[issue5397] PEP 372: OrderedDict

Changes by Matthieu Labbé : -- nosy: +matthieu.labbe ___ Python tracker ___ ___ Python-bugs-list mailing list Unsubscribe: http://mail

[issue5397] PEP 372: OrderedDict

Forest Wilkinson added the comment: Agreed here. Thanks, gents. ___ Python tracker ___ ___ Python-bugs-list mailing list Unsubscribe: http://

[issue5397] PEP 372: OrderedDict

Georg Brandl added the comment: OK, disregard my suggestions, it's better not to have a property that does almost exactly the same as keys(). ___ Python tracker ___ __

[issue5397] PEP 372: OrderedDict

Changes by Armin Ronacher : ___ Python tracker ___ ___ Python-bugs-list mailing list Unsubscribe: http://mail.python.org/mailman/options/python

[issue5397] PEP 372: OrderedDict

Armin Ronacher added the comment: Please no. We just decided to *not* extend the API. The PEP originally had a well designed list of dict API extensions that already provided exactly that. If we really want to provide access to that, we can roll back to where we came from. I don't think impl

[issue5397] PEP 372: OrderedDict

Armin Ronacher added the comment: Please no. We just decided to *not* extend the API. The PEP originally had a well designed list of dict API extensions that already provided exactly that. If we really want to provide access to that, we can roll back to where we came from. I don't think impl

[issue5397] PEP 372: OrderedDict

Changes by Raymond Hettinger : Removed file: http://bugs.python.org/file13229/od6.diff ___ Python tracker ___ ___ Python-bugs-list mailing list

[issue5397] PEP 372: OrderedDict

Changes by Raymond Hettinger : Removed file: http://bugs.python.org/file13228/od5.diff ___ Python tracker ___ ___ Python-bugs-list mailing list

[issue5397] PEP 372: OrderedDict

Changes by Raymond Hettinger : Removed file: http://bugs.python.org/file13221/od4.diff ___ Python tracker ___ ___ Python-bugs-list mailing list

[issue5397] PEP 372: OrderedDict

Forest Wilkinson added the comment: > Shouldn't popitem() allow the caller to choose which end from > which to pop? Thinking it through a bit more, and LRU cache would actually need to access the oldest item without necessarily removing it. Besides, popitem() should probably retain the signatu

[issue5397] PEP 372: OrderedDict

Raymond Hettinger added the comment: Forest, for your use case I recommend copying the code to a new class and replacing the _keys list with a deque so that you can efficiently pop from the other end. ___ Python tracker ___

[issue5397] PEP 372: OrderedDict

Raymond Hettinger added the comment: Forest, I've taken another look at what's involved and am inclined to accept the idea. It can be done without mucking-up the regular dict API and without precluding any of the other possible underlying algorithms. I like that it supports an entire new categ

[issue5397] PEP 372: OrderedDict

Georg Brandl added the comment: > The internal data structure *must* remain private so that we can build a > C replacement or switch to one of the other possible algorithms. Even then the keys list could be offered as a property. ___ Python tracker

[issue5397] PEP 372: OrderedDict

Raymond Hettinger added the comment: The internal data structure *must* remain private so that we can build a C replacement or switch to one of the other possible algorithms. ___ Python tracker _

[issue5397] PEP 372: OrderedDict

Georg Brandl added the comment: I wonder if, instead of all kinds of new APIs, the _keys list could just be made public (under a different name of course). Of course, that makes further optimization or a rewrite in C harder. ___ Python tracker

[issue5397] PEP 372: OrderedDict

Forest Wilkinson added the comment: I was just reading the PEP, and caught this bit: "Does OrderedDict.popitem() return a particular key/value pair? Yes. It pops-off the most recently inserted new key and its corresponding value." Okay, but I'd also like a convenient and fast way to find the o

[issue5397] PEP 372: OrderedDict

Armin Ronacher added the comment: Maybe premature optimization but maybe it would make sense to implement __eq__ like this: def __eq__(self, other): if isinstance(other, OrderedDict): if not dict.__eq__(self, other): return False return all(p == q for p, q in _zi

[issue5397] PEP 372: OrderedDict

Changes by Raymond Hettinger : ___ Python tracker ___ ___ Python-bugs-list mailing list Unsubscribe: http://mail.python.org/mailman/options/pyt

[issue5397] PEP 372: OrderedDict

Raymond Hettinger added the comment: Checked-in r70101 and r70102 -- resolution: -> accepted status: open -> closed ___ Python tracker ___ __

[issue5397] PEP 372: OrderedDict

Raymond Hettinger added the comment: Attaching update reflecting Guido's change to __eq__(). Added file: http://bugs.python.org/file13231/od7.diff ___ Python tracker ___

[issue5397] PEP 372: OrderedDict

Raymond Hettinger added the comment: At Antoine's request, strengthened the tests in test_copying. -- assignee: -> rhettinger Added file: http://bugs.python.org/file13229/od6.diff ___ Python tracker __

[issue5397] PEP 372: OrderedDict

Raymond Hettinger added the comment: Jim, I updated the docs to talk cover delete-reinsert. Also, added a few more tests as suggested but am leaving test_iterators, test_popitem, and test_pop unchanged. It is enough to test delete-reinsertion once or twice and know that we've separately tested

[issue5397] PEP 372: OrderedDict

Jim Jewett added the comment: I would also recommend strengthening some of the tests with regard to ordering stability across update vs delete-and-reinsert. TestOrderedDict.test_update does verify that updated items are not moved to the end, but the comment suggests it is only checking that t

[issue5397] PEP 372: OrderedDict

Jim Jewett added the comment: I would try to make it more explicit that updates do not reset the order, but deleting a item and re-inserting it *does*. (So it isn't the first insertion of the key, it is the first that hasn't yet been followed by a deletion.) Maybe change: """ When iteratin

[issue5397] PEP 372: OrderedDict

Changes by Raymond Hettinger : Removed file: http://bugs.python.org/file13220/od3.diff ___ Python tracker ___ ___ Python-bugs-list mailing list

[issue5397] PEP 372: OrderedDict

Raymond Hettinger added the comment: Added a few more tests. Added file: http://bugs.python.org/file13221/od4.diff ___ Python tracker ___ ___

[issue5397] PEP 372: OrderedDict

Changes by Raymond Hettinger : Removed file: http://bugs.python.org/file13217/od.diff ___ Python tracker ___ ___ Python-bugs-list mailing list

[issue5397] PEP 372: OrderedDict

Changes by Raymond Hettinger : Removed file: http://bugs.python.org/file13219/od2.diff ___ Python tracker ___ ___ Python-bugs-list mailing list

[issue5397] PEP 372: OrderedDict

Raymond Hettinger added the comment: Fixed. Added file: http://bugs.python.org/file13220/od3.diff ___ Python tracker ___ ___ Python-bugs-list

[issue5397] PEP 372: OrderedDict

Georg Brandl added the comment: I still see an "OrderDict" in the docs, and the TypeError in __init__ still needs to use "%" instead of ",". ___ Python tracker ___ ___

[issue5397] PEP 372: OrderedDict

Raymond Hettinger added the comment: Thanks for the review comments guys. An updated patch is attached. Added file: http://bugs.python.org/file13219/od2.diff ___ Python tracker ___

[issue5397] PEP 372: OrderedDict

Armin Ronacher added the comment: @Georg > * eval()ing the repr() will not construct the dict in the same order The alternative would be a list of dicts inside the constructor call, but that feels ugly. defaultdict from the same module is not evaluable at all, so I guess it wouldn't be that mu

[issue5397] PEP 372: OrderedDict

Antoine Pitrou added the comment: In SubclassMappingTests, MyOrderedDict should subclass OrderedDict rather than dict, shouldn't it? -- nosy: +pitrou ___ Python tracker ___ _

[issue5397] PEP 372: OrderedDict

Georg Brandl added the comment: Doc nits: * "items are returned in the order they were first added": it should be made clear that it matters when the *key* was first added * "An *OrderedDict* remembers order that entries were inserted": misses a word somewhere? * "OrderDict" should be "OrderedD

[issue5397] PEP 372: OrderedDict

New submission from Raymond Hettinger : Here is a working patch implementing PEP372 ordered dictionaries. -- components: Library (Lib) files: od.diff keywords: patch messages: 82955 nosy: rhettinger severity: normal stage: patch review status: open title: PEP 372: OrderedDict versions: