给出下表（位置）
+--------------------+---------------------+------+-----+---------+----------------+
| Field | Type | Null | Key | Default | Extra |
+--------------------+---------------------+------+-----+---------+----------------+
| position_id | bigint(20) unsigned | NO | PRI | NULL | auto_increment |
| vehicle_id | int(11) unsigned | YES | MUL | NULL | |
| location | point | NO | | NULL | |
| gps_time | datetime | NO | | NULL | |
+--------------------+---------------------+------+-----+---------+----------------+
我想从位置列的点构造一个 lineString 来计算距离。
如何从位置列构造 lineString ？就像是
select vehicle_id, lineString(location) from position group by vehicle_id order by gps_time;
这是数据
select position_id,vehicle_id,astext(location),gps_time from position order by gps_time;
+-------------+------------+-----------------------------+---------------------+
| position_id | vehicle_id | astext(location) | gps_time |
+-------------+------------+-----------------------------+---------------------+
| 3 | 1 | POINT(44.23912 15.307647) | 2014-11-23 06:02:17 |
| 2 | 1 | POINT(44.240063 15.307788) | 2014-11-23 08:54:11 |
| 1 | 1 | POINT(44.23884 15.307585) | 2014-11-23 11:50:10 |
| 4 | 1 | POINT(44.232296 15.304604) | 2014-11-24 07:24:29 |
| 5 | 1 | POINT(44.2332 15.304803) | 2014-11-24 07:24:39 |
| 6 | 1 | POINT(44.2332 15.304803) | 2014-11-24 07:24:49 |
| 7 | 1 | POINT(44.23428 15.3052) | 2014-11-24 07:24:59 |
+-------------+------------+-----------------------------+---------------------+
我想计算特定 gps_time 期间点之间的距离。即 position_id=3 的点和 position_id=2 的点之间的距离，那么结果将用 position_id=2 的点和 position_id=1 的点之间的距离计算，以此类推。
预期结果如：
+------------+------------- +--------------------+---------------------+
| vehicle_id | distance(km) | start_time | end_time |
+------------+------------- +--------------------+---------------------+
| 1 | 1.2 |2014-11-23 06:02:17 | 2014-11-24 07:24:59 |
+------------+------------- +--------------------+---------------------+
距离值（1.2）只是我心中的一个随机值。它应该从以前的点计算。