Re: [BUGS] BUG #5534: IS DOCUMENT predicate errors instead of returning false
Quoting Mike Fowler :
The following bug has been logged online:
Bug reference: 5534
Logged by: Mike Fowler
Email address: m...@mlfowler.com
PostgreSQL version: 9.0beta2
Operating system: Linux 2.6.31-14-generic #48-Ubuntu SMP
Description:IS DOCUMENT predicate errors instead of returning false
Details:
IS DOCUMENT should return false for a non-well formed document, and indeed
is coded to do such. However, the conversion to the xml type which happens
before the underlying xml_is_document function is even called fails and
exceptions out. I've mentioned this on -hackers with message ID
20100701172553.w5vdy1xbocos8...@www.mlfowler.com
--
Sent via pgsql-bugs mailing list (pgsql-bugs@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-bugs
The attached patch is a very small patch that changes parse_expr.c to
not convert everything to xml. This now means that when passed
malformed XML it will return false instead of throwing an exception.
In my mind this acceptable as I don't see anywhere in the standard
that mandates that:
xmlval IS NOT DOCUMENT == xmlval IS CONTENT
Regards,
--
Mike Fowler
Registered Linux user: 379787
*** a/src/backend/parser/parse_expr.c
--- b/src/backend/parser/parse_expr.c
***
*** 1950,1956 transformXmlExpr(ParseState *pstate, XmlExpr *x)
Assert(false);
break;
case IS_DOCUMENT:
! newe = coerce_to_specific_type(pstate, newe, XMLOID,
"IS DOCUMENT");
break;
}
--- 1950,1956
Assert(false);
break;
case IS_DOCUMENT:
! newe = coerce_to_specific_type(pstate, newe, TEXTOID,
"IS DOCUMENT");
break;
}
*** a/src/backend/utils/adt/xml.c
--- b/src/backend/utils/adt/xml.c
***
*** 795,801 xmlvalidate(PG_FUNCTION_ARGS)
bool
! xml_is_document(xmltype *arg)
{
#ifdef USE_LIBXML
bool result;
--- 795,801
bool
! xml_is_document(text *arg)
{
#ifdef USE_LIBXML
bool result;
***
*** 805,811 xml_is_document(xmltype *arg)
/* We want to catch ereport(INVALID_XML_DOCUMENT) and return false */
PG_TRY();
{
! doc = xml_parse((text *) arg, XMLOPTION_DOCUMENT, true,
GetDatabaseEncoding());
result = true;
}
--- 805,811
/* We want to catch ereport(INVALID_XML_DOCUMENT) and return false */
PG_TRY();
{
! doc = xml_parse(arg, XMLOPTION_DOCUMENT, true,
GetDatabaseEncoding());
result = true;
}
*** a/src/include/utils/xml.h
--- b/src/include/utils/xml.h
***
*** 70,76 extern xmltype *xmlelement(XmlExprState *xmlExpr, ExprContext *econtext);
extern xmltype *xmlparse(text *data, XmlOptionType xmloption, bool preserve_whitespace);
extern xmltype *xmlpi(char *target, text *arg, bool arg_is_null, bool *result_is_null);
extern xmltype *xmlroot(xmltype *data, text *version, int standalone);
! extern bool xml_is_document(xmltype *arg);
extern text *xmltotext_with_xmloption(xmltype *data, XmlOptionType xmloption_arg);
extern char *escape_xml(const char *str);
--- 70,76
extern xmltype *xmlparse(text *data, XmlOptionType xmloption, bool preserve_whitespace);
extern xmltype *xmlpi(char *target, text *arg, bool arg_is_null, bool *result_is_null);
extern xmltype *xmlroot(xmltype *data, text *version, int standalone);
! extern bool xml_is_document(text *arg);
extern text *xmltotext_with_xmloption(xmltype *data, XmlOptionType xmloption_arg);
extern char *escape_xml(const char *str);
*** a/src/test/regress/expected/xml.out
--- b/src/test/regress/expected/xml.out
***
*** 357,362 SELECT xml 'bar' IS DOCUMENT;
--- 357,378
t
(1 row)
+ SELECT xml 'barbar'
+ barbarbarbarfoo' IS DOCUMENT;
?column?
--
***
*** 376,387 SELECT xml 'abc' IS NOT DOCUMENT;
(1 row)
SELECT '<>' IS NOT DOCUMENT;
ERROR: invalid XML content
! LINE 1: SELECT '<>' IS NOT DOCUMENT;
!^
! DETAIL: Entity: line 1: parser error : StartTag: invalid element name
! <>
! ^
SELECT xmlagg(data) FROM xmltest;
xmlagg
--
--- 392,418
(1 row)
SELECT '<>' IS NOT DOCUMENT;
+ ?column?
+ --
+ t
+ (1 row)
+
+ SELECT xml 'barbar'
! barbarbarbar' IS DOCUMENT;
+ SELECT xml 'barbarbarfoo' IS DOCUMENT;
SELECT xml '' IS NOT DOCUMENT;
SELECT xml 'abc' IS NOT DOCUMENT;
SELECT '<>' IS NOT DOCUMENT;
+ SELECT xml 'barbar
--
Sent via pgsql-bugs mailing list (pgsql-bugs@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-bugs
Re: [BUGS] BUG #5533: PQexecParams in Binary Mode returns incorrect value for float4
Well don't I feel silly! Tom, thank-you for the help - I'll try and get my code debugged before bothering you again On 2/07/10 12:51 AM, "Tom Lane" wrote: > "" writes: >> Some needless speculation: >> 0.75 = 3F40 (as IEEE 32-bit float) >> 0.75 = 3FE8 (as IEEE 64-bit float) >> 1.812500 = 3FE8 (as returned by PQexecParams) >> i.e. the return value is the first 32-bits of the 64-bit representation of >> the correct value. > > Well, yeah. You declared the column as float8, not float4: > >> res = PQexec(conn, "CREATE TABLE testtbl( Intgr int4, Flt float8, PRIMARY > KEY ( Intgr ));"); > > regards, tom lane -- Sent via pgsql-bugs mailing list (pgsql-bugs@postgresql.org) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-bugs
[BUGS] BUG #5536: Disputing output for some Geometric types
The following bug has been logged online: Bug reference: 5536 Logged by: Jon Strait Email address: jstr...@moonloop.net PostgreSQL version: 8.4.4 Operating system: Linux Description:Disputing output for some Geometric types Details: Refering to Documentation section 8.8 (Geometric Types), Boxes are supposed to be output with the format of the first syntax listed, ( (x1, y1), (x2, y2) ), but the output is showing the format of the second syntax, (x1, y1), (x2, y2). Line Segments are being output in a format that is not mentioned in the documentation for Line Segments, [ (x1, y1), (x2, y2) ]. -- Sent via pgsql-bugs mailing list (pgsql-bugs@postgresql.org) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-bugs
Re: [BUGS] BUG #5534: IS DOCUMENT predicate errors instead of returning false
On fre, 2010-07-02 at 14:01 +0100, Mike Fowler wrote: > > The attached patch is a very small patch that changes parse_expr.c > to > not convert everything to xml. This now means that when passed > malformed XML it will return false instead of throwing an exception. > In my mind this acceptable as I don't see anywhere in the standard > that mandates that: > > xmlval IS NOT DOCUMENT == xmlval IS CONTENT It says that the declared type of the argument of IS DOCUMENT is XML, which would be invalidated by your patch. -- Sent via pgsql-bugs mailing list (pgsql-bugs@postgresql.org) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-bugs
[BUGS] BUG #5537: Makefile.darwin broken
The following bug has been logged online: Bug reference: 5537 Logged by: Peter Abrahamsen Email address: rainh...@gmail.com PostgreSQL version: 9.0beta2 Operating system: Darwin x86_64 (MacOS 10.6.4) Description:Makefile.darwin broken Details: Makefile.darwin references src/backend/postgres, which doesn't exist. -- Sent via pgsql-bugs mailing list (pgsql-bugs@postgresql.org) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-bugs
Re: [BUGS] BUG #5537: Makefile.darwin broken
On Fri, Jul 2, 2010 at 6:15 PM, Peter Abrahamsen wrote: > The following bug has been logged online: > > Bug reference: 5537 > Logged by: Peter Abrahamsen > Email address: rainh...@gmail.com > PostgreSQL version: 9.0beta2 > Operating system: Darwin x86_64 (MacOS 10.6.4) > Description: Makefile.darwin broken > Details: > > Makefile.darwin references src/backend/postgres, which doesn't exist. So what, specifically, is not working for you? I'm using MacOS X 10.6.4 x86_64 also, and I can build Postgres without a problem. -- Robert Haas EnterpriseDB: http://www.enterprisedb.com The Enterprise Postgres Company -- Sent via pgsql-bugs mailing list (pgsql-bugs@postgresql.org) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-bugs
Re: [BUGS] BUG #5536: Disputing output for some Geometric types
"Jon Strait" writes: > Refering to Documentation section 8.8 (Geometric Types), Boxes are supposed > to be output with the format of the first syntax listed, ( (x1, y1), (x2, > y2) ), but the output is showing the format of the second syntax, (x1, y1), > (x2, y2). Line Segments are being output in a format that is not mentioned > in the documentation for Line Segments, [ (x1, y1), (x2, y2) ]. Yeah, this documentation seems to be a bit off from the actual behavior of the code. I've adjusted the docs. regards, tom lane -- Sent via pgsql-bugs mailing list (pgsql-bugs@postgresql.org) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-bugs
