ruby_llm 1.0.0 → 1.1.0rc1

data/lib/tasks/models.rake

@@ -1,212 +1,43 @@
 # frozen_string_literal: true
 
-require 'English'
-require 'faraday'
-require 'nokogiri'
-
-# URLs to process
-PROVIDER_DOCS = {
-  openai: {
-    models: 'https://platform.openai.com/docs/models',
-    pricing: 'https://platform.openai.com/docs/pricing'
-  },
-  gemini: {
-    models: 'https://ai.google.dev/gemini-api/docs/models/gemini',
-    pricing: 'https://ai.google.dev/gemini-api/docs/pricing'
-  },
-  deepseek: {
-    models: 'https://api-docs.deepseek.com/quick_start/pricing/'
-  },
-  anthropic: {
-    models: 'https://docs.anthropic.com/en/docs/about-claude/models/all-models'
-  }
-}.freeze
-
-def fetch_page(url) # rubocop:disable Metrics/AbcSize,Metrics/MethodLength
-  if url.include?('openai.com')
-    puts "Please visit #{url} and paste the content below (type 'END' on a new line when done):"
-    original_separator = $INPUT_RECORD_SEPARATOR
-    $/ = 'END'
-    content = $stdin.gets&.chomp
-    $/ = original_separator
-
-    raise "No content provided for #{url}" unless content
-
-    content
-  else
-    response = http_client.get(url)
-    html = Nokogiri::HTML(response.body)
-
-    # Remove script tags and comments
-    html.css('script').remove
-    html.xpath('//comment()').remove
-
-    # Extract text content
-    text = html.css('body').text
-
-    # Clean up whitespace
-    text.gsub!(/\s+/, ' ')
-    text.strip!
-
-    text
-  end
-rescue StandardError => e
-  raise "Failed to fetch #{url}: #{e.message}"
-end
-
-def http_client
-  @http_client ||= Faraday.new do |f|
-    f.response :raise_error
-    f.response :logger, RubyLLM.logger, { headers: false, bodies: true }
-  end
-end
+require_relative 'model_updater'
+require_relative 'capability_scraper'
+require_relative 'capability_generator'
 
 namespace :models do # rubocop:disable Metrics/BlockLength
-  desc 'Update available models from providers'
+  desc 'Update available models from providers (API keys needed)'
   task :update do
-    require 'ruby_llm'
-
-    # Configure API keys
-    RubyLLM.configure do |config|
-      config.openai_api_key = ENV.fetch('OPENAI_API_KEY')
-      config.anthropic_api_key = ENV.fetch('ANTHROPIC_API_KEY')
-      config.gemini_api_key = ENV.fetch('GEMINI_API_KEY')
-      config.deepseek_api_key = ENV.fetch('DEEPSEEK_API_KEY')
-    end
-
-    # Refresh models (now returns self instead of models array)
-    models = RubyLLM.models.refresh!.all
-    # Write to models.json
-    File.write(File.expand_path('../ruby_llm/models.json', __dir__), JSON.pretty_generate(models.map(&:to_h)))
-
-    puts "Updated models.json with #{models.size} models:"
-    RubyLLM::Provider.providers.each do |provider_sym, provider_module|
-      provider_name = provider_module.to_s.split('::').last
-      provider_models = models.select { |m| m.provider == provider_sym.to_s }
-      puts "#{provider_name} models: #{provider_models.size}"
-    end
+    ModelUpdater.new.run
   end
 
-  desc 'Update model capabilities modules by scraping provider documentation (use PROVIDER=name to update only one)'
-  task :update_capabilities do # rubocop:disable Metrics/BlockLength
-    # Check if a specific provider was requested
-    target_provider = ENV['PROVIDER']&.to_sym
-    require 'ruby_llm'
-    require 'fileutils'
-
-    # Configure API keys
-    RubyLLM.configure do |config|
-      config.openai_api_key = ENV.fetch('OPENAI_API_KEY')
-      config.anthropic_api_key = ENV.fetch('ANTHROPIC_API_KEY')
-      config.gemini_api_key = ENV.fetch('GEMINI_API_KEY')
+  desc 'Update capabilities modules (GEMINI_API_KEY needed)'
+  task :update_capabilities, [:providers] do |_t, args|
+    gemini_key = ENV.fetch('GEMINI_API_KEY', nil)
+    unless gemini_key && !gemini_key.empty?
+      puts 'Error: GEMINI_API_KEY required'
+      exit(1)
     end
 
-    # Filter providers if a specific one was requested
-    providers_to_process = if target_provider && PROVIDER_DOCS.key?(target_provider)
-                             { target_provider => PROVIDER_DOCS[target_provider] }
-                           else
-                             PROVIDER_DOCS
-                           end
-
-    # Process each provider
-    providers_to_process.each do |provider, urls| # rubocop:disable Metrics/BlockLength
-      puts "Processing #{provider}..."
-
-      # Initialize our AI assistants
-      #
-      gemini = RubyLLM.chat(model: 'gemini-2.0-flash').with_temperature(0)
-      claude = RubyLLM.chat(model: 'claude-3-7-sonnet-20250219').with_temperature(0)
-
-      # Read existing capabilities file if present
-      existing_file = "lib/ruby_llm/providers/#{provider}/capabilities.rb"
-      existing_code = File.read(existing_file) if File.exist?(existing_file)
-
-      begin
-        # Download documentation
-        docs = urls.map do |type, url|
-          puts "  Getting #{type} documentation..."
-          content = fetch_page(url)
-
-          puts "\nHere's what I got:\n\n"
-          puts "#{content.slice(0, 500)}...\n\n"
-
-          loop do
-            print 'Does this content look correct? (y/n): '
-            answer = $stdin.gets&.chomp&.downcase
-            break if answer == 'y'
-            raise "Content verification failed for #{url}" if answer == 'n'
-          end
-
-          "#{type.to_s.upcase} DOCUMENTATION:\n\n#{content}"
-        end.join("\n\n")
-
-        # Extract relevant information with Gemini
-        puts '  Extracting model information...'
-        extraction_prompt = <<~PROMPT
-          Extract relevant model capabilities information from this documentation:
-
-          #{docs}
-
-          Focus on:
-          1. Available models and their IDs
-          2. Context window sizes
-          3. Maximum output tokens
-          4. Pricing information
-          5. Model capabilities (vision, function calling, etc)
-
-          Format the information clearly and concisely, focusing only on facts about the models.
-        PROMPT
-
-        model_info = gemini.ask(extraction_prompt).content
-
-        # Generate Ruby code with Claude
-        puts '  Generating Ruby code...'
-        code_prompt = <<~PROMPT
-          I need you to generate a Ruby module for #{provider}'s model capabilities.
-          Use this extracted model information:
-
-          #{model_info}
-
-          The module should go in lib/ruby_llm/providers/#{provider}/capabilities.rb and follow these conventions:
-
-          1. Include methods for determining context windows, token limits, pricing, and capabilities
-          2. Use consistent naming with other providers
-          3. Include detailed pricing information in a PRICES constant
-          4. Follow the existing structure in the codebase
-          5. Use Ruby idioms and clean code practices
-          6. Include module_function to make methods callable at module level
-          7. Include all necessary method documentation
-
-          Here's the existing implementation for reference (maintain similar structure and same method names):
-
-          #{existing_code}
-
-          Focus on accuracy and maintaining consistency with the existing codebase structure.
-        PROMPT
-
-        response = claude.ask(code_prompt)
-
-        # Extract Ruby code from Claude's response
-        puts "  Extracting Ruby code from Claude's response..."
-        ruby_code = nil
-
-        # Look for Ruby code block
-        ruby_code = Regexp.last_match(1).strip if response.content =~ /```ruby\s*(.*?)```/m
-
-        # Verify we found Ruby code
-        raise "No Ruby code block found in Claude's response" if ruby_code.nil? || ruby_code.empty?
+    RubyLLM.configure do |c|
+      c.gemini_api_key = gemini_key
+      c.request_timeout = 300
+    end
 
-        # Save the file
-        file_path = "lib/ruby_llm/providers/#{provider}/capabilities.rb"
-        puts "  Writing #{file_path}..."
+    target_providers = CapabilityScraper.parse_providers(args[:providers])
 
-        FileUtils.mkdir_p(File.dirname(file_path))
-        File.write(file_path, ruby_code)
-      rescue StandardError => e
-        raise "Failed to process #{provider}: #{e.message}"
+    begin
+      scraper = CapabilityScraper.new(target_providers)
+      scraper.run do |provider, docs_html|
+        generator = CapabilityGenerator.new(provider, docs_html)
+        generator.generate_and_save
       end
+    rescue StandardError => e
+      puts "Error: #{e.message}"
+      puts e.backtrace.first(5).join("\n")
+    ensure
+      puts 'Update process complete. Review generated files.'
     end
-
-    puts "Done! Don't forget to review the generated code and run the tests."
   end
 end
+
+task default: ['models:update']

data/lib/tasks/vcr.rake

@@ -0,0 +1,97 @@
+# frozen_string_literal: true
+
+# Helper functions at the top level
+def record_all_cassettes(cassette_dir)
+  # Re-record all cassettes
+  FileUtils.rm_rf(cassette_dir)
+  FileUtils.mkdir_p(cassette_dir)
+
+  puts 'Recording cassettes for all providers...'
+  run_tests
+  puts 'Done recording. Please review the new cassettes.'
+end
+
+def record_for_providers(providers, cassette_dir) # rubocop:disable Metrics/AbcSize,Metrics/MethodLength
+  # Get the list of available providers from RubyLLM itself
+  all_providers = RubyLLM::Provider.providers.keys.map(&:to_s)
+
+  # Check for valid providers
+  if providers.empty?
+    puts "Please specify providers or 'all'. Example: rake vcr:record[openai,anthropic]"
+    puts "Available providers: #{all_providers.join(', ')}"
+    return
+  end
+
+  invalid_providers = providers - all_providers
+  if invalid_providers.any?
+    puts "Invalid providers: #{invalid_providers.join(', ')}"
+    puts "Available providers: #{all_providers.join(', ')}"
+    return
+  end
+
+  # Find and delete matching cassettes
+  cassettes_to_delete = find_matching_cassettes(cassette_dir, providers)
+
+  if cassettes_to_delete.empty?
+    puts 'No cassettes found for the specified providers.'
+    puts 'Running tests to record new cassettes...'
+  else
+    delete_cassettes(cassettes_to_delete)
+    puts "\nRunning tests to record new cassettes..."
+  end
+
+  run_tests
+
+  puts "\nDone recording cassettes for #{providers.join(', ')}."
+  puts 'Please review the updated cassettes for sensitive information.'
+end
+
+def find_matching_cassettes(dir, providers) # rubocop:disable Metrics/MethodLength
+  cassettes = []
+
+  Dir.glob("#{dir}/**/*.yml").each do |file|
+    basename = File.basename(file)
+
+    # Precise matching to avoid cross-provider confusion
+    providers.each do |provider|
+      # Match only exact provider prefixes
+      next unless basename =~ /^[^_]*_#{provider}_/ || # For first section like "chat_openai_"
+                  basename =~ /_#{provider}_[^_]+_/    # For middle sections like "_openai_gpt4_"
+
+      cassettes << file
+      break
+    end
+  end
+
+  cassettes
+end
+
+def delete_cassettes(cassettes)
+  puts "Deleting #{cassettes.size} cassettes for re-recording:"
+  cassettes.each do |file|
+    puts "  - #{File.basename(file)}"
+    File.delete(file)
+  end
+end
+
+def run_tests
+  system('bundle exec rspec') || abort('Tests failed')
+end
+
+namespace :vcr do
+  desc 'Record VCR cassettes (rake vcr:record[all] or vcr:record[openai,anthropic])'
+  task :record, [:providers] do |_, args|
+    require 'fileutils'
+    require 'ruby_llm'
+
+    providers = (args[:providers] || '').downcase.split(',')
+    cassette_dir = 'spec/fixtures/vcr_cassettes'
+    FileUtils.mkdir_p(cassette_dir)
+
+    if providers.include?('all')
+      record_all_cassettes(cassette_dir)
+    else
+      record_for_providers(providers, cassette_dir)
+    end
+  end
+end

metadata

@@ -1,15 +1,29 @@
 --- !ruby/object:Gem::Specification
 name: ruby_llm
 version: !ruby/object:Gem::Version
-  version: 1.0.0
+  version: 1.1.0rc1
 platform: ruby
 authors:
 - Carmine Paolino
 autorequire:
-bindir: exe
+bindir: bin
 cert_chain: []
-date: 2025-03-11 00:00:00.000000000 Z
+date: 2025-04-02 00:00:00.000000000 Z
 dependencies:
+- !ruby/object:Gem::Dependency
+  name: base64
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
 - !ruby/object:Gem::Dependency
   name: event_stream_parser
   requirement: !ruby/object:Gem::Requirement
@@ -82,31 +96,21 @@ dependencies:
         version: '2'
 description: A delightful Ruby way to work with AI. Chat in text, analyze and generate
   images, understand audio, and use tools through a unified interface to OpenAI, Anthropic,
-  Google, and DeepSeek. Built for developer happiness with automatic token counting,
-  proper streaming, and Rails integration. No wrapping your head around multiple APIs
-  - just clean Ruby code that works.
+  Google, AWS Bedrock Anthropic, and DeepSeek. Built for developer happiness with
+  automatic token counting, proper streaming, and Rails integration. No wrapping your
+  head around multiple APIs - just clean Ruby code that works.
 email:
 - carmine@paolino.me
 executables: []
 extensions: []
 extra_rdoc_files: []
 files:
-- ".github/workflows/cicd.yml"
-- ".github/workflows/docs.yml"
-- ".gitignore"
-- ".overcommit.yml"
-- ".rspec"
-- ".rspec_status"
-- ".rubocop.yml"
-- ".yardopts"
-- Gemfile
 - LICENSE
 - README.md
-- Rakefile
-- bin/console
-- bin/setup
 - lib/ruby_llm.rb
 - lib/ruby_llm/active_record/acts_as.rb
+- lib/ruby_llm/aliases.json
+- lib/ruby_llm/aliases.rb
 - lib/ruby_llm/chat.rb
 - lib/ruby_llm/chunk.rb
 - lib/ruby_llm/configuration.rb
@@ -127,8 +131,20 @@ files:
 - lib/ruby_llm/providers/anthropic/models.rb
 - lib/ruby_llm/providers/anthropic/streaming.rb
 - lib/ruby_llm/providers/anthropic/tools.rb
+- lib/ruby_llm/providers/bedrock.rb
+- lib/ruby_llm/providers/bedrock/capabilities.rb
+- lib/ruby_llm/providers/bedrock/chat.rb
+- lib/ruby_llm/providers/bedrock/models.rb
+- lib/ruby_llm/providers/bedrock/signing.rb
+- lib/ruby_llm/providers/bedrock/streaming.rb
+- lib/ruby_llm/providers/bedrock/streaming/base.rb
+- lib/ruby_llm/providers/bedrock/streaming/content_extraction.rb
+- lib/ruby_llm/providers/bedrock/streaming/message_processing.rb
+- lib/ruby_llm/providers/bedrock/streaming/payload_processing.rb
+- lib/ruby_llm/providers/bedrock/streaming/prelude_handling.rb
 - lib/ruby_llm/providers/deepseek.rb
 - lib/ruby_llm/providers/deepseek/capabilities.rb
+- lib/ruby_llm/providers/deepseek/chat.rb
 - lib/ruby_llm/providers/gemini.rb
 - lib/ruby_llm/providers/gemini/capabilities.rb
 - lib/ruby_llm/providers/gemini/chat.rb
@@ -149,11 +165,18 @@ files:
 - lib/ruby_llm/providers/openai/tools.rb
 - lib/ruby_llm/railtie.rb
 - lib/ruby_llm/stream_accumulator.rb
+- lib/ruby_llm/streaming.rb
 - lib/ruby_llm/tool.rb
 - lib/ruby_llm/tool_call.rb
 - lib/ruby_llm/version.rb
+- lib/tasks/browser_helper.rb
+- lib/tasks/capability_generator.rb
+- lib/tasks/capability_scraper.rb
+- lib/tasks/cli_helper.rb
+- lib/tasks/code_validator.rb
+- lib/tasks/model_updater.rb
 - lib/tasks/models.rake
-- ruby_llm.gemspec
+- lib/tasks/vcr.rake
 homepage: https://rubyllm.com
 licenses:
 - MIT

data/.github/workflows/cicd.yml

@@ -1,109 +0,0 @@
-name: CI
-
-on:
-  push:
-    branches: [ "main" ]
-    paths:
-      - 'lib/**'
-      - 'spec/**'
-      - 'Gemfile'
-      - 'Rakefile'
-      - 'ruby_llm.gemspec'
-  pull_request:
-    branches: [ "main" ]
-  workflow_call:
-
-jobs:
-  test:
-    runs-on: ubuntu-latest
-    strategy:
-      matrix:
-        ruby-version: ['3.1']
-
-    steps:
-    - uses: actions/checkout@v4
-
-    - name: Set up Ruby
-      uses: ruby/setup-ruby@v1
-      with:
-        ruby-version: ${{ matrix.ruby-version }}
-        bundler-cache: true
-
-    - name: Install dependencies
-      run: bundle install
-
-    - name: Check code format
-      run: bundle exec rubocop
-
-    - name: Run tests
-      env:
-        OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
-        ANTHROPIC_API_KEY: ${{ secrets.ANTHROPIC_API_KEY }}
-        GEMINI_API_KEY: ${{ secrets.GEMINI_API_KEY }}
-        DEEPSEEK_API_KEY: ${{ secrets.DEEPSEEK_API_KEY }}
-      run: bundle exec rspec
-
-    - name: Upload coverage to Codecov
-      uses: codecov/codecov-action@v5
-      env:
-        CODECOV_TOKEN: ${{ secrets.CODECOV_TOKEN }}
-      with:
-        files: ./coverage/coverage.xml
-        fail_ci_if_error: false
-
-  publish:
-    name: Build + Publish
-    needs: test
-    if: github.event_name == 'push' && github.ref == 'refs/heads/main'
-    runs-on: ubuntu-latest
-    permissions:
-      contents: read
-      packages: write
-
-    steps:
-    - uses: actions/checkout@v4
-
-    - name: Set up Ruby
-      uses: ruby/setup-ruby@v1
-      with:
-        ruby-version: '3.3'
-        bundler-cache: true
-
-    - name: Publish to GPR
-      run: |
-        mkdir -p $HOME/.gem
-        touch $HOME/.gem/credentials
-        chmod 0600 $HOME/.gem/credentials
-        printf -- "---\n:github: ${GEM_HOST_API_KEY}\n" > $HOME/.gem/credentials
-        gem build *.gemspec
-        output=$(gem push --KEY github --host https://rubygems.pkg.github.com/${OWNER} *.gem 2>&1) || {
-          echo "$output"
-          if echo "$output" | grep -q "already been pushed"; then
-            echo "Version already exists, skipping"
-            exit 0
-          else
-            exit 1
-          fi
-        }
-      env:
-        GEM_HOST_API_KEY: "Bearer ${{secrets.GITHUB_TOKEN}}"
-        OWNER: ${{ github.repository_owner }}
-
-    - name: Publish to RubyGems
-      run: |
-        mkdir -p $HOME/.gem
-        touch $HOME/.gem/credentials
-        chmod 0600 $HOME/.gem/credentials
-        printf -- "---\n:rubygems_api_key: ${GEM_HOST_API_KEY}\n" > $HOME/.gem/credentials
-        gem build *.gemspec
-        output=$(gem push *.gem 2>&1) || {
-          echo "$output"
-          if echo "$output" | grep -q "Repushing of gem versions is not allowed"; then
-            echo "Version already exists, skipping"
-            exit 0
-          else
-            exit 1
-          fi
-        }
-      env:
-        GEM_HOST_API_KEY: "${{secrets.RUBYGEMS_AUTH_TOKEN}}"

data/.github/workflows/docs.yml

@@ -1,53 +0,0 @@
-name: Deploy docs
-
-on:
-  push:
-    branches: [main]
-    paths:
-      - 'docs/**'
-      - '.github/workflows/docs.yml'
-  workflow_dispatch:
-
-permissions:
-  contents: read
-  pages: write
-  id-token: write
-
-concurrency:
-  group: "pages"
-  cancel-in-progress: false
-
-jobs:
-  build:
-    runs-on: ubuntu-latest
-    steps:
-      - name: Checkout
-        uses: actions/checkout@v4
-      - name: Setup Ruby
-        uses: ruby/setup-ruby@v1
-        with:
-          ruby-version: '3.3'
-          bundler-cache: true
-          working-directory: docs
-      - name: Setup Pages
-        uses: actions/configure-pages@v4
-      - name: Build with Jekyll
-        working-directory: docs
-        run: bundle exec jekyll build --baseurl "${{ steps.pages.outputs.base_path }}"
-        env:
-          JEKYLL_ENV: production
-      - name: Upload artifact
-        uses: actions/upload-pages-artifact@v3
-        with:
-          path: docs/_site
-
-  deploy:
-    environment:
-      name: github-pages
-      url: ${{ steps.deployment.outputs.page_url }}
-    runs-on: ubuntu-latest
-    needs: build
-    steps:
-      - name: Deploy to GitHub Pages
-        id: deployment
-        uses: actions/deploy-pages@v4

data/.gitignore

@@ -1,58 +0,0 @@
-*.gem
-*.rbc
-/.config
-/coverage/
-/InstalledFiles
-/pkg/
-/spec/reports/
-/spec/examples.txt
-/test/tmp/
-/test/version_tmp/
-/tmp/
-
-# Used by dotenv library to load environment variables.
-.env
-
-# Ignore Byebug command history file.
-.byebug_history
-
-## Specific to RubyMotion:
-.dat*
-.repl_history
-build/
-*.bridgesupport
-build-iPhoneOS/
-build-iPhoneSimulator/
-
-## Specific to RubyMotion (use of CocoaPods):
-#
-# We recommend against adding the Pods directory to your .gitignore. However
-# you should judge for yourself, the pros and cons are mentioned at:
-# https://guides.cocoapods.org/using/using-cocoapods.html#should-i-check-the-pods-directory-into-source-control
-#
-# vendor/Pods/
-
-## Documentation cache and generated files:
-/.yardoc/
-/_yardoc/
-/doc/
-/rdoc/
-
-## Environment normalization:
-/.bundle/
-/vendor/bundle
-/lib/bundler/man/
-
-# for a library or gem, you might want to ignore these files since the code is
-# intended to run in multiple environments; otherwise, check them in:
-Gemfile.lock
-# .ruby-version
-# .ruby-gemset
-
-# unless supporting rvm < 1.11.0 or doing something fancy, ignore this:
-.rvmrc
-
-# Used by RuboCop. Remote config files pulled in from inherit_from directive.
-# .rubocop-https?--*
-
-repomix-output.txt

data/.overcommit.yml

@@ -1,26 +0,0 @@
-PreCommit:
-  RuboCop:
-    enabled: true
-    auto_correct: true
-    on_warn: fail # Treat all warnings as failures
-
-  Flay:
-    enabled: true
-
-  RSpec:
-    enabled: false
-    command: ['bundle', 'exec', 'rspec']
-    on_warn: fail
-
-  TrailingWhitespace:
-    enabled: true
-    auto_correct: true
-    exclude:
-     - '**/db/structure.sql' # Ignore trailing whitespace in generated files
-
-PostCheckout:
-  ALL: # Special hook name that customizes all hooks of this type
-    quiet: true # Change all post-checkout hooks to only display output on failure
-
-  IndexTags:
-    enabled: true # Generate a tags file with `ctags` each time HEAD changes