Skip to content

philips-software/EagleVision

Repository files navigation

EagleVision

Why use this tool

This tool will help in uncovering the details in a code repository. For example,

  • Patterns in the code

  • Type of asserts used in the test functions

  • Test functions with no assert / pattern analysed

  • Similarity in the source code/ test code functions

Note: Similarity analysis uses cosine algorithm, if the repo path given for root and repo size is relatively large, re consider to give sub-folders, Similarity analysis bound to take long time, since the computation is in memory if not provided with adequate size expect potential memory overflow

  • Cyclomatic complexity of functions

  • Repository statistics like, Type of file, LOC, Comments, Code etc.

Dependencies

  • python 3.8 : 64 bit

  • python packages (functiondefextractor, similarity-processor, lizard)

  • cloc package npm i -g [email protected] https://www.npmjs.com/package/cloc Note: cloc has a dependency for pearl, if not installed please install pearl from https://www.perl.org/

  • third party packages [Ctags, grep]

Installation

INSTALL.md

Usage

Commandline

>>>python -m eaglevision --p "path\to\input\json"
  • sample json input,
[
  {
    "path": "repo/path",
    "run_pattern_match":true,
    "run_similarity":true,
    "extraction_annotation": null,
    "extraction_delta": null,
    "extraction_exclude": "*/test_resource/*",
    "pattern_match": ["assert"],
    "pattern_seperator": ["("],
    "similarity_range": "70,100",
    "run_cloc_metric":true,
    "cloc_args": "--exclude-dir=src --exclude-ext=*.cpp,*.java",
    "run_cyclomatic_complexity":true,
    "cyclo_args": "-l java  -l python",
    "cyclo_exclude": ["*.cpp","*.java"],
    "report_folder": null
  }
]
  • Input Description,
    "path": Path of the repository to be analysed
    "run_pattern_match": On/OFF switch for running Pattern match
    "run_similarity": On/OFF switch for running Similarity check
    "extraction_annotation": Functions with this annotation will be
                             extracted else all functions
    "extraction_delta": If substring of annotation is given,
                        this input (integer value) will take number of
                        lines above and below the annotation to report,
    "extraction_exclude": Pattern to exclude for Similarity and Pattern check
    "pattern_match": Type of pattern to analyse in the source code
    "pattern_seperator": Seperator in the pattern, left side of which will
                         be used for pivot reporting
    "similarity_range": Range of similarity of interest example: "70,100",
    "run_cloc_metric": On/OFF switch for running Cloc / types of
                      files/number etc in the repo
    "cloc_args": Any additional args for Cloc tool
                 example:"--exclude-dir=src --exclude-ext=*.cpp,*.java",
    "run_cyclomatic_complexity":On/OFF switch for running cyclomatic complexity check,
    "cyclo_args": Any additional args for Lizard / cyclomatic complexity tool
                 example:"-l java  -l python"
    "cyclo_exclude": Pattern to exclude for cyclomatic complexity check
                     example: ["*.cpp","*.java"]
    "report_folder": Path where report to be placed, if null will be using the path of repo

Note:

All the inputs are taken from the json file

  1. Do not forget to have the json as list [...]

  2. Make sure pattern_match and pattern_seperator is of same length list if you are not interested in any pattern_seperator for s specific pattern_match , mark it null in pattern_seperator

  3. Make sure mark it null if a string or list parameter is not used

  4. Make sure mark it true/false for bool type

refer https://www.npmjs.com/package/cloc for cloc args
refer https://pypi.org/project/lizard/ for cyclo args
refer https://pypi.org/project/functiondefextractor/ for extraction_exclude
refer https://pypi.org/project/similarity-processor/ for similarity

Output

  • Output will be available in same folder as path given in json under EagleVisionReport

Contact

MAINTAINERS.md

License

License.md

About

No description, website, or topics provided.

Resources

License

Stars

Watchers

Forks

Packages

No packages published