Refactor: Extract duplicate log metrics finalization logic

[Copilot]

Three engine parsers (Copilot, Codex, Claude) duplicated ~15 lines of metrics finalization: appending sequences, converting tool call maps to sorted slices, counting errors via pattern matching.

Changes

• Added FinalizeToolMetrics() in metrics.go
  Handles full finalization (sequences, token usage, turns, sorting, error patterns). Used by Copilot and Codex.

• Added FinalizeToolCallsAndSequence() in metrics.go
  Lightweight variant for Claude, which extracts token/turns from result entries rather than accumulating during parse.

• Updated engine parsers
  Replaced 25-28 lines per engine with single function calls.

Before

// Copilot, Codex: Similar 25-line blocks
if len(currentSequence) > 0 {
    metrics.ToolSequences = append(metrics.ToolSequences, currentSequence)
}
metrics.TokenUsage = maxTokenUsage
metrics.Turns = turns
for _, toolInfo := range toolCallMap {
    metrics.ToolCalls = append(metrics.ToolCalls, *toolInfo)
}
sort.Slice(metrics.ToolCalls, func(i, j int) bool {
    return metrics.ToolCalls[i].Name < metrics.ToolCalls[j].Name
})
errorPatterns := e.GetErrorPatterns()
if len(errorPatterns) > 0 {
    metrics.Errors = CountErrorsAndWarningsWithPatterns(logContent, errorPatterns)
}

After

// Copilot, Codex
FinalizeToolMetrics(&metrics, toolCallMap, currentSequence, turns, tokenUsage, logContent, e.GetErrorPatterns())

// Claude
FinalizeToolCallsAndSequence(&metrics, toolCallMap, currentSequence)

Impact

• Code reduction: 54 lines across engines
• Tests: 9 cases covering edge cases, sorting, error patterns
• Pattern established: Clear template for future engines

Original prompt

---

This section details on the original issue you should resolve

<issue_title>[duplicate-code] 🔍 Duplicate Code Detected: Log Metrics Finalization Across Engines</issue_title>
<issue_description>Analysis of commit 0dffa81

Assignee: @copilot

Summary

The engine-specific ParseLogMetrics implementations all repeat the same block that finalizes tool call metrics (appending the last sequence, copying the tool call map into a slice, sorting it, and counting errors). Copilot, Codex, and Claude pipelines each copy this logic verbatim with only variable name changes.

Duplication Details

Pattern: Repeated tool call finalization logic in log parsers

• Severity: Medium
• Occurrences: 3
• Locations:
  • pkg/workflow/copilot_engine.go:492
  • pkg/workflow/codex_engine.go:407
  • pkg/workflow/claude_logs.go:340
• Code Sample:

  if len(currentSequence) > 0 {
      metrics.ToolSequences = append(metrics.ToolSequences, currentSequence)
  }
  metrics.TokenUsage = maxTokenUsage
  metrics.Turns = turns
  for _, toolInfo := range toolCallMap {
      metrics.ToolCalls = append(metrics.ToolCalls, *toolInfo)
  }
  sort.Slice(metrics.ToolCalls, func(i, j int) bool {
      return metrics.ToolCalls[i].Name < metrics.ToolCalls[j].Name
  })
  errorPatterns := e.GetErrorPatterns()
  if len(errorPatterns) > 0 {
      metrics.Errors = CountErrorsAndWarningsWithPatterns(logContent, errorPatterns)
  }

Impact Analysis

• Maintainability: Updates to how sequences, sorting, or error aggregation work must be repeated in three places, increasing the chance of drift when adding new engines.
• Bug Risk: Any bug fixes to the metric finalization (e.g., changing how warnings are counted) must be manually mirrored everywhere; missing an occurrence yields inconsistent telemetry.
• Code Bloat: The repeated blocks add ~15 lines per engine and make the already large log parsing routines harder to scan.

Refactoring Recommendations

1. Extract shared helper

   • Move the duplication into a helper such as finalizeToolMetrics(metrics *LogMetrics, toolCallMap map[string]*ToolCallInfo, currentSequence []string, turns int, tokenUsage int, logContent string, getPatterns func() []ErrorPattern).
   • Estimated effort: Medium (1-2h) because the helper can live in pkg/workflow/log_parser.go and accept function pointers for engine-specific hooks.
   • Benefits: Single source of truth for tool sequencing and error aggregation.

2. Standardize sequence tracking

   • Consider wrapping the shared state (tool call map, current sequence, turns, token totals) in a struct so engines update fields but delegate finalization to one method.
   • Benefits: Simplifies adding new engines like Claude variants without extra boilerplate.

Implementation Checklist

• Review duplication findings
• Prioritize refactoring tasks
• Create refactoring plan
• Implement changes
• Update tests
• Verify no functionality broken

Analysis Metadata

• Analyzed Files: 3
• Detection Method: Serena semantic code analysis
• Commit: 0dffa81
• Analysis Date: 2025-11-15T21:06:44Z

AI generated by Duplicate Code Detector</issue_description>

Comments on the Issue (you are @copilot in this section)

• Fixes [duplicate-code] 🔍 Duplicate Code Detected: Log Metrics Finalization Across Engines #4098

---

✨ Let Copilot coding agent set things up for you — coding agent works faster and does higher quality work when set up for your repo.

[Copilot]

Pull Request Overview

This PR refactors duplicate log metrics finalization logic that was repeated across three engine parsers (Copilot, Codex, and Claude). The refactoring extracts ~15 lines of identical finalization code (appending sequences, converting tool call maps to sorted slices, counting errors via pattern matching) into two reusable helper functions in metrics.go.

Key changes:

• Adds FinalizeToolMetrics() for full finalization (used by Copilot and Codex engines)
• Adds FinalizeToolCallsAndSequence() as a lightweight variant (used by Claude engine which manages tokens/turns separately)
• Replaces 25-28 lines per engine with single function calls

Reviewed Changes

Copilot reviewed 5 out of 5 changed files in this pull request and generated 3 comments.

Show a summary per file

File	Description
pkg/workflow/metrics.go	Adds two new helper functions for metrics finalization and imports sort package
pkg/workflow/metrics_test.go	Adds comprehensive test coverage with 9 test cases across 2 test functions
pkg/workflow/copilot_engine.go	Replaces 25 lines of finalization logic with single FinalizeToolMetrics() call
pkg/workflow/codex_engine.go	Replaces 25 lines of finalization logic with single FinalizeToolMetrics() call
pkg/workflow/claude_logs.go	Replaces 18 lines of finalization logic with FinalizeToolCallsAndSequence() call and removes unused sort import

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.