Type building metric values

[esnya]

Summary

• replace nullable/tagged building metric values with concrete Missing/Known/Invalid metric value types
• remove the duplicate BuildingMetricValueKind tag so payload availability is represented by the runtime type
• update metric consumers and parser tests to pattern-match known values instead of reading nullable payloads

Verification

• dotnet restore PlateauResoniteLink.sln --locked-mode --disable-build-servers
• dotnet format whitespace . --folder --verify-no-changes
• dotnet build PlateauResoniteLink.sln --configuration Release --no-restore --disable-build-servers -m:1 -p:UseSharedCompilation=false
• dotnet test PlateauResoniteLink.sln --configuration Release --no-restore --verbosity normal -m:1 --disable-build-servers -p:UseSharedCompilation=false
• Fake Live Sink canonical scene dump compared with baseline via git diff --no-index; no diff

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 9b86cc6a06

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: ed18dd235d

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[esnya]

このままだと Close 寄りです。Missing / Invalid を runtime 型として下流 consumer に流すのではなく、parser/import 境界で diagnostic / skip に閉じ、metric pipeline は Known だけを受ける形にしてください。

そこまでやらないなら、nullable/tag の置き換えに留まるので Close でよいと思います。

[esnya]

レビュー対応しました。Missing / Invalid を BuildingAttributeContext に流す runtime variant は削除し、parser 境界で欠損・不正メトリックを落として BuildingMetricMeasurements には known 値だけを載せる形に変えました。下流 consumer は BuildingMetricKind ごとの known metric lookup のみを行うため、metric pipeline が Missing / Invalid を受け取る経路はなくなっています。\n\n検証:\n- dotnet restore PlateauResoniteLink.sln --locked-mode --disable-build-servers\n- dotnet format whitespace . --folder --verify-no-changes\n- dotnet build PlateauResoniteLink.sln --configuration Release --no-restore --disable-build-servers -m:1 -p:UseSharedCompilation=false\n- dotnet test PlateauResoniteLink.sln --configuration Release --no-restore --verbosity normal -m:1 --disable-build-servers -p:UseSharedCompilation=false\n- Fake Live Sink canonical dump for plateau-13213-higashimurayama-shi-2020 / 53395325 / dem,bldg / grid matched the existing semantic baseline with git diff --no-index.

[esnya]

コード量を縮小していく前提では、この PR はまだ受け入れにくいです。最終差分は +219/-87 で、BuildingAttributeContext の明示フィールドを BuildingMetricMeasurements + BuildingMetricKind + dictionary へ置き換えていますが、現時点の metric 種別は固定的で、消費側も既知 metric を数個読むだけです。非有限値や invalid raw を downstream に流さない目的は parser 側で小さく達成できる一方、この形は compile-time に見えていた属性スキーマを dictionary lookup にして呼び出し側の記述も増やしています。増加分に見合う強い必然性がまだ見えません。受け入れるなら、少なくとも現在必要な不正値排除だけに絞って差分を縮めるか、固定フィールドを捨てて map 化する必要性を現在の要件で説明できる形にしてほしいです。