Fix bugs in plpgsql and ecpg caused by assuming that isspace() would only

return true for exactly the characters treated as whitespace by their flex
scanners.  Per report from Victor Snezhko and subsequent investigation.

Also fix a passel of unsafe usages of <ctype.h> functions, that is, ye olde
char-vs-unsigned-char issue.  I won't miss <ctype.h> when we are finally
able to stop using it.

12 files changed, 76 insertions, 28 deletions

diff --git a/src/backend/parser/scan.l b/src/backend/parser/scan.l
index c97c9e6ada3..303350fac9a 100644
--- a/src/backend/parser/scan.l
+++ b/src/backend/parser/scan.l
@@ -24,7 +24,7 @@
  * Portions Copyright (c) 1994, Regents of the University of California
  *
  * IDENTIFICATION
- *	  $PostgreSQL: pgsql/src/backend/parser/scan.l,v 1.137 2006/09/03 03:19:44 momjian Exp $
+ *	  $PostgreSQL: pgsql/src/backend/parser/scan.l,v 1.138 2006/09/22 21:39:57 tgl Exp $
  *
  *-------------------------------------------------------------------------
  */
@@ -145,6 +145,9 @@ static unsigned char unescape_single_char(unsigned char c);
  * did not end with a newline.
  *
  * XXX perhaps \f (formfeed) should be treated as a newline as well?
+ *
+ * XXX if you change the set of whitespace characters, fix scanner_isspace()
+ * to agree, and see also the plpgsql lexer.
  */
 
 space			[ \t\n\r\f]
diff --git a/src/backend/parser/scansup.c b/src/backend/parser/scansup.c
index c326ef0e3ab..807338652b4 100644
--- a/src/backend/parser/scansup.c
+++ b/src/backend/parser/scansup.c
@@ -9,7 +9,7 @@
  *
  *
  * IDENTIFICATION
- *	  $PostgreSQL: pgsql/src/backend/parser/scansup.c,v 1.33 2006/07/14 14:52:22 momjian Exp $
+ *	  $PostgreSQL: pgsql/src/backend/parser/scansup.c,v 1.34 2006/09/22 21:39:57 tgl Exp $
  *
  *-------------------------------------------------------------------------
  */
@@ -183,3 +183,26 @@ truncate_identifier(char *ident, int len, bool warn)
 		ident[len] = '\0';
 	}
 }
+
+/*
+ * scanner_isspace() --- return TRUE if flex scanner considers char whitespace
+ *
+ * This should be used instead of the potentially locale-dependent isspace()
+ * function when it's important to match the lexer's behavior.
+ *
+ * In principle we might need similar functions for isalnum etc, but for the
+ * moment only isspace seems needed.
+ */
+bool
+scanner_isspace(char ch)
+{
+	/* This must match scan.l's list of {space} characters */
+	/* and plpgsql's scan.l as well */
+	if (ch == ' ' ||
+		ch == '\t' ||
+		ch == '\n' ||
+		ch == '\r' ||
+		ch == '\f')
+		return true;
+	return false;
+}
diff --git a/src/backend/utils/misc/guc.c b/src/backend/utils/misc/guc.c
index 7f7ded9e6a5..9467e6412c6 100644
--- a/src/backend/utils/misc/guc.c
+++ b/src/backend/utils/misc/guc.c
@@ -10,7 +10,7 @@
  * Written by Peter Eisentraut <peter_e@gmx.net>.
  *
  * IDENTIFICATION
- *	  $PostgreSQL: pgsql/src/backend/utils/misc/guc.c,v 1.351 2006/09/22 17:41:21 petere Exp $
+ *	  $PostgreSQL: pgsql/src/backend/utils/misc/guc.c,v 1.352 2006/09/22 21:39:57 tgl Exp $
  *
  *--------------------------------------------------------------------
  */
@@ -6155,7 +6155,7 @@ assign_custom_variable_classes(const char *newval, bool doit, GucSource source)
 	initStringInfo(&buf);
 	while ((c = *cp++) != 0)
 	{
-		if (isspace(c))
+		if (isspace((unsigned char) c))
 		{
 			if (symLen > 0)
 				hasSpaceAfterToken = true;
@@ -6173,7 +6173,7 @@ assign_custom_variable_classes(const char *newval, bool doit, GucSource source)
 			continue;
 		}
 
-		if (hasSpaceAfterToken || !isalnum(c))
+		if (hasSpaceAfterToken || !isalnum((unsigned char) c))
 		{
 			/*
 			 * Syntax error due to token following space after token or non
diff --git a/src/bin/pg_dump/pg_dump.c b/src/bin/pg_dump/pg_dump.c
index 78a653fb48c..fbf131f5779 100644
--- a/src/bin/pg_dump/pg_dump.c
+++ b/src/bin/pg_dump/pg_dump.c
@@ -12,7 +12,7 @@
  *	by PostgreSQL
  *
  * IDENTIFICATION
- *	  $PostgreSQL: pgsql/src/bin/pg_dump/pg_dump.c,v 1.447 2006/08/21 00:57:25 tgl Exp $
+ *	  $PostgreSQL: pgsql/src/bin/pg_dump/pg_dump.c,v 1.448 2006/09/22 21:39:57 tgl Exp $
  *
  *-------------------------------------------------------------------------
  */
@@ -367,10 +367,10 @@ main(int argc, char **argv)
 
 					new_obj_name->next = NULL;
 					new_obj_name->name = strdup(optarg);
-					new_obj_name->is_include = islower(c) ? true : false;
+					new_obj_name->is_include = islower((unsigned char) c) ? true : false;
 
 					/* add new entry to the proper list */
-					if (tolower(c) == 'n')
+					if (tolower((unsigned char) c) == 'n')
 					{
 						if (!schemaList_tail)
 							schemaList_tail = schemaList = new_obj_name;
diff --git a/src/bin/psql/tab-complete.c b/src/bin/psql/tab-complete.c
index f8edb78b570..64b9576a662 100644
--- a/src/bin/psql/tab-complete.c
+++ b/src/bin/psql/tab-complete.c
@@ -3,7 +3,7 @@
  *
  * Copyright (c) 2000-2006, PostgreSQL Global Development Group
  *
- * $PostgreSQL: pgsql/src/bin/psql/tab-complete.c,v 1.154 2006/07/14 14:52:27 momjian Exp $
+ * $PostgreSQL: pgsql/src/bin/psql/tab-complete.c,v 1.155 2006/09/22 21:39:57 tgl Exp $
  */
 
 /*----------------------------------------------------------------------
@@ -1090,8 +1090,8 @@ psql_completion(char *text, int start, int end)
 	/* Complete "AS ON <sth with a 'T' :)>" with a "TO" */
 	else if (pg_strcasecmp(prev3_wd, "AS") == 0 &&
 			 pg_strcasecmp(prev2_wd, "ON") == 0 &&
-			 (toupper((unsigned char) prev_wd[4]) == 'T' ||
-			  toupper((unsigned char) prev_wd[5]) == 'T'))
+			 (pg_toupper((unsigned char) prev_wd[4]) == 'T' ||
+			  pg_toupper((unsigned char) prev_wd[5]) == 'T'))
 		COMPLETE_WITH_CONST("TO");
 	/* Complete "AS ON <sth> TO" with a table name */
 	else if (pg_strcasecmp(prev4_wd, "AS") == 0 &&
diff --git a/src/include/parser/scansup.h b/src/include/parser/scansup.h
index 85f100a0a43..42575e21391 100644
--- a/src/include/parser/scansup.h
+++ b/src/include/parser/scansup.h
@@ -7,7 +7,7 @@
  * Portions Copyright (c) 1996-2006, PostgreSQL Global Development Group
  * Portions Copyright (c) 1994, Regents of the University of California
  *
- * $PostgreSQL: pgsql/src/include/parser/scansup.h,v 1.19 2006/03/05 15:58:58 momjian Exp $
+ * $PostgreSQL: pgsql/src/include/parser/scansup.h,v 1.20 2006/09/22 21:39:58 tgl Exp $
  *
  *-------------------------------------------------------------------------
  */
@@ -22,4 +22,6 @@ extern char *downcase_truncate_identifier(const char *ident, int len,
 
 extern void truncate_identifier(char *ident, int len, bool warn);
 
+extern bool scanner_isspace(char ch);
+
 #endif   /* SCANSUP_H */
diff --git a/src/include/port.h b/src/include/port.h
index 792ff28d7df..ecf04941448 100644
--- a/src/include/port.h
+++ b/src/include/port.h
@@ -6,7 +6,7 @@
  * Portions Copyright (c) 1996-2006, PostgreSQL Global Development Group
  * Portions Copyright (c) 1994, Regents of the University of California
  *
- * $PostgreSQL: pgsql/src/include/port.h,v 1.98 2006/09/11 20:10:30 tgl Exp $
+ * $PostgreSQL: pgsql/src/include/port.h,v 1.99 2006/09/22 21:39:58 tgl Exp $
  *
  *-------------------------------------------------------------------------
  */
@@ -59,7 +59,7 @@ extern void get_parent_directory(char *path);
 ( \
 	((filename)[0] == '/') || \
 	(filename)[0] == '\\' || \
-	(isalpha((filename)[0]) && (filename)[1] == ':' && \
+	(isalpha((unsigned char) ((filename)[0])) && (filename)[1] == ':' && \
 	((filename)[2] == '\\' || (filename)[2] == '/')) \
 )
 #endif
diff --git a/src/interfaces/ecpg/preproc/pgc.l b/src/interfaces/ecpg/preproc/pgc.l
index e9f87fe183b..1af79f8e9e4 100644
--- a/src/interfaces/ecpg/preproc/pgc.l
+++ b/src/interfaces/ecpg/preproc/pgc.l
@@ -12,7 +12,7 @@
  *
  *
  * IDENTIFICATION
- *	  $PostgreSQL: pgsql/src/interfaces/ecpg/preproc/pgc.l,v 1.149 2006/08/18 15:59:35 meskes Exp $
+ *	  $PostgreSQL: pgsql/src/interfaces/ecpg/preproc/pgc.l,v 1.150 2006/09/22 21:39:58 tgl Exp $
  *
  *-------------------------------------------------------------------------
  */
@@ -47,6 +47,7 @@ static void addlit(char *ytext, int yleng);
 static void addlitchar (unsigned char);
 static void parse_include (void);
 static void check_escape_warning(void);
+static bool ecpg_isspace(char ch);
 
 char *token_start;
 int state_before;
@@ -245,6 +246,9 @@ param			\${integer}
  * did not end with a newline.
  *
  * XXX perhaps \f (formfeed) should be treated as a newline as well?
+ *
+ * XXX if you change the set of whitespace characters, fix ecpg_isspace()
+ * to agree.
  */
 
 ccomment		"//".*\n
@@ -872,7 +876,7 @@ cppline			{space}*#(.*\\{space})*.*{newline}
 					 *	contains at least one non-space character plus the ";"
 					 */
 					for (i = strlen(yytext)-2;
-						 i > 0 && isspace((unsigned char) yytext[i]);
+						 i > 0 && ecpg_isspace(yytext[i]);
 						 i-- )
 						;
 					yytext[i+1] = '\0';
@@ -1060,7 +1064,7 @@ cppline			{space}*#(.*\\{space})*.*{newline}
 						 *	contains at least one non-space character plus the ";"
 						 */
 						for (i = strlen(yytext)-2;
-							 i > 0 && isspace((unsigned char) yytext[i]);
+							 i > 0 && ecpg_isspace(yytext[i]);
 							 i-- )
 							;
 						yytext[i+1] = '\0';
@@ -1252,7 +1256,7 @@ parse_include(void)
 	 * yytext contains at least one non-space character plus the ";" 
 	 */
   	for (i = strlen(yytext)-2;
-		 i > 0 && isspace((unsigned char) yytext[i]);
+		 i > 0 && ecpg_isspace(yytext[i]);
 		 i--)
 		;
 
@@ -1328,3 +1332,18 @@ check_escape_warning(void)
 	       	mmerror (PARSE_ERROR, ET_WARNING, "nonstandard use of escape in a string literal");
 	warn_on_first_escape = false;   /* warn only once per string */
 }
+
+/*
+ * ecpg_isspace() --- return TRUE if flex scanner considers char whitespace
+ */
+static bool
+ecpg_isspace(char ch)
+{
+	if (ch == ' ' ||
+		ch == '\t' ||
+		ch == '\n' ||
+		ch == '\r' ||
+		ch == '\f')
+		return true;
+	return false;
+}
diff --git a/src/interfaces/libpq/fe-auth.c b/src/interfaces/libpq/fe-auth.c
index db8bbdd78da..bcc4c570bf5 100644
--- a/src/interfaces/libpq/fe-auth.c
+++ b/src/interfaces/libpq/fe-auth.c
@@ -10,7 +10,7 @@
  * exceed INITIAL_EXPBUFFER_SIZE (currently 256 bytes).
  *
  * IDENTIFICATION
- *	  $PostgreSQL: pgsql/src/interfaces/libpq/fe-auth.c,v 1.119 2006/07/14 14:52:27 momjian Exp $
+ *	  $PostgreSQL: pgsql/src/interfaces/libpq/fe-auth.c,v 1.120 2006/09/22 21:39:58 tgl Exp $
  *
  *-------------------------------------------------------------------------
  */
@@ -89,7 +89,7 @@ pg_an_to_ln(char *aname)
 		*p = '\0';
 #ifdef WIN32
 	for (p = aname; *p; p++)
-		*p = pg_tolower(*p);
+		*p = pg_tolower((unsigned char) *p);
 #endif
 
 	return aname;
diff --git a/src/pl/plpgsql/src/pl_exec.c b/src/pl/plpgsql/src/pl_exec.c
index 7b350caf155..c5e036ccb77 100644
--- a/src/pl/plpgsql/src/pl_exec.c
+++ b/src/pl/plpgsql/src/pl_exec.c
@@ -8,7 +8,7 @@
  *
  *
  * IDENTIFICATION
- *	  $PostgreSQL: pgsql/src/pl/plpgsql/src/pl_exec.c,v 1.178 2006/09/06 20:40:48 tgl Exp $
+ *	  $PostgreSQL: pgsql/src/pl/plpgsql/src/pl_exec.c,v 1.179 2006/09/22 21:39:58 tgl Exp $
  *
  *-------------------------------------------------------------------------
  */
@@ -26,6 +26,7 @@
 #include "funcapi.h"
 #include "optimizer/clauses.h"
 #include "parser/parse_expr.h"
+#include "parser/scansup.h"
 #include "tcop/tcopprot.h"
 #include "utils/array.h"
 #include "utils/builtins.h"
@@ -2527,7 +2528,7 @@ exec_stmt_dynexecute(PLpgSQL_execstate *estate,
 				char	   *ptr;
 
 				for (ptr = querystr; *ptr; ptr++)
-					if (!isspace((unsigned char) *ptr))
+					if (!scanner_isspace(*ptr))
 						break;
 				if (*ptr == 'S' || *ptr == 's')
 					ereport(ERROR,
diff --git a/src/pl/plpgsql/src/pl_funcs.c b/src/pl/plpgsql/src/pl_funcs.c
index f763e25e8d6..566480485b3 100644
--- a/src/pl/plpgsql/src/pl_funcs.c
+++ b/src/pl/plpgsql/src/pl_funcs.c
@@ -8,7 +8,7 @@
  *
  *
  * IDENTIFICATION
- *	  $PostgreSQL: pgsql/src/pl/plpgsql/src/pl_funcs.c,v 1.54 2006/08/14 21:14:41 tgl Exp $
+ *	  $PostgreSQL: pgsql/src/pl/plpgsql/src/pl_funcs.c,v 1.55 2006/09/22 21:39:58 tgl Exp $
  *
  *-------------------------------------------------------------------------
  */
@@ -381,7 +381,7 @@ plpgsql_convert_ident(const char *s, char **output, int numidents)
 			/* Normal identifier: extends till dot or whitespace */
 			const char *thisstart = s;
 
-			while (*s && *s != '.' && !isspace((unsigned char) *s))
+			while (*s && *s != '.' && !scanner_isspace(*s))
 				s++;
 			/* Downcase and truncate to NAMEDATALEN */
 			curident = downcase_truncate_identifier(thisstart, s - thisstart,
@@ -400,11 +400,11 @@ plpgsql_convert_ident(const char *s, char **output, int numidents)
 		/* If not done, skip whitespace, dot, whitespace */
 		if (*s)
 		{
-			while (*s && isspace((unsigned char) *s))
+			while (*s && scanner_isspace(*s))
 				s++;
 			if (*s++ != '.')
 				elog(ERROR, "expected dot between identifiers: %s", sstart);
-			while (*s && isspace((unsigned char) *s))
+			while (*s && scanner_isspace(*s))
 				s++;
 			if (*s == '\0')
 				elog(ERROR, "expected another identifier: %s", sstart);
diff --git a/src/port/path.c b/src/port/path.c
index 818b3212c43..4acb8046bfc 100644
--- a/src/port/path.c
+++ b/src/port/path.c
@@ -8,7 +8,7 @@
  *
  *
  * IDENTIFICATION
- *	  $PostgreSQL: pgsql/src/port/path.c,v 1.67 2006/09/11 20:10:30 tgl Exp $
+ *	  $PostgreSQL: pgsql/src/port/path.c,v 1.68 2006/09/22 21:39:58 tgl Exp $
  *
  *-------------------------------------------------------------------------
  */
@@ -69,7 +69,7 @@ skip_drive(const char *path)
 		while (*path && !IS_DIR_SEP(*path))
 			path++;
 	}
-	else if (isalpha(path[0]) && path[1] == ':')
+	else if (isalpha((unsigned char) path[0]) && path[1] == ':')
 	{
 		path += 2;
 	}