dspy 0.12.0 → 0.13.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: d40d9681a87073c6c4f5531025891bf871bee086434ee131e15a4391cba7b893
-  data.tar.gz: 1a1a4ce935e7cdf4ce9c06617f124ed8858d7074a3f6bec4484d6db24dbfa232
+  metadata.gz: d80d5b0166fe5a101e4918ffee13a70dec6ca67b493cf2e68dff1c18b2df36c1
+  data.tar.gz: 1687fe88d41c5d4627592ff5e98f87ca9f40186870386291b1ead091f51235da
 SHA512:
-  metadata.gz: 2cdd3ee7e867344b4e45dee7fd0d73dc875ec2c730a2cf75e5f620a19c535f0aac50cc6c88dfb7beb3e075b147e7467ccd4e55a7724f49f3ac111eb080fd88e0
-  data.tar.gz: f5c5ed5dd3b6aa5401c20e31fbbd9638eb1d8c0f3b7fe5adb2e2b69bbb3b051d0ecd277d6e2f3b4d4896fd19f39a89345c9503d97106a3b8e005b1f55a3e5b2b
+  metadata.gz: 7bedebf2e58243bedcf8003d25b4f55789a7e4a611f9f1997a322e93db553fe117fde155415c5a607a65b49f61f2a640c7899a0fd4793bcdf9e8672602f54755
+  data.tar.gz: cdf39e605a7550e94334c5bad13ccc5af2a255b7039559b56b918e1ba3706e671c52960558459a0c96edab318c68bc9828b1eb03530b47daa896766a2cb7aa7d

data/lib/dspy/instrumentation/token_tracker.rb

@@ -1,5 +1,7 @@
 # frozen_string_literal: true
 
+require_relative '../lm/usage'
+
 module DSPy
   module Instrumentation
     # Utility for extracting token usage from different LM adapters
@@ -9,6 +11,18 @@ module DSPy
 
       # Extract actual token usage from API responses
       def extract_token_usage(response, provider)
+        return {} unless response&.usage
+        
+        # Handle Usage struct
+        if response.usage.is_a?(DSPy::LM::Usage) || response.usage.is_a?(DSPy::LM::OpenAIUsage)
+          return {
+            input_tokens: response.usage.input_tokens,
+            output_tokens: response.usage.output_tokens,
+            total_tokens: response.usage.total_tokens
+          }
+        end
+        
+        # Fallback to legacy hash handling
         case provider.to_s.downcase
         when 'openai'
           extract_openai_tokens(response)
@@ -27,11 +41,12 @@ module DSPy
         usage = response.usage
         return {} unless usage.is_a?(Hash)
         
+        # Handle both symbol and string keys for VCR compatibility
         {
           input_tokens: usage[:prompt_tokens] || usage['prompt_tokens'],
           output_tokens: usage[:completion_tokens] || usage['completion_tokens'],
           total_tokens: usage[:total_tokens] || usage['total_tokens']
-        }
+        }.compact # Remove nil values
       end
 
       def extract_anthropic_tokens(response)
@@ -40,6 +55,7 @@ module DSPy
         usage = response.usage
         return {} unless usage.is_a?(Hash)
         
+        # Handle both symbol and string keys for VCR compatibility
         input_tokens = usage[:input_tokens] || usage['input_tokens'] || 0
         output_tokens = usage[:output_tokens] || usage['output_tokens'] || 0
         
@@ -47,7 +63,7 @@ module DSPy
           input_tokens: input_tokens,
           output_tokens: output_tokens,
           total_tokens: input_tokens + output_tokens
-        }
+        }.compact # Remove nil values
       end
     end
   end

data/lib/dspy/lm/adapters/anthropic_adapter.rb

@@ -63,9 +63,12 @@ module DSPy
             content = response.content.first.text if response.content.is_a?(Array) && response.content.first
             usage = response.usage
 
+            # Convert usage data to typed struct
+            usage_struct = UsageFactory.create('anthropic', usage)
+            
             Response.new(
               content: content,
-              usage: usage.respond_to?(:to_h) ? usage.to_h : usage,
+              usage: usage_struct,
               metadata: {
                 provider: 'anthropic',
                 model: model,

data/lib/dspy/lm/adapters/openai_adapter.rb

@@ -52,9 +52,12 @@ module DSPy
             raise AdapterError, "OpenAI refused to generate output: #{message.refusal}"
           end
 
+          # Convert usage data to typed struct
+          usage_struct = UsageFactory.create('openai', usage)
+          
           Response.new(
             content: content,
-            usage: usage.respond_to?(:to_h) ? usage.to_h : usage,
+            usage: usage_struct,
             metadata: {
               provider: 'openai',
               model: model,

data/lib/dspy/lm/response.rb

@@ -1,5 +1,7 @@
 # frozen_string_literal: true
 
+require_relative 'usage'
+
 module DSPy
   class LM
     # Normalized response format for all LM providers

data/lib/dspy/lm/usage.rb

@@ -0,0 +1,157 @@
+# frozen_string_literal: true
+
+require 'sorbet-runtime'
+
+module DSPy
+  class LM
+    # Base class for token usage information
+    class Usage < T::Struct
+      extend T::Sig
+      
+      const :input_tokens, Integer
+      const :output_tokens, Integer
+      const :total_tokens, Integer
+      
+      sig { returns(Hash) }
+      def to_h
+        {
+          input_tokens: input_tokens,
+          output_tokens: output_tokens,
+          total_tokens: total_tokens
+        }
+      end
+    end
+    
+    # OpenAI-specific usage information with additional fields
+    class OpenAIUsage < T::Struct
+      extend T::Sig
+      
+      const :input_tokens, Integer
+      const :output_tokens, Integer
+      const :total_tokens, Integer
+      const :prompt_tokens_details, T.nilable(T::Hash[Symbol, Integer]), default: nil
+      const :completion_tokens_details, T.nilable(T::Hash[Symbol, Integer]), default: nil
+      
+      sig { returns(Hash) }
+      def to_h
+        base = {
+          input_tokens: input_tokens,
+          output_tokens: output_tokens,
+          total_tokens: total_tokens
+        }
+        base[:prompt_tokens_details] = prompt_tokens_details if prompt_tokens_details
+        base[:completion_tokens_details] = completion_tokens_details if completion_tokens_details
+        base
+      end
+    end
+    
+    # Factory for creating appropriate usage objects
+    module UsageFactory
+      extend T::Sig
+      
+      sig { params(provider: String, usage_data: T.untyped).returns(T.nilable(T.any(Usage, OpenAIUsage))) }
+      def self.create(provider, usage_data)
+        return nil if usage_data.nil?
+        
+        # If already a Usage struct, return as-is
+        return usage_data if usage_data.is_a?(Usage)
+        
+        # Handle test doubles by converting to hash
+        if usage_data.respond_to?(:to_h)
+          usage_data = usage_data.to_h
+        end
+        
+        # Convert hash to appropriate struct
+        return nil unless usage_data.is_a?(Hash)
+        
+        # Normalize keys to symbols
+        normalized = usage_data.transform_keys(&:to_sym)
+        
+        case provider.to_s.downcase
+        when 'openai'
+          create_openai_usage(normalized)
+        when 'anthropic'
+          create_anthropic_usage(normalized)
+        else
+          create_generic_usage(normalized)
+        end
+      end
+      
+      private
+      
+      sig { params(data: T::Hash[Symbol, T.untyped]).returns(T.nilable(OpenAIUsage)) }
+      def self.create_openai_usage(data)
+        # OpenAI uses prompt_tokens/completion_tokens
+        input_tokens = data[:prompt_tokens] || data[:input_tokens] || 0
+        output_tokens = data[:completion_tokens] || data[:output_tokens] || 0
+        total_tokens = data[:total_tokens] || (input_tokens + output_tokens)
+        
+        # Convert prompt_tokens_details and completion_tokens_details to hashes if needed
+        prompt_details = convert_to_hash(data[:prompt_tokens_details])
+        completion_details = convert_to_hash(data[:completion_tokens_details])
+        
+        OpenAIUsage.new(
+          input_tokens: input_tokens,
+          output_tokens: output_tokens,
+          total_tokens: total_tokens,
+          prompt_tokens_details: prompt_details,
+          completion_tokens_details: completion_details
+        )
+      rescue => e
+        DSPy.logger.debug("Failed to create OpenAI usage: #{e.message}")
+        nil
+      end
+      
+      sig { params(value: T.untyped).returns(T.nilable(T::Hash[Symbol, Integer])) }
+      def self.convert_to_hash(value)
+        return nil if value.nil?
+        return value if value.is_a?(Hash) && value.keys.all? { |k| k.is_a?(Symbol) }
+        
+        # Convert object to hash if it responds to to_h
+        if value.respond_to?(:to_h)
+          hash = value.to_h
+          # Ensure symbol keys and integer values
+          hash.transform_keys(&:to_sym).transform_values(&:to_i)
+        else
+          nil
+        end
+      rescue
+        nil
+      end
+      
+      sig { params(data: T::Hash[Symbol, T.untyped]).returns(T.nilable(Usage)) }
+      def self.create_anthropic_usage(data)
+        # Anthropic uses input_tokens/output_tokens
+        input_tokens = data[:input_tokens] || 0
+        output_tokens = data[:output_tokens] || 0
+        total_tokens = data[:total_tokens] || (input_tokens + output_tokens)
+        
+        Usage.new(
+          input_tokens: input_tokens,
+          output_tokens: output_tokens,
+          total_tokens: total_tokens
+        )
+      rescue => e
+        DSPy.logger.debug("Failed to create Anthropic usage: #{e.message}")
+        nil
+      end
+      
+      sig { params(data: T::Hash[Symbol, T.untyped]).returns(T.nilable(Usage)) }
+      def self.create_generic_usage(data)
+        # Generic fallback
+        input_tokens = data[:input_tokens] || data[:prompt_tokens] || 0
+        output_tokens = data[:output_tokens] || data[:completion_tokens] || 0
+        total_tokens = data[:total_tokens] || (input_tokens + output_tokens)
+        
+        Usage.new(
+          input_tokens: input_tokens,
+          output_tokens: output_tokens,
+          total_tokens: total_tokens
+        )
+      rescue => e
+        DSPy.logger.debug("Failed to create generic usage: #{e.message}")
+        nil
+      end
+    end
+  end
+end

data/lib/dspy/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module DSPy
-  VERSION = "0.12.0"
+  VERSION = "0.13.0"
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: dspy
 version: !ruby/object:Gem::Version
-  version: 0.12.0
+  version: 0.13.0
 platform: ruby
 authors:
 - Vicente Reig Rincón de Arellano
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2025-07-24 00:00:00.000000000 Z
+date: 2025-07-25 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: dry-configurable
@@ -186,6 +186,7 @@ files:
 - lib/dspy/lm/strategies/openai_structured_output_strategy.rb
 - lib/dspy/lm/strategy_selector.rb
 - lib/dspy/lm/structured_output_strategy.rb
+- lib/dspy/lm/usage.rb
 - lib/dspy/memory.rb
 - lib/dspy/memory/embedding_engine.rb
 - lib/dspy/memory/in_memory_store.rb