lint

Author: ninthreezy

.pre-commit-config.yaml

@@ -0,0 +1,23 @@
+# Local mirror of the CI lint gates (.github/workflows build):
+#   Black --check  ->  black hook (auto-fixes on commit instead of failing)
+#   flake8 blocking pass (E9,F63,F7,F82)  ->  flake8 hook
+#
+# One-time setup per clone:
+#   brew install pre-commit   (or: pip install pre-commit)
+#   pre-commit install
+#
+# After that every `git commit` formats staged Python with Black and runs
+# the flake8 error gate; a commit can't leave the repo in a state CI would
+# reject. `pre-commit run --all-files` checks the whole tree on demand.
+repos:
+  - repo: https://github.com/psf/black
+    rev: 26.1.0
+    hooks:
+      - id: black
+  - repo: https://github.com/pycqa/flake8
+    rev: 7.1.1
+    hooks:
+      - id: flake8
+        # CI's blocking selectors only — its second, style-wide flake8 pass
+        # runs --exit-zero (advisory), so it is not enforced here either.
+        args: ['--count', '--select=E9,F63,F7,F82', '--show-source']

datatorch/api/pull.py

@@ -231,9 +231,7 @@ def _select_datasets(project: dict, dataset_name: Optional[str]) -> list:
     if len(insensitive) == 1:
         return insensitive
     names = ", ".join(sorted(d.get("name", "?") for d in nodes)) or "(none)"
-    raise PullError(
-        f"Dataset '{dataset_name}' not found. Available datasets: {names}"
-    )
+    raise PullError(f"Dataset '{dataset_name}' not found. Available datasets: {names}")
 
 
 def _resolve_version(api: _PullSession, dataset: dict, version_name: str) -> dict:
@@ -292,9 +290,7 @@ def _list_version_files(api: _PullSession, version_id: str) -> list:
             {"id": version_id, "page": page, "perPage": PAGE_SIZE},
         )
         rows = [
-            r
-            for r in (_none_safe(data.get("datasetVersion")).get("files") or [])
-            if r
+            r for r in (_none_safe(data.get("datasetVersion")).get("files") or []) if r
         ]
         for row in rows:
             files.append(
@@ -411,9 +407,7 @@ def _download_one(
                             "Run 'datatorch login' (or set DATATORCH_API_KEY) "
                             "and retry."
                         )
-                    raise PullError(
-                        api._auth_hint(f"HTTP {r.status_code} on download")
-                    )
+                    raise PullError(api._auth_hint(f"HTTP {r.status_code} on download"))
                 if r.status_code == 416:
                     # Our .part is already >= the object; discard and re-fetch.
                     _safe_remove(part)

datatorch/cli/main/pull.py

@@ -78,7 +78,9 @@ def pull(path, dataset, version_name, out, workers, force, annotations_format, h
     def on_start(dataset_name: str, file_count: int):
         if progress["bar"] is not None:
             progress["bar"].close()
-        click.echo(f"Dataset {click.style(dataset_name, bold=True)}: {file_count} file(s)")
+        click.echo(
+            f"Dataset {click.style(dataset_name, bold=True)}: {file_count} file(s)"
+        )
         progress["bar"] = tqdm(total=file_count, unit="file") if file_count else None
 
     def on_file_done(outcome: str):