Merge pull request #50 from level23/rowCount

added rowCount() method
level23 · Sep 28, 2023 · 306b712 · 306b712
2 parents 4b30a26 + d7e8763
commit 306b712
Show file tree

Hide file tree

Showing 4 changed files with 203 additions and 29 deletions.
diff --git a/README.md b/README.md
@@ -215,6 +215,7 @@ for more information.
         - [structure](#metadata-structure)
         - [timeBoundary](#metadata-timeboundary)
         - [dataSources](#metadata-datasources)
+        - [rowCount](#metadata-rowcount)
     - [Reindex/compact data/kill](#reindex--compact-data--kill)
         - [compact()](#compact)
         - [reindex()](#reindex)
@@ -3644,6 +3645,25 @@ foreach($dataSources as $dataSource) {
 }
 ```
 
+#### `metadata()->rowCount()`
+
+Retrieve the number of rows for the given dataSource and interval.
+
+The `rowCount()` method has the following parameters:
+
+| **Type**                  | **Optional/Required** | **Argument**  | **Example**      | **Description**                                                                                                                                                                           |
+|---------------------------|-----------------------|---------------|------------------|-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
+| string                    | Required              | `$dataSource` | "wikipedia"      | The name of the dataSource (table) where you want to count the rows for.                                                                                                                  |
+| string/int/DateTime       | Required              | `$start`      | "now - 24 hours" | The start date to retrieve the row count for. See [interval()](#interval) for all allowed formats.                                                                                        |
+| /string/int/DateTime/null | Optional              | `$stop`       | "now"            | The stop date to retrieve the row count for. See [interval()](#interval) for all allowed formats. When a string containing a slash is given as start date, the stop date can be left out. |
+
+Example:
+
+```php
+// Retrieve the total records for the past week.
+$numRows = $client->metadata()->rowCount("wikipedia", "now - 1 week", "now");
+```
+
 
 ## Reindex / compact data / kill
 

diff --git a/src/Metadata/MetadataBuilder.php b/src/Metadata/MetadataBuilder.php
@@ -5,6 +5,7 @@
 
 use DateTime;
 use Exception;
+use DateTimeInterface;
 use InvalidArgumentException;
 use Level23\Druid\DruidClient;
 use Level23\Druid\Types\TimeBound;
@@ -288,21 +289,74 @@ public function interval(string $dataSource, string $interval): array
      *          )
      *  )
      *
-     * @param string $dataSource
-     * @param string $interval
+     * @param string                             $dataSource
+     * @param \DateTimeInterface|int|string      $start
+     * @param \DateTimeInterface|int|string|null $stop
      *
      * @return array<int,array<string,string>>
+     * @throws \GuzzleHttp\Exception\GuzzleException
      * @throws \Level23\Druid\Exceptions\QueryResponseException
      * @throws \Exception
+     */
+    protected function getColumnsForInterval(
+        string $dataSource,
+        DateTimeInterface|int|string $start,
+        DateTimeInterface|int|string $stop = null
+    ): array {
+        $response = $this->client->query($dataSource)
+            ->interval($start, $stop)
+            ->segmentMetadata();
+
+        $columns = [];
+
+        $rows = $response->data();
+
+        if (isset($rows[0])) {
+
+            /** @var array<string,array<string,array<string,string>>> $row */
+            $row = $rows[0];
+
+            if (isset($row['columns'])) {
+                array_walk($row['columns'], function ($value, $key) use (&$columns) {
+                    $columns[] = array_merge($value, ['field' => $key]);
+                });
+            }
+        }
+
+        return $columns;
+    }
+
+    /**
+     * Return the total number of rows for the given interval
+     *
+     * @param string                             $dataSource The name of the dataSource where you want to count the
+     *                                                       rows for
+     * @param \DateTimeInterface|int|string      $start      The start of the interval.
+     * @param \DateTimeInterface|int|string|null $stop       The end of the interval, or null when it was given as a
+     *                                                       "date/date" interval in the $start parameter.
+     *
+     * @return int
      * @throws \GuzzleHttp\Exception\GuzzleException
+     * @throws \Level23\Druid\Exceptions\QueryResponseException
+     * @throws \Exception
      */
-    protected function getColumnsForInterval(string $dataSource, string $interval): array
-    {
+    public function rowCount(
+        string $dataSource,
+        DateTimeInterface|int|string $start,
+        DateTimeInterface|int|string $stop = null
+    ): int {
         $response = $this->client->query($dataSource)
-            ->interval($interval)
+            ->interval($start, $stop)
             ->segmentMetadata();
 
-        return $response->data();
+        $totalRows = 0;
+        foreach ($response->data() as $row) {
+            if (isset($row['numRows'])) {
+                $totalRows += intval($row['numRows']);
+            }
+        }
+
+        return $totalRows;
     }
 
     /**

diff --git a/src/Responses/SegmentMetadataQueryResponse.php b/src/Responses/SegmentMetadataQueryResponse.php
@@ -12,19 +12,6 @@ class SegmentMetadataQueryResponse extends QueryResponse
      */
     public function data(): array
     {
-        $columns = [];
-        if (isset($this->response[0])) {
-
-            /** @var array<string,array<string,array<string,string>>> $row */
-            $row = $this->response[0];
-
-            if (isset($row['columns'])) {
-                array_walk($row['columns'], function ($value, $key) use (&$columns) {
-                    $columns[] = array_merge($value, ['field' => $key]);
-                });
-            }
-        }
-
-        return $columns;
+        return $this->response;
     }
 }
diff --git a/tests/Metadata/MetadataBuilderTest.php b/tests/Metadata/MetadataBuilderTest.php
@@ -282,14 +282,10 @@ public function testStructure(
     }
 
     /**
-     * @testWith [[{"columns": {"__time": {"type":"LONG"}}}]]
-     *
-     * @param array<string,array<string,array<string,string>>> $segmentMetadataResponse
-     *
      * @throws \GuzzleHttp\Exception\GuzzleException
      * @throws \Level23\Druid\Exceptions\QueryResponseException
      */
-    public function testGetColumnsForInterval(array $segmentMetadataResponse): void
+    public function testGetColumnsForInterval(): void
     {
         $dataSource      = 'myDataSource';
         $interval        = '2019-08-19T13:00:00.000Z/2019-08-19T14:00:00.000Z';
@@ -304,10 +300,62 @@ public function testGetColumnsForInterval(array $segmentMetadataResponse): void
 
         $queryBuilder->shouldReceive('interval')
             ->once()
-            ->with($interval)
+            ->with($interval, null)
             ->andReturn($queryBuilder);
 
-        $responseObj = new SegmentMetadataQueryResponse($segmentMetadataResponse);
+        $rawResponse = [
+            [
+                'id'               => 'myDataSource_2019-08-19T13:00:00.000Z/2019-08-19T14:00:00.000Z_2019-08-19T00:00:03.958Z',
+                'intervals'        => ['2019-08-19T13:00:00.000Z/2019-08-19T14:00:00.000Z',],
+                'columns'          =>
+                    [
+                        '__time'              =>
+                            [
+                                'typeSignature'     => 'LONG',
+                                'type'              => 'LONG',
+                                'hasMultipleValues' => false,
+                                'hasNulls'          => false,
+                                'size'              => 0,
+                                'cardinality'       => null,
+                                'minValue'          => null,
+                                'maxValue'          => null,
+                                'errorMessage'      => null,
+                            ],
+                        'iso'                 =>
+                            [
+                                'typeSignature'     => 'STRING',
+                                'type'              => 'STRING',
+                                'hasMultipleValues' => false,
+                                'hasNulls'          => false,
+                                'size'              => 0,
+                                'cardinality'       => 42,
+                                'minValue'          => 'be',
+                                'maxValue'          => 'zw',
+                                'errorMessage'      => null,
+                            ],
+                        'counter'      =>
+                            [
+                                'typeSignature'     => 'LONG',
+                                'type'              => 'LONG',
+                                'hasMultipleValues' => false,
+                                'hasNulls'          => false,
+                                'size'              => 0,
+                                'cardinality'       => null,
+                                'minValue'          => null,
+                                'maxValue'          => null,
+                                'errorMessage'      => null,
+                            ],
+                    ],
+                'size'             => 0,
+                'numRows'          => 151,
+                'aggregators'      => null,
+                'timestampSpec'    => null,
+                'queryGranularity' => null,
+                'rollup'           => null,
+            ],
+        ];
+
+        $responseObj = new SegmentMetadataQueryResponse($rawResponse);
 
         $queryBuilder->shouldReceive('segmentMetadata')
             ->once()
@@ -317,7 +365,73 @@ public function testGetColumnsForInterval(array $segmentMetadataResponse): void
             ->shouldAllowMockingProtectedMethods()
             ->getColumnsForInterval($dataSource, $interval);
 
-        $this->assertEquals($responseObj->data(), $response);
+        $expected = [];
+        array_walk($rawResponse[0]['columns'], function ($value, $key) use (&$expected) {
+            $expected[] = array_merge($value, ['field' => $key]);
+        });
+        $this->assertEquals($expected, $response);
+    }
+
+
+    /**
+     * @throws \GuzzleHttp\Exception\GuzzleException
+     * @throws \Level23\Druid\Exceptions\QueryResponseException
+     */
+    public function testRowCount(): void
+    {
+        $dataSource      = 'myDataSource';
+        $interval        = '2019-08-19T13:00:00.000Z/2019-08-19T14:00:00.000Z';
+        $metadataBuilder = Mockery::mock(MetadataBuilder::class, [$this->client]);
+        $metadataBuilder->makePartial();
+        $queryBuilder = Mockery::mock(QueryBuilder::class, [$this->client, 'myDataSource']);
+
+        $this->client->shouldReceive('query')
+            ->with($dataSource)
+            ->once()
+            ->andReturn($queryBuilder);
+
+        $queryBuilder->shouldReceive('interval')
+            ->once()
+            ->with($interval, null)
+            ->andReturn($queryBuilder);
+
+        $rawResponse = [
+            [
+                'id'               => 'myDataSource_2019-08-19T13:00:00.000Z/2019-08-19T14:00:00.000Z_2019-08-19T00:00:03.958Z',
+                'intervals'        => ['2019-08-19T13:00:00.000Z/2019-08-19T14:00:00.000Z',],
+                'columns'          => [],
+                'size'             => 0,
+                'numRows'          => 151,
+                'aggregators'      => null,
+                'timestampSpec'    => null,
+                'queryGranularity' => null,
+                'rollup'           => null,
+            ],
+            [
+                'id'               => 'myDataSource_2019-08-19T13:00:00.000Z/2019-08-19T15:00:00.000Z_2019-08-19T00:00:03.958Z',
+                'intervals'        => ['2019-08-19T13:00:00.000Z/2019-08-19T14:00:00.000Z',],
+                'columns'          => [],
+                'size'             => 0,
+                'numRows'          => 645,
+                'aggregators'      => null,
+                'timestampSpec'    => null,
+                'queryGranularity' => null,
+                'rollup'           => null,
+            ],
+        ];
+
+        $responseObj = new SegmentMetadataQueryResponse($rawResponse);
+
+        $queryBuilder->shouldReceive('segmentMetadata')
+            ->once()
+            ->andReturn($responseObj);
+
+        $response = $metadataBuilder
+            ->shouldAllowMockingProtectedMethods()
+            ->rowCount($dataSource, $interval);
+
+
+        $this->assertEquals(645 + 151, $response);
     }
 
     /**
@@ -591,5 +705,4 @@ public function testDataSources(): void
 
         $this->assertEquals(['wikipedia', 'clicks'], $response);
     }
-
 }