Code Quality Analysis

The EmbedOps CLI provides comprehensive code quality analysis capabilities for embedded systems projects, supporting C, C++, Python, Go, and Rust.

Overview

The eo report codequality command analyzes your codebase and generates detailed metrics including:

• Cyclomatic Complexity: Measures code complexity based on control flow
• Lines of Code (LOC): Total, logical, and comment line counts
• Coupling: Efferent coupling measuring dependencies
• Cohesion (LCOM): Lack of Cohesion of Methods metric
• Health Score: Composite quality score (0-100%)

Quick Start

Basic Usage

# Analyze current directory and upload to EmbedOps
eo report codequality .

# Analyze specific directory
eo report codequality ./src

# Local analysis only (no upload)
eo report codequality ./src --local

# Filter by language
eo report codequality ./src --language go

Output Options

# Save to custom file
eo report codequality ./src --local --output quality-report.json

# Verbose output with file-by-file details
eo report codequality ./src --verbose

Command Reference

Syntax

eo report codequality [source directory] [flags]

Flags

Basic Flags

Flag	Description	Default
--language	Filter to specific language (c, cpp, python, go, rust)	All languages
--local, -l	Run locally without uploading to EmbedOps	false
--output, -o	Output file path for local reports	./code-quality-report.json
--verbose, -v	Enable verbose output with file-by-file metrics	false

Complexity Thresholds

Flag	Description	Default
--complexity-excellent	Complexity threshold for excellent quality	10
--complexity-good	Complexity threshold for good quality	20
--complexity-warning	Complexity threshold for warning	30

Lines of Code Thresholds

Flag	Description	Default
--loc-excellent	LOC threshold for excellent quality	200
--loc-good	LOC threshold for good quality	500
--loc-warning	LOC threshold for warning	1000

Cohesion (LCOM) Thresholds

Flag	Description	Default
--lcom-excellent	LCOM % threshold for excellent quality	40
--lcom-good	LCOM % threshold for good quality	65
--lcom-warning	LCOM % threshold for warning	80

Coupling Thresholds

Flag	Description	Default
--coupling-excellent	Coupling threshold for excellent quality	5
--coupling-good	Coupling threshold for good quality	10
--coupling-warning	Coupling threshold for warning	15

Examples

Standard Analysis

# Analyze and upload to EmbedOps
eo report codequality ./src

Output:

Analyzing Code Quality...
══════════════════════════════════════════════════════════════════════
Overall Health: 81% (Excellent)
──────────────────────────────────────────────────────────────────────
Analyzed: 42 files | 10,696 LOC | 328 functions | Avg Complexity: 36.6
──────────────────────────────────────────────────────────────────────
Quality Distribution:
  🟢 Excellent: 21 files (50%)
  🟡 Good:      16 files (38%)
  🟠 Warning:    5 files (12%)
  🔴 Critical:   0 files (0%)
══════════════════════════════════════════════════════════════════════

Custom Thresholds

# Use stricter thresholds
eo report codequality ./src \
  --complexity-excellent 5 \
  --complexity-good 10 \
  --complexity-warning 15 \
  --loc-excellent 100 \
  --loc-good 250 \
  --loc-warning 500

Language-Specific Analysis

# Analyze only Go code
eo report codequality ./src --language go

# Analyze only C/C++ code
eo report codequality ./src --language c
eo report codequality ./src --language cpp

Local Analysis with Verbose Output

# Generate detailed report locally
eo report codequality ./src --local --verbose --output detailed-report.json

Verbose output includes: - File-by-file metrics - Function-level complexity - Detailed health calculations - Line count breakdowns

JSON Output Format

The generated JSON report contains (all percentage metrics are expressed as decimals 0-100):

{
  "summary": {
    "OverallHealth": 81.07,
    "TotalFiles": 42,
    "TotalLOC": 10696,
    "AvgComplexity": 36.6,
    "FunctionsAnalyzed": 328,
    "Distribution": {
      "Excellent": 21,
      "Good": 16,
      "Warning": 5,
      "Critical": 0
    },
    "TopIssues": [
      {
        "Path": "src/complex.go",
        "Health": 53.46,
        "Complexity": 149,
        "LCOM": 0
      }
    ]
  },
  "file_metrics": [
    {
      "FilePath": "src/example.go",
      "Language": "go",
      "Health": 85.5,
      "Metrics": {
        "Complexity": 12,
        "LOC": 150,
        "LogicalLOC": 120,
        "CommentLOC": 30,
        "LCOM": 0.25,
        "Coupling": 3,
        "Functions": 8
      },
      "Functions": [
        {
          "Name": "ProcessData",
          "Complexity": 5,
          "LOC": 25,
          "StartLine": 10,
          "EndLine": 35
        }
      ]
    }
  ]
}

Health Scoring Algorithm

The health score is calculated using weighted factors:

Health Score = 100% - (
  0.35 × Complexity_Factor +
  0.25 × Coupling_Factor +
  0.25 × LCOM_Factor +
  0.15 × LOC_Factor
)

Where each factor is calculated based on how far the metric exceeds the configured thresholds.

Quality Categories

Health Score	Category	Description
80-100%	🟢 Excellent	Well-structured, maintainable code
60-79%	🟡 Good	Acceptable quality with minor issues
40-59%	🟠 Warning	Needs attention and refactoring
0-39%	🔴 Critical	Serious quality issues requiring immediate action

Supported Languages

Go

• Full AST-based complexity analysis
• Package-level coupling detection
• Method cohesion analysis

C/C++

• Function-level complexity (via external tools)
• Include-based coupling analysis
• Basic cohesion metrics

Python

• Function and class complexity
• Import-based coupling
• Class cohesion metrics

Rust

• Function complexity analysis
• Use-statement coupling
• Module cohesion

Integration with EmbedOps Platform

When not using --local flag, the analysis results are automatically uploaded to the EmbedOps Platform where you can:

• View historical quality trends
• Compare quality across branches
• Explore interactive treemap visualizations
• Set custom quality thresholds per repository
• Receive quality alerts in CI/CD pipelines

CI/CD Integration

Add to your CI pipeline:

# .github/workflows/quality.yml
name: Code Quality

on: [push, pull_request]

jobs:
  quality:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v2

      - name: Install EmbedOps CLI
        run: curl -sSL https://embedops.com/install.sh | sh

      - name: Analyze Code Quality
        env:
          EO_API_KEY: ${{ secrets.EO_API_KEY }}
        run: eo report codequality ./src

Troubleshooting

"No files found to analyze"

• Verify the source directory path is correct
• Check that the directory contains supported file types
• Use --verbose to see which files are being discovered

"Failed to analyze complexity"

• Ensure required external tools are installed:
• For C/C++: lizard
• For Go: gocyclo
• For Python: radon
• Check file encoding (UTF-8 recommended)

High LCOM values

LCOM (Lack of Cohesion of Methods) measures how well methods in a file work together: - 0-40%: Highly cohesive (excellent) - 40-65%: Moderately cohesive (good) - 65-80%: Low cohesion (warning) - 80-100%: Very low cohesion (critical)

Consider refactoring files with high LCOM by: - Splitting into smaller, focused modules - Grouping related functionality - Reducing unnecessary dependencies

Best Practices

1. Run regularly: Integrate into CI/CD to track quality over time
2. Set realistic thresholds: Adjust based on your project's complexity
3. Focus on trends: Monitor changes rather than absolute values
4. Address hotspots first: Focus on files with lowest health scores
5. Review critical files: Pay special attention to 🔴 critical files
6. Use language filters: Analyze specific languages for targeted improvements

Additional Resources

• EmbedOps Documentation
• Code Quality Best Practices
• API Reference