Skip to content

A reactive notebook for Python — run reproducible experiments, query with SQL, execute as a script, deploy as an app, and version with git. Stored as pure Python. All in a modern, AI-native editor.

fix(table): preserve column order in DefaultTableManager select/drop (#9235)

Kiran Gadhaveb1acb2f75a6c
diff --git a/marimo/_plugins/ui/_impl/tables/default_table.py b/marimo/_plugins/ui/_impl/tables/default_table.py
index 76022bbbb6..70e23fae91 100644
--- a/marimo/_plugins/ui/_impl/tables/default_table.py
+++ b/marimo/_plugins/ui/_impl/tables/default_table.py
@@ -141,19 +141,16 @@ class DefaultTableManager(TableManager[JsonTableData]):
return DefaultTableManager([self.data[i] for i in indices])
def select_columns(self, columns: list[str]) -> DefaultTableManager:
- column_set = set(columns)
# Column major data
if isinstance(self.data, dict):
new_data: dict[str, Any] = {
- key: value
- for key, value in self.data.items()
- if key in column_set
+ key: self.data[key] for key in columns if key in self.data
}
return DefaultTableManager(new_data)
# Row major data
return DefaultTableManager(
[
- {key: row[key] for key in column_set}
+ {key: row[key] for key in columns}
for row in self._normalize_data(self.data)
]
)
@@ -213,9 +210,11 @@ class DefaultTableManager(TableManager[JsonTableData]):
return selected_cells
def drop_columns(self, columns: list[str]) -> DefaultTableManager:
- return self.select_columns(
- list(set(self.get_column_names()) - set(columns))
- )
+ to_drop = set(columns)
+ new_cols = [
+ col for col in self.get_column_names() if col not in to_drop
+ ]
+ return self.select_columns(new_cols)
def take(self, count: int, offset: int) -> DefaultTableManager:
if count < 0:
diff --git a/tests/_plugins/ui/_impl/tables/test_default_table.py b/tests/_plugins/ui/_impl/tables/test_default_table.py
index aef8aeebf0..14748733a8 100644
--- a/tests/_plugins/ui/_impl/tables/test_default_table.py
+++ b/tests/_plugins/ui/_impl/tables/test_default_table.py
@@ -2,12 +2,14 @@ from __future__ import annotations
import datetime
import json
+import string
import unittest
from datetime import date
from pathlib import Path
from typing import Any
import pytest
+from hypothesis import given, strategies as st
from marimo._dependencies.dependencies import DependencyManager
from marimo._output.hypertext import Html
@@ -1179,3 +1181,73 @@ def test_validate_header_tooltip_invalid() -> None:
mapping = {"does_not_exist": "oops"}
with pytest.raises(ValueError):
_validate_header_tooltip(mapping, columns)
+
+
+_column_name = st.text(
+ alphabet=string.ascii_letters + string.digits + "_",
+ min_size=1,
+ max_size=10,
+)
+_column_names = st.lists(_column_name, min_size=1, max_size=10, unique=True)
+
+
+@given(cols=_column_names, data=st.data())
+def test_drop_columns_preserves_order_row_major(
+ cols: list[str], data: st.DataObject
+) -> None:
+ to_drop = data.draw(
+ st.lists(st.sampled_from(cols), unique=True, max_size=len(cols))
+ )
+ mgr = DefaultTableManager(
+ [
+ {c: i for i, c in enumerate(cols)},
+ {c: i + 1 for i, c in enumerate(cols)},
+ ]
+ )
+ to_drop_set = set(to_drop)
+ expected = [c for c in cols if c not in to_drop_set]
+ assert mgr.drop_columns(to_drop).get_column_names() == expected
+
+
+@given(cols=_column_names, data=st.data())
+def test_drop_columns_preserves_order_column_major(
+ cols: list[str], data: st.DataObject
+) -> None:
+ to_drop = data.draw(
+ st.lists(st.sampled_from(cols), unique=True, max_size=len(cols))
+ )
+ mgr = DefaultTableManager({c: [i, i + 1] for i, c in enumerate(cols)})
+ to_drop_set = set(to_drop)
+ expected = [c for c in cols if c not in to_drop_set]
+ assert mgr.drop_columns(to_drop).get_column_names() == expected
+
+
+@given(cols=_column_names, data=st.data())
+def test_select_columns_preserves_caller_order_row_major(
+ cols: list[str], data: st.DataObject
+) -> None:
+ subset = data.draw(
+ st.lists(
+ st.sampled_from(cols), unique=True, min_size=1, max_size=len(cols)
+ )
+ )
+ mgr = DefaultTableManager(
+ [
+ {c: i for i, c in enumerate(cols)},
+ {c: i + 1 for i, c in enumerate(cols)},
+ ]
+ )
+ assert mgr.select_columns(subset).get_column_names() == subset
+
+
+@given(cols=_column_names, data=st.data())
+def test_select_columns_preserves_caller_order_column_major(
+ cols: list[str], data: st.DataObject
+) -> None:
+ subset = data.draw(
+ st.lists(
+ st.sampled_from(cols), unique=True, min_size=1, max_size=len(cols)
+ )
+ )
+ mgr = DefaultTableManager({c: [i, i + 1] for i, c in enumerate(cols)})
+ assert mgr.select_columns(subset).get_column_names() == subset