Description

This is a software product line re-engineering process. The main goal of this process is to assemble and execute a feature retrieval process taking a set/family of individual products as input.

Actors/Roles

These are only recommended roles, other roles may be added if the team decides to.

Domain Engineer: possesses valid knowledge related to an area of human endeavour, an autonomous computer activity, or other specialized discipline.

Architect: has valid knowledge related to software architecture, design and architectural patterns

Analyst: requirements specialist. Must know functional and nonfunctional requirements.

Developer: possesses programming languages and technologies knowledge.

Feature Tester: must know domain glossary, domain constraints, and requirements information.

Feature Retriever: must have knowledge about feature retrieval strategies and techniques.

I/O Artifacts

For each activity/sub-process the inputs are classified as: mandatory (required), optional (can be used or not) or alternative (at least one must be used). Outputs are always mandatory. Another artifacts may be added as needed, but they have to be categorized as well.

• Input Artifacts:

  • Source Code
  • Requirements
    • Requirements List
    • Use Cases
    • Business Rules
  • Design Models
    • Class Diagrams
    • State Machine Diagrams
    • Feature Models
    • Activity Diagrams
  • Domain Information
    • Reference Architecture
    • Reference Requirements

• Output Artifacts:

  • Feature Artifacts
  • Reports
  • Modified Product Artifacts
  • Team Information Reports
  • Domain Artifacts
    • Domain Glossary
    • Domain Constraints List
  • Requirements Specification
  • Artifacts Type Specification
  • Development Information

Documentation Set

The documentation set is the main artifact generated by Assemble. It is composed by a set of different i/o artifacts and its fulled detailed here.

Assembled Process

The assembled process is the main output of the Assemble phase. It is the result of the feature retrieval process tailoring performed by the team. More information can be found here

Guidelines

Techniques

• Static Analysis Techniques

  • Clustering
  • Dependency Analysis
  • Data Flow Analysis

• Information Retrieval Techniques

  • Formal Concept Analysis
  • Latent Semantic Indexing
  • Vector Space Model

• Support Techniques

  • Expert Driven
  • Heuristics
  • Rule-Based

Generic Process

The generic process is used as base to assemble techniques and create the assembled process. Each technique contains a priority order that indicates in which activity of the generic process they most fit in.

• Extract: Features are identified and extracted from the product artifacts.
• Categorize: Features are categorized as mandatory, alternative or optional.
• Group: Features are grouped according to their similarity and dependencies.

After the execution of all activities, a the check artifacts activity is performed. If problems were found, they should be fixed. If there are no problems, the last activity will be performed:

• Create Variability Model: A Variability model of the SPL is created.