Apply +-normalization for all email domains

bin/mongodb/fix-normalized-emails.js

@@ -1,13 +1,23 @@
-function gmailNormalize(email) {
+const gmailOrProton = ['protonmail.com', 'protonmail.ch', 'pm.me', 'gmail.com', 'googlemail.com'];
+
+function normalize(email) {
   let [name, domain] = email.toLowerCase().split('@');
   [name] = name.split('+');
-  return name.replace(/\./g, '') + '@' + domain;
+
+  if (gmailOrProton.includes(domain)) {
+    return name.replace(/\./g, '') + '@' + domain;
+  } else {
+    return name + '@' + domain;
+  }
 }
 
 db.user4
-  .find({ email: /[^+.]+[+.].*@(protonmail\.com|protonmail\.ch|pm\.me|gmail\.com|googlemail\.com)$/i })
+  .find({
+    email:
+      /([^+.]+[+.].*@(protonmail\.com|protonmail\.ch|pm\.me|gmail\.com|googlemail\.com)|^[^+]+\+.*@.+)$/i,
+  })
   .forEach(user => {
-    const normalized = gmailNormalize(user.email);
+    const normalized = normalize(user.email);
     const verbatim = user.verbatimEmail || user.email;
     print(user.username, ': ', verbatim, '->', normalized);
 

modules/core/src/main/email.scala

@@ -20,10 +20,11 @@ object email:
       def normalize = NormalizedEmailAddress: // changing normalization requires database migration!
         val lower = e.toLowerCase
         lower.split('@') match
-          case Array(name, domain) if EmailAddress.gmailLikeNormalizedDomains(domain) =>
-            val normalizedName = name
-              .replace(".", "")    // remove all dots
-              .takeWhile('+' != _) // skip everything after the first '+'
+          case Array(name, domain) =>
+            val skipAfterPlus = name.takeWhile('+' != _)
+            val normalizedName =
+              if EmailAddress.gmailLikeNormalizedDomains(domain) then skipAfterPlus.replace(".", "")
+              else skipAfterPlus
             if normalizedName.isEmpty then lower else s"$normalizedName@$domain"
           case _ => lower
 

modules/core/src/test/EmailTest.scala

@@ -14,6 +14,16 @@ class EmailTest extends munit.FunSuite:
       NormalizedEmailAddress("[email protected]")
     )
 
+  test("normalize other"):
+    assertEquals(
+      EmailAddress("[email protected]").normalize,
+      NormalizedEmailAddress("[email protected]")
+    )
+    assertEquals(
+      EmailAddress("[email protected]").normalize,
+      NormalizedEmailAddress("[email protected]")
+    )
+
   test("not similar emails"):
     assert(!EmailAddress("[email protected]").similarTo(EmailAddress("[email protected]")))
     assert(!EmailAddress("[email protected]").similarTo(EmailAddress("[email protected]")))