How to combine, paginate, and sort an external API with MySQL result set

If storing the data on your DB is not an option you have to merge and sort within PHP. Here is an example (http://codepad.org/ghrLWrsW), using two unordered arrays - each one describing your data sources: $mysql = array ( array('name' => 'I', 'date' => '20110804'), array('name' => 'J', 'date' => '20110805'), array('name' => 'F', 'date' => '20110801'), array('name' => 'D', 'date' => '20110730'), array('name' => 'B', 'date' => '20110728'), array('name' => 'G', 'date' => '20110802'), ); $eventful = array ( array('name' => 'L', 'date' => '20110807'), array('name' => 'A', 'date' => '20110727'), array('name' => 'E', 'date' => '20110731'), array('name' => 'H', 'date' => '20110803'), array('name' => 'K', 'date' => '20110806'), array('name' => 'C', 'date' => '20110729'), ); You'll want to select an equal amount of maximum events to display on each data source, and merge both arrays into a single multi-dimensional array, which we will then sort using http://php.net/manual/en/function.array-multisort.php: $dates = array(); $result = array_merge($mysql, $eventful); foreach ($result as $key => $value) { $dates[$key] = $value['date']; } array_multisort($dates, SORT_ASC, $result); The $result variable will then be sorted in ascending order: Array ( [0] => Array ( [name] => A [date] => 20110727 ) [1] => Array ( [name] => B [date] => 20110728 ) [2] => Array ( [name] => C [date] => 20110729 ) [3] => Array ( [name] => D [date] => 20110730 ) [4] => Array ( [name] => E [date] => 20110731 ) [5] => Array ( [name] => F [date] => 20110801 ) [6] => Array ( [name] => G [date] => 20110802 ) [7] => Array ( [name] => H [date] => 20110803 ) [8] => Array ( [name] => I [date] => 20110804 ) [9] => Array ( [name] => J [date] => 20110805 ) [10] => Array ( [name] => K [date] => 20110806 ) [11] => Array ( [name] => L [date] => 20110807 ) ) If you need descending order you can use the SORT_DESC constant instead. To paginate this array you can just use http://php.net/manual/en/function.array-chunk.php: $result = array_chunk($result, 4); // 4 events per "page" Which, in turn, produces an array with an added dimension: Array ( [0] => Array ( [0] => Array ( [name] => A [date] => 20110727 ) [1] => Array ( [name] => B [date] => 20110728 ) [2] => Array ( [name] => C [date] => 20110729 ) [3] => Array ( [name] => D [date] => 20110730 ) ) [1] => Array ( [0] => Array ( [name] => E [date] => 20110731 ) [1] => Array ( [name] => F [date] => 20110801 ) [2] => Array ( [name] => G [date] => 20110802 ) [3] => Array ( [name] => H [date] => 20110803 ) ) [2] => Array ( [0] => Array ( [name] => I [date] => 20110804 ) [1] => Array ( [name] => J [date] => 20110805 ) [2] => Array ( [name] => K [date] => 20110806 ) [3] => Array ( [name] => L [date] => 20110807 ) ) ) Bare in mind that this is a lot less efficient than the typical database sort and limit, mostly because you have the load the whole data into memory, even those that you don't need for the active "page". I'm not sure what amount of data you're expecting to process but array_multisort() performs reasonably well up to a couple thousand records (I've actually never used it for anything larger), if you're looking for further optimizations you can destroy the unneeded indexes of the chunked array or, use http://php.net/manual/en/function.array-slice.php and some arithmetic to fetch the four specific records that belong to a certain page. Nonetheless, if you can optimize your selects (maybe with some business / date logic, not sure) you will probably make the whole sorting + chunk (or slice) process more CPU and memory friendly.