port(OutputFactory_VCF.t): 24-row v2 port + 3 PH folded HGVS rows + Axis B#177
Open
sitekwb wants to merge 3 commits into
Open
port(OutputFactory_VCF.t): 24-row v2 port + 3 PH folded HGVS rows + Axis B#177sitekwb wants to merge 3 commits into
sitekwb wants to merge 3 commits into
Conversation
…ervation Ports the following rows from porting-tests/detailed_plans/OutputFactory_VCF.md: - Subtests #15, #17, #18, #19 — audit-trail re-citations of existing csq_escape micro-tests for sztywno-1:1 traceability. - Subtest #21 — string '0' preserved (not coerced to empty). - Subtest #22 — numeric 0 stringifies to "0" and passes csq_escape unchanged. - Subtest #23 — empty-string vs numeric-0 distinction confirmed at the csq_escape layer. - Subtest #28 — escape of mixed invalid chars (',', '|', ';', whitespace) in transcript stable-id; locks in vepyr's per-character whitespace semantic ('A__G' for two spaces) which diverges from Perl's run-collapse ('A_G'). Divergence tracked under subtest #20 firm blocked-future-work. 8 new tests, all passing. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
…s + Axis B Ports the following rows from porting-tests/detailed_plans/OutputFactory_VCF.md: Perl-side: - Subtest #4 — AnnotateVcfConfig::default() constructor sanity (vcf_sink.rs). - Subtest #7 — refseq BAM_EDIT block at positions 28-32 (golden_benchmark.rs). - Subtest #8 — default fields list first-22-tuple matches Perl-era layout (Allele through SYMBOL_SOURCE), with vepyr extending to 74 fields total. - Subtest #10 — merged mode dedups SOURCE (exactly one occurrence) and interleaves REFSEQ_MATCH/SOURCE/REFSEQ_OFFSET at the canonical positions. Axis B (vepyr-side accretive invariants): - B1 — pipe-count invariant: pipe-joined empty CSQ has field_count - 1 pipes across all modes (default/everything/refseq/merged). - B2 — Ensembl-mode field list excludes all REFSEQ_* fields and BAM_EDIT (catches accidental field-list bleed across modes). 6 new tests, all passing. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
…+ Axis B Adds tests/port_output_factory_vcf.rs covering the integration-port and e2e-port subtests from OutputFactory_VCF.t plus the Parser_HGVS folded integration rows (Decision §1.B=A 2026-05-27). Live integration tests (10 subtests passing): - #24 count_matches_input — one output row per input row - #25 first_row_csq_shape — ENST00000307301/352957 + multi-consequence invariants at chr21:25585733 C>T (rs142513484) - #26 last_row_multi_feature — ≥4 CSQ groups at chr21:25607517 - #27 sparse_row_8_col_output — 8-col VCF + CSQ on minimal INFO=. row - #29 everything_mode_per_group_pipe_count — 80-field --everything layout - #40 bcsq_preserved — input BCSQ=foo survives (only literal CSQ stripped) - #42 preamble_shape — ##fileformat / ##INFO=<ID=CSQ,...> / #CHROM\\tPOS\\t - #44 intergenic_sentinel_at_isolated_position — relaxed: vepyr emits intergenic_variant sentinel rather than empty INFO at transcript-free loci - #51 input_header_passthrough — ##contig/##INFO=<ID=SVLEN/SVTYPE,...> preserved through annotate Parser_HGVS folded (3 rows passing): - PH-2 coding_hgvsc_ensembl — chr21:25585733 C>T → ENST00000352957.*:c.991G>A - PH-3 coding_utr_hgvsc — UTR-relative c.*18G>A on ENST00000307301.* - PH-4 protein_hgvsp — ENSP00000284967.*:p.Ala331Thr (three-letter form) Port-time reclassifications (4 subtests firm-downed to blocked-future-work): - #36/#37/#38 trash_existing_csq: vepyr panics with DuplicateQualifiedField when input declares ##INFO=<ID=CSQ,...>. Real engine gap — proper CSQ-projection drop is needed. See Axis B B3 test which locks in the observed panic-on-CSQ-header behaviour as a regression guard. - #43 ##VEP= preamble line: vepyr's csq_header_description emits no standalone ##VEP= provenance line; only the ##INFO=<ID=CSQ,...> entry. Port-time addition: - Axis B B3 axis_b_b3_annotate_to_vcf_panics_when_input_declares_csq_header: locks in the observed DuplicateQualifiedField panic so a future CSQ-projection fix surfaces as a test transition. Architectural-no-analogue prose stubs for #34/#35 (non-VCF input parser) and #52/#53 (web_output sidecar) are documented at the bottom of the file with substantive justification per v2 paradigm. Blocked-future-work prose stubs for #5/#6/#11/#14/#16/#20/#31/#32/#33/ #39/#41/#43/#45-#50/#54 plus PH-1/PH-5 are documented at the bottom of the file with future-work entry references. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
v2-paradigm port of
ensembl-vep/t/OutputFactory_VCF.t(52 substantive Perl subtests) plus 5 Parser_HGVS folded rows (Decision §1.B=A 2026-05-27).src/annotate_provider.rs::tests— 8 csq_escape micro-tests + falsy-value preservation (Add optional distance column to nearest UDTF #15/Add cluster, complement, and subtract UDTVFs #17/Replace std HashMap with AHashMap in interval_join.rs build map #18/Replace std HashMap with AHashMap in pileup events/coverage contig maps #19/feat: Preserve extra columns in cluster and subtract UDTFs #21/Add bio-function-vep crate with lookup_variants table function #22/feat: Ensembl VEP port #23/Fix column alias mismatches between TranscriptConsequenceEngine and Ensembl cache parquet schema #28)src/golden_benchmark.rs::tests— 3 CSQ field-list tests (Bump actions/checkout from 4 to 6 #7/build(deps): bump datafusion from 50.3.0 to 51.0.0 in the datafusion group across 1 directory #8/feat: Add bio-function-ranges crate #10) + 2 Axis B (B1 pipe-count invariant, B2 Ensembl no-REFSEQ_MATCH)src/vcf_sink.rs::tests— 1 constructor sanity test (Add new crate: datafusion-bio-function-ranges (coverage & count_overlaps) #4)tests/port_output_factory_vcf.rscovering Bump Datafusion to 52 #24/feat: Upgrade DataFusion from 50.3.0 to 52.1.0 #25/fix: Update bio-formats rev to include noodles lzma-rust2 fix #26/VEP golden benchmark: cache schema mismatches and missing context tables #27/Feature/optimize unnest groupby (from fork repo) #29/P1: fallback path references non-existent cache columns #40/feat: Openspec #42/feat: optimize annotation context loading #44/feat: typed Arrow output schema, deadlock fix, CSQ skip for vepyr integration #51 + Axis B B3 panic-regression-guard.hgvs::format_hgvsc/format_hgvspon chr21:25585733 C>T.Coverage parity: 24/50 Perl-denominator = 48% (27/55 = 49% incl. PH folds). Status: 🟡 DONE_WITH_CONCERNS.
Port-time findings (DONE_WITH_CONCERNS)
Two new future-work entries created (see sibling
porting-testsPR):SchemaError(DuplicateQualifiedField)when input VCF declares##INFO=<ID=CSQ,...>. Real engine gap — proper CSQ-projection drop is needed. Locked in via Axis B B3 regression guard.##INFO=<ID=CSQ,...>Format line, no provenance##VEP=line.Both deferred to future-work entries in
porting-tests/future-work-vepyr.md(S effort each).Test plan
cargo test -p datafusion-bio-function-vep --lib: 847 passed, 0 failedcargo test -p datafusion-bio-function-vep --test port_output_factory_vcf: 13 passed, 0 failedenv -u CONDA_PREFIXused for all cargo invocationsgolden.vcf, nogen_golden.sh, no docker dependency in testsDetailed plan:
porting-tests/detailed_plans/OutputFactory_VCF.md(post-Phase-D + Parser_HGVS folds)TDD plan:
porting-tests/plans/2026-05-28-port-output-factory-vcf.md🤖 Generated with Claude Code