Echo nested JSON array from related database tables with a single query?

问题

I have two database tables that contain information about land contracts. They are related with land_contract_annual_price.land_contract_id -> land_contract.land_contract_id.

Table 'land_contract'

Table 'land_contract_annual_price'

If a land contract has the value "Rörligt pris" in the field land_contract_price_type, there are related values in the table land_contract_annual_price. At the moment I'm doing two queries, one to each table. I then merge the results and present the land contract as a nested JSON array like this:

Version 1

[  
 {  
  "land_contract_id":118,
  "land_contract_name":"Avtalsnamn",
  "location_id":71,
  "land_contract_link":"",
  "land_contract_notes":"",
  "land_owner_id":2,
  "land_contract_start_date":"2019-07-25",
  "land_contract_end_date":"2023-07-25",
  "land_contract_terminated":"false",
  "land_contract_payment_interval":"Halv\u00e5rsvis",
  "land_contract_price_type":"R\u00f6rligt \u00e5rspris",
  "land_contract_fixed_annual_price":null,
  "land_contract_annual_prices":[  
    {"year":1, "price":873.00},
    {"year":2, "price":77289.00},
    {"year":3, "price":8.00},
    {"year":4, "price":0.00},
    {"year":5, "price":8729.00}
  ]
 }
]

If a land contract has the value "Fast pris" in the field land_contract_price_type, there are no related values in the table land_contract_annual_price. In that case I present the land contract like this (without the extra array at the end):

Version 2

[
 {
  "land_contract_id":13,
  "land_contract_name":null,
  "location_id":null,
  "land_contract_link":"https:\/\/www.something.com\/preview\/Sl%C3%A4pvdam%20Edda\/Kddal\/Bddkta\/Besika%20Markavtal%20%20Halmstad%202016-03-08.pdf?role=personal",
  "land_contract_notes":"",
  "land_owner_id":null,
  "land_contract_start_date":"2016-03-08",
  "land_contract_end_date":"2026-03-08",
  "land_contract_terminated":"true",
  "land_contract_payment_interval":"\u00c5rsvis",
  "land_contract_price_type":"Fast \u00e5rspris",
  "land_contract_fixed_annual_price":"6000.00"
 }
]

What I didn't think of, is that this solution is bad when I'm fetchin ALL the land contracts. If I'm going to do a second query to another table whenever a land contract has the value "Rörligt pris" in the field land_contract_price_type, I'm going to do hundreds of extra queries.

Is there a way to create the nested JSON array with one (1) query when a land contract has the value "Rörligt pris" in the field land_contract_price_type?

Thanks!

Below is my current code.

function read($pdo, $Id = null, $ResponseMessage = null) {

    $params = [];
    $array = [];

    $sql = "SELECT  lc.Id, lc.Name, lc.LocationId, l.Name AS LocationName, lc.Notes, lc.LandOwnerId, lo.Name AS LandOwnerName, lc.StartDate, lc.EndDate, lc.IsTerminated, lc.PaymentInterval, lc.PriceType, lc.FixedAnnualPrice, lc.Link, lc.Created, lc.Updated, lcap.AnnualPriceYear AS Year, lcap.AnnualPriceAmount AS Amount
            FROM LandContract lc
            LEFT JOIN Location l ON l.Id = lc.LocationId
            LEFT JOIN LandOwner lo ON lo.Id = lc.LandOwnerId
            LEFT JOIN LandContractAnnualPrice lcap ON lcap.LandContractId = lc.Id  
            ORDER BY lc.Id  DESC, lcap.AnnualPriceYear DESC
            ";
    if ($Id) {
        $sql .= 'WHERE lc.Id = ?';
        $params[] = $Id;
    }

    echo $sql;

    $stmt = $pdo->prepare($sql);
    $stmt->execute($params);
    while ($row = $stmt->fetch()) {
        // Fields we want to extract from the select statement into the array 
        $select_fields = ['Id', 'Name', 'LocationId', 'LocationName', 'Link', 'Notes', 'LandOwnerId', 'LandOwnerName',
                            'StartDate', 'EndDate', 'IsTerminated', 'PaymentInterval', 
                            'PriceType', 'FixedAnnualPrice ', 'Created', 'Updated'];

        if (!isset($array[$row['Id']])) {
            // initialize the subarray if it has not been set already 
            $array[$row['Id']] = array_intersect_key($row, array_flip($select_fields));

            if ($row['Year'] != null) {
                $array[$row['Id']]['AnnualPrices'] = [];
            } else {
                $array[$row['Id']]['AnnualPrice'] = $row['FixedAnnualPrice'];
            }
        }

        if ($row['Year'] != null) {
            $array[$row['Id']]['AnnualPrices'][] = ['Year' => $row['Year'], 'Amount' => $row['Amount']];
        }

    }

    if (empty($array)) {
        $ResponseMessage = new ResponseMessage();
        $ResponseMessage->Status = 'Error';
        $ResponseMessage->Message = 'No results';
        echo json_encode($ResponseMessage, JSON_UNESCAPED_UNICODE);
        exit;
    }

    $Response = array();

    if ($ResponseMessage) {
        $Response['Status'] = $ResponseMessage->Status;
        $Response['Message'] = $ResponseMessage->Message;
    }

    $Response['LandContracts'] = array_values($array);

    echo json_encode($Response, JSON_UNESCAPED_UNICODE);

    $stmt = null;
}

回答1:

You are better off using a JOIN query, and then structure your array from the result - having a query within a loop is often a very bad idea, and an indicator that you can use a JOIN instead.

You want to use a LEFT JOIN, joining them on the land_contract_id in both tables.

Then loop your results, and construct your array, which you can end up encoding into a JSON string once done.

$params = [];
$array = [];

$sql = "SELECT lc.*, 
               py.land_contract_annual_price_year AS `year`,  
               py.land_contract_annual_price_amount AS `amount`
        FROM land_contract AS lc
        LEFT JOIN land_contract_annual_price AS py 
            ON py.land_contract_id = lc.land_contract_id
        ";
if (isset($_POST['land_contract_id'])) {
    $sql .= 'WHERE lc.land_contract_id = ?';
    $params[] = $_POST["land_contract_id"];
}

$stmt = $pdo->prepare($sql);
$stmt->execute($params);
while ($row = $stmt->fetch()) {
    // Fields we want to extract from the select statement into the array 
    $select_fields = ['land_contract_id', 'land_contract_name', 'location_id', 'land_contract_link', 'land_contract_notes', 'land_owner_id', 
                        'land_contract_start_date', 'land_contract_end_date', 'land_contract_terminated', 'land_contract_payment_interval', 
                        'land_contract_price_type', 'land_contract_fixed_annual_price '];

    if (!isset($array[$row['land_contract_id']])) {
        // initialize the subarray if it has not been set already 
        $array[$row['land_contract_id']] = array_intersect_key($row, array_flip($select_fields));

        if ($row['year'] != null) {
            $array[$row['land_contract_id']]['land_contract_annual_prices'] = [];
        } else {
            $array[$row['land_contract_id']]['land_contract_annual_price'] = $row['land_contract_fixed_annual_price'];
        }
    }

    if ($row['year'] != null) {
        $array[$row['land_contract_id']]['land_contract_annual_prices'][] = ['year' => $row['year'], 'amount' => $row['amount']];
    }

}

if (empty($array)) {
    echo "No results";
    exit;
}

echo json_encode($array, JSON_UNESCAPED_UNICODE);

来源：https://stackoverflow.com/questions/57209240/echo-nested-json-array-from-related-database-tables-with-a-single-query