Use regular expressions in MySQL SELECT statements
Posted by Quinn McHenry in MySQL
A very cool and powerful capability in MySQL and other databases is the ability to incorporate regular expression syntax when selecting data. The regular expresion support in MySQL is extensive. This recipe reviews regular expression use in MySQL and lists the supported regular expression metacharacters.
The basic syntax to use regular expressions in a MySQL query is:
SELECT something FROM table WHERE column REGEXP 'regexp'
For example, to select all columns from the table events where the values in the column id end with 5587, use:
SELECT * FROM events WHERE id REGEXP '5587$'
A more elaborate example selects all columns of the table reviews where the values in the column description contain the word excellent:
SELECT * FROM reviews WHERE description REGEXP '[[:<:]]excellent[[:>:]]'
MySQL allows the following regular expression metacharacters:
. match any character
? match zero or one
* match zero or more
+ match one or more
{n} match n times
{m,n} match m through n times
{n,} match n or more times
^ beginning of line
$ end of line
[[:<:]] match beginning of words
[[:>:]] match ending of words
[:class:] match a character class
i.e., [:alpha:] for letters
[:space:] for whitespace
[:punct:] for punctuation
[:upper:] for upper case letters
[abc] match one of enclosed chars
[^xyz] match any char not enclosed
| separates alternatives
MySQL interprets a backslash (\) character as an escape character. To use a backslash in a regular expression, you must escape it with another backslash (\\).
About Quinn McHenry
View more articles by Quinn McHenry
The Conversation
Follow the reactions below and share your own thoughts.
March 15, 2009 at 10:56 pm, Lawrence said:
Thank you
July 07, 2009 at 6:05 pm, ////AE said:
Is there a way to use regexp when defining a table to SELECT $something from?
November 13, 2009 at 5:33 am, Anonymous said:
may i know Why this query showing false result
select “hellooo” REGEXP “^h[el]+o$”;
December 11, 2009 at 5:58 pm, Shane said:
santoshb,
i don’t know if you have the answer to your question yet, but your regexp is only look for one ‘o’. The $ sign means the absolute end of the string.
You could either do:
^h[el]+o
or..
^h[el]+ooo$
April 05, 2011 at 4:23 pm, Andy Mason said:
Could also use the ‘+’ operator to match one or more o’s …
select “hellooo” REGEXP “^h[el]+o+$”
January 07, 2010 at 6:33 pm, Wayne said:
I’ve been using mySQL for years, and I’ve been using regular expressions for years. Yet, somehow, I never found out until just now (from this recipe) that mySQL supported them in a manner like this. That’s amazingly useful. Thank you!
(Of course, I also see that you wrote this recipe almost six years ago – so thank you for six years ago!)
Does REGEXP have a bigger or smaller or identical performance hit when compared to using LIKE for substrings?
For instance, here’s a couple pairs of otherwise equivalent queries. Within each pair, which would be better for performance?
SELECT * FROM events WHERE id REGEXP ‘abcd’
SELECT * FROM events WHERE id like ‘%abcd%’
SELECT * FROM events WHERE id REGEXP ‘abcd.*lmnop’
SELECT * FROM events WHERE id like ‘%abcd%lmnop%’
Is LIKE just a specialize synonym for REGEXP which actually uses REGEXP in the background, or is it truly it’s own function? (Asked because if it uses REGEXP, then the performance hits should be identical within both those pairs.)
Anyway, thanks again for adding a weapon to my arsenal that should have been there years ago. I can’t imagine how I’ve overlooked it for so long.
w
January 08, 2010 at 6:29 pm, Anonymous said:
@Wayne
I ran the following queries three times each on my MySQL database which has over 200,000 articles in it:
SELECT * FROM `mos_content` WHERE title REGEXP ‘article’;
SELECT * FROM `mos_content` WHERE title LIKE ‘%article%’;
REGEXP averaged 0.047333333 seconds
LIKE averaged 0.035333333 seconds
There was hardly any deviation between the runs. But it’s obvious that LIKE is about 25% faster than REGEXP for some reason (at least on my specific MySQL configuration).
August 05, 2010 at 10:17 am, Gumuruh S said:
whoooooo… that’s cool!
so anywa, how about we just want to select N until M row only?
March 09, 2011 at 10:25 am, Josef Schalk said:
was geht denn mitr dir
September 29, 2010 at 10:36 pm, GHead said:
How to select from a column that not have character? like in php !preg_match(‘/[a-z]/i’)
February 19, 2011 at 5:25 pm, Gas said:
sepp du depp
February 19, 2011 at 5:30 pm, Sepp-seppelin_constructions said:
like und dann %
seppelinsepp
mcsepp
fußballsepp
keyboardsepp
radfahrersepp
sepp du depp
February 20, 2011 at 11:21 am, Nikn said:
vwpanzersepp
February 20, 2011 at 11:47 am, Ojio said:
ach sepp
February 23, 2011 at 9:00 am, viji said:
hi
March 09, 2011 at 10:24 am, Jo said:
moang
April 26, 2011 at 9:46 am, Gasr said:
sepp
June 04, 2011 at 5:01 am, Hallo said:
sepp du bist der depp
August 06, 2011 at 2:40 pm, Asdfasd said:
I think you use the ^ character to negate stuff:
[^a-z]
would find anything that wasn’t a to z.
July 16, 2012 at 12:40 pm, hai said:
SELECT *
FROM words
WHERE word REGEXP ‘^[able]+$’
this query willl fetch all the words having these letters alone ie ab , able, abel, be.,like tat but it is also fetching bell , label ,bee, etc ie in input there is one l only but it also fetch the word having ll , ee like tat.
Is there any way to remove duplicate letters in a word using regular exp in a query.
Please do reply
October 19, 2012 at 2:25 pm, Ssekirime Geofrey said:
Thank youuuuuuuuu. very much Quinn.
October 22, 2012 at 5:40 am, santosh said:
SELECT ‘ÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉÉ’ REGEXP ‘([ -~])*’
above Query is returning true why and if i want false then what should be my regular Expression