Chore: Initial conversion to pytest fixtures (#7110)

src/paperless_text/parsers.py

@@ -1,4 +1,4 @@
-import os
+from pathlib import Path
 
 from django.conf import settings
 from PIL import Image
@@ -15,7 +15,7 @@ class TextDocumentParser(DocumentParser):
 
     logging_name = "paperless.parsing.text"
 
-    def get_thumbnail(self, document_path, mime_type, file_name=None):
+    def get_thumbnail(self, document_path: Path, mime_type, file_name=None) -> Path:
         text = self.read_file_handle_unicode_errors(document_path)
 
         img = Image.new("RGB", (500, 700), color="white")
@@ -27,7 +27,7 @@ class TextDocumentParser(DocumentParser):
         )
         draw.text((5, 5), text, font=font, fill="black")
 
-        out_path = os.path.join(self.tempdir, "thumb.webp")
+        out_path = self.tempdir / "thumb.webp"
         img.save(out_path, format="WEBP")
 
         return out_path

src/paperless_text/tests/conftest.py

@@ -0,0 +1,30 @@
+from collections.abc import Generator
+from pathlib import Path
+
+import pytest
+
+from paperless_text.parsers import TextDocumentParser
+
+
+@pytest.fixture(scope="session")
+def sample_dir() -> Path:
+    return (Path(__file__).parent / Path("samples")).resolve()
+
+
+@pytest.fixture()
+def text_parser() -> Generator[TextDocumentParser, None, None]:
+    try:
+        parser = TextDocumentParser(logging_group=None)
+        yield parser
+    finally:
+        parser.cleanup()
+
+
+@pytest.fixture(scope="session")
+def sample_txt_file(sample_dir: Path) -> Path:
+    return sample_dir / "test.txt"
+
+
+@pytest.fixture(scope="session")
+def malformed_txt_file(sample_dir: Path) -> Path:
+    return sample_dir / "decode_error.txt"

src/paperless_text/tests/test_parser.py

@@ -1,37 +1,26 @@
 from pathlib import Path
 
-from django.test import TestCase
-
-from documents.tests.utils import DirectoriesMixin
-from documents.tests.utils import FileSystemAssertsMixin
 from paperless_text.parsers import TextDocumentParser
 
 
-class TestTextParser(DirectoriesMixin, FileSystemAssertsMixin, TestCase):
-    SAMPLE_DIR = Path(__file__).resolve().parent / "samples"
-
-    def test_thumbnail(self):
-        parser = TextDocumentParser(None)
-
+class TestTextParser:
+    def test_thumbnail(self, text_parser: TextDocumentParser, sample_txt_file: Path):
         # just make sure that it does not crash
-        f = parser.get_thumbnail(
-            self.SAMPLE_DIR / "test.txt",
-            "text/plain",
-        )
-        self.assertIsFile(f)
+        f = text_parser.get_thumbnail(sample_txt_file, "text/plain")
+        assert f.exists()
+        assert f.is_file()
 
-    def test_parse(self):
-        parser = TextDocumentParser(None)
+    def test_parse(self, text_parser: TextDocumentParser, sample_txt_file: Path):
+        text_parser.parse(sample_txt_file, "text/plain")
 
-        parser.parse(
-            self.SAMPLE_DIR / "test.txt",
-            "text/plain",
-        )
+        assert text_parser.get_text() == "This is a test file.\n"
+        assert text_parser.get_archive_path() is None
 
-        self.assertEqual(parser.get_text(), "This is a test file.\n")
-        self.assertIsNone(parser.get_archive_path())
-
-    def test_parse_invalid_bytes(self):
+    def test_parse_invalid_bytes(
+        self,
+        text_parser: TextDocumentParser,
+        malformed_txt_file: Path,
+    ):
         """
         GIVEN:
             - Text file which contains invalid UTF bytes
@@ -41,12 +30,8 @@ class TestTextParser(DirectoriesMixin, FileSystemAssertsMixin, TestCase):
             - Parsing continues
             - Invalid bytes are removed
         """
-        parser = TextDocumentParser(None)
 
-        parser.parse(
-            self.SAMPLE_DIR / "decode_error.txt",
-            "text/plain",
-        )
+        text_parser.parse(malformed_txt_file, "text/plain")
 
-        self.assertEqual(parser.get_text(), "Pantothens<EFBFBD>ure\n")
-        self.assertIsNone(parser.get_archive_path())
+        assert text_parser.get_text() == "Pantothens<EFBFBD>ure\n"
+        assert text_parser.get_archive_path() is None