Class: Verica::RubyOpenAIWrapper

Inherits:
Object
  • Object
show all
Defined in:
lib/verica/ruby_openai_wrapper.rb

Overview

The community ruby-openai gem (alexrudall) has a different API than the official openai gem: client.chat(parameters: { model:, messages: }) returning a plain Hash (string keys) instead of typed objects. Both gems define OpenAI::Client, so an app uses one or the other; this thin decorator is the ruby-openai counterpart of OpenAIWrapper. Everything delegates to the real client; only chat is intercepted to emit ONE gen_ai.* span with the SAME pinned semconv the Verica normalizer accepts. Fail-open: instrumentation errors never reach the caller; provider errors always do.

Defined Under Namespace

Classes: StreamAccumulator

Instance Method Summary collapse

Constructor Details

#initialize(client) ⇒ RubyOpenAIWrapper

Returns a new instance of RubyOpenAIWrapper.



15
16
17
 
# File 'lib/verica/ruby_openai_wrapper.rb', line 15

def initialize(client)
  @client = client
end

Dynamic Method Handling

This class handles dynamic methods through the method_missing method

#method_missing(name) ⇒ Object 



55
56
57
 
# File 'lib/verica/ruby_openai_wrapper.rb', line 55

def method_missing(name, ...)
  @client.public_send(name, ...)
end

Instance Method Details

#chat(parameters: {}) ⇒ Object 



19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
 
# File 'lib/verica/ruby_openai_wrapper.rb', line 19

def chat(parameters: {})
  model = parameters[:model] || parameters['model']
  span = safely do
    tracer = OpenTelemetry.tracer_provider.tracer('verica-observability', Verica::VERSION)
    tracer.start_span("chat #{model}", kind: :client)
  end

  # Streaming: ruby-openai delivers each chunk to the caller's proc and the
  # returned value is not a stable Hash. We wrap that proc so our accumulator
  # sees every chunk first, then hand it to the caller unchanged. The HTTP
  # call is synchronous, so by the time `@client.chat` returns the stream has
  # fully drained and the accumulator is complete.
  accumulator = nil
  call_parameters = parameters
  stream_key = stream_key_for(parameters)
  if stream_key && parameters[stream_key].respond_to?(:call)
    acc = StreamAccumulator.new
    built = safely { build_streaming_parameters(parameters, stream_key, acc) }
    if built
      accumulator = acc
      call_parameters = built
    end
  end

  begin
    response = @client.chat(parameters: call_parameters)
    safely { annotate(span, parameters, response, accumulator) } if span
    response
  rescue StandardError => e
    safely { span&.status = OpenTelemetry::Trace::Status.error(e.message) }
    raise
  ensure
    safely { span&.finish }
  end
end

#respond_to_missing?(name, include_private = false) ⇒ Boolean

Returns:

  • (Boolean) 


59
60
61
 
# File 'lib/verica/ruby_openai_wrapper.rb', line 59

def respond_to_missing?(name, include_private = false)
  @client.respond_to?(name, include_private) || super
end