def Main():
options = ParseArguments()
logger = FileLogger(options.logfile) if options.logfile else StderrLogger()
database = Database(options.dsn, readonly=True, autocommit=True)
readqueue = multiprocessing.Queue(10)
writequeue = multiprocessing.Queue(10)
writer = multiprocessing.Process(target=LinkUpdatingWorker,
args=(writequeue, options, logger))
writer.start()
processpool = [
multiprocessing.Process(target=LinkProcessingWorker,
args=(readqueue, writequeue, i, options,
logger)) for i in range(options.jobs)
]
for process in processpool:
process.start()
# base logger already passed to workers, may append prefix here
logger = logger.GetPrefixed('master: ')
prev_url = None
while True:
# Get pack of links
logger.Log('Requesting pack of urls')
urls = database.GetLinksForCheck(
after=prev_url,
prefix=options.prefix, # no limit by default
limit=options.packsize,
recheck_age=options.age * 60 * 60 * 24,
unchecked_only=options.unchecked,
checked_only=options.checked,
failed_only=options.failed,
succeeded_only=options.succeeded)
if not urls:
logger.Log(' No more urls to process')
break
# Get another pack of urls with the last hostname to ensure
# that all urls for one hostname get into a same large pack
match = re.match('([a-z]+://[^/]+/)', urls[-1])
if match:
urls += database.GetLinksForCheck(after=urls[-1],
prefix=match.group(1),
recheck_age=options.age * 60 *
60 * 24,
unchecked_only=options.unchecked,
checked_only=options.checked,
failed_only=options.failed,
succeeded_only=options.succeeded)
# Process
if options.maxpacksize and len(urls) > options.maxpacksize:
logger.Log(
'Skipping {} urls ({}..{}), exceeds max pack size'.format(
len(urls), urls[0], urls[-1]))
else:
readqueue.put(urls)
logger.Log('Enqueued {} urls ({}..{})'.format(
len(urls), urls[0], urls[-1]))
prev_url = urls[-1]
logger.Log('Waiting for child processes to exit')
# close workers
for process in processpool:
readqueue.put(None)
for process in processpool:
process.join()
# close writer
writequeue.put(None)
writer.join()
logger.Log('Done')
return 0
def Main():
parser = argparse.ArgumentParser(
formatter_class=argparse.ArgumentDefaultsHelpFormatter)
parser.add_argument('--dsn',
default=repology.config.DSN,
help='database connection params')
parser.add_argument('--logfile',
help='path to log file (log to stderr by default)')
parser.add_argument('--timeout',
type=float,
default=60.0,
help='timeout for link requests in seconds')
parser.add_argument('--delay',
type=float,
default=3.0,
help='delay between requests to one host')
parser.add_argument('--age',
type=int,
default=365,
help='min age for recheck in days')
parser.add_argument('--packsize',
type=int,
default=128,
help='pack size for link processing')
parser.add_argument(
'--maxpacksize',
type=int,
help='max pack size for link processing (useful to skip large hosts)')
parser.add_argument('--jobs',
type=int,
default=1,
help='number of parallel jobs')
parser.add_argument('--unchecked',
action='store_true',
help='only process unchecked (newly discovered) links')
parser.add_argument('--checked',
action='store_true',
help='only process old (already checked) links')
parser.add_argument('--failed',
action='store_true',
help='only process links that were checked and failed')
parser.add_argument('--succeeded',
action='store_true',
help='only process links that were checked and failed')
parser.add_argument('--prefix',
help='only process links with specified prefix')
options = parser.parse_args()
logger = FileLogger(options.logfile) if options.logfile else StderrLogger()
database = Database(options.dsn, readonly=True, autocommit=True)
queue = multiprocessing.Queue(1)
processpool = [
multiprocessing.Process(target=LinkProcessorWorker,
args=(queue, i, options, logger))
for i in range(options.jobs)
]
for process in processpool:
process.start()
# base logger already passed to workers, may append prefix here
logger = logger.GetPrefixed('master: ')
prev_url = None
while True:
# Get pack of links
logger.Log('Requesting pack of urls'.format(prev_url))
urls = database.GetLinksForCheck(
after=prev_url,
prefix=options.prefix, # no limit by default
limit=options.packsize,
recheck_age=options.age * 60 * 60 * 24,
unchecked_only=options.unchecked,
checked_only=options.checked,
failed_only=options.failed,
succeeded_only=options.succeeded)
if not urls:
logger.Log(' No more urls to process')
break
# Get another pack of urls with the last hostname to ensure
# that all urls for one hostname get into a same large pack
match = re.match('([a-z]+://[^/]+/)', urls[-1])
if match:
urls += database.GetLinksForCheck(after=urls[-1],
prefix=match.group(1),
recheck_age=options.age * 60 *
60 * 24,
unchecked_only=options.unchecked,
checked_only=options.checked,
failed_only=options.failed,
succeeded_only=options.succeeded)
# Process
if options.maxpacksize and len(urls) > options.maxpacksize:
logger.Log(
'Skipping {} urls ({}..{}), exceeds max pack size'.format(
len(urls), urls[0], urls[-1]))
else:
queue.put(urls)
logger.Log('Enqueued {} urls ({}..{})'.format(
len(urls), urls[0], urls[-1]))
prev_url = urls[-1]
logger.Log('Waiting for child processes to exit')
for process in processpool:
queue.put(None)
for process in processpool:
process.join()
logger.Log('Done')
return 0
