Functions
	get_shard_links (uri)

	print_reencoded (text)

	print_output_all (text)

	print_output_div (shard, text, keywords)

	get_output_sum (shard, text, should_print, s=[0, 0, 0, 0, 0, 0], divs=[])

	get_stats (uri, output_type, keywords, output_csv)

	main ()

Variables
	P_DIV = re.compile("(Isolate.+? !DIVERGENCE! (\n\|.)+?)Isolate ", re.MULTILINE)

	P_SUM

	P_SHARD = re.compile(r".*make_a_fuzz_shard_(\d+)")

Detailed Description

Webscraper for make_a_fuzz nightly cluster run results.

Given the uri of a make_a_fuzz run, this script will first
extract the links pointing to each of the individual shards
and then parse the output generated by each shard to
find divergences reported by the dartfuzz_test.dart program,
concatenate all output, or summarize all test results.

Example:
  collect_data.py --type sum
      https://ci.chromium.org/p/dart/builders/ci.sandbox/fuzz-linux/303

Function Documentation

◆ get_output_sum()

collect_data.get_output_sum	(	shard,
		text,
		should_print,
		s = `[0, 0, 0, 0, 0, 0]`,
		divs = `[]`
	)

Definition at line 79 of file collect_data.py.

def get_output_sum(shard, text, should_print, s=[0, 0, 0, 0, 0, 0], divs=[]):
    m = P_SUM.findall(text)
    if not m:
        sys.stderr.write("Failed to parse shard %s stdout for summary" % shard)
        return
    for test in m:
        if int(test[-1]) == 1:
            divs.append(shard)
        for i in range(len(s)):
            s[i] += int(test[i])
    if should_print:
        print(
            "Tests: %d Success: %d (Rerun: %d) Skipped: %d Timeout: %d "
            "Divergences: %d (failing shards: %s)    \r" %
            tuple(s + [", ".join(divs) if divs else "none"]),
            end="")
    return s
 
 

◆ get_shard_links()

collect_data.get_shard_links ( uri )

Definition at line 44 of file collect_data.py.

def get_shard_links(uri):
    links = []
    resp = requests.get(uri)
    soup = BeautifulSoup(resp.text, "html.parser")
    for a in soup.findAll("a"):
        if "stdout" in a.text:
            href = a["href"]
            if ("make_a_fuzz_shard" in href and "__trigger__" not in href):
                links.append(href + "?format=raw")
    return links
 
 

◆ get_stats()

collect_data.get_stats	(	uri,
		output_type,
		keywords,
		output_csv
	)

Definition at line 98 of file collect_data.py.

def get_stats(uri, output_type, keywords, output_csv):
    resp = requests.get(uri)
 
    if output_type == "all":
        print_output_all(resp.text)
    elif output_type == "div":
        shard = P_SHARD.findall(uri)[0]
        print_output_div(shard, resp.text, keywords)
    elif output_type == "sum":
        shard = P_SHARD.findall(uri)[0]
        should_print = not output_csv
        return get_output_sum(shard, resp.text, should_print)
    return None
 
 

◆ main()

collect_data.main ( )

Definition at line 113 of file collect_data.py.

def main():
    parser = argparse.ArgumentParser(description=__doc__)
    parser.add_argument(
        "--type",
        choices=("div", "sum", "all"),
        required=True,
        help=
        "Select output type (div: divergence report, sum: summary, all: complete stdout)"
    )
    parser.add_argument(
        "--filter",
        nargs="+",
        default=[],
        help="Do not include divergences containing these keywords.")
    parser.add_argument(
        "--output-csv",
        dest="output_csv",
        action="store_true",
        default=False,
        help=
        "Print output in CSV format to stdout. Only supported for --type=sum")
    parser.add_argument(
        "uri",
        type=str,
        help=
        "Uri of one make_a_fuzz run from https://ci.chromium.org/p/dart/builders/ci.sandbox/fuzz-linux."
    )
    args = parser.parse_args()
    if args.type != 'sum' and args.output_csv:
        print('Error: --output-csv can only be provided for --type=sum')
        return
 
    shard_links = get_shard_links(args.uri)
 
    if len(shard_links) == 0:
        print("Invalid run")
        sys.exit(-1)
        return
 
    for link in shard_links:
        stats = get_stats(link, args.type, args.filter, args.output_csv)
    if args.output_csv:
        print("%d,%d,%d,%d,%d,%d" % tuple(stats))
    else:
        print("")
    sys.exit(0)
 
 

◆ print_output_all()

collect_data.print_output_all ( text )

Definition at line 61 of file collect_data.py.

def print_output_all(text):
    print_reencoded(text)

◆ print_output_div()

collect_data.print_output_div	(	shard,
		text,
		keywords
	)

Definition at line 65 of file collect_data.py.

def print_output_div(shard, text, keywords):
    sys.stderr.write("Shard: " + shard + "  \r")
    m = P_DIV.findall(text)
    if m:
        for x in m:
            keep = True
            for word in keywords:
                if word in x[0]:
                    keep = False
                    break
            if keep:
                print_reencoded(x[0])
 
 

◆ print_reencoded()

collect_data.print_reencoded ( text )

Definition at line 56 of file collect_data.py.

def print_reencoded(text):
    # Re-encoding avoids breaking some terminals.
    print(text.encode("ascii", errors="ignore").decode("unicode-escape"))
 
 

Variable Documentation

◆ P_DIV

collect_data.P_DIV = re.compile("(Isolate.+? !DIVERGENCE! (\n|.)+?)Isolate ", re.MULTILINE)

Definition at line 32 of file collect_data.py.

◆ P_SHARD

collect_data.P_SHARD = re.compile(r".*make_a_fuzz_shard_(\d+)")

Definition at line 41 of file collect_data.py.

◆ P_SUM

collect_data.P_SUM

Initial value:

=  re.compile(
    r"^Tests: (\d+) Success: (\d+) "
    r"\‍(Rerun: (\d+)\‍) Skipped: (\d+) "
    r"Timeout: (\d+) Divergences: (\d+)", re.MULTILINE)

Definition at line 35 of file collect_data.py.

Functions

Variables