Technical alignment: cross-source package metadata verification as a downstream QA layer
まだ誰も着手していません。
評価
調査の方向性
This is a design discussion, not a code task: the three questions are for maintainers to answer about purl-validator's roadmap, its scope, and whether it wants a report format or ScanCode.io hook. Start by checking the repo's roadmap and existing validation code for scope, and read the linked pkgmeta-verify schemas/pmv-output.schema.json only for context. Done means the maintainers answer the scope and interface questions, which needs their input before anyone builds.
索引モデルが issue の本文から書いたものです。
説明
Technical alignment: cross-source package metadata verification as a downstream QA layer
Hi AboutCode team — I'm building a small, independent verification component that consumes existing metadata sources (registry APIs, deps.dev, ClearlyDefined, PurlDB) and produces verdicts about cross-source consistency with provenance, timestamps, and reproducible evidence. It is explicitly not another aggregator or scanner: it does not rebuild PURL databases, license scanning, or vulnerability deduplication. Think of it as a downstream QA layer that could plug into aggregation pipelines (e.g. CodeSupply's).
Before going further I wanted to check technical fit with purl-validator, since it is the closest existing piece.
What it does (PMV v0.1, Apache-2.0): https://github.com/yijigao/pkgmeta-verify
- Normalizes the same PURL across ≥2 sources per ecosystem (npm, PyPI), then adjudicates each field/version into one of:
REAL_ERROR/SYNC_DELAY/SEMANTIC_DIVERGENCE/INDETERMINATE(undecidable), each with source, timestamp, and an audit trail. - Hard rules: digest comparison only for same artifact + same algorithm; missing publication time →
undecidable, never assumed delay; empty evidence →undecidable, neverCONSISTENT. - Output schema (JSON Schema 2020-12, all instances validated):
schemas/pmv-output.schema.jsonin the repo.
Two real cases it caught (snapshots + SHA-256 manifest in evidence/):
pkg:npm/[email protected]→SEMANTIC_DIVERGENCE(license expression conjunction mismatch across sources; neither side auto-accepted).pkg:pypi/[email protected]→SYNC_DELAY(downstream lag, reported as a provable delay range only — fetch time is not treated as upstream update time).
Concrete questions:
- Does
purl-validator's roadmap include version-level existence / cross-source field-consistency checks, or is that intentionally out of scope? (I saw it currently validates base PURL existence.) - Is there an interface you'd prefer for consuming third-party verdicts like these (e.g. a specific report format, or a hook in ScanCode.io pipelines)?
- Any objection to me citing purl-validator as the "base existence layer" and this component as the "field/version consistency layer" in a funding proposal? No endorsement implied — just accurate layering.
This is a technical alignment ping only; I'm not asking for endorsement or merge commitment. Any feedback on fit or interface is welcome.
- 主要言語
- Python
- スター
- 0
- フォーク
- 2
- PR マージ指標
- 30日以内にマージされた PR はありません
環境構築
このプロジェクトには開発コンテナ、Dockerfile、コントリビューションガイドがありません。まず README を読み、一般的な手順ははじめてのコントリビューションガイドを参照してください。
はじめの一歩
- issue を最後まで読み、次にプロジェクトのコントリビューションガイドを読みます。
- 着手することを issue にコメントします — 二人が同じ作業をするのを防げます。
- リポジトリをフォークし、ブランチを切って変更します。
- issue 番号を参照したプルリクエストを送ります。
aboutcode-org/purl-validator のほかの issue
-
難易度 5/5 1週間以上 初心者へのやさしさ 30/100
aboutcode-org/purl-validator#31 · コメント 1 件 ·
-
難易度 5/5 1週間以上 初心者へのやさしさ 25/100
-
Check type of PURLオープン
難易度 3/5 1〜2日 初心者へのやさしさ 58/100
aboutcode-org/purl-validator#29 · コメント 2 件 · リアクション 2 件 ·
aboutcode-org/purl-validator の issue をすべて見る
似ている issue
-
_chunk_scan_fwd launches its Triton kernel without a device guard: NaN when a model is split across GPUs対応中かも @toddwbucy が今日担当しました。 オープン
難易度 1/5 1時間未満 初心者へのやさしさ 85/100
state-spaces/mamba#1072 ·
-
難易度 1/5 1時間未満 初心者へのやさしさ 85/100
loopj/aiovantage#414 ·
メンテナーはふだん 1 日以内に返信
-
power_mismatch_deline docstring: the rear-to-total RMAD conversion has an extra bifaciality factorオープン
難易度 1/5 1時間未満 初心者へのやさしさ 85/100
pvlib/pvlib-python#2929 ·
メンテナーはふだん 1 日以内に返信
-
data good first issue launch-day unverified
難易度 2/5 1〜3時間 初心者へのやさしさ 78/100
BradGroux/wow-forever-mission-control#8 ·
メンテナーはふだん 1 日以内に返信
-
check bilibili falsely reports read/subtitle as broken: live_test fixture BV1GJ411x7h7 is gone (404)オープン
難易度 2/5 1〜3時間 初心者へのやさしさ 78/100
Panniantong/Agent-Reach#889 ·
メンテナーはふだん 2 日以内に返信