An Exploratory Study on Detection of Cloned Code in Information Systems

Code clones are source code parts that are identical or have some degree of similarity to another part of the code. Cloning arises for a variety of reasons, including copy and paste and the reuse of ad-hoc code by programmers. Detection of information system clones is aimed at propagating changes by all clones at the development, maintenance and evolution stages, preserving data consistency, correcting errors, and so on. Clones can be classified as 1, 2, 3 and 4, depending on their similarity and characteristics that classify them as such. Several techniques and tools have been created with the objective of detecting code clones, and for this, they use techniques of representation of the source code in text, token, tree, graphic, hybrid and metrics. This systematic mapping work presents answers to the four research questions, which aim to identify, count and catalog, data from a set of 875 articles, of which 128 were selected, for the selection of relevant information seeking to provide content for the collection of data objectified. In all, 52 clone detection tools were identified, which reinforce the current theme; 26 ways of presenting source code to detect clones, where the commonly used ones stand out for ease of understanding and handling; 13 programming languages in 6 paradigms and the identification, highlighting the great presence of clones detection in object oriented information systems, of all 4 types of clones, as well as semantic and syntactic clones, which reinforces the current questioning of authors of this division search line into four types.