Extracting Traceability Information from Products' Feature Sets and Code for Reverse Engineering Software Product Lines (Bachelor's Thesis) (bibtex)
by Lukas Linsbauer
Abstract:
Quite often companies develop a range of similar software products, tailored to individual customers, by reusing code and other artifacts from related projects that share some functionality instead of writing them from scratch. Inadvertently, they do create the potential capability of reverse engineering a Software Product Line. However, tapping into the benefits ordered by this development paradigm poses several challenges. One of them is tracing the features to the code that implements them across the different product variants. In this paper, we present an algorithm to address this problem and evaluate it at the level of class methods and fields. We applied our approach to three case studies of different sizes and problem domains. Out of the more than 22,000 elements at this granularity level only 0.7% could not be traced due to code being shared by disjunctive features.
Reference:
Extracting Traceability Information from Products' Feature Sets and Code for Reverse Engineering Software Product Lines (Bachelor's Thesis) (Lukas Linsbauer), 2012.
Bibtex Entry:
@Baccthesis{Linsbauer,
  author    = {Lukas Linsbauer},
  title     = {Extracting Traceability Information from Products' Feature Sets and Code for Reverse Engineering Software Product Lines (Bachelor's Thesis)},
  year      = {2012},
  abstract  = {Quite often companies develop a range of similar software products,
	tailored to individual customers, by reusing code and other artifacts
	from related projects that share some functionality instead of writing
	them from scratch. Inadvertently, they do create the potential capability
	of reverse engineering a Software Product Line. However, tapping
	into the benefits ordered by this development paradigm poses several
	challenges. One of them is tracing the features to the code that
	implements them across the different product variants. In this paper,
	we present an algorithm to address this problem and evaluate it at
	the level of class methods and fields. We applied our approach to
	three case studies of different sizes and problem domains. Out of
	the more than 22,000 elements at this granularity level only 0.7%
	could not be traced due to code being shared by disjunctive features.},
  file      = {:BSc Theses\\2012 Lukas Linsbauer\\Lukas Linsbauer - Extracting Traceability Information from Products-preprint.pdf:PDF},
  owner     = {AK117794},
  keywords  = {FWF P23115},
  timestamp = {2015.09.21},
}
Powered by bibtexbrowser