solo: --soloCellReadStats CB writes CellReads.stats

[BenjaminDEMAILLE]

--soloCellReadStats CB writes Solo.out/<feature>/CellReads.stats.

What changed

One row per cell barcode with the fifteen counters STARsolo reports: how the barcode matched (cbMatch, cbPerfect, cbMMunique, cbMMmultiple), whether the read mapped to one genomic locus or several (genomeU/genomeM), whether it landed on a feature (featureU/featureM), where in the gene and on which strand (exonic, intronic, exonicAS, intronicAS), whether it was mitochondrial (mito), and whether it reached the matrix (countedU/countedM). Then nUMIunique and nGenesUnique, taken from the raw matrix so they agree with it by construction.

--genomeChrSetMitochondrial (new) names the chromosomes behind the mito column. Without it the column is zero throughout, which is honest: no chromosome was declared mitochondrial.

Reads whose barcode never resolved are summed into a single CBnotInPasslist row rather than dropped. That row is the reason the file is useful: it separates "these cells look thin" from "most of the input never reached a cell at all".

The region columns split by strand — an antisense read counts under exonicAS or intronicAS, never under exonic or intronic.

Divergence

DIVERGENCE.md §3.2. STAR emits these rows by walking a libc++ unordered_map, which at these map sizes is the reverse of each barcode's first appearance. That is reproduced, including across threads: the accumulator merges in read order, so a threaded run writes the same file as a serial one. It stops being reproducible past the point where libc++ rehashes, since the order then depends on the bucket count. The values never differ, only which line they sit on.

Verification

Six unit tests in src/solo/cell_reads.rs:

• each_row_fills_the_header — twenty columns, every row complete
• flags_accumulate_per_read — cbMatch counts every read; each flag counts only its own
• unresolved_reads_land_in_the_passlist_miss_row
• rows_are_emitted_in_reverse_first_appearance_order — the divergence above
• umi_and_gene_totals_come_from_the_matrix — a cell absent from the matrix prints zeros rather than inheriting a read count

Reads are folded in under a mutex held by SoloContext, so the row order is the order reads were processed in regardless of --runThreadN.

Gate: 565 lib + 26 integration tests, cargo clippy --all-targets -- -D warnings, cargo fmt --check, MSRV 1.89 — all green.

Output-neutral: the file is written only when --soloCellReadStats CB asks for it, and nothing else changes.

Split out of #152 following the one-theme rule in CONTRIBUTING.md.