From 23a08ab59563e243c9f31578628e5070e2d14ac4 Mon Sep 17 00:00:00 2001
From: Connor Shea <2977353+connorshea@users.noreply.github.com>
Date: Thu, 2 Jul 2026 15:50:12 -0600
Subject: [PATCH] Optimize Internet username, email, password, and base64

- username: pick the username shape (first-name-only vs first+last)
  before generating names, so the discarded variant is never built.
- password: reuse frozen digit/special-char tables and grow the
  character bag in place with concat instead of += (which copies the
  whole array each time).
- base64: reuse frozen urlsafe/standard character tables and append
  into a pre-sized String instead of building an intermediate array.
- sanitize_email_local_part: replace invalid characters with a single
  gsub against a character-class regex instead of rebuilding the
  allowed-character array and scanning it per character.

Benchmark (Ruby 3.4.9, arm64-darwin25, benchmark-ips):

    require 'benchmark/ips'
    require 'faker'

    Benchmark.ips do |x|
      x.config(warmup: 1, time: 2)
      x.report('username') { Faker::Internet.username }
      x.report('email') { Faker::Internet.email }
      x.report('password') { Faker::Internet.password }
      x.report('base64') { Faker::Internet.base64 }
    end

Results:

    main:        username   41.048k (+/-14.7%) i/s
                 email      21.683k (+/- 0.7%) i/s
                 password  394.792k (+/- 4.8%) i/s
                 base64    183.509k (+/- 2.4%) i/s
    this commit: username   79.313k (+/-18.2%) i/s  (~1.9x)
                 email      37.246k (+/- 1.3%) i/s  (~1.7x)
                 password  451.683k (+/- 1.0%) i/s  (~1.14x)
                 base64    462.621k (+/- 0.5%) i/s  (~2.5x)
---
 lib/faker/internet.rb | 67 ++++++++++++++++++++++++-------------------
 1 file changed, 37 insertions(+), 30 deletions(-)

diff --git a/lib/faker/internet.rb b/lib/faker/internet.rb
index 429702622d..f4f1cc54fb 100644
--- a/lib/faker/internet.rb
+++ b/lib/faker/internet.rb
@@ -18,6 +18,23 @@ class Internet < Base
       [198..198, 18..19,   0..255, 1..255]  # 198.18.0.0/15  - Used for benchmark testing of inter-network communications between subnets
     ].each(&:freeze).freeze
 
+    # Characters that are not allowed in the local part of an email address,
+    # i.e. anything other than alphanumerics and !#$%&'*+-/=?^_`{|}~.
+    # @private
+    INVALID_EMAIL_LOCAL_PART_CHARS = %r{[^0-9A-Za-z!#$%&'*+\-/=?^_`{|}~.]}
+
+    # Character sets used by #base64
+    # @private
+    BASE64_URLSAFE_CHARS = [*'0'..'9', *'A'..'Z', *'a'..'z', '-', '_'].freeze
+    # @private
+    BASE64_CHARS = [*'0'..'9', *'A'..'Z', *'a'..'z', '+', '/'].freeze
+
+    # Character sets used by #password
+    # @private
+    PASSWORD_DIGITS = Array('0'..'9').freeze
+    # @private
+    PASSWORD_SPECIAL_CHARS = %w[! @ # $ % ^ & *].freeze
+
     class << self
       ##
       # Returns the email address
@@ -96,12 +113,15 @@ def username(specifier: nil, separators: %w[. _])
             return result[0...specifier.max]
           end
 
-          sample([
-                   Char.prepare(Faker::Name.first_name),
-                   [Faker::Name.first_name, Faker::Name.last_name].map do |name|
-                     Char.prepare(name)
-                   end.join(sample(separators))
-                 ])
+          # Pick the shape of the username first so we only generate the
+          # name parts we actually need.
+          if rand(2).zero?
+            Char.prepare(Faker::Name.first_name)
+          else
+            [Faker::Name.first_name, Faker::Name.last_name].map do |name|
+              Char.prepare(name)
+            end.join(sample(separators))
+          end
         end
       end
 
@@ -154,22 +174,22 @@ def password(min_length: 8, max_length: 16, mix_case: true, special_characters:
         # use lower_chars by default and add upper_chars if mix_case
         lower_chars = self::LLetters
         password << sample(lower_chars)
-        character_bag += lower_chars
+        character_bag.concat(lower_chars)
 
-        digits = ('0'..'9').to_a
+        digits = PASSWORD_DIGITS
         password << sample(digits)
-        character_bag += digits
+        character_bag.concat(digits)
 
         if mix_case
           upper_chars = self::ULetters
           password << sample(upper_chars)
-          character_bag += upper_chars
+          character_bag.concat(upper_chars)
         end
 
         if special_characters
-          special_chars = %w[! @ # $ % ^ & *]
+          special_chars = PASSWORD_SPECIAL_CHARS
           password << sample(special_chars)
-          character_bag += special_chars
+          character_bag.concat(special_chars)
         end
 
         password << sample(character_bag) while password.length < target_length
@@ -528,14 +548,10 @@ def uuid
       #
       # @faker.version 2.11.0
       def base64(length: 16, padding: false, urlsafe: true)
-        char_range = [
-          Array('0'..'9'),
-          Array('A'..'Z'),
-          Array('a'..'z'),
-          urlsafe ? %w[- _] : %w[+ /]
-        ].flatten
-        s = Array.new(length) { sample(char_range) }.join
-        s += '=' if padding
+        char_range = urlsafe ? BASE64_URLSAFE_CHARS : BASE64_CHARS
+        s = ::String.new('', capacity: length + 1)
+        length.times { s << sample(char_range) }
+        s << '=' if padding
         s
       end
 
@@ -562,16 +578,7 @@ def user(*args)
       private
 
       def sanitize_email_local_part(local_part)
-        char_range = [
-          Array('0'..'9'),
-          Array('A'..'Z'),
-          Array('a'..'z'),
-          "!#$%&'*+-/=?^_`{|}~.".chars
-        ].flatten
-
-        local_part.chars.map do |char|
-          char_range.include?(char) ? char : '#'
-        end.join
+        local_part.gsub(INVALID_EMAIL_LOCAL_PART_CHARS, '#')
       end
 
       def construct_email(local_part, domain_name)