test_runner.py

#!/usr/bin/env python3
# Copyright (c) 2014-2016 The Bitcoin Core developers
# Copyright (c) 2017 The Bitcoin developers
# Distributed under the MIT software license, see the accompanying
# file COPYING or http://www.opensource.org/licenses/mit-license.php.
"""Run regression test suite.

This module calls down into individual test cases via subprocess. It will
forward all unrecognized arguments onto the individual test scripts.

Functional tests are disabled on Windows by default. Use --force to run them anyway.

For a description of arguments recognized by test scripts, see
`test/functional/test_framework/test_framework.py:BitcoinTestFramework.main`.

"""

import argparse
import configparser
import datetime
import os
import time
import shutil
import signal
import sys
import subprocess
import tempfile
import re
import logging
import xml.etree.ElementTree as ET
import json

# Formatting. Default colors to empty strings.
BOLD, BLUE, RED, GREY = ("", ""), ("", ""), ("", ""), ("", "")
try:
    # Make sure python thinks it can write unicode to its stdout
    "\u2713".encode("utf_8").decode(sys.stdout.encoding)
    TICK = "✓ "
    CROSS = "✖ "
    CIRCLE = "○ "
except UnicodeDecodeError:
    TICK = "P "
    CROSS = "x "
    CIRCLE = "o "

if os.name == 'posix':
    # primitive formatting on supported
    # terminal via ANSI escape sequences:
    BOLD = ('\033[0m', '\033[1m')
    BLUE = ('\033[0m', '\033[0;34m')
    RED = ('\033[0m', '\033[0;31m')
    GREY = ('\033[0m', '\033[1;30m')

TEST_EXIT_PASSED = 0
TEST_EXIT_SKIPPED = 77

NON_SCRIPTS = [
    # These are python files that live in the functional tests directory, but are not test scripts.
    "combine_logs.py",
    "create_cache.py",
    "test_runner.py",
]

TEST_PARAMS = {
    # Some test can be run with additional parameters.
    # When a test is listed here, the it  will be run without parameters
    # as well as with additional parameters listed here.
    # This:
    #    example "testName" : [["--param1", "--param2"] , ["--param3"]]
    # will run the test 3 times:
    #    testName
    #    testName --param1 --param2
    #    testname --param3
    "txn_doublespend.py": [["--mineblock"]],
    "txn_clone.py": [["--mineblock"]]
}

# Used to limit the number of tests, when list of tests is not provided on command line
# When --extended is specified, we run all tests, otherwise
# we only run a test if its execution time in seconds does not exceed EXTENDED_CUTOFF
EXTENDED_CUTOFF = 40


def on_ci():
    return os.getenv('TRAVIS') == 'true' or os.getenv('TEAMCITY_VERSION') != None


def main():
    # Read config generated by configure.
    config = configparser.ConfigParser()
    configfile = os.path.join(os.path.abspath(
        os.path.dirname(__file__)), "config.ini")
    config.read_file(open(configfile))

    src_dir = config["environment"]["SRCDIR"]
    build_dir = config["environment"]["BUILDDIR"]
    tests_dir = os.path.join(src_dir,"")

    # Parse arguments and pass through unrecognised args
    parser = argparse.ArgumentParser(add_help=False,
                                     usage='%(prog)s [test_runner.py options] [script options] [scripts]',
                                     description=__doc__,
                                     epilog='''
    Help text and arguments for individual test script:''',
                                     formatter_class=argparse.RawTextHelpFormatter)
    parser.add_argument('--coverage', action='store_true',
                        help='generate a basic coverage report for the RPC interface')
    parser.add_argument(
        '--exclude', '-x', help='specify a comma-seperated-list of scripts to exclude. Do not include the .py extension in the name.')
    parser.add_argument('--extended', action='store_true',
                        help='run the extended test suite in addition to the basic tests')
    parser.add_argument('--force', '-f', action='store_true',
                        help='run tests even on platforms where they are disabled by default (e.g. windows).')
    parser.add_argument('--help', '-h', '-?',
                        action='store_true', help='print help text and exit')
    parser.add_argument('--jobs', '-j', type=int, default=4,
                        help='how many test scripts to run in parallel. Default=4.')
    parser.add_argument('--keepcache', '-k', action='store_true',
                        help='the default behavior is to flush the cache directory on startup. --keepcache retains the cache from the previous testrun.')
    parser.add_argument('--quiet', '-q', action='store_true',
                        help='only print results summary and failure logs')
    parser.add_argument('--tmpdirprefix', '-t',
                        default=tempfile.gettempdir(), help="Root directory for datadirs")
    parser.add_argument('--junitouput', '-ju',
                        default=os.path.join(build_dir, 'junit_results.xml'), help="file that will store JUnit formated test results.")

    args, unknown_args = parser.parse_known_args()

    # Create a set to store arguments and create the passon string
    tests = set(arg for arg in unknown_args if arg[:2] != "--")
    passon_args = [arg for arg in unknown_args if arg[:2] == "--"]
    passon_args.append("--configfile=%s" % configfile)

    # Set up logging
    logging_level = logging.INFO if args.quiet else logging.DEBUG
    logging.basicConfig(format='%(message)s', level=logging_level)

    # Create base test directory
    tmpdir = os.path.join("%s", "bitcoin_test_runner_%s") % (
        args.tmpdirprefix, datetime.datetime.now().strftime("%Y%m%d_%H%M%S"))
    os.makedirs(tmpdir)

    logging.debug("Temporary test directory at %s" % tmpdir)

    enable_wallet = config["components"].getboolean("ENABLE_WALLET")
    enable_utils = config["components"].getboolean("ENABLE_UTILS")
    enable_bitcoind = config["components"].getboolean("ENABLE_BITCOIND")

    if config["environment"]["EXEEXT"] == ".exe" and not args.force:
        # https://github.com/bitcoin/bitcoin/commit/d52802551752140cf41f0d9a225a43e84404d3e9
        # https://github.com/bitcoin/bitcoin/pull/5677#issuecomment-136646964
        print(
            "Tests currently disabled on Windows by default. Use --force option to enable")
        sys.exit(0)

    if not (enable_wallet and enable_utils and enable_bitcoind):
        print(
            "No functional tests to run. Wallet, utils, and bitcoind must all be enabled")
        print(
            "Rerun `configure` with -enable-wallet, -with-utils and -with-daemon and rerun make")
        sys.exit(0)

    # Build list of tests
    all_scripts = get_all_scripts_from_disk(tests_dir, NON_SCRIPTS)

    if tests:
        # Individual tests have been specified. Run specified tests that exist
        # in the all_scripts list. Accept the name with or without .py
        # extension.
        test_list = [t for t in all_scripts if
                     (t in tests or re.sub(".py$", "", t) in tests)]
        cutoff = sys.maxsize  # do not cut off explicitly specified tests
    else:
        # No individual tests have been specified.
        # Run all tests that do not exceed
        # EXTENDED_CUTOFF, unless --extended was specified
        test_list = all_scripts
        cutoff = EXTENDED_CUTOFF
        if args.extended:
            cutoff = sys.maxsize

    # Remove the test cases that the user has explicitly asked to exclude.
    if args.exclude:
        for exclude_test in args.exclude.split(','):
            if exclude_test + ".py" in test_list:
                test_list.remove(exclude_test + ".py")

    # Use and update timings from build_dir only if separate
    # build directory is used. We do not want to pollute source directory.
    build_timings = None
    if (src_dir != build_dir):
        build_timings = Timings(os.path.join(build_dir, 'timing.json'))

    # Always use timings from scr_dir if present
    src_timings = Timings(os.path.join(
        src_dir, "test", "functional", 'timing.json'))

    # Add test parameters and remove long running tests if needed
    test_list = get_tests_to_run(
        test_list, TEST_PARAMS, cutoff, src_timings, build_timings)

    if not test_list:
        print("No valid test scripts specified. Check that your test is in one "
              "of the test lists in test_runner.py, or run test_runner.py with no arguments to run all tests")
        sys.exit(0)

    if args.help:
        # Print help for test_runner.py, then print help of the first script
        # and exit.
        parser.print_help()
        subprocess.check_call(
            [os.path.join(tests_dir, test_list[0]), '-h'])
        sys.exit(0)

    if not args.keepcache:
        shutil.rmtree(os.path.join(build_dir, "test",
                                   "cache"), ignore_errors=True)

    run_tests(test_list, build_dir, tests_dir, args.junitouput,
              config["environment"]["EXEEXT"], tmpdir, args.jobs, args.coverage, passon_args, build_timings)


def run_tests(test_list, build_dir, tests_dir, junitouput, exeext, tmpdir, jobs=1, enable_coverage=False, args=[], build_timings=None):
    # Warn if bitcoind is already running (unix only)
    try:
        pidofOutput = subprocess.check_output(["pidof", "bitcoind"])
        if pidofOutput is not None and pidofOutput != b'':
            print("%sWARNING!%s There is already a bitcoind process running on this system. Tests may fail unexpectedly due to resource contention!" % (
                BOLD[1], BOLD[0]))
    except (OSError, subprocess.SubprocessError):
        pass

    # Warn if there is a cache directory
    cache_dir = os.path.join(build_dir, "test", "cache")
    if os.path.isdir(cache_dir):
        print("%sWARNING!%s There is a cache directory here: %s. If tests fail unexpectedly, try deleting the cache directory." % (
            BOLD[1], BOLD[0], cache_dir))

    # Set env vars
    if "BITCOIND" not in os.environ:
        os.environ["BITCOIND"] = os.path.join(
            build_dir, 'src', 'bitcoind' + exeext)
        os.environ["BITCOINCLI"] = os.path.join(
            build_dir, 'src', 'bitcoin-cli' + exeext)

    flags = [os.path.join("--srcdir={}".format(build_dir), "src")] + args
    flags.append("--cachedir=%s" % cache_dir)

    if enable_coverage:
        coverage = RPCCoverage()
        flags.append(coverage.flag)
        logging.debug("Initializing coverage directory at %s" % coverage.dir)
    else:
        coverage = None

    if len(test_list) > 1 and jobs > 1:
        # Populate cache
        cache_tmpdir = os.path.join("--tmpdir=%s", "cache") % tmpdir
        create_cache_py = [os.path.join(tests_dir, 'create_cache.py')] + flags + [cache_tmpdir]
        try:
            subprocess.check_output(create_cache_py)
        except Exception:
            try:
                shutil.rmtree(cache_tmpdir, ignore_errors=True)
                subprocess.check_output(create_cache_py)
            except Exception:
                shutil.rmtree(cache_tmpdir, ignore_errors=True)
                subprocess.check_output(create_cache_py)

    # Run Tests
    job_queue = TestHandler(jobs, tests_dir, tmpdir, test_list, flags)
    time0 = time.time()
    test_results = []

    max_len_name = len(max(test_list, key=len))

    for _ in range(len(test_list)):
        test_result = job_queue.get_next()
        test_results.append(test_result)

        if test_result.status == "Passed":
            logging.debug("\n%s%s%s passed, Duration: %s s" % (
                BOLD[1], test_result.name, BOLD[0], test_result.time))
        elif test_result.status == "Skipped":
            logging.debug("\n%s%s%s skipped" %
                          (BOLD[1], test_result.name, BOLD[0]))
        else:
            print("\n%s%s%s failed, Duration: %s s\n" %
                  (BOLD[1], test_result.name, BOLD[0], test_result.time))
            print(BOLD[1] + 'stdout:\n' + BOLD[0] + test_result.stdout + '\n')
            print(BOLD[1] + 'stderr:\n' + BOLD[0] + test_result.stderr + '\n')

    runtime = int(time.time() - time0)
    print_results(test_results, max_len_name, runtime)
    save_results_as_junit(test_results, junitouput, runtime)

    if (build_timings is not None):
        build_timings.save_timings(test_results)

    if coverage:
        coverage.report_rpc_coverage()

        logging.debug("Cleaning up coverage data")
        coverage.cleanup()

    # Clear up the temp directory if all subdirectories are gone
    if not os.listdir(tmpdir):
        os.rmdir(tmpdir)

    all_passed = all(
        map(lambda test_result: test_result.status == "Passed", test_results))

    sys.exit(not all_passed)


def print_results(test_results, max_len_name, runtime):
    results = "\n" + BOLD[1] + "%s | %s | %s\n\n" % (
        "TEST".ljust(max_len_name), "STATUS   ", "DURATION") + BOLD[0]

    test_results.sort(key=lambda result: result.name.lower())
    all_passed = True
    time_sum = 0

    for test_result in test_results:
        all_passed = all_passed and test_result.status != "Failed"
        time_sum += test_result.time
        test_result.padding = max_len_name
        results += str(test_result)

    status = TICK + "Passed" if all_passed else CROSS + "Failed"
    results += BOLD[1] + "\n%s | %s | %s s (accumulated) \n" % (
        "ALL".ljust(max_len_name), status.ljust(9), time_sum) + BOLD[0]
    results += "Runtime: %s s\n" % (runtime)
    print(results)


class TestHandler:

    """
    Trigger the testscrips passed in via the list.
    """

    def __init__(self, num_tests_parallel, tests_dir, tmpdir, test_list=None, flags=None):
        assert(num_tests_parallel >= 1)
        self.num_jobs = num_tests_parallel
        self.tests_dir = tests_dir
        self.tmpdir = tmpdir
        self.test_list = test_list
        self.flags = flags
        self.num_running = 0
        # In case there is a graveyard of zombie bitcoinds, we can apply a
        # pseudorandom offset to hopefully jump over them.
        # (625 is PORT_RANGE/MAX_NODES)
        self.portseed_offset = int(time.time() * 1000) % 625
        self.jobs = []

    def get_next(self):
        while self.num_running < self.num_jobs and self.test_list:
            # Add tests
            self.num_running += 1
            t = self.test_list.pop(0)
            portseed = len(self.test_list) + self.portseed_offset
            portseed_arg = ["--portseed={}".format(portseed)]
            log_stdout = tempfile.SpooledTemporaryFile(max_size=2**16)
            log_stderr = tempfile.SpooledTemporaryFile(max_size=2**16)
            test_argv = t.split()
            tmpdir = [os.path.join("--tmpdir=%s", "%s_%s") %
                      (self.tmpdir, re.sub(".py$", "", t), portseed)]
            self.jobs.append((t,
                              time.time(),
                              subprocess.Popen([os.path.join(self.tests_dir, test_argv[0])] + test_argv[1:] + self.flags + portseed_arg + tmpdir,
                                               universal_newlines=True,
                                               stdout=log_stdout,
                                               stderr=log_stderr),
                              log_stdout,
                              log_stderr))
        if not self.jobs:
            raise IndexError('pop from empty list')
        while True:
            # Return first proc that finishes
            time.sleep(.5)
            for j in self.jobs:
                (name, time0, proc, log_out, log_err) = j
                if on_ci() and int(time.time() - time0) > 20 * 60:
                    # In travis, timeout individual tests after 20 minutes (to stop tests hanging and not
                    # providing useful output.
                    proc.send_signal(signal.SIGINT)
                if proc.poll() is not None:
                    log_out.seek(0), log_err.seek(0)
                    [stdout, stderr] = [l.read().decode('utf-8')
                                        for l in (log_out, log_err)]
                    log_out.close(), log_err.close()
                    if proc.returncode == TEST_EXIT_PASSED and stderr == "":
                        status = "Passed"
                    elif proc.returncode == TEST_EXIT_SKIPPED:
                        status = "Skipped"
                    else:
                        status = "Failed"
                    self.num_running -= 1
                    self.jobs.remove(j)

                    return TestResult(name, status, int(time.time() - time0), stdout, stderr)
            print('.', end='', flush=True)


class TestResult():
    def __init__(self, name, status, time, stdout, stderr):
        self.name = name
        self.status = status
        self.time = time
        self.padding = 0
        self.stdout = stdout
        self.stderr = stderr

    def __repr__(self):
        if self.status == "Passed":
            color = BLUE
            glyph = TICK
        elif self.status == "Failed":
            color = RED
            glyph = CROSS
        elif self.status == "Skipped":
            color = GREY
            glyph = CIRCLE

        return color[1] + "%s | %s%s | %s s\n" % (self.name.ljust(self.padding), glyph, self.status.ljust(7), self.time) + color[0]


def get_all_scripts_from_disk(test_dir, non_scripts):
    """
    Return all available test script from script directory (excluding NON_SCRIPTS)
    """
    python_files = set([t for t in os.listdir(test_dir) if t[-3:] == ".py"])
    return list(python_files - set(non_scripts))


def get_tests_to_run(test_list, test_params, cutoff, src_timings, build_timings=None):
    """
    Returns only test that will not run longer that cutoff.
    Long running tests are returned first to favor running tests in parallel
    Timings from build directory override those from src directory
    """

    def get_test_time(test):
        if build_timings is not None:
            timing = next(
                (x['time'] for x in build_timings.existing_timings if x['name'] == test), None)
            if timing is not None:
                return timing

        # try source directory. Return 0 if test is unknown to always run it
        return next(
            (x['time'] for x in src_timings.existing_timings if x['name'] == test), 0)

    # Some tests must also be run with additional parameters. Add them to the list.
    tests_with_params = []
    for test_name in test_list:
        # always execute a test without parameters
        tests_with_params.append(test_name)
        params = test_params.get(test_name)
        if params is not None:
            tests_with_params.extend(
                [test_name + " " + " ".join(p) for p in params])

    result = [t for t in tests_with_params if get_test_time(t) <= cutoff]
    result.sort(key=lambda x:  (-get_test_time(x), x))
    return result


class RPCCoverage():
    """
    Coverage reporting utilities for test_runner.

    Coverage calculation works by having each test script subprocess write
    coverage files into a particular directory. These files contain the RPC
    commands invoked during testing, as well as a complete listing of RPC
    commands per `bitcoin-cli help` (`rpc_interface.txt`).

    After all tests complete, the commands run are combined and diff'd against
    the complete list to calculate uncovered RPC commands.

    See also: test/functional/test_framework/coverage.py

    """

    def __init__(self):
        self.dir = tempfile.mkdtemp(prefix="coverage")
        self.flag = '--coveragedir={}'.format(self.dir)

    def report_rpc_coverage(self):
        """
        Print out RPC commands that were unexercised by tests.

        """
        uncovered = self._get_uncovered_rpc_commands()

        if uncovered:
            print("Uncovered RPC commands:")
            print("".join(("  - {}\n".format(i)) for i in sorted(uncovered)))
        else:
            print("All RPC commands covered.")

    def cleanup(self):
        return shutil.rmtree(self.dir)

    def _get_uncovered_rpc_commands(self):
        """
        Return a set of currently untested RPC commands.

        """
        # This is shared from `test/functional/test-framework/coverage.py`
        reference_filename = 'rpc_interface.txt'
        coverage_file_prefix = 'coverage.'

        coverage_ref_filename = os.path.join(self.dir, reference_filename)
        coverage_filenames = set()
        all_cmds = set()
        covered_cmds = set()

        if not os.path.isfile(coverage_ref_filename):
            raise RuntimeError("No coverage reference found")

        with open(coverage_ref_filename, 'r') as f:
            all_cmds.update([i.strip() for i in f.readlines()])

        for root, dirs, files in os.walk(self.dir):
            for filename in files:
                if filename.startswith(coverage_file_prefix):
                    coverage_filenames.add(os.path.join(root, filename))

        for filename in coverage_filenames:
            with open(filename, 'r') as f:
                covered_cmds.update([i.strip() for i in f.readlines()])

        return all_cmds - covered_cmds


def save_results_as_junit(test_results, file_name, time):
    """
    Save tests results to file in JUnit format

    See http://llg.cubic.org/docs/junit/ for specification of format
    """
    e_test_suite = ET.Element("testsuite",
                              {"name": "bitcoin_abc_tests",
                               "tests": str(len(test_results)),
                               #"errors":
                               "failures": str(len([t for t in test_results if t.status == "Failed"])),
                               "id": "0",
                               "skipped": str(len([t for t in test_results if t.status == "Skipped"])),
                               "time": str(time),
                               "timestamp": datetime.datetime.now().isoformat('T')
                               })

    for test_result in test_results:
        e_test_case = ET.SubElement(e_test_suite, "testcase",
                                    {"name": test_result.name,
                                     "classname": test_result.name,
                                     "time": str(test_result.time)
                                     }
                                    )
        if test_result.status == "Skipped":
            ET.SubElement(e_test_case, "skipped")
        elif test_result.status == "Failed":
            ET.SubElement(e_test_case, "failure")
        # no special element for passed tests

        ET.SubElement(e_test_case, "system-out").text = test_result.stdout
        ET.SubElement(e_test_case, "system-err").text = test_result.stderr

    ET.ElementTree(e_test_suite).write(
        file_name, "UTF-8", xml_declaration=True)


class Timings():
    """
    Takes care of loading, merging and saving tests execution times.
    """

    def __init__(self, timing_file):
        self.timing_file = timing_file
        self.existing_timings = self.load_timings()

    def load_timings(self):
        if os.path.isfile(self.timing_file):
            with open(self.timing_file) as f:
                return json.load(f)
        else:
            return []

    def get_merged_timings(self, new_timings):
        """
        Return new list containing existing timings updated with new timings
        Tests that do not exists are not removed
        """

        key = 'name'
        merged = {}
        for item in self.existing_timings + new_timings:
            if item[key] in merged:
                merged[item[key]].update(item)
            else:
                merged[item[key]] = item

        # Sort the result to preserve test ordering in file
        merged = list(merged.values())
        merged.sort(key=lambda t, key=key: t[key])
        return merged

    def save_timings(self, test_results):
        # we only save test that have passed - timings for failed test might be
        # wrong (timeouts or early fails)
        passed_results = [t for t in test_results if t.status == 'Passed']
        new_timings = list(map(lambda t: {'name':  t.name, 'time': t.time},
                               passed_results))
        merged_timings = self.get_merged_timings(new_timings)

        with open(self.timing_file, 'w') as f:
            json.dump(merged_timings, f, indent=True)


if __name__ == '__main__':
    main()