def query_tweets():
"""Handle API '/tweets'.
API Request Parameters
----------------------
ids : list of int
API Response Keys
-----------------
status : string
num_of_entries : int
tweets : dict
canonical_url : string
domain : string
id : int
date_published : string formatted datetime
site_type : {'claim', 'fact_checking'}
title : string
tweet_created_at : string formatted datetime
tweet_id : string
"""
lucene.getVMEnv().attachCurrentThread()
q_tweets_schema = Schema({
'ids':
And(Use(flask.json.loads, error="Format error of `ids`"),
lambda s: len(s) > 0,
error='Empty of `ids`'),
})
q_kwargs = copy_req_args(request.args)
try:
q_kwargs = q_tweets_schema.validate(q_kwargs)
df = db_query_tweets(engine, q_kwargs['ids'])
if len(df) == 0:
raise APINoResultError('No tweet found!')
response = dict(
status='OK',
num_of_entries=len(df),
tweets=flask.json.loads(df.to_json(**TO_JSON_KWARGS)))
except SchemaError as e:
response = dict(status='ERROR', error=str(e))
except APINoResultError as e:
response = dict(status='No result error', error=str(e))
except Exception as e:
logger.exception(e)
response = dict(status='ERROR', error='Server error, query failed')
return flask.jsonify(response)
def query_timeline():
"""Handle API '/timeline'.
API Request Parameters
----------------------
ids : list of int
resolution : character in 'HDWM'
API Response Keys
-----------------
status : string
timeline : dict
claim : dict
timestamp : list of string formatted datetime
volume : list of int
fact_checking : dict
timestamp : list of string formatted datetime
volume : list of int
"""
lucene.getVMEnv().attachCurrentThread()
q_tweets_schema = Schema({
'ids':
And(Use(flask.json.loads, error="Format error of `ids`"),
lambda s: len(s) > 0,
error='Empty of `ids`'),
Optional('resolution', default='D'):
And(Use(lambda s: s.upper()), lambda s: s in 'HDWM'),
})
q_kwargs = copy_req_args(request.args)
try:
q_kwargs = q_tweets_schema.validate(q_kwargs)
rule = '1' + q_kwargs.pop('resolution')
df = db_query_tweets(engine, q_kwargs['ids'])
if len(df) == 0:
raise APINoResultError('No tweet found!')
df = df.set_index('tweet_created_at')
df1 = df.loc[df['site_type'] == N_FACT_CHECKING]
s1 = df1['tweet_id'].drop_duplicates()
s1 = s1.resample(rule).count()
df2 = df.loc[df['site_type'] == N_CLAIM]
s2 = df2['tweet_id'].drop_duplicates()
s2 = s2.resample(rule).count()
s1, s2 = s1.align(s2, join='outer', fill_value=0)
s1 = s1.cumsum()
s2 = s2.cumsum()
response = dict(
status='OK',
timeline=dict(
fact_checking=dict(
timestamp=s1.index.strftime('%Y-%m-%dT%H:%M:%SZ').tolist(),
volume=s1.tolist()),
claim=dict(
timestamp=s2.index.strftime('%Y-%m-%dT%H:%M:%SZ').tolist(),
volume=s2.tolist())))
except SchemaError as e:
response = dict(status='ERROR', error=str(e))
except APINoResultError as e:
response = dict(status='No result error', error=str(e))
except Exception as e:
logger.exception(e)
response = dict(status='ERROR', error='Server error, query failed')
return flask.jsonify(response)