def send_to_api(self, path, post, access_token):
ctx = ndb.get_context()
try:
resp = yield ctx.urlfetch('https://alpha-api.app.net/stream/0/%s' % path, payload=json.dumps(post), deadline=30,
method='POST', headers={
'Authorization': 'Bearer %s' % access_token,
'Content-Type': 'application/json',
})
except:
logger.exception('Failed to post path: %s data: %s' % (path, post))
raise deferred.SingularTaskFailure()
parsed_resp = json.loads(resp.content)
if resp.status_code == 401:
logger.info('unauthorized')
yield self.handle_unauthorized(parsed_resp, post)
raise deferred.PermanentTaskFailure()
elif resp.status_code == 200:
self.handle_success(parsed_resp, post)
elif resp.status_code == 400:
yield self.handle_bad_response(parsed_resp, post)
raise deferred.PermanentTaskFailure()
elif resp.status_code == 403:
yield self.handle_forbidden(parsed_resp, post)
raise deferred.PermanentTaskFailure()
else:
logger.warn("Couldn't post entry key=%s. Error: %s Post:%s", self.entry_key, parsed_resp, post)
raise deferred.SingularTaskFailure()
def deferred_ratings():
""" This is the deferred ratings table calculation process """
# Disable the in-context cache to save memory
# (it doesn't give any speed advantage for this processing)
Context.disable_cache()
t0 = time.time()
try:
_create_ratings()
except DeadlineExceededError as ex:
# Hit deadline: save the stuff we already have and
# defer a new task to continue where we left off
logging.error(u"Deadline exceeded in ratings, failing permamently")
# Normal return prevents this task from being run again
raise deferred.PermanentTaskFailure()
except Exception as ex:
logging.error(
u"Exception in ratings, failing permanently: {0}".format(ex))
# Avoid having the task retried
raise deferred.PermanentTaskFailure()
t1 = time.time()
logging.info(u"Ratings calculation finished in {0:.2f} seconds".format(t1 -
t0))
StatsModel.log_cache_stats()
StatsModel.clear_cache(
) # Do not maintain the cache in memory between runs
def _put_annotations_batch(annotation_data):
try:
annotation_objects = []
qry = BookRecord.query(
BookRecord.item_id_array.IN(annotation_data.keys()))
keys = qry.fetch(MAX_OBJECTS_PER_BATCH, keys_only=True)
for key in keys:
first_item_id = key.string_id().split('|')[0]
if first_item_id not in annotation_data:
continue
short_text = annotation_data[first_item_id][0]
long_text = annotation_data[first_item_id][1]
anno_key = ndb.Key(BookAnnotation, key.string_id())
annotation = BookAnnotation(key=anno_key,
short=short_text,
long=long_text)
annotation_objects.append(annotation)
ndb.put_multi(annotation_objects)
logging.info("{} annotations put into datastore".format(
len(annotation_objects)))
except Exception as e:
logging.error(e)
raise deferred.PermanentTaskFailure()
def main(self, sequence_num):
"""Main method of the deferred task."""
with Namespace(self._namespace):
logging.info('Job started: %s w/ sequence number %d',
self._job_name, sequence_num)
time_started = time.time()
try:
db.run_in_transaction(DurableJobEntity._start_job,
self._job_name, sequence_num)
result = self.run()
db.run_in_transaction(DurableJobEntity._complete_job,
self._job_name, sequence_num,
transforms.dumps(result),
long(time.time() - time_started))
logging.info('Job completed: %s', self._job_name)
except (Exception, runtime.DeadlineExceededError) as e:
logging.error(traceback.format_exc())
logging.error('Job failed: %s\n%s', self._job_name, e)
db.run_in_transaction(DurableJobEntity._fail_job,
self._job_name, sequence_num,
traceback.format_exc(),
long(time.time() - time_started))
raise deferred.PermanentTaskFailure(e)
def main(self):
"""Main method of the deferred task."""
logging.info('Job started: %s', self._job_name)
time_started = time.time()
old_namespace = namespace_manager.get_namespace()
try:
namespace_manager.set_namespace(self._namespace)
try:
db.run_in_transaction(DurableJobEntity._start_job,
self._job_name)
result = self.run()
db.run_in_transaction(DurableJobEntity._complete_job,
self._job_name, transforms.dumps(result),
long(time.time() - time_started))
logging.info('Job completed: %s', self._job_name)
except (Exception, runtime.DeadlineExceededError) as e:
logging.error(traceback.format_exc())
logging.error('Job failed: %s\n%s', self._job_name, e)
db.run_in_transaction(DurableJobEntity._fail_job,
self._job_name, traceback.format_exc(),
long(time.time() - time_started))
raise deferred.PermanentTaskFailure(e)
finally:
namespace_manager.set_namespace(old_namespace)
def deferred_update():
""" Update all users in the datastore with lowercase nick and full name """
logging.info("Deferred user update starting")
CHUNK_SIZE = 200
count = 0
offset = 0
Context.disable_cache()
try:
q = UserModel.query()
while True:
ulist = []
chunk = 0
for um in q.fetch(CHUNK_SIZE, offset = offset):
chunk += 1
if um.nick_lc is None:
try:
um.nick_lc = um.nickname.lower()
um.name_lc = um.prefs.get("full_name", "").lower() if um.prefs else ""
ulist.append(um)
except Exception as e:
logging.info("Exception in deferred_update() when setting nick_lc: {0}".format(e))
if ulist:
try:
ndb.put_multi(ulist)
count += len(ulist)
except Exception as e:
logging.info("Exception in deferred_update() when updating ndb: {0}".format(e))
if chunk < CHUNK_SIZE:
break
offset += CHUNK_SIZE
except Exception as e:
logging.info("Exception in deferred_update(): {0}, already updated {1} records".format(e, count))
# Do not retry the task
raise deferred.PermanentTaskFailure()
logging.info("Completed updating {0} user records".format(count))
def ExtractDetailsCrawlerIssueTracker(project_name, bug_id):
"""Extract useful information for a given bug."""
logging.debug('Scraping details for bug %s in project %s.', bug_id,
project_name)
phclient = gdata.projecthosting.client.ProjectHostingClient()
try:
query = gdata.projecthosting.client.Query(issue_id=bug_id)
feed = phclient.get_issues(project_name, query=query)
except gdata.client.RequestError, e:
if ('HTTP_X_APPENGINE_TASKRETRYCOUNT' in environ and int(
environ['HTTP_X_APPENGINE_TASKRETRYCOUNT']) < _MAX_RETRIES):
if e.status == 403: # Skip 403 (Unautorized)errors.
logging.info(
'Unautorized to access this issue, skipping: %s, %s',
bug_id, project_name)
# Nuke cache data for private bugs.
url_bug_map.DeleteBugAndMappings(
bug_id, project_name, bugs_util.Provider.ISSUETRACKER)
return
else:
raise BugCrawlerError(
'Error while trying to get details for %s. Error %s' %
(str(bug_id), str(e)))
else:
raise deferred.PermanentTaskFailure(
'Error hit too many times, aborting '
'extracting details for bug %s on project %s. Error: %s' %
(str(bug_id), str(project_name), str(e)))
def main(self, sequence_num):
"""Main method of the deferred task."""
with Namespace(self._namespace):
logging.info('Job started: %s w/ sequence number %d',
self._job_name, sequence_num)
time_started = time.time()
try:
# Check we haven't been canceled before we start.
if self._already_finished(sequence_num):
logging.info(
'Job %s sequence %d already canceled or subsequent '
'run completed; not running this version.',
self._job_name, sequence_num)
return
db.run_in_transaction(DurableJobEntity._start_job,
self._job_name, sequence_num)
result = self.run()
db.run_in_transaction(DurableJobEntity._complete_job,
self._job_name, sequence_num,
transforms.dumps(result),
long(time.time() - time_started))
logging.info('Job completed: %s', self._job_name)
except (Exception, runtime.DeadlineExceededError) as e:
logging.error(traceback.format_exc())
logging.error('Job failed: %s\n%s', self._job_name, e)
db.run_in_transaction(DurableJobEntity._fail_job,
self._job_name, sequence_num,
traceback.format_exc(),
long(time.time() - time_started))
raise deferred.PermanentTaskFailure(e)
def deferred_check_user(user_id):
user = ModelProxy.User.get_by_id(user_id)
if not user:
raise deferred.PermanentTaskFailure("User id {} is invalid.".format(user_id))
unique_email_address(user)
unique_final_submission(user)
unique_group(user)
def insert_gtest_results(build_step_key):
"""Inserts GTest results into the datastore, replacing any existing ones.
Also records used parser version."""
step = BuildStep.get(build_step_key)
log_contents = ''
if step.log_gs:
with cloudstorage.open(step.log_gs) as gs_file:
log_contents = html2text(gs_file.read().decode('utf-8', 'replace'))
else:
try:
blob_reader = blobstore.BlobReader(step.log_stdio)
log_contents = html2text(blob_reader.read().decode(
'utf-8', 'replace'))
except (ValueError, blobstore.BlobNotFoundError) as e:
raise deferred.PermanentTaskFailure(e)
gtest_results = gtest_parser.parse(log_contents)
to_put = []
for fullname, result in gtest_results.iteritems():
# Only store failure results.
if result['is_successful']:
continue
if isinstance(result['log'], unicode):
log = db.Text(result['log'])
else:
log = db.Text(result['log'], encoding='utf-8')
result_entity = GTestResult(
parent=db.Key.from_path('GTestResult', str(step.key())),
build_step=step,
time_finished=step.time_finished,
gtest_parser_version=gtest_parser.VERSION,
is_crash_or_hang=result['is_crash_or_hang'],
fullname=fullname,
run_time_ms=result['run_time_ms'],
log=log)
to_put.append(result_entity)
for chunk in chunks(to_put, BATCH_SIZE):
db.put(chunk)
def tx_parser_version():
step = BuildStep.get(build_step_key)
orig_parser_version = step.gtest_parser_version
if step.gtest_parser_version < gtest_parser.VERSION:
step.gtest_parser_version = gtest_parser.VERSION
step.put()
return (orig_parser_version, step.gtest_parser_version)
_, parser_version = \
db.run_in_transaction_custom_retries(10, tx_parser_version)
query = GTestResult.all(keys_only=True)
query.filter('build_step =', build_step_key)
query.filter('gtest_parser_version <', parser_version)
db.delete(query)
def main(self, sequence_num):
logging.info('Drive job waking up')
job = self.load()
if not job:
raise deferred.PermanentTaskFailure(
'Job object for {} not found!'.format(self._job_name))
if job.has_finished:
return
try:
# pylint: disable=protected-access
drive_manager = drive_api_manager._DriveManager.from_app_context(
self._app_context)
# pylint: enable=protected-access
except errors.NotConfigured:
self.complete(sequence_num)
return
except errors.Misconfigured as error:
logging.error('%s: Drive is misconfigured in %s: %s',
self._job_name, self._app_context.get_title(), error)
raise deferred.PermanentTaskFailure('Job {} failed: {}'.format(
self._job_name, error))
try:
dto = self.get_a_valid_work_item()
except IndexError:
self.complete(sequence_num)
return
try:
logging.info('Starting download of %s', dto.title)
drive_manager.download_file(dto)
logging.info('Finished download of %s', dto.title)
except Exception as error: #pylint: disable=broad-except
# Normally errors.Error covers everything, but this covers the
# possibility of an unexpected parse error.
logging.info('Failed to sync %s (%s) from drive: %s', dto.title,
dto.key, error)
deferred.defer(self.main, sequence_num)
def update_majors():
try:
ndb.delete_multi(Major.get_majors().iter(keys_only=True))
all_majors = get_majors()
logging.info("All majors extracted: " + str(len(all_majors)))
for major in all_majors:
logging.info(major['discipline'] + " was created")
deferred.defer(create_major, major)
except Exception as e:
logging.error(e)
raise deferred.PermanentTaskFailure()
def reparse_suppression_results(build_step_key, _build_step_name):
step = BuildStep.get(build_step_key)
log_contents = ''
if step.log_gs:
with cloudstorage.open(step.log_gs) as gs_file:
log_contents = html2text(gs_file.read().decode('utf-8', 'replace'))
else:
try:
blob_reader = blobstore.BlobReader(step.log_stdio)
log_contents = html2text(blob_reader.read().decode(
'utf-8', 'replace'))
except (ValueError, blobstore.BlobNotFoundError), e:
raise deferred.PermanentTaskFailure(e)
def _run(self):
self.job_dump.update_status(self.MAPPING)
try:
self.job_mapper.run()
self.job_dump.update_status(self.REDUCING)
self.job_reducer.run(self.job_dump.map_result)
if not self.save_output:
self.clean_up()
self.job_dump.update_status(self.SUCCESS)
except JobException as e:
self.clean_up()
self.job_dump.error = '%s: %s' % (e.job_type, e.message)
self.job_dump.update_status(self.FAILURE)
raise deferred.PermanentTaskFailure()
def create_suggestions_json(suggestions_key):
suggestions = suggestions_key.get()
item_ids = suggestions.key.string_id()
assert isinstance(suggestions, SuggestionsRecord)
if not suggestions.completed:
logging.warning(
"create_suggestions_json for {} called although "
"suggestions are still not completed.".format(item_ids))
# deferred.defer(create_suggestions_json, suggestions_key,
# _countdown=3)
return
json_object = {
'version': Suggester.VERSION,
'item_ids': item_ids,
'job_started': suggestions.job_started.isoformat()
}
original_book = suggestions.original_book.get()
if not original_book:
raise deferred.PermanentTaskFailure(
"Original book not found for {} "
"when creating JSON.".format(item_ids))
assert isinstance(original_book, BookRecord)
json_object['original_book'] = {
'author': original_book.author,
'title': original_book.title,
'year': original_book.year
}
json_object['status'] = 'completed'
json_object['suggestions'] = []
book_records = ndb.get_multi(suggestions.books)
assert len(book_records) == len(suggestions.books_prediction)
for i in xrange(len(book_records)):
book = book_records[i]
assert isinstance(book, BookRecord)
json_object['suggestions'].append({
'author':
book.author,
'title':
book.title,
'item_ids':
book.key.string_id(),
'prediction':
suggestions.books_prediction[i]
})
precomputed_json = json.dumps(json_object, indent=2)
suggestions.json = precomputed_json
suggestions.put()
logging.info("Created and saved JSON for {}".format(item_ids))
def post(self):
"""Adds a BigQuery row to Datastore and streams it using a deferred task.
Raises:
deferred.PermanentTaskFailure: if we encounter any exception to avoid
adding duplicate rows during task retries.
"""
payload = pickle.loads(self.request.body)
bigquery_row_model.BigQueryRow.add(**payload)
try:
if bigquery_row_model.BigQueryRow.threshold_reached():
deferred.defer(bigquery_row_model.BigQueryRow.stream_rows)
else:
logging.info('Not streaming rows, thresholds not met.')
except Exception as e: # pylint: disable=broad-except
raise deferred.PermanentTaskFailure(
'Exception caught for BigQuery streaming: %s.' % e)
def wrapper(*args, **kwargs):
"""Wrapper for managed task decorator."""
status_entity = bootstrap_status_model.BootstrapStatus.get_or_insert(
task_function.__name__)
status_entity.description = _TASK_DESCRIPTIONS.get(
task_function.__name__, task_function.__name__)
status_entity.timestamp = datetime.datetime.utcnow()
try:
task_function(*args, **kwargs)
status_entity.success = True
status_entity.details = None
status_entity.put()
except Exception as e:
status_entity.success = False
status_entity.details = '{} {}'.format(str(type(e)), str(e))
status_entity.put()
raise deferred.PermanentTaskFailure(
'Task {} failed; error: {}'.format(task_function.__name__,
status_entity.details))
def SpawnDetailsCrawlersIssueTracker(recent_issues,
project_name,
skip_recent_check=False):
"""Queues the tasks to do the actual crawling for recent updates."""
count = 0
try:
for issue in recent_issues:
bug_id = issue['id']
logging.info(
'Adding crawler to the queue for issue_id %s, project: %s.',
bug_id, project_name)
end = bug_id.find('/')
if end > 0:
bug_id = bug_id[0:end]
bug = bugs.GetBug(bug_id=bug_id,
project=project_name,
provider=bugs_util.Provider.ISSUETRACKER)
if bug:
if not skip_recent_check and bug.last_update == issue[
'updated']:
logging.info('Bug %s is up-to-date.',
bug.key().id_or_name())
count += 1
continue
else:
logging.info('Bug %s needs to be updated.',
bug.key().id_or_name())
else:
logging.info('Bug %s seems to be a new issue.', bug_id)
deferred.defer(ExtractDetailsCrawlerIssueTracker,
project_name,
bug_id,
_queue='find-bugs-queue')
count += 1
except DeadlineExceededError:
remaining = recent_issues[count:]
deferred.defer(SpawnDetailsCrawlersIssueTracker, remaining,
project_name)
deferred.PermanentTaskFailure(
'Deadline exceeded, started a new SpawnDetailsCrawler'
' for the remaining %d urls.' % len(remaining))
return
def wrapper(*args, **kwargs):
status_entity = bootstrap_status_model.BootstrapStatus.get_or_insert(
task_function.__name__)
docstring_first_line = re.search('^.*', task_function.__doc__)
if docstring_first_line:
status_entity.description = docstring_first_line.group(0)
else:
status_entity.description = task_function.__name__
status_entity.timestamp = datetime.datetime.utcnow()
try:
task_function(*args, **kwargs)
status_entity.success = True
status_entity.details = None
status_entity.put()
except Exception as e:
status_entity.success = False
status_entity.details = '{} {}'.format(str(type(e)), str(e))
status_entity.put()
raise deferred.PermanentTaskFailure(
'Task {} failed; error: {}'.format(task_function.__name__,
status_entity.details))
def update_annotations_from_csv(filename="anotace.csv"):
logging.info("Starting to update from CSV")
my_default_retry_params = gcs.RetryParams(initial_delay=0.2,
max_delay=5.0,
backoff_factor=2,
max_retry_period=15)
gcs.set_default_retry_params(my_default_retry_params)
# bucket_name = os.environ.get('BUCKET_NAME',
# app_identity.get_default_gcs_bucket_name())
bucket_name = "tisic-knih.appspot.com"
bucket = '/' + bucket_name
filename = bucket + '/' + filename
try:
if is_dev_server():
gcs_file = open("anotace.csv")
else:
gcs_file = gcs.open(filename)
# gcs_file.seek(1000)
r = ucsv.reader(gcs_file, encoding='utf-8')
r.next() # Skip first line (header).
annotation_data = {}
for row_number, row in enumerate(r):
if row_number % 10000 == 0:
logging.info("Processing row number {} of CSV file. ".format(
row_number))
item_id = row[0]
short_text = row[1]
long_text = row[2]
annotation_data[item_id] = (short_text, long_text)
if len(annotation_data) > MAX_OBJECTS_PER_BATCH:
deferred.defer(_put_annotations_batch, annotation_data)
annotation_data = {}
deferred.defer(_put_annotations_batch, annotation_data)
gcs_file.close()
except Exception as e:
logging.error(e)
raise deferred.PermanentTaskFailure()
def __email_task():
try:
logging.info("Email task started")
yesterday = datetime.datetime.now() - datetime.timedelta(days=1)
stale_game_q = Game.query(Game.active == True,
Game.updated < yesterday)
stale_games = {}
for game in stale_game_q.iter():
player_key = game.active_player_key
if not player_key in stale_games:
stale_games[player_key] = []
stale_games[player_key].append(game.key.id())
__send_emails(stale_games)
logging.info("Email task complete")
except Exception as e:
# It can be dangerous to allow exceptions to bubble up to the job
# scheduler in this case -- depending on what happens, it can get
# stuck in a retry loop and start spamming our users.
# Instead, log the problem and raise a "stop doing that!" exception.
msg = "Unhandled exception during email task, aborting"
logging.exception(msg)
raise deferred.PermanentTaskFailure(msg)
def CommitChangeSet(change_key):
"""Attempts to commit and delete a given RuleChangeSet."""
change = change_key.get()
if change is None:
logging.info('Change no longer exists. (already committed?)')
return
logging.info('Committing a %s change set of %s rules for blockable %s',
change.change_type, len(change.rule_keys),
change.blockable_key.id())
blockable = change.blockable_key.get()
rules = ndb.get_multi(change.rule_keys)
if change.change_type == constants.RULE_POLICY.WHITELIST:
change_func = _Whitelist
elif change.change_type == constants.RULE_POLICY.BLACKLIST:
change_func = _Blacklist
elif change.change_type == constants.RULE_POLICY.REMOVE:
change_func = _Remove
elif change.change_type in constants.RULE_POLICY.SET_INSTALLER:
change_func = _ChangeInstallerState
else:
raise NotImplementedError
try:
change_func(blockable, rules)
except api.RequestError:
# For normal request errors, rely on the builtin task queue retry settings.
raise
except Exception as e: # pylint: disable=broad-except
# For all other (likely fatal) errors, make sure the task doesn't retry.
raise deferred.PermanentTaskFailure(repr(e))
else:
for rule in rules:
rule.is_committed = True
ndb.put_multi(rules)
change.key.delete()
def create_major(obj):
try:
major_info = parse_url(obj['link'])
parsed = True
if (len(major_info['first_year']['first_semester']) == 0
and len(major_info['first_year']['second_semester']) == 0
and len(major_info['second_year']['first_semester']) == 0
and len(major_info['second_year']['second_semester']) == 0
and len(major_info['third_year']['first_semester']) == 0
and len(major_info['third_year']['second_semester']) == 0
and len(major_info['fourth_year']['first_semester']) == 0
and len(major_info['fourth_year']['second_semester']) == 0):
parsed = False
obj['major_description'] = json.dumps(major_info)
obj['parsed'] = parsed
Major.add_new(obj)
except Exception as e:
logging.error(e)
raise deferred.PermanentTaskFailure()
def read_ophan(section_id=None):
client = memcache.Client()
last_read = client.get(section_id + ".epoch_seconds")
if last_read and fresh(last_read): return
ophan_json = None
if section_id == 'all':
ophan_json = ophan.popular()
else:
ophan_json = ophan.popular(section_id=section_id)
if not ophan_json:
raise deferred.PermanentTaskFailure()
ophan_data = json.loads(ophan_json)
resolved_stories = [resolve_content(entry['url']) for entry in ophan_data]
resolved_stories = [
story for story in resolved_stories if not story == None
]
client = memcache.Client()
base_key = 'all'
if section_id: base_key = section_id
if len(resolved_stories) > 0:
client.set(base_key, json.dumps(resolved_stories))
client.set(base_key + '.epoch_seconds', time.time())
logging.info("Updated data for section %s; listing %d stories" %
(section_id, len(resolved_stories)))
def main(self):
"""Main method of the deferred task."""
logging.info('Job started: %s', self._job_name)
time_started = time.time()
old_namespace = namespace_manager.get_namespace()
try:
namespace_manager.set_namespace(self._namespace)
try:
result = self.run()
DurableJobEntity.complete_job(self._job_name,
json.dumps(result),
long(time.time() - time_started))
logging.info('Job completed: %s', self._job_name)
except Exception as e:
logging.error(traceback.format_exc())
logging.error('Job failed: %s\n%s', self._job_name, e)
DurableJobEntity.fail_job(self._job_name,
traceback.format_exc(),
long(time.time() - time_started))
raise deferred.PermanentTaskFailure(e)
finally:
namespace_manager.set_namespace(old_namespace)
# Nuke cache data for private bugs.
url_bug_map.DeleteBugAndMappings(
bug_id, project_name, bugs_util.Provider.ISSUETRACKER)
return
else:
raise BugCrawlerError(
'Error while trying to get details for %s. Error %s' %
(str(bug_id), str(e)))
else:
raise deferred.PermanentTaskFailure(
'Error hit too many times, aborting '
'extracting details for bug %s on project %s. Error: %s' %
(str(bug_id), str(project_name), str(e)))
if not feed or not feed.entry:
raise deferred.PermanentTaskFailure(
'Failed to fetch full details for bug %s', bug_id)
entry = feed.entry[0]
urls = []
if entry.title.text:
urls = [(u, url_bug_map.UrlPosition.TITLE)
for u in url_util.ExtractUrls(entry.title.text)]
if entry.content.text:
urls.extend([(u, url_bug_map.UrlPosition.MAIN)
for u in url_util.ExtractUrls(entry.content.text)])
comments = GetComments(project_name, bug_id, phclient)
comments_text = GetTextInComments(comments)
if comments_text:
urls.extend([(u, url_bug_map.UrlPosition.COMMENTS)
for u in url_util.ExtractUrls(comments_text)])
def _send_to_keen(survey_key, course_key, answer_keys):
events = []
student = Student.query(Student.user == survey_key.get().participant).get()
course = course_key.get()
answers = ndb.get_multi(answer_keys)
for answer in answers:
question = answer.question.get()
if answer.string_value != '':
response = urlfetch.fetch(
'http://text-processing.com/api/sentiment',
payload=urllib.urlencode({'text': answer.string_value}),
method=urlfetch.POST)
# If we've been throttled, just give up and die
if response.status_code == 503:
continue
elif response.status_code != 200:
raise deferred.PermanentTaskFailure()
sentiment = json.loads(response.content)
answer.sentiment = sentiment['label']
answer.put()
lecturer = course.lecturer.get()
event = {
'question_key': question.key.urlsafe(),
'survey_key': answer.key.parent().urlsafe(),
'course_key': course.key.urlsafe(),
'course': {
'name': course.course.get().name,
'department': course.course.get().department.urlsafe(),
'faculty': course.course.get().faculty.urlsafe(),
'school': course.course.get().faculty.get().school.urlsafe(),
},
'question_number': question.number,
'question_text': question.question,
'lecturer': {
'key': lecturer.key.urlsafe(),
'name': lecturer.name,
'department': lecturer.department.get().name,
'faculty': lecturer.department.get().faculty.get().name,
},
'student': {
'key': student.key.urlsafe(),
'age': student.calculate_age(),
'gender': student.gender,
'status': student.status,
'year': student.year,
},
}
if question.question_type == 'closed':
event['response'] = answer.int_value
else:
event['sentiment'] = answer.sentiment
events.append(event)
keen.add_events({'answers': events})
def _send_mail_task(cls,
notification_key,
payload_key,
test_send_mail_fn=None):
exception = None
failed_permanently = False
now = datetime.datetime.utcnow()
# pylint: disable=unbalanced-tuple-unpacking,unpacking-non-sequence
notification, payload = db.get([notification_key, payload_key])
# pylint: enable=unbalanced-tuple-unpacking,unpacking-non-sequence
send_mail_fn = (test_send_mail_fn
if test_send_mail_fn else mail.send_mail)
sent = False
COUNTER_SEND_MAIL_TASK_STARTED.inc()
if not notification:
COUNTER_SEND_MAIL_TASK_FAILED_PERMANENTLY.inc()
raise deferred.PermanentTaskFailure('Notification missing: ' +
str(notification_key))
if not payload:
COUNTER_SEND_MAIL_TASK_FAILED_PERMANENTLY.inc()
raise deferred.PermanentTaskFailure('Payload missing: ' +
str(payload_key))
policy = _RETENTION_POLICIES.get(notification._retention_policy)
if not policy:
COUNTER_SEND_MAIL_TASK_FAILED_PERMANENTLY.inc()
raise deferred.PermanentTaskFailure('Unknown retention policy: ' +
notification._retention_policy)
if (cls._done(notification) or cls._failed(notification)
or cls._sent(notification)):
COUNTER_SEND_MAIL_TASK_SKIPPED.inc()
COUNTER_SEND_MAIL_TASK_SUCCESS.inc()
return
if notification._recoverable_failure_count > _RECOVERABLE_FAILURE_CAP:
message = (
'Recoverable failure cap (%s) exceeded for notification with '
'key %s') % (_RECOVERABLE_FAILURE_CAP, str(notification.key()))
_LOG.error(message)
permanent_failure = deferred.PermanentTaskFailure(message)
try:
COUNTER_SEND_MAIL_TASK_RECORD_FAILURE_CALLED.inc()
cls._record_failure(notification,
payload,
permanent_failure,
dt=now,
permanent=True,
policy=policy)
COUNTER_SEND_MAIL_TASK_RECORD_FAILURE_SUCCESS.inc()
# Must be vague. pylint: disable=broad-except
except Exception, e:
_LOG.error(
cls._get_record_failure_error_message(
notification, payload, e))
COUNTER_SEND_MAIL_TASK_RECORD_FAILURE_FAILED.inc()
COUNTER_SEND_MAIL_TASK_FAILED_PERMANENTLY.inc()
COUNTER_SEND_MAIL_TASK_FAILURE_CAP_EXCEEDED.inc()
raise permanent_failure
def _run_stats(from_time, to_time):
""" Runs a process to update user statistics and Elo ratings """
logging.info(u"Generating stats from {0} to {1}".format(
from_time, to_time))
if from_time is None or to_time is None:
# Time range must be specified
return
if from_time >= to_time:
# Null time range
return
# Iterate over all finished games within the time span in temporal order
q = GameModel.query(ndb.AND(GameModel.ts_last_move > from_time, GameModel.ts_last_move <= to_time)) \
.order(GameModel.ts_last_move) \
.filter(GameModel.over == True)
# The accumulated user statistics
users = dict()
def _init_stat(user_id, robot_level):
""" Returns the newest StatsModel instance available for the given user """
return StatsModel.newest_before(from_time, user_id, robot_level)
cnt = 0
ts_last_processed = None
try:
# Use i as a progress counter
i = 0
for gm in iter_q(q, chunk_size=250):
i += 1
lm = Alphabet.format_timestamp(gm.ts_last_move or gm.timestamp)
p0 = None if gm.player0 is None else gm.player0.id()
p1 = None if gm.player1 is None else gm.player1.id()
robot_game = (p0 is None) or (p1 is None)
if robot_game:
rl = gm.robot_level
else:
rl = 0
s0 = gm.score0
s1 = gm.score1
if (s0 == 0) and (s1 == 0):
# When a game ends by resigning immediately,
# make sure that the weaker player
# doesn't get Elo points for a draw; in fact,
# ignore such a game altogether in the statistics
continue
if p0 is None:
k0 = "robot-" + str(rl)
else:
k0 = p0
if p1 is None:
k1 = "robot-" + str(rl)
else:
k1 = p1
if k0 in users:
urec0 = users[k0]
else:
users[k0] = urec0 = _init_stat(p0, rl if p0 is None else 0)
if k1 in users:
urec1 = users[k1]
else:
users[k1] = urec1 = _init_stat(p1, rl if p1 is None else 0)
# Number of games played
urec0.games += 1
urec1.games += 1
if not robot_game:
urec0.human_games += 1
urec1.human_games += 1
# Total scores
urec0.score += s0
urec1.score += s1
urec0.score_against += s1
urec1.score_against += s0
if not robot_game:
urec0.human_score += s0
urec1.human_score += s1
urec0.human_score_against += s1
urec1.human_score_against += s0
# Wins and losses
if s0 > s1:
urec0.wins += 1
urec1.losses += 1
elif s1 > s0:
urec1.wins += 1
urec0.losses += 1
if not robot_game:
if s0 > s1:
urec0.human_wins += 1
urec1.human_losses += 1
elif s1 > s0:
urec1.human_wins += 1
urec0.human_losses += 1
# Find out whether players are established or beginners
est0 = urec0.games > ESTABLISHED_MARK
est1 = urec1.games > ESTABLISHED_MARK
# Save the Elo point state used in the calculation
gm.elo0, gm.elo1 = urec0.elo, urec1.elo
# Compute the Elo points of both players
adj = _compute_elo((urec0.elo, urec1.elo), s0, s1, est0, est1)
# When an established player is playing a beginning (provisional) player,
# leave the Elo score of the established player unchanged
# Adjust player 0
if est0 and not est1:
adj = (0, adj[1])
gm.elo0_adj = adj[0]
urec0.elo += adj[0]
# Adjust player 1
if est1 and not est0:
adj = (adj[0], 0)
gm.elo1_adj = adj[1]
urec1.elo += adj[1]
# If not a robot game, compute the human-only Elo
if not robot_game:
gm.human_elo0, gm.human_elo1 = urec0.human_elo, urec1.human_elo
adj = _compute_elo((urec0.human_elo, urec1.human_elo), s0, s1,
est0, est1)
# Adjust player 0
if est0 and not est1:
adj = (0, adj[1])
gm.human_elo0_adj = adj[0]
urec0.human_elo += adj[0]
# Adjust player 1
if est1 and not est0:
adj = (adj[0], 0)
gm.human_elo1_adj = adj[1]
urec1.human_elo += adj[1]
# Save the game object with the new Elo adjustment statistics
gm.put()
# Save the last processed timestamp
ts_last_processed = lm
cnt += 1
# Report on our progress
if i % 1000 == 0:
logging.info(u"Processed {0} games".format(i))
except DeadlineExceededError as ex:
# Hit deadline: save the stuff we already have and
# defer a new task to continue where we left off
logging.info(
u"Deadline exceeded in stats loop after {0} games and {1} users".
format(cnt, len(users)))
logging.info(u"Resuming from timestamp {0}".format(ts_last_processed))
if ts_last_processed is not None:
_write_stats(ts_last_processed, users)
deferred.defer(deferred_stats,
from_time=ts_last_processed or from_time,
to_time=to_time)
# Normal return prevents this task from being run again
return
except Exception as ex:
logging.info(u"Exception in stats loop: {0}".format(ex))
# Avoid having the task retried
raise deferred.PermanentTaskFailure()
# Completed without incident
logging.info(
u"Normal completion of stats for {1} games and {0} users".format(
len(users), cnt))
_write_stats(to_time, users)
def check_bq_job(job_id, item_ids, suggestions_key, page_token):
bq = BigQueryClient()
logging.info("Polling suggestion job {}.".format(job_id))
# TODO: catch 404 errors for jobs created 24+ hours ago, retry with new jobid
try:
bq_json = bq.get_async_job_results(job_id, page_token,
MAX_RESULTS_PER_SUGGESTIONS_QUERY)
except Exception as e:
logging.error("Error from BigQuery with item_id={}.".format(item_ids))
raise deferred.PermanentTaskFailure(e)
if not bq_json['jobComplete']:
logging.info("- job not completed yet.")
deferred.defer(check_bq_job,
job_id,
item_ids,
suggestions_key,
"",
_countdown=5)
return
if not 'rows' in bq_json:
logging.error(
u"Invalid json for BigQueryTable. Job for {} is probably "
u"invalid (bad item_id?).\n"
u"JSON:\n"
u"{}".format(item_ids, bq_json))
raise deferred.PermanentTaskFailure("No rows in BigQuery response for "
"{}.".format(item_ids))
table = BigQueryTable(bq_json)
item_ids_array = item_ids.split('|')
# Get the consolidated book for each item_id
suggestions = suggestions_key.get()
assert isinstance(suggestions, SuggestionsRecord)
for row in table.data:
item_id = row[0]
prediction = float(row[1])
if item_id in item_ids_array:
continue # Original book.
consolidated_book_key = BookRecord.query(
BookRecord.item_id_array == item_id).get(keys_only=True)
if not consolidated_book_key:
logging.info(
"No consolidated book with item_id '{}' found.".format(
item_id))
continue
if not consolidated_book_key in suggestions.books:
suggestions.books.append(consolidated_book_key)
suggestions.books_prediction.append(prediction)
if len(suggestions.books) >= 1000:
break
next_page_token = bq_json.get('pageToken', "")
if next_page_token != "" and len(suggestions.books) < 1000:
suggestions.put()
deferred.defer(check_bq_job, job_id, item_ids, suggestions_key,
next_page_token)
logging.info("Suggestions for item_ids '{}' partly fetched. "
"Running again.".format(item_ids))
else:
suggestions.completed = True
suggestions.json_generation_started = True
suggestions.put()
logging.info(
"Suggestions for item_ids '{}' completed and saved.".format(
item_ids))
deferred.defer(create_suggestions_json, suggestions.key)
