]> git.ipfire.org Git - thirdparty/sqlalchemy/sqlalchemy.git/log
thirdparty/sqlalchemy/sqlalchemy.git
18 years agosome more docstring patches for [ticket:214]
Mike Bayer [Wed, 28 Mar 2007 00:23:39 +0000 (00:23 +0000)] 
some more docstring patches for [ticket:214]

18 years ago- fix for fetchmany() "size" argument being positional in most
Mike Bayer [Wed, 28 Mar 2007 00:09:55 +0000 (00:09 +0000)] 
- fix for fetchmany() "size" argument being positional in most
dbapis [ticket:505]

18 years ago- fixes [ticket:185], join object determines primary key and removes
Mike Bayer [Tue, 27 Mar 2007 22:06:36 +0000 (22:06 +0000)] 
- fixes [ticket:185], join object determines primary key and removes
columns that are FK's to other columns in the primary key collection.
- removed workaround code from query.py get()
- removed obsolete inheritance test from mapper
- added new get() test to inheritance.py for this particular issue
- ColumnCollection has nicer string method

18 years ago- some logging cleanup
Mike Bayer [Tue, 27 Mar 2007 21:01:32 +0000 (21:01 +0000)] 
- some logging cleanup
- added 'encodedname' prop to a few ClauseElements to aid logging

18 years ago- preliminary support for unicode table and column names added.
Mike Bayer [Tue, 27 Mar 2007 16:04:34 +0000 (16:04 +0000)] 
- preliminary support for unicode table and column names added.

18 years agomsssql: more unit tests now pass
Rick Morrison [Tue, 27 Mar 2007 05:07:10 +0000 (05:07 +0000)] 
msssql: more unit tests now pass

18 years ago- improved/fixed custom collection classes when giving it "set"/
Mike Bayer [Mon, 26 Mar 2007 19:59:39 +0000 (19:59 +0000)] 
- improved/fixed custom collection classes when giving it "set"/
"sets.Set" classes or subclasses (was still looking for append()
methods on them during lazy loads)
- moved CustomCollectionsTest from unitofwork to relationships
- added more custom collections test to attributes module

18 years ago- column labels are now generated in the compilation phase, which
Mike Bayer [Sat, 24 Mar 2007 19:24:27 +0000 (19:24 +0000)] 
- column labels are now generated in the compilation phase, which
means their lengths are dialect-dependent.  So on oracle a label
that gets truncated to 30 chars will go out to 63 characters
on postgres.  Also, the true labelname is always attached as the
accessor on the parent Selectable so theres no need to be aware
of the genrerated label names [ticket:512].
- ResultProxy column targeting is greatly simplified, and relies
upon the ANSICompiler's column_labels map to translate the built-in
label on a _ColumnClause (which is now considered to be a unique
identifier of that column) to the label which was generated at compile
time.
- still need to put a baseline of ColumnClause targeting for
ResultProxy objects that originated from a textual query.

18 years agocleanup continued rel_0_3_6
Mike Bayer [Fri, 23 Mar 2007 21:38:12 +0000 (21:38 +0000)] 
cleanup continued

18 years agovarious cleanup, docs and things, getting ready for 0.3.6
Mike Bayer [Fri, 23 Mar 2007 21:33:24 +0000 (21:33 +0000)] 
various cleanup, docs and things, getting ready for 0.3.6

18 years agodocumenting generative methods on query
Mike Bayer [Thu, 22 Mar 2007 23:24:28 +0000 (23:24 +0000)] 
documenting generative methods on query

18 years ago- MetaData can bind to an engine either via "url" or "engine" kwargs
Mike Bayer [Thu, 22 Mar 2007 21:18:23 +0000 (21:18 +0000)] 
- MetaData can bind to an engine either via "url" or "engine" kwargs
to constructor, or by using connect() method.  BoundMetaData is
identical to MetaData except engine_or_url param is required.
DynamicMetaData is the same and provides thread-local connections
be default.

18 years agodan's latest patch for session.identity_key()
Mike Bayer [Thu, 22 Mar 2007 20:54:52 +0000 (20:54 +0000)] 
dan's latest patch for session.identity_key()

18 years ago- CLOB type descends from TEXT so it goes to the dialect correctly
Mike Bayer [Thu, 22 Mar 2007 17:58:30 +0000 (17:58 +0000)] 
- CLOB type descends from TEXT so it goes to the dialect correctly
- oracle CLOB has result value LOB handling

18 years ago- added explicit MSTimeStamp type which takes effect when using
Mike Bayer [Thu, 22 Mar 2007 16:50:38 +0000 (16:50 +0000)] 
- added explicit MSTimeStamp type which takes effect when using
types.TIMESTAMP.

18 years ago- slightly better support for bind params as column clauses, either
Mike Bayer [Wed, 21 Mar 2007 20:43:34 +0000 (20:43 +0000)] 
- slightly better support for bind params as column clauses, either
via bindparam() or via literal(), i.e. select([literal('foo')])
- removed "table" argument from column().  this does not add the column
to the table anyway so was misleading.
- Select _exportable_columns() only exports Selectable instances
- Select uses _exportable_columns() when searching for engines
instead of _raw_columns for similar reasons (non selectables have no engine)
- _BindParamClause no longer has a _make_proxy().  its not a ColumnElement.
- _Label detects underlying column element and will generate its own
column()._make_proxy() if the element is not a ColumnElement.  this
allows a Label to be declared for nearly anything and it can export
itself as a column on a containing Selectable.

18 years agocontextual_connection() -> contextual_connect() [ticket:515]
Mike Bayer [Tue, 20 Mar 2007 23:29:31 +0000 (23:29 +0000)] 
contextual_connection()  -> contextual_connect() [ticket:515]

18 years agothe "tack on the leftover tasks at the end" step of the "circular dependency sort"
Mike Bayer [Tue, 20 Mar 2007 15:27:34 +0000 (15:27 +0000)] 
the "tack on the leftover tasks at the end" step of the "circular dependency sort"
makes a copy of those tasks with the circular_parent marked.  this way the tasks
do not iterate through their child items polymorphically, which is necessary because
the "circular sort" stores individual subclass tasks separately (i.e. saving/deleting
should not traverse polymorhically for those tasks)

18 years ago- many-to-many table will be properly handled even for operations that
Mike Bayer [Mon, 19 Mar 2007 17:54:29 +0000 (17:54 +0000)] 
- many-to-many table will be properly handled even for operations that
occur on the "backref" side of the operation [ticket:249]

18 years agomssql: now passes still more unit tests, [ticket:481]
Rick Morrison [Mon, 19 Mar 2007 02:00:32 +0000 (02:00 +0000)] 
mssql: now passes still more unit tests, [ticket:481]

18 years ago- added db modules to genned docstrings
Mike Bayer [Sun, 18 Mar 2007 22:35:19 +0000 (22:35 +0000)] 
- added db modules to genned docstrings
- had to tweak out latest MS-SQL module change.  cant do ImportErrors right now until module
importing is moved to the connection phase across all dialects.
- took out "his" from url docstrings
- postgres doesnt do an import *

18 years ago- check for tables in the primaryjoin/secondaryjoin that arent parent of parent/child...
Mike Bayer [Sun, 18 Mar 2007 21:45:55 +0000 (21:45 +0000)] 
- check for tables in the primaryjoin/secondaryjoin that arent parent of parent/child mappers.
dont include those clauses when looking for foreign_keys (which also takes care of remote_side).
if those cols are present in foreign_keys, lazyloader makes binds out of them and tries to
target those columns on the mapper, raising either the "conflicting column" error if they have the same
name, or the "cant find column on mapping" if it has a unique name.  added tests for both.

18 years agomssql: cleanup of module importing code; specifiable DB-API module; more explicit...
Rick Morrison [Sun, 18 Mar 2007 17:14:10 +0000 (17:14 +0000)] 
mssql: cleanup of module importing code; specifiable DB-API module; more explicit ordering of module preferences. [ticket:480]

18 years agomssql: optionally use VARCHAR(max) instead of TEXT. [ticket:509]
Rick Morrison [Sun, 18 Mar 2007 16:25:19 +0000 (16:25 +0000)] 
mssql: optionally use VARCHAR(max) instead of TEXT. [ticket:509]

18 years ago- flush fixes on self-referential relationships that contain references
Mike Bayer [Sat, 17 Mar 2007 20:46:52 +0000 (20:46 +0000)] 
- flush fixes on self-referential relationships that contain references
to other instances outside of the cyclical chain, when the initial
self-referential objects are not actually part of the flush

18 years agocss tag
Mike Bayer [Sat, 17 Mar 2007 16:58:12 +0000 (16:58 +0000)] 
css tag

18 years agointegrated docutils formatting into generated documentation;
Mike Bayer [Sat, 17 Mar 2007 16:49:45 +0000 (16:49 +0000)] 
integrated docutils formatting into generated documentation;
restructuredtext fixes throughout docstrings

18 years agomerged mako doc generation branch
Mike Bayer [Sat, 17 Mar 2007 15:14:53 +0000 (15:14 +0000)] 
merged mako doc generation branch

18 years ago- added a catchall **kwargs to MSString, to help reflection of
Mike Bayer [Fri, 16 Mar 2007 01:48:28 +0000 (01:48 +0000)] 
- added a catchall **kwargs to MSString, to help reflection of
obscure types (like "varchar() binary" in MS 4.0)

18 years ago"alltests" runners call testbase.main(), which takes an optional suite,
Mike Bayer [Thu, 15 Mar 2007 18:11:05 +0000 (18:11 +0000)] 
"alltests" runners call testbase.main(), which takes an optional suite,
so that exit code is propigated

18 years agoliterals in PassiveDefault require text()
Mike Bayer [Thu, 15 Mar 2007 16:49:47 +0000 (16:49 +0000)] 
literals in PassiveDefault require text()

18 years agoMSSQL now passes still more unit tests [ticket:481]
Rick Morrison [Thu, 15 Mar 2007 02:31:15 +0000 (02:31 +0000)] 
MSSQL now passes still more unit tests [ticket:481]
Fix to null FLOAT fields in mssql-trusted.patch
MSSQL: LIMIT with OFFSET now raises an error
MSSQL: can now specify Windows authorization
MSSQL: ignores seconds on DATE columns (DATE fix, part 1)

18 years agofix CASE statement when else_ is zero
Rick Morrison [Thu, 15 Mar 2007 01:58:46 +0000 (01:58 +0000)] 
fix CASE statement when else_ is zero

18 years ago- eager loading will not "aliasize" "order by" clauses that were placed
Mike Bayer [Wed, 14 Mar 2007 23:48:07 +0000 (23:48 +0000)] 
- eager loading will not "aliasize" "order by" clauses that were placed
in the select statement by something other than the eager loader
itself, to fix possibility of dupe columns as illustrated in
[ticket:495].  however, this means you have to be more careful with
the columns placed in the "order by" of Query.select(), that you have
explicitly named them in your criterion (i.e. you cant rely on the
eager loader adding them in for you)

- query._join_to (which powers join, join_via, etc) properly takes
secondary table into account when constructing joins

18 years ago- added a handy multi-use "identity_key()" method to Session, allowing
Mike Bayer [Tue, 13 Mar 2007 03:45:28 +0000 (03:45 +0000)] 
- added a handy multi-use "identity_key()" method to Session, allowing
the generation of identity keys for primary key values, instances,
and rows, courtesy Daniel Miller

18 years agosome docs
Mike Bayer [Tue, 13 Mar 2007 02:45:40 +0000 (02:45 +0000)] 
some docs

18 years agosome formatting
Mike Bayer [Tue, 13 Mar 2007 02:37:44 +0000 (02:37 +0000)] 
some formatting

18 years ago- for hackers, refactored the "visitor" system of ClauseElement and
Mike Bayer [Sun, 11 Mar 2007 20:52:02 +0000 (20:52 +0000)] 
- for hackers, refactored the "visitor" system of ClauseElement and
SchemaItem so that the traversal of items is controlled by the
ClauseVisitor itself, using the method visitor.traverse(item).
accept_visitor() methods can still be called directly but will
not do any traversal of child items.  ClauseElement/SchemaItem now
have a configurable get_children() method to return the collection
of child elements for each parent object. This allows the full
traversal of items to be clear and unambiguous (as well as loggable),
with an easy method of limiting a traversal (just pass flags which
are picked up by appropriate get_children() methods). [ticket:501]
- accept_schema_visitor() methods removed, replaced with
get_children(schema_visitor=True)
- various docstring/changelog cleanup/reformatting

18 years ago- oracle:
Mike Bayer [Sat, 10 Mar 2007 23:31:40 +0000 (23:31 +0000)] 
- oracle:
    - got binary working for any size input !  cx_oracle works fine,
      it was my fault as BINARY was being passed and not BLOB for
      setinputsizes (also unit tests werent even setting input sizes).
    - auto_setinputsizes defaults to True for Oracle, fixed cases where
      it improperly propigated bad types.

18 years ago- Query has add_entity() and add_column() generative methods. these
Mike Bayer [Sat, 10 Mar 2007 03:41:55 +0000 (03:41 +0000)] 
- Query has add_entity() and add_column() generative methods.  these
will add the given mapper/class or ColumnElement to the query at compile
time, and apply them to the instances method.  the user is responsible
for constructing reasonable join conditions (otherwise you can get
full cartesian products).  result set is the list of tuples, non-uniqued.
- fixed multi-mapper instances() to pad out shorter results with None so
zip() gets everything

18 years ago- the full featureset of the SelectResults extension has been merged
Mike Bayer [Sat, 10 Mar 2007 02:49:12 +0000 (02:49 +0000)] 
- the full featureset of the SelectResults extension has been merged
into a new set of methods available off of Query.  These methods
all provide "generative" behavior, whereby the Query is copied
and a new one returned with additional criterion added.
The new methods include:

  filter() - applies select criterion to the query
  filter_by() - applies "by"-style criterion to the query
  avg() - return the avg() function on the given column
  join() - join to a property (or across a list of properties)
  outerjoin() - like join() but uses LEFT OUTER JOIN
  limit()/offset() - apply LIMIT/OFFSET
  range-based access which applies limit/offset:
     session.query(Foo)[3:5]
  distinct() - apply DISTINCT
  list() - evaluate the criterion and return results

no incompatible changes have been made to Query's API and no methods
have been deprecated.  Existing methods like select(), select_by(),
get(), get_by() all execute the query at once and return results
like they always did.  join_to()/join_via() are still there although
the generative join()/outerjoin() methods are easier to use.

- the return value for multiple mappers used with instances() now returns
a cartesian product of the requested list of mappers, represented
as a list of tuples.  this corresponds to the documented behavior.
So that instances match up properly, the "uniquing" is disabled when
this feature is used.
- strings and columns can also be sent to the *args of instances() where
those exact result columns will be part of the result tuples.
- query() method is added by assignmapper.  this helps with
navigating to all the new generative methods on Query.

18 years agocleanup of reversed
Mike Bayer [Fri, 9 Mar 2007 00:12:49 +0000 (00:12 +0000)] 
cleanup of reversed

18 years ago- fixed usage of 2.4-only "reversed" in topological.py [ticket:506]
Mike Bayer [Thu, 8 Mar 2007 23:14:25 +0000 (23:14 +0000)] 
- fixed usage of 2.4-only "reversed" in topological.py [ticket:506]

18 years ago- fixed use_alter flag on ForeignKeyConstraint [ticket:503]
Mike Bayer [Wed, 7 Mar 2007 18:05:39 +0000 (18:05 +0000)] 
- fixed use_alter flag on ForeignKeyConstraint [ticket:503]

18 years ago- options() method on SelectResults now implemented "generatively"
Mike Bayer [Wed, 7 Mar 2007 01:31:41 +0000 (01:31 +0000)] 
- options() method on SelectResults now implemented "generatively"
like the rest of the SelectResults methods [ticket:472]

18 years agoneed an "alias()" on map to a select
Mike Bayer [Tue, 6 Mar 2007 18:39:27 +0000 (18:39 +0000)] 
need an "alias()" on map to a select

18 years agodecorated ImportError thrown when the <database>:// module isnt found
Mike Bayer [Tue, 6 Mar 2007 17:17:27 +0000 (17:17 +0000)] 
decorated ImportError thrown when the <database>:// module isnt found

18 years agoadded "enable_typechecks=True" flag on relation so the new type check from #500 can...
Mike Bayer [Tue, 6 Mar 2007 16:47:43 +0000 (16:47 +0000)] 
added "enable_typechecks=True" flag on relation so the new type check from #500 can be disabled, since people are going to want to disable it.

18 years ago- added concept of 'require_embedded' to corresponding_column.
Mike Bayer [Mon, 5 Mar 2007 23:08:52 +0000 (23:08 +0000)] 
- added concept of 'require_embedded' to corresponding_column.
requires that the target column be present in a sub-element of the
target selectable.
- embedded logic above more appropriate for ClauseAdapter functionality
since its trying to "pull up" clauses that represent columns within
a larger union up to the level of the union itself.
- the "direction" test against the "foreign_keys" collection apparently
works for an exact "column 'x' is present in the collection", no proxy
relationships needed.  fixes the case of relating a selectable/alias
to one of its underlying tables, probably fixes other scenarios

18 years ago- put an aggressive check for "flushing object A with a collection
Mike Bayer [Mon, 5 Mar 2007 05:12:09 +0000 (05:12 +0000)] 
- put an aggressive check for "flushing object A with a collection
of B's, but you put a C in the collection" error condition -
**even if C is a subclass of B**, unless B's mapper loads polymorphically.
Otherwise, the collection will later load a "B" which should be a "C"
(since its not polymorphic) which breaks in bi-directional relationships
(i.e. C has its A, but A's backref will lazyload it as a different
instance of type "B") [ticket:500]

18 years agodont continue remote table if warning
Mike Bayer [Sun, 4 Mar 2007 18:09:43 +0000 (18:09 +0000)] 
dont continue remote table if warning

18 years ago"modernized" polymorph test, name change to "test_roundtrip"
Mike Bayer [Sat, 3 Mar 2007 21:25:44 +0000 (21:25 +0000)] 
"modernized" polymorph test, name change to "test_roundtrip"

18 years ago- bindparam() names are now repeatable! specify two
Mike Bayer [Sat, 3 Mar 2007 21:02:26 +0000 (21:02 +0000)] 
- bindparam() names are now repeatable!  specify two
distinct bindparam()s with the same name in a single statement,
and the key will be shared.  proper positional/named args translate
at compile time.  for the old behavior of "aliasing" bind parameters
with conflicting names, specify "unique=True" - this option is
still used internally for all the auto-genererated (value-based)
     bind parameters.

18 years agocleanup; removed "separate foreign key" tests, polymorphic joined-table inheritance...
Mike Bayer [Sat, 3 Mar 2007 20:24:13 +0000 (20:24 +0000)] 
cleanup; removed "separate foreign key" tests, polymorphic joined-table inheritance requires the same pk col name across tables;
added additional polymorphic load assertions

18 years ago- added "fold_equivalents" argument to Join.select(), which removes
Mike Bayer [Fri, 2 Mar 2007 21:22:14 +0000 (21:22 +0000)] 
- added "fold_equivalents" argument to Join.select(), which removes
'duplicate' columns from the resulting column clause that are known to be
equivalent based on the join condition.  this is of great usage when
constructing subqueries of joins which Postgres complains about if
duplicate column names are present.
- added support to polymorphic stuff for more than one "equivalent column",
when trying to target columns in the polymorphic union; this applies
to multi-leveled inheritance
- put above-two concepts together to get the abc_inheritance tests to work
with postgres

18 years ago- use_labels flag on select() wont auto-create labels for literal text
Mike Bayer [Thu, 1 Mar 2007 20:14:17 +0000 (20:14 +0000)] 
- use_labels flag on select() wont auto-create labels for literal text
column elements, since we can make no assumptions about the text. to
create labels for literal columns, you can say "somecol AS somelabel",
or use literal_column("somecol").label("somelabel")
- quoting wont occur for literal columns when they are "proxied" into the
column collection for their selectable (is_literal flag is propigated)

18 years agoadded a unit test for nested session transactions
Mike Bayer [Wed, 28 Feb 2007 21:42:54 +0000 (21:42 +0000)] 
added a unit test for nested session transactions

18 years agoa generating testcase that tests a three-level inheritance chain (A->B->C) and all...
Mike Bayer [Wed, 28 Feb 2007 03:35:32 +0000 (03:35 +0000)] 
a generating testcase that tests a three-level inheritance chain (A->B->C) and all one-to-many and many-to-one relationships between
all distinct A,B,C.  still needs support for foreign key to parent table not the same col as the pk col

18 years agothis test works with both one to many and many to one, but i think its intended to...
Mike Bayer [Wed, 28 Feb 2007 01:25:52 +0000 (01:25 +0000)] 
this test works with both one to many and many to one, but i think its intended to be one-to-many

18 years ago- fixed function execution with explicit connections, when you dont
Mike Bayer [Tue, 27 Feb 2007 19:04:43 +0000 (19:04 +0000)] 
- fixed function execution with explicit connections, when you dont
explicitly say "select()" off the function, i.e.
conn.execute(func.dosomething())

18 years agomigrated (most) docstrings to pep-257 format, docstring generator using straight...
Mike Bayer [Sun, 25 Feb 2007 22:44:52 +0000 (22:44 +0000)] 
migrated (most) docstrings to pep-257 format, docstring generator using straight <pre> + trim() func
for now.  applies most of [ticket:214], compliemnts of Lele Gaifax

18 years ago- more fixes to polymorphic relations, involving proper lazy-clause
Mike Bayer [Sun, 25 Feb 2007 01:27:15 +0000 (01:27 +0000)] 
- more fixes to polymorphic relations, involving proper lazy-clause
generation on many-to-one relationships to polymorphic mappers
[ticket:493]

18 years agofix typo
Mike Bayer [Sun, 25 Feb 2007 01:24:33 +0000 (01:24 +0000)] 
fix typo

18 years ago- added "refresh-expire" cascade [ticket:492]
Mike Bayer [Sat, 24 Feb 2007 21:58:38 +0000 (21:58 +0000)] 
- added "refresh-expire" cascade [ticket:492]

18 years ago- correlated subqueries work inside of ORDER BY, GROUP BY
Mike Bayer [Sat, 24 Feb 2007 00:17:08 +0000 (00:17 +0000)] 
- correlated subqueries work inside of ORDER BY, GROUP BY

18 years ago- exists() becomes useable as a standalone selectable, not just in a
Mike Bayer [Fri, 23 Feb 2007 20:46:27 +0000 (20:46 +0000)] 
- exists() becomes useable as a standalone selectable, not just in a
WHERE clause

18 years ago- a full select() construct can be passed to query.select() (which
Mike Bayer [Thu, 22 Feb 2007 22:59:51 +0000 (22:59 +0000)] 
- a full select() construct can be passed to query.select() (which
worked anyway), but also query.selectfirst(), query.selectone() which
will be used as is (i.e. no query is compiled). works similarly to
sending the results to instances().

18 years agorestored sequence back, needed by PG for the unit test rel_0_3_5
Mike Bayer [Thu, 22 Feb 2007 03:51:21 +0000 (03:51 +0000)] 
restored sequence back, needed by PG for the unit test

18 years ago- added selectfirst(), selectfirst_by() to assign_mapper [ticket:467]
Mike Bayer [Thu, 22 Feb 2007 03:16:58 +0000 (03:16 +0000)] 
- added selectfirst(), selectfirst_by() to assign_mapper [ticket:467]

18 years agoadded collengths to use varchar instead of TEXT
Mike Bayer [Thu, 22 Feb 2007 03:13:43 +0000 (03:13 +0000)] 
added collengths to use varchar instead of TEXT

18 years ago- removed deprecated method of specifying custom collections on classes;
Mike Bayer [Wed, 21 Feb 2007 21:55:45 +0000 (21:55 +0000)] 
- removed deprecated method of specifying custom collections on classes;
you must now use the "collection_class" option. the old way was
beginning to produce conflicts when people used assign_mapper(), which
now patches an "options" method, in conjunction with a relationship
named "options". (relationships take precedence over monkeypatched
assign_mapper methods).

18 years agoformatting fix
Mike Bayer [Wed, 21 Feb 2007 21:35:51 +0000 (21:35 +0000)] 
formatting fix

18 years agoformatting/cleanup
Mike Bayer [Wed, 21 Feb 2007 21:34:40 +0000 (21:34 +0000)] 
formatting/cleanup

18 years ago - eager relation loading bug fixed for eager relation on multiple
Mike Bayer [Wed, 21 Feb 2007 20:57:31 +0000 (20:57 +0000)] 
  - eager relation loading bug fixed for eager relation on multiple
  descendant classes [ticket:486]

18 years agoTested with pymssql 0.8.0
Rick Morrison [Wed, 21 Feb 2007 01:38:58 +0000 (01:38 +0000)] 
Tested with pymssql 0.8.0
mssql: added query_timeout, fixes for passing auto_insert in dburl

18 years agofix to the fix for [ticket:454], prevent other mappers in a load operation from using...
Mike Bayer [Tue, 20 Feb 2007 18:01:16 +0000 (18:01 +0000)] 
fix to the fix for [ticket:454], prevent other mappers in a load operation from using the main extension option send to the query (i.e. mappers used for eager loads etc).

18 years ago- added "contains_alias()" option for result set mapping to an alias of the mapped...
Mike Bayer [Tue, 20 Feb 2007 02:12:46 +0000 (02:12 +0000)] 
- added "contains_alias()" option for result set mapping to an alias of the mapped table

18 years ago - added "alias" argument to contains_eager(). use it to specify the string name
Mike Bayer [Tue, 20 Feb 2007 01:04:07 +0000 (01:04 +0000)] 
  - added "alias" argument to contains_eager().  use it to specify the string name
    or Alias instance of an alias used in the query for the eagerly loaded child items.
    easier to use than "decorator"

18 years ago- moved SynonymProperty to interfaces, since its more generalized and synonym-aware...
Mike Bayer [Tue, 20 Feb 2007 00:09:37 +0000 (00:09 +0000)] 
- moved SynonymProperty to interfaces, since its more generalized and synonym-aware operations
take place without knowning so much about properties
- mapper options like eagerload(), lazyload(), deferred(), will work for "synonym()" relationships [ticket:485]

18 years ago(no commit message)
Rick Morrison [Mon, 19 Feb 2007 19:03:33 +0000 (19:03 +0000)] 

18 years agoremoved not-always-applicable check for "polymorphic_identity"
Mike Bayer [Mon, 19 Feb 2007 07:06:38 +0000 (07:06 +0000)] 
removed not-always-applicable check for "polymorphic_identity"

18 years ago- documented foreign_keys argument
Mike Bayer [Mon, 19 Feb 2007 04:46:08 +0000 (04:46 +0000)] 
- documented foreign_keys argument
- 0.3.5 markers

18 years agoreverted unit test change
Mike Bayer [Mon, 19 Feb 2007 04:36:18 +0000 (04:36 +0000)] 
reverted unit test change

18 years ago- oracle issues a log warning when a related table cant be reflected due to certain
Mike Bayer [Mon, 19 Feb 2007 04:28:51 +0000 (04:28 +0000)] 
- oracle issues a log warning when a related table cant be reflected due to certain
permission errors [ticket:363]

18 years agoCompleted previously missed patches from tickets 422 and 415
Rick Morrison [Sun, 18 Feb 2007 19:43:05 +0000 (19:43 +0000)] 
Completed previously missed patches from tickets 422 and 415
get unit tests to work with pyodbc [ticket:481]
fix blank password on adodbapi [ticket:371]

18 years ago- modified patch for [ticket:379] - detecting synonyms, dblinks in reflection. test...
Mike Bayer [Sat, 17 Feb 2007 23:40:31 +0000 (23:40 +0000)] 
- modified patch for [ticket:379] - detecting synonyms, dblinks in reflection.  test passes
except for DBLINK which I cannot get to work on my oracle-xe database.
- probable (also untested) fix for [ticket:363], better error message if we get None back for
remote table information (which is due to rights)

18 years ago- fixes to tometadata() operation to propigate Constraints at column and table level
Mike Bayer [Sat, 17 Feb 2007 19:08:36 +0000 (19:08 +0000)] 
- fixes to tometadata() operation to propigate Constraints at column and table level

18 years ago- fixed generation of CHECK constraints on columns [ticket:464]
Mike Bayer [Sat, 17 Feb 2007 03:46:13 +0000 (03:46 +0000)] 
- fixed generation of CHECK constraints on columns [ticket:464]

18 years ago- extension() query option propigates to Mapper._instance() method so that
Mike Bayer [Sat, 17 Feb 2007 03:14:36 +0000 (03:14 +0000)] 
- extension() query option propigates to Mapper._instance() method so that
all loading-related methods get called [ticket:454]

18 years ago- moved change for [ticket:466] to ansisql, since thats the syntax for all databases.
Mike Bayer [Sat, 17 Feb 2007 02:39:28 +0000 (02:39 +0000)] 
- moved change for [ticket:466] to ansisql, since thats the syntax for all databases.
the change is across all dialects, not just oracle

18 years agoadded PGInet type [ticket:444]
Mike Bayer [Sat, 17 Feb 2007 02:36:11 +0000 (02:36 +0000)] 
added PGInet type [ticket:444]

18 years ago- added a Sequence to the unicode test tables to help Oracle
Mike Bayer [Sat, 17 Feb 2007 02:31:56 +0000 (02:31 +0000)] 
- added a Sequence to the unicode test tables to help Oracle
- fixed named PrimaryKeyConstraint generation on oracle [ticket:466] courtesy andrija at gmail

18 years ago[ticket:463] fix to OrderedSet
Mike Bayer [Sat, 17 Feb 2007 02:23:10 +0000 (02:23 +0000)] 
[ticket:463] fix to OrderedSet

18 years agocommented out ImportError for now; will issue new ticket for handling dialect importe...
Mike Bayer [Sat, 17 Feb 2007 01:49:52 +0000 (01:49 +0000)] 
commented out ImportError for now; will issue new ticket for handling dialect importerrors more effectively

18 years ago- fixed oracle list of binary types to check for their presence in the module (such...
Mike Bayer [Sat, 17 Feb 2007 01:18:54 +0000 (01:18 +0000)] 
- fixed oracle list of binary types to check for their presence in the module (such as BFILE not in all versions of cx_Oracle)
- removed oracle-handicap from binary unit test to test [ticket:435] fix, added an extra row containing None

18 years ago- small fix to BoundMetaData to accept unicode or string URLs
Mike Bayer [Sat, 17 Feb 2007 00:09:48 +0000 (00:09 +0000)] 
- small fix to BoundMetaData to accept unicode or string URLs

18 years agobetter MSSSQL support for implicit sequences and auto-insert, ticket 415
Rick Morrison [Fri, 16 Feb 2007 17:26:47 +0000 (17:26 +0000)] 
better MSSSQL support for implicit sequences and auto-insert, ticket 415

18 years agoFix Fix for adodbapi bug introduced by ticket 419
Rick Morrison [Fri, 16 Feb 2007 17:12:49 +0000 (17:12 +0000)] 
Fix Fix for adodbapi bug introduced by ticket 419

18 years agoFunc rewrite for better unittest compatibility
Rick Morrison [Fri, 16 Feb 2007 17:06:17 +0000 (17:06 +0000)] 
Func rewrite for better unittest compatibility
Simplified transaction handling for pymssql

18 years agodont do RAWTOHEX on None
Mike Bayer [Thu, 15 Feb 2007 17:19:13 +0000 (17:19 +0000)] 
dont do RAWTOHEX on None