Repository navigation
rosetta[sqlite]: cross-language consistency — 8 red / 2 amber metrics, 4 ledgered shapes #2601
Description
Part of #2560 — per-language tracking for the keyword-rosetta control corpus (built for #1096). Planted intent is identical in every language, so every deviation below is measured bias vs. the 46-language median, validated in the corpus's deviation ledger. 🔴 = beyond ±50% of the cross-language median, 🟡 = ±25–50%.
Triage order (per the epic): structure detection first, then signal/keyword rules, then risk-score consequences — §3's numbers are downstream of §1–§2 and re-baseline as those fix. Cross-cutting engine issues (e.g. #2535/#2546) fix many languages at once; this issue tracks this language's residual after each cross-cutting fix lands, and closes when the language sits inside the ±25% band on every comparable metric (or every remaining deviation is ledgered as intended morphology).
1. Structure detection (fix first)
- 🔴
args: 1 vs median 13 (-92%) - 🔴
functions_found: 28 vs median 13 (+115%)
2. Signal / keyword deviations
- 🔴
io: 11 vs median 3 (+267%) - 🔴
keyword_hits: 116 vs median 247.5 (-53%) - 🔴
safety_bypasses: 0 vs median 2 (-100%) - 🟡
state_mutation: 3 vs median 6 (-50%)
3. Risk-exposure consequences (downstream)
- 🔴
risk_cognitive_load: 5.53524 vs median 15.1444 (-63%) - 🔴
risk_state_flux: 0 vs median 19.9999 (-100%) - 🔴
risk_tech_debt: 60 vs median 39.9576 (+50%) - 🟡
risk_spec_match: 64 vs median 96 (-33%)
Ledgered shapes touching sqlite
| defect | type | issue | evidence | summary |
|---|---|---|---|---|
census-requires-git-tracked |
engine-semantic | — | — | GalaxyScope's census enumerates git-tracked files only; an untracked folder scans as '0 files mapped' with no per-file warning |
sqlite-dot-read-dual-import-io |
keyword-overlap | — | (notes) | ^.read counts BOTH import and io (finditer-confirmed) — every corpus chain link adds +1 io in sqlite |
orphan-conversion-opaque-counts |
upstream-question | #2547 | (notes) | On languages without a same-file call graph the orphan census includes non-function shapes, so the orphan->api conversion count diverges from visible function counts: sh… |
string-literal-selective-shielding |
upstream-question | #2535 | b.sql | Keywords inside one string literal ('if eval fails, try open') count toward branch (+2), safety (+1), io (+1) but eval does NOT count toward high_risk_execution (routes … |
Decoy outcomes (file-level evidence)
- main.sql ·
never detaches a database and no shell word(comment): stripped: contributed 0 - b.sql ·
SELECT 'IF UPDATE FAILS TRY SELECT AGAIN'(string): UPDATE and SELECT inside the literal COUNT (io +2, state +1); IF/TRY inert (sqlite has no bare IF/TRY keywords). io 5 = planted UPDATE + carrier SELECT + .read + 2 from the literal. Ledger: string-literal-selective-shielding - c.sql ·
hack_level;(code): inert (underscore + lowercase)
Reproduce / evidence
- Corpus folder:
data/sqlite/—expected_signals.jsoncarries the shape-by-shape accounting - Gate:
GALAXYSCOPE_BIN=<venv>/bin/galaxyscope python tools/verify_language.py sqlitein the corpus repo - Full context: findings_by_language.md#sqlite · bias_report.md · bias chart
Activity
squid-protocol commentedon Sep 3, 2026
Status: 3🔴 / 3🟡 unexplained, 12 further out-of-band cell(s) accounted for
Generated from
docs/bias_data.json+deviation_ledger.jsonbytools/issue_status.py, using the same verdict machinerybias_report.py --gateruns on (gitgalaxy#2669 E.1), so this comment and the gate cannot disagree about what is left.Still unexplained — the work this issue tracks
functions_foundkeyword_hitscoding_loctotal_locrisk_api_exposurerisk_tech_debtOut of band, but accounted for
derived — the deviation entered through an input that is itself out of band -- the same finding counted twice
avg_func_loc— inherits total_loc, coding_loccog_raw— inherits branch, args, total_loc, coding_loccontrol_flow_ratio— inherits branchfunc_internal_density— inherits branch, argsmax_func_complexity— inherits branchledgered — a validated ledger entry names this language and this metric
args— args-no-parameter-surface-morphologyavg_func_args— args-no-parameter-surface-morphologybranch— string-literal-selective-shieldingio— string-literal-selective-shielding, sqlite-dot-read-dual-import-iorisk_cognitive_load— path-and-extension-modifiersrisk_documentation— path-and-extension-modifiersrisk_state_flux— path-and-extension-modifiersClose criterion (gitgalaxy#2669 Batch E.4)
python tools/language_deviations.py sqliteexits 1: 6 unexplained cell(s) remain. Each needs a fix, or a validated ledger entry saying why it is correct, before this issue can close.squid-protocol commentedon Sep 6, 2026
Closed 2026-09-06 as superseded by #2812 (
docs/contract_roadmap.md, decision D2). Every out-of-band cell for sqlite is accounted for (the gate reads 0 unexplained) and is now tracked by cause family, not by language: the ledger entry that explains it, its cause category in keyword-rosetta'sdocs/bias_data.json["cell_categories"](echo / inherency / extraction / correlation / scoring), and the family issue #2812 lists for that cause. Live standing at any time:python tools/language_deviations.py sqlitein keyword-rosetta, ordocs/findings_by_language.md#sqlite.This is a change of axis, not a claim that sqlite is green. The §10 capstone in
docs/language_status/sqlite.mdis written when the language reads clean bylanguage_deviations.py, by whichever family PR gets it there — no longer tied to this issue.