Add a new GUC parameter backslash_quote, which determines whether the SQL

parser will allow "\'" to be used to represent a literal quote mark.  The
"\'" representation has been deprecated for some time in favor of the
SQL-standard representation "''" (two single quote marks), but it has been
used often enough that just disallowing it immediately won't do.  Hence
backslash_quote allows the settings "on", "off", and "safe_encoding",
the last meaning to allow "\'" only if client_encoding is a valid server
encoding.  That is now the default, and the reason is that in encodings
such as SJIS that allow 0x5c (ASCII backslash) to be the last byte of a
multibyte character, accepting "\'" allows SQL-injection attacks as per
CVE-2006-2314 (further details will be published after release).  The
"on" setting is available for backward compatibility, but it must not be
used with clients that are exposed to untrusted input.

Thanks to Akio Ishida and Yasuo Ohgaki for identifying this security issue.

1 files changed, 32 insertions, 5 deletions

diff --git a/src/backend/parser/scan.l b/src/backend/parser/scan.l
index 433e42bf6a0..f0a424a8d02 100644
--- a/src/backend/parser/scan.l
+++ b/src/backend/parser/scan.l
@@ -10,7 +10,7 @@
  * Portions Copyright (c) 1994, Regents of the University of California
  *
  * IDENTIFICATION
- *	  $PostgreSQL: pgsql/src/backend/parser/scan.l,v 1.119.4.1 2005/08/16 00:48:29 tgl Exp $
+ *	  $PostgreSQL: pgsql/src/backend/parser/scan.l,v 1.119.4.2 2006/05/21 20:11:25 tgl Exp $
  *
  *-------------------------------------------------------------------------
  */
@@ -36,6 +36,15 @@ static int		xcdepth = 0;	/* depth of nesting in slash-star comments */
 static char    *dolqstart;      /* current $foo$ quote start string */
 
 /*
+ * GUC variables.  This is a DIRECT violation of the warning given at the
+ * head of gram.y, ie flex/bison code must not depend on any GUC variables;
+ * as such, changing their values can induce very unintuitive behavior.
+ * But we shall have to live with it as a short-term thing until the switch
+ * to SQL-standard string syntax is complete.
+ */
+BackslashQuoteType backslash_quote = BACKSLASH_QUOTE_SAFE_ENCODING;
+
+/*
  * literalbuf is used to accumulate literal values when multiple rules
  * are needed to parse a single literal.  Call startlit to reset buffer
  * to empty, addlit to add text.  Note that the buffer is palloc'd and
@@ -49,6 +58,7 @@ static int		literalalloc;	/* current allocated buffer size */
 static void addlit(char *ytext, int yleng);
 static void addlitchar(unsigned char ychar);
 static char *litbufdup(void);
+static int	pg_err_position(void);
 
 /*
  * When we parse a token that requires multiple lexer rules to process,
@@ -391,6 +401,17 @@ other			.
 					addlit(yytext, yyleng);
 				}
 <xq>{xqescape}  {
+					if (yytext[1] == '\'')
+					{
+						if (backslash_quote == BACKSLASH_QUOTE_OFF ||
+							(backslash_quote == BACKSLASH_QUOTE_SAFE_ENCODING &&
+							 PG_ENCODING_IS_CLIENT_ONLY(pg_get_client_encoding())))
+							ereport(ERROR,
+									(errcode(ERRCODE_INVALID_USE_OF_ESCAPE_CHARACTER),
+									 errmsg("unsafe use of \\' in a string literal"),
+									 errhint("Use '' to write quotes in strings. \\' is insecure in client-only encodings."),
+									 errposition(pg_err_position())));
+					}
 					addlitchar(unescape_single_char(yytext[1]));
 				}
 <xq>{xqoctesc}  {
@@ -613,14 +634,20 @@ other			.
 
 %%
 
-void
-yyerror(const char *message)
+static int
+pg_err_position(void)
 {
 	const char *loc = token_start ? token_start : yytext;
-	int			cursorpos;
 
 	/* in multibyte encodings, return index in characters not bytes */
-	cursorpos = pg_mbstrlen_with_len(scanbuf, loc - scanbuf) + 1;
+	return pg_mbstrlen_with_len(scanbuf, loc - scanbuf) + 1;
+}
+
+void
+yyerror(const char *message)
+{
+	const char *loc = token_start ? token_start : yytext;
+	int			cursorpos = pg_err_position();
 
 	if (*loc == YY_END_OF_BUFFER_CHAR)
 	{