Completely refactored Content implementation.

Now uses POROs and simplifies implementation of providers.

Fixes #143

Author: crmne

lib/ruby_llm.rb

@@ -17,7 +17,8 @@
   'api' => 'API',
   'deepseek' => 'DeepSeek',
   'bedrock' => 'Bedrock',
-  'openrouter' => 'OpenRouter'
+  'openrouter' => 'OpenRouter',
+  'pdf' => 'PDF'
 )
 loader.ignore("#{__dir__}/tasks")
 loader.ignore("#{__dir__}/ruby_llm/railtie")

lib/ruby_llm/attachments.rb

@@ -0,0 +1,48 @@
+# frozen_string_literal: true
+
+module RubyLLM
+  # A module for handling attachments and file operations in RubyLLM.
+  module Attachments
+    # Base class for attachments
+    class Base
+      attr_reader :source
+
+      def initialize(source)
+        @source = source
+      end
+
+      def url?
+        @source.is_a?(String) && @source.match?(%r{^https?://})
+      end
+
+      def file?
+        @source.is_a?(String) && !url?
+      end
+
+      def content
+        @content ||= load_content if file?
+        @content ||= fetch_content if url?
+        @content
+      end
+
+      def type
+        self.class.name.demodulize.downcase
+      end
+
+      def encoded
+        Base64.strict_encode64(content)
+      end
+
+      private
+
+      def fetch_content
+        RubyLLM.logger.debug("Fetching content from URL: #{@source}")
+        Faraday.get(@source).body if url?
+      end
+
+      def load_content
+        File.read(File.expand_path(@source)) if file?
+      end
+    end
+  end
+end

lib/ruby_llm/attachments/audio.rb

@@ -0,0 +1,12 @@
+# frozen_string_literal: true
+
+module RubyLLM
+  module Attachments
+    # Represents an audio attachment
+    class Audio < Base
+      def format
+        File.extname(@source).downcase.delete('.') || 'wav'
+      end
+    end
+  end
+end

lib/ruby_llm/attachments/image.rb

@@ -0,0 +1,13 @@
+# frozen_string_literal: true
+
+module RubyLLM
+  module Attachments
+    # Represents an audio attachment
+    class Image < Base
+      def mime_type
+        ext = File.extname(@source).downcase.delete('.')
+        "image/#{ext}"
+      end
+    end
+  end
+end

lib/ruby_llm/attachments/pdf.rb

@@ -0,0 +1,12 @@
+# frozen_string_literal: true
+
+module RubyLLM
+  module Attachments
+    # Represents a PDF attachment
+    class PDF < Base
+      def mime_type
+        'application/pdf'
+      end
+    end
+  end
+end

lib/ruby_llm/content.rb

@@ -5,100 +5,58 @@ module RubyLLM
   # Stores data in a standard internal format, letting providers
   # handle their own formatting needs.
   class Content
-    def initialize(text = nil, attachments = {}) # rubocop:disable Metrics/AbcSize,Metrics/MethodLength
-      @parts = []
-      @parts << { type: 'text', text: text } unless text.nil? || text.empty?
+    attr_reader :text, :attachments
 
-      Array(attachments[:image]).each do |source|
-        @parts << attach_image(source)
-      end
-
-      Array(attachments[:audio]).each do |source|
-        @parts << attach_audio(source)
-      end
+    def initialize(text = nil, attachments = {})
+      @text = text
+      @attachments = []
 
-      Array(attachments[:pdf]).each do |source|
-        @parts << attach_pdf(source)
-      end
+      process_attachments(attachments)
+      raise ArgumentError, 'Text and attachments cannot be both nil' if @text.nil? && @attachments.empty?
     end
 
-    def to_a
-      return if @parts.empty?
-
-      @parts
+    def add_image(source)
+      @attachments << Attachments::Image.new(source)
+      self
     end
 
-    def format
-      return @parts.first[:text] if @parts.size == 1 && @parts.first[:type] == 'text'
-
-      to_a
+    def add_audio(source)
+      @attachments << Attachments::Audio.new(source)
+      self
     end
 
-    private
-
-    def attach_image(source) # rubocop:disable Metrics/MethodLength
-      source = File.expand_path(source) unless source.start_with?('http')
-
-      return { type: 'image', source: { url: source } } if source.start_with?('http')
-
-      data = Base64.strict_encode64(File.read(source))
-      mime_type = mime_type_for(source)
-
-      {
-        type: 'image',
-        source: {
-          type: 'base64',
-          media_type: mime_type,
-          data: data
-        }
-      }
+    def add_pdf(source)
+      @attachments << Attachments::PDF.new(source)
+      self
     end
 
-    def attach_audio(source)
-      source = File.expand_path(source) unless source.start_with?('http')
-      data = encode_file(source)
-      format = File.extname(source).delete('.') || 'wav'
-
-      {
-        type: 'input_audio',
-        input_audio: {
-          data: data,
-          format: format
-        }
-      }
-    end
-
-    def attach_pdf(source)
-      source = File.expand_path(source) unless source.start_with?('http')
-
-      pdf_data = {
-        type: 'pdf',
-        source: source
-      }
-
-      # For local files, validate they exist
-      unless source.start_with?('http')
-        raise Error, "PDF file not found: #{source}" unless File.exist?(source)
-
-        # Preload file content for providers that need it
-        pdf_data[:content] = File.read(source)
+    def format
+      if @text && @attachments.empty?
+        @text
+      else
+        self
       end
-
-      pdf_data
     end
 
-    def encode_file(source)
-      if source.start_with?('http')
-        response = Faraday.get(source)
-        Base64.strict_encode64(response.body)
-      else
-        Base64.strict_encode64(File.read(source))
+    # For Rails serialization
+    def as_json
+      hash = { text: @text }
+      unless @attachments.empty?
+        hash[:attachments] = @attachments.map do |a|
+          { type: a.type, source: a.source }
+        end
       end
+      hash
     end
 
-    def mime_type_for(path)
-      ext = File.extname(path).delete('.')
-      "image/#{ext}"
+    private
+
+    def process_attachments(attachments)
+      return unless attachments.is_a?(Hash)
+
+      Array(attachments[:image]).each { |source| add_image(source) }
+      Array(attachments[:audio]).each { |source| add_audio(source) }
+      Array(attachments[:pdf]).each { |source| add_pdf(source) }
     end
   end
 end

lib/ruby_llm/message.rb

@@ -7,7 +7,7 @@ module RubyLLM
   class Message
     ROLES = %i[system user assistant tool].freeze
 
-    attr_reader :role, :content, :tool_calls, :tool_call_id, :input_tokens, :output_tokens, :model_id
+    attr_reader :role, :tool_calls, :tool_call_id, :input_tokens, :output_tokens, :model_id
 
     def initialize(options = {})
       @role = options[:role].to_sym
@@ -21,6 +21,14 @@ def initialize(options = {})
       ensure_valid_role
     end
 
+    def content
+      if @content.is_a?(Content) && @content.text && @content.attachments.empty?
+        @content.text
+      else
+        @content
+      end
+    end
+
     def tool_call?
       !tool_calls.nil? && !tool_calls.empty?
     end
@@ -49,9 +57,9 @@ def to_h
 
     def normalize_content(content)
       case content
-      when Content then content.format
-      when String then Content.new(content).format
-      else content
+      when String then Content.new(content)
+      when Hash then Content.new(content[:text], content)
+      else content # Pass through nil, Content, or other types
       end
     end