fix: two CellService bugs — multi-hierarchy dict overwrite and missing kwargs

[kaizeenn]

Summary

Fixes two bugs in CellService.py that have been open for a while.

---

Fix #1068 — clear_with_dataframe overwrites hierarchies

Root cause: When dimension_mapping maps a dimension to a list of hierarchies, the loop at line ~776 does:

for hierarchy in hierarchies:
    mdx_selections[dimension] = MdxHierarchySet.tm1_subset_all(...)

Each iteration overwrites the same dict key, so only the last hierarchy survives and the generated MDX is wrong.

Fix: Accumulate a MdxHierarchySet.union() across all hierarchies so every one contributes to the MDX crossjoin:

combined = None
for hierarchy in hierarchies:
    hset = MdxHierarchySet.tm1_subset_all(dimension=dimension, hierarchy=hierarchy).filter_by_level(0)
    combined = hset if combined is None else combined.union(hset)
mdx_selections[dimension] = combined

Also fixed a minor error-message bug on the same line (reported in the same issue): the ValueError for a non-str mapping value referenced loop variable dimension instead of the DataFrame column variable column.

---

Fix #1113 — execute_mdx_dataframe_pivot swallows kwargs

Root cause: The method signature had no **kwargs, so calling it with e.g. cell_properties=["FormattedValue"] or use_compact_json=True raised TypeError even though extract_cellset_dataframe_pivot accepts those arguments.

Fix: Add **kwargs to the signature and forward it:

def execute_mdx_dataframe_pivot(self, mdx, dropna=False, fill_value=None, sandbox_name=None, **kwargs):
    cellset_id = self.create_cellset(mdx=mdx, sandbox_name=sandbox_name)
    return self.extract_cellset_dataframe_pivot(
        cellset_id=cellset_id, dropna=dropna, fill_value=fill_value, sandbox_name=sandbox_name, **kwargs
    )

---

Tests

Added Tests/test_cellservice_unit.py — 6 unit tests that run without a live TM1 server (mocking the REST layer and cube service):

• Single hierarchy str mapping still works
• Two hierarchies both appear in MDX with UNION
• Three hierarchies all appear
• **kwargs forwarded correctly to extract_cellset_dataframe_pivot
• Extra kwargs like cell_properties forwarded
• No kwargs still works

All 6 pass.

[Copilot]

Pull request overview

This PR fixes two long-standing CellService issues: (1) clear_with_dataframe incorrectly overwrote MDX selections when a dimension was mapped to multiple hierarchies, and (2) execute_mdx_dataframe_pivot did not accept/forward keyword arguments needed by downstream extraction helpers.

Changes:

• Update clear_with_dataframe to accumulate multiple hierarchy selections via MdxHierarchySet.union() so all hierarchies contribute to the MDX (issue #1068), and fix the related error message to reference the correct key (column).
• Update execute_mdx_dataframe_pivot to accept **kwargs and forward them to extract_cellset_dataframe_pivot (issue #1113).
• Add unit tests covering both regressions without requiring a live TM1 server.

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated 1 comment.

File	Description
TM1py/Services/CellService.py	Fixes multi-hierarchy handling in clear_with_dataframe and forwards **kwargs in execute_mdx_dataframe_pivot.
Tests/test_cellservice_unit.py	Adds regression tests for multi-hierarchy MDX generation and kwargs forwarding.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[github-actions]

Tests completed for environment: tm1-11. Check artifacts for details.

[github-actions]

Tests completed for environment: tm1-11. Check artifacts for details.

[MariusWirtz]

Thank you @kaizeenn
Code looks good and all tests succeed.

Please consider the feedback from Copilot. It identified 2 unused imports.
Please also add a regression test that has a real interaction with TM1.

[kaizeenn]

Hi @MariusWirtz, I've updated the PR! I removed the 2 unused imports from the unit tests, and I've also added two live integration regression tests in Tests/CellService_test.py (one for the execute_mdx_dataframe_pivot kwargs, and one for the clear_with_dataframe multiple-hierarchies logic) that run against a live TM1 instance. Ready for review!