Resolve staged renames when moving columns in UpdateSchema

[Sanjays2402]

Closes #2599

Rationale for this change

rename_column stages the new column name in self._updates, but move_first/move_before/move_after resolved the target column via self._schema.find_field(...) against the original schema. Moving a column by its new name within the same update context therefore raised ValueError: Cannot move missing column.

Reproduction (reported via AWS Glue, but reproduces fully in-memory with InMemoryCatalog — it is not Glue-specific):

with tbl.update_schema() as update:
    update.rename_column("some_column", "renamed_column")
    update.move_first("renamed_column")   # ValueError: Cannot move missing column: renamed_column

_find_for_move now checks staged renames in self._updates first (respecting _case_sensitive) before falling back to the original schema, matching the Java implementation which resolves staged renames.

Are these changes tested?

Yes. Added tests/table/test_update_schema.py covering rename-then-move by new name for move_first, move_before, and move_after using InMemoryCatalog.

• Without the fix: all 3 fail with ValueError: Cannot move missing column.
• With the fix: all 3 pass.

ruff check and ruff format --check are clean on both changed files.

Are there any user-facing changes?

No API changes. This fixes incorrect behavior: renaming a column and then moving it by its new name within the same update_schema() context now works instead of raising ValueError.