fix: preserve column orientation for vertical ranges

[hhimanshu]

Summary

Elementwise operations over a vertical range (e.g. =A1:A3*2 where A1:A3 is a 1-column, 3-row range) now spill down (an Nx1 column) instead of sideways (a 1xN row), matching Google Sheets.

resolve_range now materializes a single-column, multi-row range as a nested Nx1 array (one extra Array level per cell) instead of a flat row. Making that shape change safe required updating every function that flattens a Value::Array argument by hand to fully recurse into nested arrays, not just unwrap one level:

• AVERAGE, MAX, MIN, and the SUMIF/criterion helper (flatten_to_vec) — the 4 functions caught by pinned conformance fixtures.
• LARGE, SMALL, MODE, MODE.MULT, the PERCENTILE/PERCENTILE.INC/PERCENTILE.EXC/QUARTILE/QUARTILE.INC/QUARTILE.EXC/PERCENTRANK/PERCENTRANK.INC/PERCENTRANK.EXC family (they share one helper module), RANK, RANK.AVG, RANK.EQ, GEOMEAN, HARMEAN, TRIMMEAN, MAXA, MINA — these had no fixture coverage over a plain vertical range, so they silently regressed to #NUM!/#N/A until fixed here.

No existing conformance fixtures were touched.

How to test

From the core repo root:

cargo test -p truecalc-workbook --test vertical_range_spill_tests

Or manually, in a workbook: set A1:A3 to 1, 2, 3, then in B1 enter =A1:A3*2 — it should spill down into B1:B3 as 2, 4, 6 (not sideways into B1:D1). Then try =SMALL(A1:A3,1), =GEOMEAN(A1:A3), =RANK(A1,A1:A3), =MAXA(A1:A3) over the same vertical range — each should return its real value instead of #NUM!/#N/A.

Review

Independent review: 1 finding fixed. The review flagged that the initial fix only patched the 4 functions caught by pinned fixtures (AVERAGE, MAX, MIN, criterion::flatten_to_vec) and left ~17 other statistical functions with a one-level-only array unwrap, which the reviewer confirmed regressed (SMALL, GEOMEAN, RANK, MAXA all returned #NUM!/#N/A over a plain vertical range on the pre-fix branch). All of those functions' local flatten helpers now recurse into nested arrays, and regression tests were added pinning the previously-broken cases.

Test plan

• cargo build --workspace
• cargo clippy --workspace -- -D warnings — clean
• cargo test --workspace — 3613 passed
• cargo test -p truecalc-workbook --test vertical_range_spill_tests — 2 passed (new regression coverage for SMALL/LARGE/GEOMEAN/RANK/MAXA/MINA/MODE/PERCENTRANK over a vertical range)

closes #724

---

^{Need help on this PR? Tag /codesmith with what you need. Autofix is disabled.}

[github-actions]

Test Coverage by Category

Category	Unit Tests	Google Sheets Conformance	Property Cases	Total
Array	42	547/547 ✓	1,000 (2×500)	1,589
Database	35	182/182 ✓	3,500 (7×500)	3,717
Date	369	418/418 ✓	2,500 (5×500)	3,287
Engineering	245	886/888 ⚠	5,500 (11×500)	6,633
Filter	11	80/80 ✓	4,500 (9×500)	4,591
Financial	149	1,208/1,208 ✓	2,000 (4×500)	3,357
Info	0	256/256 ✓	4,500 (9×500)	4,756
Logical	121	263/263 ✓	3,500 (7×500)	3,884
Lookup	69	392/392 ✓	1,000 (2×500)	1,461
Math	536	2,006/2,006 ✓	8,000 (16×500)	10,542
Operator	87	250/250 ✓	7,500 (15×500)	7,837
Parser	83	92/92 ✓	4,000 (8×500)	4,175
Statistical	483	3,156/3,156 ✓	5,000 (10×500)	8,639
Text	298	729/733 ⚠	4,000 (8×500)	5,031
Timezone	47	—	—	47
Volatile	0	—	3,500 (7×500)	3,500
Web	29	56/56 ✓	6,000 (12×500)	6,085
Total	2,835	10,521/10,527	66,000 (132×500)	~79,362

_{✓ = 100% passing · ⚠ = known deviation · The ~79,362 total counts formula evaluations (each conformance row and each property case = 1). GitHub Checks reports 3,606 Rust test functions: 2,835 unit + 159 property functions (shown as cases above) + 612 conformance/integration.}