如何在Laravel中按question_id的唯一性对数组排序

我有一个包含多个子数组的数组列表，每个子数组包含若干字段。需要基于question_id字段对该列表排序，要求每个question_id对应的首个出现的数组排在前面，重复的同question_id数组放在后面，且不删除任何数组元素。

原始数组 #

array:6 [▼
  0 => array:8 [▼
    "id" => 35
    "answer" => null
    "type" => 3
    "user_id" => 9
    "question_id" => 36
    "answer_id" => 1
    "created_at" => "2023-04-07T09:03:39.000000Z"
    "updated_at" => "2023-04-07T09:03:39.000000Z"
  ]
  1 => array:8 [▼
    "id" => 36
    "answer" => null
    "type" => 3
    "user_id" => 9
    "question_id" => 36
    "answer_id" => 1
    "created_at" => "2023-04-07T09:03:39.000000Z"
    "updated_at" => "2023-04-07T09:03:39.000000Z"
  ]
  2 => array:8 [▼
    "id" => 37
    "answer" => "Gello"
    "type" => 0
    "user_id" => 9
    "question_id" => 37
    "answer_id" => 1
    "created_at" => "2023-04-07T09:03:39.000000Z"
    "updated_at" => "2023-04-07T09:03:39.000000Z"
  ]
  3 => array:8 [▼
    "id" => 38
    "answer" => "Uho"
    "type" => 0
    "user_id" => 9
    "question_id" => 37
    "answer_id" => 1
    "created_at" => "2023-04-07T09:03:39.000000Z"
    "updated_at" => "2023-04-07T09:03:39.000000Z"
  ]
  4 => array:8 [▼
    "id" => 39
    "answer" => "Who"
    "type" => 1
    "user_id" => 9
    "question_id" => 38
    "answer_id" => 1
    "created_at" => "2023-04-07T09:03:39.000000Z"
    "updated_at" => "2023-04-07T09:03:39.000000Z"
  ]
  5 => array:8 [▼
    "id" => 40
    "answer" => "Window"
    "type" => 1
    "user_id" => 9
    "question_id" => 38
    "answer_id" => 1
    "created_at" => "2023-04-07T09:03:40.000000Z"
    "updated_at" => "2023-04-07T09:03:40.000000Z"
  ]
]

期望排序后的数组 #

array:6 [▼
  0 => array:8 [▼
    "id" => 35
    "answer" => null
    "type" => 3
    "user_id" => 9
    "question_id" => 36
    "answer_id" => 1
    "created_at" => "2023-04-07T09:03:39.000000Z"
    "updated_at" => "2023-04-07T09:03:39.000000Z"
  ]
  2 => array:8 [▼
    "id" => 37
    "answer" => "Hello"
    "type" => 0
    "user_id" => 9
    "question_id" => 37
    "answer_id" => 1
    "created_at" => "2023-04-07T09:03:39.000000Z"
    "updated_at" => "2023-04-07T09:03:39.000000Z"
  ]
  4 => array:8 [▼
    "id" => 39
    "answer" => "Who"
    "type" => 1
    "user_id" => 9
    "question_id" => 38
    "answer_id" => 1
    "created_at" => "2023-04-07T09:03:39.000000Z"
    "updated_at" => "2023-04-07T09:03:39.000000Z"
  ]
  1 => array:8 [▼
    "id" => 36
    "answer" => null
    "type" => 3
    "user_id" => 9
    "question_id" => 36
    "answer_id" => 1
    "created_at" => "2023-04-07T09:03:39.000000Z"
    "updated_at" => "2023-04-07T09:03:39.000000Z"
  ]
  3 => array:8 [▼
    "id" => 38
    "answer" => "Uho"
    "type" => 0
    "user_id" => 9
    "question_id" => 37
    "answer_id" => 1
    "created_at" => "2023-04-07T09:03:39.000000Z"
    "updated_at" => "2023-04-07T09:03:39.000000Z"
  ]
  5 => array:8 [▼
    "id" => 40
    "answer" => "Window"
    "type" => 1
    "user_id" => 9
    "question_id" => 38
    "answer_id" => 1
    "created_at" => "2023-04-07T09:03:40.000000Z"
    "updated_at" => "2023-04-07T09:03:40.000000Z"
  ]
]

实现方案（PHP） #

推荐采用分组重组的方式，逻辑简单且效率较高：

1. 遍历原始数组，将每个question_id的首个元素单独收集，重复元素按question_id分组存储
2. 先将所有首个元素按原始顺序排列，再依次追加每个question_id的重复元素

代码示例：

// 原始数组，替换为你的数组变量
$originalArray = [
    ["id" => 35, "answer" => null, "type" => 3, "user_id" => 9, "question_id" => 36, "answer_id" => 1, "created_at" => "2023-04-07T09:03:39.000000Z", "updated_at" => "2023-04-07T09:03:39.000000Z"],
    ["id" => 36, "answer" => null, "type" => 3, "user_id" => 9, "question_id" => 36, "answer_id" => 1, "created_at" => "2023-04-07T09:03:39.000000Z", "updated_at" => "2023-04-07T09:03:39.000000Z"],
    ["id" => 37, "answer" => "Gello", "type" => 0, "user_id" => 9, "question_id" => 37, "answer_id" => 1, "created_at" => "2023-04-07T09:03:39.000000Z", "updated_at" => "2023-04-07T09:03:39.000000Z"],
    ["id" => 38, "answer" => "Uho", "type" => 0, "user_id" => 9, "question_id" => 37, "answer_id" => 1, "created_at" => "2023-04-07T09:03:39.000000Z", "updated_at" => "2023-04-07T09:03:39.000000Z"],
    ["id" => 39, "answer" => "Who", "type" => 1, "user_id" => 9, "question_id" => 38, "answer_id" => 1, "created_at" => "2023-04-07T09:03:39.000000Z", "updated_at" => "2023-04-07T09:03:39.000000Z"],
    ["id" => 40, "answer" => "Window", "type" => 1, "user_id" => 9, "question_id" => 38, "answer_id" => 1, "created_at" => "2023-04-07T09:03:40.000000Z", "updated_at" => "2023-04-07T09:03:40.000000Z"],
];

$firstItems = [];
$duplicateGroups = [];

foreach ($originalArray as $item) {
    $qid = $item['question_id'];
    if (!isset($duplicateGroups[$qid])) {
        // 首次出现的元素加入单独列表
        $firstItems[] = $item;
        $duplicateGroups[$qid] = [];
    } else {
        // 重复元素加入对应分组
        $duplicateGroups[$qid][] = $item;
    }
}

// 合并得到排序后的数组：先首个元素，再各分组的重复元素
$sortedArray = array_merge($firstItems, ...array_values($duplicateGroups));

// 输出结果
print_r($sortedArray);

方案说明 #

• 该方法时间复杂度为O(n)，仅需遍历一次原始数组，效率远高于排序函数多次比较的方式
• 严格保留了原始数组中question_id的首次出现顺序，以及重复元素的原始顺序

内容的提问来源于stack exchange，提问作者Amirali Pourmoshir