Implement #130 (#160) (#161)

Only create new file version if OCR result was not empty Signed-off-by: Robin Windey <ro.windey@gmail.com> Signed-off-by: Robin Windey <ro.windey@gmail.com> Signed-off-by: Robin Windey <ro.windey@gmail.com>
R0Wi-DEV · Nov 1, 2022 · ffce79d · ffce79d
1 parent 221f3f2
commit ffce79d
Show file tree

Hide file tree

Showing 2 changed files with 41 additions and 7 deletions.
diff --git a/lib/BackgroundJobs/ProcessFileJob.php b/lib/BackgroundJobs/ProcessFileJob.php
@@ -184,19 +184,21 @@ private function processFile(string $filePath, WorkflowSettings $settings) : voi
 			return;
 		}
 
-
 		$fileContent = $ocrFile->getFileContent();
 		$nodeId = $node->getId();
 		$originalFileExtension = $node->getExtension();
 		$newFileExtension = $ocrFile->getFileExtension();
 
-		if ($originalFileExtension === $newFileExtension) {
-			$this->createNewFileVersion($filePath, $fileContent, $nodeId);
-			$this->eventService->textRecognized($ocrFile, $node);
-		} else {
-			$this->createNewFileVersion($filePath.".pdf", $fileContent, $nodeId);
-			$this->eventService->textRecognized($ocrFile, $node);
+		// Only create a new file version if the file OCR result was not empty #130
+		if ($ocrFile->getRecognizedText() !== '') {
+			$newFilePath = $originalFileExtension === $newFileExtension ?
+				$filePath :
+				$filePath . ".pdf";
+
+			$this->createNewFileVersion($newFilePath, $fileContent, $nodeId);
 		}
+
+		$this->eventService->textRecognized($ocrFile, $node);
 	}
 
 	private function getNode(string $filePath) : ?Node {

diff --git a/tests/Unit/BackgroundJobs/ProcessFileJobTest.php b/tests/Unit/BackgroundJobs/ProcessFileJobTest.php
@@ -403,6 +403,38 @@ public function testCallsProcessingFileAccessor(array $arguments, string $user,
 		$this->assertEquals(1, $calledWithNull);
 	}
 
+	/**
+	 * @dataProvider dataProvider_ValidArguments
+	 */
+	public function testDoesNotCreateNewFileVersionIfOcrContentWasEmpty(array $arguments, string $user, string $rootFolderPath, string $originalFileExtension, string $expectedOcrFilename) {
+		$this->processFileJob->setArgument($arguments);
+		$mimeType = 'application/pdf';
+		$content = 'someFileContent';
+		$ocrContent = '';
+		$ocrResult = new OcrProcessorResult($ocrContent, "pdf", $ocrContent);
+
+		$fileMock = $this->createValidFileMock($mimeType, $content);
+		$this->rootFolder->method('get')
+			->willReturn($fileMock);
+
+		$this->ocrService->expects($this->once())
+			->method('ocrFile')
+			->willReturn($ocrResult);
+
+		$viewMock = $this->createMock(IView::class);
+		$this->viewFactory->expects($this->never())
+			->method('create')
+			->willReturn($viewMock);
+
+		$this->processingFileAccessor->expects($this->never())
+			->method('setCurrentlyProcessedFileId');
+
+		$this->eventService->expects($this->once())
+			->method('textRecognized');
+
+		$this->processFileJob->execute($this->jobList);
+	}
+
 	public function dataProvider_InvalidArguments() {
 		$arr = [
 			[null, 1],