Merge pull request #236 from PrimozGodec/replace-cachecontrol

[ENH] Replace cachecontrol with requests_cache

orangecontrib/imageanalytics/local_embedder.py

@@ -1,9 +1,3 @@
-from os.path import join
-
-import cachecontrol.caches
-import requests
-
-from Orange.canvas.config import cache_dir
 from Orange.misc.utils.embedder_utils import EmbedderCache
 from Orange.util import dummy_callback
 from orangecontrib.imageanalytics.utils.embedder_utils import ImageLoader
@@ -18,13 +12,6 @@ def __init__(self, model, model_settings):
 
         self._target_image_size = model_settings["target_image_size"]
 
-        self._session = cachecontrol.CacheControl(
-            requests.session(),
-            cache=cachecontrol.caches.FileCache(
-                join(cache_dir(), __name__ + ".ImageEmbedder.httpcache")
-            ),
-        )
-
         self._image_loader = ImageLoader()
         self._cache = EmbedderCache(model)
 

orangecontrib/imageanalytics/utils/embedder_utils.py

@@ -1,32 +1,33 @@
 import ftplib
 import logging
+from datetime import timedelta
 from io import BytesIO
 from os.path import join
 from urllib.error import URLError
 from urllib.parse import urlparse
 from urllib.request import urlopen
 
-import cachecontrol.caches
-import numpy as np
-import requests
+from AnyQt.QtCore import QStandardPaths
 from PIL import ImageFile
 from PIL.Image import LANCZOS
 from PIL.Image import open as open_image
 from requests.exceptions import RequestException
-
-from Orange.misc.environ import cache_dir
+from requests_cache import CachedSession
 
 log = logging.getLogger(__name__)
 ImageFile.LOAD_TRUNCATED_IMAGES = True
 
 
 class ImageLoader:
     def __init__(self):
-        self._session = cachecontrol.CacheControl(
-            requests.session(),
-            cache=cachecontrol.caches.FileCache(
-                join(cache_dir(), __name__ + ".ImageEmbedder.httpcache")
-            ),
+        cache_dir = QStandardPaths.writableLocation(QStandardPaths.CacheLocation)
+        cache_path = join(cache_dir, "networkcache", "image_loader.sqlite")
+        self._session = CachedSession(
+            cache_path,
+            backend="sqlite",
+            cache_control=True,
+            expire_after=timedelta(days=1),
+            stale_if_error=True,
         )
 
     def load_image_or_none(self, file_path, target_size=None):
@@ -65,7 +66,7 @@ def _load_image_from_url_or_local_path(self, file_path):
         urlparts = urlparse(file_path)
         if urlparts.scheme in ("http", "https"):
             try:
-                file = self._session.get(file_path, stream=True).raw
+                file = BytesIO(self._session.get(file_path).content)
             except RequestException:
                 log.warning("Image skipped", exc_info=True)
                 return None

orangecontrib/imageanalytics/utils/tests/test_embedder_utils.py

@@ -3,7 +3,9 @@
 from unittest.mock import patch
 from urllib.error import URLError
 
+import numpy as np
 from PIL.Image import Image
+from numpy.testing import assert_array_equal
 from requests import RequestException
 
 from orangecontrib.imageanalytics.utils.embedder_utils import ImageLoader
@@ -49,6 +51,10 @@ def test_load_images_url(self) -> None:
         """
         image = self.image_loader.load_image_or_none(self.im_url)
         self.assertTrue(isinstance(image, Image))
+        image_array = np.array(image)
+        # manually checked values
+        assert_array_equal(image_array[147, 181], [103, 10, 0])
+        assert_array_equal(image_array[305, 331], [21, 1, 2])
 
         image = self.image_loader.load_image_or_none(self.im_paths[0],
                                                      target_size=(255, 255))
@@ -59,7 +65,7 @@ def test_load_images_url(self) -> None:
         image = self.image_loader.load_image_or_none(self.im_url + "a")
         self.assertIsNone(image)
 
-    @patch("requests.sessions.Session.get", side_effect=RequestException)
+    @patch("requests_cache.CachedSession.get", side_effect=RequestException)
     def test_load_images_url_request_exception(self, _) -> None:
         """
         Handle loading images from http, https type urls

setup.py

@@ -93,7 +93,6 @@ def include_documentation(local_dir, install_dir):
         classifiers=CLASSIFIERS,
         install_requires=[
             "AnyQt",
-            "cachecontrol[filecache]",
             "ndf >=0.1.4",
             "numpy >=1.16",
             "Orange3 >=3.34.0",
@@ -104,6 +103,7 @@ def include_documentation(local_dir, install_dir):
             # constraint when requiring Orange> 3.35
             "pandas <2.1",
             "requests",
+            "requests_cache",
             "scipy",
         ],
         extras_require={