I have a very simple about sets. This is a minimal example:
#!/usr/bin/python
class x():
def __init__(self, y):
self.__y = y
def __eq__(self, other):
return self.__y == other.__y
def __hash__(self):
return hash(self.__y)
a = x("foo")
s = set([x("bar"), x("moo"), a])
z = x("foo")
print("z = ", z)
print(s)
for i in s:
print(i, i == a, i is a, i == z, i is z)
The problem is: two instances of x() are equal (__eq__ returns true),
but they are not identical. I have an equal element ("z"), but want to
get the *actual* element ("a") in the set. I.d. in the above example,
i'd like something like:
print(s.getelement(z) is a)
True
Is there something like the "getelement" function? How can I do what I want?
Kind regards,
Johannes
--
"Meine Gegenklage gegen dich lautet dann auf bewusste Verlogenheit,
verlästerung von Gott, Bibel und mir und bewusster Blasphemie."
-- Prophet und Visionär Hans Joss aka HJP in de.sci.physik
<48d8bf1d$0$7510$5402...@news.sunrise.ch>
> Is there something like the "getelement" function? How can I do what I want?
One side note: The obvious trivial solution:
def findset(s, e):
for i in s:
if e == i:
return i
return None
is because of its complexity of O(n) against the native O(log n) out of
the question...
> I have a very simple about sets. This is a minimal example:
> The problem is: two instances of x() are equal (__eq__ returns true),
> but they are not identical. I have an equal element ("z"), but want to
> get the *actual* element ("a") in the set. I.d. in the above example,
> i'd like something like:
>
> print(s.getelement(z) is a)
> True
>
> Is there something like the "getelement" function? How can I do what I
> want?
Here's a trick to find the actual element. I think Raymond Hettinger posted
an implementation of this idea recently, but I can't find it at the moment.
class X:
def __init__(self, y):
self.__y = y
def __eq__(self, other):
try:
return self.__y == other.__y
except AttributeError:
return NotImplemented
def __hash__(self):
return hash(self.__y)
a = X("foo")
s = set([X("bar"), X("moo"), a])
z = X("foo")
class Grab:
def __init__(self, value):
self.search_value = value
def __hash__(self):
return hash(self.search_value)
def __eq__(self, other):
if self.search_value == other:
self.actual_value = other
return True
return False
assert a == z
assert a is not z
grab = Grab(z)
grab in s
assert grab.actual_value is a
Peter
> class Grab:
> def __init__(self, value):
> self.search_value = value
> def __hash__(self):
> return hash(self.search_value)
> def __eq__(self, other):
> if self.search_value == other:
> self.actual_value = other
> return True
> return False
>
> assert a == z
> assert a is not z
> grab = Grab(z)
> grab in s
> assert grab.actual_value is a
Wow, this is truly amazing! I'd never have come up with that solution.
Just wonderful, thank you very much! :-))
That's very nice, and I may add to my tricks. Probably Raymond has
more brain than me :-)
But some other times you may accept to change the class and the set/
dict, making it tell apart the keys only when they are different
object:
class Some(object):
def __init__(self, y):
self._y = y
def __eq__(self, other):
return self is other
def __hash__(self):
return hash(self._y)
Now your set/dict keeps all the instances, even when they contain the
same _y.
Bye,
bearophile
Your code is inverted from Raymond's:
http://code.activestate.com/recipes/499299/
class _CaptureEq:
'Object wrapper that remembers "other" for successful equality tests.'
def __init__(self, obj):
self.obj = obj
self.match = None
def __eq__(self, other):
result = (self.obj == other)
if result:
self.match = other
return result
# support hash() or anything else needed by __ contains__
def __getattr__(self, name):
return getattr(self.obj, name)
def get_equivalent(container, item, default=None):
'''Gets the specific container element matched by: "item in container".
Useful for retreiving a canonical value equivalent to "item". For
example, a caching or interning application may require fetching a
single representativ e instance from many possible equivalent
instances).
>>> get_equivalent(set([1, 2, 3]), 2.0) # 2.0 is equivalent to 2
2
>>> get_equivalent([1, 2, 3], 4, default=0)
0
'''
t = _CaptureEq(item)
if t in container:
return t.match
return default
--
Aahz (aa...@pythoncraft.com) <*> http://www.pythoncraft.com/
"If you think it's expensive to hire a professional to do the job, wait
until you hire an amateur." --Red Adair
No. but you pratically wrote it: in your code, when i == z, then i is
the element you want.
def getelement(sset, item):
for i in sset:
if i == item: return i
The only use for this is see, though, is getting the representative of
an equivalence class from any member thereof.
tjr
> In article <gt1kb7$jqg$03$1...@news.t-online.com>,
> Peter Otten <__pet...@web.de> wrote:
>>
>>Here's a trick to find the actual element. I think Raymond Hettinger
>>posted an implementation of this idea recently, but I can't find it at the
>>moment.
>
> Your code is inverted from Raymond's:
I can't see the inversion.
> http://code.activestate.com/recipes/499299/
Thanks for providing the reference.
Peter
You were wrapping the objects inserted into the set; Raymond's trick
involved only wrapping the comparison object. It's therefore much more
convenient.
> In article <gt8v37$kib$03$2...@news.t-online.com>,
> Peter Otten <__pet...@web.de> wrote:
>>Aahz wrote:
>>> In article <gt1kb7$jqg$03$1...@news.t-online.com>,
>>> Peter Otten <__pet...@web.de> wrote:
>>>>
>>>>Here's a trick to find the actual element. I think Raymond Hettinger
>>>>posted an implementation of this idea recently, but I can't find it at
>>>>the moment.
>>>
>>> Your code is inverted from Raymond's:
>>
>>I can't see the inversion.
>
> You were wrapping the objects inserted into the set; Raymond's trick
> involved only wrapping the comparison object. It's therefore much more
> convenient.
I think you are misreading my code. I took the items (of class X) as they
were specified by the OP.
The reason I changed their __eq__() method is not that I did not understand
Raymond's trick, but rather a quirk in the set's item lookup:
>>> class A(object):
... def __init__(self, value):
... self.value = value
... def __hash__(self):
... return hash(self.value)
... def __eq__(self, other):
... return self.value == other.value
...
>>> item = A("a")
>>> container = map(A, "abc")
>>> from get_equivalent import get_equivalent
TestResults(failed=0, attempted=0)
get_equivalent() is Raymond's implementation from the recipe. Let's try it:
>>> get_equivalent(container, item)
<__main__.A object at 0x2091e50>
>>> _ is not item
True
Works. Now the same with a set:
>>> container = set(container)
>>> print get_equivalent(container, item)
None
Oops.
wanted in some_list
performs wanted.__eq__(candidate)
where candidate is an item in the list.
wanted in some_set
tries
candidate.__eq__(wanted)
first. You must ensure that this fails in an orderly manner for
wanted.__eq__(candidate)
to be tried at all:
>>> def __eq__(self, other):
... if not isinstance(other, A):
... return NotImplemented
... return self.value == other.value
...
>>> A.__eq__ = __eq__
>>> print get_equivalent(container, item)
<__main__.A object at 0x2091e50>
>>> _ is not item
True
Peter
Gotcha -- thanks for the explanation!
David Mertz suggested something like this in one of his Developer
Works
articles. Essentially, he was describing an IdentityDict or
IdentitySet.
I don't really see how those would be useful in regular python.
Why treat equal but not identical objects as distinct in an
environment where you have so little control over object identity?
>>> s = IdentitySet(['abc'])
>>> e = 'ab' + 'c' # distinct element, equal to 'abc', but not identical
>>> e in s
False
For the most part, we can't even count on equal integers having the
same identity:
>>> x = 1000
>>> y = 1001 - 1
>>> [id(o) for o in (x, y)]
[16675616, 16676240]
I'm curious about your use cases for the Some() class in conjunction
with a dict or set.
Raymond