def fetch_all(cls, filter_expr=None, filter_value=None, sort=None):
# TODO: Make this use less resources. Read options online and then use profiler to test each one.
assert (filter_expr is None)==(filter_value is None)
import settings, helpers
def get_query():
query = cls.all()
if filter_expr is not None:
assert filter_value is not None
query = query.filter(filter_expr, filter_value)
if sort is not None:
query.order(sort)
return query
items = get_query().fetch(EXPECTED_UPPER_BOUND)
if len(items) >= EXPECTED_UPPER_BOUND:
if settings.DEBUG:
assert False, "Upper bound is apparently not big enough."
else:
helpers.log("ERROR: Upper bound is apparently not big enough.")
items = list( get_query() )
if cls.default_sort_key_fn is not None:
items.sort(key=cls.default_sort_key_fn)
return tuple(items)
def set_up_environment (self, settings, ontodir):
"""
Sets up the XDG_*_HOME variables and make sure the directories exist
Settings should be a dict mapping schema names to dicts that hold the
settings that should be changed in those schemas. The contents dicts
should map key->value, where key is a key name and value is a suitable
GLib.Variant instance.
"""
helpers.log ("[Conf] Setting test environment...")
for var, directory in TEST_ENV_DIRS.iteritems ():
helpers.log ("export %s=%s" %(var, directory))
self.__recreate_directory (directory)
os.environ [var] = directory
for directory in EXTRA_DIRS:
self.__recreate_directory (directory)
if ontodir:
helpers.log ("export %s=%s" % ("TRACKER_DB_ONTOLOGIES_DIR", ontodir))
os.environ ["TRACKER_DB_ONTOLOGIES_DIR"] = ontodir
for var, value in TEST_ENV_VARS.iteritems ():
helpers.log ("export %s=%s" %(var, value))
os.environ [var] = value
# Previous loop should have set DCONF_PROFILE to the test location
if settings is not None:
self._apply_settings(settings)
helpers.log ("[Conf] environment ready")
def extract_sold_by(product_page_lxml):
"""Extracts who the product is sold by given a product page html in
unicode"""
sold_by = u'Not Sold by Amazon'
try:
results = product_page_lxml.cssselect('div.buying')
for search_result in results:
search_result = search_result.text_content()
if 'Fulfilled by Amazon' in search_result:
sold_by = u'Fulfilled by Amazon'
break
elif 'Ships from and sold by Amazon.com' in search_result:
sold_by = u'Sold by Amazon'
break
elif 'Ships from and sold by Amazon Digital Services' in \
search_result:
sold_by = u'Sold by Amazon'
break
else:
continue
except:
helpers.log('Sold by error on page')
sold_by = 'Error'
return sold_by
def send_statistics():
if GlobalVars.metasmoke_down:
log('warning', "Metasmoke is down, not sending statistics")
return
GlobalVars.posts_scan_stats_lock.acquire()
if GlobalVars.post_scan_time != 0:
posts_per_second = GlobalVars.num_posts_scanned / GlobalVars.post_scan_time
payload = {'key': GlobalVars.metasmoke_key,
'statistic': {'posts_scanned': GlobalVars.num_posts_scanned, 'api_quota': GlobalVars.apiquota,
'post_scan_rate': posts_per_second}}
else:
payload = {'key': GlobalVars.metasmoke_key,
'statistic': {'posts_scanned': GlobalVars.num_posts_scanned, 'api_quota': GlobalVars.apiquota}}
GlobalVars.post_scan_time = 0
GlobalVars.num_posts_scanned = 0
GlobalVars.posts_scan_stats_lock.release()
headers = {'Content-type': 'application/json'}
if GlobalVars.metasmoke_host is not None:
log('info', 'Sent statistics to metasmoke: ', payload['statistic'])
Metasmoke.post("/statistics.json",
data=json.dumps(payload), headers=headers)
def _check_batch(saved):
if time.time() < DeletionWatcher.next_request_time:
time.sleep(DeletionWatcher.next_request_time - time.time())
for site, posts in saved.items():
ids = ";".join(post_id for post_id in posts if not DeletionWatcher._ignore((post_id, site)))
uri = "https://api.stackexchange.com/2.2/posts/{}".format(ids)
params = {
'site': site,
'key': 'IAkbitmze4B8KpacUfLqkw(('
}
res = requests.get(uri, params=params)
json = res.json()
if "items" not in json:
log('warning',
'DeletionWatcher API request received no items in response (code {})'.format(res.status_code))
log('warning', res.text)
return
if 'backoff' in json:
DeletionWatcher.next_request_time = time.time() + json['backoff']
for post in json['items']:
if time.time() - post["creation_date"] < 7200:
yield to_protocol_relative(post["link"]).replace("/q/", "/questions/")
def getRandomIdeas(questionObj, ideas, size=5):
numIdeas = len(ideas) if ideas else 0
if numIdeas >= size:
return random.sample(ideas, size)
else:
helpers.log("WARNING: Cannot return {0} random ideas since only {1} ideas available".format(size, numIdeas))
return []
def getPerson(question=None, nickname=None):
person = None
# check if person id stored in session
# if so use to retrieve logged in user
session = gaesessions.get_current_session()
person_id = session.pop("new_person_id") if session.has_key("new_person_id") else None
if person_id:
person = Person.get_by_id(person_id)
# check if person id stored in session corresponds to inputs
if not person:
person = None
helpers.log("WARNING: Person not found by id {0}".format(person_id))
elif question and question != person.question:
person = None
elif nickname and nickname != person.nickname:
person = None
if not person:
user = users.get_current_user()
if question:
if question.nicknameAuthentication:
# if no nickname provided, check session
if not nickname:
questionSessionValues = session.get(question.code)
nickname = questionSessionValues["nickname"] if questionSessionValues else None
if nickname:
person = Person.all().filter("question =", question).filter("nickname =", nickname).get()
elif user is not None:
person = Person.all().filter("question =", question).filter("user ="******"question =", None).filter("user =", user).get()
return person
def _send_update(from_person, to_person, *updates):
from model import Student, Teacher
from client_id_utils import timestamp_for_client_id
# from settings import CHANNEL_LIMIT_PER_STUDENT
from helpers import log
assert isinstance(to_person, (Student,Teacher)), repr(to_person)
import datetime
timestamp = datetime.datetime.now()
updates_list = list(updates)
updates_list[0]['timestamp'] = timestamp.strftime('%B %d, %Y %H:%M:%S')
updates_json = json.dumps(updates_list)
# Sort and dedupe client_ids by timestamp, descending
client_ids = set(to_person.client_ids)
key_fn = lambda client_id:timestamp_for_client_id(client_id)
client_ids = sorted(client_ids, key=key_fn, reverse=True)
# if CHANNEL_LIMIT_PER_STUDENT is not None and len(client_ids) > CHANNEL_LIMIT_PER_STUDENT and isinstance(to_person, Student):
# client_ids = client_ids[:CHANNEL_LIMIT_PER_STUDENT]
# log( "=> WARNING: Found %d client IDs for %r but only using %d"%(len(to_person.client_ids), to_person, CHANNEL_LIMIT_PER_STUDENT) )
if len(client_ids)==0:
log("=> MESSAGE NOT SENT. No current client IDs for {0}.".format(to_person))
for client_id in client_ids:
log( "client ID : %r : sent %s"%(client_id," + ".join(u["type"] for u in updates)) )
channel.send_message(client_id, updates_json)
def main():
while True:
data = read(API_URL)
result = parse(data)
log(repr(result))
save_to_db(result)
sleep(TIMEOUT)
def send_stats_on_post(title, link, reasons, body, username, user_link, why, owner_rep,
post_score, up_vote_count, down_vote_count):
if GlobalVars.metasmoke_host is None:
log('info', "Metasmoke location not defined, not reporting")
return
metasmoke_key = GlobalVars.metasmoke_key
try:
if len(why) > 1024:
why = why[:512] + '...' + why[-512:]
post = {'title': title, 'link': link, 'reasons': reasons,
'body': body, 'username': username, 'user_link': user_link,
'why': why, 'user_reputation': owner_rep, 'score': post_score,
'upvote_count': up_vote_count, 'downvote_count': down_vote_count}
# Remove None values (if they somehow manage to get through)
post = dict((k, v) for k, v in post.items() if v)
payload = {'post': post, 'key': metasmoke_key}
headers = {'Content-type': 'application/json'}
requests.post(GlobalVars.metasmoke_host + "/posts.json", data=json.dumps(payload), headers=headers)
except Exception as e:
log('error', e)
def _openFile(self):
fileName = QtGui.QFileDialog.getOpenFileName(self, "OpenImage", "src", "Bitmaps (*.bmp)")
if fileName == "":
h.warn("No file selected.")
return
h.log("Loaded!")
#self.pixmapItem = QtGui.QPixmap(fileName)
#item = QtGui.QGraphicsPixmapItem(self.pixmapItem)
#self.scene.addItem(item)
item = QtGui.QStandardItem(os.path.basename(str(fileName)))
item.imageFileName = str(fileName)
orgImg = Image.open(str(fileName), mode='r').convert()
item.image = MImage.pil_to_array(orgImg.convert('L'))
item.pixmap = QtGui.QPixmap(fileName)
#item.setCheckable(True)
self.model.appendRow(item)
self.ui.imageListView.setModel(self.model)
self._checkButtons()
def reload():
commit = git_commit_info()
censored_committer_names = GlobalVars.censored_committer_names
if md5(commit['author'][0].encode('utf-8')).hexdigest() in censored_committer_names:
commit['author'] = censored_committer_names[md5(commit['author'][0].encode('utf-8')).hexdigest()]
GlobalVars.commit = commit
GlobalVars.commit_with_author = "`{}` ({}: {})".format(
commit['id'],
commit['author'][0] if type(commit['author']) in {list, tuple} else commit['author'],
commit['message'])
GlobalVars.on_master = "HEAD detached" not in git_status()
GlobalVars.s = "[ {} ] SmokeDetector started at [rev {}]({}/commit/{}) (running on {}, Python {})".format(
GlobalVars.chatmessage_prefix, GlobalVars.commit_with_author, GlobalVars.bot_repository,
GlobalVars.commit['id'], GlobalVars.location, platform.python_version())
GlobalVars.s_reverted = \
"[ {} ] SmokeDetector started in [reverted mode](" \
"https://charcoal-se.org/smokey/SmokeDetector-Statuses#reverted-mode) " \
"at [rev {}]({}/commit/{}) (running on {})".format(
GlobalVars.chatmessage_prefix, GlobalVars.commit_with_author, GlobalVars.bot_repository,
GlobalVars.commit['id'], GlobalVars.location)
GlobalVars.s_norestart = "[ {} ] Blacklists reloaded at [rev {}]({}/commit/{}) (running on {})".format(
GlobalVars.chatmessage_prefix, GlobalVars.commit_with_author, GlobalVars.bot_repository,
GlobalVars.commit['id'], GlobalVars.location)
GlobalVars.s_norestart2 = "[ {} ] FindSpam module reloaded at [rev {}]({}/commit/{}) (running on {})".format(
GlobalVars.chatmessage_prefix, GlobalVars.commit_with_author, GlobalVars.bot_repository,
GlobalVars.commit['id'], GlobalVars.location)
GlobalVars.standby_message = \
"[ {} ] SmokeDetector started in [standby mode](" \
"https://charcoal-se.org/smokey/SmokeDetector-Statuses#standby-mode) " \
"at [rev {}]({}/commit/{}) (running on {})".format(
GlobalVars.chatmessage_prefix, GlobalVars.commit_with_author, GlobalVars.bot_repository,
GlobalVars.commit['id'], GlobalVars.location)
log('debug', "GlobalVars loaded")
def send(self, dstip, data):
debug2('UDP: sending to %r port %d\n' % dstip)
try:
self.sock.sendto(data, dstip)
except socket.error, e:
log('UDP send to %r port %d: %s\n' % (dstip[0], dstip[1], e))
return
def set_up_environment(self, gsettings, ontodir):
"""
Sets up the XDG_*_HOME variables and make sure the directories exist
gsettings is a list of triplets (schema, key, value) that will be set/unset in gsetting
"""
assert not gsettings or type(gsettings) is list
helpers.log("[Conf] Setting test environment...")
for var, directory in TEST_ENV_DIRS.iteritems():
helpers.log("export %s=%s" % (var, directory))
self.__recreate_directory(directory)
os.environ[var] = directory
for directory in EXTRA_DIRS:
self.__recreate_directory(directory)
if ontodir:
helpers.log("export %s=%s" % ("TRACKER_DB_ONTOLOGIES_DIR", ontodir))
os.environ["TRACKER_DB_ONTOLOGIES_DIR"] = ontodir
for var, value in TEST_ENV_VARS.iteritems():
helpers.log("export %s=%s" % (var, value))
os.environ[var] = value
# Previous loop should have set DCONF_PROFILE to the test location
if gsettings:
self.dconf = DConfClient()
self.dconf.reset()
for (schema, key, value) in gsettings:
self.dconf.write(schema, key, value)
helpers.log("[Conf] environment ready")
def begin_crawl():
# explode out all of our category `start_urls` into subcategories
with open(settings.start_file, "r") as f:
for line in f:
line = line.strip()
if not line or line.startswith("#"):
continue # skip blank and commented out lines
page, html = make_request(line)
count = 0
# look for subcategory links on this page
subcategories = page.findAll("div", "bxc-grid__image") # downward arrow graphics
subcategories.extend(page.findAll("li", "sub-categories__list__item")) # carousel hover menu
sidebar = page.find("div", "browseBox")
if sidebar:
subcategories.extend(sidebar.findAll("li")) # left sidebar
for subcategory in subcategories:
link = subcategory.find("a")
if not link:
continue
link = link["href"]
count += 1
enqueue_url(link)
log("Found {} subcategories on {}".format(count, line))
def reload():
GlobalVars.commit = commit = git_commit_info()
GlobalVars.commit_with_author = "`{}` ({}: {})".format(
commit.id, commit.author, commit.message)
GlobalVars.on_branch = git_ref()
GlobalVars.s = "[ {} ] SmokeDetector started at [rev {}]({}/commit/{}) (running on {}, Python {})".format(
GlobalVars.chatmessage_prefix, GlobalVars.commit_with_author, GlobalVars.bot_repository,
GlobalVars.commit.id, GlobalVars.location, platform.python_version())
GlobalVars.s_reverted = \
"[ {} ] SmokeDetector started in [reverted mode](" \
"https://charcoal-se.org/smokey/SmokeDetector-Statuses#reverted-mode) " \
"at [rev {}]({}/commit/{}) (running on {})".format(
GlobalVars.chatmessage_prefix, GlobalVars.commit_with_author, GlobalVars.bot_repository,
GlobalVars.commit.id, GlobalVars.location)
GlobalVars.s_norestart = "[ {} ] Blacklists reloaded at [rev {}]({}/commit/{}) (running on {})".format(
GlobalVars.chatmessage_prefix, GlobalVars.commit_with_author, GlobalVars.bot_repository,
GlobalVars.commit.id, GlobalVars.location)
GlobalVars.s_norestart2 = "[ {} ] FindSpam module reloaded at [rev {}]({}/commit/{}) (running on {})".format(
GlobalVars.chatmessage_prefix, GlobalVars.commit_with_author, GlobalVars.bot_repository,
GlobalVars.commit.id, GlobalVars.location)
GlobalVars.standby_message = \
"[ {} ] SmokeDetector started in [standby mode](" \
"https://charcoal-se.org/smokey/SmokeDetector-Statuses#standby-mode) " \
"at [rev {}]({}/commit/{}) (running on {})".format(
GlobalVars.chatmessage_prefix, GlobalVars.commit_with_author, GlobalVars.bot_repository,
GlobalVars.commit.id, GlobalVars.location)
log('debug', "GlobalVars loaded")
def handle_slice_sampler_exception(exception, starting_point, proposal_measure, opt_compwise=False):
'''
Handles slice sampler exceptions. If the slice sampler shrank to zero the slice sampler will be restarted
a few times. If this fails or if the exception was another this method will raise the given exception.
Args:
exception: the exception that occured
starting_point: the starting point that was used
proposal_measure: the used proposal measure
opt_compwise: how to set the compwise option
Returns:
the output of the slice sampelr
Raises:
Exception: the first argument
'''
if exception.message == "Slice sampler shrank to zero!":
log("Slice sampler shrank to zero! Action: trying to restart " + str(NUMBER_OF_RESTARTS)
+ " times with same starting point")
restarts_left = NUMBER_OF_RESTARTS
while restarts_left > 0:
try:
return slice_sample(starting_point, proposal_measure, compwise=opt_compwise)
except Exception as e:
log("Restart failed. " + str(restarts_left) + " restarts left. Exception was: " + e.message)
restarts_left = restarts_left - 1
# if we leave the while loop we will raise the exception we got
raise exception
def __init__(self, progress, total, message, newline=True):
if 'linux' in sys.platform:
message = "{message:<20}".format(message=message)
elif 'darwin' in sys.platform:
message = """ \033[38;5;204m{message:<20}\033[0m """.format(message=message)
import time
time.sleep(0.01)
progress += 1
percentage = (progress * 10 / total) # (divided by 10)
percentage_left = 10 - percentage
bar = '['
bar += percentage * log('*', colour="white", ret=True)
bar += percentage_left * log('*', colour="black", ret=True)
bar += ']'
bar += ' {}'.format(progress)
string = ''
if progress != total:
if progress == 1 and newline == True:
string = '\n\r {message} {bar}'.format(message=message, bar=bar)
else:
string = '\r {message} {bar}'.format(message=message, bar=bar)
else:
string = '\r {message} {bar}'.format(message=message, bar=bar)
sys.stdout.write(string)
sys.stdout.flush()
def sample_hyperparameters_gp(mcmc_iters, noiseless, input_points, func_values, cov_func, noise, amp2, ls):
'''
Samples hyper parameters for Gaussian processes.
Args:
mcmc_iters: the number of hyper-parameter samples required
noiseless: the modeled function is noiseless
input_points: all the points that have been evaluated so far
func_values: the corresponding observed function values
cov_func: the covariance function the Gaussian process uses
noise: a starting value for the noise
amp2: a starting value for the amplitude
ls: an array of starting values for the length scales (size has to be the dimension of the input points)
Returns:
a list of hyper-parameter tuples
the tuples are of the form (mean, noise, amplitude, [length-scales])
'''
mean = np.mean(func_values)
hyper_samples = []
# sample hyper parameters
for i in xrange(0, mcmc_iters):
if noiseless:
noise = 1e-3
[mean, amp2] = _sample_mean_amp_noise(input_points, func_values, cov_func, np.array([mean, amp2]), ls)
else:
[mean, amp2, noise] = _sample_mean_amp_noise(input_points, func_values, cov_func, np.array([mean, amp2, noise]), ls)
ls = _sample_ls(input_points, func_values, cov_func, ls, mean, amp2, noise)
#This is the order as expected
#log("mean: " + str(mean) + ", noise: " + str(noise) + " amp: " + str(amp2) + ", ls: " + str(ls))
hyper_samples.append((mean, noise, amp2, ls))
samples = []
for i in xrange(0, mcmc_iters - 1, mcmc_iters / 10):
samples.append(hyper_samples[i])
log("mean: " + str(hyper_samples[i][0]) + ", noise: " + str(hyper_samples[i][1]) + " amp: " + str(hyper_samples[i][2]) + ", ls: " + str(hyper_samples[i][3]))
return samples
def determine_if_autoflagged(post_url):
"""
Given the URL for a post, determine whether or not it has been autoflagged.
"""
payload = {
'key': GlobalVars.metasmoke_key,
'filter': 'GKNJKLILHNFMJLFKINGJJHJOLGFHJF', # id and autoflagged
'urls': post_url
}
try:
response = Metasmoke.get("/api/v2.0/posts/urls", params=payload).json()
except Exception as e:
log('error', e)
return False, []
if len(response["items"]) > 0 and response["items"][0]["autoflagged"]:
# get flagger names
id = str(response["items"][0]["id"])
payload = {'key': GlobalVars.metasmoke_key}
flags = Metasmoke.get("/api/v2.0/posts/" + id + "/flags", params=payload).json()
if len(flags["items"]) > 0:
return True, [user["username"] for user in flags["items"][0]["autoflagged"]["users"]]
return False, []
def subscribe(self, post_url, callback=None, pickle=True, timeout=None):
post_id, post_site, post_type = fetch_post_id_and_site_from_url(post_url)
if post_site not in GlobalVars.site_id_dict:
log("warning", "unknown site {} when subscribing to {}".format(post_site, post_url))
return
if post_type == "answer":
question_id = datahandling.get_post_site_id_link((post_id, post_site, post_type))
if question_id is None:
return
else:
question_id = post_id
site_id = GlobalVars.site_id_dict[post_site]
action = "{}-question-{}".format(site_id, question_id)
max_time = (time.time() + timeout) if timeout else None
if action not in self.posts:
self.posts[action] = (post_id, post_site, post_type, post_url, [(callback, max_time)] if callback else [])
try:
self.socket.send(action)
except websocket.WebSocketException:
log('error', 'DeletionWatcher failed on sending {}'.format(action))
elif callback:
_, _, _, _, callbacks = self.posts[action]
callbacks.append((callback, max_time))
else:
return
if pickle:
Tasks.do(self._save)
def determine_if_autoflagged(post_url):
"""
Given the URL for a post, determine whether or not it has been autoflagged.
"""
payload = {
'key': GlobalVars.metasmoke_key,
'filter': 'GFGJGHFMHGOLMMJMJJJGHIGOMKFKKILF', # id and autoflagged
'urls': post_url
}
try:
response = Metasmoke.get("/api/v2.0/posts/urls", params=payload).json()
except Exception as e:
log('error', e)
return False, []
# The first report of a URL is the only one that will be autoflagged. MS responses to the
# /posts/urls endpoint have the oldest report last.
first_report_index = len(response["items"]) - 1
if first_report_index > -1 and response["items"][first_report_index]["autoflagged"]:
# get flagger names
id = str(response["items"][first_report_index]["id"])
payload = {'key': GlobalVars.metasmoke_key}
flags = Metasmoke.get("/api/v2.0/posts/" + id + "/flags", params=payload).json()
if len(flags["items"]) > 0:
return True, [user["username"] for user in flags["items"][0]["autoflagged"]["users"]]
return False, []
def setup_websocket(attempt, max_attempts):
try:
ws = websocket.create_connection("wss://qa.sockets.stackexchange.com/")
ws.send("155-questions-active")
return ws
except websocket.WebSocketException:
log('warning', 'WS failed to create websocket connection. Attempt {} of {}.'.format(attempt, max_attempts))
return None
def log_exception(exctype, value, tb):
now = datetime.utcnow()
tr = '\n'.join((traceback.format_tb(tb)))
exception_only = ''.join(traceback.format_exception_only(exctype, value)).strip()
logged_msg = "{exception}\n{now} UTC\n{row}\n\n".format(exception=exception_only, now=now, row=tr)
log('error', logged_msg)
with open("errorLogs.txt", "a") as f:
f.write(logged_msg)
def reset (self):
profile = os.environ ["DCONF_PROFILE"]
assert profile == "trackertest"
# XDG_CONFIG_HOME is useless
dconf_db = os.path.join (os.environ ["HOME"], ".config", "dconf", profile)
if os.path.exists (dconf_db):
log ("[Conf] Removing dconf-profile: " + dconf_db)
os.remove (dconf_db)
def tracker_store_restart_with_new_ontologies (self, ontodir):
self.store.stop ()
if ontodir:
helpers.log ("[Conf] Setting %s - %s" % ("TRACKER_DB_ONTOLOGIES_DIR", ontodir))
os.environ ["TRACKER_DB_ONTOLOGIES_DIR"] = ontodir
try:
self.store.start ()
except dbus.DBusException, e:
raise UnableToBootException ("Unable to boot the store \n(" + str(e) + ")")
def send_message(self, text, length_check=True):
if "no-chat" in sys.argv:
log('info', "Blocked message to {0} due to no-chat setting: {1}".format(self.name, text))
return
if "charcoal-hq-only" not in sys.argv or int(self.id) == 11540:
return rooms.Room.send_message(self, text, length_check)
else:
log('info', "Blocked message to {0} due to charcoal-hq-only setting: {1}".format(self.name, text))
def callback(self):
log('--no callback defined-- %r\n' % self)
(r, w, x) = select.select(self.socks, [], [], 0)
for s in r:
v = s.recv(4096)
if not v:
log('--closed-- %r\n' % self)
self.socks = []
self.ok = False
def check_if_spam_json(json_data):
try:
post = Post(json_data=json_data)
except PostParseError as err:
log('error', 'Parse error {0} when parsing json_data {1!r}'.format(
err, json_data))
return False, '', ''
is_spam, reason, why = check_if_spam(post)
return is_spam, reason, why
def read(url):
try:
response = urlopen(url)
content = response.read()
return json.loads(content.decode('utf8'))
except URLError as e:
log("Could not read data!", str(e))
return {}
def __init__(self, name, t):
if name not in ('master', 'slave'):
helpers.log("HA controller must either be 'master' or 'slave'")
self.t = t
self._name = name
self.rest = HaBsnRestClient(name, t)
def make_api_call_for_site(self, site):
if site not in self.queue:
return
self.queue_modify_lock.acquire()
new_posts = self.queue.pop(site)
store_bodyfetcher_queue()
self.queue_modify_lock.release()
new_post_ids = [int(k) for k in new_posts.keys()]
if GlobalVars.flovis is not None:
for post_id in new_post_ids:
GlobalVars.flovis.stage('bodyfetcher/api_request', site,
post_id, {
'site': site,
'posts': list(new_posts.keys())
})
self.queue_timing_modify_lock.acquire()
post_add_times = [v for k, v in new_posts.items()]
pop_time = datetime.utcnow()
for add_time in post_add_times:
try:
seconds_in_queue = (pop_time - add_time).total_seconds()
if site in self.queue_timings:
self.queue_timings[site].append(seconds_in_queue)
else:
self.queue_timings[site] = [seconds_in_queue]
except KeyError: # XXX: Any other possible exception?
continue # Skip to next item if we've got invalid data or missing values.
store_queue_timings()
self.queue_timing_modify_lock.release()
self.max_ids_modify_lock.acquire()
if site in self.previous_max_ids and max(
new_post_ids) > self.previous_max_ids[site]:
previous_max_id = self.previous_max_ids[site]
intermediate_posts = range(previous_max_id + 1, max(new_post_ids))
# We don't want to go over the 100-post API cutoff, so take the last
# (100-len(new_post_ids)) from intermediate_posts
intermediate_posts = intermediate_posts[(100 - len(new_post_ids)):]
# new_post_ids could contain edited posts, so merge it back in
combined = chain(intermediate_posts, new_post_ids)
# Could be duplicates, so uniquify
posts = list(set(combined))
else:
posts = new_post_ids
try:
if max(new_post_ids) > self.previous_max_ids[site]:
self.previous_max_ids[site] = max(new_post_ids)
store_bodyfetcher_max_ids()
except KeyError:
self.previous_max_ids[site] = max(new_post_ids)
store_bodyfetcher_max_ids()
self.max_ids_modify_lock.release()
log('debug', "New IDs / Hybrid Intermediate IDs for {}:".format(site))
if len(new_post_ids) > 30:
log(
'debug', "{} +{} more".format(
sorted(new_post_ids)[:30],
len(new_post_ids) - 30))
else:
log('debug', sorted(new_post_ids))
if len(new_post_ids) == len(posts):
log('debug', "[ *Identical* ]")
elif len(posts) > 30:
log('debug',
"{} +{} more".format(sorted(posts)[:30],
len(posts) - 30))
else:
log('debug', sorted(posts))
question_modifier = ""
pagesize_modifier = {}
if site == "stackoverflow.com":
# Not all SO questions are shown in the realtime feed. We now
# fetch all recently modified SO questions to work around that.
if self.last_activity_date != 0:
pagesize = "50"
else:
pagesize = "25"
pagesize_modifier = {
'pagesize': pagesize,
'min': str(self.last_activity_date)
}
else:
question_modifier = "/{0}".format(";".join(
[str(post) for post in posts]))
url = "https://api.stackexchange.com/2.2/questions{}".format(
question_modifier)
params = {
'filter':
'!*xq08dCDNr)PlxxXfaN8ntivx(BPlY_8XASyXLX-J7F-)VK*Q3KTJVkvp*',
'key': 'IAkbitmze4B8KpacUfLqkw((',
'site': site
}
params.update(pagesize_modifier)
# wait to make sure API has/updates post data
time.sleep(3)
GlobalVars.api_request_lock.acquire()
# Respect backoff, if we were given one
if GlobalVars.api_backoff_time > time.time():
time.sleep(GlobalVars.api_backoff_time - time.time() + 2)
try:
time_request_made = datetime.now().strftime('%H:%M:%S')
response = requests.get(url, params=params, timeout=20).json()
except (requests.exceptions.Timeout, requests.ConnectionError,
Exception):
# Any failure in the request being made (timeout or otherwise) should be added back to
# the queue.
self.queue_modify_lock.acquire()
if site in self.queue:
self.queue[site].update(new_posts)
else:
self.queue[site] = new_posts
self.queue_modify_lock.release()
GlobalVars.api_request_lock.release()
return
self.api_data_lock.acquire()
add_or_update_api_data(site)
self.api_data_lock.release()
message_hq = ""
if "quota_remaining" in response:
if response[
"quota_remaining"] - GlobalVars.apiquota >= 5000 and GlobalVars.apiquota >= 0:
tell_rooms_with(
"debug",
"API quota rolled over with {0} requests remaining. "
"Current quota: {1}.".format(GlobalVars.apiquota,
response["quota_remaining"]))
sorted_calls_per_site = sorted(
GlobalVars.api_calls_per_site.items(),
key=itemgetter(1),
reverse=True)
api_quota_used_per_site = ""
for site_name, quota_used in sorted_calls_per_site:
sanatized_site_name = site_name.replace(
'.com', '').replace('.stackexchange', '')
api_quota_used_per_site += sanatized_site_name + ": {0}\n".format(
str(quota_used))
api_quota_used_per_site = api_quota_used_per_site.strip()
tell_rooms_with("debug", api_quota_used_per_site)
clear_api_data()
if response["quota_remaining"] == 0:
tell_rooms_with(
"debug", "API reports no quota left! May be a glitch.")
tell_rooms_with("debug",
str(response)) # No code format for now?
if GlobalVars.apiquota == -1:
tell_rooms_with(
"debug", "Restart: API quota is {quota}.".format(
quota=response["quota_remaining"]))
GlobalVars.apiquota = response["quota_remaining"]
else:
message_hq = "The quota_remaining property was not in the API response."
if "error_message" in response:
message_hq += " Error: {} at {} UTC.".format(
response["error_message"], time_request_made)
if "error_id" in response and response["error_id"] == 502:
if GlobalVars.api_backoff_time < time.time(
) + 12: # Add a backoff of 10 + 2 seconds as a default
GlobalVars.api_backoff_time = time.time() + 12
message_hq += " Backing off on requests for the next 12 seconds."
message_hq += " Previous URL: `{}`".format(url)
if "backoff" in response:
if GlobalVars.api_backoff_time < time.time() + response["backoff"]:
GlobalVars.api_backoff_time = time.time() + response["backoff"]
GlobalVars.api_request_lock.release()
if len(message_hq) > 0 and "site is required" not in message_hq:
tell_rooms_with("debug", message_hq.strip())
if "items" not in response:
return
if site == "stackoverflow.com":
items = response["items"]
if len(items) > 0 and "last_activity_date" in items[0]:
self.last_activity_date = items[0]["last_activity_date"]
num_scanned = 0
start_time = time.time()
for post in response["items"]:
pnb = copy.deepcopy(post)
if 'body' in pnb:
pnb['body'] = 'Present, but truncated'
if 'answers' in pnb:
del pnb['answers']
if "title" not in post or "body" not in post:
if GlobalVars.flovis is not None and 'question_id' in post:
GlobalVars.flovis.stage(
'bodyfetcher/api_response/no_content', site,
post['question_id'], pnb)
continue
post['site'] = site
try:
post['edited'] = (post['creation_date'] !=
post['last_edit_date'])
except KeyError:
post[
'edited'] = False # last_edit_date not present = not edited
try:
post_ = Post(api_response=post)
except PostParseError as err:
log('error',
'Error {0} when parsing post: {1!r}'.format(err, post_))
if GlobalVars.flovis is not None and 'question_id' in post:
GlobalVars.flovis.stage('bodyfetcher/api_response/error',
site, post['question_id'], pnb)
continue
num_scanned += 1
is_spam, reason, why = check_if_spam(post_)
if is_spam:
try:
if GlobalVars.flovis is not None and 'question_id' in post:
GlobalVars.flovis.stage(
'bodyfetcher/api_response/spam', site,
post['question_id'], {
'post': pnb,
'check_if_spam': [is_spam, reason, why]
})
handle_spam(post=post_, reasons=reason, why=why)
except Exception as e:
log('error', "Exception in handle_spam:", e)
elif GlobalVars.flovis is not None and 'question_id' in post:
GlobalVars.flovis.stage(
'bodyfetcher/api_response/not_spam', site,
post['question_id'], {
'post': pnb,
'check_if_spam': [is_spam, reason, why]
})
try:
if "answers" not in post:
pass
else:
for answer in post["answers"]:
anb = copy.deepcopy(answer)
if 'body' in anb:
anb['body'] = 'Present, but truncated'
num_scanned += 1
answer["IsAnswer"] = True # Necesssary for Post object
answer[
"title"] = "" # Necessary for proper Post object creation
answer[
"site"] = site # Necessary for proper Post object creation
try:
answer['edited'] = (answer['creation_date'] !=
answer['last_edit_date'])
except KeyError:
answer[
'edited'] = False # last_edit_date not present = not edited
answer_ = Post(api_response=answer, parent=post_)
is_spam, reason, why = check_if_spam(answer_)
if is_spam:
try:
if GlobalVars.flovis is not None and 'answer_id' in answer:
GlobalVars.flovis.stage(
'bodyfetcher/api_response/spam', site,
answer['answer_id'], {
'post': anb,
'check_if_spam':
[is_spam, reason, why]
})
handle_spam(answer_, reasons=reason, why=why)
except Exception as e:
log('error', "Exception in handle_spam:", e)
elif GlobalVars.flovis is not None and 'answer_id' in answer:
GlobalVars.flovis.stage(
'bodyfetcher/api_response/not_spam', site,
answer['answer_id'], {
'post': anb,
'check_if_spam': [is_spam, reason, why]
})
except Exception as e:
log('error', "Exception handling answers:", e)
end_time = time.time()
GlobalVars.posts_scan_stats_lock.acquire()
GlobalVars.num_posts_scanned += num_scanned
GlobalVars.post_scan_time += end_time - start_time
GlobalVars.posts_scan_stats_lock.release()
return
