用于获取地理位置记录的 MySQL 查询优化

Posted 2023-04-15

【中文标题】用于获取地理位置记录的 MySQL 查询优化

【英文标题】：MySQL Query optimisation for fetching geolocation records

【发布时间】：2013-12-09 12:07:24

【问题描述】：

我有一个地方的地理位置意味着我有一个CenterLatitude 和CenterLongitude 和一个radius。在这个半径内有很多不同类型的区域记录RegionType。我想通过匹配radius 中每个RegionType 的记录强度来获取正确的RegionType 记录。

我目前的逻辑是这样的，

第 1 步：找到最小和最大 lat 和 long 值

$milesToKM = 112.654; $minLat = $data['lat'] - ($data['rad'] / $milesToKM); $maxLat = $data['lat'] + ($data['rad'] / $milesToKM); $minLong = $data['lon'] - ($data['rad'] / $milesToKM); $maxLong = $data['lon'] + ($data['rad'] / $milesToKM);

第 2 步：选择 Regiontype 和 count

$query = $pdo->prepare("SELECT `RegionType`, count(`RegionType`) as count from (SELECT RegionType FROM data_api_region WHERE (CenterLatitude between $minLat and $maxLat) AND (CenterLongitude between $minLong and $maxLong) AND languageCode = '" . $data['locale'] . "') q GROUP BY `RegionType`");
$query->execute();
$regionCounts = $query->fetchAll(PDO::FETCH_ASSOC);
unset($query);

$finalCountArr = array();
foreach ($regionCounts as $regionCountData) 
$finalCountArr[$regionCountData['RegionType']] = $regionCountData['count'];

第三步：决定相关RegionType

switch (true) 
            case (isset($finalCountArr['Continent']) && $finalCountArr['Continent'] > 5):
                $regionTypeFinal = 'Continent';
                break;
            case (isset($finalCountArr['Country']) && $finalCountArr['Country'] > 5):
                $regionTypeFinal = 'Country';
                break;
            case (isset($finalCountArr['Multi-Region (within a country)']) && $finalCountArr['Multi-Region (within a country)'] > 5):
                $regionTypeFinal = 'Multi-Region (within a country)';
                break;
            case (isset($finalCountArr['Province (State)']) && $finalCountArr['Province (State)'] > 5):
                $regionTypeFinal = 'Province (State)';
                break;
            case (isset($finalCountArr['Multi-City (Vicinity)']) && $finalCountArr['Multi-City (Vicinity)'] > 5):
                $regionTypeFinal = 'Multi-City (Vicinity)';
                break;
            case (isset($finalCountArr['City']) && $finalCountArr['City'] > 5):
                $regionTypeFinal = 'City';
                break;
            case (isset($finalCountArr['Neighborhood']) && $finalCountArr['Neighborhood'] > 5):
                $regionTypeFinal = 'Neighborhood';
                break;
            default:
                $regionTypeFinal = 'Point of Interest';
                break;
        

第 4 步：获取正确的记录

$query = $pdo->prepare("SELECT `id` as regionId, `RegionName` as regionName,`RegionNameLong` as regionNameLong, `RegionType` as regionType,`CenterLongitude` as centerLongitude, `CenterLatitude` as centerLatitude ,`LanguageCode` as languageCode FROM data_api_region WHERE (CenterLatitude between $minLat and $maxLat) AND (CenterLongitude between $minLong and $maxLong) AND languageCode = '" . $data['locale'] . "' AND RegionType = '$regionTypeFinal'");
        $query->execute();
        $finalResultRegions = $query->fetchAll(PDO::FETCH_ASSOC);

谁能告诉我如何优化这个过程，因为表中缺少记录，所以这个过程需要时间..是否可以使用一个 SQL 查询？

【参考方案1】：

没有测试数据或表格布局，所以我无法真正测试它，但它可能可以在一条 SQL 中完成，如下所示：-

SELECT `id` as regionId, `RegionName` as regionName,`RegionNameLong` as regionNameLong, `RegionType` as regionType,`CenterLongitude` as centerLongitude, `CenterLatitude` as centerLatitude ,`LanguageCode` as languageCode 
FROM data_api_region 
INNER JOIN
(
    SELECT CASE
            WHEN RegionType = 'Continent' AND `count` > 5
                THEN 'Continent'
            WHEN RegionType = 'Country' AND `count`  > 5
                THEN 'Country'
            WHEN RegionType = 'Multi-Region (within a country)' AND `count`  > 5
                THEN 'Multi-Region (within a country)'
            WHEN RegionType = 'Province (State)' AND `count`  > 5
                THEN 'Province (State)'
            WHEN RegionType = 'Multi-City (Vicinity)' AND `count`  > 5
                THEN 'Multi-City (Vicinity)'
            WHEN RegionType = 'City' AND `count`  > 5
                THEN 'City'
            WHEN RegionType = 'Neighborhood' AND `count`  > 5
                THEN 'Neighborhood'
            ELSE 'Point of Interest'
        END AS regionTypeFinal
    FROM 
    (
        SELECT `RegionType`, count(`RegionType`) as `count` 
        FROM data_api_region 
        WHERE (CenterLatitude between $minLat and $maxLat) 
        AND (CenterLongitude between $minLong and $maxLong) AND languageCode = '" . $data['locale'] . "'
        GROUP BY `RegionType`
    ) q 
) Sub1
ON data_api_region.RegionType = Sub1.regionTypeFinal
WHERE (CenterLatitude between $minLat and $maxLat) 
AND (CenterLongitude between $minLong and $maxLong) 
AND languageCode = '" . $data['locale'] . "' 

【讨论】：

它确实有效..但是查询仍然需要 2-3 秒才能执行..无论如何谢谢。

你能发布表格声明吗？