def setup(opts={}, args=[], conf_path=None):
conf_root = None
conf_over = {}
# Process command line
if hasattr(opts, 'option'):
for o in opts.option:
p = o.split('=')
if len(p) == 2:
conf_over[p[0]] = p[1]
if hasattr(opts, 'days') and opts.days is not None:
conf_over['days'] = opts.days
if hasattr(opts, 'formatter') and opts.formatter is not None:
conf_over['formatter'] = opts.formatter
if hasattr(opts, 'grabber') and opts.grabber is not None:
conf_over['grabber'] = opts.grabber
if hasattr(opts, 'debug') and opts.debug is not None:
conf_over['debug_level'] = opts.debug
if hasattr(opts, 'logpath') and opts.logpath is not None:
conf_over['log_path'] = opts.logpath
if hasattr(opts, 'syslog') and opts.syslog is not None:
conf_over['syslog'] = opts.syslog
if hasattr(opts, 'config') and opts.config is not None:
conf_path = opts.config
if hasattr(opts, 'confdir') and opts.confdir is not None:
conf_root = opts.confdir
# Defaults
if conf_root is None:
conf_root = os.path.expanduser('~/.pyepg')
if conf_path is None:
conf_path = os.path.join(conf_root, 'config')
cache_path = os.path.join(conf_root, 'cache')
# Load configuration
conf.init(conf_path, conf_over)
# Initialise log
log.init(conf.get('log_path', None), conf.get('syslog', False),
conf.get('debug_level', -1))
# Initialise the cache
cache.init(cache_path)
def setup ( opts = {}, args = [], conf_path = None ):
conf_root = None
conf_over = {}
# Process command line
if hasattr(opts, 'option'):
for o in opts.option:
p = o.split('=')
if len(p) == 2:
conf_over[p[0]] = p[1]
if hasattr(opts, 'days') and opts.days is not None:
conf_over['days'] = opts.days
if hasattr(opts, 'formatter') and opts.formatter is not None:
conf_over['formatter'] = opts.formatter
if hasattr(opts, 'grabber') and opts.grabber is not None:
conf_over['grabber'] = opts.grabber
if hasattr(opts, 'debug') and opts.debug is not None:
conf_over['debug_level'] = opts.debug
if hasattr(opts, 'logpath') and opts.logpath is not None:
conf_over['log_path'] = opts.logpath
if hasattr(opts, 'syslog') and opts.syslog is not None:
conf_over['syslog'] = opts.syslog
if hasattr(opts, 'config') and opts.config is not None:
conf_path = opts.config
if hasattr(opts, 'confdir') and opts.confdir is not None:
conf_root = opts.confdir
# Defaults
if conf_root is None:
conf_root = os.path.expanduser('~/.pyepg')
if conf_path is None:
conf_path = os.path.join(conf_root, 'config')
cache_path = os.path.join(conf_root, 'cache')
# Load configuration
conf.init(conf_path, conf_over)
# Initialise log
log.init(conf.get('log_path', None), conf.get('syslog', False),
conf.get('debug_level', -1))
# Initialise the cache
cache.init(cache_path)
def get_channels(package=None):
# Get defaults
if package is None: package = get_package()
# Map channels
channels = map(lambda x: Channel(x), conf.get('channel[]', []))
if package:
channels = package.channels(channels)
return channels
def get_channels ( package = None ):
# Get defaults
if package is None: package = get_package()
# Map channels
channels = map(lambda x: Channel(x), conf.get('channel[]', []))
if package:
channels = package.channels(channels)
return channels
def configure ():
print ''
print 'Atlas Configuration'
print '-' * 60
# API key
apikey = conf.get('atlas_apikey', '')
print ''
print 'API Key [%s]: ' % apikey,
apikey = sys.stdin.readline().strip()
if apikey: conf.set('atlas_apikey', apikey)
# Publishers to be used
p_pubs = [ 'bbc.co.uk', 'five.tv', 'channel4.com', 'itv.com', 'tvblob.com' ]
s_pubs = [ 'pressassociation.com' ]
conf.set('atlas_primary_publishers', conf.get('atlas_primary_publishers', p_pubs))
conf.set('atlas_secondary_publishers', conf.get('atlas_secondary_publishers', s_pubs))
# Hidden settings
conf.set('atlas_channel_chunk', conf.get('atlas_channel_chunk', 32))
conf.set('atlas_time_chunk', conf.get('atlas_time_chunk', 86400))
def configure():
print ''
print 'Atlas Configuration'
print '-' * 60
# API key
apikey = conf.get('atlas_apikey', '')
print ''
print 'API Key [%s]: ' % apikey,
apikey = sys.stdin.readline().strip()
if apikey: conf.set('atlas_apikey', apikey)
# Publishers to be used
p_pubs = ['bbc.co.uk', 'five.tv', 'channel4.com', 'itv.com', 'tvblob.com']
s_pubs = ['pressassociation.com']
conf.set('atlas_primary_publishers',
conf.get('atlas_primary_publishers', p_pubs))
conf.set('atlas_secondary_publishers',
conf.get('atlas_secondary_publishers', s_pubs))
# Hidden settings
conf.set('atlas_channel_chunk', conf.get('atlas_channel_chunk', 32))
conf.set('atlas_time_chunk', conf.get('atlas_time_chunk', 86400))
def publisher_overlay(a, b, pubs):
ignore_keys = conf.get('atlas_overlay_ignore',
['uri']) #, 'transmission_end_time' ])
pa = a['publisher']['key']
pb = b['publisher']['key']
ia = -1
ib = -1
try:
ia = pubs.index(pa)
except:
pass
try:
ib = pubs.index(pb)
except:
pass
def _overlay(a, b):
if type(b) == dict:
for k in b:
if k not in a:
a[k] = b[k]
elif k not in ignore_keys:
a[k] = _overlay(a[k], b[k])
return a
elif type(b) == list:
for i in range(len(b)):
if i < len(a):
a[i] = _overlay(a[i], b[i])
else:
a.append(b[i])
return a
else:
return b
ret = None
if ib < ia:
t = a
a = b
b = t
args = (a['uri'],
a['broadcasts'][0]['transmission_time'].strftime('%H:%M'),
a['broadcasts'][0]['transmission_end_time'].strftime('%H:%M'),
b['uri'],
b['broadcasts'][0]['transmission_time'].strftime('%H:%M'),
b['broadcasts'][0]['transmission_end_time'].strftime('%H:%M'))
log.debug('overlay %s @ %s-%s with %s @ %s-%s' % args, 6)
ret = _overlay(a, b)
return ret
def grab ( opts, args ):
# Initialise EPG
epg = EPG()
# Get config
days = conf.get('days', 7)
today = datetime.datetime.today()
# Get grabber/formatter
grabber = get_grabber()
formatter = get_formatter()
# Channels
channels = get_channels()
# Get EPG
log.info('grabbing EPG for %d days' % days)
grabber.grab(epg, channels, today, today + datetime.timedelta(days=days))
# Attempt to deal with missing +N channels
fix_plus_n(epg, channels)
# Finish the EPG (will tidy it up)
epg.finish()
# Output
formatter.format(epg, sys.stdout)
# Stats
log.info('')
log.info('Statistics:')
log.info('--------------------------------------')
log.info('Channel Count: %d' % len(epg.get_channels()))
log.info('Brand Count: %d' % len(epg.get_brands()))
log.info('Series Count: %d' % len(epg.get_series()))
log.info('Episode Count: %d' % len(epg.get_episodes()))
log.info('Schedule Count: %d' % epg.get_sched_count())
def grab(opts, args):
# Initialise EPG
epg = EPG()
# Get config
days = conf.get('days', 7)
today = datetime.datetime.today()
# Get grabber/formatter
grabber = get_grabber()
formatter = get_formatter()
# Channels
channels = get_channels()
# Get EPG
log.info('grabbing EPG for %d days' % days)
grabber.grab(epg, channels, today, today + datetime.timedelta(days=days))
# Attempt to deal with missing +N channels
fix_plus_n(epg, channels)
# Finish the EPG (will tidy it up)
epg.finish()
# Output
formatter.format(epg, sys.stdout)
# Stats
log.info('')
log.info('Statistics:')
log.info('--------------------------------------')
log.info('Channel Count: %d' % len(epg.get_channels()))
log.info('Brand Count: %d' % len(epg.get_brands()))
log.info('Series Count: %d' % len(epg.get_series()))
log.info('Episode Count: %d' % len(epg.get_episodes()))
log.info('Schedule Count: %d' % epg.get_sched_count())
def publisher_overlay ( a, b, pubs ):
ignore_keys = conf.get('atlas_overlay_ignore', [ 'uri' ])#, 'transmission_end_time' ])
pa = a['publisher']['key']
pb = b['publisher']['key']
ia = -1
ib = -1
try:
ia = pubs.index(pa)
except: pass
try:
ib = pubs.index(pb)
except: pass
def _overlay ( a, b ):
if type(b) == dict:
for k in b:
if k not in a:
a[k] = b[k]
elif k not in ignore_keys:
a[k] = _overlay(a[k], b[k])
return a
elif type(b) == list:
for i in range(len(b)):
if i < len(a):
a[i] = _overlay(a[i], b[i])
else:
a.append(b[i])
return a
else:
return b
ret = None
if ib < ia:
t = a
a = b
b = t
args = (a['uri'], a['broadcasts'][0]['transmission_time'].strftime('%H:%M'), a['broadcasts'][0]['transmission_end_time'].strftime('%H:%M'), b['uri'], b['broadcasts'][0]['transmission_time'].strftime('%H:%M'), b['broadcasts'][0]['transmission_end_time'].strftime('%H:%M'))
log.debug('overlay %s @ %s-%s with %s @ %s-%s' % args, 6)
ret = _overlay(a, b)
return ret
def _get_file(name, ttl=None):
import time
log.debug('cache: get file %s' % name, 3)
ok = False
data = None
meta = None
valid = False
path = CACHE_PATH + os.path.sep + name
# Default TTL
if ttl is None: ttl = conf.get('default_cache_ttl', 7 * 86400)
# Check age
if os.path.exists(path) and os.path.exists(path + '.meta'):
log.debug('cache: %s in cache' % name, 4)
st = os.stat(path)
meta = eval(open(path + '.meta').read())
data = open(path).read()
# OK
if (st.st_mtime + ttl) > time.time():
log.debug('cache: %s ttl ok' % name, 4)
ok = True
# TTL passed
else:
log.debug('cache: %s ttl expired' % name, 4)
# Validate
if 'md5' in meta and meta['md5'] == md5(data):
log.debug('cache: %s md5 ok' % name, 4)
valid = True
else:
log.debug('cache: %s md5 mismatch' % name)
# Return data
return (data, meta, ok, valid)
def _get_file ( name, ttl = None ):
import time
log.debug('cache: get file %s' % name, 3)
ok = False
data = None
meta = None
valid = False
path = CACHE_PATH + os.path.sep + name
# Default TTL
if ttl is None: ttl = conf.get('default_cache_ttl', 7*86400)
# Check age
if os.path.exists(path) and os.path.exists(path + '.meta'):
log.debug('cache: %s in cache' % name, 4)
st = os.stat(path)
meta = eval(open(path + '.meta').read())
data = open(path).read()
# OK
if (st.st_mtime + ttl) > time.time():
log.debug('cache: %s ttl ok' % name, 4)
ok = True
# TTL passed
else:
log.debug('cache: %s ttl expired' % name, 4)
# Validate
if 'md5' in meta and meta['md5'] == md5(data):
log.debug('cache: %s md5 ok' % name, 4)
valid = True
else:
log.debug('cache: %s md5 mismatch' % name)
# Return data
return (data, meta, ok, valid)
def get_package ():
ret = None
package = conf.get('package', None)
if package:
ret = _import('pyepg.package.%s', package)
return ret
def grab(epg, channels, start, stop):
import multiprocessing as mp
# Filter the channel list (only include those we have listing for)
channels = filter_channels(channels)
days = util.total_seconds(stop - start) / 86400
channels = sorted(channels, cmp=lambda a, b: cmp(a.number, b.number))
log.info('atlas - epg grab %d channels for %d days' %
(len(channels), days))
# Config
grab_thread_cnt = conf.get('atlas_grab_threads', 32)
data_thread_cnt = conf.get('atlas_data_threads', 0)
if grab_thread_cnt <= 0:
grab_thread_cnt = len(channels)
if data_thread_cnt <= 0:
data_thread_cnt = mp.cpu_count() * 2
data_thread_cnt = min(data_thread_cnt, len(channels))
grab_thread_cnt = min(grab_thread_cnt, len(channels))
# Create input/output queues
inq = ChannelQueue(channels)
outq = DataQueue(len(channels))
# Create grab threads
grab_threads = []
for i in range(grab_thread_cnt):
t = GrabThread(i, inq, outq, start, stop)
grab_threads.append(t)
# Create data threads
data_threads = []
for i in range(data_thread_cnt):
t = DataThread(i, outq, epg)
data_threads.append(t)
# Start threads
for t in grab_threads:
t.start()
for t in data_threads:
t.start()
# Wait for completion (inq first)
ins = outs = len(channels)
while True:
s = inq.remain()
if s != ins:
ins = s
log.info('atlas - grab %3d/%3d channels remain' %
(s, len(channels)))
s = outq.remain()
if s != outs:
outs = s
log.info('atlas - proc %3d/%3d channels remain' %
(s, len(channels)))
if not ins and not outs: break
# Safety checks
i = 0
for t in grab_threads:
if t.isAlive(): i = i + 1
if not i and ins:
log.error('atlas - grab threads have died prematurely')
break
i = 0
for t in data_threads:
if t.isAlive(): i = i + 1
if not i and outs:
log.error('atlas - proc threads have died prematurely')
break
time.sleep(1.0)
def run(self):
conn = None
log.debug('atlas - grab thread %3d started' % self._idx, 0)
# Create connection
import httplib
retry = conf.get('atlas_conn_retry_limit', 5)
while not conn and retry:
try:
conn = httplib.HTTPConnection(ATLAS_API_HOST)
log.debug('atlas - grab thread %3d conn created' % self._idx,
1)
except:
retry = retry - 1
time.sleep(conf.get('atlas_conn_retry_period', 2.0))
if not conn:
log.error('atlas - grab thread %3d failed to connect')
return
# Config
key = conf.get('atlas_apikey', None)
p_pubs = conf.get('atlas_primary_publishers',\
[ 'bbc.co.uk', 'itv.com' 'tvblob.com',\
'channel4.com' ])
s_pubs = conf.get('atlas_secondary_publishers',\
[ 'pressassociation.com' ])
anno = [ 'broadcasts', 'extended_description', 'series_summary',\
'brand_summary', 'people' ]
tsize = conf.get('atlas_time_chunk', self._stop - self._start)
# Time
tm_from = time.mktime(self._start.timetuple())
tm_to = time.mktime(self._stop.timetuple())
# URL base
url = 'schedule.json?'
url = url + 'annotations=' + ','.join(anno)
if key: url = url + '&apiKey=' + key
# Until queue exhausted
while True:
# Get next entry
c = None
try:
c = self._inq.get_nowait()
except Empty:
break
log.debug(
'atlas - grab thread %3d fetch %s' % (self._idx, c.title), 0)
sched = []
# By time
tf = tm_from
while tf < tm_to:
tt = min(tf + tsize, tm_to)
a = (time.strftime('%Y-%m-%d %H:%M', time.localtime(tf)),\
time.strftime('%Y-%m-%d %H:%M', time.localtime(tt)))
#log.info('atlas - period %s to %s' % a)
# Process each publisher
pubs = []
for p in s_pubs:
pubs.append(p)
for p in p_pubs:
if p in c.publisher: pubs.append(p)
log.debug('PUBS: %s' % pubs, 0)
for p in pubs:
#log.info('atlas - publisher %s' % p)
u = url + '&from=%d&to=%d' % (tf, tt)
u = u + '&publisher=' + p
u = u + '&channel_id=' + c.shortid
# Fetch data
data = atlas_fetch(u, conn)
# Processs
if data and 'schedule' in data:
for s in data['schedule']:
if 'items' in s:
sched.extend(s['items'])
# Update
tf = tf + tsize
# Put into the output queue
log.debug(
'atlas - grab thread %3d fetched %s' % (self._idx, c.title), 1)
self._outq.put((c, pubs, sched))
self._inq.task_done()
# Done
if conn: conn.close()
log.debug('atlas - grab thread %3d complete' % self._idx, 0)
def configure ( opts, args, conf_path = None ):
#
# Global
#
print 'System Configuration'
print '-' * 60
# Number of days to grab
days = conf.get('days', 7)
while True:
print 'Days to grab [%d]: ' % days,
t = sys.stdin.readline().strip()
if not t: break
try:
days = int(t)
break
except: pass
conf.set('days', days)
# Postcode
print '\nPostcode (for regional TV) [%s]: ' % conf.get('postcode', ''),
pc = sys.stdin.readline().strip()
if pc:
conf.set('postcode', pc)
#
# Grabber
#
grabbers = get_grabbers()
if not grabbers:
log.error('no grabbers available')
sys.exit(1)
options = map(lambda x: x[0], grabbers)
idx = get_select('\nSelect grabber:', options)
grabber = grabbers[idx][1]
conf.set('grabber', grabbers[idx][0])
print ''
print 'Grabber: %s' % grabbers[idx][0]
#
# Formatter
#
formatters = get_formatters()
if not formatters:
log.error('no formatters available')
sys.exit(1)
options = map(lambda x: x[0], formatters)
idx = get_select('\nSelect formatter:', options)
formatter = formatters[idx][1]
conf.set('formatter', formatters[idx][0])
print ''
print 'Formatter: %s' % formatters[idx][0]
#
# Grabber/Formatter config
#
if hasattr(grabber, 'configure'):
grabber.configure()
if hasattr(formatter, 'configure'):
formatter.configure()
#
# Channels
#
channels = []
print ''
print 'Channel Configuration'
print '-' * 60
# Get packages
packages = grabber.packages()
options = []
options.extend(['Skip'])
options.extend(map(lambda x: x.title(), packages))
idx = get_select('Select Platform:', options)
# Platform
if idx:
idx = idx - 1
package = packages[idx]
conf.set('package', package.id())
# Exclusions
a = None
while a not in [ 'y', 'n', 'yes', 'no' ]:
print '\nWould you like to add exclusions (y/n)? ',
a = sys.stdin.readline().strip().lower()
# Get
if a in [ 'y', 'yes' ]:
for c in package.channels():
a = None
while a not in [ 'y', 'n', 'yes', 'no' ]:
print '\n %s (y/n)? ' % c.title,
a = sys.stdin.readline().strip().lower()
if a in [ 'y', 'yes' ]: channels.append(c.title)
# Store
channels = []
for c in package.channels():
channels.append(c.uri)
conf.set('channel[]', channels)
#
# Output summary and get confirmation
#
# TODO
#
# Save
#
conf.save()
def get_grabber():
return _import('pyepg.grabber.%s', conf.get('grabber', 'atlas'))
def get_formatter():
return _import('pyepg.formatter.%s', conf.get('formatter', 'epg'))
continue
# Regional channel
if p[7] == '1':
regional.append(c)
# Store
elif c.extra['stream'][0][0]:
chns.append(c)
except Exception, e:
log.error('failed to process [%s] [e=%s]' % (l, str(e)))
# Process regional channels
regions = process_region_data(reg_data)
pc = conf.get('postcode', '')
for c in regional:
t = find_regional(c, pc, chns, regions)
if t:
c.uri = t.uri
c.extra['stream'] = t.extra['stream']
chns.insert(0, c)
# Filter duplicates
ret = []
for c in chns:
if c not in ret:
ret.append(c)
return ret
def configure(opts, args, conf_path=None):
#
# Global
#
print 'System Configuration'
print '-' * 60
# Number of days to grab
days = conf.get('days', 7)
while True:
print 'Days to grab [%d]: ' % days,
t = sys.stdin.readline().strip()
if not t: break
try:
days = int(t)
break
except:
pass
conf.set('days', days)
# Postcode
print '\nPostcode (for regional TV) [%s]: ' % conf.get('postcode', ''),
pc = sys.stdin.readline().strip()
if pc:
conf.set('postcode', pc)
#
# Grabber
#
grabbers = get_grabbers()
if not grabbers:
log.error('no grabbers available')
sys.exit(1)
options = map(lambda x: x[0], grabbers)
idx = get_select('\nSelect grabber:', options)
grabber = grabbers[idx][1]
conf.set('grabber', grabbers[idx][0])
print ''
print 'Grabber: %s' % grabbers[idx][0]
#
# Formatter
#
formatters = get_formatters()
if not formatters:
log.error('no formatters available')
sys.exit(1)
options = map(lambda x: x[0], formatters)
idx = get_select('\nSelect formatter:', options)
formatter = formatters[idx][1]
conf.set('formatter', formatters[idx][0])
print ''
print 'Formatter: %s' % formatters[idx][0]
#
# Grabber/Formatter config
#
if hasattr(grabber, 'configure'):
grabber.configure()
if hasattr(formatter, 'configure'):
formatter.configure()
#
# Channels
#
channels = []
print ''
print 'Channel Configuration'
print '-' * 60
# Get packages
packages = grabber.packages()
options = []
options.extend(['Skip'])
options.extend(map(lambda x: x.title(), packages))
idx = get_select('Select Platform:', options)
# Platform
if idx:
idx = idx - 1
package = packages[idx]
conf.set('package', package.id())
# Exclusions
a = None
while a not in ['y', 'n', 'yes', 'no']:
print '\nWould you like to add exclusions (y/n)? ',
a = sys.stdin.readline().strip().lower()
# Get
if a in ['y', 'yes']:
for c in package.channels():
a = None
while a not in ['y', 'n', 'yes', 'no']:
print '\n %s (y/n)? ' % c.title,
a = sys.stdin.readline().strip().lower()
if a in ['y', 'yes']: channels.append(c.title)
# Store
channels = []
for c in package.channels():
channels.append(c.uri)
conf.set('channel[]', channels)
#
# Output summary and get confirmation
#
# TODO
#
# Save
#
conf.save()
def get_grabber():
return _import('pyepg.grabber.%s', conf.get('grabber', 'atlas'))
def get_package():
ret = None
package = conf.get('package', None)
if package:
ret = _import('pyepg.package.%s', package)
return ret
def get_formatter():
return _import('pyepg.formatter.%s', conf.get('formatter', 'epg'))
def run ( self ):
conn = None
log.debug('atlas - grab thread %3d started' % self._idx, 0)
# Create connection
import httplib
retry = conf.get('atlas_conn_retry_limit', 5)
while not conn and retry:
try:
conn = httplib.HTTPConnection(ATLAS_API_HOST)
log.debug('atlas - grab thread %3d conn created' % self._idx, 1)
except:
retry = retry - 1
time.sleep(conf.get('atlas_conn_retry_period', 2.0))
if not conn:
log.error('atlas - grab thread %3d failed to connect')
return
# Config
key = conf.get('atlas_apikey', None)
p_pubs = conf.get('atlas_primary_publishers',\
[ 'bbc.co.uk', 'itv.com' 'tvblob.com',\
'channel4.com' ])
s_pubs = conf.get('atlas_secondary_publishers',\
[ 'pressassociation.com' ])
anno = [ 'broadcasts', 'extended_description', 'series_summary',\
'brand_summary', 'people' ]
tsize = conf.get('atlas_time_chunk', self._stop - self._start)
# Time
tm_from = time.mktime(self._start.timetuple())
tm_to = time.mktime(self._stop.timetuple())
# URL base
url = 'schedule.json?'
url = url + 'annotations=' + ','.join(anno)
if key: url = url + '&apiKey=' + key
# Until queue exhausted
while True:
# Get next entry
c = None
try:
c = self._inq.get_nowait()
except Empty:
break
log.debug('atlas - grab thread %3d fetch %s' % (self._idx, c.title), 0)
sched = []
# By time
tf = tm_from
while tf < tm_to:
tt = min(tf + tsize, tm_to)
a = (time.strftime('%Y-%m-%d %H:%M', time.localtime(tf)),\
time.strftime('%Y-%m-%d %H:%M', time.localtime(tt)))
#log.info('atlas - period %s to %s' % a)
# Process each publisher
pubs = []
for p in s_pubs: pubs.append(p)
for p in p_pubs:
if p in c.publisher: pubs.append(p)
log.debug('PUBS: %s' % pubs, 0)
for p in pubs:
#log.info('atlas - publisher %s' % p)
u = url + '&from=%d&to=%d' % (tf, tt)
u = u + '&publisher=' + p
u = u + '&channel_id=' + c.shortid
# Fetch data
data = atlas_fetch(u, conn)
# Processs
if data and 'schedule' in data:
for s in data['schedule']:
if 'items' in s:
sched.extend(s['items'])
# Update
tf = tf + tsize
# Put into the output queue
log.debug('atlas - grab thread %3d fetched %s' % (self._idx, c.title), 1)
self._outq.put((c, pubs, sched))
self._inq.task_done()
# Done
if conn: conn.close()
log.debug('atlas - grab thread %3d complete' % self._idx, 0)
def grab ( epg, channels, start, stop ):
import multiprocessing as mp
# Filter the channel list (only include those we have listing for)
channels = filter_channels(channels)
days = util.total_seconds(stop - start) / 86400
channels = sorted(channels, cmp=lambda a,b: cmp(a.number,b.number))
log.info('atlas - epg grab %d channels for %d days' % (len(channels), days))
# Config
grab_thread_cnt = conf.get('atlas_grab_threads', 32)
data_thread_cnt = conf.get('atlas_data_threads', 0)
if grab_thread_cnt <= 0:
grab_thread_cnt = len(channels)
if data_thread_cnt <= 0:
data_thread_cnt = mp.cpu_count() * 2
data_thread_cnt = min(data_thread_cnt, len(channels))
grab_thread_cnt = min(grab_thread_cnt, len(channels))
# Create input/output queues
inq = ChannelQueue(channels)
outq = DataQueue(len(channels))
# Create grab threads
grab_threads = []
for i in range(grab_thread_cnt):
t = GrabThread(i, inq, outq, start, stop)
grab_threads.append(t)
# Create data threads
data_threads = []
for i in range(data_thread_cnt):
t = DataThread(i, outq, epg)
data_threads.append(t)
# Start threads
for t in grab_threads: t.start()
for t in data_threads: t.start()
# Wait for completion (inq first)
ins = outs = len(channels)
while True:
s = inq.remain()
if s != ins:
ins = s
log.info('atlas - grab %3d/%3d channels remain' % (s, len(channels)))
s = outq.remain()
if s != outs:
outs = s
log.info('atlas - proc %3d/%3d channels remain' % (s, len(channels)))
if not ins and not outs: break
# Safety checks
i = 0
for t in grab_threads:
if t.isAlive(): i = i + 1
if not i and ins:
log.error('atlas - grab threads have died prematurely')
break
i = 0
for t in data_threads:
if t.isAlive(): i = i + 1
if not i and outs:
log.error('atlas - proc threads have died prematurely')
break
time.sleep(1.0)
def get_data ( name, ttl = None ):
url = conf.get('data_url', 'http://cloud.github.com/downloads/adamsutton/PyEPG')
return get_url(url + '/' + name, True, ttl)
continue
# Regional channel
if p[7] == '1':
regional.append(c)
# Store
elif c.extra['stream'][0][0]:
chns.append(c)
except Exception, e:
log.error('failed to process [%s] [e=%s]' % (l, str(e)))
# Process regional channels
regions = process_region_data(reg_data)
pc = conf.get('postcode', '')
for c in regional:
t = find_regional(c, pc, chns, regions)
if t:
c.uri = t.uri
c.extra['stream'] = t.extra['stream']
chns.insert(0, c)
# Filter duplicates
ret = []
for c in chns:
if c not in ret:
ret.append(c)
return ret
def get_data(name, ttl=None):
url = conf.get('data_url',
'http://cloud.github.com/downloads/adamsutton/PyEPG')
return get_url(url + '/' + name, True, ttl)