codes, per discussion from last March. parse.h should now be included
*only* by gram.y, scan.l, keywords.c, parser.c. This prevents surprising
misbehavior after seemingly-trivial grammar adjustments.
Also applied patch from Lars Stenberg to make callable statements use the form
select * from func() when running against a 7.3 server instead of select func() to allow for set returning functions to be called.
Modified Files:
jdbc/org/postgresql/errors.properties
jdbc/org/postgresql/core/Encoding.java
jdbc/org/postgresql/jdbc1/AbstractJdbc1Statement.java
rid of the assumption that sizeof(Oid)==sizeof(int). This is one small
step towards someday supporting 8-byte OIDs. For the moment, it doesn't
do much except get rid of a lot of unsightly casts.
locParam lists can be converted to bitmapsets to speed updating. Also,
replace 'locParam' with 'allParam', which contains all the paramIDs
relevant to the node (i.e., the union of extParam and locParam); this
saves a step during SetChangedParamList() without costing anything
elsewhere.
Instead of grovelling through pg_class to find them, make use of the
handy dandy dependency mechanism: just delete everything that depends
on our temp schema. Unlike the pg_class scan, the dependency mechanism
is smart enough to delete things in an order that doesn't fall foul of
any dependency restrictions. Fixes problem reported by David Heggie:
a temp table with a serial column may cause a backend FATAL exit at
shutdown time, if it chances to try to delete the temp sequence first.
expression accepted by the regex operators, per discussion yesterday.
Along the way, reduce deadlock_timeout from PGC_POSTMASTER to PGC_SIGHUP
category. It is probably best to insist that all backends share the same
setting, but that doesn't mean it has to be frozen at startup.
(extracted from Tcl 8.4.1 release, as Henry still hasn't got round to
making it a separate library). This solves a performance problem for
multibyte, as well as upgrading our regexp support to match recent Tcl
and nearly match recent Perl.
and the other fixes a NPE in Statement.toString() under some circumstances.
The second patch was originally submitted by Oliver Jowett and updated by Kris
startup, not in the parser; this allows ALTER DOMAIN to work correctly
with domain constraint operations stored in rules. Rod Taylor;
code review by Tom Lane.
nodes where it's not really necessary. In many cases where the scan node
is not the topmost plan node (eg, joins, aggregation), it's possible to
just return the table tuple directly instead of generating an intermediate
projection tuple. In preliminary testing, this reduced the CPU time
needed for 'SELECT COUNT(*) FROM foo' by about 10%.
restriction was debatable to begin with, but it has now become obvious
that it breaks forward-porting of user-defined types; contrib/lo being
the most salient example.
columns of type lo (see contrib/lo). Rather than hacking the function
definitions on-the-fly, just modify the queries issued by FixupBlobRefs
so that they work even if CREATE CAST hasn't been issued.
connection shutdown. This is a grotty workaround for a Tcl bug, but
said bug has been there long enough that I'm not holding my breath
for a real fix. Per discussions and testing from ljb and g.hintermayer.
bison 1.875 and later as we did from earlier bison releases. Eventually
we will probably want to adopt the newer message spelling ... but not yet.
Per recent discussion on pgpatches.
Note: I didn't change the build rules for bootstrap, ecpg, or plpgsql
grammars, since these do not affect regression test results.
for type 'time without time zone', as we already did for type
'timestamp without time zone'. This patch was proposed by Tom Lockhart
on 7-Nov-02, but he never got around to applying it. Adjust regression
tests and documentation to match.
cannot actually happen at present because ArrayCount() is only called
on strings beginning with '{', but seems best to prevent it going forward.
Per report from Yichen Xie.
value of MAX_TIME_PRECISION in floating-point-timestamp-storage case
from 13 to 10, which is as much as time_out is actually willing to print.
(The alternative of increasing the number of digits we are willing to
print looks risky; we might find ourselves printing roundoff garbage.)
passed to join selectivity estimators. Make use of this in eqjoinsel
to derive non-bogus selectivity for IN clauses. Further tweaking of
cost estimation for IN.
initdb forced because of pg_proc.h changes.
Try to model the effect of rescanning input tuples in mergejoins;
account for JOIN_IN short-circuiting where appropriate. Also, recognize
that mergejoin and hashjoin clauses may now be more than single operator
calls, so we have to charge appropriate execution costs.
necessarily following the JOIN syntax to develop the query plan. The old
behavior is still available by setting GUC variable JOIN_COLLAPSE_LIMIT
to 1. Also create a GUC variable FROM_COLLAPSE_LIMIT to control the
similar decision about when to collapse sub-SELECT lists into their parent
lists. (This behavior existed already, but the limit was always
GEQO_THRESHOLD/2; now it's separately adjustable.)
of the socket file and socket lock file; this should prevent both of them
from being removed by even the stupidest varieties of /tmp-cleaning
script. Per suggestion from Giles Lean.
rather than actually opening the files. This eliminates some corner cases
where the file indeed exists but open() fails for another reason, such
as being out of file descriptors. The net reliability gain is probably
tiny, since xlog.c is full of other file open calls that will elog(PANIC)
if they fail for any reason; but this specific failure mode has been
observed in the field, so we may as well fix it.
of known-equal expressions includes any constant expressions (including
Params from outer queries), we actively suppress any 'var = var'
clauses that are or could be deduced from the set, generating only the
deducible 'var = const' clauses instead. The idea here is to push down
the restrictions implied by the equality set to base relations whenever
possible. Once we have applied the 'var = const' clauses, the 'var = var'
clauses are redundant, and should be suppressed both to save work at
execution and to avoid double-counting restrictivity.
that's selecting into a RECORD variable returns zero rows, make it
assign an all-nulls row to the RECORD; this is consistent with what
happens when the SELECT INTO target is not a RECORD. In support of
this, tweak the SPI code so that a valid tuple descriptor is returned
even when a SPI select returns no rows.
There are two implementation techniques: the executor understands a new
JOIN_IN jointype, which emits at most one matching row per left-hand row,
or the result of the IN's sub-select can be fed through a DISTINCT filter
and then joined as an ordinary relation.
Along the way, some minor code cleanup in the optimizer; notably, break
out most of the jointree-rearrangement preprocessing in planner.c and
put it in a new file prep/prepjointree.c.
that used to do it in planner. That was an ancient kluge that was
never satisfactory; errors should be detected at parse time when possible.
But at the time we didn't have the support mechanism (expression_tree_walker
et al) to make it convenient to do in the parser.
simplify callers. It turns out the common case is that the caller
does want to recurse into sub-queries, so push support for that into
these subroutines.
datetime token tables. Even more embarrassing, the regression tests
revealed some of the problems --- but evidently the bogus output wasn't
questioned. Add code to postmaster startup to directly check the tables
for correct ordering, in hopes of not being embarrassed like this again.
join_references(), it's practical to consolidate all join_references()
processing into the set_plan_references traversal in setrefs.c. This
seems considerably cleaner than the old way where we did it for join
quals in createplan.c and for targetlists in setrefs.c.
containing a volatile function), rather than only on 'Var = Var' clauses
as before. This makes it practical to do flatten_join_alias_vars at the
start of planning, which in turn eliminates a bunch of klugery inside the
planner to deal with alias vars. As a free side effect, we now detect
implied equality of non-Var expressions; for example in
SELECT ... WHERE a.x = b.y and b.y = 42
we will deduce a.x = 42 and use that as a restriction qual on a. Also,
we can remove the restriction introduced 12/5/02 to prevent pullup of
subqueries whose targetlists contain sublinks.
Still TODO: make statistical estimation routines in selfuncs.c and costsize.c
smarter about expressions that are more complex than plain Vars. The need
for this is considerably greater now that we have to be able to estimate
the suitability of merge and hash join techniques on such expressions.
the code would only capture milliseconds where as both postgres and the java
Timestamp object support greater resolution.
Also fixed a bug reported by Rhett Sutphin where the last digit of the
fractional seconds was lost when using timestamp without time zone
Modified Files:
jdbc/org/postgresql/jdbc1/AbstractJdbc1ResultSet.java
jdbc/org/postgresql/test/jdbc2/TimestampTest.java
Use explicit joins to avoid using the genetic query optimizer. Also fixed
a regression test that was failing to compile. This change also cleans up
how key names are reported as per:
A change to the value of the FK_NAME column. Currently the returned value
is the triggers arguments which look like
"<unnamed>\000t2\000t1\000UNSPECIFIED\000a\000a\000"
This was required for server versions < 7.3 when a user did not supply
constraint names. Every constraint was named "<unnamed>"
. 7.3 has enforced unique constraint names per table so unnamed foreign
keys will have different names "$1", "$2" and so on. I've used logic
along the lines of the following to preserve the unique names in the
original scheme, but allow people who go to the trouble of naming their
constraints to see them:
if (triggerargs.startsWith("<unnamed>")) {
fkname = [the whole ugly trigger args name originally used];
} else {
fkname = [the actual fk name];
}
Modified Files:
jdbc/org/postgresql/jdbc1/AbstractJdbc1DatabaseMetaData.java
jdbc/org/postgresql/test/jdbc2/optional/ConnectionPoolTest.java
a qualification clause (and hence can get away with being sloppy about
distinguishing FALSE from UNKNOWN). We need to know this in subselect.c;
marking the subplans in setrefs.c is too late.
HAVING quals. Normally this is an insignificant effect --- but it
will not be insignificant when these clauses contain sub-selects.
The added costs cannot affect the planning of the query containing
them, but they might have an impact when the query is a sub-query
of a larger one.
costs for expression evaluation, not only per-tuple cost as before.
This extension is needed in order to deal realistically with hashed or
materialized sub-selects.
>
> I'd suggest that the runtime.sgml description explicitly say "values of
> at least a few thousand are recommended for production installations".
Neil Conway
Simplify SubLink by storing just a List of operator OIDs, instead of
a list of incomplete OpExprs --- that was a bizarre and bulky choice,
with no redeeming social value since we have to build new OpExprs
anyway when forming the plan tree.
'NOT (x IN (subselect))', that is 'NOT (x = ANY (subselect))',
rather than 'x <> ALL (subselect)' as we formerly did. This
opens the door to optimizing NOT IN the same way as IN, whereas
there's no hope of optimizing the expression using <>. Also,
convert 'x <> ALL (subselect)' to the NOT(IN) style, so that
the optimization will be available when processing rules dumped
by older Postgres versions.
initdb forced due to small change in SubLink node representation.
per gripe from Csaba Nagy. There is still potential for platform-specific
behavior for values that are exactly halfway between integers, but at
least we now get the expected answer for all other cases.
causes interval rounding not to work as expected in 7.3, for example
SELECT '18:17:15.6'::interval(0) does not round the value.
I did not force initdb, but one is needed to install the added row.
believe I didn't notice this before -- once 64k was sent to/from the
server the client would crash. Basicly, in 7.3 the server SSL code set
the initial state to "about to renegotiate" without actually starting
the renegotiation. In addition, the server and client didn't properly
handle the SSL_ERROR_WANT_(READ|WRITE) error. This is fixed in the
second patch.
Nathan Mueller
first, that I missed when checking over 7.3.1, was that the client
method was switched to SSLv23 along with the server. The SSLv23 client
method does SSLv2 by default, but can also understand SSLv3. In our
situation the SSLv2 backwords compatibility is really only needed on the
server. This is the first patch.
The last thing is that I found a way for the server to understand SSLv2
HELLO messages (sent by pre-7.3 clients) but then get them to talk
SSLv3. This is the last one.
Nathan Mueller
> The big problem is that while pg_dump's dump_trigger() looks at
> tginitdeferred and dumps accordingly, pg_get_constraintdef doesn't look
> at tginitdeferred, and therefore doesn't record the requirement as part
> of ALTER TABLE ADD CONSTRAINT.
pg_get_constraintdef should probably be looking at condeferrable and
condeferred in the pg_constraint row it's looking at. Maybe something
like the attached.
(Added, output only non-default values.)
Stephan Szabo
proposal on -hackers a couple days ago. The 'tgenabled' status of the
trigger is consulted when the trigger is added to the deferred queue
(i.e. when the event occurs that fires the trigger), not when the
deferred event is executed.
No regression tests, as another bug prevents them (the pg_trigger row
for a trigger is only loaded once per session, so any changes to it are
not noticed unless the client disconnects and reconnects).
Neil Conway
the index AM when we know we are fetching a unique row. However, this
logic did not consider the possibility that it would be asked to fetch
backwards. Also fix mark/restore to work correctly in this scenario.
previously determined not to be the last segment of a relation.
This reduces the expected cost to one seek, rather than one seek per
segment. We can get away with this because truncation of a relation
will cause a relcache flush and so the md.c file descriptor will be
closed; when it is re-opened we will re-determine the last segment.
match parent table. This used to work, but was broken in 7.3 by
rearrangement of code that handles targetlist sorting. Add a regression
test to catch future breakage.
patches of 9-Dec (permissions fix) and 13-Dec (performance) as well as
a partial fix for locking issues: concurrent DROP COLUMN should not
create trouble anymore. But concurrent DROP TABLE is still a risk, and
there is no protection at all against creating a column of a domain while
we are altering the domain.
columns in DefineIndex. So, ALTER TABLE ... PRIMARY KEY will now
automatically add the NOT NULL constraint. It appeared the alter_table
regression test wanted this to occur, as after the change the regression
test better matched in inline 'fails'/'succeeds' comments.
Rod Taylor
beginning/end of cursor.
Have MOVE return 0/1 depending on cursor position.
Matches SQL spec.
Pass cursor counter from parser as a long rather than int.
Doc updates.
computation: reduce the bucket number mod nbatch. This changes the
association between original bucket numbers and batches, but that
doesn't matter. Minor other cleanups in hashjoin code to help
centralize decisions.
Also, tweak -C option (emit CREATE DATABASE command) to emit encoding
name rather than encoding number, for consistency with pg_dumpall
and better cross-version portability.
"SSLv23_method(void), SSLv23_server_method(void), SSLv23_client_method(void)
A TLS/SSL connection established with these methods will understand the SSLv2,
SSLv3, and TLSv1 protocol. A client will send out SSLv2 client hello messages
and will indicate that it also understands SSLv3 and TLSv1. A server will
understand SSLv2, SSLv3, and TLSv1 client hello messages. This is the best
choice when compatibility is a concern."
This will maintain backwards compatibility for those us that don't use
TLS connections ...
allocation in best_inner_indexscan(). While at it, simplify GEQO's
interface to the main planner --- make_join_rel() offers exactly the
API it really wants, whereas calling make_rels_by_clause_joins() and
make_rels_by_clauseless_joins() required jumping through hoops.
Rewrite gimme_tree for clarity (sometimes iteration is much better than
recursion), and approximately halve GEQO's runtime by recognizing that
tours of the forms (a,b,c,d,...) and (b,a,c,d,...) are equivalent
because of symmetry in make_join_rel().
disallowed by CREATE TABLE (eg, pseudo-types); also disallow these types
from being introduced by the range-function syntax. While at it, allow
CREATE TABLE to create zero-column tables, per recent pghackers discussion.
I am back-patching this into 7.3 since failure to disallow pseudo-types
is arguably a security hole.
practice of evaluating MemSet's arguments multiple times, except for
the special case of newNode(), where we can assume the argument is
a constant sizeof() operator.
Also, add GetMemoryChunkContext() to mcxt.c's API, in preparation for
fixing recent GEQO breakage.
given any malloc block until something is first allocated in it; but
thereafter, MemoryContextReset won't release that first malloc block.
This preserves the quick-reset property of the original policy, without
forcing 8K to be allocated to every context whether any of it is ever
used or not. Also, remove some more no-longer-needed explicit freeing
during ExecEndPlan.
a per-query memory context created by CreateExecutorState --- and destroyed
by FreeExecutorState. This provides a final solution to the longstanding
problem of memory leaked by various ExecEndNode calls.
failing to find pg_hba.conf should be a fatal error anyway, so I
increased the priority of the elog() from LOG to FATAL and refactored
the code a little bit.
Neil Conway
in the planned representation of a subplan at all any more, only SubPlan.
This means subselect.c doesn't scribble on its input anymore, which seems
like a good thing; and there are no longer three different possible
interpretations of a SubLink. Simplify node naming and improve comments
in primnodes.h. No change to stored rules, though.
produce which output in the geometry test, even with the problem narrowed
down to only whether they print minus zero or not. Instead, use
pg_regress' locale-variant mechanism to automatically consider the test
to pass if it matches either supplied comparison file. geometry_1.out
replaces the former geometry-positive-zeros.out.
execution state trees, and ExecEvalExpr takes an expression state tree
not an expression plan tree. The plan tree is now read-only as far as
the executor is concerned. Next step is to begin actually exploiting
this property.
completion. Note that it's based on 7.3 tarball, not CVS HEAD, or 7.3rel
branch. Damn, looking at CVS, this will patch into 7.3rel (just tested,
it does) probably collide with Rod Taylor's patch adding ALTER TRIGGER
stuff. O.K, second patch attached against HEAD - not tested, hand
merged.
Ross Reedstrom
>
> In pg.py the attributes of DB are defined as being the same as
> the attributes of the corresponding pgobject "db", using the following
...
> The problem is that the attributes of db (which are read only)
> are not static (they are actually function calls to PostgreSQL),
> especially "status" and "error", but those attributes are copied
> and this is done only once when initializing the DB object.
>
> So, in effect, only the attribute "db.error" of a DB instance
> will be updated, but not the attribute "error". Same with "status".
> Don't copy the (read only) attributes of the pgobject to the
> DB object, but only the methods, and all of them, like this:
>
> --------------- change in pg.py ------------------
> # Create convience methods, in a way that is still overridable.
> for e in self.db.__methods__:
> setattr(self, e, getattr(self.db, e))
> ----------------------------------------------------
>
> Furthermore, make an addition to the documentation of the
> DB wrapper class (i.e. in pygresql-pg-db.html):
> After the sentence "All pgobject methods are included in this class also."
> add the following sentence "The pgobject read-only attributes can be
> accessed py adding the prefix 'db.' to them."
Christoph Zwerschke
* Add schema, cast, and conversion backslash commands to psql
I had to create a new publically available function,
pg_conversion_is_visible, as it seemed to be missing from the catalogs.
This required me to do no small amount of hacking around in namespace.c
I have updated the \? help and sgml docs.
\dc - list conversions [PATTERN]
\dC - list casts
\dn list schemas
I didn't support patterns with casts as there's nothing obvious to match
against.
Catalog version incremented --- initdb required.
Christopher Kings-Lynne
make VALUE a non-reserved word again, use less invasive method of passing
ConstraintTestValue into transformExpr, fix problems with nested constraint
testing, do correct thing with NULL result from a constraint expression,
remove memory leak. Domain checks still need much more work if we are going
to allow ALTER DOMAIN, however.
so that all executable expression nodes inherit from a common supertype
Expr. This is somewhat of an exercise in code purity rather than any
real functional advance, but getting rid of the extra Oper or Func node
formerly used in each operator or function call should provide at least
a little space and speed improvement.
initdb forced by changes in stored-rules representation.
documentation and regression test mods. It seemed small and unobtrusive enough
to not require a specific proposal on the hackers list -- but if not, let me
know and I'll make a pitch. Otherwise, if there are no objections please apply.
Joe Conway
postgresql version 7.3, but yea... this patch adds full IPv6
support to postgres. I've tested it out on 7.2.3 and has
been running perfectly stable.
CREDITS:
The KAME Project (Initial patch)
Nigel Kukard <nkukard@lbsd.net>
Johan Jordaan <johanj@lando.co.za>
=====================
I suggested an improvement of the inserttable in the PyGreSQL interface
already in January, but seemingly it was never implemented. I was told this
is the right place to get patches in for PyGreSQL, so I'm reposting my patch
here.
I consider the inserttable methode essential in populating the database
because of its benefits in performance compared to insert, so I think this
patch is quite essential. The attachment is an improved version of the
corresponding pg_inserttable function in pgmodule.c, which fixes the
following problems:
* The function raised exceptions because PyList_GetItem was used beyond the
size of the list. This was checked by comparing the result with NULL, but
the exception was not cleaned up, which could result in mysterious errors in
the following Python code. Instead of clearing the exception using
PyErr_Clear or something like that, I avoided throwing the exception at all
by at first requesting the size of the list. Using this opportunity, I also
checked the uniformity of the size of the rows passed in the lists/tuples.
The function also accepts (and silently ignores) empty lists and sublists.
* Python "None" values are now accepted and properly converted to PostgreSQL
NULL values
* The function now generates an error message in case of a line buffer
overflow
* It copes with tabulators, newlines and backslashes in strings now
* Rewrote the buffer filling code which should now run faster by avoiding
unnecessary string copy operations forth and back
Christoph Zwerschke
cleaning up locale names and nothing else. Since all the locale names
are in plain ASCII I think it will be safe to use ASCII-only lower-case
conversion.
Nicolai Tufar
problems that occur if sublink is referenced via a join alias variable.
Perhaps this can be improved later, but a simple and safe fix is needed
for 7.3.1.
to plan nodes, not vice-versa. All executor state nodes now inherit from
struct PlanState. Copying of plan trees has been simplified by not
storing a list of SubPlans in Plan nodes (eliminating duplicate links).
The executor still needs such a list, but it can build it during
ExecutorStart since it has to scan the plan tree anyway.
No initdb forced since no stored-on-disk structures changed, but you
will need a full recompile because of node-numbering changes.
('SELECT expression') inline, like macros, during the constant-folding
phase of planning. The actual expansion is not difficult, but checking
that we're not changing the semantics of the call turns out to be more
subtle than one might think; in particular must pay attention to
permissions issues, strictness, and volatility.
well as function calls. This is needed for cases where the planner has
constant-folded or inlined the original function call. Possibly we should
back-patch this change into 7.3 branch as well.
and eliminate its manual pfree() calls. This solves the encoding-conversion
bug recently reported, and should be faster and more robust than the
original coding anyway. For example, we are no longer at risk if
datatype output routines leak memory or choose to return a constant string.
logic, dissuade planner from thinking that 'x IS DISTINCT FROM 42' may
be optimized into 'x = 42' (!!), cause dependency on = operator to be
recorded correctly, minor other improvements.
instead of only one. This should speed up planning (only one hash path
to consider for a given pair of relations) as well as allow more effective
hashing, when there are multiple hashable joinclauses.
operations: make sure we use operators that are compatible, as determined
by a mergejoin link in pg_operator. Also, add code to planner to ensure
we don't try to use hashed grouping when the grouping operators aren't
marked hashable.
sublink results and COPY's domain constraint checking. A Const that
isn't really constant is just a Bad Idea(tm). Remove hacks in
parse_coerce and other places that were needed because of the former
klugery.
just done for copyfuncs/equalfuncs. Read functions in particular get
a lot shorter than before, and it's much easier to compare an out function
with the corresponding read function to make sure they agree.
initdb forced due to small changes in nodestring format (regularizing
a few cases that were formerly idiosyncratic).
reliance on macros, in hopes of eliminating silly typos (like copying
to the wrong field) and just generally making it easier to see the forest
instead of the trees. As an example, here is the new code for A_Indices:
static A_Indices *
_copyAIndices(A_Indices *from)
{
A_Indices *newnode = makeNode(A_Indices);
COPY_NODE_FIELD(lidx);
COPY_NODE_FIELD(uidx);
return newnode;
}
static bool
_equalAIndices(A_Indices *a, A_Indices *b)
{
COMPARE_NODE_FIELD(lidx);
COMPARE_NODE_FIELD(uidx);
return true;
}
I plan to redo outfuncs/readfuncs in a similar style, but am committing
what I've got.
raises pgdb.DatabaseError when any of the fetch*
methods was invoked but previous call to execute* did
not produce any result set or no call was issued yet.
Also, raises pgdb.NotSupportedError when .nextset() is
invoked, instead of NameError.
This behaviour complies with DB-API 2.0.
Thanks for your work!
Timur Irmatov.
used for the primary key lookup. This will prevent a database lookup
for each connection object that gets created. This could be a significant
optimization on a busy system.
Similarly, the get_attnames method allows for the attributes dictionary
to be installed directly.
debug output is managed. The user can continue to use the current method
of passing a formatting string to have a replacement done and output will
be sent to the standard output exactly as it did before. In addition they
can set it to a file object, sys.stderr for example, and the query string
will be printed to it. Thay can also set it to a method (function) and the
query string will be passed to that method giving them the maximum flexibility
to do whatever they want with the query string.
I will be working with the PyGreSQL documentation shortly and at that time
will properly document this feature.
joinclauses is determined accurately for each join. Formerly, the code only
considered joinclauses that used all of the rels from the outer side of the
join; thus for example
FROM (a CROSS JOIN b) JOIN c ON (c.f1 = a.x AND c.f2 = b.y)
could not exploit a two-column index on c(f1,f2), since neither of the
qual clauses would be in the joininfo list it looked in. The new code does
this correctly, and also is able to eliminate redundant clauses, thus fixing
the problem noted 24-Oct-02 by Hans-Jürgen Schönig.
-hackers a couple days ago.
Notes/caveats:
- added regression tests for the new functionality, all
regression tests pass on my machine
- added pg_dump support
- updated PL/PgSQL to support per-statement triggers; didn't
look at the other procedural languages.
- there's (even) more code duplication in trigger.c than there
was previously. Any suggestions on how to refactor the
ExecXXXTriggers() functions to reuse more code would be
welcome -- I took a brief look at it, but couldn't see an
easy way to do it (there are several subtly-different
versions of the code in question)
- updated the documentation. I also took the liberty of
removing a big chunk of duplicated syntax documentation in
the Programmer's Guide on triggers, and moving that
information to the CREATE TRIGGER reference page.
- I also included some spelling fixes and similar small
cleanups I noticed while making the changes. If you'd like
me to split those into a separate patch, let me know.
Neil Conway
one more row from the subplan than the COUNT would appear to require.
This costs a little more logic but a number of people have complained
about the old implementation.
results due to doing arithmetic on uninitialized values. Add some
documentation about the AT TIME ZONE construct. Update some other
date/time documentation that seemed out of date for 7.3.
database access outside a transaction; revert bogus performance improvement
in SIBackendInit(); improve comments; add documentation (this part courtesy
Neil Conway).
parameter to allow it to be forced off for comparison purposes.
Add ORDER BY clauses to a bunch of regression test queries that will
otherwise produce randomly-ordered output in the new regime.
on a preparedStatement would reset the prepared statment causing subsequent
uses of the preparedStatement to fail (i.e. the following series of calls
would fail: addBatch() executeBatch() addBatch() executBatch()). This is
a regression from 7.2 where this worked correctly. The regression test has
also been modified to explicitly test for this case.
Modified Files:
jdbc/org/postgresql/jdbc1/AbstractJdbc1Statement.java
jdbc/org/postgresql/jdbc2/AbstractJdbc2Statement.java
jdbc/org/postgresql/test/jdbc2/BatchExecuteTest.java
of groups produced by GROUP BY. This improves the accuracy of planning
estimates for grouped subselects, and is needed to check whether a
hashed aggregation plan risks memory overflow.
"canceled", so I changed the one remaining usage of the British
spelling ("cancelled") over to the former, and updated the translation
files appropriately.
Neil Conway
The code was not making TupleConstr structs for such catalogs in
several places; with the consequence that the not-null constraint
wasn't actually enforced. With this change,
INSERT INTO pg_proc VALUES('sdf');
generates a 'Fail to add null value' error instead of a core dump.
- CLUSTER ALL clusters all the tables that have some index with
indisclustered set and the calling user owns.
- CLUSTER tablename clusters the named table, using the index with
indisclustered set. If no index has the bit set, throws elog(ERROR).
- The multi-relation version (CLUSTER ALL) uses a multitransaction
approach, similar to what VACUUM does.
Alvaro Herrera
Add simple ALTER DATABASE, ALTER TRIGGER, CHECK POINT, CREATE
CONVERSION, CREATE DOMAIN, CREATE LANGUAGE, DEALLOCATE, DROP CONVERSION,
DROP DOMAIN, DROP LANGUAGE, EXECUTE, PREPARE
Complete CAST in CREATE CAST and DROP CAST but doesn't suggest what
should follow.
Add many more SET / SHOW variables to the list. Taken from SHOW ALL
output.
Complete a case sensitive search to allow \dD, \dd, \dS, \ds, \h, \H to
complete properly. But there are no matches, then try a case
insensitive search to allow case conversion. Add all missing help
options.
\Q<tab> -> \q
\dD<tab> -> \dD
\dd<tab> -> \dd
\D<tab><tab><tab> -> \d (with listing of \d? commands)
sel<tab> -> SELECT
Rod Taylor
Fixed bug with using setNull()(or setXXX(x, null)) and serverside prepare statements.
Improved error message when using a connection object that has already been closed.
Modified Files:
jdbc/org/postgresql/errors.properties
jdbc/org/postgresql/core/Encoding.java
jdbc/org/postgresql/core/QueryExecutor.java
jdbc/org/postgresql/jdbc1/AbstractJdbc1Connection.java
jdbc/org/postgresql/jdbc1/AbstractJdbc1Statement.java
anymore given the mktime() workaround now done in DetermineLocalTimeZone.
This has now been confirmed by Robert Bruccoleri for Irix, and I'm going
to extrapolate to AIX as well.
before commit, not after :-( --- the original coding is not only unsafe
if an error occurs while it's processing, but it generates an invalid
sequence of WAL entries. Resurrect 7.2 logic for deleting items when
no longer needed. Use an enum instead of random macros. Editorialize
on names used for routines and constants. Teach backend/nodes routines
about new field in CreateTable struct. Add a regression test.
regression test to test for this case. Patch submitted by Kris Jurka.
Modified Files:
jdbc/org/postgresql/jdbc1/AbstractJdbc1DatabaseMetaData.java
jdbc/org/postgresql/test/jdbc2/DatabaseMetaDataTest.java
contains the patches to Makefile.global.in and Makefile.unixware. The
Makefile.unixware patch has been updated to include the contents of
LD_LIBRARY_PATH, if present, to the -rpath (-R) option. This change
will simplify configuring and building PostgreSQL on systems that
support LD_LIBRARY_PATH. You can set LD_LIBRARY_PATH to include all
the directorys you want to have searched for additional libraries, run
configure, then run make. The paths in LD_LIBRARY_PATH will then be
embedded in the executables via the -rpath (-R) option to the linker,
and so will not require LD_LIBRARY_PATH in order to run.
Billy G. Allie
PL/PgSQL. Previously, it had been bundled together with the assign
statement implementation, for some reason that wasn't clear to me
(they certainly don't share any code with one another). So I separated
them and made PERFORM a statement like any other. No changes in
functionality.
Along the way, I added some regression tests for PERFORM, added a
bunch more SGML tags to the PL/PgSQL docs, and removed an obsolete
comment relating to the implementation of RETURN NEXT.
Neil Conway
>
> ... he is now about to write an inlined version that can go into
> s_lock.h . I'll send the new patch later on...
OK, here it comes:
An inlined version of tas(), that works for both, powerpc and
powerpc64. The patch is against 7.3b5 and passes the test suite on
both architectures.
Reinhard Max
postgres.h or c.h includes a system header (such as stdio.h or
stdlib.h), there's no need to specifically include it in any of the .c
files in the backend.
Neil Conway
variation. To do this, set extra_float_digits to -3 in the geometry
test, and tweak the CIRCLE_TBL dataset to avoid values that suffer
from severe cancellation error (eg, circles that just touch an axis).
We still need two geometry 'expected' files to account for the
difference between platforms that display minus zero as '-0' and those
that just say '0', but with luck that's all we'll need.
"traditional" behavior, so the change should be transparent. Use the
command "\pset pager always" to turn it on. Anything else does the
normal toggle between "on" and "off"
Greg Sabino Mullane
in hopes of reducing platform-to-platform variations in its results.
This will cause the geometry regression test to start failing on some
platforms. I plan to update the test later today.
precision for float4, float8, and geometric types. Set it in pg_dump
so that float data can be dumped/reloaded exactly (at least on platforms
where the float I/O support is properly implemented). Initial patch by
Pedro Ferreira, some additional work by Tom Lane.
now)" item on the open items, and subsequent plpgsql function I sent in,
made me realize it was too hard to get the upper and lower bound of an
array. The attached creates two functions that I think will be very
useful when combined with the ability of plpgsql to return sets.
array_lower(array, dim_num)
- and -
array_upper(array, dim_num)
They return the value (as an int) of the upper and lower bound of the
requested dim in the provided array.
Joe Conway
not read until after we've read the port-specific header file. In
particular this should make it safer to #include system headers for
inet_aton; in general it seems that the port header file ought to be
in a position to set definitions before we do stuff based on having
a definition or not.
node now does its own grouping of the input rows, and has no need for a
preceding GROUP node in the plan pipeline. This allows elimination of
the misnamed tuplePerGroup option for GROUP, and actually saves more code
in nodeGroup.c than it costs in nodeAgg.c, as well as being presumably
faster. Restructure the API of query_planner so that we do not commit to
using a sorted or unsorted plan in query_planner; instead grouping_planner
makes the decision. (Right now it isn't any smarter than query_planner
was, but that will change as soon as it has the option to select a hash-
based aggregation step.) Despite all the hackery, no initdb needed since
only in-memory node types changed.
- First fixes a problem with a recent patch allowing setNull on updateable
resultsets
- Second removed toLower() calls on database object names. Leave it to
the caller to correctly pass lower, upper or mixed case. The driver
already has methods that the caller can use to determine that postgres
stores identifiers in lowercase. (unless the identifier was quoted when
created).
Modified Files:
jdbc/org/postgresql/jdbc1/AbstractJdbc1DatabaseMetaData.java
jdbc/org/postgresql/jdbc2/AbstractJdbc2ResultSet.java
a child table already has a matching column. Acquire appropriate
lock on child table; do the right thing with any CHECK constraints
attached to the new parent column.
where it's safe to do database access. Along the way, fix core dump
for 'DEFAULT' parameters to CREATE DATABASE. initdb forced due to
change in pg_proc entry.
(usually bison output files), not as standalone files. This hack
works around flex's insistence on including <stdio.h> before we are
able to include postgres.h; postgres.h will already be read before
the compiler starts to read the flex output file. Needed for largefile
support on some platforms.
core file to be produced for debugging, and avoids trying to run the
normal proc-exit cleanup hooks, which are likely to cause additional
problems if the system is hosed.
between signal handler and enable/disable code, avoid accumulation of
timing error due to trying to maintain remaining-time instead of
absolute-end-time, disable timeout before commit not after.
Only affects machines where MAXALIGN > 4, and is a boundary-condition
case even there, but still surprising that it's not been identified
before. Also reduce tuple chain move give-up messages from WARNING
to DEBUG1, since they are not unexpected conditions.
'empty declaration' warnings from compilers that care about such things.
Per discussion back before 7.2 release; we didn't do it then because
we'd already missed all the beta cycle ...
that tv_sec is signed; return a useful error message on timeout failure;
honor PGCONNECT_TIMEOUT environment variable in PQsetdbLogin; make code
obey documentation statement that timeout=0 means no timeout.
coercions, not implicit ones. For example, 'select abstime(1035497293)'
should succeed because there is an explicit binary coercion from int4
to abstime.
principled order; in particular ensure that all shared resources
are released before we release transaction locks. The code used
to release locks before buffer pins, which might explain an ancient
note I have about a bufmgr assertion failure I'd seen once several
years ago, and been unable to reproduce since. (Theory: someone
trying to drop a relation might be able to reach FlushRelationBuffers
before the last user of the relation had gotten around to dropping
his buffer pins.)
On Sat, Oct 19, 2002 at 12:11:32AM +0200, Peter Eisentraut wrote:
> $ ./clusterdb
> psql: could not connect to server: No such file or directory
> Is the server running locally and accepting
> connections on Unix domain socket "/tmp/.s.PGSQL.5432"?
> psql: could not connect to server: No such file or directory
> Is the server running locally and accepting
> connections on Unix domain socket "/tmp/.s.PGSQL.5432"?
> clusterdb: While clustering peter, the following failed:
> $
>
> This could probably handled a little more gracefully.
Yes, sorry. A patch for this is attached. Please apply.
Alvaro Herrera
item, if the page containing the current item is split while the indexscan
is stopped and holds no read-lock on the page. The current item might
move right onto a page that the indexscan holds no pin on. In the prior
code this would allow btbulkdelete to reach and possibly delete the item,
causing 'my bits moved right off the end of the world!' when the indexscan
finally resumes. Fix by chaining read-locks to the right during
_bt_restscan and requiring btbulkdelete to LockBufferForCleanup on every
page it scans, not only those with deletable items. Per my pghackers
message of 25-May-02. (Too bad no one could think of a better way.)
Also removed some unused files and fixed the which needed a small change
after the previous patch to build.xml.
Modified Files:
jdbc/Makefile jdbc/org/postgresql/core/Encoding.java
jdbc/org/postgresql/jdbc1/AbstractJdbc1Connection.java
Removed Files:
jdbc/utils/CheckVersion.java jdbc/utils/buildDriver
jdbc/utils/changelog.pl
whose conditions might yield NULL. The negated qual to attach to the
original query is properly 'x IS NOT TRUE', not 'NOT x'. This fix
produces correct behavior, but we may be taking a performance hit because
the planner is much stupider about IS NOT TRUE than it is about NOT
clauses. Future TODO: teach prepqual, other parts of planner how to
cope with BooleanTest clauses more effectively.
configure hasn't been run before trying to build.
Also cleaned up the README file and removed some obsolete files.
Modified Files:
jdbc/README jdbc/build.xml
Removed Files:
jdbc/CHANGELOG jdbc/Implementation jdbc/jdbc.jpx
SPI_prepare: they all save the prepared plan into topCxt, and so the
procCxt copy that's actually returned by SPI_prepare ought to be freed.
Diagnosis and plpython fix by Nigel Andrews, followup for other PLs
by Tom Lane.
in such a way that indexes on int8 columns would be used (by quoting the value)
caused other problems. Will need to wait for the backend to properly fix
the root problem.
Modified Files:
jdbc/org/postgresql/jdbc1/AbstractJdbc1Statement.java
