[What sort algorithm is used to discuss the C implementation within the function of the]php sort series?]

Ext/standard/php_array.h Https://github.com/php/php-src/blob/master/ext/standard/php_array.h #ifndef Php_array_h #define Php_array_h Php_minit_function (array); Php_mshutdown_function (array); Php_function (Ksort); Php_function (Krsort); Php_function (Natsort); Php_function (Natcasesort); Php_function (Asort); Php_function (Arsort); Php_function (sort); Php_function (Rsort); Php_function (Usort); Php_function (Uasort); Php_function (Uksort); ...... Copy Code The sort function defined above: Arsort--Reverse-sort the array and keep the index relationship Asort--sort the array and keep the index relationship Krsort--the array is reversed by the key name Ksort--Sorts the array by key name Natcasesort--using the "natural sort" algorithm to Sort Natsort of case-sensitive letters--sort by the "natural sort" algorithm for arrays Rsort--reverse sorting of arrays sort--sort an arrayUasort--Use a user-defined comparison function to sort the values in the array and keep the index associated uksort--sorts the key names in the array using a user-defined comparison function Usort--sorts the values in the array using a user-defined comparison function for simplicity, only the sort function is parsed : https://github.com/php/php-src/blob/master/ext/standard/array.c /* {{proto bool sort (array &array_arg [, int sort_flags]) Sort an array */ Php_function (sort) { Zval *array; Long sort_type = Php_sort_regular; if (Zend_parse_parameters (Zend_num_args () tsrmls_cc, "A|l", &array, &sort_type) = = FAILURE) { Return_false; } Php_set_compare_func (Sort_type tsrmls_cc); if (Zend_hash_sort (z_arrval_p (array), Zend_qsort, Php_array_data_compare, 1 tsrmls_cc) = = FAILURE) { Return_false; } Return_true; } /* }}} */ Copy Code In the code, you see the if (Zend_hash_sort (z_arrval_p (array), Zend_qsort, Php_array_data_compare, ...). The possibility of using a quick sort is large. Continue the analysis. Zend/zend_hash.c Https://github.com/php/php-src/blob/master/Zend/zend_hash.c (*sort_func) (void *) artmp, I, sizeof (Bucket *), compar tsrmls_cc); Handle_block_interruptions (); Ht->plisthead = artmp[0]; Ht->plisttail = NULL; Ht->pinternalpointer = ht->plisthead; Artmp[0]->plistlast = NULL; if (i > 1) { Artmp[0]->plistnext = artmp[1]; for (j = 1; j < I-1; J + +) { Artmp[j]->plistlast = Artmp[j-1]; Artmp[j]->plistnext = artmp[j+1]; } Artmp[j]->plistlast = Artmp[j-1]; Artmp[j]->plistnext = NULL; } else { Artmp[0]->plistnext = NULL; } Ht->plisttail = Artmp[i-1]; Pefree (artmp, ht->persistent); Handle_unblock_interruptions (); if (renumber) { p = ht->plisthead; i=0; while (P! = NULL) { p->nkeylength = 0; P->h = i++; p = p->plistnext; } Ht->nnextfreeelement = i; Zend_hash_rehash (HT); } Copy Code In the algorithm, faster than the fast sort is no doubt the Cardinal sort, a cursory look at the algorithm, may be the basis of sorting in the hash bucket . Bucket sequencing is a stable bucket sort is the fastest one in a common sort, faster than a fast line ... In most cases, the buckets are sorted very quickly, but at the same time they are very space-consuming (change time with space）

Used, which, PHP

