UPDATE statement with a cross join in mysql - mysql

Question

I have a table which has a column which contains a text blob. I want to update some of the values in the text fields by using a mapping table, but I'm not sure if there is a way to do it without cursors. Here is an example:
USE Temp;
CREATE TABLE `Temp`.`test_text` (
`some_text` text
) ENGINE=InnoDB DEFAULT CHARSET=latin1 COMMENT=' ';
CREATE TABLE `Temp`.`mapping` (
`src` VARCHAR(1024) NULL,
`dest` VARCHAR(1024) NULL) ENGINE=InnoDB DEFAULT CHARSET=latin1 COMMENT=' ';
-- test_text has a single column with some_text
INSERT INTO `Temp`.`test_text`
(`some_text`)
VALUES
('There once was a man named BobFrank. He changed his name to BobDude.');
-- path_mapping has two columns, which contain all the mappings I'd like to do
INSERT INTO `Temp`.`mapping`
(`src`,
`dest`)
VALUES
('BobFrank', 'BobsNewFrank'),
('BobDude', 'BobsNewDude');
UPDATE `Temp`.`test_text` tt, `Temp`.`mapping` mp
SET tt.some_text = REPLACE(tt.some_text, mp.src, mp.dest);
SELECT *
FROM `Temp`.`test_text`
Result:
There once was a man named BobsNewFrank. He changed his name to BobDude."
The code above only seems to replace BobFrank in the output.
What I would like to see is the value in test_text to be after the update:
There once was a man named BobsNewFrank. He changed his name to BobsNewDude.

Answer 1

Your current table structure is less than ideal, because the test_text table has unnormalized CSV data in it. This will preclude using most of the database operations which we would think to use here. I was able to come up with a solution, but it required storing each CSV term in test_text in a separate row. That is, I used the following table:
CREATE TABLE test_text (some_text varchar(55));
INSERT INTO test_text (some_text)
VALUES
('"/tmp/BobFrank"'),
('"/tmp/BobDude/"');
I left the path_mapping identical to what you had. Then, I only needed a fairly simple query:
SELECT
GROUP_CONCAT(REPLACE(t1.some_text, t2.src, t2.dest)) AS output
FROM test_text t1
INNER JOIN path_mapping t2
ON t1.some_text LIKE CONCAT('%', t2.src, '%');
This generated the following output:
Demo