Description

Abstract

At present, there are many software tools that analyse static code from software programs: sonar, PMD, Clang... These tools analyse source code of a programme without execute it. However, it does not exist a tool to perform this kind of analysis for Fortran.

It is clear that a more complete and updated tool to check Fortran code is currently needed, as the options available do not apply the metrics that have others code analysis tools in other programming languages. Some softwares do a lexical analysis from the code, but only a few do a semantical analysis of Fortran code.

FortranAnalyser, has developed and that adapts to the characteristics of the parsing that is wanted to do. It finds common programming flaws on the source code in Fortran for any standard. The program performs an average of the individual scores of the code that will make up the final note of the project.

Introduction

FortranAnalyser is developed in Java. It is multilanguage and cross-platform with GPLv3 licence. When the analysis is done, a quality report is generated in PDF and this file is saved in the temp directory in root directory.

Quality metrics

Here they are exposed the quality metrics that this software check to analyse the quality from the Fortran project selected:

• the use of the "implicit none" statement
• the use of comments in files
• the use of the "exit" statement
• the number of nested loops
• the number of lines with comments
• the use of the "cycle" statement

additionally, the report includes more information about each file that it can be useful for the user:

• full file directory
• file name
• number of lines
• number of funtions
• Number of calls to subroutines
• number of comments
• number of declared variables
• comments:
• on declared functions
• at the beginning of the file
• on declaration of variables
• on declaration of subroutines
• on control structures
• maximum nesting complexity
• number of subroutines
• use or not of the "exit" statement
• use or not of the "cycle" statement
• cyclomatic complexity (McCabe, 1976)

Example of quality report