您可以使用递归CTE（对于这样一个简单的任务，这有点复杂）来完成此操作：

-- Create and populate example table:
CREATE TABLE "testwords" ("word" text);
INSERT INTO "testwords"("word") VALUES ('Hello'), ('world!'), ('foo bar baz');

-- CTE to select the reversed characters from the "testwords" table:
WITH RECURSIVE "reverse" ("stringno", "string", "charno", "char") AS (

    SELECT -- start at the last character:
        _rowid_
    ,   "word"
    ,   length("word")
    ,   substr("word", -1, 1)
    FROM "testwords"
    
    UNION ALL

    SELECT -- recurse: move one letter to the left
        "stringno"
    ,   "string"
    ,   "charno" - 1
    ,   substr("string", "charno" - 1, 1)
    FROM "reverse"
    WHERE "charno" > 1 -- stop condition
)
SELECT
    "string"
,   group_concat("char", '') AS "reversed"
FROM "reverse"
GROUP BY "stringno";

递归CTE选择每个字符串的每个字符（从右到左），并且

• stringno是适合于对每个字符串的字符进行分组的数字
• string是要反转的字符串
• charno是此字符的位置（从length()向下计数到1）
• char是字符本身

最后，使用group_concat()重新组合各个字符，得到所需的结果：

| Hello       | olleH       |
| world!      | !dlrow      |
| foo bar baz | zab rab oof |