cellpycore 0.2.1__tar.gz → 0.2.2__tar.gz

cellpycore-0.2.2/.gitattributes

@@ -0,0 +1,14 @@
+# GitHub Linguist: keep language stats focused on library source.
+# Without this, graphify-out/graph.html (~3.8MB) dominates as HTML.
+
+graphify-out/** linguist-generated
+docs/** linguist-documentation
+tests/** linguist-documentation
+.issueflows/** linguist-documentation
+dev/** linguist-documentation
+scripts/** linguist-documentation
+
+# Cross-platform line endings for shell helpers
+.aliases     text eol=lf
+*.sh         text eol=lf
+*.lock       text eol=lf

cellpycore-0.2.2/.issueflows/03-solved-issues/issue136_original.md

@@ -0,0 +1,60 @@
+# Issue #136: Legacy bridge: preserve test_id through steps/summary output; legacy-schema support for merge_data
+
+Source: https://github.com/cellpy/cellpy-core/issues/136
+
+## Original issue text
+
+## Summary
+
+The `OldCellpyCellCore` legacy bridge strips `test_id` from the outbound
+step/summary frames, which degrades multi-test (campaign-merged) objects to
+cycle-only summary grouping. cellpy #507 works around it; these requests make
+the workaround unnecessary and unblock campaign merges that keep original
+per-test cycle numbers.
+
+## Details
+
+1. **Outbound `test_id` stripping.** `_native_steps_to_legacy` re-projects to a
+   fixed legacy column order (cell_core.py ~781-815) and the summary path
+   filters to `_legacy_summary_column_order` (~928-1065); `test_id` is in
+   `NATIVE_ONLY_STEP` / `NATIVE_ONLY_CYCLE` (mapping.py ~191-196, 227-231). The
+   inbound side is fine: a raw `test_id` column passes through (exception
+   lists, mapping.py ~129-144) and `make_step_table` groups on
+   `(test_id, cycle_num, step_num)` (proven by the cycler_small golden).
+2. **Bridge summary rebuilds native steps from the stripped legacy steps**
+   (`make_core_summary`, cell_core.py ~1010-1012), so
+   `use_tid` (summarizers.py ~792) is always False through the bridge —
+   cycle-only grouping, no per-test windowing, and cross-test collision when
+   cycle numbers overlap. cellpy #507's workaround: re-stamp `test_id` onto the
+   legacy steps frame post-bridge (depends on the inbound step rename staying
+   rename-only with no column filtering — please treat that as a contract or
+   provide the fix below).
+3. **`_add_legacy_summary_cruft`** (cell_core.py ~960-978) does global pandas
+   cumsums (cumulated CE, shifted_*, cumulated_ric*) even when the engine
+   windowed per test — these columns are wrong for multi-test objects.
+4. **`merge_data(schema=<legacy Schema>)` raises `AttributeError`** —
+   `HeadersNormal` lacks the native attribute names (`datapoint_num`,
+   `cycle_num`, `test_id`), and passing the native schema instead silently
+   skips offsets on legacy-named frames (`_max_column` returns 0 for missing
+   columns). Either document legacy schemas as unsupported for merge/update, or
+   add attribute aliases so bridge consumers can call
+   `merge_data`/`update_data` directly (cellpy #507 mirrors the semantics in
+   pandas instead).
+
+## Requests
+
+- [ ] Carry `test_id` through the legacy step output (or have
+      `make_core_summary` reuse the native steps it already computed instead of
+      rebuilding from the stripped legacy frame).
+- [ ] Optionally expose `test_id` in the legacy summary output behind a flag.
+- [ ] Window `_add_legacy_summary_cruft` per `test_id` when present.
+- [ ] Resolve the `merge_data` legacy-schema story (docs or aliases).
+- [ ] Together these unblock `renumber_cycles=False` (original per-test cycle
+      numbers) for cellpy's campaign merge ahead of the native-schema path
+      (cellpy #511).
+
+## Refs
+
+cellpy jepegit/cellpy#507 (campaign merge + workaround), jepegit/cellpy#511
+(native schema opt-in), core tests test_golden.py (cycler_small composite-key
+golden), test_merge.py / test_e2e.py (native multi-test invariants).

cellpycore-0.2.2/.issueflows/03-solved-issues/issue136_plan.md

@@ -0,0 +1,87 @@
+# Issue #136 — plan
+
+Source: https://github.com/cellpy/cellpy-core/issues/136
+
+## Goal
+
+Legacy bridge keeps `test_id` on outbound steps (and summary when present) so
+campaign-merged multi-test objects get per-test summary windowing through the
+bridge — no cellpy #507 re-stamp workaround. Also fix legacy-cruft cumsums and
+clarify `merge_data` + legacy schema.
+
+## Constraints
+
+- Back-compat: single-test objects already have `test_id=0` on raw; emitting
+  `test_id` on steps/summary is **additive** (extra column). Goldens that pin
+  exact column sets may need updates.
+- Mapping totality (`NATIVE_ONLY_*` / attr maps) must stay consistent.
+- Do not change native engine behaviour; bridge-only (+ headers/mapping).
+- Unblocks cellpy #510 V2-15 pin after a **new** cellpycore PyPI release.
+
+### Prior art
+
+| Hit | Role | Plan |
+|-----|------|------|
+| [`OldCellpyCellCore._native_steps_to_legacy`](../../src/cellpycore/cell_core.py) / `_legacy_step_column_order` | Drops unknown cols via fixed order | **Extend** order with `test_id` when present |
+| [`make_core_summary`](../../src/cellpycore/cell_core.py) rebuilds native steps from legacy `data.steps` | `use_tid` false after strip | Fixed once steps carry `test_id` through rename |
+| [`_add_legacy_summary_cruft`](../../src/cellpycore/cell_core.py) | Global pandas cumsums | **Window** by `test_id` when column present |
+| [`NATIVE_ONLY_STEP` / `NATIVE_ONLY_CYCLE`](../../src/cellpycore/legacy/mapping.py) | Declares `test_id` intentionally unbridged | **Remove** `test_id`; add identity scalar/cycle pairs |
+| [`HeadersStepTable` / `HeadersSummary`](../../src/cellpycore/legacy/headers.py) | No `test_id` field today | **Add** `test_id: str = "test_id"` |
+| cellpy [`cellreader.make_step_table` re-stamp](../../../cellpy/cellpy/readers/cellreader.py) (~2293) | Workaround | Becomes redundant after this; leave for a cellpy follow-up |
+| [`merge_data`](../../src/cellpycore/merge.py) | Expects native `Schema` attrs | **Document + hard error** on legacy header schema (cellpy already uses pandas merge) |
+| Toolbox `00-tools/` | Empty | None |
+| Graph | Bridge / mapping communities | Touch `cell_core` + `legacy/mapping` + `legacy/headers` |
+
+## Approach
+
+1. **Steps outbound:** Add `HeadersStepTable.test_id`. Put `test_id` early in
+   `_legacy_step_column_order` (before cycle/step). Add identity
+   `STEP_SCALAR_PAIRS` entry `("test_id", "test_id")`; drop `test_id` from
+   `NATIVE_ONLY_STEP`. Update attr maps (`LEGACY_ATTR_TO_SCHEMA` /
+   `LEGACY_ATTR_UNMAPPED`) for the new header field.
+2. **Summary outbound:** Add `HeadersSummary.test_id`. Include in
+   `_legacy_summary_column_order` (near `cycle_index`) when present after rename.
+   Add identity `CYCLE_PAIRS` entry; drop `test_id` from `NATIVE_ONLY_CYCLE`.
+   No flag — always pass through when the native summary has it (single-test →
+   all `0`).
+3. **Cruft:** In `_add_legacy_summary_cruft`, if `test_id` in summary columns and
+   `nunique > 1` (or always when column present): compute cumsums / shifted /
+   ric **per group**; else keep today's global behaviour.
+4. **`merge_data` + legacy schema:** At entry, if `schema.raw` is
+   `HeadersNormal` (or lacks `datapoint_num` / `cycle_num` / `test_id`), raise
+   `TypeError` with a clear message: use `config.default_schema()` / native
+   `Schema`. Document in `merge_data` docstring. No full legacy-schema merge
+   implementation in this issue (cellpy does not call it).
+5. **Tests:** Bridge multi-test fixture (or reuse schema/e2e multi-test raw) →
+   `make_core_step_table` / `make_core_summary` → assert `test_id` on steps and
+   summary `{0,1}`; assert summary cumulatives restart per test. Mapping
+   totality tests update with the exception-set moves. One test that
+   `merge_data(..., schema=legacy.schema)` raises the new TypeError.
+
+## Files to touch
+
+| Path | Change |
+|------|--------|
+| `src/cellpycore/legacy/headers.py` | `test_id` on step + summary headers |
+| `src/cellpycore/legacy/mapping.py` | pairs / `NATIVE_ONLY_*` / attr maps + comments |
+| `src/cellpycore/cell_core.py` | column order, cruft windowing |
+| `tests/test_*.py` (bridge / mapping / merge) | new + update totality |
+| `CHANGELOG` / HISTORY if present | note for 0.2.2 release |
+
+## Test strategy
+
+```bash
+uv run pytest
+uv run ruff check && uv run ruff format --check
+```
+
+Focus: new bridge multi-test test + `tests/test_header_mapping.py` totality +
+`tests/test_merge.py` legacy-schema error.
+
+## Decisions (confirmed 2026-07-17)
+
+1. **Summary `test_id`:** always when present (no flag).
+2. **`merge_data` legacy schema:** hard `TypeError` + docstring (also
+   `update_data`).
+3. **Cruft windowing:** when `test_id` column exists (groupby).
+4. **Goldens:** regenerate stage-B step/summary snapshots (additive `test_id`).

cellpycore-0.2.2/.issueflows/03-solved-issues/issue136_status.md

@@ -0,0 +1,18 @@
+# Issue #136 — status
+
+- [x] Done
+
+## What's done
+
+- Plan confirmed (2026-07-17): carry `test_id` on steps+summary; window cruft;
+  hard-error legacy schema on merge/update; regenerate goldens.
+- Implemented on `136-legacy-bridge-preserve-test-id`.
+- Version bumped `0.2.1` → `0.2.2` (static `pyproject.toml`); HISTORY promoted.
+- `uv run pytest` 261 passed; ruff clean.
+- Close: commit + PR; PyPI release via `release` from `main` after merge
+  (see `release-procedure.md`).
+
+## Remaining work
+
+- None for the code issue. Post-merge: `iflow cleanup` + cut GitHub/PyPI
+  release for `0.2.2`, then re-pin cellpy.

cellpycore-0.2.2/.issueflows/04-designs-and-guides/legacy-bridge-test-id.md

@@ -0,0 +1,15 @@
+# Legacy bridge carries `test_id` (#136)
+
+**Context.** `OldCellpyCellCore` used to drop `test_id` from outbound steps /
+summary via fixed column orders + `NATIVE_ONLY_*` exceptions. That forced cellpy
+#507 to re-stamp steps so campaign summaries could window per test.
+
+**Decision.** Bridge `test_id` as an identity column on steps and summary when
+present. Window `_add_legacy_summary_cruft` by `test_id`. `merge_data` /
+`update_data` require native `config.Schema` (hard `TypeError` on `Headers*`).
+
+**Alternatives.** Opt-in flag for summary `test_id` — rejected (additive is
+enough). Full legacy-schema merge aliases — rejected (step/cycle attrs incomplete;
+cellpy uses pandas merge).
+
+**Refs.** cellpy #507 / #510 / #511; release after this → re-pin cellpy.

{cellpycore-0.2.1 → cellpycore-0.2.2}/.issueflows/04-designs-and-guides/this-project.md

@@ -79,3 +79,13 @@ uv run pre-commit install
   the public API.
 - Per-step stat column names (`<signal>_<stat>`) are a fixed engine contract,
   not schema-injected (issue #70).
+
+## Release & version bump
+
+Static version in ``pyproject.toml`` (``project.version``). Bump with::
+
+    uv version --bump patch|minor|major
+
+Cut a GitHub release from ``main`` after merge (``release`` alias / see
+``release-procedure.md``) so ``release.yml`` publishes ``cellpycore`` to PyPI.
+

{cellpycore-0.2.1 → cellpycore-0.2.2}/HISTORY.md

@@ -4,6 +4,15 @@ All notable changes to this project will be documented in this file.
 
 ## [Unreleased]
 
+## [0.2.2] - 2026-07-17
+
+- Legacy bridge carries ``test_id`` through step and summary frames (identity
+  mapping) so campaign-merged multi-test objects keep per-test summary
+  windowing without a consumer re-stamp. ``_add_legacy_summary_cruft`` cumsums
+  are windowed per ``test_id`` when present. ``merge_data`` / ``update_data``
+  raise ``TypeError`` on legacy ``Headers*`` schemas (require native
+  ``config.Schema``). Golden step/summary snapshots regenerated. (#136)
+
 ## [0.2.1] - 2026-07-15
 
 - Fix `OldCellpyCellCore.make_core_summary` to pass `test_mode` into

{cellpycore-0.2.1 → cellpycore-0.2.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: cellpycore
-Version: 0.2.1
+Version: 0.2.2
 Summary: Core engine for cellpy: fast, thread-safe step and cycle summarization of battery-cycling raw data
 License-Expression: MIT
 License-File: LICENSE

{cellpycore-0.2.1 → cellpycore-0.2.2}/graphify-out/.graphify_labels.json

@@ -297,15 +297,18 @@
   "295": "Community 295",
   "296": "Community 296",
   "297": "Community 297",
+  "298": "Community 298",
   "299": "Community 299",
   "300": "Community 300",
   "301": "Community 301",
   "302": "Community 302",
+  "303": "Community 303",
   "304": "Community 304",
   "305": "Community 305",
   "306": "Community 306",
   "307": "Community 307",
   "308": "Community 308",
+  "309": "Community 309",
   "310": "Community 310",
   "311": "Community 311",
   "312": "Community 312",
@@ -313,13 +316,16 @@
   "314": "Community 314",
   "315": "Community 315",
   "316": "Community 316",
+  "317": "Community 317",
   "318": "Community 318",
   "319": "Community 319",
   "320": "Community 320",
+  "321": "Community 321",
   "322": "Community 322",
   "323": "Community 323",
   "324": "Community 324",
   "325": "Community 325",
+  "326": "Community 326",
   "327": "Community 327",
   "328": "Community 328",
   "329": "Community 329",
@@ -337,5 +343,29 @@
   "341": "Community 341",
   "342": "Community 342",
   "343": "Community 343",
-  "344": "Community 344"
+  "344": "Community 344",
+  "345": "Community 345",
+  "346": "Community 346",
+  "347": "Community 347",
+  "348": "Community 348",
+  "349": "Community 349",
+  "350": "Community 350",
+  "351": "Community 351",
+  "352": "Community 352",
+  "353": "Community 353",
+  "354": "Community 354",
+  "355": "Community 355",
+  "356": "Community 356",
+  "357": "Community 357",
+  "358": "Community 358",
+  "359": "Community 359",
+  "360": "Community 360",
+  "361": "Community 361",
+  "362": "Community 362",
+  "363": "Community 363",
+  "364": "Community 364",
+  "365": "Community 365",
+  "366": "Community 366",
+  "367": "Community 367",
+  "368": "Community 368"
 }