Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

ChainMap should preserve order of underlying mappings #76973

Closed
rhettinger opened this issue Feb 8, 2018 · 10 comments
Closed

ChainMap should preserve order of underlying mappings #76973

rhettinger opened this issue Feb 8, 2018 · 10 comments
Labels
3.7 (EOL) end of life 3.8 only security fixes stdlib Python modules in the Lib dir type-bug An unexpected behavior, bug, or error

Comments

@rhettinger
Copy link
Contributor

BPO 32792
Nosy @rhettinger, @ned-deily, @serhiy-storchaka
PRs
  • bpo-32792: Preserve mapping order in ChainMap() #5586
  • [3.7] bpo-32792: Preserve mapping order in ChainMap() (GH-5586) #5617
  • Note: these values reflect the state of the issue at the time it was migrated and might not reflect the current state.

    Show more details

    GitHub fields:

    assignee = None
    closed_at = <Date 2018-02-11.09:29:52.688>
    created_at = <Date 2018-02-08.08:46:17.632>
    labels = ['3.7', '3.8', 'type-bug', 'library']
    title = 'ChainMap should preserve order of underlying mappings'
    updated_at = <Date 2018-02-11.09:29:52.687>
    user = 'https://github.com/rhettinger'

    bugs.python.org fields:

    activity = <Date 2018-02-11.09:29:52.687>
    actor = 'rhettinger'
    assignee = 'none'
    closed = True
    closed_date = <Date 2018-02-11.09:29:52.688>
    closer = 'rhettinger'
    components = ['Library (Lib)']
    creation = <Date 2018-02-08.08:46:17.632>
    creator = 'rhettinger'
    dependencies = []
    files = []
    hgrepos = []
    issue_num = 32792
    keywords = ['patch']
    message_count = 10.0
    messages = ['311817', '311822', '311852', '311969', '311984', '311985', '311987', '311988', '311989', '311990']
    nosy_count = 3.0
    nosy_names = ['rhettinger', 'ned.deily', 'serhiy.storchaka']
    pr_nums = ['5586', '5617']
    priority = 'normal'
    resolution = 'fixed'
    stage = 'resolved'
    status = 'closed'
    superseder = None
    type = 'behavior'
    url = 'https://bugs.python.org/issue32792'
    versions = ['Python 3.7', 'Python 3.8']

    @rhettinger
    Copy link
    Contributor Author

    This also applies to 3.6 because ChainMap can be used with OrderedDict.

    @rhettinger rhettinger added 3.7 (EOL) end of life 3.8 only security fixes stdlib Python modules in the Lib dir type-bug An unexpected behavior, bug, or error labels Feb 8, 2018
    @serhiy-storchaka
    Copy link
    Member

    An alternate implementation:

    d = {}
    for mapping in reversed(self.maps):
        d.update(mapping)
    return iter(d)

    Unfortunately both implementations work only with hashable keys. In general case mappings may be not hashtables.

    @ned-deily
    Copy link
    Member

    See discussion on PR 5586 regarding backporting to 3.6.x.

    @serhiy-storchaka
    Copy link
    Member

    Sorry, I was wrong. reversed() is not needed here.

    The advantage of this implementation is that it can be faster because of iterating mappings in C code instead of Python code. But the side effect of it is that the iterator keeps references to all values. If this is not desirable, the code can be written something like:

        return iter(dict.fromkeys(itertools.chain.from_iterable(self.maps)))

    @rhettinger
    Copy link
    Contributor Author

    New changeset 3793f95 by Raymond Hettinger in branch 'master':
    bpo-32792: Preserve mapping order in ChainMap() (GH-5586)
    3793f95

    @serhiy-storchaka
    Copy link
    Member

    On PR 5586 we discussed reverting the order of the iteration by mappings. There are reasons of doing this. But this adds a subtle behavior change. Currently list(ChainMap({1: int}, {1.0: float})) returns [1]. With reversed() it will return [1.0]. This change LGTM.

    @serhiy-storchaka
    Copy link
    Member

    The code in msg311969 doesn't reuse hash values. The following implementations do this:

    return iter({**m for m in reversed(self.maps)})
    

    or, without keeping reverences to values:

    return iter(list({**m for m in reversed(self.maps)}))
    

    @rhettinger
    Copy link
    Contributor Author

    New changeset 170b3f7 by Raymond Hettinger (Miss Islington (bot)) in branch '3.7':
    bpo-32792: Preserve mapping order in ChainMap() (GH-5586) (#GH-5617)
    170b3f7

    @rhettinger
    Copy link
    Contributor Author

    The code in msg311969 doesn't reuse hash values.

    That doesn't make sense. The dict.update() method reuses the hashes of the input mappings when possible.

        >>> from collections import ChainMap
        >>> class Int(int):
                def __hash__(self):
                        print(f'Hashing {self}', file=sys.stderr)
                        return int.__hash__(self)
    
                
        >>> import sys
        >>> d = { Int(1): 'f1', Int(2): 'f2' }
        Hashing 1
        Hashing 2
        >>> e = { Int(1): 's1', Int(3): 's3' }
        Hashing 1
        Hashing 3
        >>> c = ChainMap(d, e)
        >>> list(c)                     # Note, no calls to hash() were made
        [1, 3, 2]

    @serhiy-storchaka
    Copy link
    Member

    I referred to msg311969, not msg311822.

    @ezio-melotti ezio-melotti transferred this issue from another repository Apr 10, 2022
    Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
    Labels
    3.7 (EOL) end of life 3.8 only security fixes stdlib Python modules in the Lib dir type-bug An unexpected behavior, bug, or error
    Projects
    None yet
    Development

    No branches or pull requests

    3 participants