Code coverage

#!/usr/bin/env python3

import argparse
import json
from collections import defaultdict


def lines2phabricator(filename, lines):
    ret = ""
    numlines = 0
    with open(filename) as f:
        for row in f:
            numlines += 1
    for i in range(1, numlines + 1):
        if i not in lines:
            ret += "N"
        elif lines[i] == 0:
            ret += "U"
        else:
            ret += "C"
    return ret


parser = argparse.ArgumentParser(description='Parse llvm-cov export data.')
parser.add_argument('input', help='input file')
parser.add_argument('coverage', help='coverage output file')
parser.add_argument('summary', help='summary output file')
parser.add_argument('files', nargs='+', help='files to process')
args = parser.parse_args()

# for specification of the format see:
# https://github.com/llvm-mirror/llvm/blob/master/tools/llvm-cov/CoverageExporterJson.cpp

data = json.load(open(args.input, "r"))

totals = defaultdict(lambda: defaultdict(int))
sources = defaultdict(lambda: defaultdict(int))
for export in data["data"]:
    for cfile in export["files"]:
        for segment in cfile["segments"]:
            filename = cfile["filename"]
            if filename not in args.files:
                continue
            line, col, count, has_count, is_region_entry, is_gap_region = \
                segment
            sources[filename][line] += count
    for function in export["functions"]:
        for region in function["regions"]:
            line_start, column_start, line_end, column_end, execution_count, \
                file_id, expanded_file_id, kind = region
            filename = function["filenames"][file_id]
            if filename not in args.files:
                continue
            for i in range(line_start, line_end + 1):
                sources[filename][i] += execution_count
    for total, values in export["totals"].items():
        for key, value in values.items():
            totals[total][key] += value

coverage = {}
for filename, lines in sources.items():
    path = "/".join(filename.split("/")[3:])
    coverage[path] = lines2phabricator(filename, lines)

with open(args.coverage, "w") as f:
    json.dump(coverage, f)

summary = """<!DOCTYPE html>
<h1>Code coverage</h1>
<table>
<tr><th>Coverage</th><th>Total</th></tr>
"""
ROW = "<tr><td>{name}</td><td>{covered} / {count} ({percent:.2%})</td></tr>\n"
for what in ["functions", "instantiations", "lines", "regions"]:
    now = totals[what]
    now["percent"] = now["covered"] / now["count"]
    summary += ROW.format(**dict(totals[what], name=what.capitalize()))
summary += "</table>\n"

with open(args.summary, "w") as f:
    f.write(summary)