From c991a22b405a255824b4cc2e10a14801dc0cda48 Mon Sep 17 00:00:00 2001
From: Josh Carmichael <carmichael@princeton.edu>
Date: Thu, 15 May 2025 15:38:20 -0400
Subject: [PATCH 01/11] proceed with download on hash mismatch, with warning.

---
 src/aspire/downloader/__init__.py     |  1 +
 src/aspire/downloader/data_fetcher.py | 17 ++++++++++++++++-
 2 files changed, 17 insertions(+), 1 deletion(-)

diff --git a/src/aspire/downloader/__init__.py b/src/aspire/downloader/__init__.py
index be0d375878..714bba0e53 100644
--- a/src/aspire/downloader/__init__.py
+++ b/src/aspire/downloader/__init__.py
@@ -4,6 +4,7 @@
 # isort: on
 from .data_fetcher import (
     available_downloads,
+    download_all,
     emdb_2484,
     emdb_2660,
     emdb_2824,
diff --git a/src/aspire/downloader/data_fetcher.py b/src/aspire/downloader/data_fetcher.py
index f655d06ab3..4b81092e6c 100644
--- a/src/aspire/downloader/data_fetcher.py
+++ b/src/aspire/downloader/data_fetcher.py
@@ -1,4 +1,5 @@
 import shutil
+import warnings
 
 import numpy as np
 import pooch
@@ -40,7 +41,21 @@ def fetch_data(dataset_name):
     :return: The absolute path (including the file name) of the file in
         local storage.
     """
-    return _data_fetcher.fetch(dataset_name)
+    try:
+        return _data_fetcher.fetch(dataset_name)
+    except ValueError as e:
+        warnings.warn(
+            f"Hash mismatch for {dataset_name}, proceeding with download. "
+            "Source file may have been updated."
+        )
+        # force download without hash check
+        url = _data_fetcher.get_url(dataset_name)
+        return pooch.retrieve(
+            url=url,
+            known_hash=None,
+            fname=dataset_name,
+            path=_data_fetcher.path,
+        )
 
 
 def download_all():

From 2cc6eb845ff4937bcb642f976d3cda63bf0d89ca Mon Sep 17 00:00:00 2001
From: Josh Carmichael <carmichael@princeton.edu>
Date: Thu, 15 May 2025 15:45:17 -0400
Subject: [PATCH 02/11] update hash in registry

---
 src/aspire/downloader/registry.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/src/aspire/downloader/registry.py b/src/aspire/downloader/registry.py
index 467ad3c772..d96536d068 100644
--- a/src/aspire/downloader/registry.py
+++ b/src/aspire/downloader/registry.py
@@ -10,7 +10,7 @@
     "emdb_5778.map": "877cbe37b86561c3dfb255aa2308fefcdd8f51f91928b17c2ef5c8dd3afaaef7",
     "emdb_6287.map": "81463aa6d024c80efcd19aa9b5ac58f3b3464af56e1ef0f104bd25071acc9204",
     "emdb_2824.map": "7682e1ef6e5bc9f2de9edcf824a03e454ef9cb1ca33bc12920633559f7f826e4",
-    "emdb_14621.map": "b45774245c2bd5e1a44e801b8fb1705a44d5850631838d060294be42e34a6900",
+    "emdb_14621.map": "98363ae950229243131025995b5ba0486857ccb1256b3df8d25c1c282155238c",
     "emdb_2484.map": "6a324e23352bea101c191d5e854026162a5a9b0b8fc73ac5a085cc22038e1999",
     "emdb_6458.map": "645208af6d36bbd3d172c549e58d387b81142fd320e064bc66105be0eae540d1",
     "simulated_channelspin.npz": "c0752674acb85417f6a77a28ac55280c1926c73fda9e25ce0a9940728b1dfcc8",

From b58cada8d46ab491fe9dac7818f19a84a24dd3eb Mon Sep 17 00:00:00 2001
From: Josh Carmichael <carmichael@princeton.edu>
Date: Fri, 16 May 2025 10:57:59 -0400
Subject: [PATCH 03/11] stack_level. unused variable

---
 src/aspire/downloader/data_fetcher.py | 9 ++++++---
 1 file changed, 6 insertions(+), 3 deletions(-)

diff --git a/src/aspire/downloader/data_fetcher.py b/src/aspire/downloader/data_fetcher.py
index 4b81092e6c..df194c9432 100644
--- a/src/aspire/downloader/data_fetcher.py
+++ b/src/aspire/downloader/data_fetcher.py
@@ -34,7 +34,7 @@ def fetch_data(dataset_name):
     file in local storage doesn’t match the one in the registry, will download a
     new copy of the file. This is considered a sign that the file was updated in
     the remote storage. If the hash of the downloaded file still doesn’t match the
-    one in the registry, will raise an exception to warn of possible file corruption.
+    one in the registry, will warn user of possible file corruption.
 
     :param dataset_name: The file name (as appears in the registry) to
         fetch from local storage.
@@ -43,11 +43,14 @@ def fetch_data(dataset_name):
     """
     try:
         return _data_fetcher.fetch(dataset_name)
-    except ValueError as e:
+    except ValueError:
         warnings.warn(
             f"Hash mismatch for {dataset_name}, proceeding with download. "
-            "Source file may have been updated."
+            "Source file may have been updated.",
+            UserWarning,
+            stacklevel=1,
         )
+
         # force download without hash check
         url = _data_fetcher.get_url(dataset_name)
         return pooch.retrieve(

From 23a8b1e7c67fdc7238b06e813164f79bd38b4846 Mon Sep 17 00:00:00 2001
From: Josh Carmichael <carmichael@princeton.edu>
Date: Fri, 16 May 2025 14:27:45 -0400
Subject: [PATCH 04/11] add scheduled_workflow and scheduled downloader test.

---
 .github/workflows/scheduled_workflow.yml | 38 ++++++++++++++++++++++++
 tests/test_downloader.py                 |  9 ++++++
 tox.ini                                  |  3 +-
 3 files changed, 49 insertions(+), 1 deletion(-)
 create mode 100644 .github/workflows/scheduled_workflow.yml
 create mode 100644 tests/test_downloader.py

diff --git a/.github/workflows/scheduled_workflow.yml b/.github/workflows/scheduled_workflow.yml
new file mode 100644
index 0000000000..ac8bac0976
--- /dev/null
+++ b/.github/workflows/scheduled_workflow.yml
@@ -0,0 +1,38 @@
+name: ASPIRE Python Scheduled Workflow
+
+on:
+  workflow_dispatch:  # Manual "Run workflow" button
+  schedule:
+    - cron: '0 0 * * 0'  # Every Sunday at 00:00 UTC
+
+
+jobs:
+  check:
+    runs-on: ubuntu-latest
+    steps:
+    - uses: actions/checkout@v4
+    - name: Set up Python
+      uses: actions/setup-python@v5
+      with:
+        python-version: '3.9'
+    - name: Install dependencies
+      run: |
+        pip install tox
+    - name: Run Tox Check
+      run: tox -e check
+
+  scheduled-tests:
+    needs: check
+    runs-on: ubuntu-latest
+
+    steps:
+    - uses: actions/checkout@v4
+    - name: Set up Python
+      uses: actions/setup-python@v5
+      with:
+        python-version: '3.10'
+    - name: Install Dependencies
+      run: |
+        pip install -e ".[dev]"
+    - name: Scheduled Tests
+      run: PYTHONWARNINGS=error python -m pytest -m scheduled
diff --git a/tests/test_downloader.py b/tests/test_downloader.py
new file mode 100644
index 0000000000..cfb32666c8
--- /dev/null
+++ b/tests/test_downloader.py
@@ -0,0 +1,9 @@
+import pytest
+
+from aspire.downloader import download_all
+
+
+@pytest.mark.scheduled
+def test_download_all():
+    """This test will throw a warning if any hashes have changed"""
+    _ = download_all()
diff --git a/tox.ini b/tox.ini
index afa6003cc5..3183b3c0ef 100644
--- a/tox.ini
+++ b/tox.ini
@@ -92,7 +92,8 @@ line_length = 88
 testpaths = tests
 markers =
     expensive: mark a test as a long running test.
-addopts = -m "not expensive"
+    scheduled: tests that should only run in the scheduled workflow
+addopts = -m "not expensive and not scheduled"
 
 [gh-actions]
 python =

From cb237b778bae6ddea7c7a54380bc6ed2d419c571 Mon Sep 17 00:00:00 2001
From: Josh Carmichael <carmichael@princeton.edu>
Date: Fri, 16 May 2025 15:21:24 -0400
Subject: [PATCH 05/11] same py version

---
 .github/workflows/scheduled_workflow.yml | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/.github/workflows/scheduled_workflow.yml b/.github/workflows/scheduled_workflow.yml
index ac8bac0976..0763244184 100644
--- a/.github/workflows/scheduled_workflow.yml
+++ b/.github/workflows/scheduled_workflow.yml
@@ -14,7 +14,7 @@ jobs:
     - name: Set up Python
       uses: actions/setup-python@v5
       with:
-        python-version: '3.9'
+        python-version: '3.10'
     - name: Install dependencies
       run: |
         pip install tox

From fa1b431e34a09a652a4daa1cda31fc036d4645a3 Mon Sep 17 00:00:00 2001
From: Josh Carmichael <carmichael@princeton.edu>
Date: Tue, 20 May 2025 10:31:04 -0400
Subject: [PATCH 06/11] Use logger warning. Update test to fail on hash
 mismatch warning.

---
 src/aspire/downloader/data_fetcher.py | 11 ++++++-----
 tests/test_downloader.py              | 11 ++++++++---
 2 files changed, 14 insertions(+), 8 deletions(-)

diff --git a/src/aspire/downloader/data_fetcher.py b/src/aspire/downloader/data_fetcher.py
index df194c9432..146eb97448 100644
--- a/src/aspire/downloader/data_fetcher.py
+++ b/src/aspire/downloader/data_fetcher.py
@@ -1,5 +1,5 @@
+import logging
 import shutil
-import warnings
 
 import numpy as np
 import pooch
@@ -11,6 +11,9 @@
 from aspire.utils import Rotation
 from aspire.volume import Volume
 
+logger = logging.getLogger(__name__)
+
+
 # Initialize pooch data fetcher instance.
 _data_fetcher = pooch.create(
     # Set the cache path defined in the config. By default, the cache
@@ -44,11 +47,9 @@ def fetch_data(dataset_name):
     try:
         return _data_fetcher.fetch(dataset_name)
     except ValueError:
-        warnings.warn(
+        logger.warning(
             f"Hash mismatch for {dataset_name}, proceeding with download. "
-            "Source file may have been updated.",
-            UserWarning,
-            stacklevel=1,
+            "Source file may have been updated."
         )
 
         # force download without hash check
diff --git a/tests/test_downloader.py b/tests/test_downloader.py
index cfb32666c8..68e20b51fa 100644
--- a/tests/test_downloader.py
+++ b/tests/test_downloader.py
@@ -4,6 +4,11 @@
 
 
 @pytest.mark.scheduled
-def test_download_all():
-    """This test will throw a warning if any hashes have changed"""
-    _ = download_all()
+def test_download_all(caplog):
+    """Fail if a hash mismatch warning is logged during download."""
+    caplog.clear()
+    with caplog.at_level("WARNING"):
+        _ = download_all()
+
+    if "Hash mismatch" in caplog.text:
+        pytest.fail(f"Hash mismatch warning was logged.\nCaptured logs:\n{caplog.text}")

From 6cbbf33000081b0adac8b011d13dc2b769394d96 Mon Sep 17 00:00:00 2001
From: Josh Carmichael <carmichael@princeton.edu>
Date: Tue, 20 May 2025 10:39:11 -0400
Subject: [PATCH 07/11] Workflow updates: Run on develop, remove check, remove
 fail on warnings.

---
 .github/workflows/scheduled_workflow.yml | 20 +++++---------------
 1 file changed, 5 insertions(+), 15 deletions(-)

diff --git a/.github/workflows/scheduled_workflow.yml b/.github/workflows/scheduled_workflow.yml
index 0763244184..e4a1507a9e 100644
--- a/.github/workflows/scheduled_workflow.yml
+++ b/.github/workflows/scheduled_workflow.yml
@@ -2,25 +2,15 @@ name: ASPIRE Python Scheduled Workflow
 
 on:
   workflow_dispatch:  # Manual "Run workflow" button
+    branches:
+      - develop
   schedule:
     - cron: '0 0 * * 0'  # Every Sunday at 00:00 UTC
+      branches:
+        - develop
 
 
 jobs:
-  check:
-    runs-on: ubuntu-latest
-    steps:
-    - uses: actions/checkout@v4
-    - name: Set up Python
-      uses: actions/setup-python@v5
-      with:
-        python-version: '3.10'
-    - name: Install dependencies
-      run: |
-        pip install tox
-    - name: Run Tox Check
-      run: tox -e check
-
   scheduled-tests:
     needs: check
     runs-on: ubuntu-latest
@@ -35,4 +25,4 @@ jobs:
       run: |
         pip install -e ".[dev]"
     - name: Scheduled Tests
-      run: PYTHONWARNINGS=error python -m pytest -m scheduled
+      run: pytest -m scheduled

From c48ad9fc2a0b67a3b155823eab36add9f0ff5e72 Mon Sep 17 00:00:00 2001
From: Josh Carmichael <carmichael@princeton.edu>
Date: Tue, 20 May 2025 10:59:29 -0400
Subject: [PATCH 08/11] Workflow updates: Checkout develop, set cron off-hour,
 remove needs field.

---
 .github/workflows/scheduled_workflow.yml | 6 ++----
 1 file changed, 2 insertions(+), 4 deletions(-)

diff --git a/.github/workflows/scheduled_workflow.yml b/.github/workflows/scheduled_workflow.yml
index e4a1507a9e..d4d82f80df 100644
--- a/.github/workflows/scheduled_workflow.yml
+++ b/.github/workflows/scheduled_workflow.yml
@@ -5,14 +5,12 @@ on:
     branches:
       - develop
   schedule:
-    - cron: '0 0 * * 0'  # Every Sunday at 00:00 UTC
-      branches:
-        - develop
+    - cron: '15 0 * * 0'  # Every Sunday at 00:15 UTC
 
 
 jobs:
   scheduled-tests:
-    needs: check
+    if: github.ref == 'refs/heads/develop'  # Ensure only runs on develop
     runs-on: ubuntu-latest
 
     steps:

From 679b2bf50e83df305aa56a87fcf2e7442abf9ab5 Mon Sep 17 00:00:00 2001
From: Josh Carmichael <carmichael@princeton.edu>
Date: Tue, 20 May 2025 14:38:49 -0400
Subject: [PATCH 09/11] test hash mismatch warning works

---
 tests/test_downloader.py | 31 +++++++++++++++++++++++++++++++
 1 file changed, 31 insertions(+)

diff --git a/tests/test_downloader.py b/tests/test_downloader.py
index 68e20b51fa..dfbc135c61 100644
--- a/tests/test_downloader.py
+++ b/tests/test_downloader.py
@@ -1,6 +1,9 @@
+from pathlib import Path
+
 import pytest
 
 from aspire.downloader import download_all
+from aspire.downloader.data_fetcher import _data_fetcher, fetch_data
 
 
 @pytest.mark.scheduled
@@ -12,3 +15,31 @@ def test_download_all(caplog):
 
     if "Hash mismatch" in caplog.text:
         pytest.fail(f"Hash mismatch warning was logged.\nCaptured logs:\n{caplog.text}")
+
+
+def test_fetch_data_warning(caplog):
+    """Test that we get expected warning on hash mismatch."""
+    # Use the smallest dataset in the registry
+    dataset_name = "emdb_3645.map"
+
+    # Remove file from cache if it exists
+    cached_path = Path(_data_fetcher.path) / dataset_name
+    if cached_path.exists():
+        cached_path.unlink()
+
+    # Save original hash from the registry
+    original_hash = _data_fetcher.registry.get(dataset_name)
+    assert original_hash is not None
+
+    # Temporarily override the hash to simulate a mismatch
+    _data_fetcher.registry[dataset_name] = "md5:invalidhash123"
+
+    try:
+        caplog.clear()
+        with caplog.at_level("WARNING"):
+            path = fetch_data(dataset_name)
+        assert path  # Should return the path to the downloaded file
+        assert f"Hash mismatch for {dataset_name}" in caplog.text
+    finally:
+        # Restore original hash
+        _data_fetcher.registry[dataset_name] = original_hash

From 77c9ad7d1ffb72998e7c3c775253241be26ad6cd Mon Sep 17 00:00:00 2001
From: Josh Carmichael <carmichael@princeton.edu>
Date: Wed, 21 May 2025 08:48:44 -0400
Subject: [PATCH 10/11] checkout on develop. mark test as scheduled.

---
 .github/workflows/scheduled_workflow.yml | 3 ++-
 tests/test_downloader.py                 | 1 +
 2 files changed, 3 insertions(+), 1 deletion(-)

diff --git a/.github/workflows/scheduled_workflow.yml b/.github/workflows/scheduled_workflow.yml
index d4d82f80df..8f96eaa30e 100644
--- a/.github/workflows/scheduled_workflow.yml
+++ b/.github/workflows/scheduled_workflow.yml
@@ -10,11 +10,12 @@ on:
 
 jobs:
   scheduled-tests:
-    if: github.ref == 'refs/heads/develop'  # Ensure only runs on develop
     runs-on: ubuntu-latest
 
     steps:
     - uses: actions/checkout@v4
+      with:
+        ref: develop
     - name: Set up Python
       uses: actions/setup-python@v5
       with:
diff --git a/tests/test_downloader.py b/tests/test_downloader.py
index dfbc135c61..cddc1cf177 100644
--- a/tests/test_downloader.py
+++ b/tests/test_downloader.py
@@ -17,6 +17,7 @@ def test_download_all(caplog):
         pytest.fail(f"Hash mismatch warning was logged.\nCaptured logs:\n{caplog.text}")
 
 
+@pytest.mark.scheduled
 def test_fetch_data_warning(caplog):
     """Test that we get expected warning on hash mismatch."""
     # Use the smallest dataset in the registry

From 363f02ea11a97a211dae0098ed255ac0c0d707ef Mon Sep 17 00:00:00 2001
From: Josh Carmichael <carmichael@princeton.edu>
Date: Wed, 21 May 2025 11:17:57 -0400
Subject: [PATCH 11/11] remove manual run-workflow button

---
 .github/workflows/scheduled_workflow.yml | 3 ---
 1 file changed, 3 deletions(-)

diff --git a/.github/workflows/scheduled_workflow.yml b/.github/workflows/scheduled_workflow.yml
index 8f96eaa30e..77a789597e 100644
--- a/.github/workflows/scheduled_workflow.yml
+++ b/.github/workflows/scheduled_workflow.yml
@@ -1,9 +1,6 @@
 name: ASPIRE Python Scheduled Workflow
 
 on:
-  workflow_dispatch:  # Manual "Run workflow" button
-    branches:
-      - develop
   schedule:
     - cron: '15 0 * * 0'  # Every Sunday at 00:15 UTC