tests/test_io/test_io_core.py (1)

14-17: Minor: Redundant imports (dascore and dc)

Both dascore and dascore as dc are imported; only one is necessary. You can use dc for both attribute patching and functional calls.

Apply this diff for brevity:

-import rich.progress as prog
-
-import dascore
-import dascore as dc
+import rich.progress as prog
+import dascore as dc

And adjust the monkeypatch line below:

-        monkeypatch.setattr(dascore, "_debug", False)
+        monkeypatch.setattr(dc, "_debug", False)

tests/test_io/test_sintela_binary/test_sintela_binary.py (1)

18-35: Solid negative test; consider asserting on message and adding a positive control

This proves mis-sized files raise InvalidFiberFileError. Two optional enhancements:

• Assert the error message contains a hint (e.g., “not divisible by packet size”) to lock in UX.
• Add a positive control test using the unmodified .raw file to assert scan succeeds, guarding against false positives in fixture setup.

Example tweak inside the test:

-        with pytest.raises(InvalidFiberFileError):
+        with pytest.raises(InvalidFiberFileError, match="divisible|packet size"):
             fiber_io.scan(extra_bytes_file)

dascore/utils/progress.py (2)

57-61: Docstring nit: clarify that a Progress instance or subclass is accepted

Minor clarity: explicitly say “instance or subclass” to reflect the behavior after the above change.

Suggested wording:

• “…can also accept a rich.progress.Progress instance or subclass.”

Also, “emmit” -> “emit” below (typo).

To update the typo outside the changed hunk, adjust this line:

# The minimum length to emit a progress bar.

---

36-37: Optional: consider constructing columns via compat to keep all rich touchpoints centralized

You import Progress from compat but still construct columns via rich.progress as prog. This is fine, but if you’re aiming to isolate ‘rich’ behind compat, consider mirroring columns or providing a small factory there. Not required for this PR.

dascore/io/core.py (8)

23-23: Decouple runtime from rich by avoiding a hard Progress import

With from future import annotations, you can reference Progress as a string literal and avoid importing it at runtime. This keeps rich fully optional and prevents accidental ImportError during import if rich isn’t installed. Also aligns with type-only usage here.

Apply:

-from dascore.compat import Progress
+from typing import TYPE_CHECKING
+
+if TYPE_CHECKING:
+    from dascore.compat import Progress

And update the annotation to a forward ref:

-def scan(..., progress: PROGRESS_LEVELS | Progress = "standard", ...) -> list[dc.PatchAttrs]:
+def scan(..., progress: PROGRESS_LEVELS | "Progress" = "standard", ...) -> list[dc.PatchAttrs]:

Also applies to: 790-790

---

363-365: Catching Exception in _get_format risks masking bugs; add debug logging

Swallowing every exception improves resilience to third-party FiberIOs, but completely silences programmer errors. Suggest logging at debug level to aid diagnosis without spamming users.

Minimal change:

-                except Exception:
-                    continue
+                except Exception as ex:
+                    # Keep robust, but surface details in debug logs.
+                    import logging as _logging
+                    _logging.getLogger(__name__).debug(
+                        "get_format suppressed for %s: %r",
+                        getattr(fiber_io, "name", "?"),
+                        ex,
+                    )
+                    continue

---

811-813: Docstring typo: “acceted” → “accepts”; clarify phrasing

Small doc polish for correctness.

-        "basic" is best for low latency scenarios. Can also acceted a subclass
-        of rich.progress.Progress.
+        "basic" is best for low-latency scenarios. Can also accept a subclass
+        of rich.progress.Progress.

---

882-887: Include exception detail in warning for failed scans

Currently the warning hides the cause, making triage harder. Appending the exception message keeps logs actionable without being noisy.

-                    except (OSError, InvalidFiberFileError, ValueError, TypeError):
-                        warnings.warn(f"Failed to scan {resource}", UserWarning)
+                    except (OSError, InvalidFiberFileError, ValueError, TypeError) as ex:
+                        warnings.warn(f"Failed to scan {resource}: {ex}", UserWarning)
                         continue

---

888-890: Deriving package name from ex.msg is brittle

ImportError/MissingOptionalDependencyError usually carries .name; .msg parsing can misidentify the package. Prefer name, then fall back to message parsing.

-                    except MissingOptionalDependencyError as ex:
-                        missing_optional_deps[ex.msg.split(" ")[0]] += 1
+                    except MissingOptionalDependencyError as ex:
+                        pkg = getattr(ex, "name", None)
+                        if not pkg:
+                            # Fallback: first token from message or str(ex)
+                            txt = getattr(ex, "msg", "") or str(ex)
+                            pkg = (txt.split()[0] if txt else "unknown")
+                        missing_optional_deps[pkg] += 1
                         continue

---

891-892: Avoid re-parsing PatchAttrs objects

If source already yields PatchAttrs, re-creating via from_dict is unnecessary.

-                for attr in source:
-                    out.append(dc.PatchAttrs.from_dict(attr))
+                for attr in source:
+                    out.append(attr if isinstance(attr, dc.PatchAttrs)
+                               else dc.PatchAttrs.from_dict(attr))

---

893-895: Comment typo

Minor typo and casing.

-    # Ensure ctl + c exists scan.
+    # Ensure Ctrl+C exits scan.

---

654-662: Align scan_to_df progress type with scan’s new annotation

scan_to_df passes progress through to scan but its type hint still excludes a Progress instance. Align for consistency.

-def scan_to_df(..., progress: PROGRESS_LEVELS = "standard", ...) -> pd.DataFrame:
+def scan_to_df(..., progress: PROGRESS_LEVELS | "Progress" = "standard", ...) -> pd.DataFrame:

Configuration used: CodeRabbit UI

Review profile: CHILL

Plan: Pro

💡 Knowledge Base configuration:

• MCP integration is disabled by default for public repositories
• Jira integration is disabled by default for public repositories
• Linear integration is disabled by default for public repositories

You can enable these sources in your CodeRabbit configuration.