cleanup tmp files due to pdftohtml tmp images

document_clipper/pdf.py

@@ -29,6 +29,17 @@ def __init__(self, pdf_file, pdf_image_to_text_method=None):
         self._pdf_to_xml = None
         self._pdf_image_to_text_method = pdf_image_to_text_method
 
+    def __enter__(self):
+        return self
+
+    def __exit__(self, exc_type, exc_value, traceback):
+        if self._pdf_to_xml:
+            try:
+                for image in self._pdf_to_xml.findAll('image'):
+                    os.remove(image['src'])
+            except:
+                logging.exception(u"Error cleaning up '%s'" % self.pdf_file.name)
+
     def _read_file(self):
         """
         :return: the contents of the PDF file loaded into the instance

tests/test_document_clipper_pdf.py

@@ -35,7 +35,6 @@ def _images_to_text_method_mocked(self):
 
     def test_pdf_to_xml_ok(self):
         pdf_to_xml = self.document_clipper_pdf_reader.pdf_to_xml()
-
         self.assertTrue(pdf_to_xml.is_xml)
         self.assertIsNotNone(pdf_to_xml.contents)
 
@@ -320,3 +319,24 @@ def test_fix_pdf_error(self, mock_os_remove):
         ret_file_path = self.document_clipper_pdf_writer.fix_pdf(invalid_path)
         self.assertEqual(ret_file_path, invalid_path)
         mock_os_remove.assert_not_called()
+
+    @patch('os.remove')
+    def test_cleaning_up_beacuse_of_pdf_to_xml_tmp_images(self, mock_remove):
+        self.pdf_file = open(PATH_TO_PDF_FILE_WITH_IMAGES)
+        with DocumentClipperPdfReader(self.pdf_file) as document_clipper_pdf_reader:
+            pdf_to_xml = document_clipper_pdf_reader.pdf_to_xml()
+
+        self.assertTrue(pdf_to_xml.is_xml)
+        self.assertIsNotNone(pdf_to_xml.contents)
+        # remove for images from tmp
+        self.assertEqual(len(mock_remove.call_args_list), 4)
+
+    @patch('os.remove')
+    def test_cleaning_up_beacuse_of_pdf_to_xml_tmp_images_nothing_to_clean(self, mock_remove):
+        with DocumentClipperPdfReader(self.pdf_file) as document_clipper_pdf_reader:
+            pdf_to_xml = document_clipper_pdf_reader.pdf_to_xml()
+
+        self.assertTrue(pdf_to_xml.is_xml)
+        self.assertIsNotNone(pdf_to_xml.contents)
+        # remove for images from tmp
+        self.assertEqual(len(mock_remove.call_args_list), 0)