Merge pull request #5 from HelgeSverre/feature/cleanup-textract-s3

Add hook to enable deletion of files after processing with Textract

README.md

@@ -426,6 +426,28 @@ return [
 TEXTRACT_DISK="uploads"
 ```
 
+### Delete the files after processing them with Textract
+
+#### Using S3 Lifecycle Rules
+
+You can configure a lifecycle rule on your S3 bucket to delete the files after a certain amount of time, see the AWS
+docs for more info:
+
+https://repost.aws/knowledge-center/s3-empty-bucket-lifecycle-rule
+
+#### Using the `cleanupFileUsing` hook
+
+By default, the package will __NOT__ delete the files that has been uploaded in the textract S3 bucket, if you want to
+delete these files, you can implement this using the `TextractUsingS3Upload::cleanupFileUsing(Closure)` hook.
+
+```php
+// Delete the file from the S3 bucket
+TextractUsingS3Upload::cleanupFileUsing(function (string $filePath) {
+    Storage::disk('textract')->delete($filePath);
+}
+```
+
+
 **Note**
 
 Textract is not available in all regions:

composer.json

@@ -34,6 +34,7 @@
     },
     "require-dev": {
         "laravel/pint": "^1.0",
+        "mockery/mockery": "^1.6",
         "nunomaduro/collision": "^7.8",
         "orchestra/testbench": "^8.8",
         "pestphp/pest": "^2.0",

src/Text/Loaders/Textract/TextractUsingS3Upload.php

@@ -18,10 +18,17 @@ class TextractUsingS3Upload implements TextLoader
 {
     protected static mixed $generateFilePathUsing = null;
 
+    protected static mixed $cleanupFileUsing = null;
+
     public function __construct(protected TextractService $textractService)
     {
     }
 
+    public static function cleanupFileUsing(callable $callback): void
+    {
+        static::$cleanupFileUsing = $callback;
+    }
+
     public static function generateFilePathUsing(callable $callback): void
     {
         static::$generateFilePathUsing = $callback;
@@ -41,6 +48,16 @@ public function getFilePath(): string
         return $this->defaultFilePathGenerator();
     }
 
+    /** @noinspection PhpInconsistentReturnPointsInspection */
+    public function cleanup(string $path)
+    {
+        if (static::$cleanupFileUsing) {
+            return (static::$cleanupFileUsing)($path);
+        }
+
+        // No cleanup by default
+    }
+
     /**
      * @throws TextractTimedOut
      * @throws TextractConfigNotFoundException
@@ -70,12 +87,14 @@ public function load(mixed $data): ?TextContent
             throw new TextractStorageException("Could not create the file in the textract s3 bucket with path '{$path}'.");
         }
 
-        return new TextContent(
-            $this->textractService->s3ObjectToText(
-                s3Object: new S3Object(bucket: $bucket, name: $path),
-                timeoutInSeconds: config('extractor.textract_timeout'),
-                pollingIntervalInSeconds: config('extractor.textract_polling_interval')
-            )
+        $result = $this->textractService->s3ObjectToText(
+            s3Object: new S3Object(bucket: $bucket, name: $path),
+            timeoutInSeconds: config('extractor.textract_timeout'),
+            pollingIntervalInSeconds: config('extractor.textract_polling_interval')
         );
+
+        $this->cleanup($path);
+
+        return new TextContent($result);
     }
 }

tests/Text/TextLoaderTest.php

@@ -1,7 +1,10 @@
 <?php
 
 use HelgeSverre\Extractor\Facades\Text;
+use HelgeSverre\Extractor\Text\Loaders\Textract\TextractService;
+use HelgeSverre\Extractor\Text\Loaders\Textract\TextractUsingS3Upload;
 use HelgeSverre\Extractor\Text\TextContent;
+use Illuminate\Support\Facades\Storage;
 
 it('Can load Text', function () {
     $text = Text::text(file_get_contents(__DIR__.'/../samples/wolt-pizza-norwegian.txt'));
@@ -94,3 +97,47 @@
         'Termination of the Agreement',
     );
 });
+
+it('removes the file from S3 using the provided cleanup callback', function () {
+    Storage::fake('s3');
+
+    $mock = Mockery::mock(TextractService::class);
+    $mock->shouldReceive('s3ObjectToText');
+
+    $textractUsingS3Upload = new TextractUsingS3Upload($mock);
+
+    $testFilePath = 'extractor/test-file.pdf';
+
+    Storage::disk('s3')->put($testFilePath, 'Test content');
+
+    Storage::disk('s3')->assertExists($testFilePath);
+
+    TextractUsingS3Upload::cleanupFileUsing(function ($path) {
+        Storage::disk('s3')->delete($path);
+    });
+
+    $textractUsingS3Upload->cleanup($testFilePath);
+
+    Storage::disk('s3')->assertMissing($testFilePath);
+});
+
+it('overrides the default file path generation with a custom callback', function () {
+    Storage::fake('s3');
+
+    $mock = Mockery::mock(TextractService::class);
+    $mock->shouldReceive('s3ObjectToText');
+
+    $textractUsingS3Upload = new TextractUsingS3Upload($mock);
+
+    $customFilePath = 'custom-path/custom-file.pdf';
+
+    TextractUsingS3Upload::generateFilePathUsing(function () use ($customFilePath) {
+        return $customFilePath;
+    });
+
+    expect($textractUsingS3Upload->getFilePath())->toBe($customFilePath);
+
+    Storage::disk('s3')->put($textractUsingS3Upload->getFilePath(), 'Test content');
+
+    Storage::disk('s3')->assertExists($customFilePath);
+});