implement pdf content extraction

nextcloud · Dec 20, 2024 · f616b07 · f616b07
1 parent 3e4893e
commit f616b07
Show file tree

Hide file tree

Showing 4 changed files with 143 additions and 3 deletions.
diff --git a/composer.json b/composer.json
@@ -15,7 +15,8 @@
 		"erusev/parsedown": "^1.7",
 		"henck/rtf-to-html": "^1.2",
 		"html2text/html2text": "^4.3",
-		"phpoffice/phpword": "^1.2"
+		"phpoffice/phpword": "^1.2",
+		"smalot/pdfparser": "^2.10"
 	},
 	"scripts": {
 		"lint": "find . -name \\*.php -not -path './vendor/*' -print0 | xargs -0 -n1 php -l",

diff --git a/composer.lock b/composer.lock
diff --git a/lib/Service/AssistantService.php b/lib/Service/AssistantService.php
@@ -46,6 +46,7 @@
 use RtfHtmlPhp\Document;
 use RtfHtmlPhp\Html\HtmlFormatter;
 use RuntimeException;
+use Smalot\PdfParser\Parser;
 
 /**
  * @psalm-import-type AssistantTaskProcessingTaskType from ResponseDefinitions
@@ -613,6 +614,13 @@ public function parseTextFromFile(string $userId, ?string $filePath = null, ?int
 					$this->tempManager->clean();
 					break;
 				}
+			case 'application/pdf':
+				{
+					$parser = new Parser();
+					$pdf = $parser->parseContent($fileContent);
+					$text = $pdf->getText();
+					break;
+				}
 		}
 		return $text;
 	}

diff --git a/src/components/fields/TextInput.vue b/src/components/fields/TextInput.vue
@@ -60,7 +60,7 @@ const VALID_MIME_TYPES = [
 	'application/msword', // doc
 	'application/vnd.openxmlformats-officedocument.wordprocessingml.document', // docx
 	'application/vnd.oasis.opendocument.text', // odt
-	// 'application/pdf', // Not yet supported
+	'application/pdf', // pdf
 ]
 
 const picker = (callback, target) => getFilePickerBuilder(t('assistant', 'Choose a text file'))