System Functions

GaussDB is compatible with most MySQL system functions, but there are some differences. Only system functions in M-compatible mode can be used. System functions of the original GaussDB cannot be used in case of unexpected results. Currently, some system functions in GaussDB with the same names as those in MySQL are not supported in M-compatible mode. For some of them, the message indicating that they are not supported in M-compatible mode is displayed. Other functions still retain the behaviors of the original GaussDB system functions. Do not use these functions in case of unexpected results. The following table lists the functions with the same name.

**Table 1** Same-name functions for which a message indicating that they are not supported in M-compatible mode is displayed
cot	isEmpty	json_append	json_array	json_array_append
json_array_insert	json_contains	json_contains_path	json_depth	json_extract
json_insert	json_keys	json_length	json_merge	json_merge_patch
json_merge_preserve	json_object	json_quote	json_remove	json_valid
json_replace	json_search	json_set	json_type	json_unquote
last_insert_id	md5	mod	octet_length	overlaps
point	radians	stddev_pop	stddev_samp	var_pop
var_samp	variance	lag	lead	-

**Table 2** Same-name functions that retain the behaviors of the original GaussDB system functions in M-compatible mode
ceil	decode	encode	format	instr
position	round	stddev	regexp_instr	regexp_like
regexp_replace	regexp_substr	row_num	-	-

In M-compatible mode, system functions have the following common differences:

The return value type of a system function is the same as that of MySQL only when the node type of the input parameter is Var (table data) or Const (constant input). In other cases (for example, the input parameter is a calculation expression or function expression), the return value type may be different from that of MySQL.
When an aggregate function uses an expression such as another function, operator, or SELECT clause as the input parameter (for example, SELECT sum(abs(n)) FROM t;), the aggregate function cannot obtain the precision transferred by the input parameter expression. As a result, the precision of the function result is different from that of MySQL.