Class: Deepsearch::Engine::Steps::Summarization::Process

Inherits:
Object
  • Object
show all
Defined in:
lib/deepsearch/engine/steps/summarization/process.rb

Overview

Generates a final, synthesized answer to the user’s query based on relevant text chunks. It constructs a detailed prompt for an LLM, including the query, context from chunks, and instructions for citing sources, then returns the LLM’s response.

Instance Attribute Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize(query:, relevant_chunks:) ⇒ Process

Returns a new instance of Process.



15
16
17
18
 
# File 'lib/deepsearch/engine/steps/summarization/process.rb', line 15

def initialize(query:, relevant_chunks:)
  @query = query
  @relevant_chunks = relevant_chunks
end

Instance Attribute Details

#query ⇒ Object (readonly)

Returns the value of attribute query.



13
14
15
 
# File 'lib/deepsearch/engine/steps/summarization/process.rb', line 13

def query
  @query
end

#relevant_chunks ⇒ Object (readonly)

Returns the value of attribute relevant_chunks.



13
14
15
 
# File 'lib/deepsearch/engine/steps/summarization/process.rb', line 13

def relevant_chunks
  @relevant_chunks
end

Instance Method Details

#execute ⇒ Object 



20
21
22
23
24
25
26
27
28
29
30
31
32
 
# File 'lib/deepsearch/engine/steps/summarization/process.rb', line 20

def execute
  return Result.new(summary: "No relevant content found to summarize.") if relevant_chunks.empty?

  prompt = build_summary_prompt
  Deepsearch.configuration.logger.debug("Summarizing content with LLM...")
  response = RubyLLM.chat.ask(prompt)
  Deepsearch.configuration.logger.debug("Summarization complete.")

  Result.new(summary: response.content)
rescue StandardError => e
  Deepsearch.configuration.logger.debug("Error during summarization: #{e.message}")
  Result.new(summary: nil, error: e.message)
end