log-coverage-analyzer

This skill performs comprehensive log coverage analysis for code repositories. It identifies logging deficiencies in call chains and detects high-frequency log risks that may impact performance.

Analysis Principles

Log Level Definitions

Log Level	Behavior	Use Case
LOGE	Always prints, never lost	Critical - Must exist on all error return paths
LOGI	Always prints, may be lost if high-frequency triggered at same location	Important - Success paths, key operations
LOGD	Only prints when debug switch enabled	Optional - Debug information only
LOGW	Only prints when debug switch enabled	Optional - Warning information only

Core Analysis Rules

1. Log Coverage Rule: Each branch in a function call chain must have LOGE or LOGI level logging
2. High-Frequency Risk Rule: Functions that may be called frequently (loops, timers, callbacks, event handlers) with LOGE/LOGI logs pose performance risks
3. Error Path Rule: All error return paths MUST have LOGE logging
4. Context Rule: All logs must include sufficient context information (IDs, states, parameters)

Execution Workflow

Step 00: Create Analysis Plan

Description: Create detailed analysis plan before starting

Actions:

• Use TodoWrite tool to create task list
• Plan analysis scope:
  • Source file discovery and filtering
  • Function call chain extraction
  • Log coverage analysis per branch
  • High-frequency risk detection
  • Report generation

Output: Analysis plan created via TodoWrite tool

Step 01: Discover Source Files

Description: Find all relevant source files in the repository

Actions:

• Use Glob tool to find source files based on language patterns:
  • C/C++: **/*.cpp, **/*.cc, **/*.cxx, **/*.h, **/*.hpp
  • Java: **/*.java
  • Python: **/*.py
  • JavaScript/TypeScript: **/*.js, **/*.ts
  • Go: **/*.go
  • Rust: **/*.rs
• Exclude test files if specified (patterns: **/test/**, **/tests/**, **/*_test.*)
• Exclude build directories (patterns: **/build/**, **/out/**, **/target/**)

Output: List of source files to analyze

Step 02: Scan Logging Patterns

Description: Scan all source files for logging macros/functions

Actions:

• Use Grep tool to find log statements with pattern:
  • Case-insensitive search for: LOG[DEIW], HILOG[DEIW], ALOG[DEIW]
  • Also search for: \.log[deiw]\(, Log\.[deiw], LOG\.
• Count occurrences per file
• Identify log level distribution

Output: Log statistics per file

Step 03: Extract Functions and Call Chains

Description: Parse source files to extract function definitions and call relationships

Actions:

• For each source file:
  • Identify function definitions (various patterns per language)
  • Extract function bodies including all branches
  • Identify function calls within each function
  • Build call chain graph
• Map branch conditions (if/else, switch/case, try/catch, early returns)

Output: Function list with call relationships

Step 04: Analyze Log Coverage per Branch

Description: Check each branch in call chains for required logging

Actions:

• For each function in call chain:
  • Identify all branches:
    • Early returns
    • if/else branches
    • switch/case branches
    • try/catch blocks
    • loop exits with error conditions
  • Check each branch for LOGE or LOGI presence
  • Flag branches without required logging as Log Deficiency
  • Check error return paths specifically for LOGE

Deficiency Classification:

• High Priority: Error return path without LOGE
• Medium Priority: Success path without LOGI for multi-step operations
• Low Priority: Branch without any logging (non-critical)

Output: List of log deficiencies with locations

Step 05: Detect High-Frequency Log Risks

Description: Identify functions with LOGE/LOGI that may be called frequently

Actions:

• Identify high-frequency function patterns:
  • Network data callbacks (OnDataReceived, OnPacketReceived, OnMessage)
  • Timer callbacks (OnTimer, Tick, Update)
  • Event handlers (HandleEvent, ProcessEvent, OnEvent)
  • Loop operations (ProcessItem, HandlePacket, SendData)
  • Stream processing (ProcessStream, HandleFrame, EncodeFrame)
  • Message queue handlers (OnMessage, Dispatch)
• For functions with LOGE/LOGI in these patterns:
  • Analyze call frequency potential
  • Classify risk level: High/Medium/Low
  • Flag as High-Frequency Risk

Risk Classification:

• High Risk: Per-packet/per-frame LOGE/LOGI in data path
• Medium Risk: Per-message LOGI in messaging path
• Low Risk: Low-frequency callbacks with logs

Output: List of high-frequency risks with locations

Step 06: Generate Analysis Report

Description: Create comprehensive analysis report with all findings

Actions:

• Generate Markdown report with:
  1. Summary statistics
  2. Log deficiencies with fix suggestions
  3. High-frequency risks with fix suggestions
  4. Call chain analysis
  5. Detailed code examples
• Use Write tool to save report to log-coverage-report-YYYYMMDD-HHmmss.md

Output: Complete analysis report

Language-Specific Patterns

C/C++

Function Definition Patterns:

[return_type] [class::]function_name([parameters]) {
[\w\s:*,]*\{

Log Patterns:

• LOG[DEIW]\(, HILOG[DEIW]\(, ALOG[DEIW]\(
• __android_log_print
• OHOS::HiviewDFX::HiLog::[Error|Warn|Info|Debug]

Java

Function Definition Patterns:

(public|private|protected)?(\s+static)?\s+\w+\s+\w+\s*\(.*\)\s*(throws\s+[\w\s,]+)?\s*\{

Log Patterns:

• Log\.[deiw]\(
• Logger\.(error|warn|info|debug)\(
• Timber\.[deiw]\(

Python

Function Definition Patterns:

def\s+\w+\s*\(.*\)\s*->?\s*.*:

Log Patterns:

• logger\.(error|warning|info|debug)\(
• logging\.(error|warning|info|debug)\(
• print\(

JavaScript/TypeScript

Function Definition Patterns:

function\s+\w+\s*\(.*\)\s*\{
|\w+\s*\([^)]*\)\s*(=>|\{)

Log Patterns:

• console\.(error|warn|info|log)\(
• logger\.(error|warn|info|debug)\(

Fix Recommendations

For Log Deficiencies

Error Return Path (Must Fix):

// BEFORE:
if (remote == nullptr) {
    return ERR_NULL_OBJECT;  // ❌ No LOGE
}

// AFTER:
if (remote == nullptr) {
    HILOGE("FunctionName: remote is null, context=%{public}d", context);
    return ERR_NULL_OBJECT;
}

Success Path (Should Fix):

// BEFORE:
int32_t CreateSession(...) {
    // ... initialization code
    return ERR_OK;  // ❌ No LOGI on success
}

// AFTER:
int32_t CreateSession(...) {
    // ... initialization code
    HILOGI("CreateSession: success, sessionId=%{public}d, name=%{public}s", id, name);
    return ERR_OK;
}

For High-Frequency Risks

Remove High-Frequency LOGI:

// BEFORE:
void OnPacketReceived(int socketId, const void* data, uint32_t len) {
    HILOGI("packet received: socket=%{public}d, len=%{public}u", socketId, len);  // ⚠️ Per-packet
    // ... process packet
}

// AFTER:
void OnPacketReceived(int socketId, const void* data, uint32_t len) {
    // Removed per-packet LOGI
    static std::atomic<uint64_t> packetCount{0};
    if (++packetCount % 1000 == 1) {
        HILOGI("Packet stats: count=%{public}llu, socket=%{public}d",
            packetCount.load(), socketId);
    }
    // ... process packet
}

Use Throttled LOGE for Errors:

// BEFORE:
void ProcessPacket(const Packet* pkt) {
    if (pkt == nullptr) {
        HILOGE("packet is null");  // ⚠️ Per-packet error
        return;
    }
}

// AFTER:
void ProcessPacket(const Packet* pkt) {
    if (pkt == nullptr) {
        static std::atomic<uint32_t> errorCount{0};
        if (++errorCount % 100 == 1) {
            HILOGE("ProcessPacket: null packet, count=%{public}u", errorCount.load());
        }
        return;
    }
}

Output Format

================================================================================
Log Coverage Analysis Report
================================================================================

Repository: <repository_path>
Analysis Date: <timestamp>
Files Analyzed: <count>

## Summary Statistics

| Metric | Count |
|--------|-------|
| Total Source Files | <number> |
| Total Functions | <number> |
| Functions Analyzed | <number> |
| Log Deficiencies | <number> |
| High-Frequency Risks | <number> |

## Log Deficiencies

### [High/Medium/Low] Priority

**File**: `<file_path>`
**Function**: `<function_name>`
**Lines**: `<line_range>`
**Issue**: `<description>`

**Evidence**:
```cpp
<code snippet>

Impact: <explain impact on debugging/troubleshooting>

Fix:

<fixed code>

---

High-Frequency Risks

[High/Medium/Low] Risk

File: <file_path> Function: <function_name> Lines: <line_range> Risk:

Evidence:

<code snippet>

Analysis:

Fix:

<fixed code with throttling/statistics>

---

Call Chain Analysis

Call Chain: <chain_name>

<function_1>()
    ↓ [✓/✗/⚠️] <log_status>
<function_2>()
    ↓ [✓/✗/⚠️] <log_status>
<function_3>()
    ├─ [✓/✗/⚠️] branch_1
    ├─ [✓/✗/⚠️] branch_2
    └─ [✓/✗/⚠️] branch_3

Legend:

• ✓ : Has required LOGE/LOGI
• ✗ : Missing required logging
• ⚠️ : Has high-frequency log risk

---

Recommendations

Immediate Actions (P0)

Follow-up Actions (P1)

Long-term Improvements (P2)

================================================================================

## Parameters

| Parameter | Required | Description |
|-----------|----------|-------------|
| `--path` | No | Repository path (default: current directory) |
| `--exclude-tests` | No | Exclude test files (default: true) |
| `--lang` | No | Language filter (cpp, java, python, js, all) |
| `--output` | No | Output report path |

## Usage Examples

Analyze current directory

/log-coverage-analyzer

Analyze specific repository

/log-coverage-analyzer --path /path/to/repo

Analyze C++ code only

/log-coverage-analyzer --lang cpp --path /path/to/repo

Include test files in analysis

/log-coverage-analyzer --exclude-tests false

Custom output location

/log-coverage-analyzer --output /path/to/report.md

## Tips

- Use `Grep` with `output_mode: content` and `-B/-C` flags for context
- Use `Read` tool with `offset` and `limit` for large files
- For large repositories, focus on critical directories first
- High-frequency risks should be prioritized over minor log deficiencies
- Always include context (IDs, states, parameters) in LOGE messages