]> granicus.if.org Git - postgresql/log
postgresql
23 years agoGive a more intelligible and useful error message for SELECT with no
Tom Lane [Tue, 13 Nov 2001 02:05:27 +0000 (02:05 +0000)]
Give a more intelligible and useful error message for SELECT with no
destination in plpgsql.

23 years agoUpdate TODO list.
Bruce Momjian [Tue, 13 Nov 2001 01:14:57 +0000 (01:14 +0000)]
Update TODO list.

23 years agoTweak parser so that there is a defined representation for datatypes
Tom Lane [Mon, 12 Nov 2001 21:04:46 +0000 (21:04 +0000)]
Tweak parser so that there is a defined representation for datatypes
bpchar, bit, numeric with typmod -1.  Alter format_type so that this
representation is printed when the typmod is -1.  This ensures that
tables having such columns can be pg_dump'd and reloaded correctly.
Also, remove the rather useless and non-SQL-compliant default
precision and scale for type NUMERIC.  A numeric column declared as
such (with no precision/scale) will now have typmod -1 which means
that numeric values of any precision/scale can be stored in it,
without conversion to a uniform scale.  This seems significantly
more useful than the former behavior.  Part of response to bug #513.

23 years agoUpdate TODO list.
Bruce Momjian [Mon, 12 Nov 2001 20:26:39 +0000 (20:26 +0000)]
Update TODO list.

23 years agoIf the alternatives for a CASE construct all have the same typmod,
Tom Lane [Mon, 12 Nov 2001 20:05:24 +0000 (20:05 +0000)]
If the alternatives for a CASE construct all have the same typmod,
use that typmod not -1 as the typmod of the CASE result.
Part of response to bug#513.

23 years agoIf the inputs of a UNION/INTERSECT/EXCEPT construct all agree on the
Tom Lane [Mon, 12 Nov 2001 20:04:20 +0000 (20:04 +0000)]
If the inputs of a UNION/INTERSECT/EXCEPT construct all agree on the
typmod of a particular column, mark the output with that same typmod,
not -1 as formerly.  -1 is still used if there is any disagreement.
Part of response to bug#513.

23 years agofixed bug in ResultSet. Version 1.29 backed out two previous fixes (1.26 and 1.25...
Barry Lind [Mon, 12 Nov 2001 19:59:46 +0000 (19:59 +0000)]
fixed bug in ResultSet.  Version 1.29 backed out two previous fixes (1.26 and 1.25).  This checkin add back those two previous fixes.  Problem reported by Daniel Germain

23 years agoAdd some more index entries.
Peter Eisentraut [Mon, 12 Nov 2001 19:19:39 +0000 (19:19 +0000)]
Add some more index entries.

23 years agoCommit to support MD5 passwords as per the backend for 7.2. This patch was submitted...
Barry Lind [Mon, 12 Nov 2001 19:11:56 +0000 (19:11 +0000)]
Commit to support MD5 passwords as per the backend for 7.2.  This patch was submitted by Jeremy Wohl jeremyw-pgjdbc@igmus.org

23 years agoAdd port number to pg_dumplo.
Bruce Momjian [Mon, 12 Nov 2001 17:44:14 +0000 (17:44 +0000)]
Add port number to pg_dumplo.

andrea gelmini

23 years agoRepair crash in EvalPlanQual of query involving nestloop with inner
Tom Lane [Mon, 12 Nov 2001 17:18:06 +0000 (17:18 +0000)]
Repair crash in EvalPlanQual of query involving nestloop with inner
index scan.  Problem was that link to outer tuple wasn't being stored
everyplace it needed to be.

23 years agoRemove duplicate extern declaration.
Tom Lane [Mon, 12 Nov 2001 16:34:58 +0000 (16:34 +0000)]
Remove duplicate extern declaration.

23 years agopsql's \do was going out of its way to lie about the result type of
Tom Lane [Mon, 12 Nov 2001 15:57:08 +0000 (15:57 +0000)]
psql's \do was going out of its way to lie about the result type of
operators.  Should report the declared oprresult type, not the return type
of the underlying proc, which might be only binary-compatible (cf.
textcat entries).

23 years agoUpdate FAQ.
Bruce Momjian [Mon, 12 Nov 2001 07:38:57 +0000 (07:38 +0000)]
Update FAQ.

23 years agoUpdate TODO list.
Bruce Momjian [Mon, 12 Nov 2001 07:10:58 +0000 (07:10 +0000)]
Update TODO list.

23 years ago Tables without oids wouldn't be able to be
Bruce Momjian [Mon, 12 Nov 2001 06:09:09 +0000 (06:09 +0000)]
    Tables without oids wouldn't be able to be
used inside fk constraints, since some of the checks
in the trigger did a SELECT oid.  Since the oid wasn't
actually used, I changed this to SELECT 1.  My test
case with non-oid tables now works and fk regression
appears to run fine on my machine.

Stephan Szabo

23 years agoRemove a bunch more no-longer-used stuff in libpq-be.h.
Tom Lane [Mon, 12 Nov 2001 05:43:25 +0000 (05:43 +0000)]
Remove a bunch more no-longer-used stuff in libpq-be.h.

23 years agoUpdate TODO list.
Bruce Momjian [Mon, 12 Nov 2001 05:35:48 +0000 (05:35 +0000)]
Update TODO list.

23 years agoSuppress duplicate error messages in pq_flush. Write error messages to
Tom Lane [Mon, 12 Nov 2001 04:54:08 +0000 (04:54 +0000)]
Suppress duplicate error messages in pq_flush.  Write error messages to
postmaster log with elog(DEBUG) so that they will be timestamped etc.
Once upon a time I think elog() was unsafe here, but it shouldn't be anymore.

23 years agosend() attempt for IDENT communication should retry on EINTR.
Tom Lane [Mon, 12 Nov 2001 04:29:23 +0000 (04:29 +0000)]
send() attempt for IDENT communication should retry on EINTR.

23 years agoThe PacketReceive/PacketSend routines aren't used anymore.
Tom Lane [Mon, 12 Nov 2001 04:19:15 +0000 (04:19 +0000)]
The PacketReceive/PacketSend routines aren't used anymore.

23 years agoUpdate MD5 documentation.
Bruce Momjian [Mon, 12 Nov 2001 03:57:16 +0000 (03:57 +0000)]
Update MD5 documentation.

23 years agoAdd more MD5 documentation now that odbc/jdbc support it.
Bruce Momjian [Mon, 12 Nov 2001 02:35:16 +0000 (02:35 +0000)]
Add more MD5 documentation now that odbc/jdbc support it.

23 years agoSync up both files.
Bruce Momjian [Mon, 12 Nov 2001 02:05:17 +0000 (02:05 +0000)]
Sync up both files.

23 years ago Add comments and remove CVS tag from md5.c so they remain identical.
Bruce Momjian [Mon, 12 Nov 2001 02:00:18 +0000 (02:00 +0000)]
 Add comments and remove CVS tag from md5.c so they remain identical.

23 years agoAdd comments and remove CVS tag from md5.c so they remain identical.
Bruce Momjian [Mon, 12 Nov 2001 01:59:43 +0000 (01:59 +0000)]
Add comments and remove CVS tag from md5.c so they remain identical.

23 years agoUpdate md5 to match.
Bruce Momjian [Mon, 12 Nov 2001 01:56:12 +0000 (01:56 +0000)]
Update md5 to match.

23 years agoAdd comments of duplicate definitions in interfaces/odbc/md5.h.
Bruce Momjian [Mon, 12 Nov 2001 01:52:46 +0000 (01:52 +0000)]
Add comments of duplicate definitions in interfaces/odbc/md5.h.

23 years agoAdd code to check that md5.c files are in sync.
Bruce Momjian [Mon, 12 Nov 2001 01:42:03 +0000 (01:42 +0000)]
Add code to check that md5.c files are in sync.

23 years agoIndent new rename.c for Tom Lane.
Bruce Momjian [Mon, 12 Nov 2001 01:34:50 +0000 (01:34 +0000)]
Indent new rename.c for Tom Lane.

23 years agoAdd md5 authentication support thanks to Bruce Momjian.
Hiroshi Inoue [Mon, 12 Nov 2001 00:54:28 +0000 (00:54 +0000)]
Add md5 authentication support thanks to Bruce Momjian.

23 years agoMake ALTER TABLE RENAME update foreign-key trigger arguments correctly.
Tom Lane [Mon, 12 Nov 2001 00:46:36 +0000 (00:46 +0000)]
Make ALTER TABLE RENAME update foreign-key trigger arguments correctly.
Brent Verner, with review and kibitzing from Tom Lane.

23 years agoClean up a bunch of ScanKeyEntryInitialize calls that weren't bothering
Tom Lane [Mon, 12 Nov 2001 00:00:55 +0000 (00:00 +0000)]
Clean up a bunch of ScanKeyEntryInitialize calls that weren't bothering
to apply the proper Datum conversion macros to search key values.

23 years agoTest program needs to declare MaxBackends, per Bernd Tegge.
Tom Lane [Sun, 11 Nov 2001 22:12:00 +0000 (22:12 +0000)]
Test program needs to declare MaxBackends, per Bernd Tegge.

23 years agoDefend against brain-dead QNX implementation of qsort().
Tom Lane [Sun, 11 Nov 2001 22:00:25 +0000 (22:00 +0000)]
Defend against brain-dead QNX implementation of qsort().
Per report from Bernd Tegge, 10-Nov-01.

23 years agoIn find_mergeclauses_for_pathkeys, it's okay to return multiple merge
Tom Lane [Sun, 11 Nov 2001 20:33:53 +0000 (20:33 +0000)]
In find_mergeclauses_for_pathkeys, it's okay to return multiple merge
clauses per path key.  Indeed, we *must* do so or we will be unable to
form a valid plan for FULL JOIN with overlapping join conditions, eg
select * from a full join b on
a.v1 = b.v1 and a.v2 = b.v2 and a.v1 = b.v2.

23 years agoFix typo pointed out by bruno@wolff.to
Bruce Momjian [Sun, 11 Nov 2001 19:24:28 +0000 (19:24 +0000)]
Fix typo pointed out by bruno@wolff.to

23 years agoFix for compiling libpq++ on Solaris with Sun SPRO6U2.
Bruce Momjian [Sun, 11 Nov 2001 19:20:53 +0000 (19:20 +0000)]
Fix for compiling libpq++ on Solaris with Sun SPRO6U2.

Denis A Ustimenko

23 years agosort_inner_and_outer needs a check to ensure that it's consumed all the
Tom Lane [Sun, 11 Nov 2001 19:18:54 +0000 (19:18 +0000)]
sort_inner_and_outer needs a check to ensure that it's consumed all the
mergeclauses in RIGHT/FULL join cases, just like the other routines have.
I'm not quite sure why I thought it didn't need one --- but Nick
Fankhauser's recent bug report proves that it does.

23 years agoUse abbreviated connection string more widely.
Hiroshi Inoue [Sun, 11 Nov 2001 07:24:36 +0000 (07:24 +0000)]
Use abbreviated connection string more widely.
This seems to fix the trouble with PowerBuilder
reported by Magbus Weber.

23 years agoDefend against possibility that SSL error reporting mechanism returns
Tom Lane [Sun, 11 Nov 2001 02:09:05 +0000 (02:09 +0000)]
Defend against possibility that SSL error reporting mechanism returns
a NULL pointer.  Per report from Stephen Pillinger 8-Nov-01.

23 years agoClean up usage-statistics display code (ShowUsage and friends). StatFp
Tom Lane [Sat, 10 Nov 2001 23:51:14 +0000 (23:51 +0000)]
Clean up usage-statistics display code (ShowUsage and friends).  StatFp
is gone, usage messages now go through elog(DEBUG).

23 years agoRestructure child-exit logging messages for easier translation,
Tom Lane [Sat, 10 Nov 2001 23:06:12 +0000 (23:06 +0000)]
Restructure child-exit logging messages for easier translation,
per suggestion from Peter.

23 years agoAllow TIMESTAMP, VARCHAR, et al to be used as unquoted column names,
Tom Lane [Sat, 10 Nov 2001 22:31:49 +0000 (22:31 +0000)]
Allow TIMESTAMP, VARCHAR, et al to be used as unquoted column names,
though alas not as unquoted function names.  De-reserve a bunch of
keywords that could have been in ColId rather than ColLabel all along.
Per recent proposal in pgsql-patches.

23 years agoRepair erroneous ALTER DROP CONSTRAINT example.
Tom Lane [Sat, 10 Nov 2001 20:13:37 +0000 (20:13 +0000)]
Repair erroneous ALTER DROP CONSTRAINT example.

23 years agoJason Davies patch to getImported/getExported keys
Dave Cramer [Fri, 9 Nov 2001 02:57:50 +0000 (02:57 +0000)]
Jason Davies patch to getImported/getExported keys

23 years agoSupport for images in jadetex and pdfjadetex. Formatting improvements for
Peter Eisentraut [Thu, 8 Nov 2001 23:46:29 +0000 (23:46 +0000)]
Support for images in jadetex and pdfjadetex.  Formatting improvements for
lists, bibliography, index, page numbering.  Add comments.

23 years agoUse <sect1> rather than <section> to simplify formatting.
Peter Eisentraut [Thu, 8 Nov 2001 23:44:01 +0000 (23:44 +0000)]
Use <sect1> rather than <section> to simplify formatting.

23 years agoAdd missing spaces
Peter Eisentraut [Thu, 8 Nov 2001 23:42:23 +0000 (23:42 +0000)]
Add missing spaces

23 years agoFix indexterms
Peter Eisentraut [Thu, 8 Nov 2001 23:41:12 +0000 (23:41 +0000)]
Fix indexterms

23 years agoMove the indexterms around so they don't affect formatting
Peter Eisentraut [Thu, 8 Nov 2001 23:40:40 +0000 (23:40 +0000)]
Move the indexterms around so they don't affect formatting

23 years agoUse <important> rather than <caution> when the condition isn't hazardous,
Peter Eisentraut [Thu, 8 Nov 2001 23:39:22 +0000 (23:39 +0000)]
Use <important> rather than <caution> when the condition isn't hazardous,
because the result looks to "scary" otherwise.

23 years agoSmall tuning
Peter Eisentraut [Thu, 8 Nov 2001 23:37:50 +0000 (23:37 +0000)]
Small tuning

23 years agoSpacing improvements
Peter Eisentraut [Thu, 8 Nov 2001 23:36:55 +0000 (23:36 +0000)]
Spacing improvements

23 years agoDon't use dot in id attribute, which confuses the index generation code in
Peter Eisentraut [Thu, 8 Nov 2001 23:36:18 +0000 (23:36 +0000)]
Don't use dot in id attribute, which confuses the index generation code in
this particular instance.

23 years agoCleanup of bibliography
Peter Eisentraut [Thu, 8 Nov 2001 23:34:33 +0000 (23:34 +0000)]
Cleanup of bibliography

23 years agoFix indenting for 'extern "C"' cases.
Bruce Momjian [Thu, 8 Nov 2001 20:37:52 +0000 (20:37 +0000)]
Fix indenting for 'extern "C"' cases.

23 years agoUpdate TODO list.
Bruce Momjian [Thu, 8 Nov 2001 18:24:55 +0000 (18:24 +0000)]
Update TODO list.

23 years agoMake extern C handling more flexible.
Bruce Momjian [Thu, 8 Nov 2001 17:03:23 +0000 (17:03 +0000)]
Make extern C handling more flexible.

23 years agoWhen given oversized key, encrypt/decrypt corrupted
Bruce Momjian [Thu, 8 Nov 2001 15:56:58 +0000 (15:56 +0000)]
When given oversized key, encrypt/decrypt corrupted
memory.  This fixes it.  Also a free() was missing.

marko

23 years agoPostgreSQL works again on Mac OS X 10.1. Hold your nose before
Tom Lane [Thu, 8 Nov 2001 04:24:03 +0000 (04:24 +0000)]
PostgreSQL works again on Mac OS X 10.1.  Hold your nose before
investigating the kluge that makes it so...

23 years agoAdd casts to suppress compiler warnings observed on Darwin platform
Tom Lane [Thu, 8 Nov 2001 04:05:13 +0000 (04:05 +0000)]
Add casts to suppress compiler warnings observed on Darwin platform
(surprised no one has reported these yet...)

23 years agoFix replacement of extern C string.
Bruce Momjian [Wed, 7 Nov 2001 22:10:02 +0000 (22:10 +0000)]
Fix replacement of extern C string.

23 years agoFix typo.
Bruce Momjian [Wed, 7 Nov 2001 21:29:04 +0000 (21:29 +0000)]
Fix typo.

23 years agoPrevent indenting of 'extern "C"' blocks.
Bruce Momjian [Wed, 7 Nov 2001 21:24:28 +0000 (21:24 +0000)]
Prevent indenting of 'extern "C"' blocks.

23 years agoMention new =NULL change in Migration section.
Bruce Momjian [Wed, 7 Nov 2001 17:27:36 +0000 (17:27 +0000)]
Mention new =NULL change in Migration section.

23 years agoUpdate TODO list.
Bruce Momjian [Wed, 7 Nov 2001 05:27:46 +0000 (05:27 +0000)]
Update TODO list.

23 years agoFix wrapping.
Bruce Momjian [Wed, 7 Nov 2001 03:12:20 +0000 (03:12 +0000)]
Fix wrapping.

23 years agoCorrect, not too ambitious, not too
Bruce Momjian [Wed, 7 Nov 2001 03:10:59 +0000 (03:10 +0000)]
Correct, not too ambitious, not too
long for an enhancement history item? :)

--
Serguei A. Mokhov
~

23 years agoSome post pgident run updates,
Bruce Momjian [Wed, 7 Nov 2001 03:09:15 +0000 (03:09 +0000)]
Some post pgident run updates,
one fuzzy translation fix, some
other messages tweaking. Theoretically,
should be up-to-date by now.

Please apply to /src/interfaces/libpq/ru.po

--
Serguei A. Mokhov

23 years agoUpdate TODO list.
Bruce Momjian [Wed, 7 Nov 2001 01:34:03 +0000 (01:34 +0000)]
Update TODO list.

23 years agoAdd some notes about conditional rules used with views.
Tom Lane [Tue, 6 Nov 2001 23:54:32 +0000 (23:54 +0000)]
Add some notes about conditional rules used with views.

23 years agoUpdate TODO list.
Bruce Momjian [Tue, 6 Nov 2001 23:42:44 +0000 (23:42 +0000)]
Update TODO list.

23 years agoAdd mention of another MySQL conversion tool.
Bruce Momjian [Tue, 6 Nov 2001 22:25:53 +0000 (22:25 +0000)]
Add mention of another MySQL conversion tool.

23 years agotag'd and branded as b2 ...
PostgreSQL Daemon [Tue, 6 Nov 2001 20:12:07 +0000 (20:12 +0000)]
tag'd and branded as b2 ...

23 years agoUpdate for latest version of horology test. REL7_2_BETA2
Tom Lane [Tue, 6 Nov 2001 18:03:48 +0000 (18:03 +0000)]
Update for latest version of horology test.

23 years agoClean up formatting of child process exit-status reports so that they
Tom Lane [Tue, 6 Nov 2001 18:02:48 +0000 (18:02 +0000)]
Clean up formatting of child process exit-status reports so that they
are correct, consistent, and complete ... motivated by gripe from
Oliver Elphick, but I see someone had already made an incomplete stab
at this.

23 years agoFix last (?) problem with sensitivity to daylight savings time status
Thomas G. Lockhart [Tue, 6 Nov 2001 16:31:13 +0000 (16:31 +0000)]
Fix last (?) problem with sensitivity to daylight savings time status
 when running the regression test. Reported by Tom Lane.

23 years agoAdd British Double Standard Time (BDST) per mailing list report.
Thomas G. Lockhart [Tue, 6 Nov 2001 16:29:51 +0000 (16:29 +0000)]
Add British Double Standard Time (BDST) per mailing list report.

23 years agoUse PostgreSQL's standard declaration for a 32-bit integer to define
Thomas G. Lockhart [Tue, 6 Nov 2001 16:27:31 +0000 (16:27 +0000)]
Use PostgreSQL's standard declaration for a 32-bit integer to define
 the cash data type.

23 years agoAdd Darwin to list of known systems for ps_status.
Tom Lane [Tue, 6 Nov 2001 01:15:29 +0000 (01:15 +0000)]
Add Darwin to list of known systems for ps_status.

23 years agoThe extra semaphore that proc.c now allocates for checkpoint processes
Tom Lane [Tue, 6 Nov 2001 00:38:26 +0000 (00:38 +0000)]
The extra semaphore that proc.c now allocates for checkpoint processes
should be accounted for in the PROC_SEM_MAP_ENTRIES() macro.  Otherwise
the ports that rely on this macro to size data structures are broken.
Mea culpa.

23 years agoFix coredump in plpgsql when trying to return a rowtype result.
Tom Lane [Mon, 5 Nov 2001 19:41:56 +0000 (19:41 +0000)]
Fix coredump in plpgsql when trying to return a rowtype result.
Need to return a TupleTableSlot, not just a bare tuple.

23 years agoNew pgindent run with fixes suggested by Tom. Patch manually reviewed,
Bruce Momjian [Mon, 5 Nov 2001 17:46:40 +0000 (17:46 +0000)]
New pgindent run with fixes suggested by Tom.  Patch manually reviewed,
initdb/regression tests pass.

23 years agoPreparation for the parameter array handling.
Hiroshi Inoue [Mon, 5 Nov 2001 10:35:14 +0000 (10:35 +0000)]
Preparation for the parameter array handling.

23 years ago1) Fix a few bugs about SQLGetData()
Hiroshi Inoue [Mon, 5 Nov 2001 09:46:17 +0000 (09:46 +0000)]
1) Fix a few bugs about SQLGetData()
   reported by Mika Mantyla.
2) Timestamp precision.
3) Separate ODBC3.0 files.

23 years agoMore cleanup for stuff after closing brace in first column.
Bruce Momjian [Mon, 5 Nov 2001 06:37:51 +0000 (06:37 +0000)]
More cleanup for stuff after closing brace in first column.

23 years agoHandle tabs after closing brace in first column with less indenting.
Bruce Momjian [Mon, 5 Nov 2001 05:47:50 +0000 (05:47 +0000)]
Handle tabs after closing brace in first column with less indenting.

23 years agoImprove readability of script.
Bruce Momjian [Mon, 5 Nov 2001 05:18:43 +0000 (05:18 +0000)]
Improve readability of script.

23 years agoCREATE TABLE foo (x,y,z) AS SELECT ... can't apply target column names
Tom Lane [Mon, 5 Nov 2001 05:00:14 +0000 (05:00 +0000)]
CREATE TABLE foo (x,y,z) AS SELECT ... can't apply target column names
to the target list in gram.y; it must wait till after expansion of the
target list in analyze.c.  Per bug report 4-Nov:
lx=# CREATE TABLE abc (a char, b char, c char);
CREATE
lx=# CREATE TABLE xyz (x, y, z) AS SELECT * FROM abc;
ERROR:  CREATE TABLE/AS SELECT has mismatched column count

23 years agoRemove ill-considered Assert.
Tom Lane [Mon, 5 Nov 2001 01:34:37 +0000 (01:34 +0000)]
Remove ill-considered Assert.

23 years agoPull in variables defined in structs; had too many tabs.
Bruce Momjian [Sun, 4 Nov 2001 21:27:41 +0000 (21:27 +0000)]
Pull in variables defined in structs;  had too many tabs.

23 years agoFix now-obsolete comment.
Tom Lane [Sun, 4 Nov 2001 20:12:57 +0000 (20:12 +0000)]
Fix now-obsolete comment.

23 years agoVersion was 3.3 but last released version was 3.1. Setting to match rest
D'Arcy J.M. Cain [Sun, 4 Nov 2001 20:11:01 +0000 (20:11 +0000)]
Version was 3.3 but last released version was 3.1.  Setting to match rest
of the documentation in preparation for upcoming release.

23 years agoMerge three existing ways of signaling postmaster from child processes,
Tom Lane [Sun, 4 Nov 2001 19:55:31 +0000 (19:55 +0000)]
Merge three existing ways of signaling postmaster from child processes,
so that only one signal number is used not three.  Flags in shared
memory tell the reason(s) for the current signal.  This method is
extensible to handle more signal reasons without chewing up even more
signal numbers, but the immediate reason is to keep pg_pwd reloads
separate from SIGHUP processing in the postmaster.
Also clean up some problems in the postmaster with delayed response to
checkpoint status changes --- basically, it wouldn't schedule a checkpoint
if it wasn't getting connection requests on a regular basis.

23 years agoNote that PyGreSQL has been checked against Python 2.1 now.
D'Arcy J.M. Cain [Sun, 4 Nov 2001 19:47:16 +0000 (19:47 +0000)]
Note that PyGreSQL has been checked against Python 2.1 now.

23 years agoThe "%d", while syntactically correct, was confusing. Added a space to
D'Arcy J.M. Cain [Sun, 4 Nov 2001 19:42:13 +0000 (19:42 +0000)]
The "%d", while syntactically correct, was confusing.  Added a space to
make it clearer that d was the argument to the format operator.

23 years ago- Fix compiler warning in pg_restore
Philip Warner [Sun, 4 Nov 2001 04:05:36 +0000 (04:05 +0000)]
- Fix compiler warning in pg_restore
- Fix handling of {data/schema}-only restores when using a full
  backup file; prior version was restoring schema in data-only
  restores. Added enum to make code easier to understand.

23 years agoGot "ADD" to appear only in ALTER TABLE and not CREATE TABLE
Bruce Momjian [Sun, 4 Nov 2001 03:08:11 +0000 (03:08 +0000)]
Got "ADD" to appear only in ALTER TABLE and not CREATE TABLE
UNIQUE-PRIMARY KEY notice message.  This is what Christopher wanted from
his patch.

23 years agoRemove "ADD" from TABLE / ADD UNIQUE-PRIMARY error message because the
Bruce Momjian [Sun, 4 Nov 2001 02:41:09 +0000 (02:41 +0000)]
Remove "ADD" from TABLE / ADD UNIQUE-PRIMARY error message because the
same code is called for both creation and alter.  Not worth worrying
about.