fix(plotter): raise TypeError on extra args to nodes()/edges() with DataFrame

[Lay-quq]

Summary

• Linked issue(s): Refs [FEA] move more bindings to nodes()/edges() and stronger binding error checking #398 (intentionally Refs, not Closes — only the first checklist item of [FEA] move more bindings to nodes()/edges() and stronger binding error checking #398 is addressed)
• Scope is explicit (non-goals included when relevant)

Problem

nodes() / edges() only forward extra positional/keyword arguments when the data argument is a callable. When given a DataFrame (or any other non-callable), extra arguments were silently dropped. For example, graphistry.nodes(df, 'id', point_title='some_col') succeeded but left point_title unbound.

Change

• PlotterBase.nodes() / PlotterBase.edges() with non-callable data now raise TypeError naming the unexpected positional and/or keyword arguments.
• When an unexpected keyword is a bind() binding name (e.g. point_title, edge_title), the message suggests using .bind(...) instead. Misspelled / unknown names do not get that hint.
• The callable path (argument forwarding) and bind() are unchanged. Formal parameters such as node=, source=, destination=, edge= are unaffected, as is nodes(None).
• Module-level graphistry.nodes() / graphistry.edges() delegate to PlotterBase, so they behave the same; both entry points are tested.

⚠️ Breaking / behavior change

Calls that previously silently ignored extra arguments now raise TypeError. This is recorded under ### Breaking 🔥 in the [Development] section of CHANGELOG.md.

Evidence that existing code relied on the silent behavior: graphistry/tests/compute/test_hop_column_conflicts_solution.py called .edges(edges_df, 'src', 'dst', src_renamed='id', dst_renamed='id'). edges() has no such parameters, so they were silently dropped. Unmodified, this test now fails with edges() got unexpected keyword argument(s) 'src_renamed', 'dst_renamed'. In this PR only those two arguments were removed (and an inaccurate comment on that line corrected); the test's intent is unchanged and it passes on both base and head. A repo-wide search for keyword calls to .nodes(...) / .edges(...) found no other such usages, and no other existing test changed outcome.

Non-goals

• The second checklist item of [FEA] move more bindings to nodes()/edges() and stronger binding error checking #398 is not addressed here; this PR only makes unknown arguments raise a helpful exception.

Validation

• Local lint/type/tests run for touched scope
• CI is green (not yet run at time of opening)

Local results (Python 3.12, pip install -e .[test]):

• New TestNodesEdgesExtraArgs in graphistry/tests/test_plotter.py (11 tests, covering both PlotterBase and module-level entry points): on base 0b6f88d8 8 failed / 3 passed; on this branch 11 passed.
• Full pytest, run in the style of bin/test-minimal.sh plus the GFQL tests (modules requiring optional dependencies or network skipped):
  • base: 10359 passed / 8730 skipped / 0 failed
  • head: 10370 passed / 8730 skipped / 0 failed (+11 = the new tests)
• bin/lint.sh: pass
• bin/typecheck.sh (mypy): pass

Note

Extra positional arguments are shown via repr in the error message, which can be long if a DataFrame is passed positionally. Happy to truncate or adjust the message if you prefer.

Cypher Frontend CI Evidence (when PR touches cypher frontend / IR scope)

N/A — this PR does not touch the cypher frontend / IR scope.

---

AI disclosure: This change was produced with AI assistance (implementation and tests), then independently reviewed and tested by a separate AI review step before submission.