From 3960b44f99199a939a80886b96b11008c4c5f613 Mon Sep 17 00:00:00 2001
From: Jonathan Kingston <jkingston@duckduckgo.com>
Date: Fri, 19 Sep 2025 10:53:48 +0100
Subject: [PATCH 1/4] Migrate active selection language script

---
 injected/src/features/page-context.js       | 10 ++++++
 injected/src/features/page-context/utils.js | 40 +++++++++++++++++++++
 2 files changed, 50 insertions(+)
 create mode 100644 injected/src/features/page-context/utils.js

diff --git a/injected/src/features/page-context.js b/injected/src/features/page-context.js
index 797daf5da1..f18cb510c4 100644
--- a/injected/src/features/page-context.js
+++ b/injected/src/features/page-context.js
@@ -1,6 +1,7 @@
 import ContentFeature from '../content-feature.js';
 import { getFaviconList } from './favicon.js';
 import { isDuckAi, isBeingFramed, getTabUrl } from '../utils.js';
+import { getActiveSelectionLanguage } from './page-context/utils.js';
 const MSG_PAGE_CONTEXT_RESPONSE = 'collectionResult';
 
 function collapseWhitespace(str) {
@@ -76,12 +77,21 @@ export default class PageContext extends ContentFeature {
     listenForUrlChanges = true;
 
     init() {
+        this.setupActiveSelectionLanguageListener();
         if (!this.shouldActivate()) {
             return;
         }
         this.setupListeners();
     }
 
+    setupActiveSelectionLanguageListener() {
+        this.messaging.subscribe('getActiveSelectionLanguage', () => {
+            const activeSelectionLanguage = getActiveSelectionLanguage();
+            this.log.info('Active selection language', activeSelectionLanguage);
+            this.messaging.notify('getActiveSelectionLanguageResult', { activeSelectionLanguage });
+        });
+    }
+
     setupListeners() {
         this.observeContentChanges();
         if (this.getFeatureSettingEnabled('subscribeToCollect', 'enabled')) {
diff --git a/injected/src/features/page-context/utils.js b/injected/src/features/page-context/utils.js
new file mode 100644
index 0000000000..9d5c980659
--- /dev/null
+++ b/injected/src/features/page-context/utils.js
@@ -0,0 +1,40 @@
+export function getActiveSelectionLanguage() {
+    // BCP 47 language tag regex
+    const bcp47 = /^(?:en-GB-oed|i-ami|i-bnn|i-default|i-enochian|i-hak|i-klingon|i-lux|i-mingo|i-navajo|i-pwn|i-tao|i-tay|i-tsu|sgn-BE-FR|sgn-BE-NL|sgn-CH-DE|art-lojban|cel-gaulish|no-bok|no-nyn|zh-guoyu|zh-hakka|zh-min|zh-min-nan|zh-xiang|(?<langtag>(?<language>[A-Za-z]{2,3}(?<extlang>-[A-Za-z]{3}(-[A-Za-z]{3}){0,2})?|[A-Za-z]{4,8})(?:-(?<script>[A-Za-z]{4}))?(?:-(?<region>[A-Za-z]{2}|[0-9]{3}))?(?<variants>(?:-(?:[0-9A-Za-z]{5,8}|[0-9][0-9A-Za-z]{3}))*)(?<extensions>(?:-(?:[0-9A-WY-Za-wy-z](?:-[0-9A-Za-z]{2,8})+))*)(?:-(?<privateuse>x(?:-[0-9A-Za-z]{1,8})+))?)|(?<privateuse>x(?:-[0-9A-Za-z]{1,8})+))$/mgi
+    
+    // Get the Selection from the currently focused document (top or same-origin iframe)
+    function getActiveSelection() {
+        const activeElement = document.activeElement;
+        if (activeElement && activeElement.tagName === 'IFRAME') {
+            try {
+                return activeElement.contentWindow?.getSelection() || null;
+            } catch {
+                // Cross-origin iframe: cannot access its selection
+            }
+        }
+        return window.getSelection?.() || null;
+    }
+    
+    const selection = getActiveSelection();
+    const startContainer = selection?.rangeCount ? selection.getRangeAt(0).startContainer : null;
+
+    // If no selection, use the document's lang (or null if absent)
+    if (!startContainer) {
+        return document.documentElement.getAttribute('lang') || null;
+    }
+
+    // If the start is a text node, step up to its parent element
+    const startElement =
+        startContainer.nodeType === Node.ELEMENT_NODE
+            ? startContainer
+            : startContainer.parentElement;
+
+    // Look up the DOM tree for a lang attribute; fallback to the document's lang
+    const lang = startElement?.closest('[lang]')?.getAttribute('lang') ||
+        document.documentElement.getAttribute('lang');
+    
+    // Validate the lang attribute against the BCP 47 regex
+    if (lang && bcp47.test(lang)) return lang;
+    
+    return null;
+}
\ No newline at end of file

From 3820290c93fedd7abd47e15c9e5f9ebcad581e9c Mon Sep 17 00:00:00 2001
From: Jonathan Kingston <jkingston@duckduckgo.com>
Date: Fri, 19 Sep 2025 10:58:50 +0100
Subject: [PATCH 2/4] lint fixes

---
 injected/src/features/page-context/utils.js | 24 ++++++++++-----------
 1 file changed, 12 insertions(+), 12 deletions(-)

diff --git a/injected/src/features/page-context/utils.js b/injected/src/features/page-context/utils.js
index 9d5c980659..f550bf2267 100644
--- a/injected/src/features/page-context/utils.js
+++ b/injected/src/features/page-context/utils.js
@@ -1,20 +1,23 @@
 export function getActiveSelectionLanguage() {
     // BCP 47 language tag regex
-    const bcp47 = /^(?:en-GB-oed|i-ami|i-bnn|i-default|i-enochian|i-hak|i-klingon|i-lux|i-mingo|i-navajo|i-pwn|i-tao|i-tay|i-tsu|sgn-BE-FR|sgn-BE-NL|sgn-CH-DE|art-lojban|cel-gaulish|no-bok|no-nyn|zh-guoyu|zh-hakka|zh-min|zh-min-nan|zh-xiang|(?<langtag>(?<language>[A-Za-z]{2,3}(?<extlang>-[A-Za-z]{3}(-[A-Za-z]{3}){0,2})?|[A-Za-z]{4,8})(?:-(?<script>[A-Za-z]{4}))?(?:-(?<region>[A-Za-z]{2}|[0-9]{3}))?(?<variants>(?:-(?:[0-9A-Za-z]{5,8}|[0-9][0-9A-Za-z]{3}))*)(?<extensions>(?:-(?:[0-9A-WY-Za-wy-z](?:-[0-9A-Za-z]{2,8})+))*)(?:-(?<privateuse>x(?:-[0-9A-Za-z]{1,8})+))?)|(?<privateuse>x(?:-[0-9A-Za-z]{1,8})+))$/mgi
-    
+    const bcp47 =
+        /^(?:en-GB-oed|i-ami|i-bnn|i-default|i-enochian|i-hak|i-klingon|i-lux|i-mingo|i-navajo|i-pwn|i-tao|i-tay|i-tsu|sgn-BE-FR|sgn-BE-NL|sgn-CH-DE|art-lojban|cel-gaulish|no-bok|no-nyn|zh-guoyu|zh-hakka|zh-min|zh-min-nan|zh-xiang|(?<langtag>(?<language>[A-Za-z]{2,3}(?<extlang>-[A-Za-z]{3}(-[A-Za-z]{3}){0,2})?|[A-Za-z]{4,8})(?:-(?<script>[A-Za-z]{4}))?(?:-(?<region>[A-Za-z]{2}|[0-9]{3}))?(?<variants>(?:-(?:[0-9A-Za-z]{5,8}|[0-9][0-9A-Za-z]{3}))*)(?<extensions>(?:-(?:[0-9A-WY-Za-wy-z](?:-[0-9A-Za-z]{2,8})+))*)(?:-(?<privateuse>x(?:-[0-9A-Za-z]{1,8})+))?)|(?<privateuse>x(?:-[0-9A-Za-z]{1,8})+))$/gim;
+
     // Get the Selection from the currently focused document (top or same-origin iframe)
     function getActiveSelection() {
         const activeElement = document.activeElement;
         if (activeElement && activeElement.tagName === 'IFRAME') {
             try {
-                return activeElement.contentWindow?.getSelection() || null;
+                // Type assertion to access contentWindow safely
+                const iframe = /** @type {HTMLIFrameElement} */ (activeElement);
+                return iframe.contentWindow?.getSelection() || null;
             } catch {
                 // Cross-origin iframe: cannot access its selection
             }
         }
         return window.getSelection?.() || null;
     }
-    
+
     const selection = getActiveSelection();
     const startContainer = selection?.rangeCount ? selection.getRangeAt(0).startContainer : null;
 
@@ -25,16 +28,13 @@ export function getActiveSelectionLanguage() {
 
     // If the start is a text node, step up to its parent element
     const startElement =
-        startContainer.nodeType === Node.ELEMENT_NODE
-            ? startContainer
-            : startContainer.parentElement;
+        startContainer.nodeType === Node.ELEMENT_NODE ? /** @type {Element} */ (startContainer) : startContainer.parentElement;
 
     // Look up the DOM tree for a lang attribute; fallback to the document's lang
-    const lang = startElement?.closest('[lang]')?.getAttribute('lang') ||
-        document.documentElement.getAttribute('lang');
-    
+    const lang = startElement?.closest('[lang]')?.getAttribute('lang') || document.documentElement.getAttribute('lang');
+
     // Validate the lang attribute against the BCP 47 regex
     if (lang && bcp47.test(lang)) return lang;
-    
+
     return null;
-}
\ No newline at end of file
+}

From 616b45a4930f2533f7533257d4356a613efdb774 Mon Sep 17 00:00:00 2001
From: Jonathan Kingston <jkingston@duckduckgo.com>
Date: Fri, 19 Sep 2025 11:00:11 +0100
Subject: [PATCH 3/4] Add feature gating

---
 injected/src/features/page-context.js | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/injected/src/features/page-context.js b/injected/src/features/page-context.js
index f18cb510c4..08276dfea4 100644
--- a/injected/src/features/page-context.js
+++ b/injected/src/features/page-context.js
@@ -85,6 +85,9 @@ export default class PageContext extends ContentFeature {
     }
 
     setupActiveSelectionLanguageListener() {
+        if (!this.getFeatureSettingEnabled('subscribeToActiveSelectionLanguage', 'enabled')) {
+            return;
+        }
         this.messaging.subscribe('getActiveSelectionLanguage', () => {
             const activeSelectionLanguage = getActiveSelectionLanguage();
             this.log.info('Active selection language', activeSelectionLanguage);

From 57786d764265582c564e2bb93774b5f766f03544 Mon Sep 17 00:00:00 2001
From: Jonathan Kingston <jkingston@duckduckgo.com>
Date: Fri, 19 Sep 2025 11:05:34 +0100
Subject: [PATCH 4/4] Add schema stuff

---
 .../page-context/collect.subscribe.json       |  8 ++
 .../page-context/collectionResult.notify.json | 91 +++++++++++++++++++
 .../getActiveSelectionLanguage.subscribe.json |  8 ++
 ...tActiveSelectionLanguageResult.notify.json | 15 +++
 4 files changed, 122 insertions(+)
 create mode 100644 injected/src/messages/page-context/collect.subscribe.json
 create mode 100644 injected/src/messages/page-context/collectionResult.notify.json
 create mode 100644 injected/src/messages/page-context/getActiveSelectionLanguage.subscribe.json
 create mode 100644 injected/src/messages/page-context/getActiveSelectionLanguageResult.notify.json

diff --git a/injected/src/messages/page-context/collect.subscribe.json b/injected/src/messages/page-context/collect.subscribe.json
new file mode 100644
index 0000000000..90609085ca
--- /dev/null
+++ b/injected/src/messages/page-context/collect.subscribe.json
@@ -0,0 +1,8 @@
+{
+  "$schema": "http://json-schema.org/draft-07/schema#",
+  "type": "object",
+  "title": "Collect",
+  "description": "Subscription to collect page content data",
+  "properties": {},
+  "additionalProperties": false
+}
diff --git a/injected/src/messages/page-context/collectionResult.notify.json b/injected/src/messages/page-context/collectionResult.notify.json
new file mode 100644
index 0000000000..8587f3c032
--- /dev/null
+++ b/injected/src/messages/page-context/collectionResult.notify.json
@@ -0,0 +1,91 @@
+{
+  "$schema": "http://json-schema.org/draft-07/schema#",
+  "type": "object",
+  "title": "CollectionResult",
+  "description": "Result of page content collection, either success or error",
+  "oneOf": [
+    {
+      "type": "object",
+      "title": "Collection Success",
+      "required": ["success", "data"],
+      "properties": {
+        "success": {
+          "const": true
+        },
+        "data": {
+          "type": "object",
+          "description": "Collected page data",
+          "required": ["favicon", "title", "content", "truncated", "timestamp", "url"],
+          "properties": {
+            "favicon": {
+              "type": "array",
+              "description": "List of favicon URLs",
+              "items": {
+                "type": "string"
+              }
+            },
+            "title": {
+              "type": ["string", "null"],
+              "description": "Page title"
+            },
+            "metaDescription": {
+              "type": ["string", "null"],
+              "description": "Page meta description"
+            },
+            "content": {
+              "type": "string",
+              "description": "Main page content in markdown format"
+            },
+            "truncated": {
+              "type": "boolean",
+              "description": "Whether content was truncated due to size limits"
+            },
+            "headings": {
+              "type": "array",
+              "description": "List of page headings",
+              "items": {
+                "type": "string"
+              }
+            },
+            "links": {
+              "type": "array",
+              "description": "List of page links",
+              "items": {
+                "type": "string"
+              }
+            },
+            "images": {
+              "type": "array",
+              "description": "List of image URLs",
+              "items": {
+                "type": "string"
+              }
+            },
+            "timestamp": {
+              "type": "number",
+              "description": "Unix timestamp when data was collected"
+            },
+            "url": {
+              "type": "string",
+              "description": "Page URL"
+            }
+          }
+        }
+      }
+    },
+    {
+      "type": "object",
+      "title": "Collection Error",
+      "required": ["success", "error"],
+      "properties": {
+        "success": {
+          "const": false
+        },
+        "error": {
+          "type": "string",
+          "description": "Error message describing what went wrong"
+        }
+      }
+    }
+  ]
+}
diff --git a/injected/src/messages/page-context/getActiveSelectionLanguage.subscribe.json b/injected/src/messages/page-context/getActiveSelectionLanguage.subscribe.json
new file mode 100644
index 0000000000..af75541f18
--- /dev/null
+++ b/injected/src/messages/page-context/getActiveSelectionLanguage.subscribe.json
@@ -0,0 +1,8 @@
+{
+  "$schema": "http://json-schema.org/draft-07/schema#",
+  "type": "object",
+  "title": "GetActiveSelectionLanguage",
+  "description": "Subscription to get the active selection's language",
+  "properties": {},
+  "additionalProperties": false
+}
diff --git a/injected/src/messages/page-context/getActiveSelectionLanguageResult.notify.json b/injected/src/messages/page-context/getActiveSelectionLanguageResult.notify.json
new file mode 100644
index 0000000000..6bafe6f60a
--- /dev/null
+++ b/injected/src/messages/page-context/getActiveSelectionLanguageResult.notify.json
@@ -0,0 +1,15 @@
+{
+  "$schema": "http://json-schema.org/draft-07/schema#",
+  "type": "object",
+  "title": "GetActiveSelectionLanguageResult",
+  "description": "Result of getting the active selection's language as BCP 47 language tag",
+  "required": ["activeSelectionLanguage"],
+  "properties": {
+    "activeSelectionLanguage": {
+      "type": ["string", "null"],
+      "description": "BCP 47 language tag of the active selection (e.g., 'en', 'en-US', 'fr-FR') or null if no language detected",
+      "pattern": "^(?:en-GB-oed|i-ami|i-bnn|i-default|i-enochian|i-hak|i-klingon|i-lux|i-mingo|i-navajo|i-pwn|i-tao|i-tay|i-tsu|sgn-BE-FR|sgn-BE-NL|sgn-CH-DE|art-lojban|cel-gaulish|no-bok|no-nyn|zh-guoyu|zh-hakka|zh-min|zh-min-nan|zh-xiang|(?:(?:[A-Za-z]{2,3}(?:-[A-Za-z]{3}(?:-[A-Za-z]{3}){0,2})?|[A-Za-z]{4,8})(?:-[A-Za-z]{4})?(?:-[A-Za-z]{2}|[0-9]{3})?(?:(?:-(?:[0-9A-Za-z]{5,8}|[0-9][0-9A-Za-z]{3}))*)(?:(?:-(?:[0-9A-WY-Za-wy-z](?:-[0-9A-Za-z]{2,8})+))*)(?:-(?:x(?:-[0-9A-Za-z]{1,8})+))?)|(?:x(?:-[0-9A-Za-z]{1,8})+))$",
+      "examples": ["en", "en-US", "fr-FR", "zh-CN", null]
+    }
+  }
+}