root/django/trunk/django/db/models/query.py

try:
    set
except NameError:
    from sets import Set as set     # Python 2.3 fallback

from django.db import connection, transaction, IntegrityError
from django.db.models.fields import DateField
from django.db.models.query_utils import Q, select_related_descend
from django.db.models import signals, sql
from django.utils.datastructures import SortedDict


# Used to control how many objects are worked with at once in some cases (e.g.
# when deleting objects).
CHUNK_SIZE = 100
ITER_CHUNK_SIZE = CHUNK_SIZE

# The maximum number of items to display in a QuerySet.__repr__
REPR_OUTPUT_SIZE = 20

# Pull into this namespace for backwards compatibility.
EmptyResultSet = sql.EmptyResultSet


class CyclicDependency(Exception):
    """
    An error when dealing with a collection of objects that have a cyclic
    dependency, i.e. when deleting multiple objects.
    """
    pass


class CollectedObjects(object):
    """
    A container that stores keys and lists of values along with remembering the
    parent objects for all the keys.

    This is used for the database object deletion routines so that we can
    calculate the 'leaf' objects which should be deleted first.
    """

    def __init__(self):
        self.data = {}
        self.children = {}

    def add(self, model, pk, obj, parent_model, nullable=False):
        """
        Adds an item to the container.

        Arguments:
        * model - the class of the object being added.
        * pk - the primary key.
        * obj - the object itself.
        * parent_model - the model of the parent object that this object was
          reached through.
        * nullable - should be True if this relation is nullable.

        Returns True if the item already existed in the structure and
        False otherwise.
        """
        d = self.data.setdefault(model, SortedDict())
        retval = pk in d
        d[pk] = obj
        # Nullable relationships can be ignored -- they are nulled out before
        # deleting, and therefore do not affect the order in which objects
        # have to be deleted.
        if parent_model is not None and not nullable:
            self.children.setdefault(parent_model, []).append(model)
        return retval

    def __contains__(self, key):
        return self.data.__contains__(key)

    def __getitem__(self, key):
        return self.data[key]

    def __nonzero__(self):
        return bool(self.data)

    def iteritems(self):
        for k in self.ordered_keys():
            yield k, self[k]

    def items(self):
        return list(self.iteritems())

    def keys(self):
        return self.ordered_keys()

    def ordered_keys(self):
        """
        Returns the models in the order that they should be dealt with (i.e.
        models with no dependencies first).
        """
        dealt_with = SortedDict()
        # Start with items that have no children
        models = self.data.keys()
        while len(dealt_with) < len(models):
            found = False
            for model in models:
                if model in dealt_with:
                    continue
                children = self.children.setdefault(model, [])
                if len([c for c in children if c not in dealt_with]) == 0:
                    dealt_with[model] = None
                    found = True
            if not found:
                raise CyclicDependency(
                    "There is a cyclic dependency of items to be processed.")

        return dealt_with.keys()

    def unordered_keys(self):
        """
        Fallback for the case where is a cyclic dependency but we don't  care.
        """
        return self.data.keys()


class QuerySet(object):
    """
    Represents a lazy database lookup for a set of objects.
    """
    def __init__(self, model=None, query=None):
        self.model = model
        self.query = query or sql.Query(self.model, connection)
        self._result_cache = None
        self._iter = None
        self._sticky_filter = False

    ########################
    # PYTHON MAGIC METHODS #
    ########################

    def __getstate__(self):
        """
        Allows the QuerySet to be pickled.
        """
        # Force the cache to be fully populated.
        len(self)

        obj_dict = self.__dict__.copy()
        obj_dict['_iter'] = None
        return obj_dict

    def __repr__(self):
        data = list(self[:REPR_OUTPUT_SIZE + 1])
        if len(data) > REPR_OUTPUT_SIZE:
            data[-1] = "...(remaining elements truncated)..."
        return repr(data)

    def __len__(self):
        # Since __len__ is called quite frequently (for example, as part of
        # list(qs), we make some effort here to be as efficient as possible
        # whilst not messing up any existing iterators against the QuerySet.
        if self._result_cache is None:
            if self._iter:
                self._result_cache = list(self._iter)
            else:
                self._result_cache = list(self.iterator())
        elif self._iter:
            self._result_cache.extend(list(self._iter))
        return len(self._result_cache)

    def __iter__(self):
        if self._result_cache is None:
            self._iter = self.iterator()
            self._result_cache = []
        if self._iter:
            return self._result_iter()
        # Python's list iterator is better than our version when we're just
        # iterating over the cache.
        return iter(self._result_cache)

    def _result_iter(self):
        pos = 0
        while 1:
            upper = len(self._result_cache)
            while pos < upper:
                yield self._result_cache[pos]
                pos = pos + 1
            if not self._iter:
                raise StopIteration
            if len(self._result_cache) <= pos:
                self._fill_cache()

    def __nonzero__(self):
        if self._result_cache is not None:
            return bool(self._result_cache)
        try:
            iter(self).next()
        except StopIteration:
            return False
        return True

    def __getitem__(self, k):
        """
        Retrieves an item or slice from the set of results.
        """
        if not isinstance(k, (slice, int, long)):
            raise TypeError
        assert ((not isinstance(k, slice) and (k >= 0))
                or (isinstance(k, slice) and (k.start is None or k.start >= 0)
                    and (k.stop is None or k.stop >= 0))), \
                "Negative indexing is not supported."

        if self._result_cache is not None:
            if self._iter is not None:
                # The result cache has only been partially populated, so we may
                # need to fill it out a bit more.
                if isinstance(k, slice):
                    if k.stop is not None:
                        # Some people insist on passing in strings here.
                        bound = int(k.stop)
                    else:
                        bound = None
                else:
                    bound = k + 1
                if len(self._result_cache) < bound:
                    self._fill_cache(bound - len(self._result_cache))
            return self._result_cache[k]

        if isinstance(k, slice):
            qs = self._clone()
            if k.start is not None:
                start = int(k.start)
            else:
                start = None
            if k.stop is not None:
                stop = int(k.stop)
            else:
                stop = None
            qs.query.set_limits(start, stop)
            return k.step and list(qs)[::k.step] or qs
        try:
            qs = self._clone()
            qs.query.set_limits(k, k + 1)
            return list(qs)[0]
        except self.model.DoesNotExist, e:
            raise IndexError, e.args

    def __and__(self, other):
        self._merge_sanity_check(other)
        if isinstance(other, EmptyQuerySet):
            return other._clone()
        combined = self._clone()
        combined.query.combine(other.query, sql.AND)
        return combined

    def __or__(self, other):
        self._merge_sanity_check(other)
        combined = self._clone()
        if isinstance(other, EmptyQuerySet):
            return combined
        combined.query.combine(other.query, sql.OR)
        return combined

    ####################################
    # METHODS THAT DO DATABASE QUERIES #
    ####################################

    def iterator(self):
        """
        An iterator over the results from applying this QuerySet to the
        database.
        """
        fill_cache = self.query.select_related
        if isinstance(fill_cache, dict):
            requested = fill_cache
        else:
            requested = None
        max_depth = self.query.max_depth
        extra_select = self.query.extra_select.keys()
        index_start = len(extra_select)
        for row in self.query.results_iter():
            if fill_cache:
                obj, _ = get_cached_row(self.model, row, index_start,
                        max_depth, requested=requested)
            else:
                obj = self.model(*row[index_start:])
            for i, k in enumerate(extra_select):
                setattr(obj, k, row[i])
            yield obj

    def count(self):
        """
        Performs a SELECT COUNT() and returns the number of records as an
        integer.

        If the QuerySet is already fully cached this simply returns the length
        of the cached results set to avoid multiple SELECT COUNT(*) calls.
        """
        if self._result_cache is not None and not self._iter:
            return len(self._result_cache)

        return self.query.get_count()

    def get(self, *args, **kwargs):
        """
        Performs the query and returns a single object matching the given
        keyword arguments.
        """
        clone = self.filter(*args, **kwargs)
        num = len(clone)
        if num == 1:
            return clone._result_cache[0]
        if not num:
            raise self.model.DoesNotExist("%s matching query does not exist."
                    % self.model._meta.object_name)
        raise self.model.MultipleObjectsReturned("get() returned more than one %s -- it returned %s! Lookup parameters were %s"
                % (self.model._meta.object_name, num, kwargs))

    def create(self, **kwargs):
        """
        Creates a new object with the given kwargs, saving it to the database
        and returning the created object.
        """
        obj = self.model(**kwargs)
        obj.save(force_insert=True)
        return obj

    def get_or_create(self, **kwargs):
        """
        Looks up an object with the given kwargs, creating one if necessary.
        Returns a tuple of (object, created), where created is a boolean
        specifying whether an object was created.
        """
        assert kwargs, \
                'get_or_create() must be passed at least one keyword argument'
        defaults = kwargs.pop('defaults', {})
        try:
            return self.get(**kwargs), False
        except self.model.DoesNotExist:
            try:
                params = dict([(k, v) for k, v in kwargs.items() if '__' not in k])
                params.update(defaults)
                obj = self.model(**params)
                sid = transaction.savepoint()
                obj.save(force_insert=True)
                transaction.savepoint_commit(sid)
                return obj, True
            except IntegrityError, e:
                transaction.savepoint_rollback(sid)
                try:
                    return self.get(**kwargs), False
                except self.model.DoesNotExist:
                    raise e

    def latest(self, field_name=None):
        """
        Returns the latest object, according to the model's 'get_latest_by'
        option or optional given field_name.
        """
        latest_by = field_name or self.model._meta.get_latest_by
        assert bool(latest_by), "latest() requires either a field_name parameter or 'get_latest_by' in the model"
        assert self.query.can_filter(), \
                "Cannot change a query once a slice has been taken."
        obj = self._clone()
        obj.query.set_limits(high=1)
        obj.query.add_ordering('-%s' % latest_by)
        return obj.get()

    def in_bulk(self, id_list):
        """
        Returns a dictionary mapping each of the given IDs to the object with
        that ID.
        """
        assert self.query.can_filter(), \
                "Cannot use 'limit' or 'offset' with in_bulk"
        assert isinstance(id_list, (tuple,  list)), \
                "in_bulk() must be provided with a list of IDs."
        if not id_list:
            return {}
        qs = self._clone()
        qs.query.add_filter(('pk__in', id_list))
        return dict([(obj._get_pk_val(), obj) for obj in qs.iterator()])

    def delete(self):
        """
        Deletes the records in the current QuerySet.
        """
        assert self.query.can_filter(), \
                "Cannot use 'limit' or 'offset' with delete."

        del_query = self._clone()

        # Disable non-supported fields.
        del_query.query.select_related = False
        del_query.query.clear_ordering()

        # Delete objects in chunks to prevent the list of related objects from
        # becoming too long.
        while 1:
            # Collect all the objects to be deleted in this chunk, and all the
            # objects that are related to the objects that are to be deleted.
            seen_objs = CollectedObjects()
            for object in del_query[:CHUNK_SIZE]:
                object._collect_sub_objects(seen_objs)

            if not seen_objs:
                break
            delete_objects(seen_objs)

        # Clear the result cache, in case this QuerySet gets reused.
        self._result_cache = None
    delete.alters_data = True

    def update(self, **kwargs):
        """
        Updates all elements in the current QuerySet, setting all the given
        fields to the appropriate values.
        """
        assert self.query.can_filter(), \
                "Cannot update a query once a slice has been taken."
        query = self.query.clone(sql.UpdateQuery)
        query.add_update_values(kwargs)
        rows = query.execute_sql(None)
        transaction.commit_unless_managed()
        self._result_cache = None
        return rows
    update.alters_data = True

    def _update(self, values):
        """
        A version of update that accepts field objects instead of field names.
        Used primarily for model saving and not intended for use by general
        code (it requires too much poking around at model internals to be
        useful at that level).
        """
        assert self.query.can_filter(), \
                "Cannot update a query once a slice has been taken."
        query = self.query.clone(sql.UpdateQuery)
        query.add_update_fields(values)
        self._result_cache = None
        return query.execute_sql(None)
    _update.alters_data = True

    ##################################################
    # PUBLIC METHODS THAT RETURN A QUERYSET SUBCLASS #
    ##################################################

    def values(self, *fields):
        return self._clone(klass=ValuesQuerySet, setup=True, _fields=fields)

    def values_list(self, *fields, **kwargs):
        flat = kwargs.pop('flat', False)
        if kwargs:
            raise TypeError('Unexpected keyword arguments to values_list: %s'
                    % (kwargs.keys(),))
        if flat and len(fields) > 1:
            raise TypeError("'flat' is not valid when values_list is called with more than one field.")
        return self._clone(klass=ValuesListQuerySet, setup=True, flat=flat,
                _fields=fields)

    def dates(self, field_name, kind, order='ASC'):
        """
        Returns a list of datetime objects representing all available dates for
        the given field_name, scoped to 'kind'.
        """
        assert kind in ("month", "year", "day"), \
                "'kind' must be one of 'year', 'month' or 'day'."
        assert order in ('ASC', 'DESC'), \
                "'order' must be either 'ASC' or 'DESC'."
        return self._clone(klass=DateQuerySet, setup=True,
                _field_name=field_name, _kind=kind, _order=order)

    def none(self):
        """
        Returns an empty QuerySet.
        """
        return self._clone(klass=EmptyQuerySet)

    ##################################################################
    # PUBLIC METHODS THAT ALTER ATTRIBUTES AND RETURN A NEW QUERYSET #
    ##################################################################

    def all(self):
        """
        Returns a new QuerySet that is a copy of the current one. This allows a
        QuerySet to proxy for a model manager in some cases.
        """
        return self._clone()

    def filter(self, *args, **kwargs):
        """
        Returns a new QuerySet instance with the args ANDed to the existing
        set.
        """
        return self._filter_or_exclude(False, *args, **kwargs)

    def exclude(self, *args, **kwargs):
        """
        Returns a new QuerySet instance with NOT (args) ANDed to the existing
        set.
        """
        return self._filter_or_exclude(True, *args, **kwargs)

    def _filter_or_exclude(self, negate, *args, **kwargs):
        if args or kwargs:
            assert self.query.can_filter(), \
                    "Cannot filter a query once a slice has been taken."

        clone = self._clone()
        if negate:
            clone.query.add_q(~Q(*args, **kwargs))
        else:
            clone.query.add_q(Q(*args, **kwargs))
        return clone

    def complex_filter(self, filter_obj):
        """
        Returns a new QuerySet instance with filter_obj added to the filters.

        filter_obj can be a Q object (or anything with an add_to_query()
        method) or a dictionary of keyword lookup arguments.

        This exists to support framework features such as 'limit_choices_to',
        and usually it will be more natural to use other methods.
        """
        if isinstance(filter_obj, Q) or hasattr(filter_obj, 'add_to_query'):
            clone = self._clone()
            clone.query.add_q(filter_obj)
            return clone
        else:
            return self._filter_or_exclude(None, **filter_obj)

    def select_related(self, *fields, **kwargs):
        """
        Returns a new QuerySet instance that will select related objects.

        If fields are specified, they must be ForeignKey fields and only those
        related objects are included in the selection.
        """
        depth = kwargs.pop('depth', 0)
        if kwargs:
            raise TypeError('Unexpected keyword arguments to select_related: %s'
                    % (kwargs.keys(),))
        obj = self._clone()
        if fields:
            if depth:
                raise TypeError('Cannot pass both "depth" and fields to select_related()')
            obj.query.add_select_related(fields)
        else:
            obj.query.select_related = True
        if depth:
            obj.query.max_depth = depth
        return obj

    def dup_select_related(self, other):
        """
        Copies the related selection status from the QuerySet 'other' to the
        current QuerySet.
        """
        self.query.select_related = other.query.select_related

    def order_by(self, *field_names):
        """
        Returns a new QuerySet instance with the ordering changed.
 &nb