Specify the encoding of input to fmtId()

anarazel · anarazel · commit 5df4e1632e65 · 2025-02-10T10:03:40.000-05:00
This commit adds fmtIdEnc() and fmtQualifiedIdEnc(), which allow to specify the encoding as an explicit argument. Additionally setFmtEncoding() is provided, which defines the encoding when no explicit encoding is provided, to avoid breaking all code using fmtId(). All users of fmtId()/fmtQualifiedId() are either converted to the explicit version or a call to setFmtEncoding() has been added. This commit does not yet utilize the now well-defined encoding, that will happen in a subsequent commit. Reviewed-by: Noah Misch <noah@leadboat.com> Reviewed-by: Tom Lane <tgl@sss.pgh.pa.us> Backpatch-through: 13 Security: CVE-2025-1094
diff --git a/src/bin/pg_dump/pg_backup_archiver.c b/src/bin/pg_dump/pg_backup_archiver.c
@@ -2725,6 +2725,7 @@ processEncodingEntry(ArchiveHandle *AH, TocEntry *te)
 			fatal("unrecognized encoding \"%s\"",
 				  ptr1);
 		AH->public.encoding = encoding;
+		setFmtEncoding(encoding);
 	}
 	else
 		fatal("invalid ENCODING item: %s",
diff --git a/src/bin/pg_dump/pg_dump.c b/src/bin/pg_dump/pg_dump.c
@@ -1094,6 +1094,7 @@ setup_connection(Archive *AH, const char *dumpencoding,
 	 * we know how to escape strings.
 	 */
 	AH->encoding = PQclientEncoding(conn);
+	setFmtEncoding(AH->encoding);
 
 	std_strings = PQparameterStatus(conn, "standard_conforming_strings");
 	AH->std_strings = (std_strings && strcmp(std_strings, "on") == 0);
diff --git a/src/bin/pg_dump/pg_dumpall.c b/src/bin/pg_dump/pg_dumpall.c
@@ -507,6 +507,7 @@ main(int argc, char *argv[])
 	 * we know how to escape strings.
 	 */
 	encoding = PQclientEncoding(conn);
+	setFmtEncoding(encoding);
 	std_strings = PQparameterStatus(conn, "standard_conforming_strings");
 	if (!std_strings)
 		std_strings = "off";
diff --git a/src/bin/psql/command.c b/src/bin/psql/command.c
@@ -1220,6 +1220,7 @@ exec_command_encoding(PsqlScanState scan_state, bool active_branch)
 				/* save encoding info into psql internal data */
 				pset.encoding = PQclientEncoding(pset.db);
 				pset.popt.topt.encoding = pset.encoding;
+				setFmtEncoding(pset.encoding);
 				SetVariable(pset.vars, "ENCODING",
 							pg_encoding_to_char(pset.encoding));
 			}
@@ -3606,6 +3607,8 @@ SyncVariables(void)
 	pset.popt.topt.encoding = pset.encoding;
 	pset.sversion = PQserverVersion(pset.db);
 
+	setFmtEncoding(pset.encoding);
+
 	SetVariable(pset.vars, "DBNAME", PQdb(pset.db));
 	SetVariable(pset.vars, "USER", PQuser(pset.db));
 	SetVariable(pset.vars, "HOST", PQhost(pset.db));
diff --git a/src/bin/scripts/common.c b/src/bin/scripts/common.c
@@ -430,8 +430,9 @@ appendQualifiedRelation(PQExpBuffer buf, const char *spec,
 		exit(1);
 	}
 	appendPQExpBufferStr(buf,
-						 fmtQualifiedId(PQgetvalue(res, 0, 1),
-										PQgetvalue(res, 0, 0)));
+						 fmtQualifiedIdEnc(PQgetvalue(res, 0, 1),
+										   PQgetvalue(res, 0, 0),
+										   PQclientEncoding(conn)));
 	appendPQExpBufferStr(buf, columns);
 	PQclear(res);
 	termPQExpBuffer(&sql);
diff --git a/src/bin/scripts/createdb.c b/src/bin/scripts/createdb.c
@@ -190,6 +190,8 @@ main(int argc, char *argv[])
 
 	conn = connectMaintenanceDatabase(&cparams, progname, echo);
 
+	setFmtEncoding(PQclientEncoding(conn));
+
 	initPQExpBuffer(&sql);
 
 	appendPQExpBuffer(&sql, "CREATE DATABASE %s",
diff --git a/src/bin/scripts/createuser.c b/src/bin/scripts/createuser.c
@@ -266,6 +266,8 @@ main(int argc, char *argv[])
 
 	conn = connectMaintenanceDatabase(&cparams, progname, echo);
 
+	setFmtEncoding(PQclientEncoding(conn));
+
 	initPQExpBuffer(&sql);
 
 	printfPQExpBuffer(&sql, "CREATE ROLE %s", fmtId(newuser));
diff --git a/src/bin/scripts/dropdb.c b/src/bin/scripts/dropdb.c
@@ -127,13 +127,6 @@ main(int argc, char *argv[])
 			exit(0);
 	}
 
-	initPQExpBuffer(&sql);
-
-	appendPQExpBuffer(&sql, "DROP DATABASE %s%s%s;",
-					  (if_exists ? "IF EXISTS " : ""),
-					  fmtId(dbname),
-					  force ? " WITH (FORCE)" : "");
-
 	/* Avoid trying to drop postgres db while we are connected to it. */
 	if (maintenance_db == NULL && strcmp(dbname, "postgres") == 0)
 		maintenance_db = "template1";
@@ -147,6 +140,12 @@ main(int argc, char *argv[])
 
 	conn = connectMaintenanceDatabase(&cparams, progname, echo);
 
+	initPQExpBuffer(&sql);
+	appendPQExpBuffer(&sql, "DROP DATABASE %s%s%s;",
+					  (if_exists ? "IF EXISTS " : ""),
+					  fmtIdEnc(dbname, PQclientEncoding(conn)),
+					  force ? " WITH (FORCE)" : "");
+
 	if (echo)
 		printf("%s\n", sql.data);
 	result = PQexec(conn, sql.data);
diff --git a/src/bin/scripts/dropuser.c b/src/bin/scripts/dropuser.c
@@ -143,7 +143,8 @@ main(int argc, char *argv[])
 
 	initPQExpBuffer(&sql);
 	appendPQExpBuffer(&sql, "DROP ROLE %s%s;",
-					  (if_exists ? "IF EXISTS " : ""), fmtId(dropuser));
+					  (if_exists ? "IF EXISTS " : ""),
+					  fmtIdEnc(dropuser, PQclientEncoding(conn)));
 
 	if (echo)
 		printf("%s\n", sql.data);
diff --git a/src/bin/scripts/reindexdb.c b/src/bin/scripts/reindexdb.c
@@ -532,7 +532,8 @@ run_reindex_command(PGconn *conn, ReindexType type, const char *name,
 	{
 		case REINDEX_DATABASE:
 		case REINDEX_SYSTEM:
-			appendPQExpBufferStr(&sql, fmtId(name));
+			appendPQExpBufferStr(&sql,
+								 fmtIdEnc(name, PQclientEncoding(conn)));
 			break;
 		case REINDEX_INDEX:
 		case REINDEX_TABLE:
@@ -702,8 +703,9 @@ get_parallel_object_list(PGconn *conn, ReindexType type,
 	for (i = 0; i < ntups; i++)
 	{
 		appendPQExpBufferStr(&buf,
-							 fmtQualifiedId(PQgetvalue(res, i, 1),
-											PQgetvalue(res, i, 0)));
+							 fmtQualifiedIdEnc(PQgetvalue(res, i, 1),
+											   PQgetvalue(res, i, 0),
+											   PQclientEncoding(conn)));
 
 		simple_string_list_append(tables, buf.data);
 		resetPQExpBuffer(&buf);
diff --git a/src/bin/scripts/vacuumdb.c b/src/bin/scripts/vacuumdb.c
@@ -611,8 +611,9 @@ vacuum_one_database(const ConnParams *cparams,
 	for (i = 0; i < ntups; i++)
 	{
 		appendPQExpBufferStr(&buf,
-							 fmtQualifiedId(PQgetvalue(res, i, 1),
-											PQgetvalue(res, i, 0)));
+							 fmtQualifiedIdEnc(PQgetvalue(res, i, 1),
+											   PQgetvalue(res, i, 0),
+											   PQclientEncoding(conn)));
 
 		if (tables_listed && !PQgetisnull(res, i, 2))
 			appendPQExpBufferStr(&buf, PQgetvalue(res, i, 2));
diff --git a/src/fe_utils/string_utils.c b/src/fe_utils/string_utils.c
@@ -19,13 +19,16 @@
 
 #include "common/keywords.h"
 #include "fe_utils/string_utils.h"
+#include "mb/pg_wchar.h"
 
 static PQExpBuffer defaultGetLocalPQExpBuffer(void);
 
 /* Globals exported by this file */
 int			quote_all_identifiers = 0;
 PQExpBuffer (*getLocalPQExpBuffer) (void) = defaultGetLocalPQExpBuffer;
 
+static int	fmtIdEncoding = -1;
+
 
 /*
  * Returns a temporary PQExpBuffer, valid until the next call to the function.
@@ -54,14 +57,48 @@ defaultGetLocalPQExpBuffer(void)
 	return id_return;
 }
 
+/*
+ * Set the encoding that fmtId() and fmtQualifiedId() use.
+ *
+ * This is not safe against multiple connections having different encodings,
+ * but there is no real other way to address the need to know the encoding for
+ * fmtId()/fmtQualifiedId() input for safe escaping. Eventually we should get
+ * rid of fmtId().
+ */
+void
+setFmtEncoding(int encoding)
+{
+	fmtIdEncoding = encoding;
+}
+
+/*
+ * Return the currently configured encoding for fmtId() and fmtQualifiedId().
+ */
+static int
+getFmtEncoding(void)
+{
+	if (fmtIdEncoding != -1)
+		return fmtIdEncoding;
+
+	/*
+	 * In assertion builds it seems best to fail hard if the encoding was not
+	 * set, to make it easier to find places with missing calls. But in
+	 * production builds that seems like a bad idea, thus we instead just
+	 * default to UTF-8.
+	 */
+	Assert(fmtIdEncoding != -1);
+
+	return PG_UTF8;
+}
+
 /*
  *	Quotes input string if it's not a legitimate SQL identifier as-is.
  *
- *	Note that the returned string must be used before calling fmtId again,
+ *	Note that the returned string must be used before calling fmtIdEnc again,
  *	since we re-use the same return buffer each time.
  */
 const char *
-fmtId(const char *rawid)
+fmtIdEnc(const char *rawid, int encoding)
 {
 	PQExpBuffer id_return = getLocalPQExpBuffer();
 
@@ -134,25 +171,42 @@ fmtId(const char *rawid)
 }
 
 /*
- * fmtQualifiedId - construct a schema-qualified name, with quoting as needed.
+ *	Quotes input string if it's not a legitimate SQL identifier as-is.
+ *
+ *	Note that the returned string must be used before calling fmtId again,
+ *	since we re-use the same return buffer each time.
+ *
+ *  NB: This assumes setFmtEncoding() previously has been called to configure
+ *  the encoding of rawid. It is preferable to use fmtIdEnc() with an
+ *  explicit encoding.
+ */
+const char *
+fmtId(const char *rawid)
+{
+	return fmtIdEnc(rawid, getFmtEncoding());
+}
+
+/*
+ * fmtQualifiedIdEnc - construct a schema-qualified name, with quoting as
+ * needed.
  *
  * Like fmtId, use the result before calling again.
  *
  * Since we call fmtId and it also uses getLocalPQExpBuffer() we cannot
  * use that buffer until we're finished with calling fmtId().
  */
 const char *
-fmtQualifiedId(const char *schema, const char *id)
+fmtQualifiedIdEnc(const char *schema, const char *id, int encoding)
 {
 	PQExpBuffer id_return;
 	PQExpBuffer lcl_pqexp = createPQExpBuffer();
 
 	/* Some callers might fail to provide a schema name */
 	if (schema && *schema)
 	{
-		appendPQExpBuffer(lcl_pqexp, "%s.", fmtId(schema));
+		appendPQExpBuffer(lcl_pqexp, "%s.", fmtIdEnc(schema, encoding));
 	}
-	appendPQExpBufferStr(lcl_pqexp, fmtId(id));
+	appendPQExpBufferStr(lcl_pqexp, fmtIdEnc(id, encoding));
 
 	id_return = getLocalPQExpBuffer();
 
@@ -162,6 +216,24 @@ fmtQualifiedId(const char *schema, const char *id)
 	return id_return->data;
 }
 
+/*
+ * fmtQualifiedId - construct a schema-qualified name, with quoting as needed.
+ *
+ * Like fmtId, use the result before calling again.
+ *
+ * Since we call fmtId and it also uses getLocalPQExpBuffer() we cannot
+ * use that buffer until we're finished with calling fmtId().
+ *
+ * NB: This assumes setFmtEncoding() previously has been called to configure
+ * the encoding of schema/id. It is preferable to use fmtQualifiedIdEnc()
+ * with an explicit encoding.
+ */
+const char *
+fmtQualifiedId(const char *schema, const char *id)
+{
+	return fmtQualifiedIdEnc(schema, id, getFmtEncoding());
+}
+
 
 /*
  * Format a Postgres version number (in the PG_VERSION_NUM integer format
diff --git a/src/include/fe_utils/string_utils.h b/src/include/fe_utils/string_utils.h
@@ -25,7 +25,10 @@ extern PQExpBuffer (*getLocalPQExpBuffer) (void);
 
 /* Functions */
 extern const char *fmtId(const char *identifier);
+extern const char *fmtIdEnc(const char *identifier, int encoding);
 extern const char *fmtQualifiedId(const char *schema, const char *id);
+extern const char *fmtQualifiedIdEnc(const char *schema, const char *id, int encoding);
+extern void setFmtEncoding(int encoding);
 
 extern char *formatPGVersionNumber(int version_number, bool include_minor,
 								   char *buf, size_t buflen);

Original file line number	Diff line number	Diff line change
`@@ -2725,6 +2725,7 @@ processEncodingEntry(ArchiveHandle AH, TocEntry te)`
`2725`	`2725`	`fatal("unrecognized encoding \"%s\"",`
`2726`	`2726`	`ptr1);`
`2727`	`2727`	`AH->public.encoding = encoding;`
	`2728`	`+ setFmtEncoding(encoding);`
`2728`	`2729`	`}`
`2729`	`2730`	`else`
`2730`	`2731`	`fatal("invalid ENCODING item: %s",`