Doctrine ORM serialization after join - php

Question

I guys I am working diwth doctrine JOINS and I came across a problem.
I have done a query with joins and it seems ok, the results are good! but the output is terrible
public function getMarketcapData($page = -1, $limit = -1){
$pageAndCount = $page>-1 && $limit>0;
$qb = $this->getEntityManager()->createQueryBuilder();
$q = $qb->select('cc, count(cr) as coinRawsCount, cmc as relatedCategory')
->from('AppBundle:CoinClean', 'cc')
->leftJoin("AppBundle:CoinRaw", "cr", 'WITH', 'cc = cr.coinClean')
->leftJoin("AppBundle:CoinMapCategory", "cmc", 'WITH', 'cc.relatedCategory = cmc')
->andWhere('cc = cr.coinClean')
->andWhere('cc.relatedCategory = cmc')
->groupBy('cc.id')
;
if($pageAndCount) $q = $q->setFirstResult($page*$limit)->setMaxResults($limit);
$q= $q->orderBy('cc.rank', "ASC")->getQuery();
$result = $q->getResult();
if($pageAndCount){
$qb = $this->getEntityManager()->createQueryBuilder();
$q = $qb->select('count(u.id)')->from('AppBundle:CoinClean', 'u')->getQuery();
return new PageResult($result, (int)$q->getSingleScalarResult(), $page, $limit );
}else{
return $result;
}
}
here you have the output: it is an array made by array + obejcts
data:[
[ // first query result
{data from cc} // ony one object
],
{// first query result from join
"coinRawsCount": 4,
"relatedCategory": {...}
},
[ // second query result
{data from cc2} // ony one object
],
{// second query result from join
"coinRawsCount": 4,
"relatedCategory": {...}
}
]
my goal is to condense all in
"data":[
{
"data": {...},
"coinRawsCount": 4,
"relatedCategory": {...}
},
{
"data": {...},
"coinRawsCount": 4,
"relatedCategory": {...}
},
]
Any ideas?

Answer 1

You could loop over the result and merge on every other row, e.g. like this:
$rows = $q->getResult();
$result = []
foreach ($rows as $index => $row) {
if ($row instanceof App\Entity\CoinClean) {
// Skip first row and merge data when reading next row
continue;
}
$result[] = array_merge(
['data' => $rows[$index - 1]],
$row
);
}
return $result;
Assuming the first row is always the entity, we will skip this entry using continue;. The next row should contain an array with the additional data. We can then take the previous offset ($index - 1) to fetch the entry and the current row and merge them together. You might have to do additional checks.
Keep in mind that you will loop over a potentially large array and since you copy the data over to a new array this might use up a considerable amount of memory. So you might have to profile and optimize it, if you notice poor performance.
Other options would be to perform a native SQL query and then just map the output using a ResultSetMappingBuilder or to use custom hydration, i.e. create a new custom DTO-object for that query, that better fits the resulting data than the original entity structure, using DQL SELECT NEW App\Dto\MyModel FROM ...your query...'.