Skip to content

V2PipelineOrchestrator

reference
1 min readUpdated

Kind: Class

Source: atloria-monorepo/packages/doc-automation/src/v2/orchestrator.ts

Methods

MethodSignatureReturns
getStoregetStore()StageResultStore
runrun(progressCallback: V2ProgressCallback)Promise<V2PipelineResult>
cancelcancel()void

Where it refuses work

  • V2PipelineOrchestrator stops the work with Error when !hasOpenAI — “OpenAI provider selected but no API key provided (openaiApiKey or azureOpenaiApiKey)”.
  • V2PipelineOrchestrator stops the work with Error when !endpoint || !apiKey — “Azure Responses provider selected but endpoint or apiKey not provided”.
  • V2PipelineOrchestrator stops the work with Error when !this.parseResult — “Parse result not available”.
  • V2PipelineOrchestrator stops the work with Error when !this.chunkResult — “Chunk result not available”.
  • V2PipelineOrchestrator stops the work with Error when !this.chunkResult || !this.featureResult — “Chunk or feature result not available”.
  • V2PipelineOrchestrator stops the work with Error when !this.analysisResult — “Analysis result not available”.

When something fails

  • V2PipelineOrchestrator handles failure in 44 places: it logs it and continues in 22, lets it reach the caller in 13, discards it silently in 5, and turns it into a return value in 4. A failure discarded silently leaves no trace for whoever debugs this later.

Was this page helpful?

Download as PDF