Extend sortsupport for text to more opclasses.

Have varlena.c expose an interface that allows the char(n), bytea, and
bpchar types to piggyback on a now-generalized SortSupport for text.
This pushes a little more knowledge of the bpchar/char(n) type into
varlena.c than might be preferred, but that seems like the approach
that creates least friction.  Also speed things up for index builds
that use text_pattern_ops or varchar_pattern_ops.

This patch does quite a bit of renaming, but it seems likely to be
worth it, so as to avoid future confusion about the fact that this code
is now more generally used than the old names might have suggested.

Peter Geoghegan, reviewed by Álvaro Herrera and Andreas Karlsson,
with small tweaks by me.

doc/src/sgml/datatype.sgml

@@ -1140,8 +1140,7 @@ SELECT '52093.89'::money::numeric::float8;
      advantages in some other database systems, there is no such advantage in
      <productname>PostgreSQL</productname>; in fact
      <type>character(<replaceable>n</>)</type> is usually the slowest of
-     the three because of its additional storage costs and slower
-     sorting.  In most situations
+     the three because of its additional storage costs.  In most situations
      <type>text</type> or <type>character varying</type> should be used
      instead.
     </para>

src/backend/utils/adt/varchar.c

@@ -17,6 +17,7 @@
 
 #include "access/hash.h"
 #include "access/tuptoaster.h"
+#include "catalog/pg_collation.h"
 #include "libpq/pqformat.h"
 #include "nodes/nodeFuncs.h"
 #include "utils/array.h"
@@ -649,14 +650,21 @@ varchartypmodout(PG_FUNCTION_ARGS)
  *****************************************************************************/
 
 /* "True" length (not counting trailing blanks) of a BpChar */
-static int
+static inline int
 bcTruelen(BpChar *arg)
 {
-	char	   *s = VARDATA_ANY(arg);
+	return bpchartruelen(VARDATA_ANY(arg), VARSIZE_ANY_EXHDR(arg));
+}
+
+int
+bpchartruelen(char *s, int len)
+{
 	int			i;
-	int			len;
 
-	len = VARSIZE_ANY_EXHDR(arg);
+	/*
+	 * Note that we rely on the assumption that ' ' is a singleton unit on
+	 * every supported multibyte server encoding.
+	 */
 	for (i = len - 1; i >= 0; i--)
 	{
 		if (s[i] != ' ')
@@ -858,6 +866,23 @@ bpcharcmp(PG_FUNCTION_ARGS)
 	PG_RETURN_INT32(cmp);
 }
 
+Datum
+bpchar_sortsupport(PG_FUNCTION_ARGS)
+{
+	SortSupport ssup = (SortSupport) PG_GETARG_POINTER(0);
+	Oid			collid = ssup->ssup_collation;
+	MemoryContext oldcontext;
+
+	oldcontext = MemoryContextSwitchTo(ssup->ssup_cxt);
+
+	/* Use generic string SortSupport */
+	varstr_sortsupport(ssup, collid, true);
+
+	MemoryContextSwitchTo(oldcontext);
+
+	PG_RETURN_VOID();
+}
+
 Datum
 bpchar_larger(PG_FUNCTION_ARGS)
 {
@@ -926,8 +951,9 @@ hashbpchar(PG_FUNCTION_ARGS)
 /*
  * The following operators support character-by-character comparison
  * of bpchar datums, to allow building indexes suitable for LIKE clauses.
- * Note that the regular bpchareq/bpcharne comparison operators are assumed
- * to be compatible with these!
+ * Note that the regular bpchareq/bpcharne comparison operators, and
+ * regular support functions 1 and 2 with "C" collation are assumed to be
+ * compatible with these!
  */
 
 static int
@@ -1030,3 +1056,20 @@ btbpchar_pattern_cmp(PG_FUNCTION_ARGS)
 
 	PG_RETURN_INT32(result);
 }
+
+
+Datum
+btbpchar_pattern_sortsupport(PG_FUNCTION_ARGS)
+{
+	SortSupport ssup = (SortSupport) PG_GETARG_POINTER(0);
+	MemoryContext oldcontext;
+
+	oldcontext = MemoryContextSwitchTo(ssup->ssup_cxt);
+
+	/* Use generic string SortSupport, forcing "C" collation */
+	varstr_sortsupport(ssup, C_COLLATION_OID, true);
+
+	MemoryContextSwitchTo(oldcontext);
+
+	PG_RETURN_VOID();
+}

src/include/catalog/catversion.h

@@ -53,6 +53,6 @@
  */
 
 /*							yyyymmddN */
-#define CATALOG_VERSION_NO	201601281
+#define CATALOG_VERSION_NO	201602031
 
 #endif

src/include/catalog/pg_amproc.h

@@ -80,7 +80,9 @@ DATA(insert (	421   702 702 1 357 ));
 DATA(insert (	423   1560 1560 1 1596 ));
 DATA(insert (	424   16 16 1 1693 ));
 DATA(insert (	426   1042 1042 1 1078 ));
+DATA(insert (	426   1042 1042 2 3328 ));
 DATA(insert (	428   17 17 1 1954 ));
+DATA(insert (	428   17 17 2 3331 ));
 DATA(insert (	429   18 18 1 358 ));
 DATA(insert (	434   1082 1082 1 1092 ));
 DATA(insert (	434   1082 1082 2 3136 ));
@@ -128,7 +130,9 @@ DATA(insert (	1996   1083 1083 1 1107 ));
 DATA(insert (	2000   1266 1266 1 1358 ));
 DATA(insert (	2002   1562 1562 1 1672 ));
 DATA(insert (	2095   25 25 1 2166 ));
+DATA(insert (	2095   25 25 2 3332 ));
 DATA(insert (	2097   1042 1042 1 2180 ));
+DATA(insert (	2097   1042 1042 2 3333 ));
 DATA(insert (	2099   790 790 1  377 ));
 DATA(insert (	2233   703 703 1  380 ));
 DATA(insert (	2234   704 704 1  381 ));

src/include/catalog/pg_proc.h

@@ -1130,6 +1130,8 @@ DATA(insert OID = 1064 (  bpchar_smaller   PGNSP PGUID 12 1 0 0 0 f f f f t f i
 DESCR("smaller of two");
 DATA(insert OID = 1078 (  bpcharcmp		   PGNSP PGUID 12 1 0 0 0 f f f f t f i s 2 0 23 "1042 1042" _null_ _null_ _null_ _null_ _null_ bpcharcmp _null_ _null_ _null_ ));
 DESCR("less-equal-greater");
+DATA(insert OID = 3328 ( bpchar_sortsupport PGNSP PGUID 12 1 0 0 0 f f f f t f i s 1 0 2278 "2281" _null_ _null_ _null_ _null_ _null_ bpchar_sortsupport _null_ _null_ _null_ ));
+DESCR("sort support");
 DATA(insert OID = 1080 (  hashbpchar	   PGNSP PGUID 12 1 0 0 0 f f f f t f i s 1 0 23 "1042" _null_ _null_ _null_ _null_ _null_	hashbpchar _null_ _null_ _null_ ));
 DESCR("hash");
 DATA(insert OID = 1081 (  format_type	   PGNSP PGUID 12 1 0 0 0 f f f f f f s s 2 0 25 "26 23" _null_ _null_ _null_ _null_ _null_ format_type _null_ _null_ _null_ ));
@@ -2861,6 +2863,8 @@ DATA(insert OID = 1952 (  byteage		   PGNSP PGUID 12 1 0 0 0 f f f t t f i s 2 0
 DATA(insert OID = 1953 (  byteane		   PGNSP PGUID 12 1 0 0 0 f f f t t f i s 2 0 16 "17 17" _null_ _null_ _null_ _null_ _null_ byteane _null_ _null_ _null_ ));
 DATA(insert OID = 1954 (  byteacmp		   PGNSP PGUID 12 1 0 0 0 f f f f t f i s 2 0 23 "17 17" _null_ _null_ _null_ _null_ _null_ byteacmp _null_ _null_ _null_ ));
 DESCR("less-equal-greater");
+DATA(insert OID = 3331 (  bytea_sortsupport PGNSP PGUID 12 1 0 0 0 f f f f t f i s 1 0 2278 "2281" _null_ _null_ _null_ _null_ _null_ bytea_sortsupport _null_ _null_ _null_ ));
+DESCR("sort support");
 
 DATA(insert OID = 3917 (  timestamp_transform PGNSP PGUID 12 1 0 0 0 f f f f t f i s 1 0 2281 "2281" _null_ _null_ _null_ _null_ _null_ timestamp_transform _null_ _null_ _null_ ));
 DESCR("transform a timestamp length coercion");
@@ -3373,6 +3377,8 @@ DATA(insert OID = 2163 ( text_pattern_ge	 PGNSP PGUID 12 1 0 0 0 f f f f t f i s
 DATA(insert OID = 2164 ( text_pattern_gt	 PGNSP PGUID 12 1 0 0 0 f f f f t f i s 2 0 16 "25 25" _null_ _null_ _null_ _null_ _null_ text_pattern_gt _null_ _null_ _null_ ));
 DATA(insert OID = 2166 ( bttext_pattern_cmp  PGNSP PGUID 12 1 0 0 0 f f f f t f i s 2 0 23 "25 25" _null_ _null_ _null_ _null_ _null_ bttext_pattern_cmp _null_ _null_ _null_ ));
 DESCR("less-equal-greater");
+DATA(insert OID = 3332 ( bttext_pattern_sortsupport PGNSP PGUID 12 1 0 0 0 f f f f t f i s 1 0 2278 "2281" _null_ _null_ _null_ _null_ _null_ bttext_pattern_sortsupport _null_ _null_ _null_ ));
+DESCR("sort support");
 
 DATA(insert OID = 2174 ( bpchar_pattern_lt	  PGNSP PGUID 12 1 0 0 0 f f f f t f i s 2 0 16 "1042 1042" _null_ _null_ _null_ _null_ _null_ bpchar_pattern_lt _null_ _null_ _null_ ));
 DATA(insert OID = 2175 ( bpchar_pattern_le	  PGNSP PGUID 12 1 0 0 0 f f f f t f i s 2 0 16 "1042 1042" _null_ _null_ _null_ _null_ _null_ bpchar_pattern_le _null_ _null_ _null_ ));
@@ -3380,6 +3386,8 @@ DATA(insert OID = 2177 ( bpchar_pattern_ge	  PGNSP PGUID 12 1 0 0 0 f f f f t f
 DATA(insert OID = 2178 ( bpchar_pattern_gt	  PGNSP PGUID 12 1 0 0 0 f f f f t f i s 2 0 16 "1042 1042" _null_ _null_ _null_ _null_ _null_ bpchar_pattern_gt _null_ _null_ _null_ ));
 DATA(insert OID = 2180 ( btbpchar_pattern_cmp PGNSP PGUID 12 1 0 0 0 f f f f t f i s 2 0 23 "1042 1042" _null_ _null_ _null_ _null_ _null_ btbpchar_pattern_cmp _null_ _null_ _null_ ));
 DESCR("less-equal-greater");
+DATA(insert OID = 3333 ( btbpchar_pattern_sortsupport PGNSP PGUID 12 1 0 0 0 f f f f t f i s 1 0 2278 "2281" _null_ _null_ _null_ _null_ _null_ btbpchar_pattern_sortsupport _null_ _null_ _null_ ));
+DESCR("sort support");
 
 DATA(insert OID = 2188 ( btint48cmp			PGNSP PGUID 12 1 0 0 0 f f f f t f i s 2 0 23 "23 20" _null_ _null_ _null_ _null_ _null_ btint48cmp _null_ _null_ _null_ ));
 DESCR("less-equal-greater");

src/include/utils/builtins.h

@@ -16,6 +16,7 @@
 
 #include "fmgr.h"
 #include "nodes/parsenodes.h"
+#include "utils/sortsupport.h"
 
 /*
  *		Defined in adt/
@@ -761,8 +762,10 @@ extern Datum bpcharle(PG_FUNCTION_ARGS);
 extern Datum bpchargt(PG_FUNCTION_ARGS);
 extern Datum bpcharge(PG_FUNCTION_ARGS);
 extern Datum bpcharcmp(PG_FUNCTION_ARGS);
+extern Datum bpchar_sortsupport(PG_FUNCTION_ARGS);
 extern Datum bpchar_larger(PG_FUNCTION_ARGS);
 extern Datum bpchar_smaller(PG_FUNCTION_ARGS);
+extern int	bpchartruelen(char *s, int len);
 extern Datum bpcharlen(PG_FUNCTION_ARGS);
 extern Datum bpcharoctetlen(PG_FUNCTION_ARGS);
 extern Datum hashbpchar(PG_FUNCTION_ARGS);
@@ -771,6 +774,7 @@ extern Datum bpchar_pattern_le(PG_FUNCTION_ARGS);
 extern Datum bpchar_pattern_gt(PG_FUNCTION_ARGS);
 extern Datum bpchar_pattern_ge(PG_FUNCTION_ARGS);
 extern Datum btbpchar_pattern_cmp(PG_FUNCTION_ARGS);
+extern Datum btbpchar_pattern_sortsupport(PG_FUNCTION_ARGS);
 
 extern Datum varcharin(PG_FUNCTION_ARGS);
 extern Datum varcharout(PG_FUNCTION_ARGS);
@@ -808,6 +812,7 @@ extern Datum text_pattern_le(PG_FUNCTION_ARGS);
 extern Datum text_pattern_gt(PG_FUNCTION_ARGS);
 extern Datum text_pattern_ge(PG_FUNCTION_ARGS);
 extern Datum bttext_pattern_cmp(PG_FUNCTION_ARGS);
+extern Datum bttext_pattern_sortsupport(PG_FUNCTION_ARGS);
 extern Datum textlen(PG_FUNCTION_ARGS);
 extern Datum textoctetlen(PG_FUNCTION_ARGS);
 extern Datum textpos(PG_FUNCTION_ARGS);
@@ -818,6 +823,7 @@ extern Datum textoverlay_no_len(PG_FUNCTION_ARGS);
 extern Datum name_text(PG_FUNCTION_ARGS);
 extern Datum text_name(PG_FUNCTION_ARGS);
 extern int	varstr_cmp(char *arg1, int len1, char *arg2, int len2, Oid collid);
+extern void	varstr_sortsupport(SortSupport ssup, Oid collid, bool bpchar);
 extern int varstr_levenshtein(const char *source, int slen,
 				   const char *target, int tlen,
 				   int ins_c, int del_c, int sub_c,

src/include/utils/bytea.h

@@ -42,6 +42,7 @@ extern Datum byteale(PG_FUNCTION_ARGS);
 extern Datum byteagt(PG_FUNCTION_ARGS);
 extern Datum byteage(PG_FUNCTION_ARGS);
 extern Datum byteacmp(PG_FUNCTION_ARGS);
+extern Datum bytea_sortsupport(PG_FUNCTION_ARGS);
 extern Datum byteacat(PG_FUNCTION_ARGS);
 extern Datum byteapos(PG_FUNCTION_ARGS);
 extern Datum bytea_substr(PG_FUNCTION_ARGS);