DXR is a code search and navigation tool aimed at making sense of large projects. It supports full-text and regex searches as well as structural queries.

Mercurial (c68fe15a81fc)

VCS Links

Line Code
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250
#!/usr/bin/env python
"""
Usage: extract_and_run_command.py [-j N] [command to run] -- [files and/or directories]
Usage: extract_and_run_command.py [-j N] [command to run] -- [files and/or directories]
    -j is the number of workers to start, defaulting to 1.
    [command to run] must be a command that can accept one or many files
    to process as arguments.

WARNING: This script does NOT respond to SIGINT. You must use SIGQUIT or SIGKILL to
WARNING: This script does NOT respond to SIGINT. You must use SIGQUIT or SIGKILL to
         terminate it early.
"""

### The canonical location for this file is
###   https://hg.mozilla.org/build/tools/file/default/stage/extract_and_run_command.py
###   https://hg.mozilla.org/build/tools/file/default/stage/extract_and_run_command.py
###
### Please update the copy in puppet to deploy new changes to
### stage.mozilla.org, see
# https://wiki.mozilla.org/ReleaseEngineering/How_To/Modify_scripts_on_stage


import logging
import os
import shutil
import subprocess
import sys
import sys
import tempfile
import time
from os import path
from threading import Thread


try:
    from Queue import Queue
except ImportError:
    from queue import Queue
    from queue import Queue

logging.basicConfig(
    stream=sys.stdout, level=logging.INFO, format="%(message)s")
log = logging.getLogger(__name__)


try:
    # the future - https://github.com/mozilla/build-mar via a venv
    from mardor.marfile import BZ2MarFile
except:
    # the past - http://hg.mozilla.org/build/tools/file/default/buildfarm/utils/mar.py
    sys.path.append(
        path.join(path.dirname(path.realpath(__file__)), "../buildfarm/utils"))
    from mar import BZ2MarFile


SEVENZIP = "7za"


def extractMar(filename, tempdir):
def extractMar(filename, tempdir):
    m = BZ2MarFile(filename)
    m.extractall(path=tempdir)


def extractExe(filename, tempdir):
def extractExe(filename, tempdir):
    try:
        # We don't actually care about output, put we redirect to a tempfile
        # to avoid deadlocking in wait() when stdout=PIPE
        fd = tempfile.TemporaryFile()
        proc = subprocess.Popen([SEVENZIP, 'x', '-o%s' % tempdir, filename],
                                stdout=fd, stderr=subprocess.STDOUT)
        proc.wait()
        proc.wait()
    except subprocess.CalledProcessError:
        # Not all EXEs are 7-zip files, so we have to ignore extraction errors
        pass


# The keys here are matched against the last 3 characters of filenames.
# The values are callables that accept two string arguments.
EXTRACTORS = {
    '.mar': extractMar,
    '.exe': extractExe,
}



def find_files(d):
    """yields all of the files in `d'"""
    for root, _, files in os.walk(d):
        for f in files:
        for f in files:
            yield path.abspath(path.join(root, f))


def rchmod(d, mode=0o755):
    """chmods everything in `d' to `mode', including `d' itself"""
    os.chmod(d, mode)
    for root, dirs, files in os.walk(d):
        for item in dirs:
        for item in dirs:
            os.chmod(path.join(root, item), mode)
        for item in files:
            os.chmod(path.join(root, item), mode)



def maybe_extract(filename):
    """If an extractor is found for `filename', extracts it to a temporary
       directory and chmods it. The consumer is responsible for removing
       the extracted files, if desired."""
    ext = path.splitext(filename)[1]
    ext = path.splitext(filename)[1]
    if ext not in EXTRACTORS.keys():
        return None
    # Append the full filepath to the tempdir
    tempdir_root = tempfile.mkdtemp()
    tempdir_root = tempfile.mkdtemp()
    tempdir = path.join(tempdir_root, filename.lstrip('/'))
    os.makedirs(tempdir)
    EXTRACTORS[ext](filename, tempdir)
    rchmod(tempdir_root)
    return tempdir_root
    return tempdir_root


def process(item, command):
    def format_time(t):
        return time.strftime("%H:%M:%S", time.localtime(t))
        return time.strftime("%H:%M:%S", time.localtime(t))
    # Buffer output to avoid interleaving of multiple workers'
    logs = []
    args = [item]
    proc = None
    start = time.time()
    start = time.time()
    logs.append("START %s: %s" % (format_time(start), item))
    # If the file was extracted, we need to process all of its files, too.
    tempdir = maybe_extract(item)
    if tempdir:
    if tempdir:
        for f in find_files(tempdir):
            args.append(f)

    try:
        fd = tempfile.TemporaryFile()
        fd = tempfile.TemporaryFile()
        proc = subprocess.Popen(command + args, stdout=fd)
        proc.wait()
        if proc.returncode != 0:
            raise Exception("returned %s" % proc.returncode)
    finally:
    finally:
        if tempdir:
            shutil.rmtree(tempdir)
        fd.seek(0)
        fd.seek(0)
        # rstrip() here to avoid an unnecessary newline, if it exists.
        logs.append(fd.read().rstrip())
        end = time.time()
        elapsed = end - start
        logs.append("END %s (%d seconds elapsed): %s\n" % (
        logs.append("END %s (%d seconds elapsed): %s\n" % (
            format_time(end), elapsed, item))
        # Now that we've got all of our output, print it. It's important that
        # the logging module is used for this, because "print" is not
        # thread-safe.
        log.info("\n".join(logs))
        log.info("\n".join(logs))


def worker(command, errors):
    item = q.get()
    item = q.get()
    while item != None:
        try:
            process(item, command)
        except:
            errors.put(item)
        item = q.get()
        item = q.get()

if __name__ == '__main__':
    # getopt is used in favour of optparse to enable "--" as a separator
    # between the command and list of files. optparse doesn't allow that.
    from getopt import getopt
    from getopt import getopt
    options, args = getopt(sys.argv[1:], 'j:h', ['help'])

    concurrency = 1
    for o, a in options:
        if o == '-j':
        if o == '-j':
            concurrency = int(a)
        elif o in ('-h', '--help'):
            log.info(__doc__)
            log.info(__doc__)
            sys.exit(0)

    if len(args) < 3 or '--' not in args:
        log.error(__doc__)
        sys.exit(1)
        sys.exit(1)

    command = []
    while args[0] != "--":
        command.append(args.pop(0))
    args.pop(0)
    args.pop(0)

    q = Queue()
    errors = Queue()
    threads = []
    for i in range(concurrency):
    for i in range(concurrency):
        t = Thread(target=worker, args=(command, errors))
        t.start()
        threads.append(t)

    # find_files is a generator, so work will begin prior to it finding
    # find_files is a generator, so work will begin prior to it finding
    # all of the files
    for arg in args:
        if path.isfile(arg):
            q.put(arg)
        else:
        else:
            for f in find_files(arg):
                q.put(f)
    # Because the workers are started before we start populating the q
    # they can't use .empty() to determine whether or not their done.
    # We also can't use q.join() or j.task_done(), because we need to
    # We also can't use q.join() or j.task_done(), because we need to
    # support Python 2.4. We know that find_files won't yield None,
    # so we can detect doneness by having workers die when they get None
    # as an item.
    for i in range(concurrency):
        q.put(None)


    for t in threads:
        t.join()

    if not errors.empty():
        log.error("Command failed for the following files:")
        log.error("Command failed for the following files:")
        while not errors.empty():
            log.error("  %s" % errors.get())
        sys.exit(1)