feat: Support a wider range of iterables in SchemaBase.to_dict (#3501)

* feat: Support a wider range of iterables in `SchemaBase.to_dict`

* test: Update validation errors to use non-iterable element type

Previously `list[set[str]]`

* fix: Prevent `to_dict` method being called on `pd.Series`

AFAIK, this is the intended case for `Parameter | Expression` not for converting arbitrary objects

* test: Add tests for iterables and ranges

* fix(typing): Ignore type errors for tests

* refactor: Use `narwhals.stable.v1`

#3501 (comment)

* test: Increase coverage in `test_to_dict_iterables`

The original test obscured the fact that this change applies anywhere a `Sequence` is annotated.

* docs: Add a doc for `test_to_dict_iterables`

* revert: Change `test_chart_validation_errors` back to demonstrate failures to @joelostblom

Reverting 1f3dfd3

#3501 (comment)

* docs: Update User Guide to use `Sequence`

#3501 (comment)

* test: Fix `test_chart_validation_errors` failure verbosity

#3501 (comment)

* refactor: Move `inspect.cleandoc` inside of `test_chart_validation_errors`

All 18 cases use this, saves 30 lines

* revert: Restore original fix to `test_chart_validation_errors`

#3501 (comment)
#3501 (comment)
#3501 (comment)

* docs: Remove "ordered" descriptor from `Sequence`

Co-authored-by: Joel Ostblom <[email protected]>

* test: Remove missed `inspect.cleandoc`

#3501 (comment)

* test: Only modify message, not input to `test_chart_validation_errors`

#3501 (comment)

* style: fix oddly formatted `test_multiple_field_strings_in_condition`

---------

Co-authored-by: Joel Ostblom <[email protected]>

altair/utils/schemapi.py

@@ -32,6 +32,7 @@
 import jsonschema
 import jsonschema.exceptions
 import jsonschema.validators
+import narwhals.stable.v1 as nw
 from packaging.version import Version
 
 # This leads to circular imports with the vegalite module. Currently, this works
@@ -488,6 +489,14 @@ def _subclasses(cls: type[Any]) -> Iterator[type[Any]]:
             yield cls
 
 
+def _from_array_like(obj: Iterable[Any], /) -> list[Any]:
+    try:
+        ser = nw.from_native(obj, strict=True, series_only=True)
+        return ser.to_list()
+    except TypeError:
+        return list(obj)
+
+
 def _todict(obj: Any, context: dict[str, Any] | None, np_opt: Any, pd_opt: Any) -> Any:
     """Convert an object to a dict representation."""
     if np_opt is not None:
@@ -512,10 +521,16 @@ def _todict(obj: Any, context: dict[str, Any] | None, np_opt: Any, pd_opt: Any)
             for k, v in obj.items()
             if v is not Undefined
         }
-    elif hasattr(obj, "to_dict"):
+    elif (
+        hasattr(obj, "to_dict")
+        and (module_name := obj.__module__)
+        and module_name.startswith("altair")
+    ):
         return obj.to_dict()
     elif pd_opt is not None and isinstance(obj, pd_opt.Timestamp):
         return pd_opt.Timestamp(obj).isoformat()
+    elif _is_iterable(obj, exclude=(str, bytes)):
+        return _todict(_from_array_like(obj), context, np_opt, pd_opt)
     else:
         return obj
 
@@ -1232,6 +1247,12 @@ def _is_list(obj: Any | list[Any]) -> TypeIs[list[Any]]:
     return isinstance(obj, list)
 
 
+def _is_iterable(
+    obj: Any, *, exclude: type | tuple[type, ...] = (str, bytes)
+) -> TypeIs[Iterable[Any]]:
+    return not isinstance(obj, exclude) and isinstance(obj, Iterable)
+
+
 def _passthrough(*args: Any, **kwds: Any) -> Any | dict[str, Any]:
     return args[0] if args else kwds
 

doc/user_guide/encodings/index.rst

@@ -420,7 +420,7 @@ options available to change the sort order:
 - Passing the name of an encoding channel to ``sort``, such as ``"x"`` or ``"y"``, allows for
   sorting by that channel. An optional minus prefix can be used for a descending
   sort. For example ``sort='-x'`` would sort by the x channel in descending order.
-- Passing a list to ``sort`` allows you to explicitly set the order in which
+- Passing a `Sequence <https://docs.python.org/3/library/stdtypes.html#sequence-types-list-tuple-range>`_ to ``sort`` allows you to explicitly set the order in which
   you would like the encoding to appear
 - Using the ``field`` and ``op`` parameters to specify a field and aggregation operation to sort by.