the basic idea is to bring topological back down to the raw function,
then the whole UOW constructs itself as very fine grained elements with
full dependencies to each other. then a straight execute with a straight sort.
the hope is that the mechanism here would be vastly simpler. while
the presence of a large number of fine-grained records may be expensive
it still is potentially a lot easier to distill into C code, as the
uow's structure now consists of data.
- Deprecated the hardcoded TIMESTAMP function, which when
used as func.TIMESTAMP(value) would render "TIMESTAMP value".
This breaks on some platforms as Postgres doesn't allow
bind parameters to be used in this context. The hard-coded
uppercase is also inappropriate and there's lots of other
PG casts that we'd need to support. So instead, use
text constructs i.e. select(["timestamp '12/05/09'"]).
and also addresses a significant chunk of py3k deprecations. It's mainly
expicit __hash__ methods. Additionally, most usage of sets/dicts to store columns uses
util-based placeholder names.
- added 2.4-style binops to util.Set on 2.3
- OrderedSets pickle on 2.3
- more lib/sqlalchemy set vs Set corrections
- fixed InstrumentedSet.discard for 2.3
- set, sorted compatibility for test suite
- added testing.fails_if decorator
- Part one of test suite fixes to run on 2.3
Lots of failures still around sets; sets.Set differs from __builtin__.set
particularly in the binops. We depend on set extensively now and may need to
provide a corrected sets.Set subclass on 2.3.
- Importing testenv has no side effects- explicit functions provide similar behavior to the old immediate behavior of testbase
- testing.db has the configured db
- Fixed up the perf/* scripts
- cleanup of mapper._instance, query.instances(). mapper identifies objects which are part of the
current load using a app-unique id on the query context.
- attributes refactor; attributes now mostly use copy-on-modify instead of copy-on-load behavior,
simplified get_history(), added a new set of tests
- fixes to OrderedSet such that difference(), intersection() and others can accept an iterator
- OrderedIdentitySet passes in OrderedSet to the IdentitySet superclass for usage in difference/intersection/etc. operations so that these methods actually work with ordering behavior.
- query.order_by() takes into account aliased joins, i.e. query.join('orders', aliased=True).order_by(Order.id)
- cleanup etc.
- Collections gain a @converter framework for flexible validation and adaptation of bulk assignment
- Bogus bulk assignments now raise TypeError instead of exceptions.ArgumentError
- topological.py cleaned up, presents three public facing functions which
return list/tuple based structures, without exposing any internals. only
the third function returns the "hierarchical" structure. when results
include "cycles" or "child" items, 2- or 3- tuples are used to represent
results.
- unitofwork uses InstanceState almost exclusively now. new and deleted lists
are now dicts which ref the actual object to provide a strong ref for the
duration that they're in those lists. IdentitySet is only used for the public
facing versions of "new" and "deleted".
- unitofwork topological sort no longer uses the "hierarchical" version of the sort
for the base sort, only for the "per-object" secondary sort where it still
helps to group non-dependent operations together and provides expected insert
order. the default sort deals with UOWTasks in a straight list and is greatly
simplified. Tests all pass but need to see if svilen's stuff still works,
one block of code in _sort_cyclical_dependencies() seems to not be needed anywhere
but i definitely put it there for a reason at some point; if not hopefully we
can derive more test coverage from that.
- the UOWEventHandler is only applied to object-storing attributes, not
scalar (i.e. column-based) ones. cuts out a ton of overhead when setting
non-object based attributes.
- InstanceState also used throughout the flush process, i.e. dependency.py,
mapper.save_obj()/delete_obj(), sync.execute() all expect InstanceState objects
in most cases now.
- mapper/property cascade_iterator() takes InstanceState as its argument,
but still returns lists of object instances so that they are not dereferenced.
- a few tricks needed when dealing with InstanceState, i.e. when loading a list
of items that are possibly fresh from the DB, you *have* to get the actual objects
into a strong-referencing datastructure else they fall out of scope immediately.
dependency.py caches lists of dependent objects which it loads now (i.e. history
collections).
- AttributeHistory is gone, replaced by a function that returns a 3-tuple of
added, unchanged, deleted. these collections still reference the object
instances directly for the strong-referencing reasons mentiontioned, but
it uses less IdentitySet logic to generate.
- also omitted all modules and classes that aren't expicitly public
- omitted 'Smallinteger' (small i), but it's still in schema
- omitted NullType-related items from types.__all__
- patched up a few tests to use sql.table and sql.column, other related.
Mike Bayer