02671nam a22001097a 4500100006500000245006500065260003000130300000900160500231800169700004102487856003302528  aIndher, Zahid a15MSIT15aSupervisor - Dr. Abdul Wahid Memon  aPersisting comments using custom annotation inside byte code  aNawabshah:bQuest,c2018.  a42p.  aABSTRACT

Developer comments assist newer programmers to understand the code rapidly, the main benefit being their importance during the maintenance phase of Software Development Life Cycle (SDLC). If high level code base is lost due to any reason and only the lower level code base is available, the code base can he converted to high level language by using the reverse engineering technique. There can be problems, if the software package is old or out dated and the programmer who wrote the initial code is either not onboard or not immediately accessible. I such a case, the new developer will inevitable face have a lot of problems and hindrances while trying to understand the source code and specific business terms. In addition, i f the code is recovered using any decompiler then it would be much more difficult for new programmer to fully understand the code within a short time. All of these issues will consequently lead to an increase in development cost to the new programmer. The absence of the comments inside recovered source code will also ad<l to the heap of problems. This research proposes the comments recovery approach which helps recover the code along with the comments which were added by initial developer. This research mainly focuses on java program mining language, however java discards the comments when it compiles a high level code to byte code. "Custom annotations" feature of Java is used to pinpoint the comments as specific location. Comments and annotation both are types of Metadata. Java programming does not let the comments be the part of bytecode. These are only allowed to go with bytecode. Comments are converted to custom annotation and pinpointed at specific location in source code, as soon process finishes and compilation completes and then the class file is generated. The comments will be inside bytecode. Existing technique helps the decompiler recover the comments by extracting comments in separate file and code in class file and finally merge them when decompiled. Using existing technique will lend to  a dependency of compiler as well as decompiler to recover the developer comments.
This research purposes universal technique embedding the comments with improved file size,  correctness  of  code,  correctness  or comments  and  improved  conversion
  aDepartment of Information Technology  uhttps://tinyurl.com/2d4y2wp5