Skip to content

Optimize array_slice for packed/unpacked arrays with large offsets #4860

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
wants to merge 2 commits into from
Closed

Optimize array_slice for packed/unpacked arrays with large offsets #4860

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
c6c0eb8
Optimize array_slice for packed arrays with large offsets
TysonAndre Oct 25, 2019
8ad8a8c
Use /* instead of /**
TysonAndre Oct 26, 2019
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
110 changes: 73 additions & 37 deletions ext/standard/array.c
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -3455,6 +3455,32 @@ PHP_FUNCTION(array_splice)
}
/* }}} */

/* {{{ find_bucket_at_offset(HashTable* ht, zend_long offset)
Finds the bucket at the given valid offset */
static inline Bucket* find_bucket_at_offset(HashTable* ht, zend_long offset)
{
zend_long pos;
Bucket *bucket;
ZEND_ASSERT(offset >= 0 && offset <= ht->nNumOfElements);
if (HT_IS_WITHOUT_HOLES(ht)) {
/* There's no need to iterate over the array to filter out holes if there are no holes */
/* This properly handles both packed and unpacked arrays. */
return ht->arData + offset;
}
/* Otherwise, this code has to iterate over the HashTable and skip holes in the array. */
pos = 0;
ZEND_HASH_FOREACH_BUCKET(ht, bucket) {
if (pos >= offset) {
/* This is the bucket of the array element at the requested offset */
return bucket;
}
++pos;
} ZEND_HASH_FOREACH_END();

/* Return a pointer to the end of the bucket array. */
return ht->arData + ht->nNumUsed;
}

/* {{{ proto array array_slice(array input, int offset [, int length [, bool preserve_keys]])
Returns elements specified by offset and length */
PHP_FUNCTION(array_slice)
Expand All @@ -3466,7 +3492,6 @@ PHP_FUNCTION(array_slice)
zend_bool length_is_null = 1; /* Whether an explicit length has been omitted */
zend_bool preserve_keys = 0; /* Whether to preserve keys while copying to the new array */
uint32_t num_in; /* Number of elements in the input array */
uint32_t pos; /* Current position in the array */
zend_string *string_key;
zend_ulong num_key;

Expand Down Expand Up @@ -3507,50 +3532,61 @@ PHP_FUNCTION(array_slice)
/* Initialize returned array */
array_init_size(return_value, (uint32_t)length);

/* Start at the beginning and go until we hit offset */
pos = 0;
if (HT_IS_PACKED(Z_ARRVAL_P(input)) &&
(!preserve_keys ||
(offset == 0 && HT_IS_WITHOUT_HOLES(Z_ARRVAL_P(input))))) {
zend_hash_real_init_packed(Z_ARRVAL_P(return_value));
ZEND_HASH_FILL_PACKED(Z_ARRVAL_P(return_value)) {
ZEND_HASH_FOREACH_VAL(Z_ARRVAL_P(input), entry) {
pos++;
if (pos <= offset) {
// Contains modified variants of ZEND_HASH_FOREACH_VAL
{
HashTable *ht = Z_ARRVAL_P(input);
Bucket *p = find_bucket_at_offset(ht, offset);
Bucket *end = ht->arData + ht->nNumUsed;

/* Start at the beginning and go until we hit offset */
if (HT_IS_PACKED(Z_ARRVAL_P(input)) &&
(!preserve_keys ||
(offset == 0 && HT_IS_WITHOUT_HOLES(Z_ARRVAL_P(input))))) {

zend_hash_real_init_packed(Z_ARRVAL_P(return_value));
ZEND_HASH_FILL_PACKED(Z_ARRVAL_P(return_value)) {
for (; p != end; p++) {
if (__fill_idx >= length) {
break;
}
entry = &p->val;
if (UNEXPECTED(Z_TYPE_P(entry) == IS_UNDEF)) {
continue;
}
if (UNEXPECTED(Z_ISREF_P(entry)) &&
UNEXPECTED(Z_REFCOUNT_P(entry) == 1)) {
entry = Z_REFVAL_P(entry);
}
Z_TRY_ADDREF_P(entry);
ZEND_HASH_FILL_ADD(entry);
}
} ZEND_HASH_FILL_END();
} else {
zend_long n = 0; /* Current number of elements */
for (; p != end; p++) {
entry = &p->val;
if (UNEXPECTED(Z_TYPE_P(entry) == IS_UNDEF)) {
continue;
}
if (pos > offset + length) {
if (n >= length) {
break;
}
if (UNEXPECTED(Z_ISREF_P(entry)) &&
UNEXPECTED(Z_REFCOUNT_P(entry) == 1)) {
entry = Z_REFVAL_P(entry);
}
Z_TRY_ADDREF_P(entry);
ZEND_HASH_FILL_ADD(entry);
} ZEND_HASH_FOREACH_END();
} ZEND_HASH_FILL_END();
} else {
ZEND_HASH_FOREACH_KEY_VAL(Z_ARRVAL_P(input), num_key, string_key, entry) {
pos++;
if (pos <= offset) {
continue;
}
if (pos > offset + length) {
break;
}
n++;
num_key = p->h;
string_key = p->key;

if (string_key) {
entry = zend_hash_add_new(Z_ARRVAL_P(return_value), string_key, entry);
} else {
if (preserve_keys) {
entry = zend_hash_index_add_new(Z_ARRVAL_P(return_value), num_key, entry);
if (string_key) {
entry = zend_hash_add_new(Z_ARRVAL_P(return_value), string_key, entry);
} else {
entry = zend_hash_next_index_insert_new(Z_ARRVAL_P(return_value), entry);
if (preserve_keys) {
entry = zend_hash_index_add_new(Z_ARRVAL_P(return_value), num_key, entry);
} else {
entry = zend_hash_next_index_insert_new(Z_ARRVAL_P(return_value), entry);
}
}
zval_add_ref(entry);
}
zval_add_ref(entry);
} ZEND_HASH_FOREACH_END();
}
}
}
/* }}} */
Expand Down
105 changes: 105 additions & 0 deletions ext/standard/tests/array/array_slice_variation11.phpt
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,105 @@
--TEST--
Test array_slice() function : usage variations - array has holes in buckets
--FILE--
<?php
/* Prototype : array array_slice(array $input, int $offset [, int $length [, bool $preserve_keys]])
* Description: Returns elements specified by offset and length
* Source code: ext/standard/array.c
*/

/*
* Check that results of array_slice are correct when there are holes in buckets caused by unset()
*/

echo "*** Testing array_slice() : usage variations ***\n";

function dump_slice(array $input, $offsetToUnset, int $offset, int $length) {
unset($input[$offsetToUnset]);
var_dump(array_slice($input, $offset, $length));
}

echo "\n-- Call array_slice() on array with string keys--\n";
$input = ['one' => 'un', 'two' => 'deux', 23 => 'twenty-three', 'zero'];
dump_slice($input, 'two', 0, 1);
dump_slice($input, 'two', 0, 2);
dump_slice($input, 'two', 0, 3);
dump_slice($input, 23, 1, 2);

echo "\n-- Call array_slice() on array with packed keys--\n";
$input = [10, 11, 12, 'thirteen'];
dump_slice($input, 0, 0, 1);
dump_slice($input, 1, 0, 1);
dump_slice($input, 1, 0, 3);
dump_slice($input, 1, -1, 1);
dump_slice($input, 1, 0, 3);
dump_slice($input, 1, -3, 3);

echo "Done";
?>
--EXPECT--
*** Testing array_slice() : usage variations ***

-- Call array_slice() on array with string keys--
array(1) {
["one"]=>
string(2) "un"
}
array(2) {
["one"]=>
string(2) "un"
[0]=>
string(12) "twenty-three"
}
array(3) {
["one"]=>
string(2) "un"
[0]=>
string(12) "twenty-three"
[1]=>
string(4) "zero"
}
array(2) {
["two"]=>
string(4) "deux"
[0]=>
string(4) "zero"
}

-- Call array_slice() on array with packed keys--
array(1) {
[0]=>
int(11)
}
array(1) {
[0]=>
int(10)
}
array(3) {
[0]=>
int(10)
[1]=>
int(12)
[2]=>
string(8) "thirteen"
}
array(1) {
[0]=>
string(8) "thirteen"
}
array(3) {
[0]=>
int(10)
[1]=>
int(12)
[2]=>
string(8) "thirteen"
}
array(3) {
[0]=>
int(10)
[1]=>
int(12)
[2]=>
string(8) "thirteen"
}
Done