首先mysql更新資料的某個欄位，一般這樣寫：

UPDATE mytable SET myfield = 'value' WHERE other_field = 'other_value';

也可以這樣用in指定要更新的記錄：

UPDATE mytable SET myfield = 'value' WHERE other_field in ('other_values');

這裡注意 ‘other_values' 是一個逗號（，）分隔的字串，如：1,2,3

如果更新多條資料而且每條記錄要更新的值不同，可能很多人會這樣寫：

foreach ($values as $id => $myvalue) {
 $sql = "UPDATE mytable SET myfield = $myvalue WHERE id = $id";
 mysql_query($sql);
}
 

即是迴圈一條一條的更新記錄。一條記錄update一次，這樣效能很差，也很容易造成阻塞。

那麼能不能一條sql語句實現批量更新呢？

mysql並沒有提供直接的方法來實現批量更新，但是可以用點小技巧來實現。

UPDATE mytable
 SET myfield = CASE id
 WHEN 1 THEN 'myvalue1'
 WHEN 2 THEN 'myvalue2'
 WHEN 3 THEN 'myvalue3'
 END
WHERE other_field ('other_values')

如果where條件查詢出記錄的id不在CASE範圍內，myfield將被設定為空。

如果更新多個值的話，只需要稍加修改：

UPDATE mytable
 SET myfield1 = CASE id
 WHEN 1 THEN 'myvalue11'
 WHEN 2 THEN 'myvalue12'
 WHEN 3 THEN 'myvalue13'
 END,
 myfield2 = CASE id
 WHEN 1 THEN 'myvalue21'
 WHEN 2 THEN 'myvalue22'
 WHEN 3 THEN 'myvalue23'
 END
WHERE id IN (1,2,3)

這裡以php為例，構造這兩條mysql語句：

1、更新多條單個欄位為不同值, mysql模式

$ids_values = array(
 1 => 11,
 2 => 22,
 3 => 33,
 4 => 44,
 5 => 55,
 6 => 66,
 7 => 77,
 8 => 88,
);
 
$ids = implode(',', array_keys($ids_values ));
$sql = "UPDATE mytable SET myfield = CASE id ";
foreach ($ids_values as $id=> $myvalue) {
 $sql .= sprintf("WHEN %d THEN %d ", $id, $myvalue);
}
$sql .= "END WHERE id IN ($ids)";
echo $sql.";<br/>";

輸出：

UPDATE mytable SET myfield = CASE id WHEN 1 THEN 11 WHEN 2 THEN 22 WHEN 3 THEN 33 WHEN 4 THEN 44 WHEN 5 THEN 55 WHEN 6 THEN 66 WHEN 7 THEN 77 WHEN 8 THEN 88 END WHERE id IN (1,2,3,4,5,6,7,8);

2、更新多個欄位為不同值, PDO模式

$data = array(array('id' => 1, 'myfield1val' => 11, 'myfield2val' => 111), array('id' => 2, 'myfield1val' => 22, 'myfield2val' => 222));
$where_in_ids = implode(',', array_map(function($v) {return ":id_" . $v['id'];}, $data));
$update_sql = 'UPDATE mytable SET';
$params = array();

$update_sql .= ' myfield1 = CASE id';
foreach($data as $key => $item) {
 $update_sql .= " WHEN :id_" . $key . " THEN :myfield1val_" . $key . " ";
 $params[":id_" . $key] = $item['id'];
 $params[":myfield1val_" . $key] = $item['myfield1val'];
}
$update_sql .= " END";

$update_sql .= ',myfield2 = CASE id';
foreach($data as $key => $item) {
 $update_sql .= " WHEN :id_" . $key . " THEN :myfield2val_" . $key . " ";
 $params[":id_" . $key] = $item['id'];
 $params[":myfield1va2_" . $key] = $item['myfield2val'];
}
$update_sql .= " END";

$update_sql .= " WHERE id IN (" . $where_in_ids . ")";
echo $update_sql.";<br/>";
var_dump($params);

輸出：

UPDATE mytable SET myfield1 = CASE id WHEN :id_0 THEN :myfield1val_0 WHEN :id_1 THEN :myfield1val_1 END,myfield2 = CASE id WHEN :id_0 THEN :myfield2val_0 WHEN :id_1 THEN :myfield2val_1 END WHERE id IN (:id_1,:id_2);

array (size=6)
 ':id_0' => int 1
 ':myfield1val_0' => int 11
 ':id_1' => int 2
 ':myfield1val_1' => int 22
 ':myfield1va2_0' => int 111
 ':myfield1va2_1' => int 222

另外三種批量更新方式

1. replace into 批量更新

replace into mytable(id, myfield) values (1,'value1'),(2,'value2'),(3,'value3');

2. insert into ...on duplicate key update批量更新

insert into mytable(id, myfield1, myfield2) values (1,'value11','value21'),(2,'value12','value22'),(3,'value13','value23') on duplicate key update myfield1=values(myfield2),values(myfield2)+values(id);

3. 臨時表

DROP TABLE IF EXISTS `tmptable`;
create temporary table tmptable(id int(4) primary key,myfield varchar(50));
insert into tmptable values (1,'value1'),(2,'value2'),(3,'value3');
update mytable, tmptable set mytable.myfield = tmptable.myfield where mytable.id = tmptable.id;

1. 【replace into】和【insert into】更新都依賴於主鍵或唯一值，並都可能造成新增記錄的操作的結構隱患
   
2. 【replace into】操作本質是對重複記錄先delete然後insert，如果更新的欄位不全缺失的欄位將被設定成預設值
   
3. 【insert into】則只是update重複的記錄，更改的欄位只能依循公式值
   
4. 【臨時表】方式需要使用者有temporary 表的create 許可權
   
5.  數量較少時【replace into】和【insert into】效能最好，數量大時【臨時表】最好，【CASE】則具有通用型也不具結構隱患 、

Solution 1:  修改1列

update student s, city c
set s.city_name = c.name
where s.city_code = c.code;

Solution 2:  修改多個列

update  a,  b

set a.title=b.title, a.name=b.name
where a.id=b.id

Solution 3: 採用子查詢

update student s set city_name = (select name from city where code = s.city_code);