def org_timeseries(obj, org_id=None, metrics_lookup=None, commit=True):
"""
Ingest Timeseries Metrics for an organization.
"""
cmd_kwargs = {"org_id": org_id}
# parse datetime.
if "datetime" not in obj:
cmd_kwargs["datetime"] = dates.floor_now(unit="hour", value=1).isoformat()
else:
ds = obj.pop("datetime")
dt = dates.parse_iso(ds)
cmd_kwargs["datetime"] = dates.floor(dt, unit="hour", value=1).isoformat()
metrics = ingest_util.prepare_metrics(obj, metrics_lookup, valid_levels=["org", "all"], check_timeseries=True)
# upsert command
cmd = """SELECT upsert_org_metric_timeseries(
{org_id},
'{datetime}',
'{metrics}')
""".format(
metrics=obj_to_json(metrics), **cmd_kwargs
)
if commit:
try:
db.session.execute(cmd)
except Exception as err:
raise RequestError(err.message)
cmd_kwargs["metrics"] = metrics
return cmd_kwargs
return cmd
def setup(self):
# get max publish date of the last time we ran this.
max_date_last_run = self.last_job.get('max_date_last_run', None)
if max_date_last_run:
self.max_date_last_run = dates.parse_iso(max_date_last_run)
else:
self.max_date_last_run = None
def setup(self):
# get max publish date of the last time we ran this.
max_date_last_run = self.last_job.get('max_date_last_run', None)
if max_date_last_run:
self.max_date_last_run = dates.parse_iso(max_date_last_run)
else:
self.max_date_last_run = None
self.log.info('Got max date: {}'.format(self.max_date_last_run))
self.publish_dates = []
def valid_datetime(self, key, opt):
"""
Validate a iso-datetime option.
"""
v_opt = dates.parse_iso(opt)
if not v_opt:
return RecipeSchemaError(
"{} should be a 'datetime' field but was passed '{}'."
.format(key, opt))
return v_opt
def valid_datetime(self, key, opt):
"""
Validate a iso-datetime option.
"""
v_opt = dates.parse_iso(opt)
if not v_opt:
return RecipeSchemaError(
"{} should be a 'datetime' field but was passed '{}'."
.format(key, opt))
return v_opt
def arg_date(name, default=None):
""" Fetch a query argument, as a datetime object."""
v = request.args.get(name, '')
if not len(v):
return default
v = dates.parse_iso(v)
if not v:
raise RequestError(
'Invalid value for "{}". '
'It should be an iso8601 datetime string.'.format(name))
return v
def arg_date(name, default=None):
""" Fetch a query argument, as a datetime object."""
v = request.args.get(name, '')
if not len(v):
return default
v = dates.parse_iso(v)
if not v:
raise RequestError('Invalid value for "{}". '
'It should be an iso8601 datetime string.'
.format(name))
return v
def arg_list(name, default=None, typ=str, exclusions=False):
""" get a comma-separated list of args, asserting a type.
includes the ability to parse out exclusions via '!' or '-' prefix"""
type_string = str(typ).split("'")[1]
include_values = []
exclude_values = []
vv = arg_str(name, default=None)
if not vv:
if exclusions:
return default, default
return default
for i, value in enumerate(vv.split(','), start=1):
if value.startswith('!') or value.startswith('-'):
value = value[1:]
exclude = True
else:
exclude = False
try:
if 'date' in type_string:
v = dates.parse_iso(value)
if not v:
raise ValueError
else:
v = typ(value)
except:
raise RequestError(
'"{}", element {} of "{}" is invalid. '
'It should be a {} type.'
.format(value), i, name, type_string)
if exclusions:
if exclude:
exclude_values.append(v)
else:
include_values.append(v)
else:
include_values.append(v)
if exclusions:
return include_values, exclude_values
return include_values
def arg_list(name, default=None, typ=str, exclusions=False):
""" get a comma-separated list of args, asserting a type.
includes the ability to parse out exclusions via '!' or '-' prefix"""
type_string = str(typ).split("'")[1]
include_values = []
exclude_values = []
vv = arg_str(name, default=None)
if not vv:
if exclusions:
return default, default
return default
for i, value in enumerate(vv.split(','), start=1):
if value.startswith('!') or value.startswith('-'):
value = value[1:]
exclude = True
else:
exclude = False
try:
if 'date' in type_string:
v = dates.parse_iso(value)
if not v:
raise ValueError
else:
v = typ(value)
except:
raise RequestError('"{}", element {} of "{}" is invalid. '
'It should be a {} type.'.format(
value, i, name, type_string))
if exclusions:
if exclude:
exclude_values.append(v)
else:
include_values.append(v)
else:
include_values.append(v)
if exclusions:
return include_values, exclude_values
return include_values
def prepare_date(o, field):
"""
Prepare a date
"""
if field not in o:
return None
if o[field] is None:
return None
dt = dates.parse_iso(o[field])
if not dt:
raise RequestError('{}: {} is an invalid date.'.format(
field, o[field]))
return dt
def get(self, *args, **kw):
"""
The main get/cache function.
"""
# get a custom ttl, fallback on default
ttl = kw.pop('ttl', self.ttl)
# format the key
key = self.format_key(*args, **kw)
# last modified key
lm_key = "{}:last_modified".format(key)
# attempt to get the object from redis
if not self.debug:
obj = self.redis.get(key)
else:
obj = None
# if it doesn't exist, proceed with work
if not obj:
# not cached
is_cached = False
obj = self.work(*args, **kw)
# if the worker returns None, break out
if not obj:
return CacheResponse(key, obj, None, False)
# set the object in redis at the specified
# key with the specified ttl
self.redis.set(key, self.serialize(obj), ex=ttl)
# set the last modified time
last_modified = dates.now()
self.redis.set(lm_key, last_modified.isoformat(), ex=ttl)
else:
# is cached
is_cached = True
# if it does exist, deserialize it.
obj = self.deserialize(obj)
# get the cached last modified time
last_modified = dates.parse_iso(self.redis.get(lm_key))
return CacheResponse(key, obj, last_modified, is_cached)
def _prepare_metric_date(obj):
"""
Round a metric to a configurable unit.
"""
u = settings.METRICS_MIN_DATE_UNIT
v = settings.METRICS_MIN_DATE_VALUE
# set current time if no time exists.
if 'datetime' not in obj:
return dates.floor_now(unit=u, value=v).isoformat()
ds = obj.pop('datetime')
dt = dates.parse_iso(ds, enforce_tz=True)
return dates.floor(dt, unit=u, value=v).isoformat()
def run(self):
"""
Count shares for all content items.
"""
for content_item in self.api.orgs.simple_content():
created = dates.parse_iso(content_item['created'])
if created < self.max_age:
continue
url = content_item.get('url')
if url:
data = shares.count(url)
data.pop('url', None)
data['content_item_id'] = content_item.get('id')
yield data
def prepare_date(o, field):
"""
Prepare a date
"""
if field not in o:
return None
if o[field] is None:
return None
dt = dates.parse_iso(o[field])
if not dt:
raise RequestError(
'{}: {} is an invalid date.'
.format(field, o[field]))
return dt
def get(self, *args, **kw):
"""
The main get/cache function.
"""
# get a custom ttl, fallback on default
ttl = kw.pop('ttl', self.ttl)
# format the key
key = self.format_key(*args, **kw)
# last modified key
lm_key = "{}:last_modified".format(key)
# attempt to get the object from redis
if not self.debug:
obj = self.redis.get(key)
else:
obj = None
# if it doesn't exist, proceed with work
if not obj:
# not cached
is_cached = False
obj = self.work(*args, **kw)
# if the worker returns None, break out
if not obj:
return CacheResponse(key, obj, None, False)
# set the object in redis at the specified
# key with the specified ttl
self.redis.set(key, self.serialize(obj), ex=ttl)
# set the last modified time
last_modified = dates.now()
self.redis.set(lm_key, last_modified.isoformat(), ex=ttl)
else:
# is cached
is_cached = True
# if it does exist, deserialize it.
obj = self.deserialize(obj)
# get the cached last modified time
last_modified = dates.parse_iso(self.redis.get(lm_key))
return CacheResponse(key, obj, last_modified, is_cached)
def _prepare_metric_date(obj):
"""
Round a metric to a configurable unit.
"""
u = settings.METRICS_MIN_DATE_UNIT
v = settings.METRICS_MIN_DATE_VALUE
# set current time if no time exists.
if 'datetime' not in obj:
return dates.floor_now(unit=u, value=v).isoformat()
ds = obj.pop('datetime')
dt = dates.parse_iso(ds, enforce_tz=True)
return dates.floor(dt, unit=u, value=v).isoformat()
def content_timeseries(obj,
org_id=None,
metrics_lookup=None,
content_item_ids=None,
commit=True):
"""
Ingest Timeseries Metrics for a content item.
"""
# if not content_item_id or not org or not metrics_lookup:
# raise RequestError('Missing required kwargs.')
content_item_id = obj.pop('content_item_id')
if not content_item_id:
raise RequestError('Object is missing a "content_item_id"')
if not content_item_id in content_item_ids:
raise RequestError(
'Content Item with ID {} doesnt exist'.format(content_item_id))
cmd_kwargs = {"org_id": org_id, "content_item_id": content_item_id}
# parse datetime.
if 'datetime' not in obj:
cmd_kwargs['datetime'] = dates.floor_now(unit='hour',
value=1).isoformat()
else:
ds = obj.pop('datetime')
dt = dates.parse_iso(ds)
cmd_kwargs['datetime'] = dates.floor(dt, unit='hour',
value=1).isoformat()
metrics = ingest_util.prepare_metrics(obj,
metrics_lookup,
valid_levels=['content_item', 'all'],
check_timeseries=True)
# upsert command
cmd = """SELECT upsert_content_metric_timeseries(
{org_id},
{content_item_id},
'{datetime}',
'{metrics}')
""".format(metrics=obj_to_json(metrics), **cmd_kwargs)
if commit:
try:
db.session.execute(cmd)
except Exception as err:
raise RequestError(err.message)
cmd_kwargs['metrics'] = metrics
return cmd
def content_timeseries(obj, org_id=None, metrics_lookup=None, content_item_ids=None, commit=True):
"""
Ingest Timeseries Metrics for a content item.
"""
# if not content_item_id or not org or not metrics_lookup:
# raise RequestError('Missing required kwargs.')
content_item_id = obj.pop("content_item_id")
if not content_item_id:
raise RequestError('Object is missing a "content_item_id"')
if not content_item_id in content_item_ids:
raise RequestError("Content Item with ID {} doesnt exist".format(content_item_id))
cmd_kwargs = {"org_id": org_id, "content_item_id": content_item_id}
# parse datetime.
if "datetime" not in obj:
cmd_kwargs["datetime"] = dates.floor_now(unit="hour", value=1).isoformat()
else:
ds = obj.pop("datetime")
dt = dates.parse_iso(ds)
cmd_kwargs["datetime"] = dates.floor(dt, unit="hour", value=1).isoformat()
metrics = ingest_util.prepare_metrics(
obj, metrics_lookup, valid_levels=["content_item", "all"], check_timeseries=True
)
# upsert command
cmd = """SELECT upsert_content_metric_timeseries(
{org_id},
{content_item_id},
'{datetime}',
'{metrics}')
""".format(
metrics=obj_to_json(metrics), **cmd_kwargs
)
if commit:
try:
db.session.execute(cmd)
except Exception as err:
raise RequestError(err.message)
cmd_kwargs["metrics"] = metrics
return cmd
def org_timeseries(obj, org_id=None, metrics_lookup=None, commit=True):
"""
Ingest Timeseries Metrics for an organization.
"""
cmd_kwargs = {'org_id': org_id}
# parse datetime.
if 'datetime' not in obj:
cmd_kwargs['datetime'] = dates.floor_now(unit='hour',
value=1).isoformat()
else:
ds = obj.pop('datetime')
dt = dates.parse_iso(ds)
cmd_kwargs['datetime'] = dates.floor(dt, unit='hour',
value=1).isoformat()
metrics = ingest_util.prepare_metrics(obj,
metrics_lookup,
valid_levels=['org', 'all'],
check_timeseries=True)
# upsert command
cmd = """SELECT upsert_org_metric_timeseries(
{org_id},
'{datetime}',
'{metrics}')
""".format(metrics=obj_to_json(metrics), **cmd_kwargs)
if commit:
try:
db.session.execute(cmd)
except Exception as err:
raise RequestError(err.message)
cmd_kwargs['metrics'] = metrics
return cmd_kwargs
return cmd
def publish_date(soup, source_url=None):
"""
Extract publish date from meta / source_url.
"""
# try isodate first
for tag in PUBLISH_DATE_TAGS:
ds = _extract_tag_data(soup, tag)
if ds:
dt = dates.parse_iso(ds, enforce_tz=False)
if dt:
return dt
# try a timestamp next.
for tag in PUBLISH_DATE_TAGS:
ds = _extract_tag_data(soup, tag)
if ds:
dt = dates.parse_ts(ds)
if dt:
return dt
# try any date next.
for tag in PUBLISH_DATE_TAGS:
ds = _extract_tag_data(soup, tag)
if ds:
dt = dates.parse_any(ds, enforce_tz=False)
if dt:
return dt
# fallback on url regex
if source_url:
dm = re_url_date.search(source_url)
if dm:
ds = dm.group(0)
dt = dates.parse_any(ds, enforce_tz=False)
if dt:
return dt
def publish_date(soup, source_url=None):
"""
Extract publish date from meta / source_url.
"""
# try isodate first
for tag in PUBLISH_DATE_TAGS:
ds = _extract_tag_data(soup, tag)
if ds:
dt = dates.parse_iso(ds, enforce_tz=False)
if dt:
return dt
# try a timestamp next.
for tag in PUBLISH_DATE_TAGS:
ds = _extract_tag_data(soup, tag)
if ds:
dt = dates.parse_ts(ds)
if dt:
return dt
# try any date next.
for tag in PUBLISH_DATE_TAGS:
ds = _extract_tag_data(soup, tag)
if ds:
dt = dates.parse_any(ds, enforce_tz=False)
if dt:
return dt
# fallback on url regex
if source_url:
dm = re_url_date.search(source_url)
if dm:
ds = dm.group(0)
dt = dates.parse_any(ds, enforce_tz=False)
if dt:
return dt
def created(self):
if self.post.get('updated_time'):
return dates.parse_iso(self.post['updated_time'])
return None
def created(self):
if self.post.get('updated_time'):
return dates.parse_iso(self.post['updated_time'])
return None
def get_status(user, job_id):
"""
Get the status of a queued job.
"""
# parse args.
queue = request.args.get('queue')
if not queue:
raise RequestError(
'You must pass in the queue name to fetch a job\'s status')
if not queue in queues:
raise RequestError(
'"{}" is not a valid queue.'
.format(queue))
q = queues.get(queue)
job = q.fetch_job(job_id)
if not job:
raise RequestError(
'A job with ID {} does not exist'
.format(job_id))
# fetch metadata about this job
# from the session
# parse args.
started = request.args.get('started')
orig_url = request.args.get('orig_url')
if started:
started = dates.parse_iso(started)
# format return value
ret = {
'job_id': job_id,
'queue': queue,
'status': None,
'started': started,
'orig_url': orig_url
}
# determine time since start
if started:
ret['time_since_start'] = (dates.now() - started).seconds
# determine status
if job.is_queued:
ret['status'] = 'queued'
if job.is_started:
ret['status'] = 'running'
if job.is_failed:
ret['status'] = 'error'
ret['message'] = "An unknown error occurred."
if job.is_finished:
rv = job.return_value
# job will return true if successful
if rv is True:
ret['status'] = 'success'
# job will return an error if unsuccessful
else:
ret['status'] = 'error'
ret['message'] = str(rv.message)
return jsonify(ret)
def get_status(user, job_id):
"""
Get the status of a queued job.
"""
# parse args.
queue = request.args.get('queue')
if not queue:
raise RequestError(
'You must pass in the queue name to fetch a job\'s status')
if not queue in queues:
raise RequestError('"{}" is not a valid queue.'.format(queue))
q = queues.get(queue)
job = q.fetch_job(job_id)
if not job:
raise RequestError('A job with ID {} does not exist'.format(job_id))
# fetch metadata about this job
# from the session
# parse args.
started = request.args.get('started')
orig_url = request.args.get('orig_url')
if started:
started = dates.parse_iso(started)
# format return value
ret = {
'job_id': job_id,
'queue': queue,
'status': None,
'started': started,
'orig_url': orig_url
}
# determine time since start
if started:
ret['time_since_start'] = (dates.now() - started).seconds
# determine status
if job.is_queued:
ret['status'] = 'queued'
if job.is_started:
ret['status'] = 'running'
if job.is_failed:
ret['status'] = 'error'
ret['message'] = "An unknown error occurred."
if job.is_finished:
rv = job.return_value
# job will return true if successful
if rv is True:
ret['status'] = 'success'
# job will return an error if unsuccessful
else:
ret['status'] = 'error'
ret['message'] = rv.message
return jsonify(ret)
