Question

I have a database where I store keywords grouped into projects and data related to each keyword, then I display datagrids foreach project with one row per keyword and several columns all retrieved from the same table “data”. I have 4 tables, keywords, projects, group_keywords and data. “keywords” only stores the keyword, “projects” the project name, “group_keywords” the keywords ids for the keywords assigned to that project, and “data” is where all the data foreach keyword goes, identified by a foreign key for the keywords.id, and a name column to identify the data name.

Now to retrieve the keywords + all the data for a project I use this query:

SELECT * FROM `group_keywords` 
INNER JOIN keywords on keywords.id = keyword_id 
INNER JOIN data ON data.id = keywords.id 
WHERE `group_id` = (SELECT `id` FROM `projects` WHERE `name` = 'ProjectName'

This gives me something like

    id  group_id    keyword_id  id  keyword     id  name    value
    12  5   52  52  absorption food     52  data_name_x1    6
    12  5   52  52  absorption food     52  data_name_x2    8
    12  5   52  52  absorption food     52  data_name_x3    26
    12  5   52  52  absorption food     52  data_name_x4    2
...

But what I want is to get:

id  group_id    keyword_id  id  keyword id  data_name_x1    data_name_x2    data_name_x3    data_name_x4
12  5   52  52  absorption food     52      6               8               26              2
...

So I can sort and use pagination for the datagrids easly, otherwise I have no idea how to do it, because when using big data sets I can’t just dump everything into an array, too much data.

This is the schema:

-- --------------------------------------------------------
-- Table structure for table `keywords`

CREATE TABLE IF NOT EXISTS `keywords` (
  `id` int(10) unsigned NOT NULL auto_increment,
  `keyword` varchar(255) NOT NULL,
  UNIQUE KEY `id` (`id`),
  UNIQUE KEY `keyword` (`keyword`)
) ENGINE=InnoDB  DEFAULT CHARSET=latin1 AUTO_INCREMENT=115386 ;


-- --------------------------------------------------------
-- Table structure for table `data`

CREATE TABLE IF NOT EXISTS `data` (
  `id` int(10) unsigned NOT NULL,
  `name` varchar(100) NOT NULL,
  `value` varchar(15) NOT NULL,
  UNIQUE KEY `id` (`id`,`name`)
) ENGINE=InnoDB DEFAULT CHARSET=latin1;


-- --------------------------------------------------------
-- Table structure for table `projects`
--

CREATE TABLE IF NOT EXISTS `projects` (
  `id` int(10) NOT NULL auto_increment,
  `name` varchar(100) NOT NULL,
  `parent` varchar(100) default NULL,
  PRIMARY KEY  (`id`)
) ENGINE=InnoDB  DEFAULT CHARSET=latin1 AUTO_INCREMENT=21 ;


-- --------------------------------------------------------
-- Table structure for table `group_keywords`

CREATE TABLE IF NOT EXISTS `group_keywords` (
  `id` int(10) NOT NULL auto_increment,
  `group_id` int(10) NOT NULL,
  `keyword_id` int(10) unsigned NOT NULL,
  PRIMARY KEY  (`id`),
  UNIQUE KEY `group_id` (`group_id`,`keyword_id`),
  KEY `keyword_id` (`keyword_id`)
) ENGINE=InnoDB  DEFAULT CHARSET=latin1 AUTO_INCREMENT=119503 ;


-- --------------------------------------------------------
-- Constraints for table `data`
--
ALTER TABLE `data`
  ADD CONSTRAINT `data_ibfk_1` FOREIGN KEY (`id`) REFERENCES `keywords` (`id`) ON DELETE CASCADE ON UPDATE CASCADE;

-- --------------------------------------------------------
-- Constraints for table `group_keywords`
--
ALTER TABLE `group_keywords`
  ADD CONSTRAINT `group_keywords_ibfk_1` FOREIGN KEY (`keyword_id`) REFERENCES `keywords` (`id`) ON DELETE CASCADE ON UPDATE CASCADE;

Answer 1

This operation is traditionally called “unpivot” and a few RDBMSs support it, but MySQL doesn’t appear to be one of them. You have two options, do it in SQL or do it in PHP. In MySQL it looks something like this with self-joins (I don’t know which field qualifies as an ID field for you, so forgive me creating my own example). From a performance perspective, make sure you index both the ID and Column Name, otherwise these joins will crawl.

shapes
ID  Name   Value
1   Color  Red
1   Shape  Circle
... for more "columns"
2   Color  Green
2   Shape  Square
... for more "columns"

SELECT
  A.ID,
  B.Value as Color,
  C.Value as Shape
  ... for more "columns"
FROM shapes A
LEFT JOIN shapes B ON B.ID = A.ID AND B.Name = 'Color'
LEFT JOIN shapes C ON C.ID = A.ID AND C.Name = 'Shape'
... for more "columns"

Which should net us (unless my head-SQL-parser is misrunning tonight):

ID  Color   Shape
1   Red     Circle
2   Green   Square

For the PHP version, you don’t necessarily have to load up an array, you can stream it. Sort by the PK and walk down it setting the properties. In pseudocode:

Set X to undefined
Get a Record
   Check the ID property, if it's different than X, create a new object, set X to the new ID, and yield the previous object
   Set the property of the object based on the "Name" column of our result

Hope this helps!