Mysql – What’s the difference between POINT(X,Y) and GeomFromText(“POINT(X Y)”)

MySQLperformancespatial

I'd like to store some geometric positions in my MySQL database. For this I use the POINT datatype. Almost everywhere I read that the function GeomFromText should be used to insert data in the table.

However, I found out that POINT(X,Y) also works. I didn't find any description why GeomFromText should be used instead of POINT.

For example I have the following simple relation:

CREATE TABLE Site (
    SiteID      BIGINT UNSIGNED,
    Position    POINT
);

And I can insert values using the following two variants:

INSERT INTO Site (
    1,
    GeomFromText( 'POINT(48.19976 16.45572)' )
);

INSERT INTO Site (
    2,
    POINT(48.19976, 16.45572)
);

When I view the table (SELECT * FROM Site) I see the same binary blob for the location, and when I view the coordinates (SELECT *, AsText(Position) FROM Site) I also see the same values.

So why should GeomFromText be used? Are there any (known) performance differences between these two variants? How is this solved in other database systems than MySQL?

Best Answer

There are two different binary formats related to the MySQL spatial extensions, the "well-known binary" (WKB) format from the standards, and the MySQL internal GEOMETRY data type.

Prior to MySQL 5.1.35, functions like POINT() didn't return the MySQL internal data type; they returned WKB... so prior to then, you had to do this:

INSERT INTO t1 (pt_col) VALUES (GeomFromWKB(Point(1,2)));

But now, as in your example, this works:

INSERT INTO t1 (pt_col) VALUES(Point(1,2));

To the developers' credit, when they changed Point() and similar functions to (more sanely) return GEOMETRY objects, they allowed GeomFromWKB() and similar functions to actually accept either WKB or MySQL Geometry data as input even though the functions are intended to accept WKB as input.

The fact that the 1st method works (in spite of being technically wrong) on newer servers and the 2nd method doesn't work at all prior to MySQL 5.1.35 might explain why examples were written using the approach you've seen -- to avoid the issue entirely. Otherwise... I've got nothing, here.

Concatenating and then parsing text seems intuitively slower and more error-prone than functions that accept proper variables as input, so I can't think of any reason to craft concatenated strings and use the text-based functions.

http://dev.mysql.com/doc/refman/5.1/en/creating-spatial-values.html#gis-wkb-functions

http://dev.mysql.com/doc/relnotes/mysql/5.1/en/news-5-1-35.html

Related Solutions

MySQL – How to Get Next Unique Value Without Auto Increment

What is interesting about this situation can be solved using the MyISAM storage.

I answedred a question like this back in April 2012 : How can you have two auto-incremental columns in one table? You need to create one table whose sole purpose is the create sequences of work order for each site

CREATE TABLE site_workorder_seq
(
    SiteID int not null,
    SiteWorkorderNum int not null auto_increment,
    PRIMARY KEY (SiteID,SiteWorkorderNum)
) ENGINE=MyISAM;

Here is a sample loading into this table:

mysql>     DROP DATABASE david;
Query OK, 1 row affected (0.01 sec)

mysql>     CREATE DATABASE david;
Query OK, 1 row affected (0.00 sec)

mysql>     USE david
Database changed
mysql>     CREATE TABLE site_workorder_seq
    ->     (
    ->         SiteID int not null,
    ->         SiteWorkorderNum int not null auto_increment,
    ->         PRIMARY KEY (SiteID,SiteWorkorderNum)
    ->     ) ENGINE=MyISAM;
Query OK, 0 rows affected (0.01 sec)

mysql>     INSERT INTO site_workorder_seq (SiteID) VALUES
    ->     (1),(1),(2),(3),(3),(3),(3),(4),(4),(4),
    ->     (5),(5),(4),(2),(2),(2);
Query OK, 16 rows affected (0.00 sec)
Records: 16  Duplicates: 0  Warnings: 0

mysql>     SELECT * FROM site_workorder_seq;
+--------+------------------+
| SiteID | SiteWorkorderNum |
+--------+------------------+
|      1 |                1 |
|      1 |                2 |
|      2 |                1 |
|      2 |                2 |
|      2 |                3 |
|      2 |                4 |
|      3 |                1 |
|      3 |                2 |
|      3 |                3 |
|      3 |                4 |
|      4 |                1 |
|      4 |                2 |
|      4 |                3 |
|      4 |                4 |
|      5 |                1 |
|      5 |                2 |
+--------+------------------+
16 rows in set (0.00 sec)

mysql>

Let's look at the the last WorkorderNum from each site

mysql> SELECT SiteID,MAX(SiteWorkorderNum) SiteWorkorderNum
    -> FROM site_workorder_seq GROUP BY SiteID;
+--------+------------------+
| SiteID | SiteWorkorderNum |
+--------+------------------+
|      1 |                2 |
|      2 |                4 |
|      3 |                4 |
|      4 |                4 |
|      5 |                2 |
+--------+------------------+
5 rows in set (0.05 sec)

mysql>

Now, suppose you want to get the next SiteWorkorderNum for SiteID 3. You could do this:

INSERT INTO site_workorder_seq (SiteID) VALUES (3);
SELECT MAX(SiteWorkorderNum) INTO @nextworkordernum
FROM site_workorder_seq WHERE SiteID=3;
SELECT @nextworkordernum;

Let's run this and see what happens

mysql> INSERT INTO site_workorder_seq (SiteID) VALUES (3);
Query OK, 1 row affected (0.00 sec)

mysql> SELECT MAX(SiteWorkorderNum) INTO @nextworkordernum
    -> FROM site_workorder_seq WHERE SiteID=3;
Query OK, 1 row affected (0.00 sec)

mysql> SELECT @nextworkordernum;
+-------------------+
| @nextworkordernum |
+-------------------+
|                 5 |
+-------------------+
1 row in set (0.03 sec)

mysql> SELECT * FROM site_workorder_seq;
+--------+------------------+
| SiteID | SiteWorkorderNum |
+--------+------------------+
|      1 |                1 |
|      1 |                2 |
|      2 |                1 |
|      2 |                2 |
|      2 |                3 |
|      2 |                4 |
|      3 |                1 |
|      3 |                2 |
|      3 |                3 |
|      3 |                4 |
|      3 |                5 |
|      4 |                1 |
|      4 |                2 |
|      4 |                3 |
|      4 |                4 |
|      5 |                1 |
|      5 |                2 |
+--------+------------------+
17 rows in set (0.00 sec)

mysql> SELECT SiteID,MAX(SiteWorkorderNum) SiteWorkorderNum
    -> FROM site_workorder_seq GROUP BY SiteID;
+--------+------------------+
| SiteID | SiteWorkorderNum |
+--------+------------------+
|      1 |                2 |
|      2 |                4 |
|      3 |                5 |
|      4 |                4 |
|      5 |                2 |
+--------+------------------+
5 rows in set (0.00 sec)

mysql>

As long as you use this one MyISAM apart from all InnoDB tables, you can generate workordernums per site to your hearts content.

Mysql – Get data from two tables with recursive relationships

BAD NEWS

MySQL does not support recursive data or table structures. Neither does it support recursive SQL.

GOOD NEWS

There is a hope. I have written some stored functions on how to retrieve hierarchies of data

Oct 24, 2011 : Find highest level of a hierarchical field: with vs without CTEs
Dec 10, 2012 : MySQL: Tree-Hierarchical query

CAVEAT

You would have to implement this against your tables.
You also need to make sure you are entering data as a Tree and not as a Graph.
For any row in the position_inherits table that does not have a parent, simply set that row's inherit_privs_from to zero(0). That way, you have establish root positions.