def __get_schedule(schedule_string):
if schedule_string is not None:
start_time, end_time = _TimeParser.parse_time_interval(
schedule_string)
return DaySchedule(start_time, end_time)
else:
return AlwaysActiveSchedule()
def __init__(self,
navigators,
sentinel,
activity_schedule=None,
log_file_path=None,
state_file_path=None,
save_period=None,
logging_level=logging.ERROR):
"""
@param navigators: list of navigators to be used by the crawler.
Each navigator will be run in a separate thread, thus the
number of the threads is equal to the number of navigators.
@type navigators: list of L{AbstractTreeNavigator}s
@param sentinel: a technical node which will be made parent of the
root node.
@type sentinel: L{AbstractNode}
@param activity_schedule: if C{None}, no schedule is used and the
program works until it finishes crawling.
@type activity_schedule: L{AbstractActivitySchedule}
@param log_file_path: path to the log file. If C{None}, no log file
will be used.
@param state_file_path: path to the file where the state of the
program will be saved. If C{None}, the state will not be saved.
@param save_period: time between saving the tree state. If
C{state_file_path} is C{None}, this value is ignored.
@param logging_level: one of the logging level constants from C{logging}
"""
if log_file_path is not None:
lenient_makedir(os.path.dirname(log_file_path))
if state_file_path is not None:
if os.path.exists(state_file_path):
print "State file already exists. Loading the tree from this "\
"file and changing nodes with state PROCESSING to OPEN ... ",
self.__load_state_file(state_file_path, sentinel)
print "Done."
else:
lenient_makedir(os.path.dirname(state_file_path))
self.__tree = RWLockTreeAccessor(sentinel)
self.__navigators = navigators
self.__manager = None
self.__state_file_path = state_file_path
self.__save_period = save_period
self.__activity_schedule = activity_schedule
if activity_schedule is None:
self.__activity_schedule = AlwaysActiveSchedule()
self.__logging_level = logging_level
self.__log_file_path = log_file_path
