Extract detection into a testable module, fix an over-matching bug

.github/workflows/test.yml

@@ -0,0 +1,17 @@
+name: test
+
+on:
+  push:
+    branches: [main]
+  pull_request:
+
+jobs:
+  test:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+      - uses: actions/setup-node@v4
+        with:
+          node-version: 22
+      - run: npm install --no-fund --no-audit
+      - run: npm test

.gitignore

@@ -0,0 +1,4 @@
+node_modules/
+package-lock.json
+*.zip
+.DS_Store

manifest.json

@@ -8,7 +8,7 @@
   "content_scripts": [
     {
       "matches": ["https://mail.google.com/*"],
-      "js": ["src/strip.js"],
+      "js": ["src/detect.js", "src/strip.js"],
       "run_at": "document_idle",
       "all_frames": true
     }

package.json

@@ -0,0 +1,13 @@
+{
+  "name": "mailsuitefree",
+  "version": "0.1.0",
+  "private": true,
+  "description": "Chrome extension that removes the Mailsuite promotional signature from outgoing Gmail messages",
+  "license": "MIT",
+  "scripts": {
+    "test": "node --test \"test/**/*.test.js\""
+  },
+  "devDependencies": {
+    "jsdom": "^24.1.0"
+  }
+}

src/detect.js

@@ -0,0 +1,168 @@
+/*
+ * Detection rules for the Mailsuite signature.
+ *
+ * Deliberately free of chrome.* APIs and of globals beyond the nodes handed in,
+ * so the same code runs inside Gmail and under jsdom in test/. src/strip.js
+ * holds the Gmail glue.
+ */
+(function (root, factory) {
+  const api = factory();
+  if (typeof module === 'object' && module.exports) module.exports = api;
+  else root.MailSuiteFreeDetect = api;
+})(typeof globalThis !== 'undefined' ? globalThis : this, function () {
+  'use strict';
+
+  const ELEMENT_NODE = 1;
+  const TEXT_NODE = 3;
+
+  const PROMO_HOST = /(^|\.)(mailtrack\.io|mailsuite\.com)$/i;
+
+  /* Visible signature wording, lifted from Mailsuite's own popup.bundle.js i18n
+     table (senderNotifiedSignatureText and its 12 / 13 / 15 / 17 / 18 variants,
+     every locale the extension ships). Only ever used to subtract known text
+     while deciding whether a block is nothing but signature. Never on its own
+     as grounds to delete something. */
+  const PHRASES = [
+    'sender notified by', 'sender notified with mailtrack',
+    'email delivery certified by', 'sent with mailtrack', 'sent with',
+    'remitente notificado con', 'entrega de email certificada por', 'enviado con mailtrack',
+    'remetente notificado com', 'enviado com mailtrack',
+    'expéditeur notifié par', "livraison d'e-mail certifiée par", 'envoyé avec mailtrack',
+    'absender benachrichtigt von', 'e-mail-zustellung zertifiziert von', 'gesendet mit mailtrack',
+    'mittente notificato da', 'consegna email certificata da', 'inviato con mailtrack',
+    'nadawca powiadomiony przez', 'dostarczenie e-maila potwierdzone przez', 'wysłano z mailtrack',
+    '送信者への通知', 'メール配信証明', 'mailtrackで送信',
+    '발신자 알림', '이메일 수신 확인', 'mailtrack로 전송',
+    '寄件者已透過', '已透過mailtrack寄出',
+    'प्रेषक को सूचित किया गया',
+  ];
+
+  const UNSUB = /unsubscribe|opt[-\s]?out|darse de baja|se désabonner|abmelden|annulla iscrizione|cancelar subscri|wypisz|取消訂閱|配信停止|수신 거부|सदस्यता समाप्त/i;
+
+  const PUNCTUATION = /[\s .,;:·|—–()[\]-]/g;
+
+  /** A link out to Mailsuite's own site, which is what the signature always is. */
+  const promoAnchor = (a) => Boolean(a.getAttribute('href')) && PROMO_HOST.test(a.hostname || '');
+
+  /** The read-tracking pixel, which has to survive whatever else we delete. */
+  const looksLikeBeacon = (img) => {
+    const src = img.getAttribute('src') || '';
+    if (/mailtrack\.io|mailsuite\.com/i.test(src)) return true;
+    const w = parseInt(img.getAttribute('width') || '', 10);
+    const h = parseInt(img.getAttribute('height') || '', 10);
+    return (w > 0 && w <= 3) || (h > 0 && h <= 3);
+  };
+
+  /** True when this element holds the signature and nothing of the user's own. */
+  const signatureOnly = (el) => {
+    /* An <img> contributes no textContent, so without this an element wrapping
+       somebody's inline screenshot next to the signature would read as
+       signature-only and take the screenshot with it. A real image is never
+       just a signature, whatever the text around it says. */
+    for (const img of el.querySelectorAll('img')) {
+      if (!looksLikeBeacon(img)) return false;
+    }
+
+    let text = el.textContent || '';
+    for (const a of el.querySelectorAll('a')) {
+      const label = a.textContent;
+      if (label && promoAnchor(a)) text = text.split(label).join(' ');
+    }
+    text = text.toLowerCase();
+    for (const phrase of PHRASES) text = text.split(phrase).join(' ');
+    return text.replace(PUNCTUATION, '') === '';
+  };
+
+  /**
+   * The block to delete for a given promo link, or null to leave it alone.
+   *
+   * The link's immediate parent has to be signature-only. That single check is
+   * what stops the extension eating a sentence in which somebody genuinely
+   * linked to mailsuite.com. From there it climbs as far as it can while every
+   * parent is still signature-only, so the whole block goes rather than a bare
+   * <a> stranded in leftover text.
+   */
+  const signatureBlock = (anchor, root) => {
+    const parent = anchor.parentElement;
+    if (!parent || parent === root || !signatureOnly(parent)) return null;
+    let block = parent;
+    while (
+      block.parentElement &&
+      block.parentElement !== root &&
+      signatureOnly(block.parentElement)
+    ) {
+      block = block.parentElement;
+    }
+    return block;
+  };
+
+  /**
+   * Move any tracking beacon out of the block before the block is deleted.
+   * Parked where the block stood rather than at the end of the body, so a
+   * signature that was not last does not send the pixel past what followed it.
+   */
+  const rescueBeacons = (block, root) => {
+    const parent = block.parentNode || root;
+    let rescued = 0;
+    for (const img of Array.from(block.querySelectorAll('img'))) {
+      if (!looksLikeBeacon(img)) continue;
+      parent.insertBefore(img, block);
+      rescued += 1;
+    }
+    return rescued;
+  };
+
+  /** Drop the blank lines Mailsuite leaves sitting above its signature. */
+  const trimBefore = (block) => {
+    let prev = block.previousSibling;
+    while (prev) {
+      const blank =
+        (prev.nodeType === TEXT_NODE && !prev.textContent.trim()) ||
+        (prev.nodeType === ELEMENT_NODE &&
+          (prev.tagName === 'BR' ||
+            (/^(DIV|P)$/.test(prev.tagName) &&
+              !prev.textContent.trim() &&
+              !prev.querySelector('img'))));
+      if (!blank) break;
+      const doomed = prev;
+      prev = prev.previousSibling;
+      doomed.remove();
+    }
+  };
+
+  /**
+   * Strip every signature under `root`, returning how many were removed.
+   * `keepUnsubscribe` leaves blocks carrying an opt-out link alone, since bulk
+   * mail is required to carry one under CAN-SPAM and GDPR.
+   */
+  const scrubRoot = (root, options) => {
+    const keepUnsubscribe = !options || options.keepUnsubscribe !== false;
+    let removed = 0;
+    for (const a of Array.from(root.querySelectorAll('a'))) {
+      if (!root.contains(a) || !promoAnchor(a)) continue;
+      const block = signatureBlock(a, root);
+      if (!block) continue;
+      if (keepUnsubscribe && UNSUB.test(block.textContent || '')) continue;
+      /* Trim first. rescueBeacons parks the pixel immediately before the block,
+         which would otherwise stop trimBefore seeing the blank line above it. */
+      trimBefore(block);
+      rescueBeacons(block, root);
+      block.remove();
+      removed += 1;
+    }
+    return removed;
+  };
+
+  return {
+    PROMO_HOST,
+    PHRASES,
+    UNSUB,
+    promoAnchor,
+    looksLikeBeacon,
+    signatureOnly,
+    signatureBlock,
+    rescueBeacons,
+    trimBefore,
+    scrubRoot,
+  };
+});

src/strip.js

@@ -1,42 +1,14 @@
 /*
- * Mailsuite Signature Stripper
- *
- * Removes the promotional signature that Mailsuite (formerly Mailtrack) appends
- * to outgoing Gmail messages.
- *
- * Two things it deliberately does NOT remove:
- *   1. The read-tracking beacon. It is lifted out of the signature block before
- *      the block is deleted, so open tracking keeps working.
- *   2. Any block containing an unsubscribe / opt-out link. Stripping those from
- *      bulk mail is a CAN-SPAM and GDPR problem. Off by default, see settings.
+ * Gmail glue. The rules for what counts as a signature live in src/detect.js,
+ * which is loaded first by the manifest.
  *
  * Nothing leaves the browser. This only edits the compose window's DOM.
  */
 (() => {
   'use strict';
 
-  const PROMO_HOST = /(^|\.)(mailtrack\.io|mailsuite\.com)$/i;
-
-  /* Visible signature wording, lifted from Mailsuite's own popup.bundle.js i18n
-     table (senderNotifiedSignatureText / 12 / 13 / 15 / 17 / 18, every locale
-     the extension ships). Used to decide whether a block is nothing but the
-     signature, never on its own to delete something. */
-  const PHRASES = [
-    'sender notified by', 'sender notified with mailtrack',
-    'email delivery certified by', 'sent with mailtrack', 'sent with',
-    'remitente notificado con', 'entrega de email certificada por', 'enviado con mailtrack',
-    'remetente notificado com', 'enviado com mailtrack',
-    'expéditeur notifié par', "livraison d'e-mail certifiée par", 'envoyé avec mailtrack',
-    'absender benachrichtigt von', 'e-mail-zustellung zertifiziert von', 'gesendet mit mailtrack',
-    'mittente notificato da', 'consegna email certificata da', 'inviato con mailtrack',
-    'nadawca powiadomiony przez', 'dostarczenie e-maila potwierdzone przez', 'wysłano z mailtrack',
-    '送信者への通知', 'メール配信証明', 'mailtrackで送信',
-    '발신자 알림', '이메일 수신 확인', 'mailtrack로 전송',
-    '寄件者已透過', '已透過mailtrack寄出',
-    'प्रेषक को सूचित किया गया',
-  ];
-
-  const UNSUB = /unsubscribe|opt[-\s]?out|darse de baja|se désabonner|abmelden|annulla iscrizione|cancelar subscri|wypisz|取消訂閱|配信停止|수신 거부|सदस्यता समाप्त/i;
+  const detect = globalThis.MailSuiteFreeDetect;
+  if (!detect) return;
 
   /* Gmail's compose body, across layouts and locales. */
   const COMPOSE = [
@@ -53,91 +25,12 @@
     for (const key of Object.keys(changes)) settings[key] = changes[key].newValue;
   });
 
-  const promoAnchor = (a) => {
-    const href = a.getAttribute('href');
-    if (!href) return false;
-    try {
-      return PROMO_HOST.test(new URL(href, location.href).hostname);
-    } catch {
-      return false;
-    }
-  };
-
-  const looksLikeBeacon = (img) => {
-    const src = img.getAttribute('src') || '';
-    if (/mailtrack\.io|mailsuite\.com/i.test(src)) return true;
-    const w = parseInt(img.getAttribute('width') || '', 10);
-    const h = parseInt(img.getAttribute('height') || '', 10);
-    return (w > 0 && w <= 3) || (h > 0 && h <= 3);
-  };
-
-  /* True when this element holds the signature and none of the user's own text. */
-  const signatureOnly = (el) => {
-    let text = el.textContent || '';
-    for (const a of el.querySelectorAll('a')) {
-      const label = a.textContent;
-      if (label && promoAnchor(a)) text = text.split(label).join(' ');
-    }
-    text = text.toLowerCase();
-    for (const phrase of PHRASES) text = text.split(phrase).join(' ');
-    return text.replace(/[\s .,;:·|—–()[\]-]/g, '') === '';
-  };
-
-  /* Climb from the promo link to the outermost wrapper that is still nothing but
-     signature, so the whole block goes and not just the <a>. Stops the moment a
-     parent contains something the user wrote. */
-  const signatureBlock = (anchor, root) => {
-    let block = anchor;
-    while (block.parentElement && block.parentElement !== root) {
-      if (!signatureOnly(block.parentElement)) break;
-      block = block.parentElement;
-    }
-    return block;
-  };
-
-  /* Move the read-tracking beacon out of the block before deleting it. */
-  const rescueBeacons = (block, root) => {
-    for (const img of [...block.querySelectorAll('img')]) {
-      if (looksLikeBeacon(img)) root.appendChild(img);
-    }
-  };
-
-  /* Drop the blank lines Mailsuite leaves sitting above its signature. */
-  const trimBefore = (block) => {
-    let prev = block.previousSibling;
-    while (prev) {
-      const blank =
-        (prev.nodeType === Node.TEXT_NODE && !prev.textContent.trim()) ||
-        (prev.nodeType === Node.ELEMENT_NODE &&
-          (prev.tagName === 'BR' ||
-            (/^(DIV|P)$/.test(prev.tagName) &&
-              !prev.textContent.trim() &&
-              !prev.querySelector('img'))));
-      if (!blank) break;
-      const doomed = prev;
-      prev = prev.previousSibling;
-      doomed.remove();
-    }
-  };
-
-  const scrubRoot = (root) => {
-    let removed = 0;
-    for (const a of [...root.querySelectorAll('a')]) {
-      if (!root.contains(a) || !promoAnchor(a)) continue;
-      const block = signatureBlock(a, root);
-      if (settings.keepUnsubscribe && UNSUB.test(block.textContent || '')) continue;
-      rescueBeacons(block, root);
-      trimBefore(block);
-      block.remove();
-      removed += 1;
-    }
-    return removed;
-  };
-
   const scrubAll = () => {
     if (!settings.enabled) return 0;
     let removed = 0;
-    for (const root of document.querySelectorAll(COMPOSE)) removed += scrubRoot(root);
+    for (const root of document.querySelectorAll(COMPOSE)) {
+      removed += detect.scrubRoot(root, { keepUnsubscribe: settings.keepUnsubscribe });
+    }
     if (removed) {
       settings.removed += removed;
       chrome.storage.local.set({ removed: settings.removed });
@@ -146,7 +39,7 @@
   };
 
   /* Mailsuite drops the signature into the compose box while you are still
-     writing, so watching for it is the main mechanism. Debounced because Gmail
+     writing, so watching for it is the main mechanism. Debounced, because Gmail
      mutates its DOM constantly. */
   let timer = 0;
   const schedule = () => {
@@ -161,7 +54,7 @@
   }).observe(document.body, { childList: true, subtree: true });
 
   /* Belt and braces for a signature injected at send time. Capture fires before
-     every other click handler, the document-level bubble listener fires after
+     every other click handler and the document-level bubble listener fires after
      them, so one of the two lands on the right side of Mailsuite's own handler
      whichever order they run in. */
   const SEND_LABEL = /^(send|enviar|envoyer|senden|invia|wyślij|送信|보내기|傳送|发送|भेजें)\b/i;