PHP Recursive Loop using UASORT on Multidimensional Array - php

I am writing a script that loops through a multidimensional array and it's working as hoped (sort of) but I get errors that I just can't remedy.
I am still not that comfortable building loops to manage nested arrays.
Here is my code. The goal is to sort each layer by the value of the sequence key and in the end I export the array as json.
The sequence key may or may not exist in every sub array so that may need some sort of if clause
<?php
$list = [
"key" => "book",
"sequence" => 1,
"items" => [
[
"key" => "verse",
"sequence" => 2,
"items" => [
["sequence" => 3],
["sequence" => 1],
["sequence" => 2],
],
],
[
"key" => "page",
"sequence" => 1,
"items" => [
[
"key" => "page",
"sequence" => 2,
"items" => [
["sequence" => 2],
["sequence" => 1],
["sequence" => 3],
],
],
[
"key" => "paragraph",
"sequence" => 1,
"items" => [
["sequence" => 2],
["sequence" => 1],
["sequence" => 3],
],
],
],
],
],
];
function sortit(&$array){
foreach($array as $key => &$value){
//If $value is an array.
if(is_array($value)){
if($key == "items"){
uasort($value, function($a,&$b) {
return $a["sequence"] <=> $b["sequence"];
});
}
//We need to loop through it.
sortit($value);
} else{
//It is not an array, so print it out.
echo $key . " : " . $value . "<br/>";
}
}
}
sortit($list);
echo "<pre>";
print_r($list);
?>
Here is the output and error I am getting, and I think I understand why the error is being thrown but at the same time I can not implement the proper checks needed to fix the error.
key : book
sequence : 1
key : page
sequence : 1
E_WARNING : type 2 -- Illegal string offset 'sequence' -- at line 39
E_NOTICE : type 8 -- Undefined index: sequence -- at line 39
sequence : 1
sequence : 2
sequence : 3
sequence : 1
key : page
E_WARNING : type 2 -- Illegal string offset 'sequence' -- at line 39
E_NOTICE : type 8 -- Undefined index: sequence -- at line 39
sequence : 1
sequence : 2
sequence : 3
sequence : 2
key : verse
Not that I am worried to much but another thing that I would like is the array to still be structured in the original order, ie: key, sequence, items

Using usort and array references makes it straightforward. If we're dealing with an array with a set item key, sort the item array and recurse on its children, otherwise, we're at a leaf node and can return.
function seqSort(&$arr) {
if (is_array($arr) && array_key_exists("items", $arr)) {
usort($arr["items"], function ($a, $b) {
return $a["sequence"] - $b["sequence"];
});
foreach ($arr["items"] as &$item) {
$item = seqSort($item);
}
}
return $arr;
}
Result:
array (
'key' => 'book',
'sequence' => 1,
'items' =>
array (
0 =>
array (
'key' => 'page',
'sequence' => 1,
'items' =>
array (
0 =>
array (
'key' => 'page',
'sequence' => 1,
'items' =>
array (
0 =>
array (
'sequence' => 1,
),
1 =>
array (
'sequence' => 2,
),
2 =>
array (
'sequence' => 3,
),
),
),
),
),
1 =>
array (
'key' => 'verse',
'sequence' => 2,
'items' =>
array (
0 =>
array (
'sequence' => 1,
),
1 =>
array (
'sequence' => 2,
),
2 =>
array (
'sequence' => 3,
),
),
),
),
)
Try it!
Note that the outermost structure is a root node that isn't part of an array and can't be sorted (this may be unintentional and causing confusion).

Related

How to extract each level in nested array to separated arrays with specific key-value pairs?

I've got a 4-level nested array, each level represent a hiereachy in an organization. Each level's key is the hierarchy value, and inside it there is the id of it in the array. The last level is only the value-id pair:
$data = [
"top_level_data1" => [
'top_level_id' => 0,
'sub_level_1' => [
'sub_level_1_id' => 0,
'sub_level_2_data1' => [
'sub_level_2_id' => 0,
'sub_level_2_data' => [
0 => "some_val"
1 => "some_other_val"
]
]
]
],
"top_level_data2" => [
'level_1_id' => 1,
'sub_level_1_other' => [
'sub_level_1_id' => 1,
'sub_level_2_data2' => [
'sub_level_2_id' => 1,
'sub_level_3_data1' => [
2 => "another_val"
3 => "bar"
4 => "foo"
]
],
'sub_level_2_data3' => [
'sub_level_2_id' => 2,
'sub_level_3_data2' => [
5 => "foobar"
6 => "hello"
7 => "goodbye"
]
]
]
]
];
I want to extract it to separate arrays that would contain the hierarchy value-id pairs.
Expected output from the above example (without the ids of the above level):
$top_level = [
0 => "top_level_data1",
1 => "top_level_data2"
]
$sub_level_1 = [
0 => "sub_level_1",
1 => "sub_level_1_other"
]
$sub_level_2 = [
0 => "sub_level_2_data1",
1 => "sub_level_2_data2",
2 => "sub_level_2_data3"
]
$sub_level_3 = [
0 => "some_val"
1 => "some_other_val"
2 => "another_val"
3 => "bar"
4 => "foo"
5 => "foobar"
6 => "hello"
7 => "goodbye"
]
I've tried using the RecursiveIterator but that does not work (At first I wanted to use it just to check if it iterates correctly, even before assignting the key-values as I wanted, but it just doesn't do it as I wanted):
$it = new RecursiveIteratorIterator(new RecursiveArrayIterator($data));
foreach($it as $key => $val) {
$level[$it->getDepth()][] = $val;
}
This can do it:
$result = [];
$recursive = null;
$recursive = function($args,$level=0) use (&$recursive,&$result){
foreach($args as $k=>$v){
if(is_array($v)){
$result[$level][]=$k;
$recursive($v,$level+1);
} else if(is_string($v)){
$result[$level][]=$v;
}
}
};
$recursive($data);
var_export($result);
But your expected output isn't ok. Because i think sub_level_2_data must be sub_level_3_data (in the $data array above). And the hole sub_level_3*level is missing in your expected output.
With this code you get an result array instead of separate filled variables.
Will not explain it in detail, but try to understand whats happening here and you can learn some new php stuff.

How can I manage some irregular array to regular array?

I write code with some array that have different structure, but I must extract the data to do something else. How can I manager these array?
The array's structure are as follow:
$a = [
'pos1' => 'somedata',
'pos2' => ['data2', 'data3'],
'pos3' => '';
];
$b = [
[
'pos1' => ['data1', 'data2', ['nest1', 'nest2']],
'pos2' => ['data1', 'data2', 'data3'],
],
['data1', 'data2'],
'data4',
];
The array's Index can be a key or a position, and the value of the corresponding index may be a array with the same structure. More tough problem is that the subarray can be nesting, and the time of the nesting has different length.
Fortunately, every array has it's owe fixed structure.
I want to convert the these array to the format as follow. When the index is a value, change it to the keyword; and if the index is a keyword, nothing changed.
$a = [
'pos1' => 'somedata',
'pos2' => [
'pos2_1' => 'data2',
'pos2_2' => 'data3'
],
'pos3' => '';
];
$b = [
'pos1' => [
'pos1_1' => [
'pos1_1_1' => 'data1',
'pos1_1_2' => 'data2',
'pos1_1_3' => [
'pos1_1_3_1' => 'nest1',
'pos1_1_3_2' => 'nest2',
],
],
'pos1_2' => [
'pos1_2_1' => 'data1',
'pos1_2_2' => 'data2',
'pos1_2_3' => 'data3',
],
],
'pos2' => ['data1', 'data2'],
'pos3' => 'data4',
];
My first solution is for every array, write the function to convert the format(the keyword will specify in function). But it is a huge task and diffcult to manage.
The second solution is write a common function, with two argument: the source array and the configuration that specify the keyword to correspondent value index. For example:
$a = [0, ['pos10' => 1]];
$conf = [
// It means that when the value index is 0, it will change it into 'pos1'
'pos1' => 0,
'pos2' => 1,
];
The common funciton will generate the result of:
$result = [
'pos1' => 0,
'pos2' => ['pos10' => 1],
]
But this solution will lead to a problem: the config is diffcult to understand and design, and other people will spend a lot of time to understand the format after conversion.
Is there are some better solution to manage these array that other people can easy to use these array?
Thanks.

Array sort menu

I have the following array to show menu's based on the order the user specified.
The array is as follows:
$menuArray = [
'Main Street' => [
['/index.php', 'Home'],
['/city.php', $cityData[$user->city][0]],
['/travel.php', 'Travel'],
['/bank.php', 'Bank'],
['/inventory.php', 'Inventory'],
['/dailies.php', 'Dailies'],
],
'Activities' => [
(!$my->hospital) ? ['/hospital.php', 'Hospital'] : [],
(!$my->hospital && !$my->prison) ? ['/crime.php', 'Crime'] : [],
['/missions.php', 'Missions'],
['/achievements.php', 'Achievements'],
],
'Services' => [
['/hospital.php', 'Hospital'],
['/prison.php', 'Prison'],
['/search.php', 'Search'],
],
'Account' => [
['/edit_account.php', 'Edit Account'],
['/notepad.php', 'Notepad'],
['/logout.php', 'Logout'],
]
];
I have a column menu_order stored in the database, which has a default value of 0,1,2,3,4, but this can change per user as they will be able to change their menu to their likes.
What I'd like to achieve:
0 => Main Street
1 => Activities
2 => Services
3 => Account
4 => Communication
To get the menu order, I do
$menuOrder = explode(',', $user->menu_order);
But I'm not sure how to handle the foreach for displaying the menu.
Here's one way to do it -- use replacement rather than a sorting algorithm.
Code: (Demo)
$menuArray = [
'Main Street' => [],
'Activities' => [],
'Services' => [],
'Account' => []
];
$lookup = [
0 => 'Main Street',
1 => 'Activities',
2 => 'Services',
3 => 'Account',
4 => 'Communication'
];
$customsort = '4,2,1,3,0';
$keys = array_flip(explode(',', $customsort)); convert string to keyed array
//var_export($keys);
$ordered_keys = array_flip(array_replace($keys, $lookup)); // apply $lookup values to keys, then invert key-value relationship
//var_export($ordered_keys);
$filtered_keys = array_intersect_key($ordered_keys, $menuArray); // remove items not on the current menu ('Communication" in this case)
//var_export($filtered_keys);
$final = array_replace($filtered_keys, $menuArray); // apply menu data to ordered&filtered keys
var_export($final);
Output:
array (
'Services' =>
array (
),
'Activities' =>
array (
),
'Account' =>
array (
),
'Main Street' =>
array (
),
)
And here's another way using uksort() and a spaceship operator:
$ordered_keys = array_flip(array_values(array_replace(array_flip(explode(',', $customsort)), $lookup)));
uksort($menuArray, function($a, $b) use ($ordered_keys) {
return $ordered_keys[$a] <=> $ordered_keys[$b];
});
var_export($menuArray);
As a consequence of how your are storing your custom sort order, most of the code involved is merely to set up the "map"/"lookup" data.
You could try something like this to produce the menu:
function display_menu($menus, $m) {
if (!isset($menus[$m])) return;
echo "<ul>";
foreach ($menus[$m] as $item) {
if (!count($item)) continue;
echo "<li>{$item[1]}\n";
}
echo "</ul>";
}
$menuMap = array(0 => 'Main Street',
1 => 'Activities',
2 => 'Services',
3 => 'Account',
4 => 'Communication');
$menuOrder = explode(',', $user->menu_order);
foreach ($menuOrder as $menuIndex) {
$thisMenu = $menuMap[$menuIndex];
display_menu($menuArray, $thisMenu);
}
Small demo on 3v4l.org

Keep array rows where a column value is found in a second flat array

** I have edited this to show how I got my code to work using array_search
I have an array, $arr1 with 5 columns as such:
key id name style age whim
0 14 bob big 33 no
1 72 jill big 22 yes
2 39 sue yes 111 yes
3 994 lucy small 23 no
4 15 sis med 24 no
5 16 maj med 87 yes
6 879 Ike larg 56 no
7 286 Jed big 23 yes
This array is in a cache, not a database.
I then have a second array with a list of id values -
$arr2 = array(0=>14, 1=>72, 2=>8790)
How do I filter $arr1 so it returns only the rows with the id values in $arr2?
I got my code to work as follows:
$arr1 = new CachedStuff(); // get cache
$resultingArray = []; // create an empty array to hold rows
$filter_function = function ($row) use ($arr2) {
return (array_search($row['id'], $arr2));
};
$resultingArrayIDs = $arr1->GetIds($filter_function, $resultingArray);
This gives me two outputs: $resultingArray & $resultingArrayIDs both of which represent the intersection of the $arr1 and $arr2.
This whole task can be accomplished with just one slick, native function call -- array_uintersect().
Because the two compared parameters in the custom callback may come either input array, try to access from the id column and if there isn't one declered, then fallback to the parameter's value.
Under the hood, this function performs sorting while evaluating as a means to improve execution time / processing speed. I expect this approach to outperform iterated calls of in_array() purely from a point of minimized function calls.
Code: (Demo)
var_export(
array_uintersect(
$arr1,
$arr2,
fn($a, $b) =>
($a['id'] ?? $a)
<=>
($b['id'] ?? $b)
)
);
Something like this should do it, provided I've understood your question and data structure correctly:
$dataArray = [
[ 'key' => 0, 'id' => 14 , 'name' => 'bob' , 'style' => 'big' , 'age' => 33 , 'whim' => 'no' ],
[ 'key' => 1, 'id' => 72 , 'name' => 'jill' , 'style' => 'big' , 'age' => 22 , 'whim' => 'yes' ],
[ 'key' => 2, 'id' => 39 , 'name' => 'sue' , 'style' => 'yes' , 'age' => 111 , 'whim' => 'yes' ],
[ 'key' => 3, 'id' => 994 , 'name' => 'lucy' , 'style' => 'small' , 'age' => 23 , 'whim' => 'no' ],
[ 'key' => 4, 'id' => 15 , 'name' => 'sis' , 'style' => 'med' , 'age' => 24 , 'whim' => 'no' ],
[ 'key' => 5, 'id' => 16 , 'name' => 'maj' , 'style' => 'med' , 'age' => 87 , 'whim' => 'yes' ],
[ 'key' => 6, 'id' => 879 , 'name' => 'Ike' , 'style' => 'larg' , 'age' => 56 , 'whim' => 'no' ],
[ 'key' => 7, 'id' => 286 , 'name' => 'Jed' , 'style' => 'big' , 'age' => 23 , 'whim' => 'yes' ]
];
$filterArray = [14, 72, 879];
$resultArray = array_filter( $dataArray, function( $row ) use ( $filterArray ) {
return in_array( $row[ 'id' ], $filterArray );
} );
View this example on eval.in
However, your question appears to suggest this data might be coming from a database; is that correct? If so, perhaps it's more efficient to pre-filter the results at the database-level. Either by adding a field in the SELECT query, that represents a boolean value whether a row matched your filter ids, or by simply not returning the other rows at all.
One way is with foreach loop with array_search()
$result = [];
foreach ($arr1 as $value) { // Loop thru $arr1
if (array_search($value['id'], $arr2) !== false) { // Check if id is in $arr2
$result[] = $value; // Push to result if true
}
}
// print result
print_r($result);
As #DecentDabbler mentioned - if the data is coming out of a database, using an IN on your WHERE will allow you to retrieve only the relevant data.
Another way to filter is to use array functions
array_column extracts the value of the id column into an array
array_intersect returns the elements which are in both $arr1['id'] and $arr2
array_flip flips the resulting array such that the indices into $arr1 indicate the elements in both $arr1 and $arr2
$arr1 = [ [ 'id' => 14, 'name' => 'bob'],
['id' => 72, 'name' => 'jill'],
['id' => 39, 'name' => 'sue'],
['id' => 994, 'name' => 'lucy'],
['id' => 879, 'name'=> 'large']];
$arr2 = [ 14,72,879 ];
$intersection = array_flip(array_intersect(array_column($arr1,'id'),$arr2));
foreach ($intersection as $i) {
var_dump($arr1[$i]);;
}

Grouping element array php based first character value

I have an array based MySql database.
This is the array.
[
0 => [
'id' => '1997'
'lokasi_terakhir' => 'YA4121'
]
1 => [
'id' => '1998'
'lokasi_terakhir' => 'PL2115'
]
2 => [
'id' => '1999'
'lokasi_terakhir' => 'PL4111'
]
]
How can I get the element lokasi_terakhir that grouped by the first character ? What the best way ?
This is the goal :
[
"Y" => 1,
"P" => 2
]
Please advise
Here are two refined methods. Which one you choose will come down to your personal preference (you won't find better methods).
In the first, I am iterating the array, declaring the first character of the lokasi_terakhir value as the key in the $result declaration. If the key doesn't yet exist in the output array then it must be declared / set to 1. After it has been instantiated, it can then be incremented -- I am using "pre-incrementation".
The second method first maps a new array using the first character of the lokasi_terakhir value from each subarray, then counts each occurrence of each letter.
(Demonstrations Link)
Method #1: (foreach)
foreach($array as $item){
if(!isset($result[$item['lokasi_terakhir'][0]])){
$result[$item['lokasi_terakhir'][0]]=1; // instantiate
}else{
++$result[$item['lokasi_terakhir'][0]]; // increment
}
}
var_export($result);
Method #2: (functional)
var_export(array_count_values(array_map(function($a){return $a['lokasi_terakhir'][0];},$array)));
// generate array of single-character elements, then count occurrences
Output: (from either)
array (
'Y' => 1,
'P' => 2,
)
You can group those items like this:
$array = [
0 => [
'id' => '1997',
'lokasi_terakhir' => 'YA4121'
],
1 => [
'id' => '1998',
'lokasi_terakhir' => 'PL2115'
],
2 => [
'id' => '1999',
'lokasi_terakhir' => 'PL4111'
]
];
$result = array();
foreach($array as $item) {
$char = substr($item['lokasi_terakhir'], 0, 1);
if(!isset($result[$char])) {
$result[$char] = array();
}
$result[$char][] = $item;
}
<?php
$array=[
0 => [
'id' => '1997',
'lokasi_terakhir' => 'YA4121'
],
1 => [
'id' => '1998',
'lokasi_terakhir' => 'PL2115'
],
2 => [
'id' => '1999',
'lokasi_terakhir' => 'PL4111'
]
];
foreach($array as $row){
$newArray[]=$row['lokasi_terakhir'][0];
}
print_r(array_flip(array_unique($newArray)));
this code gets the first letter of the fields lokasi_terakhir , get the unique values to avoid duplicates and just flips the array to get the outcome you want.
The output is this :
Array ( [Y] => 0 [P] => 1 )

Categories