Shell header identity: device-inline evidence flips nav overlay mode; style whitespace breaks equivalence
メンテナーはふだん 1 日以内に返信
まだ誰も着手していません。
評価
調査の方向性
Start with WordPressSitePlan/ShellExtraction.php, especially absorbResponsiveVariantLandmarks() and normalizeNestedChromeMarkup(), then trace NavigationToggleSuppressor::navigationOverlayMenu() and its source-variant checks. Add neutral regressions for descendant tablet-only inline evidence and whitespace differences in inline style declarations. Done means both route pairs project equivalent navigation and extract one shared header part.
索引モデルが issue の本文から書いたものです。
説明
Problem
Shell extraction (WordPressSitePlan/ShellExtraction.php, absorbResponsiveVariantLandmarks() → per-variant identity_markup signatures) excludes routes as non_equivalent when their source header is identical but incidental capture evidence differs. Two independent causes were observed on one multi-route capture (7 routes; only 4 got the shared header part):
1. Per-device inline evidence on one menu item flips the whole navigation's overlay mode
On one route, a single <li> inside the desktop nav carries tablet-only inline evidence (width: 93px instead of the desktop 88px):
<li … style="position: relative;…;width: 88px;height: 38px"
data-blocks-engine-capture-device-inline="{"tablet":"position: relative;…;width: 93px;height: 38px"}">
That route projects the nav as:
overlayMenu:"mobile"blocks-engine-native-responsive-navigation blocks-engine-navigation-overlay-…- the parent group gets
blocks-engine-sidebar-navigation-carrier
The other routes project overlayMenu:"never" with no carrier. The source has no toggle for this nav; the phone layout is a separate device document. Removing only that one attribute makes the route project never and join the shared header.
A related route that is missing one coalesced device stylesheet (a separate DLA issue) flips the same way. That suggests NavigationToggleSuppressor::navigationOverlayMenu() (hasEquivalentSourceNavigationVariant() or navigationToggleControl()) is reading evidence that shouldn't decide the overlay mode.
2. The identity signature is sensitive to whitespace in serialized style values
After normalization, the home route's desktop signature differs from the others only in serialized inline styles inside block attributes:
"style":"visibility:inherit;overflow-x:visible" vs "style":"visibility: inherit;overflow-x: visible"
"itemStyle":"position:relative;box-sizing:border-box;…" vs "position: relative;box-sizing: border-box;…"
Same declarations, different whitespace. normalizeNestedChromeMarkup() does not canonicalize declaration text stored in block attribute JSON.
Expected
- A navigation's overlay mode is decided by real collapse evidence: a source toggle, or an equivalent source navigation variant within the same document. Device-inline style evidence on a descendant, or stylesheets scoped to another device document, don't turn a non-collapsing nav into a native responsive overlay.
- Shell identity treats CSS declaration lists that differ only in whitespace as equal (or the projection serializes them canonically, so both routes emit the same text).
Acceptance
- Neutral regression with two routes sharing an identical header nav, where one route has a descendant item with
data-blocks-engine-capture-device-inlinetablet evidence. Both project the sameoverlayMenuand the header becomes one shared part. - Neutral regression where two routes' headers differ only in whitespace inside an inline
style. They extract one shared header part.
AI disclosure: investigated and drafted by Claude (Opus) via the Pi coding agent (Franklin) for Chris Huber, from a Brandspell Wix import (DLA 0.33.169 / php-transformer 0.45.305).
- 主要言語
- PHP
- スター
- 25
- フォーク
- 3
- 平均マージ
- 2時間 23分
- マージ済み PR（30日）
- 942
環境構築
- Dockerfile・Docker Compose ファイルなし
- プルリクエストのテンプレートなし
- コントリビューションガイドを読む
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。
Automattic/blocks-engine のほかの issue
-
難易度 2/5 1〜3時間 初心者へのやさしさ 78/100
Automattic/blocks-engine#5003 ·
メンテナーはふだん 1 日以内に返信
-
難易度 2/5 1〜3時間 初心者へのやさしさ 70/100
Automattic/blocks-engine#4943 ·
メンテナーはふだん 1 日以内に返信
-
難易度 2/5 1〜3時間 初心者へのやさしさ 70/100
Automattic/blocks-engine#4942 ·
メンテナーはふだん 1 日以内に返信
-
難易度 2/5 1〜3時間 初心者へのやさしさ 75/100
Automattic/blocks-engine#4921 ·
メンテナーはふだん 1 日以内に返信
-
難易度 2/5 1〜3時間 初心者へのやさしさ 82/100
Automattic/blocks-engine#775 ·
メンテナーはふだん 1 日以内に返信
Automattic/blocks-engine の issue をすべて見る
似ている issue
-
enhancement priority Sales
難易度 2/5 1〜3時間 初心者へのやさしさ 78/100
novalo-dev/Prestashop-MCP-Server#201 ·
メンテナーはふだん 1 日以内に返信
-
enhancement
難易度 2/5 1〜3時間 初心者へのやさしさ 74/100
web-token/jwt-framework#757 ·
メンテナーはふだん 1 日以内に返信
-
難易度 2/5 1〜3時間 初心者へのやさしさ 75/100
メンテナーはふだん 1 日以内に返信
-
Include flows spanning the full report time window対応中かも @somethingwithproof が今日担当しました。 オープン
難易度 2/5 1〜3時間 初心者へのやさしさ 77/100
Cacti/plugin_flowview#306 ·
メンテナーはふだん 1 日以内に返信
-
bug origin: upstream priority: low
難易度 2/5 1〜3時間 初心者へのやさしさ 78/100
mattytap/Mattytap-Sermons#105 ·