find_paths.py

import collections
import re
import sys


ends = collections.Counter()
paths = collections.Counter()

def clean(path):
    parts = path.split(",")
    result = []
    for part in parts:
        if not part:
            continue
        if part.startswith("WIDE"):
            continue
        if part == "NONE":
            continue
        result.append(part[:-1] if part.endswith("*") else part)
    return ",".join(result)

for line in sys.stdin:
    time, dir, message = line.rstrip("\n").split("\t")
    m = re.match(r"^(?P<src>[^>]+)>(?P<desc>[^,:]+)(?P<path>(,[^,:]+)*):(?P<info>.*)$", message)
    assert m
    src, path = m.group("src", "path")
    cleaned = clean(path)
    parts = cleaned.split(",")
    end = parts[-1] if cleaned else src
    ends[end] += 1
    paths[cleaned] += 1

print("# Ends")
for end, count in ends.most_common():
    print(f"{count:4}: {end}")

print()
print("# Paths")
for path, count in paths.most_common():
    print(f"{count:4}: {path}")