[PATCHES] SQL/XML publishing function experimental patch
Hello
This patch contains SQL/XML public function XMLFOREST, XMLELEMENT, ... based
on CString type. This patch hasn't high quality, but can be usefull for
testing funcionality, and maybe for some people. Patch needs initdb and
following registration:
CREATE AGGREGATE pg_catalog.fast_cstring_concat (
BASETYPE = cstring,
SFUNC = xmlagg_trans,
STYPE = text,
FINALFUNC = xmlagg_final);
I invite all notes, ..
Best regards
Pavel Stehule
_
Chcete sdilet sve obrazky a hudbu s prateli? http://messenger.msn.cz/
diff -c -r pgsql.clean/src/backend/executor/execQual.c pgsql2/src/backend/executor/execQual.c
*** pgsql.clean/src/backend/executor/execQual.c 2005-06-27 00:05:36.0 +0200
--- pgsql2/src/backend/executor/execQual.c 2005-09-05 17:41:48.0 +0200
***
*** 53,59
#include utils/lsyscache.h
#include utils/memutils.h
#include utils/typcache.h
!
/* static function decls */
static Datum ExecEvalArrayRef(ArrayRefExprState *astate,
--- 53,60
#include utils/lsyscache.h
#include utils/memutils.h
#include utils/typcache.h
! #include lib/stringinfo.h
! #include string.h
/* static function decls */
static Datum ExecEvalArrayRef(ArrayRefExprState *astate,
***
*** 111,116
--- 112,119
static Datum ExecEvalNullIf(FuncExprState *nullIfExpr,
ExprContext *econtext,
bool *isNull, ExprDoneCond *isDone);
+ static Datum ExecEvalXml(XmlExprState *xmlExpr, ExprContext *econtext,
+ bool *isNull, ExprDoneCond *isDone);
static Datum ExecEvalNullTest(GenericExprState *nstate,
ExprContext *econtext,
bool *isNull, ExprDoneCond *isDone);
***
*** 218,223
--- 221,228
* have to know the difference (that's what they need refattrlength for).
*--
*/
+
+
static Datum
ExecEvalArrayRef(ArrayRefExprState *astate,
ExprContext *econtext,
***
*** 2304,2312
}
}
! return result;
}
/*
* ExecEvalNullIf
*
--- 2309,2498
}
}
! return result;
! }
!
! /*
! * ExecEvalXml
! *
! */
!
! static char *
! getXmlParam(XmlParams *params, XmlParamOp op)
! {
! if (params == NULL)
! return NULL;
! switch (op)
! {
! case IS_XMLNAME:
! return params-name;
! case IS_XMLVERSION:
! return params-version;
! case IS_XMLENCODING:
! return params-encoding;
! case IS_XMLSTANDALONE:
! return params-standalone;
! }
! return NULL;
! }
!
! static Datum ExecEvalXml(XmlExprState *xmlExpr, ExprContext *econtext,
! bool *isNull, ExprDoneCond *isDone)
! {
! StringInfoData buf;
! bool isnull;
! ListCell *arg;
!
! initStringInfo(buf);
!
! *isNull = false;
!
! if (isDone)
! *isDone = ExprSingleResult;
!
!
! switch (xmlExpr-op)
! {
! case IS_XMLCONCAT:
! {
! *isNull = true;
! foreach(arg, xmlExpr-xml_args)
! {
! ExprState *e = (ExprState *) lfirst(arg);
! Datum value = ExecEvalExpr(e, econtext, isnull, NULL);
! if (!isnull)
! {
! appendStringInfo(buf, %s, (char *)value);
! *isNull = false;
! }
! }
! }
! break;
! case IS_XMLELEMENT:
! {
! appendStringInfo(buf, %s, getXmlParam(xmlExpr-params, IS_XMLNAME));
! int state = 0;
! int i = 0;
! foreach(arg, xmlExpr-nargs)
! {
! GenericExprState *gstate = (GenericExprState *) lfirst(arg);
! Datum value = ExecEvalExpr(gstate-arg, econtext, isnull, NULL);
! if (!isnull)
! {
! char *outstr = DatumGetCString(OidFunctionCall1(xmlExpr-nargs_tcache[i], value));
! appendStringInfo(buf, %s=\%s\, xmlExpr-nargs_ncache[i], outstr);
! pfree(outstr);
! }
! i += 1;
! }
! foreach(arg, xmlExpr-xml_args)
! {
! ExprState *e = (ExprState *) lfirst(arg);
! Datum value = ExecEvalExpr(e, econtext, isnull, NULL);
! if (!isnull)
! {
! if (state == 0)
! {
! appendStringInfo(buf, );
! state = 1;
! }
! appendStringInfo(buf, %s, (char *)value);
! }
! }
! if (xmlExpr-args)
! {
! ExprState *expr = linitial(xmlExpr-args);
! Datum value = ExecEvalExpr(expr, econtext, isnull, NULL);
!
! if (!isnull)
! {
! char *outstr = DatumGetCString(OidFunctionCall1(xmlExpr-arg_typeId, value));
! if (state == 0)
! {
! appendStringInfoChar(buf, '');
! state = 1;
! }
!
Re: [PATCHES] [INTERFACES] [HACKERS] ECPG ignores SAVEPOINT if first statement of a transaction
On Sat, Aug 13, 2005 at 09:01:45AM -0600, Michael Fuhr wrote: If no statements have been executed in this transaction then committed is true, so ECPGtrans() ignores the current statement. It looks like the code should make an extra consideration for SAVEPOINT in case it's the first statement. I rearranged the whole code as it seemed to have more holes. This code really was old. :-) So could you please check with the lates CVS, either HEAD or 8.0? I'm also wondering if the check for a NULL return value from PQexec() is sufficient. Shouldn't it also check for a non-NULL result that's anything other than PGRES_COMMAND_OK? Yes, it should. I changed this as well. Michael -- Michael Meskes Email: Michael at Fam-Meskes dot De, Michael at Meskes dot (De|Com|Net|Org) ICQ: 179140304, AIM/Yahoo: michaelmeskes, Jabber: [EMAIL PROTECTED] Go SF 49ers! Go Rhein Fire! Use Debian GNU/Linux! Use PostgreSQL! ---(end of broadcast)--- TIP 3: Have you checked our extensive FAQ? http://www.postgresql.org/docs/faq
Re: [PATCHES] SQL/XML publishing function experimental patch
Pavel Stehule [EMAIL PROTECTED] writes: This patch contains SQL/XML public function XMLFOREST, XMLELEMENT, ... based on CString type. This patch hasn't high quality, but can be usefull for testing funcionality, and maybe for some people. 1. Why did you base the datatype on cstring? That's inappropriate for anything except guaranteed-short strings, because it's not toastable. 2. Is it really necessary to hack up the grammar for this? Seems like you should create ordinary functions to manipulate the datatype(s). If every datatype thought it could do this sort of thing, we'd have a completely unmanageable grammar. regards, tom lane ---(end of broadcast)--- TIP 3: Have you checked our extensive FAQ? http://www.postgresql.org/docs/faq
Re: [PATCHES] SQL/XML publishing function experimental patch
Am Montag, 12. September 2005 16:05 schrieb Tom Lane: 2. Is it really necessary to hack up the grammar for this? Yes, because the syntax defined in the SQL standard is completely weird. -- Peter Eisentraut http://developer.postgresql.org/~petere/ ---(end of broadcast)--- TIP 4: Have you searched our list archives? http://archives.postgresql.org
[PATCHES] Autovacuum docs
Hackers, I have prepared some docs for autovacuum. I attach the patch. Comments on grammar, style etc. are welcome. I will not apply it yet because it conflicts with Neil's runtime-config restructuring, so I'll wait until he applies his patch. -- Alvaro Herrera -- Valdivia, Chile Architect, www.EnterpriseDB.com www.google.com: interfaz de línea de comando para la web. Index: catalogs.sgml === RCS file: /home/alvherre/cvs/pgsql/doc/src/sgml/catalogs.sgml,v retrieving revision 2.113 diff -c -r2.113 catalogs.sgml *** catalogs.sgml 8 Sep 2005 20:07:41 - 2.113 --- catalogs.sgml 12 Sep 2005 13:56:35 - *** *** 1137,1145 primarypg_autovacuum/primary /indexterm para The catalog structnamepg_autovacuum/structname stores optional !per-relation configuration parameters for quoteautovacuum/. If there is an entry here for a particular relation, the given parameters will be used for autovacuuming that table. If no entry is present, the system-wide defaults will be used. --- 1137,1150 primarypg_autovacuum/primary /indexterm + indexterm zone=catalog-pg-autovacuum +primaryautovacuum/primary +secondarytable-specific configuration/secondary + /indexterm + para The catalog structnamepg_autovacuum/structname stores optional !per-relation configuration parameters for xref linkend=autovacuum endterm=autovacuum-title. If there is an entry here for a particular relation, the given parameters will be used for autovacuuming that table. If no entry is present, the system-wide defaults will be used. Index: maintenance.sgml === RCS file: /home/alvherre/cvs/pgsql/doc/src/sgml/maintenance.sgml,v retrieving revision 1.45 diff -c -r1.45 maintenance.sgml *** maintenance.sgml22 Jun 2005 21:14:28 - 1.45 --- maintenance.sgml12 Sep 2005 14:10:17 - *** *** 99,104 --- 99,110 xref linkend=runtime-config-resource-vacuum-cost. /para + para +An automated mechanism for performing the necessary commandVACUUM/ +operations has been added in productnamePostgreSQL/productname 8.1. +See xref linkend=autovacuum endterm=autovacuum-title. + /para + sect2 id=vacuum-for-space-recovery titleRecovering disk space/title *** *** 177,189 the program filenamevacuumdb/ may be helpful. /para -tip - para - The filenamecontrib/pg_autovacuum/ program can be useful for - automating high-frequency vacuuming operations. - /para -/tip - para commandVACUUM FULL/ is recommended for cases where you know you have deleted the majority of rows in a table, so that the --- 183,188 *** *** 456,461 --- 455,551 /para /warning /sect2 + + sect2 id=autovacuum +title id=autovacuum-titleThe Auto-Vacuum Daemon/title + +indexterm + primaryautovacuum/primary + secondarygeneral information/secondary +/indexterm +para + Beginning in productnamePostgreSQL /productname 8.1, there is a + separate optional server process called the firsttermautovacuum + daemon/firstterm, whose purpose is to automate the issuance of + commandVACUUM/command and commandANALYZE /command commands. + When enabled, the autovacuum daemon runs periodically and checks for + tables that have had a large number of inserted, updated or deleted + tuples. These checks use the row-level statistics collection facility; + therefore, the autovacuum daemon cannot be used unless xref + linkend=guc-stats-start-collector and xref + linkend=guc-stats-row-level are set literaltrue/literal. Also, it's + important to allow a slot for the autovacuum process when choosing the + value of xref linkend=guc-superuser-reserved-connections. +/para + +para + The autovacuum daemon, when enabled, runs every xref + linkend=guc-autovacuum-naptime seconds and determines which database + to process. Any database which is close to transaction ID wraparound + is immediately processed. In this case, autovacuum issues a + database-wide commandVACUUM/command call, or commandVACUUM + FREEZE/command if it's a template database, and then terminates. If + no database fulfills this criterion, the one that was least recently + processed by autovacuum itself is chosen. In this mode, each table in + the database is checked for new and obsolete tuples, according to the + applicable autovacuum parameters. If a link linkend=catalog-pg-autovacuum + structnamepg_autovacuum/structname/link tuple is found for this + table, these settings are applied; otherwise the global values in + filenamepostgresql.conf/filename are used. See xref
Re: [PATCHES] Autovacuum docs
Alvaro Herrera wrote: I have prepared some docs for autovacuum. I attach the patch. Comments on grammar, style etc. are welcome. Looks very good to me. -- Peter Eisentraut http://developer.postgresql.org/~petere/ ---(end of broadcast)--- TIP 9: In versions below 8.0, the planner will ignore your desire to choose an index scan if your joining column's datatypes do not match
Re: [PATCHES] SQL/XML publishing function experimental patch
1. Why did you base the datatype on cstring? That's inappropriate for anything except guaranteed-short strings, because it's not toastable. It's only my laziness. I can adapt patch for TEXT. It's only temporary solution. The goal is true xml type. 2. Is it really necessary to hack up the grammar for this? Seems like you should create ordinary functions to manipulate the datatype(s). If every datatype thought it could do this sort of thing, we'd have a completely unmanageable grammar. regards, tom lane I prefere any general solution. But there isn't. I hope SQL/XML syntax is exception and the solution SQL/XML function is relative clean. There was ideas about pretranslators from different syntaxs. Is this idea dead? regards Pavel Stehule _ Emotikony a pozadi programu MSN Messenger ozivi vasi konverzaci. http://messenger.msn.cz/ ---(end of broadcast)--- TIP 2: Don't 'kill -9' the postmaster
Re: [PATCHES] docs: config chapter
Peter Eisentraut wrote: Moreover, I don't agree with the premise. This patch is blocking other work. If you still object to it, can you elaborate why? Otherwise I'll apply it this evening (EST). -Neil ---(end of broadcast)--- TIP 3: Have you checked our extensive FAQ? http://www.postgresql.org/docs/faq
Re: [PATCHES] pg_dump feature patch to allow selection of rows to
Seth Robertson wrote: In order to support some automation, I needed the ability to pull out subsets of postgres tables in a format which could be loaded into other postgres databases. One alternative would be using CREATE TABLE AS to create a separate table containing the desired subset, and then using pg_dump to dump that table. Instead of writing a utility to replicate the functionality in pg_dump, I created a new flag, --where, which allows you to postpend where (or order or limit or whatever) sql fragments at the end of the select done by pg_dump. Neat, but IMHO too much of a kludge for mainline. -Neil ---(end of broadcast)--- TIP 9: In versions below 8.0, the planner will ignore your desire to choose an index scan if your joining column's datatypes do not match
Re: [PATCHES] SPI_getbinval() documentation correction
Michael Fuhr wrote: The synopsis for SPI_getbinval() shows the third argument being colnumber but in the Arguments section it appears as rownumber. This patch corrects that. Thanks for the patch -- applied to HEAD, 8.0, and 7.4 -Neil ---(end of broadcast)--- TIP 5: don't forget to increase your free space map settings
Re: [PATCHES] pg_dump feature patch to allow selection of rows to
Neil Conway [EMAIL PROTECTED] writes: Seth Robertson wrote: Instead of writing a utility to replicate the functionality in pg_dump, I created a new flag, --where, which allows you to postpend where (or order or limit or whatever) sql fragments at the end of the select done by pg_dump. Neat, but IMHO too much of a kludge for mainline. Yeah, I felt the same --- the fact that it doesn't work for the normal COPY case renders it a bit too ugly. regards, tom lane ---(end of broadcast)--- TIP 9: In versions below 8.0, the planner will ignore your desire to choose an index scan if your joining column's datatypes do not match
Re: [PATCHES] Autovacuum docs
Alvaro Herrera [EMAIL PROTECTED] writes: I have prepared some docs for autovacuum. I attach the patch. Comments on grammar, style etc. are welcome. A few piddling suggestions: +See xref linkend=autovacuum endterm=autovacuum-title. I think we mostly don't use endterm on cross-references to sections. Not sure if that's a project standard, or just laziness, though. (Peter?) + sect2 id=autovacuum +title id=autovacuum-titleThe Auto-Vacuum Daemon/title The other sect2 titles in this chapter capitalize only the first word. Again, I'm not sure how fully that convention is followed, but I think this would look better as The auto-vacuum daemon. + for more details on the global settings. Note that the autovacuum + daemon can be instructed to skip particular tables, by setting the + structnamepg_autovacuum/.structfieldenabled/ field to literalfalse/literal. +/para This note seems a bit awkwardly placed. I'd drop it and instead add enabled to the list of additional pg_autovacuum parameters here: + Additionally to the base threshold values and scale factors, there are + two parameters that can be set for each table in structnamepg_autovacuum/structname: Additionally to is not good style ... maybe Besides? variablelist *** *** 3470,3477 commandVACUUM/ operations. If -1 is specified (which is the default), the regular xref linkend=guc-vacuum-cost-delay value will be used. - This setting can be overridden for individual tables by entries in - structnamepg_autovacuum/. /para /listitem /varlistentry --- 3467,3472 *** *** 3487,3494 commandVACUUM/ operations. If -1 is specified (which is the default), the regular xref linkend=guc-vacuum-cost-limit value will be used. - This setting can be overridden for individual tables by entries in - structnamepg_autovacuum/. /para /listitem /varlistentry --- 3482,3487 I don't think you should remove those sentences. The information about GUC variables is intentionally somewhat redundant with discussions elsewhere, and in this case it's important as a cross-reference. Otherwise looks great. regards, tom lane ---(end of broadcast)--- TIP 4: Have you searched our list archives? http://archives.postgresql.org
Re: [PATCHES] docs: config chapter
Neil Conway wrote: This patch is blocking other work [...] Otherwise I'll apply it this evening (EST). Applied. -Neil ---(end of broadcast)--- TIP 2: Don't 'kill -9' the postmaster
Re: [PATCHES] Autovacuum docs
Tom Lane wrote: + Additionally to the base threshold values and scale factors, there are + two parameters that can be set for each table in structnamepg_autovacuum/structname: Additionally to is not good style ... maybe Besides? I think In addition to was what Alvaro meant, which works. I just gave the patch a quick eye-balling and it looks good to me. Alvaro, thanks for doing this, I know I said I would get to it, but I just don't seem to be able to find time for Postgres work these days. If fact I can barely find enough time to keep up with the lists these days. Anyway, it looks good and thanks! Matt ---(end of broadcast)--- TIP 5: don't forget to increase your free space map settings
Re: [PATCHES] Autovacuum docs
On Mon, Sep 12, 2005 at 04:17:29PM -0400, Tom Lane wrote: Alvaro Herrera [EMAIL PROTECTED] writes: I have prepared some docs for autovacuum. I attach the patch. Comments on grammar, style etc. are welcome. Applied, updated per your comments. +See xref linkend=autovacuum endterm=autovacuum-title. I think we mostly don't use endterm on cross-references to sections. Not sure if that's a project standard, or just laziness, though. (Peter?) I removed it, but personally I think it's better to put it where it makes the text read naturally. For example, that text reads See _The auto-vacuum daemon_, instead of See _Section 21.1.4_ -- I think the former is better. There are instances where it doesn't read so good, for example in the runtime configuration chapter, the text is These settings control the default behavior for the autovacuum daemon. Please refer to _Section 22.1.4_ for more information. I don't see how to cleanly fit The auto-vacuum daemon in that sentence in a way that looks like a reference. So, in short, if there is a policy on this, I propose to update it to use endterm wherever the surrounding text allows it. Please check the new version. Bruce's site is already up to date: http://candle.pha.pa.us/main/writings/pgsql/sgml/maintenance.html#AUTOVACUUM -- Alvaro Herrera -- Valdivia, Chile Architect, www.EnterpriseDB.com One man's impedance mismatch is another man's layer of abstraction. (Lincoln Yeoh) ---(end of broadcast)--- TIP 2: Don't 'kill -9' the postmaster
