email_data 1770511189.0.0 → 1771115673.0.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: be394107e0b3e494dd5f5f0d7decf2eb98f5f395e1ad52987c3e41c695af243b
-  data.tar.gz: 9db3dbcbc89537ef5e3b157af8446adb318fcc3ad3026dbad8245841a6abd478
+  metadata.gz: 2779ed93e4463ec6a2768c0b70f3859cfd8efdcfa522e9eb9ace0384640529e2
+  data.tar.gz: e61a7582023106b090c89e0b2e2807e7eb44e7ac41d468df954a54b621f1c941
 SHA512:
-  metadata.gz: d3adf44778124266c0f88881d78e8771c5f96472d2c0d818ebd39416c19bde579293d5179c606c1d04e3dc07e6f1f37b8ff6d8784bdefd6ec3a1b71a31c5caf8
-  data.tar.gz: 9312405e054179d66cfacc7e627df8138988476dc014ddd3859b6a77ac09ac37d0cad43dc146483fe3f242a979da3dcadd3c05b5620a23d6808edcbe8664db0f
+  metadata.gz: 3b2524c334539c45e8cfe23c7f813f5747a953d9c966a03e9ab84d98d764ead0ebcba36ba032b9ffb8ac04a3a8276502f0a604cda4fe6f7702c828fc0b103472
+  data.tar.gz: cbb93a1ea1d6093f8cecf8bf8165ee83fd0343ea902269fcea993795524fa224b4a46ff911745b196f267cbb6c3828a1f03b292d6bf852a92b2e2af432a4cd45

data/.github/workflows/ruby-tests.yml

@@ -23,7 +23,7 @@ jobs:
     strategy:
       fail-fast: false
       matrix:
-        ruby: ["3.1", "3.2", "3.3"]
+        ruby: ["3.1", "3.2", "3.3", "3.4", "4.0"]
         gemfile:
           - Gemfile
     services:
@@ -34,9 +34,7 @@ jobs:
           POSTGRES_HOST: localhost
           POSTGRES_PORT: 5432
           POSTGRES_HOST_AUTH_METHOD: trust
-        options:
-          --health-cmd pg_isready --health-interval 10s --health-timeout 5s
-          --health-retries 5
+        options: --health-cmd pg_isready --health-interval 10s --health-timeout 5s --health-retries 5
 
     steps:
       - uses: actions/checkout@v6
@@ -45,9 +43,7 @@ jobs:
         with:
           path: vendor/bundle
           key: >
-            ${{ runner.os }}-${{ matrix.ruby }}-gems-${{
-            hashFiles(matrix.gemfile) }}
-
+            ${{ runner.os }}-${{ matrix.ruby }}-gems-${{ hashFiles(matrix.gemfile) }}
       - name: Install PostgreSQL client
         run: |
           sudo apt -y install libpq-dev

data/bin/helpers.rb

@@ -16,6 +16,7 @@ require "timeout"
 require "active_support/all"
 require "root_domain"
 require "email_data"
+require "base64"
 
 EXCEPTIONS = File.read("#{__dir__}/../data/manual/exceptions.txt").lines.map(&:chomp)
 

data/bin/sync-disposable-emails

@@ -3,25 +3,23 @@
 
 require_relative "helpers"
 
-def temp_mail_address
-  path = "disposable/tempmailaddress.txt"
-  url = "https://www.tempmailaddress.com/index/index"
+def tempmail_io
+  path = "disposable/tempmail_io.txt"
+  url = "https://api.internal.temp-mail.io/api/v2/domains"
 
   refresh_list(url: url, path: path) do |response|
-    data = JSON.parse(
-      response.body.gsub(/[^-,:\w@.{}"]/, ""),
-      symbolize_names: true
-    )
-    [data[:email].split("@").last]
+    JSON.parse(response.data)["domains"]
   end
 end
 
-def tempmail_io
-  path = "disposable/tempmail_io.txt"
-  url = "https://api.internal.temp-mail.io/api/v2/domains"
+def load_10minute_mail
+  path = "disposable/10minutemail.txt"
+  url = "https://10minutemail.net/history.html"
 
   refresh_list(url: url, path: path) do |response|
-    JSON.parse(response.data)["domains"]
+    JSON.parse(
+      Base64.decode64(response.body[/var domain_json_base64="(.*?)";/, 1])
+    )
   end
 end
 
@@ -115,7 +113,7 @@ def process_scraping(element, value_selector)
   when /^attr\((.*?)\)/
     value = element[Regexp.last_match(1)]
   else
-    element.attributes.each do |_name, attr|
+    element.attributes.each_value do |attr|
       attr = attr.value.to_s
       value = attr if attr&.match?(host_regex)
     end
@@ -139,20 +137,14 @@ threads << thread { load_github_url("https://raw.githubusercontent.com/jespernis
 threads << thread { load_github_url("https://raw.githubusercontent.com/wesbos/burner-email-providers/master/emails.txt") }
 threads << thread { load_github_url("https://gist.github.com/fnando/dafe542cac13f831bbf5521a55248116/raw/disposable.txt") }
 threads << thread { load_github_url("https://raw.githubusercontent.com/bitbar/disposable-email-domains/master/domains.txt") }
-threads << thread { temp_mail_address }
 threads << thread { tempmail_io }
 threads << thread { moakt }
 threads << thread { load_file("disposable/disposable_manually_added.txt") }
-
-10.times do
-  threads << thread { domain_scraping("10minutemail", "https://10minutemail.net/?lang=en", "#fe_text::attr(value)") }
-end
-
+threads << thread { load_10minute_mail }
 threads << thread { domain_scraping("forwardemail", "https://forwardemail.net/en/disposable-addresses", "code::text()") }
 threads << thread { domain_scraping("guerrillamail", "https://www.guerrillamail.com/", "select option::attr(value)") }
-threads << thread { domain_scraping("tempr", "https://tempr.email/", "select[name=DomainId] option::text()") }
-threads << thread { domain_scraping("fake_email_generator", "https://fakemailgenerator.net", "[data-mailhost]::attr(data-mailhost)") }
-threads << thread { domain_scraping("1secmail", "https://www.1secmail.com/", "select[id=domain] option::attr(value)") }
+threads << thread { domain_scraping("tempr", "https://trashmailr.com", "#login-domains option::attr(value)") }
+threads << thread { domain_scraping("fake_email_generator", "https://www.temporary-mail.net", ".dropdown-menu li b:last-of-type::text()") }
 threads << thread { domain_scraping("emailfake", "https://generator.email", ".tt-suggestion p::text()") }
 threads << thread { domain_scraping("emailfake", "https://emailfake.com/", ".tt-suggestion p::text()") }
 threads << thread { domain_scraping("emailfake", "https://email-fake.com/", ".tt-suggestion p::text()") }
@@ -186,11 +178,11 @@ end
 
 ignore_domains = normalize_list(File.read("#{__dir__}/../data/free_email_domains.txt").lines)
                  .map do |domain|
-  RootDomain.call(domain)
+                   RootDomain.call(domain)
 end
 ignore_domains += normalize_list(File.read("#{__dir__}/../data/private_relays.txt").lines)
                   .map do |line|
-  RootDomain.call(line)
+                    RootDomain.call(line)
 end
 
 puts "=> Normalize domains (count: #{domains.size})"