fix #346 scan robust to corrput file (#426)

dascore/io/core.py

@@ -7,6 +7,7 @@
 
 import inspect
 import os.path
+import warnings
 from collections import defaultdict
 from collections.abc import Generator
 from functools import cache, cached_property, wraps
@@ -712,7 +713,7 @@ def _get_fiber_io_and_req_type(
 
     Raises
     ------
-    UnkownFileFormatError if no format is determinable from the
+    UnknownFileFormatError if no format is determinable from the
     patch_source
 
     """
@@ -837,7 +838,11 @@ def scan(
                 # contents should be returned.
                 source = fiber_io.scan(resource, timestamp=timestamp, _pre_cast=True)
             else:
-                source = fiber_io.scan(resource, _pre_cast=True)
+                try:
+                    source = fiber_io.scan(resource, _pre_cast=True)
+                except OSError:  # This happens if the file is corrupt see #346.
+                    warnings.warn(f"Failed to scan {resource}", UserWarning)
+                    continue
             for attr in source:
                 out.append(dc.PatchAttrs.from_dict(attr))
     return out

tests/test_io/test_io_core.py

@@ -343,6 +343,23 @@ def test_can_raise(self):
         with pytest.raises(NotImplementedError):
             fio.scan(bad_input)
 
+    def test_bad_checksum(self, monkeypatch, terra15_v6_path):
+        """Test for when format is identified but can't read part of file #346"""
+        # Monkey patch scan to raise OSError. This simulates observed behavior.
+        fname, ver = FiberIO.manager._get_format(path=terra15_v6_path)
+        fiber_io = FiberIO.manager.get_fiberio(format=fname, version=ver)
+
+        def raise_os_error(*args, **kwargs):
+            raise OSError("Simulated OS issue")
+
+        monkeypatch.setattr(fiber_io, "scan", raise_os_error)
+
+        # Ensure scanning doesn't raise and warns
+        msg = "Failed to scan"
+        with pytest.warns(UserWarning, match=msg):
+            scan = dc.scan(terra15_v6_path)
+        assert not len(scan)
+
 
 class TestCastType:
     """Test suite to ensure types are intelligently cast to type hints."""