unknownin/out patch (was [HACKERS] PQescapeBytea is not multibyte aware)

Tom Lane wrote:
> Joe Conway  writes:
> 
>>I think you're correct that in a client/database encoding mismatch 
>>scenario, there would be bigger problems. Thoughts on this?
> 
> 
> This scenario is probably why Tatsuo wants PQescapeBytea to octalize
> everything with the high bit set; I'm not sure there's any lesser way
> out.  Nonetheless, if UNKNOWN conversion introduces additional failures
> then it makes sense to fix that.
> 
> 			regards, tom lane
> 

Here's a patch to add unknownin/unknownout support. I also poked around 
looking for places that assume UNKNOWN == TEXT. One of those was the 
"SET" type in pg_type.h, which was using textin/textout. This one I took 
care of in this patch. The other suspicious place was in 
string_to_dataum (which is defined in both selfuncs.c and indxpath.c). I 
wasn't too sure about those, so I left them be.

Regression tests all pass with the exception of horology, which also 
fails on CVS tip. It looks like that is a daylight savings time issue 
though.

Also as a side note, I can't get make check to get past initdb if I 
configure with --enable-multibyte on CVS tip. Is there a known problem 
or am I just being clueless . . .wait, let's qualify that -- am I being 
clueless on this one issue? ;-)

Joe
diff -Ncr pgsql.orig/src/backend/utils/adt/varlena.c pgsql/src/backend/utils/adt/varlena.c
*** pgsql.orig/src/backend/utils/adt/varlena.c	Sun Apr  7 10:21:25 2002
--- pgsql/src/backend/utils/adt/varlena.c	Sun Apr  7 11:44:54 2002
***************
*** 228,233 ****
--- 228,273 ----
  }
  
  
+ /*
+  *		unknownin			- converts "..." to internal representation
+  */
+ Datum
+ unknownin(PG_FUNCTION_ARGS)
+ {
+ 	char	   *inputStr = PG_GETARG_CSTRING(0);
+ 	unknown	   *result;
+ 	int			len;
+ 
+ 	len = strlen(inputStr) + VARHDRSZ;
+ 
+ 	result = (unknown *) palloc(len);
+ 	VARATT_SIZEP(result) = len;
+ 
+ 	memcpy(VARDATA(result), inputStr, len - VARHDRSZ);
+ 
+ 	PG_RETURN_UNKNOWN_P(result);
+ }
+ 
+ 
+ /*
+  *		unknownout			- converts internal representation to "..."
+  */
+ Datum
+ unknownout(PG_FUNCTION_ARGS)
+ {
+ 	unknown	   *t = PG_GETARG_UNKNOWN_P(0);
+ 	int			len;
+ 	char	   *result;
+ 
+ 	len = VARSIZE(t) - VARHDRSZ;
+ 	result = (char *) palloc(len + 1);
+ 	memcpy(result, VARDATA(t), len);
+ 	result[len] = '\0';
+ 
+ 	PG_RETURN_CSTRING(result);
+ }
+ 
+ 
  /* ========== PUBLIC ROUTINES ========== */
  
  /*
diff -Ncr pgsql.orig/src/include/c.h pgsql/src/include/c.h
*** pgsql.orig/src/include/c.h	Sun Apr  7 10:21:29 2002
--- pgsql/src/include/c.h	Sun Apr  7 11:40:59 2002
***************
*** 389,394 ****
--- 389,395 ----
   */
  typedef struct varlena bytea;
  typedef struct varlena text;
+ typedef struct varlena unknown;
  typedef struct varlena BpChar;	/* blank-padded char, ie SQL char(n) */
  typedef struct varlena VarChar; /* var-length char, ie SQL varchar(n) */
  
diff -Ncr pgsql.orig/src/include/catalog/pg_proc.h pgsql/src/include/catalog/pg_proc.h
*** pgsql.orig/src/include/catalog/pg_proc.h	Sun Apr  7 10:21:29 2002
--- pgsql/src/include/catalog/pg_proc.h	Sun Apr  7 11:56:09 2002
***************
*** 235,240 ****
--- 235,245 ----
  DATA(insert OID = 108 (  scalargtjoinsel   PGNSP PGUID 12 f t t s 3 f 701 "0 26 0" 100 0 0 100  scalargtjoinsel - _null_ ));
  DESCR("join selectivity of > and related operators on scalar datatypes");
  
+ DATA(insert OID =  109 (  unknownin			   PGNSP PGUID 12 f t t i 1 f 705 "0" 100 0 0 100	unknownin - _null_ ));
+ DESCR("(internal)");
+ DATA(insert OID =  110 (  unknownout		   PGNSP PGUID 12 f t t i 1 f 23  "0" 100 0 0 100	unknownout - _null_ ));
+ DESCR("(internal)");
+ 
  DATA(insert OID = 112 (  text			   PGNSP PGUID 12 f t t i 1 f  25 "23" 100 0 0 100  int4_text - _null_ ));
  DESCR("convert int4 to text");
  DATA(insert OID = 113 (  text			   PGNSP PGUID 12 f t t i 1 f  25 "21" 100 0 0 100  int2_text - _null_ ));
diff -Ncr pgsql.orig/src/include/catalog/pg_type.h pgsql/src/include/catalog/pg_type.h
*** pgsql.orig/src/include/catalog/pg_type.h	Sun Apr  7 10:21:29 2002
--- pgsql/src/include/catalog/pg_type.h	Sun Apr  7 11:57:36 2002
***************
*** 302,308 ****
  DESCR("array of INDEX_MAX_KEYS oids, used in system tables");
  #define OIDVECTOROID	30
  
! DATA(insert OID = 32 (	SET		   PGNSP PGUID -1  -1 f b t \054 0   0 textin textout textin textout i p f 0 -1 0 _null_ _null_ ));
  DESCR("set of tuples");
  
  DATA(insert OID = 71 (	pg_type		 PGNSP PGUID 4 4 t c t \054 1247 0 int4in int4out int4in int4out i p f 0 -1 0 _null_ _null_ ));
--- 302,308 ----
  DESCR("array of INDEX_MAX_KEYS oids, used in system tables");
  #define OIDVECTOROID	30
  
! DATA(insert OID = 32 (	SET		   PGNSP PGUID -1  -1 f b t \054 0   0 unknownin unknownout unknownin unknownout i p f 0 -1 0 _null_ _null_ ));
  DESCR("set of tuples");
  
  DATA(insert OID = 71 (	pg_type		 PGNSP PGUID 4 4 t c t \054 1247 0 int4in int4out int4in int4out i p f 0 -1 0 _null_ _null_ ));
***************
*** 366,372 ****
  DATA(insert OID = 704 (  tinterval PGNSP PGUID 12  47 f b t \054 0   0 tintervalin tintervalout tintervalin tintervalout i p f 0 -1 0 _null_ _null_ ));
  DESCR("(abstime,abstime), time interval");
  #define TINTERVALOID	704
! DATA(insert OID = 705 (  unknown   PGNSP PGUID -1  -1 f b t \054 0   0 textin textout textin textout i p f 0 -1 0 _null_ _null_ ));
  DESCR("");
  #define UNKNOWNOID		705
  
--- 366,372 ----
  DATA(insert OID = 704 (  tinterval PGNSP PGUID 12  47 f b t \054 0   0 tintervalin tintervalout tintervalin tintervalout i p f 0 -1 0 _null_ _null_ ));
  DESCR("(abstime,abstime), time interval");
  #define TINTERVALOID	704
! DATA(insert OID = 705 (  unknown   PGNSP PGUID -1  -1 f b t \054 0   0 unknownin unknownout unknownin unknownout i p f 0 -1 0 _null_ _null_ ));
  DESCR("");
  #define UNKNOWNOID		705
  
diff -Ncr pgsql.orig/src/include/fmgr.h pgsql/src/include/fmgr.h
*** pgsql.orig/src/include/fmgr.h	Sun Apr  7 10:21:29 2002
--- pgsql/src/include/fmgr.h	Sun Apr  7 12:11:30 2002
***************
*** 185,190 ****
--- 185,191 ----
  /* DatumGetFoo macros for varlena types will typically look like this: */
  #define DatumGetByteaP(X)			((bytea *) PG_DETOAST_DATUM(X))
  #define DatumGetTextP(X)			((text *) PG_DETOAST_DATUM(X))
+ #define DatumGetUnknownP(X)			((unknown *) PG_DETOAST_DATUM(X))
  #define DatumGetBpCharP(X)			((BpChar *) PG_DETOAST_DATUM(X))
  #define DatumGetVarCharP(X)			((VarChar *) PG_DETOAST_DATUM(X))
  /* And we also offer variants that return an OK-to-write copy */
***************
*** 200,205 ****
--- 201,207 ----
  /* GETARG macros for varlena types will typically look like this: */
  #define PG_GETARG_BYTEA_P(n)		DatumGetByteaP(PG_GETARG_DATUM(n))
  #define PG_GETARG_TEXT_P(n)			DatumGetTextP(PG_GETARG_DATUM(n))
+ #define PG_GETARG_UNKNOWN_P(n)		DatumGetUnknownP(PG_GETARG_DATUM(n))
  #define PG_GETARG_BPCHAR_P(n)		DatumGetBpCharP(PG_GETARG_DATUM(n))
  #define PG_GETARG_VARCHAR_P(n)		DatumGetVarCharP(PG_GETARG_DATUM(n))
  /* And we also offer variants that return an OK-to-write copy */
***************
*** 239,244 ****
--- 241,247 ----
  /* RETURN macros for other pass-by-ref types will typically look like this: */
  #define PG_RETURN_BYTEA_P(x)   PG_RETURN_POINTER(x)
  #define PG_RETURN_TEXT_P(x)    PG_RETURN_POINTER(x)
+ #define PG_RETURN_UNKNOWN_P(x) PG_RETURN_POINTER(x)
  #define PG_RETURN_BPCHAR_P(x)  PG_RETURN_POINTER(x)
  #define PG_RETURN_VARCHAR_P(x) PG_RETURN_POINTER(x)
  
diff -Ncr pgsql.orig/src/include/utils/builtins.h pgsql/src/include/utils/builtins.h
*** pgsql.orig/src/include/utils/builtins.h	Sun Apr  7 10:21:29 2002
--- pgsql/src/include/utils/builtins.h	Sun Apr  7 12:26:17 2002
***************
*** 414,419 ****
--- 414,422 ----
  extern bool SplitIdentifierString(char *rawstring, char separator,
  								  List **namelist);
  
+ extern Datum unknownin(PG_FUNCTION_ARGS);
+ extern Datum unknownout(PG_FUNCTION_ARGS);
+ 
  extern Datum byteain(PG_FUNCTION_ARGS);
  extern Datum byteaout(PG_FUNCTION_ARGS);
  extern Datum byteaoctetlen(PG_FUNCTION_ARGS);