Code Metrics - Lines of Code, Coupling

Code Metrics - Lines of Code, Coupling

Code metrics provide quantitative measurements of software characteristics. They help teams assess code quality, identify problem areas, and track improvement over time.

Size Metrics

  +---------------------------------------------------------+
  |              SIZE METRICS                                |
  +---------------------------------------------------------+
  |                                                         |
  |   Lines of Code (LOC)                                   |
  |   - Total lines in the source file                      |
  |   - Simple but includes comments and whitespace          |
  |                                                         |
  |   Physical Lines of Code (PLOC)                         |
  |   - Non-blank, non-comment lines                        |
  |   - More accurate measure of actual code                 |
  |                                                         |
  |   Logical Lines of Code (LLOC)                          |
  |   - Counted by statements                               |
  |   - Language-dependent                                   |
  |                                                         |
  |   Non-Comment Lines of Code (NCLOC)                     |
  |   - All lines excluding comments                        |
  |   - Commonly used in industry                           |
  |                                                         |
  +---------------------------------------------------------+

Coupling Metrics

  +---------------------------------------------------------+
  |              COUPLING TYPES                              |
  +---------------------------------------------------------+
  |                                                         |
  |   CONTENT COUPLING (worst)                              |
  |   - Module A accesses internal data of Module B         |
  |                                                         |
  |   COMMON/SHARED COUPLING                                |
  |   - Both modules share global data                      |
  |                                                         |
  |   CONTROL COUPLING                                      |
  |   - Module A passes control flag to Module B            |
  |                                                         |
  |   STAMP COUPLING                                        |
  |   - Module A passes data structure to Module B          |
  |                                                         |
  |   DATA COUPLING (best)                                  |
  |   - Module A passes data items to Module B              |
  |   - Minimal dependency                                  |
  |                                                         |
  +---------------------------------------------------------+

Cohesion Metrics

Cohesion measures how closely related the elements within a module are. Higher cohesion is better — it means the module has a clear, focused purpose.

  Cohesion Levels (best to worst):
  1. Functional  - All elements contribute to a single task
  2. Sequential  - Output of one is input of next
  3. Communicational - Elements operate on same data
  4. Procedural  - Elements are part of a sequence
  5. Temporal    - Elements happen at same time
  6. Logical     - Elements are logically related
  7. Coincidental - No meaningful relationship (worst)

Key Takeaways

• LOC measures size but not quality
• Low coupling and high cohesion are desirable
• Use metrics to identify code that needs attention
• Metrics are tools for improvement, not goals in themselves