note 95176 added to function.array-merge

I am in the middle of recoding a very complex CMS I wrote in the past, and I often used array_merge() to determine default values (example: paths, keys which refer to objects elsewhere, etc). I was trying to decide if I should use array_merge or individual calls to isset in combination with an $opts array, to supply default values in functions. For example, here would be two simple examples: <?php function foo( array $opts = array() ) { if ( isset($opts['val1']) ) $opts['val1'] = 'value'; if ( isset($opts['val2']) ) $opts['val2'] = 'value'; if ( isset($opts['val3']) ) $opts['val3'] = 'value'; # ... continue coding } function bar( array $opts = array() ) { $opts = array_merge( array( 'val1' => 'value', 'val2' => 'value', 'val3' => 'value', ), $opts); # ... continue coding } ?> Hypothesis: isset() will be dramatically faster in both circumstances, as benchmarks I have performed in the past usually demonstrate that array iterations are typically slow and should be avoided when and where possible. (ex: using strrpos and substr rather than explode and array_pop to get a file name from a path, of course you can just use basename() but this is just for explanation purposes) --- I performed numerous benchmarks using loops, and then hard-coded examples and got fairly similar results. My four focus points however, were supplying the values in the $opts array, and not supplying the values, with both methods. Calls to isset() WITH the values supplied was ALWAYS the fastest (which makes sense, as no assignments are made) Calls to isset() WITH the values supplied, and array_merge() with NO values set came very close in times, and fluctuated which was faster or slower... (This surprised me as I expected that working with array_merge would simply take more time; But also makes sense as I think about it more as you are still utilizing an array and array_merge is probably making isset() calls in itself) Calls to array_merge() with NO values supplied was always the SLOWEST. (which also makes sense, because all values are replaced) --- Conclusion: Calls to isset() are faster than using array_merge(); But not by much. However, the more values supplied when calling the function determine how much closer the execution time will take. Therefor, with this type of functionality in mind, supplying DEFAULT values mean often NO, SOME, or ALL values will be included when calling the function... Because I am worried about the execution time over easy / less code, I'll stick with the isset() method of supplying default values to an array. --- Now that the test results are in, I thought I would share as I spent quite a bit of time testing and going over results. ---- Server IP: 64.71.164.2 Probable Submitter: 76.167.60.11 ---- Manual Page -- http://www.php.net/manual/en/function.array-merge.php Edit -- https://master.php.net/note/edit/95176 Del: integrated -- https://master.php.net/note/delete/95176/integrated Del: useless -- https://master.php.net/note/delete/95176/useless Del: bad code -- https://master.php.net/note/delete/95176/bad+code Del: spam -- https://master.php.net/note/delete/95176/spam Del: non-english -- https://master.php.net/note/delete/95176/non-english Del: in docs -- https://master.php.net/note/delete/95176/in+docs Del: other reasons-- https://master.php.net/note/delete/95176 Reject -- https://master.php.net/note/reject/95176 Search -- https://master.php.net/manage/user-notes.php