sql 統計每行屬性值的出現次數，區分大小寫

Question

我有表（ notes_subject ）結構 -

+--------------+-------------+------+-----+---------+----------------+
| Field        | Type        | Null | Key | Default | Extra          |
+--------------+-------------+------+-----+---------+----------------+
| id           | int         | NO   | PRI | NULL    | auto_increment |
| user_id      | int         | NO   |     | NULL    |                |
| note_id      | varchar(25) | NO   | MUL | NULL    |                |
| subject_name | text        | NO   |     | NULL    |                |
+--------------+-------------+------+-----+---------+----------------+

和存儲在此表中的數據 -

+----+---------+--------------+--------------+
| id | user_id | note_id      | subject_name |
+----+---------+--------------+--------------+
| 10 |       2 | UdMs870BSswp | CN           |
| 12 |       2 | 8stMvslwIGr2 | CN           |
| 13 |       2 | PB3KNbbFkaUm | cn           |
+----+---------+--------------+--------------+

注：CN 和 cn（小寫）不同。

我想通過此表中的user_id計算每個subject_name的出現次數。 所以我運行查詢 -

SELECT subject_name, COUNT(subject_name) 
FROM notes_subject where user_id=2 GROUP BY subject_name;

它取得了-

+--------------+---------------------+
| subject_name | COUNT(subject_name) |
+--------------+---------------------+
| CN           |                   3 |
+--------------+---------------------+

但這不是正確的結果，因為 CN 和 cn 不同。
我還想要結果中的 id、user_id、note_id。

Answer 1

如果你的數據庫支持window functions ，我想你想要這樣的東西。 嘗試使用collate utf8mb4_bin看看這是否有助於您需要的區分大小寫

select id, 
       user_id, 
       note_id, 
       subject_name, 
       count(subject_name collate utf8mb4_bin) over (partition by user_id, subject_name collate utf8mb4_bin) 
from notes_subject; 

如果window functions了，也可以單獨聚合，再join回主表。

演示

Answer 2

發布另一種選擇，即selecting the group by column as a binary field 。

使用 MYSQL 8.0 版本在dbfiddle中測試

SELECT cast(subject_name as binary) subject_name, COUNT(subject_name)  cnt
  FROM notes_subject 
 WHERE user_id=2 
GROUP BY cast(subject_name as binary)

Answer 3

您可以簡單地使用 Binary 來轉換它。

SELECT count(*), CAST(subject_name as BINARY) AS lastname_cs 
FROM notes_subject where user_id=2 
GROUP BY CAST(subject_name as BINARY);