backup/warcprox
1
0
Fork 0
mirror of https://github.com/internetarchive/warcprox.git synced 2025-01-18 13:22:09 +01:00
Code Issues Projects Releases Activity

new option --plugin

Browse Source
This commit is contained in:
Noah Levitt 2017-06-28 12:53:34 -07:00
parent e31302a6e3
commit 4c32394256
1 changed files with 30 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

33
warcprox/main.py
View File

@ -42,6 +42,7 @@ import warcprox
import re
import doublethink
import cryptography.hazmat.backends.openssl
import importlib
class BetterArgumentDefaultsHelpFormatter(
argparse.ArgumentDefaultsHelpFormatter,
@ -125,8 +126,21 @@ def _build_arg_parser(prog=os.path.basename(sys.argv[0])):
help=argparse.SUPPRESS)
arg_parser.add_argument('--profile', action='store_true', default=False,
help=argparse.SUPPRESS)
arg_parser.add_argument('--onion-tor-socks-proxy', dest='onion_tor_socks_proxy',
default=None, help='host:port of tor socks proxy, used only to connect to .onion sites')
arg_parser.add_argument(
'--onion-tor-socks-proxy', dest='onion_tor_socks_proxy',
default=None, help=(
'host:port of tor socks proxy, used only to connect to '
'.onion sites'))
arg_parser.add_argument(
'--plugin', metavar='PLUGIN_CLASS', dest='plugins',
action='append', help=(
'Qualified name of plugin class, e.g. "mypkg.mymod.MyClass". '
'May be used multiple times to register multiple plugins. '
'Plugin classes are loaded from the regular python module '
'search path. They will be instantiated with no arguments and '
'must have a method `notify(self, recorded_url, records)` '
'which will be called for each url, after warc records have '
'been written.'))
arg_parser.add_argument('--version', action='version',
version="warcprox {}".format(warcprox.__version__))
arg_parser.add_argument('-v', '--verbose', dest='verbose', action='store_true')
@ -210,7 +224,8 @@ def init_controller(args):
stats_db=stats_db, options=options)
if args.playback_port is not None:
playback_index_db = warcprox.playback.PlaybackIndexDb(args.playback_index_db_file, options=options)
playback_index_db = warcprox.playback.PlaybackIndexDb(
args.playback_index_db_file, options=options)
playback_proxy = warcprox.playback.PlaybackProxy(
server_address=(args.address, args.playback_port), ca=ca,
playback_index_db=playback_index_db, warcs_dir=args.directory,
@ -220,6 +235,18 @@ def init_controller(args):
playback_index_db = None
playback_proxy = None
for qualname in args.plugins or []:
try:
(module_name, class_name) = qualname.rsplit('.', 1)
module_ = importlib.import_module(module_name)
class_ = getattr(module_, class_name)
listener = class_()
listener.notify # make sure it has this method
listeners += listener
except Exception as e:
logging.fatal('problem with plugin class %r: %s', qualname, e)
sys.exit(1)
writer_pool = warcprox.writer.WarcWriterPool(options=options)
# number of warc writer threads = sqrt(proxy.max_threads)
# I came up with this out of thin air because it strikes me as reasonable