ASCII：表达式只能包含 ASCII 字符；也就是说，字符范围在 Unicode 范围 U+0000 到 U+007F 之间。ASCII: The expression can contain only ASCII characters; that is, characters in the Unicode range U+0000 to U+007F.

UNICODE：表达式可以包含 Unicode 范围内的字符 U+0000 到 U+10FFFF。这包括基本多语言平面（BMP）范围内的字符 (U+0000 到 U+FFFF) 以及 BMP 范围外的补充字符 (U+10000 到 U+10FFFF)。UNICODE: The expression can contain characters in the Unicode range U+0000 to U+10FFFF. This includes characters in the Basic Multilingual Plane (BMP) range (U+0000 to U+FFFF) and supplementary characters outside the BMP range (U+10000 to U+10FFFF).

字符串常量的表达集取决于字符串内容，并可能与字符串字符集的表达集不同。请考虑以下语句：The repertoire for a string constant depends on string content and may differ from the repertoire of the string character set. Consider these statements:

SET NAMES utf8mb4; SELECT 'abc';
SELECT _utf8mb4'def';

虽然在前面的每种情况下字符集都是utf8mb4，但这些字符串实际上并不包含任何超出 ASCII 范围的字符，因此它们的字符集是ASCII而不是UNICODE。Although the character set is utf8mb4 in each of the preceding cases, the strings do not actually contain any characters outside the ASCII range, so their repertoire is ASCII rather than UNICODE.

一个具有ascii字符集的列因为其字符集而具有ASCII字符集。在下表中，c1具有ASCII字符集：A column having the ascii character set has ASCII repertoire because of its character set. In the following table, c1 has ASCII repertoire:

CREATE TABLE t1 (c1 CHAR(1) CHARACTER SET ascii);

以下示例说明了在发生错误时，如果没有储备如何确定结果：The following example illustrates how repertoire enables a result to be determined in a case where an error occurs without repertoire:

CREATE TABLE t1 (
  c1 CHAR(1) CHARACTER SET latin1,
  c2 CHAR(1) CHARACTER SET ascii
);
INSERT INTO t1 VALUES ('a','b');
SELECT CONCAT(c1,c2) FROM t1;

没有储备时，会发生此错误：Without repertoire, this error occurs:

ERROR 1267 (HY000): Illegal mix of collations (latin1_swedish_ci,IMPLICIT)
and (ascii_general_ci,IMPLICIT) for operation 'concat'

使用字符集，从子集到超集（ascii 到 latin1）可以进行转换并返回结果：Using repertoire, subset to superset (ascii to latin1) conversion can occur and a result is returned:

+---------------+
| CONCAT(c1,c2) |
+---------------+
| ab            |
+---------------+

带有一个字符串参数的函数继承其参数的字符集。UPPER(_utf8mb4'abc') 的结果具有 ASCII 字符集，因为其参数具有 ASCII 字符集。（尽管有 _utf8mb4 前缀，字符串 'abc' 中不包含 ASCII 范围之外的字符。）Functions with one string argument inherit the repertoire of their argument. The result of UPPER(_utf8mb4'abc') has ASCII repertoire because its argument has ASCII repertoire. (Despite the _utf8mb4 introducer, the string 'abc' contains no characters outside the ASCII range.)

对于返回字符串但没有字符串参数且使用 character_set_connection 作为结果字符集的函数，如果 character_set_connection 是 ascii，则结果字符集为 ASCII，否则为 UNICODE：For functions that return a string but do not have string arguments and use character_set_connection as the result character set, the result repertoire is ASCII if character_set_connection is ascii, and UNICODE otherwise:

FORMAT(numeric_column, 4);

使用曲目会改变 MySQL 对以下示例的评估方式：Use of repertoire changes how MySQL evaluates the following example:

SET NAMES ascii;
CREATE TABLE t1 (a INT, b VARCHAR(10) CHARACTER SET latin1);
INSERT INTO t1 VALUES (1,'b');
SELECT CONCAT(FORMAT(a, 4), b) FROM t1;

没有储备时，会发生此错误：Without repertoire, this error occurs:

ERROR 1267 (HY000): Illegal mix of collations (ascii_general_ci,COERCIBLE)
and (latin1_swedish_ci,IMPLICIT) for operation 'concat'

使用曲目时，将返回一个结果：With repertoire, a result is returned:

+-------------------------+
| CONCAT(FORMAT(a, 4), b) |
+-------------------------+
| 1.0000b                 |
+-------------------------+

具有两个或更多字符串参数的函数使用结果字符集的“最宽”参数集，其中UNICODE比ASCII更宽。考虑以下CONCAT()调用：Functions with two or more string arguments use the “widest” argument repertoire for the result repertoire, where UNICODE is wider than ASCII. Consider the following CONCAT() calls:

CONCAT(_ucs2 X'0041', _ucs2 X'0042')
CONCAT(_ucs2 X'0041', _ucs2 X'00C2')

对于第一次调用，字符集是ASCII，因为两个参数都在 ASCII 范围内。对于第二次调用，字符集是UNICODE，因为第二个参数在 ASCII 范围之外。For the first call, the repertoire is ASCII because both arguments are within the ASCII range. For the second call, the repertoire is UNICODE because the second argument is outside the ASCII range.

函数返回值的字符集和校对规则取决于仅影响结果字符集和校对规则的参数集合。The repertoire for function return values is determined based on the repertoire of only those arguments that affect the result's character set and collation.

IF(column1 < column2, 'smaller', 'greater')

结果编码是 ASCII，因为两个字符串参数（第二个参数和第三个参数）都使用 ASCII 编码。第一个参数对于结果编码没有影响，即使表达式使用了字符串值。The result repertoire is ASCII because the two string arguments (the second argument and the third argument) both have ASCII repertoire. The first argument does not matter for the result repertoire, even if the expression uses string values.