feat(frames): Polars Phase A — de-index raw/summary/journal + warn-only lint (Stage 1.12, #457)

[jepegit]

Closes #457 (Stage 1.12 / polars plan Phase A + D-preparation).

Keys live in columns, never in an index — dissolved in place while still on pandas, v1.x-safe:

• A1 raw: arbin_res stops promoting data_point to the index. The frozen v8 storage keeps its stored index (write-side promotion and the HDF5 where-clauses are storage-boundary concerns inside cellpy_file/); the read path de-indexes right after select, so in-memory raw always carries a RangeIndex + the data_point column.
• A2 summary: cycle_index_as_index flipped off in both summary builders; split/with_cycles stop re-promoting; the read path restores cycle_index as a column when loading old files (their stored index was created with drop=True).
• A3 journal: the four pages.set_index(filename) sites gain drop=False — the key now also lives in a column; removing the index itself rides utils wave 1 as planned.
• Warn-only index lint: tests/test_index_lint.py counts index idioms outside the sanctioned boundary modules (readers/cellpy_file/, parameters/legacy/) and warns — never fails; polars plan Phase D flips it to a hard ban.

Deliberate characterization updates (each documented in-test per the Stage-0 exit criteria): index-name assertions inverted, golden lookups moved from .loc[cycle] to column filters, journal missing-header count 1 → 0 (filename no longer missing from columns — an improvement), remove_outliers_from_summary_on_index made column-aware (still keyed by cycle numbers), and the pipeline_smoke goldens regenerated via dev/regenerate_goldens.py (the summary legitimately gains the cycle_index column).

Full suite: 585 passed, 0 failed. The tiered benchmark gate on this PR is the #457 acceptance check (de-indexing is where perf could silently move).

🤖 Generated with Claude Code

[cursor]

Bugbot is not enabled for your account, so this pull request was not reviewed.

Enable Bugbot in the Cursor dashboard to get automatic reviews on future PRs.

[cursor]

Did not approve: Cursor Bugbot is not enabled and did not review this PR, so there is no automated review signal. Human review is needed for the de-indexing changes across readers and utils; reviewers have been assigned.

  

_{Sent by Cursor Approval Agent: Pull Request Router and Approver}