paradizelost/paperless-ngx
1
0
Fork 0
mirror of https://github.com/paperless-ngx/paperless-ngx.git synced 2025-07-30 18:27:45 -05:00
Code Issues Packages Projects Releases Wiki Activity

fixed most of the tests

Browse Source
This commit is contained in:
Jonas Winkler
2020-11-02 19:42:23 +01:00
parent 3a08a2d206
commit f182709fdd
3 changed files with 22 additions and 52 deletions
Download Patch File Download Diff File Expand all files Collapse all files

37
src/paperless_tesseract/tests/test_date.py
View File

@@ -5,7 +5,7 @@ from unittest import mock
from uuid import uuid4
from dateutil import tz
from django.test import TestCase
from django.test import TestCase, override_settings
from ..parsers import RasterisedDocumentParser
from django.conf import settings
@@ -16,36 +16,34 @@ class TestDate(TestCase):
SAMPLE_FILES = os.path.join(os.path.dirname(__file__), "samples")
SCRATCH = "/tmp/paperless-tests-{}".format(str(uuid4())[:8])
MOCK_SCRATCH = "paperless_tesseract.parsers.RasterisedDocumentParser.SCRATCH" # NOQA: E501
def setUp(self):
os.makedirs(self.SCRATCH, exist_ok=True)
def tearDown(self):
shutil.rmtree(self.SCRATCH)
@mock.patch(MOCK_SCRATCH, SCRATCH)
@override_settings(SCRATCH_DIR=SCRATCH)
def test_date_format_1(self):
input_file = os.path.join(self.SAMPLE_FILES, "")
document = RasterisedDocumentParser(input_file, None)
document._text = "lorem ipsum 130218 lorem ipsum"
self.assertEqual(document.get_date(), None)
@mock.patch(MOCK_SCRATCH, SCRATCH)
@override_settings(SCRATCH_DIR=SCRATCH)
def test_date_format_2(self):
input_file = os.path.join(self.SAMPLE_FILES, "")
document = RasterisedDocumentParser(input_file, None)
document._text = "lorem ipsum 2018 lorem ipsum"
self.assertEqual(document.get_date(), None)
@mock.patch(MOCK_SCRATCH, SCRATCH)
@override_settings(SCRATCH_DIR=SCRATCH)
def test_date_format_3(self):
input_file = os.path.join(self.SAMPLE_FILES, "")
document = RasterisedDocumentParser(input_file, None)
document._text = "lorem ipsum 20180213 lorem ipsum"
self.assertEqual(document.get_date(), None)
@mock.patch(MOCK_SCRATCH, SCRATCH)
@override_settings(SCRATCH_DIR=SCRATCH)
def test_date_format_4(self):
input_file = os.path.join(self.SAMPLE_FILES, "")
document = RasterisedDocumentParser(input_file, None)
@@ -59,7 +57,7 @@ class TestDate(TestCase):
)
)
@mock.patch(MOCK_SCRATCH, SCRATCH)
@override_settings(SCRATCH_DIR=SCRATCH)
def test_date_format_5(self):
input_file = os.path.join(self.SAMPLE_FILES, "")
document = RasterisedDocumentParser(input_file, None)
@@ -76,7 +74,7 @@ class TestDate(TestCase):
)
)
@mock.patch(MOCK_SCRATCH, SCRATCH)
@override_settings(SCRATCH_DIR=SCRATCH)
def test_date_format_6(self):
input_file = os.path.join(self.SAMPLE_FILES, "")
document = RasterisedDocumentParser(input_file, None)
@@ -93,7 +91,7 @@ class TestDate(TestCase):
)
self.assertEqual(document.get_date(), None)
@mock.patch(MOCK_SCRATCH, SCRATCH)
@override_settings(SCRATCH_DIR=SCRATCH)
def test_date_format_7(self):
input_file = os.path.join(self.SAMPLE_FILES, "")
document = RasterisedDocumentParser(input_file, None)
@@ -111,7 +109,7 @@ class TestDate(TestCase):
)
)
@mock.patch(MOCK_SCRATCH, SCRATCH)
@override_settings(SCRATCH_DIR=SCRATCH)
def test_date_format_8(self):
input_file = os.path.join(self.SAMPLE_FILES, "")
document = RasterisedDocumentParser(input_file, None)
@@ -135,7 +133,7 @@ class TestDate(TestCase):
)
)
@mock.patch(MOCK_SCRATCH, SCRATCH)
@override_settings(SCRATCH_DIR=SCRATCH)
def test_date_format_9(self):
input_file = os.path.join(self.SAMPLE_FILES, "")
document = RasterisedDocumentParser(input_file, None)
@@ -157,7 +155,7 @@ class TestDate(TestCase):
"paperless_tesseract.parsers.RasterisedDocumentParser.get_text",
return_value="01-07-0590 00:00:00"
)
@mock.patch(MOCK_SCRATCH, SCRATCH)
@override_settings(SCRATCH_DIR=SCRATCH)
def test_crazy_date_past(self, *args):
document = RasterisedDocumentParser("/dev/null", None)
document.get_text()
@@ -167,7 +165,7 @@ class TestDate(TestCase):
"paperless_tesseract.parsers.RasterisedDocumentParser.get_text",
return_value="01-07-2350 00:00:00"
)
@mock.patch(MOCK_SCRATCH, SCRATCH)
@override_settings(SCRATCH_DIR=SCRATCH)
def test_crazy_date_future(self, *args):
document = RasterisedDocumentParser("/dev/null", None)
document.get_text()
@@ -177,7 +175,7 @@ class TestDate(TestCase):
"paperless_tesseract.parsers.RasterisedDocumentParser.get_text",
return_value="20 408000l 2475"
)
@mock.patch(MOCK_SCRATCH, SCRATCH)
@override_settings(SCRATCH_DIR=SCRATCH)
def test_crazy_date_with_spaces(self, *args):
document = RasterisedDocumentParser("/dev/null", None)
document.get_text()
@@ -187,13 +185,8 @@ class TestDate(TestCase):
"paperless_tesseract.parsers.RasterisedDocumentParser.get_text",
return_value="No date in here"
)
@mock.patch(
"paperless_tesseract.parsers.RasterisedDocumentParser."
"FILENAME_DATE_ORDER",
new_callable=mock.PropertyMock,
return_value="YMD"
)
@mock.patch(MOCK_SCRATCH, SCRATCH)
@override_settings(FILENAME_DATE_ORDER="YMD")
@override_settings(SCRATCH_DIR=SCRATCH)
def test_filename_date_parse_invalid(self, *args):
document = RasterisedDocumentParser("/tmp/20 408000l 2475 - test.pdf", None)
document.get_text()

6
src/paperless_tesseract/tests/test_ocr.py
View File

@@ -62,10 +62,6 @@ class TestOCR(TestCase):
)
@skipIf(not TESSERACT_INSTALLED, "Tesseract not installed. Skipping")
@mock.patch(
"paperless_tesseract.parsers.RasterisedDocumentParser.SCRATCH",
SAMPLE_FILES
)
@mock.patch("paperless_tesseract.parsers.pyocr", FakePyOcr)
def test_image_to_string_with_text_free_page(self):
"""
@@ -77,4 +73,4 @@ class TestOCR(TestCase):
text-free pages are now handled correctly so long as we work around
this weird exception.
"""
image_to_string(["no-text.png", "en"])
image_to_string([os.path.join(self.SAMPLE_FILES, "no-text.png"), "en"])