Code Similarity

Checks for identical or similar code. Currently looks for similar or identical functions, if, for, switch, while, do/while, try and block expressions. (available since 2.6.0)

Configuration

{
   "type": "CodeSimilarity",
    "props": {
        "severity": "INFO",
        "severityIdentical": "WARNING",
        "thresholdIdentical": 8,
        "thresholdSimilar": 12
    }
}

Options

severityIdentical - Severity level for identical code blocks (severity is used for reporting similar code blocks)

thresholdIdentical - number of tokens up to which where identical code is tolerated

thresholdSimilar - number of tokens up to which similar code is tolerated

(tokens are keywords, constants, operators, etc.)

Similar and identical code

CodeSimilarity determines similar and identical code parts by generating and comparing hashes for each eligible code section.

Code is considered identical if names, keywords, operations, values and overall structure are identical in two or more places.

Code is considered similar if keywords and overall structure are identical, names and values may vary, and operators need to fall into identical categories.

If e.g. code A has a + in one place, whereas code B has a - in the same location and code C uses &, but everything else is identical then A and B show up as similar, C will not. For operator categories see Haxe manual.

Comments and whitespace have no influence to detecting similar or identical code fragments.

CodeSimilarity will only look at function, if, for, switch, while, do/while, try and { as starting points for similar or identical code.

Current version of check will not look at individiual lines outside of or surrounding keywords / tokens of that list. So there might be code fragments that CodeSimilarity will not mark as similar or identical, even though they might preceed or follow a section of code that was found to be similar.