[tests] Improve access to datadir.montgomery rc variable throughout the test units

doc/catalog.json

 {
   "clapp": {
     "versions": {
-      "stable": "https://www.idiap.ch/software/biosignal/docs/software/clapp/stable/sphinx/",
-      "latest": "https://www.idiap.ch/software/biosignal/docs/software/clapp/main/sphinx/"
+      "stable": "https://clapp.readthedocs.io/en/stable/",
+      "latest": "https://clapp.readthedocs.io/en/latest/"
     },
     "sources": {}
   }

src/ptbench/data/hivtb/__init__.py

@@ -34,12 +34,12 @@ _protocols = [
     importlib.resources.files(__name__).joinpath("fold_9.json.bz2"),
 ]
 
-_root_path = load_rc().get("datadir.hivtb", os.path.realpath(os.curdir))
+_datadir = load_rc().get("datadir.hivtb", os.path.realpath(os.curdir))
 
 
 def _raw_data_loader(sample):
     return dict(
-        data=load_pil_baw(os.path.join(_root_path, sample["data"])),
+        data=load_pil_baw(os.path.join(_datadir, sample["data"])),
         label=sample["label"],
     )
 

src/ptbench/data/indian/__init__.py

@@ -34,12 +34,12 @@ _protocols = [
     importlib.resources.files(__name__).joinpath("fold_9.json.bz2"),
 ]
 
-_root_path = load_rc().get("datadir.indian", os.path.realpath(os.curdir))
+_datadir = load_rc().get("datadir.indian", os.path.realpath(os.curdir))
 
 
 def _raw_data_loader(sample):
     return dict(
-        data=load_pil_baw(os.path.join(_root_path, sample["data"])),
+        data=load_pil_baw(os.path.join(_datadir, sample["data"])),
         label=sample["label"],
     )
 

src/ptbench/data/montgomery/__init__.py

@@ -40,25 +40,19 @@ _protocols = [
     importlib.resources.files(__name__).joinpath("fold_9.json.bz2"),
 ]
 
-_root_path = None
+_datadir = load_rc().get("datadir.montgomery", os.path.realpath(os.curdir))
 
 
 def _raw_data_loader(sample):
-    # hack to allow tests to change "datadir.montgomery"
-    global _root_path
-    _root_path = _root_path or load_rc().get(
-        "datadir.montgomery", os.path.realpath(os.curdir)
-    )
-
     return dict(
-        data=load_pil_baw(os.path.join(_root_path, sample["data"])),  # type: ignore
+        data=load_pil_baw(os.path.join(_datadir, sample["data"])),  # type: ignore
         label=sample["label"],
     )
 
 
 def _loader(context, sample):
     # "context" is ignored in this case - database is homogeneous
-    # we returned delayed samples to avoid loading all images at once
+    # we return delayed samples to avoid loading all images at once
     return make_delayed(sample, _raw_data_loader)
 
 

src/ptbench/data/nih_cxr14_re/__init__.py

@@ -38,7 +38,7 @@ _protocols = [
     importlib.resources.files(__name__).joinpath("cardiomegaly.json.bz2"),
 ]
 
-_root_path = load_rc().get("datadir.nih_cxr14_re", os.path.realpath(os.curdir))
+_datadir = load_rc().get("datadir.nih_cxr14_re", os.path.realpath(os.curdir))
 _idiap_folders = load_rc().get("nih_cxr14_re.idiap_folder_structure", False)
 
 
@@ -51,7 +51,7 @@ def _raw_data_loader(sample):
         return dict(
             data=load_pil_rgb(
                 os.path.join(
-                    _root_path,
+                    _datadir,
                     os.path.dirname(sample["data"]),
                     basename[:5],
                     basename,
@@ -61,7 +61,7 @@ def _raw_data_loader(sample):
         )
     else:
         return dict(
-            data=load_pil_rgb(os.path.join(_root_path, sample["data"])),
+            data=load_pil_rgb(os.path.join(_datadir, sample["data"])),
             label=sample["label"],
         )
 

src/ptbench/data/padchest/__init__.py

@@ -236,12 +236,12 @@ _protocols = [
     importlib.resources.files(__name__).joinpath("cardiomegaly_idiap.json.bz2"),
 ]
 
-_root_path = load_rc().get("datadir.padchest", os.path.realpath(os.curdir))
+_datadir = load_rc().get("datadir.padchest", os.path.realpath(os.curdir))
 
 
 def _raw_data_loader(sample):
     return dict(
-        data=load_pil(os.path.join(_root_path, sample["data"])),
+        data=load_pil(os.path.join(_datadir, sample["data"])),
         label=sample["label"],
     )
 

src/ptbench/data/shenzhen/__init__.py

@@ -41,12 +41,12 @@ _protocols = [
     importlib.resources.files(__name__).joinpath("fold_9.json.bz2"),
 ]
 
-_root_path = load_rc().get("datadir.shenzhen", os.path.realpath(os.curdir))
+_datadir = load_rc().get("datadir.shenzhen", os.path.realpath(os.curdir))
 
 
 def _raw_data_loader(sample):
     return dict(
-        data=load_pil_baw(os.path.join(_root_path, sample["data"])),
+        data=load_pil_baw(os.path.join(_datadir, sample["data"])),
         label=sample["label"],
     )
 

src/ptbench/data/tbpoc/__init__.py

@@ -34,12 +34,12 @@ _protocols = [
     importlib.resources.files(__name__).joinpath("fold_9.json.bz2"),
 ]
 
-_root_path = load_rc().get("datadir.tbpoc", os.path.realpath(os.curdir))
+_datadir = load_rc().get("datadir.tbpoc", os.path.realpath(os.curdir))
 
 
 def _raw_data_loader(sample):
     return dict(
-        data=load_pil_baw(os.path.join(_root_path, sample["data"])),
+        data=load_pil_baw(os.path.join(_datadir, sample["data"])),
         label=sample["label"],
     )
 

tests/conftest.py

@@ -2,9 +2,12 @@
 #
 # SPDX-License-Identifier: GPL-3.0-or-later
 
+import os
 import pathlib
+import typing
 
 import pytest
+import tomli_w
 
 
 @pytest.fixture
@@ -67,30 +70,56 @@ def temporary_basedir(tmp_path_factory):
     return tmp_path_factory.mktemp("test-cli")
 
 
-@pytest.fixture(scope="session")
-def montgomery_datadir(tmp_path_factory) -> pathlib.Path:
-    from ptbench.utils.rc import load_rc
-
-    database_dir = load_rc().get("datadir.montgomery")
-    if database_dir is not None:
-        return pathlib.Path(database_dir)
-
-    # else, we must extract the LFS component
-    archive = (
-        pathlib.Path(__file__).parents[0] / "data" / "lfs" / "test-database.zip"
-    )
-    assert archive.exists(), (
-        f"Neither datadir.montgomery is set on the global configuration, "
-        f"(typically ~/.config/ptbench.toml), or it is possible to detect "
-        f"the presence of {archive}' (did you git submodule init --update "
-        f"this submodule?)"
-    )
-
-    database_dir = tmp_path_factory.mktemp("montgomery_datadir")
+@pytest.fixture(scope="session", autouse=True)
+def ensure_montgomery(
+    tmp_path_factory,
+) -> typing.Generator[None, None, None]:
+    """A pytest fixture that ensures that datadir.montgomery is always
+    available."""
 
+    import tempfile
     import zipfile
 
-    with zipfile.ZipFile(archive) as zf:
-        zf.extractall(database_dir)
+    from ptbench.utils.rc import load_rc
+
+    rc = load_rc()
 
-    return database_dir
+    database_dir = rc.get("datadir.montgomery")
+    if database_dir is not None:
+        # if the user downloaded it, use that copy
+        yield
+
+    else:
+        # else, we must extract the LFS component (we are likely on the CI)
+        archive = (
+            pathlib.Path(__file__).parents[0]
+            / "data"
+            / "lfs"
+            / "test-database.zip"
+        )
+        assert archive.exists(), (
+            f"Neither datadir.montgomery is set on the global configuration, "
+            f"(typically ~/.config/ptbench.toml), or it is possible to detect "
+            f"the presence of {archive}' (did you git submodule init --update "
+            f"this submodule?)"
+        )
+
+        database_dir = tmp_path_factory.mktemp("montgomery_datadir")
+        rc.setdefault("datadir.montgomery", str(database_dir))
+
+        with zipfile.ZipFile(archive) as zf:
+            zf.extractall(database_dir)
+
+        with tempfile.TemporaryDirectory() as tmpdir:
+            config_filename = "ptbench.toml"
+            with open(os.path.join(tmpdir, config_filename), "wb") as f:
+                tomli_w.dump(rc.data, f)
+                f.flush()
+
+            old_config_home = os.environ.get("XDG_CONFIG_HOME")
+            os.environ["XDG_CONFIG_HOME"] = tmpdir
+            yield
+            if old_config_home is None:
+                del os.environ["XDG_CONFIG_HOME"]
+            else:
+                os.environ["XDG_CONFIG_HOME"] = old_config_home

tests/test_config.py

@@ -2,13 +2,8 @@
 #
 # SPDX-License-Identifier: GPL-3.0-or-later
 
-import contextlib
-import os
-import tempfile
-
 import numpy as np
 import pytest
-import tomli_w
 import torch
 
 from torch.utils.data import ConcatDataset
@@ -21,23 +16,6 @@ from ptbench.configs.datasets import get_positive_weights, get_samples_weights
 N = 10
 
 
-@contextlib.contextmanager
-def rc_context(**new_config):
-    with tempfile.TemporaryDirectory() as tmpdir:
-        config_filename = "ptbench.toml"
-        with open(os.path.join(tmpdir, config_filename), "wb") as f:
-            tomli_w.dump(new_config, f)
-            f.flush()
-        old_config_home = os.environ.get("XDG_CONFIG_HOME")
-        os.environ["XDG_CONFIG_HOME"] = tmpdir
-        yield
-        if old_config_home is None:
-            del os.environ["XDG_CONFIG_HOME"]
-        else:
-            os.environ["XDG_CONFIG_HOME"] = old_config_home
-
-
-@pytest.mark.skip_if_rc_var_not_set("datadir.montgomery")
 def test_montgomery():
     def _check_subset(samples, size):
         assert len(samples) == size
@@ -60,20 +38,15 @@ def test_montgomery():
     _check_subset(dataset["test"], 28)
 
 
-def test_get_samples_weights(montgomery_datadir):
-    # Temporarily modify Montgomery datadir
-    new_value = {"datadir.montgomery": str(montgomery_datadir)}
-    with rc_context(**new_value):
-        from ptbench.configs.datasets.montgomery.default import dataset
+def test_get_samples_weights():
+    from ptbench.configs.datasets.montgomery.default import dataset
 
-        train_samples_weights = get_samples_weights(
-            dataset["__train__"]
-        ).numpy()
+    train_samples_weights = get_samples_weights(dataset["__train__"]).numpy()
 
-        unique, counts = np.unique(train_samples_weights, return_counts=True)
+    unique, counts = np.unique(train_samples_weights, return_counts=True)
 
-        np.testing.assert_equal(counts, np.array([51, 37]))
-        np.testing.assert_equal(unique, np.array(1 / counts, dtype=np.float32))
+    np.testing.assert_equal(counts, np.array([51, 37]))
+    np.testing.assert_equal(unique, np.array(1 / counts, dtype=np.float32))
 
 
 @pytest.mark.skip_if_rc_var_not_set("datadir.nih_cxr14_re")
@@ -87,22 +60,17 @@ def test_get_samples_weights_multi():
     )
 
 
-def test_get_samples_weights_concat(montgomery_datadir):
-    # Temporarily modify Montgomery datadir
-    new_value = {"datadir.montgomery": str(montgomery_datadir)}
-    with rc_context(**new_value):
-        from ptbench.configs.datasets.montgomery.default import dataset
+def test_get_samples_weights_concat():
+    from ptbench.configs.datasets.montgomery.default import dataset
 
-        train_dataset = ConcatDataset(
-            (dataset["__train__"], dataset["__train__"])
-        )
+    train_dataset = ConcatDataset((dataset["__train__"], dataset["__train__"]))
 
-        train_samples_weights = get_samples_weights(train_dataset).numpy()
+    train_samples_weights = get_samples_weights(train_dataset).numpy()
 
-        unique, counts = np.unique(train_samples_weights, return_counts=True)
+    unique, counts = np.unique(train_samples_weights, return_counts=True)
 
-        np.testing.assert_equal(counts, np.array([102, 74]))
-        np.testing.assert_equal(unique, np.array(2 / counts, dtype=np.float32))
+    np.testing.assert_equal(counts, np.array([102, 74]))
+    np.testing.assert_equal(unique, np.array(2 / counts, dtype=np.float32))
 
 
 @pytest.mark.skip_if_rc_var_not_set("datadir.nih_cxr14_re")
@@ -127,19 +95,14 @@ def test_get_samples_weights_multi_concat():
     np.testing.assert_equal(train_samples_weights, ref_samples_weights)
 
 
-def test_get_positive_weights(montgomery_datadir):
-    # Temporarily modify Montgomery datadir
-    new_value = {"datadir.montgomery": str(montgomery_datadir)}
-    with rc_context(**new_value):
-        from ptbench.configs.datasets.montgomery.default import dataset
+def test_get_positive_weights():
+    from ptbench.configs.datasets.montgomery.default import dataset
 
-        train_positive_weights = get_positive_weights(
-            dataset["__train__"]
-        ).numpy()
+    train_positive_weights = get_positive_weights(dataset["__train__"]).numpy()
 
-        np.testing.assert_equal(
-            train_positive_weights, np.array([51.0 / 37.0], dtype=np.float32)
-        )
+    np.testing.assert_equal(
+        train_positive_weights, np.array([51.0 / 37.0], dtype=np.float32)
+    )
 
 
 @pytest.mark.skip_if_rc_var_not_set("datadir.nih_cxr14_re")
@@ -204,21 +167,16 @@ def test_get_positive_weights_multi():
     )
 
 
-def test_get_positive_weights_concat(montgomery_datadir):
-    # Temporarily modify Montgomery datadir
-    new_value = {"datadir.montgomery": str(montgomery_datadir)}
-    with rc_context(**new_value):
-        from ptbench.configs.datasets.montgomery.default import dataset
+def test_get_positive_weights_concat():
+    from ptbench.configs.datasets.montgomery.default import dataset
 
-        train_dataset = ConcatDataset(
-            (dataset["__train__"], dataset["__train__"])
-        )
+    train_dataset = ConcatDataset((dataset["__train__"], dataset["__train__"]))
 
-        train_positive_weights = get_positive_weights(train_dataset).numpy()
+    train_positive_weights = get_positive_weights(train_dataset).numpy()
 
-        np.testing.assert_equal(
-            train_positive_weights, np.array([51.0 / 37.0], dtype=np.float32)
-        )
+    np.testing.assert_equal(
+        train_positive_weights, np.array([51.0 / 37.0], dtype=np.float32)
+    )
 
 
 @pytest.mark.skip_if_rc_var_not_set("datadir.nih_cxr14_re")