Wrap lizard's CSV complexity output in a usable object, and give usage example

lizard_complexity.py

import sys
from collections import namedtuple


# A friendlier named object for working with each output line
Analysis = namedtuple("Analysis", [
    "NLOC", # Count of number of active LOC (excl whitespace)
    "CCN",  # Cyclomatic Complexity Number
    "token", # token count: how many syntactic tokens ("if", "[", "5", etc) are used in the functions
    "PARAM", # number of parameters to the function
    "length", # total number of file lines
    "blob", # string detailing file, function, and line numbers
    "file",
    "function",
    "signature",
    "line_start",
    "line_end",
])


results = []

def parse_analysis(lizard_csv_lines):
    """ Given an iterable of the lines of lizard CSV output, parse each to an Analysis object.
    """
    for line in lizard_csv_lines:
        line = line.strip()
        line_parts = eval("["+line+"]")
        results.append(Analysis(*line_parts ))

    return results


def sorted_results(results, keyname="CCN"):
    """ Given a list of Analysis objects, sort according to the named key.

    Typically this would be "CCN" for the cyclomatic complexity, or "NLOC" for the number of lins of code
    """
    results = results[:]
    results.sort(key=lambda a: a.__getattribute__(keyname))
    return results


def filter_upper_threshold(results, keyname="CCN", value=20):
    return [res for res in results if res.__getattribute__(keyname) >= value]


def main():
    """ An example implementation of how to use the functions herein.
    """
    max_complexity=20

    results = parse_analysis(sys.stdin)
    results = filter_upper_threshold(results, keyname="NLOC", value=max_complexity)

    for res in results:
        print(f"CCN={res.CCN} --> NLOC={res.NLOC} : {res.file} : {res.line_start} : {res.signature}")

    res_count = len(results)
    assert res_count == 0, f"Found {res_count} items of complexity higher than {max_complexity}"


if __name__ == "__main__":
    main()