fix: handle dupe column labels (#27)

.pre-commit-config.yaml

@@ -4,7 +4,7 @@ repos:
     hooks:
       - id: black
   - repo: https://github.com/charliermarsh/ruff-pre-commit
-    rev: v0.0.291
+    rev: v0.1.4
     hooks:
       - id: ruff
         args: [ --fix, --exit-non-zero-on-fix ]

CHANGELOG.md

@@ -4,6 +4,11 @@ All notable changes to this project will be documented in this file.
 
 ## [Unreleased]
 
+### Bug Fixes
+
+-   Tables with the ArrowBackend no longer display incorrect output when column labels are duplicated ([#26](https://github.com/tconbeer/textual-fastdatatable/issues/26)).
+
+
 ## [0.2.0] - 2023-11-08
 
 ### Features

src/textual_fastdatatable/backend.py

@@ -128,7 +128,21 @@ def sort(
 
 class ArrowBackend(DataTableBackend):
     def __init__(self, data: pa.Table) -> None:
-        self.data: pa.Table = data
+        # Arrow allows duplicate field names, but a table's to_pylist() and
+        # to_pydict() methods will drop duplicate-named fields!
+        field_names: list[str] = []
+        renamed = False
+        for field in data.column_names:
+            n = 0
+            while field in field_names:
+                field = f"{field}{n}"
+                renamed = True
+                n += 1
+            field_names.append(field)
+        if renamed:
+            self.data: pa.Table = data.rename_columns(field_names)
+        else:
+            self.data = data
         self._string_data: pa.Table | None = None
         self._column_content_widths: list[int] = []
 

stubs/pyarrow/__init__.pyi

@@ -2,6 +2,8 @@ from __future__ import annotations
 
 from typing import Any, Iterable, Iterator, Literal, Mapping, Sequence, Type, TypeVar
 
+import pandas as pd
+
 from .compute import CastOptions
 from .types import DataType as DataType
 from .types import string as string
@@ -71,6 +73,7 @@ class _Tabular:
     def column(self, i: int | str) -> _PandasConvertible: ...
     def equals(self: T, other: T, check_metadata: bool = False) -> bool: ...
     def itercolumns(self) -> Iterator[_PandasConvertible]: ...
+    def rename_columns(self: T, names: list[str]) -> T: ...
     def select(self: T, columns: Sequence[str | int]) -> T: ...
     def set_column(
         self: T, i: int, field_: str | Field, column: Array | ChunkedArray
@@ -107,11 +110,20 @@ def array(
     safe: bool = True,
     memory_pool: MemoryPool | None = None,
 ) -> Array | ChunkedArray: ...
+def concat_arrays(
+    arrays: Iterable[Array], memory_pool: MemoryPool | None = None
+) -> Array: ...
 def nulls(
     size: int,
     type: DataType | None = None,  # noqa: A002
     memory_pool: MemoryPool | None = None,
 ) -> Array: ...
-def concat_arrays(
-    arrays: Iterable[Array], memory_pool: MemoryPool | None = None
-) -> Array: ...
+def table(
+    data: pd.DataFrame
+    | Mapping[str, _PandasConvertible | list]
+    | list[_PandasConvertible],
+    names: list[str] | None = None,
+    schema: Schema | None = None,
+    metadata: Mapping | None = None,
+    nthreads: int | None = None,
+) -> Table: ...

tests/unit_tests/test_arrow_backend.py

@@ -163,3 +163,12 @@ def test_empty_query() -> None:
     data: dict[str, list] = {"a": []}
     backend = ArrowBackend.from_pydict(data)
     assert backend.column_content_widths == [0]
+
+
+def test_dupe_column_labels() -> None:
+    arr = pa.array([0, 1, 2, 3])
+    tab = pa.table([arr] * 3, names=["a", "a", "a"])
+    backend = ArrowBackend(data=tab)
+    assert backend.column_count == 3
+    assert backend.row_count == 4
+    assert backend.get_row_at(2) == [2, 2, 2]