Tom Lane [Wed, 31 Jul 2002 17:19:54 +0000 (17:19 +0000)]
Instead of having a configure-time DEFAULT_ATTSTATTARGET, store -1 in
attstattarget to indicate 'use the default'. The default is now a GUC
variable default_statistics_target, and so may be changed on the fly. Along
the way we gain the ability to have pg_dump dump the per-column statistics
target when it's not the default. Patch by Neil Conway, with some kibitzing
from Tom Lane.
Tom Lane [Tue, 30 Jul 2002 21:56:04 +0000 (21:56 +0000)]
Teach pg_dump to dump user-defined operator classes. For the moment,
this only works against 7.3 or later databases; the pushups required
to do it without regprocedure/regtype/etc seem more trouble than they're
worth, considering that existing users aren't expecting pg_dump support
for this.
Bruce Momjian [Tue, 30 Jul 2002 16:55:45 +0000 (16:55 +0000)]
The attached patch removes the last remnants of support for
'tioga recipes', whatever those are -- Peter E. killed most
of it a couple days ago, but this patch removes the rest. Most
of it was #ifdef'ed out anyway.
Bruce Momjian [Tue, 30 Jul 2002 16:55:06 +0000 (16:55 +0000)]
IMPROVED VERSION APPLIED:
The attached patch completes the following TODO item:
* Generate failure on short COPY lines rather than pad NULLs
I also restructed a lot of the existing COPY code, did some code
review on the column list patch sent in by Brent Verner a little
while ago, and added some regression tests. I also added an
explicit check (and resultant error) for extra data before
the end-of-line.
Bruce Momjian [Tue, 30 Jul 2002 16:35:05 +0000 (16:35 +0000)]
Added support for schemas and quotes in tab-complete.c, as well as
a few other things:
* Made all references to the pg_* tables absolute, by specifying
the pg_catalog schema.
* Added SCHEMA as a create/delete completion option.
* Added SCHEMA completion as: SELECT nspname FROM
pg_catalog.pg_namespace
WHERE substr(nspname,1,%d)='%s'
* Added completion of "INSERT INTO <table> (" with attribute names.
* Added completion of "INSERT INTO <table> (attribs)" with
VALUES or SELECT
* Added limited locking completion: only for one table:
"LOCK" and "LOCK TABLE" now both get a completion list of tables
Complete with "IN" for LOCK [TABLE] <table>
Complete LOCK [TABLE] <table> IN with a lock mode
* Added a very simple WHERE finisher that uses the previous word
as a table lookup for attributes.
* Added quote support when parsing "previous words". In other words,
hitting tab after INSERT INTO "foo bar baby"
now does the right thing and recognizes "foo bar baby" as one word.
Letting tab-complete quote things that should be quoted seems to be
temporarily ifdef'ed out due to readline compatibility problems.
Can anyone elaborate on this?
Bruce Momjian [Tue, 30 Jul 2002 16:33:21 +0000 (16:33 +0000)]
This should fix a bug where a row that was updated or
deleted that had another row inserted/updated to its old
value during the same statement or other statements before the
integrity check for noaction would incorrectly error. This
could happen in deferred constraints or due to triggers or
functions. It's effectively a reworking of the previous patch that
did a not exists to instead do a separate check.
Tom Lane [Tue, 30 Jul 2002 16:33:08 +0000 (16:33 +0000)]
Since we're depending on %option noyywrap in the main scanner now,
we may as well use it in all our flex files. Make all the flex files
have a consistent set of options.
Bruce Momjian [Tue, 30 Jul 2002 16:31:11 +0000 (16:31 +0000)]
As mentioned above, here is my contrib/tablefunc patch. It includes
three functions which exercise the tablefunc API.
show_all_settings()
- returns the same information as SHOW ALL, but as a query result
normal_rand(int numvals, float8 mean, float8 stddev, int seed)
- returns a set of normally distributed float8 values
- This routine implements Algorithm P (Polar method for normal
deviates) from Knuth's _The_Art_of_Computer_Programming_, Volume 2,
3rd ed., pages 122-126. Knuth cites his source as "The polar
method", G. E. P. Box, M. E. Muller, and G. Marsaglia,
_Annals_Math,_Stat._ 29 (1958), 610-611.
crosstabN(text sql)
- returns a set of row_name plus N category value columns
- crosstab2(), crosstab3(), and crosstab4() are defined for you,
but you can create additional crosstab functions per directions
in the README.
Bruce Momjian [Tue, 30 Jul 2002 16:20:03 +0000 (16:20 +0000)]
Here are two patches. The guc_and_tablefunc patch addresses the two
changes mentioned above, and also adds a new function to the tablefunc
API. The tablefunc API change adds the following function:
* Oid foidGetTypeId(Oid foid) - Get a function's typeid given the
* function Oid. Use this together with TypeGetTupleDesc() to get a
* TupleDesc which is derived from the function's declared return type.
In the next post I'll send the contrib/tablefunc patch, which
illustrates the usage of this new function. Also attached is a doc patch
for this change. The doc patch also adds a function that I failed to
document previously.
Tom Lane [Tue, 30 Jul 2002 05:24:56 +0000 (05:24 +0000)]
Rewrite xindex.sgml for CREATE OPERATOR CLASS. catalogs.sgml finally
contains descriptions of every single system table. Update 'complex'
tutorial example too.
Tom Lane [Mon, 29 Jul 2002 23:46:35 +0000 (23:46 +0000)]
Centralize code for interpreting schema references, which had gotten
copied more places than I first thought it would. This fixes a bug:
a couple of these places were neglecting to enforce USAGE access on
explicitly-referenced schemas.
Tom Lane [Mon, 29 Jul 2002 22:14:11 +0000 (22:14 +0000)]
Implement CREATE/DROP OPERATOR CLASS. Work still remains: need more
documentation (xindex.sgml should be rewritten), need to teach pg_dump
about it, need to update contrib modules that currently build pg_opclass
entries by hand. Original patch by Bill Studenmund, grammar adjustments
and general update for 7.3 by Tom Lane.
Attempts to use the obsoleted attributes of pg_operator or pg_proc
in the CREATE commands will be greeted by a warning. For pg_type,
there is no warning (yet) because pg_dump scripts still contain these
attributes.
Also remove new but already obsolete spellings
isVolatile, isStable, isImmutable in WITH clause. (Use new syntax
instead.)
Barry Lind [Tue, 23 Jul 2002 03:59:55 +0000 (03:59 +0000)]
Initial restructuring to add jdbc3 support. There was a significant amount
of duplicated code between the jdbc1 and jdbc2. This checkin restructures
the code so that the duplication is removed so that the jdbc3 support
can be added without adding yet another copy of everything. Also many
classes were renamed to avoid confusion with multiple different objects
having the same name. The timestamp tests were also updated to add support
for testing timestamp without time zone in addition to timestamp with time zone
Tom Lane [Sat, 20 Jul 2002 19:55:38 +0000 (19:55 +0000)]
Tweak CreateTrigger() so that the OID used in the name of an
RI_ConstraintTrigger is the same OID assigned to the pg_trigger row.
This reduces consumption of OIDs and may ease debugging.
Bruce Momjian [Sat, 20 Jul 2002 16:45:07 +0000 (16:45 +0000)]
Mark items as done:
> * -Add GUC parameter for DATESTYLE
> o -Allow specification of column names
> o -Change syntax to WITH DELIMITER, (keep old syntax around?)
> o -Remove SET KSQO option now that OR processing is improved (Tom)
> o -Allow SHOW to output as a query result, like EXPLAIN
> * -Add SQL92 schemas (Tom)
Bruce Momjian [Sat, 20 Jul 2002 05:57:31 +0000 (05:57 +0000)]
I can't remember who said they were working on schema related psql
changes, but I kept finding myself wishing I could see what schema a
table or view exists in when I use \dt, \dv, etc. So, here is a patch
which does just that.
It sorts on "Schema" first, and "Name" second.
It also changes the test for system objects to key off the namespace
name starting with 'pg_' instead of the object name.
Sample output:
test=# create schema testschema;
CREATE SCHEMA
test=# create view testschema.ts_view as select 1;
CREATE VIEW
test=# \dv
List of relations
Name | Schema | Type | Owner
--------------------+------------+------+----------
__testpassbyval | public | view | postgres
fooview | public | view | postgres
master_pg_proc | public | view | postgres
rmt_pg_proc | public | view | postgres
vw_dblink_get_pkey | public | view | postgres
vw_dblink_replace | public | view | postgres
ts_view | testschema | view | postgres
(7 rows)
Bruce Momjian [Sat, 20 Jul 2002 05:49:28 +0000 (05:49 +0000)]
> 2. This patch includes the same Table Function API fixes that I
> submitted on July 9:
>
> http://archives.postgresql.org/pgsql-patches/2002-07/msg00056.php
>
> Please disregard that one *if* this one is applied. If this one is
> rejected please go ahead with the July 9th patch.
The July 9th Table Function API patch mentioned above is now in CVS, so
here is an updated version of the guc patch which should apply cleanly
against CVS tip.
Bruce Momjian [Sat, 20 Jul 2002 05:43:31 +0000 (05:43 +0000)]
Hello, i noticed that win32 native stopped working/compiling after the SSL merge
.
So i took the opportunity to fix some stuff:
1. Made the thing compile (typos & needed definitions) with the new pqsecure_* s
tuff, and added fe-secure.c to the win32.mak makefile.
2. Fixed some MULTIBYTE compile errors (when building without MB support).
3. Made it do that you can build with debug info: "nmake -f win32.mak DEBUG=1".
4. Misc small compiler speedup changes.
The resulting .dll has been tested in production, and everything seems ok.
I CC:ed -hackers because i'm not sure about two things:
1. In libpq-int.h I typedef ssize_t as an int because Visual C (v6.0)
doesn't de fine ssize_t. Is that ok, or is there any standard about what
type should be use d for ssize_t?
2. To keep the .dll api consistent regarding MULTIBYTE I just return -1
in fe-connect.c:PQsetClientEncoding() instead of taking away the whole
function. I wonder if i should do any compares with the
conn->client_encoding and return 0 if not hing would have changed (if so
how do i check that?).
Bruce Momjian [Sat, 20 Jul 2002 05:16:59 +0000 (05:16 +0000)]
oid is needed, it is added at the end of the struct (after the null
bitmap, if present).
Per Tom Lane's suggestion the information whether a tuple has an oid
or not is carried in the tuple descriptor. For debugging reasons
tdhasoid is of type char, not bool. There are predefined values for
WITHOID, WITHOUTOID and UNDEFOID.
This patch has been generated against a cvs snapshot from last week
and I don't expect it to apply cleanly to current sources. While I
post it here for public review, I'm working on a new version against a
current snapshot. (There's been heavy activity recently; hope to
catch up some day ...)
This is a long patch; if it is too hard to swallow, I can provide it
in smaller pieces:
Part 1: Accessor macros
Part 2: tdhasoid in TupDesc
Part 3: Regression test
Part 4: Parameter withoid to heap_addheader
Part 5: Eliminate t_oid from HeapTupleHeader
Part 2 is the most hairy part because of changes in the executor and
even in the parser; the other parts are straightforward.
Up to part 4 the patched postmaster stays binary compatible to
databases created with an unpatched version. Part 5 is small (100
lines) and finally breaks compatibility.
Bruce Momjian [Sat, 20 Jul 2002 04:59:10 +0000 (04:59 +0000)]
The attached patch fixes a build problem with GEQO when using the
PX recombination operator, changes some elog() messages from LOG
to DEBUG1, puts some debugging functions inside the appropriate
#ifdef (not enabled by default), and makes a few other minor
cleanups.
BTW, the elog() change is motivated by at least one user who
has sent a concerned email to -general asking exactly what the
"ERX recombination operator" is, and what it is doing to their
DBMS.
Bruce Momjian [Sat, 20 Jul 2002 04:57:13 +0000 (04:57 +0000)]
This patch fixes a regression caused by my recent changes to heap
tuple header. The fix is based on the thought that HEAP_MOVED_IN is
not needed any more as soon as HEAP_XMIN_COMMITTED has been set. So
in tqual.c and vacuum.c the HEAP_MOVED bits are cleared when
HEAP_XMIN_COMMITTED is set.
Vacuum robustness is enhanced by rearranging ifs, so that we have a
chance to elog(ERROR, ...) before an assertion fails.
> Hi Tatsuo,
>
> I've attached a patch for the version of pgbench in CVS. It includes the
> following changes:
>
> - fix some spelling mistakes, indentation stuff, etc.
>
> - minor code cleanup -- (void) args instead of (), etc.
>
> - allocate the state array dynamically, so that it is only as
> large as needed. This reduces the memory consumption of pgbench
> slightly, and makes a larger MAXCLIENTS setting possible
>
> - (the only controversial change) add an option "-l" to log
> transaction latencies to a file. The "transaction latency"
> is the time between when the BEGIN is issued and the transaction
> commits. This is written to a file, along with the client #
> and the transaction #. The data in the file can then be used
> for things like:
>
> - consistency analysis: is the TPS the same through the
> entire run of pgbench, or does it change?
>
> - more detailed stats: what is the average latency, worse-case
> latency, best-case latency?
>
> - graphs: feed the data to gnuplot, graph latency versus. time
>
> - etc.
>
> I was going to store this data in memory and write it to disk
> at the end of the pgbench run, but that isn't feasible because
> the data can be very large: for example, ~70MB if benchmarking
> 128 clients doing 100,000 transactions each.
>
> Cheers,
>
> Neil
Peter Eisentraut [Thu, 18 Jul 2002 23:11:32 +0000 (23:11 +0000)]
pg_cast table, and standards-compliant CREATE/DROP CAST commands, plus
extension to create binary compatible casts. Includes dependency tracking
as well.
pg_proc.proimplicit is now defunct, but will be removed in a separate
commit.
pg_dump provides a migration path from the previous scheme to declare
casts. Dumping binary compatible casts is currently impossible, though.
projects / postgresql / log