|
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224 |
- """
- Code to manage the creation and SQL rendering of 'where' constraints.
- """
-
- from django.db.models.sql.datastructures import EmptyResultSet
- from django.utils import tree
- from django.utils.functional import cached_property
-
- # Connection types
- AND = 'AND'
- OR = 'OR'
-
-
- class WhereNode(tree.Node):
- """
- Used to represent the SQL where-clause.
-
- The class is tied to the Query class that created it (in order to create
- the correct SQL).
-
- A child is usually an expression producing boolean values. Most likely the
- expression is a Lookup instance.
-
- However, a child could also be any class with as_sql() and either
- relabeled_clone() method or relabel_aliases() and clone() methods and
- contains_aggregate attribute.
- """
- default = AND
-
- def split_having(self, negated=False):
- """
- Returns two possibly None nodes: one for those parts of self that
- should be included in the WHERE clause and one for those parts of
- self that must be included in the HAVING clause.
- """
- if not self.contains_aggregate:
- return self, None
- in_negated = negated ^ self.negated
- # If the effective connector is OR and this node contains an aggregate,
- # then we need to push the whole branch to HAVING clause.
- may_need_split = (
- (in_negated and self.connector == AND) or
- (not in_negated and self.connector == OR))
- if may_need_split and self.contains_aggregate:
- return None, self
- where_parts = []
- having_parts = []
- for c in self.children:
- if hasattr(c, 'split_having'):
- where_part, having_part = c.split_having(in_negated)
- if where_part is not None:
- where_parts.append(where_part)
- if having_part is not None:
- having_parts.append(having_part)
- elif c.contains_aggregate:
- having_parts.append(c)
- else:
- where_parts.append(c)
- having_node = self.__class__(having_parts, self.connector, self.negated) if having_parts else None
- where_node = self.__class__(where_parts, self.connector, self.negated) if where_parts else None
- return where_node, having_node
-
- def as_sql(self, compiler, connection):
- """
- Returns the SQL version of the where clause and the value to be
- substituted in. Returns '', [] if this node matches everything,
- None, [] if this node is empty, and raises EmptyResultSet if this
- node can't match anything.
- """
- result = []
- result_params = []
- if self.connector == AND:
- full_needed, empty_needed = len(self.children), 1
- else:
- full_needed, empty_needed = 1, len(self.children)
-
- for child in self.children:
- try:
- sql, params = compiler.compile(child)
- except EmptyResultSet:
- empty_needed -= 1
- else:
- if sql:
- result.append(sql)
- result_params.extend(params)
- else:
- full_needed -= 1
- # Check if this node matches nothing or everything.
- # First check the amount of full nodes and empty nodes
- # to make this node empty/full.
- # Now, check if this node is full/empty using the
- # counts.
- if empty_needed == 0:
- if self.negated:
- return '', []
- else:
- raise EmptyResultSet
- if full_needed == 0:
- if self.negated:
- raise EmptyResultSet
- else:
- return '', []
- conn = ' %s ' % self.connector
- sql_string = conn.join(result)
- if sql_string:
- if self.negated:
- # Some backends (Oracle at least) need parentheses
- # around the inner SQL in the negated case, even if the
- # inner SQL contains just a single expression.
- sql_string = 'NOT (%s)' % sql_string
- elif len(result) > 1:
- sql_string = '(%s)' % sql_string
- return sql_string, result_params
-
- def get_group_by_cols(self):
- cols = []
- for child in self.children:
- cols.extend(child.get_group_by_cols())
- return cols
-
- def relabel_aliases(self, change_map):
- """
- Relabels the alias values of any children. 'change_map' is a dictionary
- mapping old (current) alias values to the new values.
- """
- for pos, child in enumerate(self.children):
- if hasattr(child, 'relabel_aliases'):
- # For example another WhereNode
- child.relabel_aliases(change_map)
- elif hasattr(child, 'relabeled_clone'):
- self.children[pos] = child.relabeled_clone(change_map)
-
- def clone(self):
- """
- Creates a clone of the tree. Must only be called on root nodes (nodes
- with empty subtree_parents). Childs must be either (Contraint, lookup,
- value) tuples, or objects supporting .clone().
- """
- clone = self.__class__._new_instance(
- children=[], connector=self.connector, negated=self.negated)
- for child in self.children:
- if hasattr(child, 'clone'):
- clone.children.append(child.clone())
- else:
- clone.children.append(child)
- return clone
-
- def relabeled_clone(self, change_map):
- clone = self.clone()
- clone.relabel_aliases(change_map)
- return clone
-
- @classmethod
- def _contains_aggregate(cls, obj):
- if isinstance(obj, tree.Node):
- return any(cls._contains_aggregate(c) for c in obj.children)
- return obj.contains_aggregate
-
- @cached_property
- def contains_aggregate(self):
- return self._contains_aggregate(self)
-
-
- class NothingNode(object):
- """
- A node that matches nothing.
- """
- contains_aggregate = False
-
- def as_sql(self, compiler=None, connection=None):
- raise EmptyResultSet
-
-
- class ExtraWhere(object):
- # The contents are a black box - assume no aggregates are used.
- contains_aggregate = False
-
- def __init__(self, sqls, params):
- self.sqls = sqls
- self.params = params
-
- def as_sql(self, compiler=None, connection=None):
- sqls = ["(%s)" % sql for sql in self.sqls]
- return " AND ".join(sqls), list(self.params or ())
-
-
- class SubqueryConstraint(object):
- # Even if aggregates would be used in a subquery, the outer query isn't
- # interested about those.
- contains_aggregate = False
-
- def __init__(self, alias, columns, targets, query_object):
- self.alias = alias
- self.columns = columns
- self.targets = targets
- self.query_object = query_object
-
- def as_sql(self, compiler, connection):
- query = self.query_object
-
- # QuerySet was sent
- if hasattr(query, 'values'):
- if query._db and connection.alias != query._db:
- raise ValueError("Can't do subqueries with queries on different DBs.")
- # Do not override already existing values.
- if query._fields is None:
- query = query.values(*self.targets)
- else:
- query = query._clone()
- query = query.query
- if query.can_filter():
- # If there is no slicing in use, then we can safely drop all ordering
- query.clear_ordering(True)
-
- query_compiler = query.get_compiler(connection=connection)
- return query_compiler.as_subquery_condition(self.alias, self.columns, compiler)
-
- def relabel_aliases(self, change_map):
- self.alias = change_map.get(self.alias, self.alias)
-
- def clone(self):
- return self.__class__(
- self.alias, self.columns, self.targets,
- self.query_object)
|