up our sleeve. Let’s
examine the internal C structure of bound methods,
PyMethodObject:
PyMethodObject<main.A object at 0xdeadbeef>struct _object* _ob_nextstruct _object* _ob_prevPy_ssize_t ob_refcntstruct _typeobject* ob_typePyObject* im_funcPyObject* im_selfPyObject* im_classPyObject* im_weakreflist
The four gray fields of the struct come from
PyObject_HEAD,
which exist in all Python objects. The first two fields are from
_PyObject_HEAD_EXTRA,
and only exist when the debugging macro Py_TRACE_REFS is defined, in order to
support more advanced reference counting. We can see that the im_self field,
which mantains the reference to our target object, is either forth or sixth in
the struct depending on Py_TRACE_REFS. If we can figure out the size of the
field and its offset from the start of the struct, then we can set its value
directly using ctypes.memmove():
{% highlight python %}
ctypes.memmove(id(f) + offset, ctypes.byref(ctypes.py_object(b)), field_size)
{% endhighlight %}
Here, id(f) is the memory location of our method, which refers to the start
of the C struct from above. offset is the number of bytes between this memory
location and the start of the im_self field. We use
ctypes.byref()
to create a reference to the replacement object, b, which will be copied over
the existing reference to a. Finally, field_size is the number of bytes
we’re copying, equal to the size of the im_self field.
Well, all but one of these fields are pointers to structure types, meaning they
have the same size,[5] equal to
ctypes.sizeof(ctypes.py_object).
This is (probably) 4 or 8 bytes, depending on whether you’re on a 32-bit or a
64-bit system. The other field is a Py_ssize_t object—possibly the same size
as the pointers, but we can’t be sure—which is equal to
ctypes.sizeof(ctypes.c_ssize_t).
We know that field_size must be ctypes.sizeof(ctypes.py_object), since we
are copying a structure pointer. offset is this value multiplied by the
number of structure pointers before im_self (4 if Py_TRACE_REFS is defined
and 2 otherwise), plus ctypes.sizeof(ctypes.c_ssize_t) for ob_type. But how
do we determine if Py_TRACE_REFS is defined? We can’t check the value of a
macro at runtime, but we can check for the existence of
sys.getobjects(),
which is
only defined when that macro is.
Therefore, we can make our replacement like so:
{% highlight pycon %}
import ctypes import sys field_size = ctypes.sizeof(ctypes.py_object) ptrs_in_struct = 4 if hasattr(sys, “getobjects”) else 2 offset = ptrs_in_struct * field_size + ctypes.sizeof(ctypes.c_ssize_t) ctypes.memmove(id(f) + offset, ctypes.byref(ctypes.py_object(b)), field_size) 4470258440 f.self is b True f() <__main__.B object at 0x10a8af290>
{% endhighlight %}
Excellent—it worked!
There’s another kind of bound method, which is the built-in variety as opposed
to the user-defined variety we saw above. An example is a.__sizeof__():
{% highlight pycon %}
a, b = A(), B() f = a.sizeof f <built-in method sizeof of A object at 0x10ab44b50> replace(a, b) f.self <__main__.A object at 0x10ab44b50>
{% endhighlight %}
This is stored internally as a
PyCFunctionObject.
Let’s take a look at its layout:
PyCFunctionObject<main.A object at 0xdeadbeef>struct _object* _ob_nextstruct _object* _ob_prevPy_ssize_t ob_refcntstruct _typeobject* ob_typePyMethodDef* m_mlPyObject* m_selfPyObject* m_module
Fortunately, m_self here has the same offset as im_self from before, so we
can just use the same code:
{% highlight pycon %}
ctypes.memmove(id(f) + offset, ctypes.byref(ctypes.py_object(b)), field_size) 4474703768 f.self is b True f <built-in method sizeof of B object at 0x10ab4f150>
{% endhighlight %}
Dictionary keys
Dictionary keys have a different reference relation type than values, but the
replacement works mostly the same way. We pop the value of the old key from the
dictionary, and then insert it in again under the new key. Here’s the code,
which we’ll stick into the main block in replace():
{% highlight python %}
elif isinstance(relation, Path.R_INDEXKEY): source = path.src.theone source[new] = source.pop(source.keys()[relation.r])
{% endhighlight %}
And, a demonstration:
{% highlight pycon %}
a, b = A(), B() d = {a: 1} replace(a, b) d {<__main__.B object at 0x10fb47950>: 1}
{% endhighlight %}
Closure cells
We’ll cover just one more case, this time involving a closure. Here’s our test function:
{% highlight python %}
def wrapper(obj): def inner(): return obj return inner
{% endhighlight %}
As we can see, an instance of the inner function keeps references to the locals
of the wrapper function, even after using our current
version of replace():
{% highlight pycon %}
a, b = A(), B() f = wrapper(a) f() <__main__.A object at 0x109446090> replace(a, b) f() <__main__.A object at 0x109446090>
{% endhighlight %}
Internally, CPython implements this using things called
cells. We notice that
f.func_closure gives us a tuple of cell objects, and we can examine an
individual cell’s contents with .cell_contents:
{% highlight pycon %}
f.func_closure (<cell at 0x10ad9f478: instance object at 0x109446090>,) f.func_closure[0].cell_contents <__main__.A object at 0x109446090>
{% endhighlight %}
As expected, we can’t just modify it...
{% highlight pycon %}
f.func_closure[0].cell_contents = b Traceback (most recent call last): File “”, line 1, in AttributeError: attribute ‘cell_contents’ of ‘cell’ objects is not writable
{% endhighlight %}
...because that would be too easy. So, how can we replace it? Well, we could
go back to memmove, but there’s an easier way thanks to the ctypes module
also exposing Python’s C API. Specifically, the
PyCell_Set function
(which seems to lack a pure Python equivalent) does exactly what we want. Since
the function expects PyObject*s as arguments, we’ll need to use
ctypes.py_object as a wrapper. Here it is:
{% highlight pycon %}
from ctypes import py_object, pythonapi pythonapi.PyCell_Set(py_object(f.func_closure[0]), py_object(b)) 0 f() <__main__.B object at 0x10ad94dd0>
{% endhighlight %}
Perfect – the replacement worked. To tie it together with replace(), we’ll
note that Guppy represents the cell contents relationship with
Based_R_INTERATTR, for what I assume to be “internal attribute”. We can use
this to find the cell object within the inner function that references our
target object, and then use the method above to make the change:
{% highlight python %}
elif isinstance(relation, Path.R_INTERATTR): if isinstance(source, CellType): pythonapi.PyCell_Set(py_object(source), py_object(new)) return
{% endhighlight %}
Other cases
There are many, many more types of possible replacements. I’ve written a more
extensible version of replace() with some test cases, which can be viewed
on Gist here.
Certainly, not every case is handled by it, but it seems to cover the majority
that I’ve found through testing. There are a number of reference relations in
Guppy that I couldn’t figure out how to replicate without doing something
insane (R_HASATTR, R_CELL, and R_STACK), so some obscure replacements are
likely unimplemented.
Some other kinds of replacements are known, but impossible. For example,
replacing a class object that uses __slots__ with another class will not work
if the replacement class has a different slot layout and instances of the old
class exist. More generally, replacing a class with a non-class object won’t
work if instances of the class exist. Furthermore, references stored in data
structures managed by C extensions cannot be changed, since there’s no good way
for us to track these.
Footnotes
-
^ This post relies heavily on implementation details of CPython 2.7. While it could be adapted for Python 3 by examining changes to the internal structures of objects that we used above, that would be a lost cause if you wanted to replicate this on Jython or some other implementation. We are so dependent on concepts specific to CPython that you would need to start from scratch, beginning with a language-specific replacement for Guppy.
-
^ The DOT files used to generate graphs in this post are available on Gist.
-
^ They’re actually grouped together by clodo (“class or dict object”), which is similar to type, but groups
__dict__s separately by their owner’s type. -
^ Python’s documentation tells us not to modify the locals dictionary, but screw that; we’re gonna do it anyway.
-
^ According to the C99 and C11 standards; section 6.2.5.27 in the former and 6.2.5.28 in the latter: “All pointers to structure types shall have the same representation and alignment requirements as each other.”