Add a new DataFrame::renameEach()

Author: stloyd

src/core/etl/src/Flow/ETL/DataFrame.php

@@ -41,12 +41,13 @@
     OrderEntriesTransformer,
     OrderEntries\Comparator,
     OrderEntries\TypeComparator,
-    RenameAllCaseTransformer,
+    RenameEachTransformer,
     RenameEntryTransformer,
     RenameStrReplaceAllEntriesTransformer,
     ScalarFunctionFilterTransformer,
     ScalarFunctionTransformer,
     SelectEntriesTransformer,
+    StyleConverter\RenameStrategy,
     UntilTransformer,
     WindowFunctionTransformer};
 use Flow\Filesystem\Path\Filter;
@@ -84,7 +85,8 @@ public function autoCast() : self
      * Merge/Split Rows yielded by Extractor into batches of given size.
      * For example, when Extractor is yielding one row at time, this method will merge them into batches of given size
      * before passing them to the next pipeline element.
-     * Similarly when Extractor is yielding batches of rows, this method will split them into smaller batches of given size.
+     * Similarly when Extractor is yielding batches of rows, this method will split them into smaller batches of given
+     * size.
      *
      * In order to merge all Rows into a single batch use DataFrame::collect() method or set size to -1 or 0.
      *
@@ -210,7 +212,8 @@ public function crossJoin(self $dataFrame, string $prefix = '') : self
 
     /**
      * @param int $limit maximum numbers of rows to display
-     * @param bool|int $truncate false or if set to 0 columns are not truncated, otherwise default truncate to 20 characters
+     * @param bool|int $truncate false or if set to 0 columns are not truncated, otherwise default truncate to 20
+     *                           characters
      * @param Formatter $formatter
      *
      * @trigger
@@ -258,7 +261,8 @@ public function dropDuplicates(string|Reference ...$entries) : self
     }
 
     /**
-     * Drop all partitions from Rows, additionally when $dropPartitionColumns is set to true, partition columns are also removed.
+     * Drop all partitions from Rows, additionally when $dropPartitionColumns is set to true, partition columns are
+     * also removed.
      *
      * @lazy
      */
@@ -636,10 +640,12 @@ public function renameAll(string $search, string $replace) : self
 
     /**
      * @lazy
+     *
+     * @deprecated use DataFrame::renameEach() with a selected RenameStrategy
      */
     public function renameAllLowerCase() : self
     {
-        $this->pipeline->add(new RenameAllCaseTransformer(lower: true));
+        $this->renameEach(RenameStrategy::LOWER);
 
         return $this;
     }
@@ -658,30 +664,43 @@ public function renameAllStyle(StringStyles|string $style) : self
 
     /**
      * @lazy
+     *
+     * @deprecated use DataFrame::renameEach() with a selected RenameStrategy
      */
     public function renameAllUpperCase() : self
     {
-        $this->pipeline->add(new RenameAllCaseTransformer(upper: true));
+        $this->renameEach(RenameStrategy::UPPER);
 
         return $this;
     }
 
     /**
      * @lazy
+     *
+     * @deprecated use DataFrame::renameEach() with a selected RenameStrategy
      */
     public function renameAllUpperCaseFirst() : self
     {
-        $this->pipeline->add(new RenameAllCaseTransformer(ucfirst: true));
+        $this->renameEach(RenameStrategy::UCFIRST);
 
         return $this;
     }
 
     /**
      * @lazy
+     *
+     * @deprecated use DataFrame::renameEach() with a selected RenameStrategy
      */
     public function renameAllUpperCaseWord() : self
     {
-        $this->pipeline->add(new RenameAllCaseTransformer(ucwords: true));
+        $this->renameEach(RenameStrategy::UCWORDS);
+
+        return $this;
+    }
+
+    public function renameEach(RenameStrategy $strategy) : self
+    {
+        $this->pipeline->add(new RenameEachTransformer($strategy));
 
         return $this;
     }
@@ -825,8 +844,8 @@ public function transform(Transformer|Transformation|Transformations|WithEntry $
     }
 
     /**
-     * The difference between filter and until is that filter will keep filtering rows until extractors finish yielding rows.
-     * Until will send a STOP signal to the Extractor when the condition is not met.
+     * The difference between filter and until is that filter will keep filtering rows until extractors finish yielding
+     * rows. Until will send a STOP signal to the Extractor when the condition is not met.
      *
      * @lazy
      */

src/core/etl/src/Flow/ETL/Pipeline/Optimizer/LimitOptimization.php

@@ -8,7 +8,16 @@
 use Flow\ETL\Function\ScalarFunction\ExpandResults;
 use Flow\ETL\{Loader, Pipeline, Transformer};
 use Flow\ETL\Pipeline\{BatchingPipeline, CollectingPipeline, LinkedPipeline, SynchronousPipeline, VoidPipeline};
-use Flow\ETL\Transformer\{CallbackRowTransformer, DropEntriesTransformer, EntryNameStyleConverterTransformer, LimitTransformer, RenameAllCaseTransformer, RenameEntryTransformer, RenameStrReplaceAllEntriesTransformer, ScalarFunctionTransformer, SelectEntriesTransformer};
+use Flow\ETL\Transformer\{CallbackRowTransformer,
+    DropEntriesTransformer,
+    EntryNameStyleConverterTransformer,
+    LimitTransformer,
+    RenameAllCaseTransformer,
+    RenameEachTransformer,
+    RenameEntryTransformer,
+    RenameStrReplaceAllEntriesTransformer,
+    ScalarFunctionTransformer,
+    SelectEntriesTransformer};
 
 final class LimitOptimization implements Optimization
 {
@@ -27,6 +36,7 @@ final class LimitOptimization implements Optimization
         SelectEntriesTransformer::class,
         DropEntriesTransformer::class,
         RenameAllCaseTransformer::class,
+        RenameEachTransformer::class,
         RenameEntryTransformer::class,
         RenameStrReplaceAllEntriesTransformer::class,
         LimitTransformer::class,

src/core/etl/src/Flow/ETL/Transformer/RenameAllCaseTransformer.php

@@ -6,6 +6,9 @@
 
 use Flow\ETL\{FlowContext, Row, Rows, Transformer};
 
+/**
+ * @deprecated use RenameAllTransformer with a selected RenameStrategy
+ */
 final readonly class RenameAllCaseTransformer implements Transformer
 {
     public function __construct(

src/core/etl/src/Flow/ETL/Transformer/RenameEachTransformer.php

@@ -0,0 +1,74 @@
+<?php
+
+declare(strict_types=1);
+
+namespace Flow\ETL\Transformer;
+
+use Flow\ETL\{FlowContext, Row, Rows, Transformer, Transformer\StyleConverter\RenameStrategy};
+
+final readonly class RenameEachTransformer implements Transformer
+{
+    public function __construct(
+        private RenameStrategy $strategy,
+    ) {
+    }
+
+    public function transform(Rows $rows, FlowContext $context) : Rows
+    {
+        return $rows->map(function (Row $row) : Row {
+            foreach ($row->entries()->all() as $entry) {
+                $row = match ($this->strategy) {
+                    RenameStrategy::LOWER => $row->rename($entry->name(), \mb_strtolower($entry->name())),
+                    RenameStrategy::UPPER => $row->rename($entry->name(), \mb_strtoupper($entry->name())),
+                    RenameStrategy::UCFIRST => $row->rename($entry->name(), $this->ucFirst($entry->name())),
+                    RenameStrategy::UCWORDS => $row->rename($entry->name(), $this->ucWords($entry->name())),
+                    RenameStrategy::TRANSLITERATE => $row->rename($entry->name(), $this->transliterate($entry->name())),
+                };
+            }
+
+            return $row;
+        });
+    }
+
+    private function transliterate(string $string) : string
+    {
+        if (\function_exists('transliterator_transliterate')) {
+            return (string) \transliterator_transliterate('Any-Latin; Latin-ASCII; Lower()', $string);
+        }
+
+        return $string;
+    }
+
+    private function ucFirst(string $string) : string
+    {
+        // Available from PHP 8.4+
+        if (\function_exists('mb_ucfirst')) {
+            return \mb_ucfirst($string);
+        }
+
+        $encoding = \mb_internal_encoding();
+
+        return \mb_strtoupper(\mb_substr($string, 0, 1, $encoding), $encoding) . \mb_substr($string, 1, null, $encoding);
+    }
+
+    private function ucWords(string $string) : string
+    {
+        $result = '';
+        $previousCharacter = ' ';
+
+        $encoding = \mb_internal_encoding();
+
+        for ($i = 0, $length = \mb_strlen($string, $encoding); $i < $length; $i++) {
+            $currentCharacter = \mb_substr($string, $i, 1, $encoding);
+
+            if (' ' === $previousCharacter) {
+                $currentCharacter = \mb_strtoupper($currentCharacter, $encoding);
+            }
+
+            $result .= $currentCharacter;
+            $previousCharacter = $currentCharacter;
+        }
+
+        return $result;
+    }
+}

src/core/etl/src/Flow/ETL/Transformer/StyleConverter/RenameStrategy.php

@@ -0,0 +1,14 @@
+<?php
+
+declare(strict_types=1);
+
+namespace Flow\ETL\Transformer\StyleConverter;
+
+enum RenameStrategy : string
+{
+    case LOWER = 'LOWER';
+    case TRANSLITERATE = 'TRANSLITERATE';
+    case UCFIRST = 'UCFIRST';
+    case UCWORDS = 'UCWORDS';
+    case UPPER = 'UPPER';
+}

src/core/etl/tests/Flow/ETL/Tests/Integration/DataFrame/RenameTest.php

@@ -6,7 +6,7 @@
 
 use function Flow\ETL\DSL\{bool_entry, df, from_rows, int_entry, json_entry, ref, str_entry};
 use function Flow\ETL\DSL\{row, rows};
-use Flow\ETL\{Function\StyleConverter\StringStyles};
+use Flow\ETL\{Function\StyleConverter\StringStyles, Transformer\StyleConverter\RenameStrategy};
 use Flow\ETL\Tests\FlowIntegrationTestCase;
 
 final class RenameTest extends FlowIntegrationTestCase
@@ -98,6 +98,24 @@ public function test_rename_all_to_snake_case() : void
         );
     }
 
+    public function test_rename_all_transliterate() : void
+    {
+        $rows = rows(row(int_entry('ÓSMY', 8)), row(int_entry('DZIEWIĄTY', 9)));
+
+        $ds = df()
+            ->read(from_rows($rows))
+            ->renameEach(RenameStrategy::TRANSLITERATE)
+            ->getEachAsArray();
+
+        self::assertEquals(
+            [
+                ['osmy' => 8],
+                ['dziewiaty' => 9],
+            ],
+            \iterator_to_array($ds)
+        );
+    }
+
     public function test_rename_all_upper_case() : void
     {
         $rows = rows(row(int_entry('id', 1), str_entry('name', 'name'), bool_entry('active', true)), row(int_entry('id', 2), str_entry('name', 'name'), bool_entry('active', false)));