def main():
config_file = CONFIG_FILE
offline = 0
verbose = 0
for arg in sys.argv[1:]:
if arg == "-h" or arg == "--help":
print "Usage: planet [options] [CONFIGFILE]"
print
print "Options:"
print " -v, --verbose DEBUG level logging during update"
print " -o, --offline Update the Planet from the cache only"
print " -h, --help Display this help message and exit"
print
sys.exit(0)
elif arg == "-v" or arg == "--verbose":
verbose = 1
elif arg == "-o" or arg == "--offline":
offline = 1
elif arg.startswith("-"):
print >>sys.stderr, "Unknown option:", arg
sys.exit(1)
else:
config_file = arg
# Read the configuration file
config = ConfigParser()
config.read(config_file)
if not config.has_section("Planet"):
print >>sys.stderr, "Configuration missing [Planet] section."
sys.exit(1)
# Read the [Planet] config section
planet_name = config_get(config, "Planet", "name", PLANET_NAME)
planet_link = config_get(config, "Planet", "link", PLANET_LINK)
planet_feed = config_get(config, "Planet", "feed", PLANET_FEED)
owner_name = config_get(config, "Planet", "owner_name", OWNER_NAME)
owner_email = config_get(config, "Planet", "owner_email", OWNER_EMAIL)
if verbose:
log_level = "DEBUG"
else:
log_level = config_get(config, "Planet", "log_level", LOG_LEVEL)
feed_timeout = config_get(config, "Planet", "feed_timeout", FEED_TIMEOUT)
template_files = config_get(config, "Planet", "template_files",
TEMPLATE_FILES).split(" ")
# Default feed to the first feed for which there is a template
if not planet_feed:
for template_file in template_files:
name = os.path.splitext(os.path.basename(template_file))[0]
if name.find('atom')>=0 or name.find('rss')>=0:
planet_feed = urlparse.urljoin(planet_link, name)
break
# Define locale
if config.has_option("Planet", "locale"):
# The user can specify more than one locale (separated by ":") as
# fallbacks.
locale_ok = False
for user_locale in config.get("Planet", "locale").split(':'):
user_locale = user_locale.strip()
try:
locale.setlocale(locale.LC_ALL, user_locale)
except locale.Error:
pass
else:
locale_ok = True
break
if not locale_ok:
print >>sys.stderr, "Unsupported locale setting."
sys.exit(1)
# Activate logging
planet.logging.basicConfig()
planet.logging.getLogger().setLevel(planet.logging.getLevelName(log_level))
log = planet.logging.getLogger("planet.runner")
try:
log.warning
except:
log.warning = log.warn
# timeoutsocket allows feedparser to time out rather than hang forever on
# ultra-slow servers. Python 2.3 now has this functionality available in
# the standard socket library, so under 2.3 you don't need to install
# anything. But you probably should anyway, because the socket module is
# buggy and timeoutsocket is better.
if feed_timeout:
try:
feed_timeout = float(feed_timeout)
except:
log.warning("Feed timeout set to invalid value '%s', skipping", feed_timeout)
feed_timeout = None
if feed_timeout and not offline:
try:
from planet import timeoutsocket
timeoutsocket.setDefaultSocketTimeout(feed_timeout)
log.debug("Socket timeout set to %d seconds", feed_timeout)
except ImportError:
import socket
if hasattr(socket, 'setdefaulttimeout'):
log.debug("timeoutsocket not found, using python function")
socket.setdefaulttimeout(feed_timeout)
log.debug("Socket timeout set to %d seconds", feed_timeout)
else:
log.error("Unable to set timeout to %d seconds", feed_timeout)
# run the planet
my_planet = planet.Planet(config)
my_planet.run(planet_name, planet_link, template_files, offline)
my_planet.generate_all_files(template_files, planet_name,
planet_link, planet_feed, owner_name, owner_email)
def spiderPlanet(only_if_new = False):
""" Spider (fetch) an entire planet """
# log = planet.getLogger(config.log_level(),config.log_format())
log = planet.getLogger(config.log_level(),config.log_format())
global index
index = True
timeout = config.feed_timeout()
try:
socket.setdefaulttimeout(float(timeout))
log.info("Socket timeout set to %d seconds", timeout)
except:
try:
from planet import timeoutsocket
timeoutsocket.setDefaultSocketTimeout(float(timeout))
log.info("Socket timeout set to %d seconds", timeout)
except:
log.warning("Timeout set to invalid value '%s', skipping", timeout)
from Queue import Queue
from threading import Thread
fetch_queue = Queue()
parse_queue = Queue()
threads = {}
http_cache = config.http_cache_directory()
# Should this be done in config?
if http_cache and not os.path.exists(http_cache):
os.makedirs(http_cache)
if int(config.spider_threads()):
# Start all the worker threads
for i in range(int(config.spider_threads())):
threads[i] = Thread(target=httpThread,
args=(i,fetch_queue, parse_queue, log))
threads[i].start()
else:
log.info("Building work queue")
# Load the fetch and parse work queues
for uri in config.subscriptions():
# read cached feed info
sources = config.cache_sources_directory()
feed_source = filename(sources, uri)
feed_info = feedparser.parse(feed_source)
if feed_info.feed and only_if_new:
log.info("Feed %s already in cache", uri)
continue
if feed_info.feed.get('planet_http_status',None) == '410':
log.info("Feed %s gone", uri)
continue
if threads and _is_http_uri(uri):
fetch_queue.put(item=(uri, feed_info))
else:
parse_queue.put(item=(uri, feed_info, uri))
# Mark the end of the fetch queue
for thread in threads.keys():
fetch_queue.put(item=(None, None))
# Process the results as they arrive
while fetch_queue.qsize() or parse_queue.qsize() or threads:
while parse_queue.qsize() == 0 and threads:
time.sleep(0.1)
while parse_queue.qsize():
(uri, feed_info, feed) = parse_queue.get(False)
try:
if not hasattr(feed,'headers') or int(feed.headers.status)<300:
options = {}
if hasattr(feed_info,'feed'):
options['etag'] = \
feed_info.feed.get('planet_http_etag',None)
try:
modified=time.strptime(
feed_info.feed.get('planet_http_last_modified',
None))
except:
pass
data = feedparser.parse(feed, **options)
else:
data = feedparser.FeedParserDict({'version': None,
'headers': feed.headers, 'entries': [], 'feed': {},
'bozo': 0, 'status': int(feed.headers.status)})
writeCache(uri, feed_info, data)
except Exception, e:
import sys, traceback
type, value, tb = sys.exc_info()
log.error('Error processing %s', uri)
for line in (traceback.format_exception_only(type, value) +
traceback.format_tb(tb)):
log.error(line.rstrip())
for index in threads.keys():
if not threads[index].isAlive():
del threads[index]
if not threads:
log.info("Finished threaded part of processing.")
def main():
config_file = CONFIG_FILE
offline = 0
verbose = 0
for arg in sys.argv[1:]:
if arg == "-h" or arg == "--help":
print "Usage: planet [options] [CONFIGFILE]"
print
print "Options:"
print " -v, --verbose DEBUG level logging during update"
print " -o, --offline Update the Planet from the cache only"
print " -h, --help Display this help message and exit"
print
sys.exit(0)
elif arg == "-v" or arg == "--verbose":
verbose = 1
elif arg == "-o" or arg == "--offline":
offline = 1
elif arg.startswith("-"):
print >> sys.stderr, "Unknown option:", arg
sys.exit(1)
else:
config_file = arg
# Read the configuration file
config = ConfigParser()
config.read(config_file)
if not config.has_section("Planet"):
print >> sys.stderr, "Configuration missing [Planet] section."
sys.exit(1)
# Read the [Planet] config section
planet_name = config_get(config, "Planet", "name", PLANET_NAME)
planet_link = config_get(config, "Planet", "link", PLANET_LINK)
planet_feed = config_get(config, "Planet", "feed", PLANET_FEED)
owner_name = config_get(config, "Planet", "owner_name", OWNER_NAME)
owner_email = config_get(config, "Planet", "owner_email", OWNER_EMAIL)
if verbose:
log_level = "DEBUG"
else:
log_level = config_get(config, "Planet", "log_level", LOG_LEVEL)
feed_timeout = config_get(config, "Planet", "feed_timeout", FEED_TIMEOUT)
template_files = config_get(config, "Planet", "template_files",
TEMPLATE_FILES).split(" ")
# Default feed to the first feed for which there is a template
if not planet_feed:
for template_file in template_files:
name = os.path.splitext(os.path.basename(template_file))[0]
if name.find('atom') >= 0 or name.find('rss') >= 0:
planet_feed = urlparse.urljoin(planet_link, name)
break
# Define locale
if config.has_option("Planet", "locale"):
# The user can specify more than one locale (separated by ":") as
# fallbacks.
locale_ok = False
for user_locale in config.get("Planet", "locale").split(':'):
user_locale = user_locale.strip()
try:
locale.setlocale(locale.LC_ALL, user_locale)
except locale.Error:
pass
else:
locale_ok = True
break
if not locale_ok:
print >> sys.stderr, "Unsupported locale setting."
sys.exit(1)
# Activate logging
planet.logging.basicConfig()
planet.logging.getLogger().setLevel(planet.logging.getLevelName(log_level))
log = planet.logging.getLogger("planet.runner")
try:
log.warning
except:
log.warning = log.warn
# timeoutsocket allows feedparser to time out rather than hang forever on
# ultra-slow servers. Python 2.3 now has this functionality available in
# the standard socket library, so under 2.3 you don't need to install
# anything. But you probably should anyway, because the socket module is
# buggy and timeoutsocket is better.
if feed_timeout:
try:
feed_timeout = float(feed_timeout)
except:
log.warning("Feed timeout set to invalid value '%s', skipping",
feed_timeout)
feed_timeout = None
if feed_timeout and not offline:
try:
from planet import timeoutsocket
timeoutsocket.setDefaultSocketTimeout(feed_timeout)
log.debug("Socket timeout set to %d seconds", feed_timeout)
except ImportError:
import socket
if hasattr(socket, 'setdefaulttimeout'):
log.debug("timeoutsocket not found, using python function")
socket.setdefaulttimeout(feed_timeout)
log.debug("Socket timeout set to %d seconds", feed_timeout)
else:
log.error("Unable to set timeout to %d seconds", feed_timeout)
# run the planet
my_planet = planet.Planet(config)
my_planet.run(planet_name, planet_link, template_files, offline)
my_planet.generate_all_files(template_files, planet_name, planet_link,
planet_feed, owner_name, owner_email)
def run_planet(config_file=None,
PLANET_NAME="Unconfigured Planet",
PLANET_LINK="Unconfigured Planet",
PLANET_FEED=None,
OWNER_NAME="Anonymous Coward",
OWNER_EMAIL="",
verbose=0,
offline=0,
LOG_LEVEL="WARNING",
FEED_TIMEOUT=20,
TEMPLATE_FILES="examples/basic/planet.html.tmpl"):
# Read the configuration file
config = ConfigParser()
config.read(config_file)
assert config.has_section("Planet"), \
"Configuration missing [Planet] section."
# Read the [Planet] config section
planet_name = config_get(config, "Planet", "name", PLANET_NAME)
planet_link = config_get(config, "Planet", "link", PLANET_LINK)
planet_feed = config_get(config, "Planet", "feed", PLANET_FEED)
owner_name = config_get(config, "Planet", "owner_name", OWNER_NAME)
owner_email = config_get(config, "Planet", "owner_email", OWNER_EMAIL)
if verbose:
log_level = "DEBUG"
else:
log_level = config_get(config, "Planet", "log_level", LOG_LEVEL)
feed_timeout = config_get(config, "Planet", "feed_timeout", FEED_TIMEOUT)
template_files = config_get(config, "Planet", "template_files",
TEMPLATE_FILES).split(" ")
# Default feed to the first feed for which there is a template
if not planet_feed:
for template_file in template_files:
name = os.path.splitext(os.path.basename(template_file))[0]
if name.find('atom')>=0 or name.find('rss')>=0:
planet_feed = urlparse.urljoin(planet_link, name)
break
# Define locale
if config.has_option("Planet", "locale"):
# The user can specify more than one locale (separated by ":") as
# fallbacks.
locale_ok = False
for user_locale in config.get("Planet", "locale").split(':'):
user_locale = user_locale.strip()
try:
locale.setlocale(locale.LC_ALL, user_locale)
except locale.Error:
pass
else:
locale_ok = True
break
if not locale_ok:
print >>sys.stderr, "Unsupported locale setting."
sys.exit(1)
# Activate logging
planet.logging.basicConfig()
planet.logging.getLogger().setLevel(planet.logging.getLevelName(log_level))
log = planet.logging.getLogger("planet.runner")
try:
log.warning
except:
log.warning = log.warn
# timeoutsocket allows feedparser to time out rather than hang forever on
# ultra-slow servers. Python 2.3 now has this functionality available in
# the standard socket library, so under 2.3 you don't need to install
# anything. But you probably should anyway, because the socket module is
# buggy and timeoutsocket is better.
if feed_timeout:
try:
feed_timeout = float(feed_timeout)
except:
log.warning("Feed timeout set to invalid value '%s', skipping", feed_timeout)
feed_timeout = None
if feed_timeout and not offline:
try:
from planet import timeoutsocket
timeoutsocket.setDefaultSocketTimeout(feed_timeout)
log.debug("Socket timeout set to %d seconds", feed_timeout)
except ImportError:
import socket
if hasattr(socket, 'setdefaulttimeout'):
log.debug("timeoutsocket not found, using python function")
socket.setdefaulttimeout(feed_timeout)
log.debug("Socket timeout set to %d seconds", feed_timeout)
else:
log.error("Unable to set timeout to %d seconds", feed_timeout)
# run the planet
my_planet = planet.Planet(config)
my_planet.run(planet_name, planet_link, template_files, offline)
my_planet.generate_all_files(template_files, planet_name,
planet_link, planet_feed, owner_name, owner_email)
