Remove pattern from database - mysql

I have a MySQL table with over 3 million rows that includes peoples names and the state they are from e.g. "Smith (SC)" I would like to remove the state info and have "Smith". Since there are many difference last names and states used I was wondering if there was any trick to make this easier and just remove the parentheses and the text contained inside of them.
Any advice would be greatly appreciated.

Test with a select to make sure you're getting good results:
SELECT LastName, LEFT(LastName, LOCATE('(', LastName) - 1) AS CorrectedLastName
FROM YourTable
WHERE LOCATE('(', LastName) <> 0
Once you're confident in the results:
UPDATE YourTable
SET LastName = LEFT(LastName, LOCATE('(', LastName) - 1)
WHERE LOCATE('(', LastName) <> 0

Make a backup first, then try
Update <tablename>
set name =substr(name,1, instr(name,'('));
substr is for oracle but check the mysql equivalent

Related

remove whitespace in between Column values in mySQL

How to remove all spaces between a column field?. The spaces occur in the middle of the text so trim won't work and also replace is not working.
my code is
UPDATE temp_emp t1, master_employee t2
SET t1.lm= t2.emp_id
where REPLACE(t1.lm, ' ', '') = REPLACE(CONCAT(t2.first_name,'',t2.last_name), ' ', '');
for example when i run the query ,
select REPLACE(lm, ' ', '') AS concat from temp_emp1
i get the output as follows
concat
----------------------------------------
rick joe
james cole
albert Th
i want the output to be ;like this
concat
----------------------------------------
rickjoe
jamescole
albertTh
Without knowing the table structures and data, it is difficult for me to follow what you are doing. However, to accomplish the ouput of two concatenated columns is very straightforward.
Assume you have a table master_employee with just two columns and you want to output the FIRST and LAST names concatenated with no spaces in between. You simply use the function concat()for MySQL:
SELECT CONCAT(first_name, last_name)
from master_employee;
In Oracle, the concatenation is two pipes (||):
SELECT first_name || last_name
from master_employee;
Hope this helps.
If you want to update the existing column which has multiple spaces into one then this update query will be helpful:
UPDATE your_table SET column_that_you_want_to_change=
REGEXP_REPLACE(column_that_you_want_to_change, '[[:space:]]+', ' ');
If you don't want any spaces then this should work:
UPDATE your_table SET column_that_you_want_to_change=
REGEXP_REPLACE(column_that_you_want_to_change, '[[:space:]]+', '');

Update that splits one column into two?

I'm in a bit of a time crunch. I have a table with 2 columns. firstName, and lastName.
The data imported into SQL had both first and last name inside of the firstName column.
Is there a quick way to update the table to put everything before the first space into the lastName, and everything after the space into the firstName column? I know I could export, and do this via excel, but it in close to production time so I would rather not risk issues.
I have looked at a couple different split posts on here, but those do not involve updating at the same time.
Try to use:
UPDATE
table
SET
lastname = SUBSTRING(firstName, 1, CHARINDEX(' ', firstName) - 1),
firstName= SUBSTRING(firstName, CHARINDEX(' ', firstName) + 1, LEN(firstName))
MySQL
Demonstration:
SET #str := 'Robert Optional Dickey';
SELECT
SUBSTRING_INDEX(#str,SUBSTRING_INDEX(#str,' ',-1),1) AS lastName,
SUBSTRING_INDEX(#str,' ',-1) AS firstName;
Output:
lastName firstName
Robert Optional Dickey
Update Query:
UPDATE your_table
SET lastName = SUBSTRING_INDEX(#str,SUBSTRING_INDEX(#str,' ',-1),1),
firstName = SUBSTRING_INDEX(#str,' ',-1);
Note: It will work for any number spaces inside the full name.It just considers the string after the last space as first name.

Query for first result in a column - substring_index function - MySQL

I can't seem to get the substring_index() to work:
I have created a simple table as follows:
CREATE TABLE ContactList(
cont_id int(11) NOT NULL AUTO_INCREMENT,
last_name varchar(30),
first_name varchar(20),
interests varchar(100),
PRIMARY KEY(cont_id));
I then populated the ContactList table as follows:
INSERT INTO ContactList (last_name, first_name, interests)
VALUES
('Murphy', 'Dave', 'Golf, Pets, Basketball'),
('Murphy', 'Ben', 'Pets, Gym, Basketball'),
('Finn', 'Belinda', 'Pets, Tennis, Knitting'),
('Murphy', 'Steve', 'Pets, Archery, Fishing');
I ran a quick SELECT to ensure the data was entered correctly:
SELECT * FROM ContactList;
Then I ran the following query:
SELECT * FROM ContactList
WHERE last_name = 'Murphy'
AND SUBSTRING_INDEX(interests, ',' ,1) = 'Pets';
I was expecting to get two records back (which I did for Ben & Steve), however, for the 'Interests' column I was assuming I should only get one interest back if it equaled 'pets' (due to the substring_index) however, I got all interests back. How can I use the SUBSTRING_INDEX() to run the query and only get the first interest listed back for each record if it says 'Pets'?
BTW I am using MySQL Version 5.5.24 and I know the Interests would be best suited in their own table - I just want to see why substring_index is not picking the first item from the list if it equals 'pets'.
Thanks for any input,
Andy R ;-)
You're using SUBSTRING_INDEX in the WHERE clause, which determines which rows to include. That's good, but you also need to use it in the SELECT clause, which determines which columns to include.
Try this:
SELECT
last_name,
first_name,
SUBSTRING_INDEX(interests, ',' ,1) AS FirstInterestInList
FROM ContactList
WHERE last_name = 'Murphy'
AND SUBSTRING_INDEX(interests, ',' ,1) = 'Pets';
Although substring_index() will work for the first element, you really want find_in_list():
SELECT last_name, first_name, SUBSTRING_INDEX(interests, ',' ,1) AS FirstInterestInList
FROM ContactList
WHERE last_name = 'Murphy' and
find_in_set('pets', interests) = 1
The advantage of find_inset() is that it will work for arbitrary positions.
Just as a note, though, your delimiter is ', '. For find_in_set() to work best, you should have no space after the column.
Also, if you are doing queries like this, you should fix your data structure. It really wants a table called something like ContactInterests which contains one row for each contact and each interest.

Database - the best way, how to search names

I am solving in these days following situation:
In my DB table I have columns for name and surname.
On my page I have input for searching people, and I am struggling with a problem, how to search the name in database that is stored in two columns and I have the name as string ("Joe Green").
For example, in database I have followings:
Joe New
Joe Blue
Joe Green
Joe Francois Green
What could be the best way, how this problem to solve? I am currently working with MySQL database and Rails 3.
EDIT: Thank you guys for you replies, but I don't know, how to make the query in Rails 3 notation, is it possible to use "concat"?
if your database table engine is myISAM then use FULLTEXT search
first create FULLTEXT index by
CREATE FULLTEXT INDEX fx_name ON pages (name, surname)
then use below query to retrieve required data
SELECT *
FROM table
WHERE MATCH(name,surname) AGAINST ('keyword' IN BOOLEAN MODE)
update
alternative:use CONCAT
first create index
CREATE INDEX users_firstandlast ON users(name, surname);
option1: repeat the CONCAT in your WHERE clause (because AS doesn't create a name you can use in the WHERE clause):
SELECT CONCAT(name, ' ', surname) as fullname
FROM users
WHERE CONCAT(name, ' ', surname) LIKE '%joe green%';
option 2:
SELECT CONCAT(name, ' ', surname) as fullname
FROM users
WHERE name LIKE '%joegreen%' or surname LIKE '%joegreen%';
More information
A lot of this would be significantly simpler if you had given some indication of what language is being used at the interface between the input and the database (since Rails is usually associated with Ruby I'll assume you mean that), however....
SELECT *
FROM yourtable
WHERE CONCAT(name, ' ', surname)
LIKE CONCAT('%',REPLACE(?,' ', '%'),'%')
ORDER BY LEVENSTEIN(CONCAT(name, ' ', surname), ?)
(where '?' is replaced by your search string, and the levenstein function is described here)
Performance will be poor - a better solution would be to split the string Ruby and attempt matches of varying combinations.
One way is:
* split user's input to words
* search each word in name and surname columns
first use the split function to split the string
namearr = name.split
namestr = ""
namearr.each do |i|
namestr = namestr + i
end
then use the values in your query
stmt = "select * from mytable where CONCAT(firstname,surename)= " + namestr
you can use the sql server full text search which i do not recommend because of the performance is very bad ,or use the lucin .net which is more powerful and her are some links :
http://incubator.apache.org/lucene.net/
http://msdn.microsoft.com/en-us/library/ms142571.aspx
mark as answered if it helps :)

mysql: how can i remove the street number from a streetaddress column?

I have a column named streetaddress that contains
<Street Number> <Street Name>
for example:
15 rue Gontier-Patin
4968 Hillcrest Circle
how can i remove the numbers from the beginning of the row purely in sql?
How about something like this - trim off everything up to and including the first space in strings which start with a number
UPDATE mytable
SET addresscol=SUBSTRING(addresscol, LOCATE(' ', addresscol)+1)
WHERE addresscol REGEXP '^[0-9]';
This is based on #Paul Dixon above but is just for returning results without updating the data for things like sorting:
SELECT IF(address REGEXP '^[0-9]', SUBSTRING(address, LOCATE(' ', address)+1), address) AS addrSort FROM table;
MySQL does not have regexp replace functions so it depends on the values in your column. If the string always begins with a number and then a space, you can do it with the SQl query Paul Dixon posted (doh, he was faster than me :D ).
If you need regular expressions to solve it, you have to do it in application code outside the database.
I think this should do the job on the basis that each entry has it's street address seperated from the house number by a space (" ");
UPDATE table
SET streetaddress = MID(streetaddress, LOCATE(' ', streetaddress) + 1);
I've tested this and it works fine.
You can also use the following if you'd like to extract the house number to a new column previous to the update;
UPDATE table
SET housenumber = MID(streetaddress, 1, LOCATE(' ', streetaddress) - 1);