def import_data(plattform: str, year: str):
index = ElasticIndex("emanus-{}-data-base-{}".format(plattform, year),
"doc")
with open("emanus-{}-{}.json".format(plattform, year), 'r') as fp:
text = json.load(fp)
metadata = dict()
data = list()
for key in text:
if key == "data":
for item in text[key]:
result = item
result["identifier"] = item["dimensions"]["pagestem"]
data.append(result)
else:
metadata[key] = text[key]
index.bulk(data, identifier_key="identifier")
index.index_into(metadata, 0)
class SimpleElasticConsumer(AbstractBaseConsumer):
"""
A KafkaConsumer which consumes messages and indexes them into a ElasticIndex one by one.
Requires the following configs:
Consumer:
bootstrap_servers: localhost:9092
client_id: test
group_id: elastic-consumer-test
auto_offset_reset: earliest
Topics:
- test
ElasticIndex:
index: name-of-index
doc_type: _doc (default value for elasticsearch 6)
url: http://localhost:9200
timeout: 300
"""
def __init__(self,
config,
config_class=ElasticConsumerConfig,
logger=logging.getLogger(__name__)):
super().__init__(config, config_class, logger=logger)
self._index = ElasticIndex(**self.configuration.elastic_settings)
def consume(self) -> bool:
"""
Consumes a single message from the subscribed topic and indexes it into the elasticsearch index.
Returns True if successful, False otherwise.
"""
message = next(self._consumer)
key = message.key.decode('utf-8')
try:
value = json.loads(message.value.decode('utf-8'))
except JSONDecodeError as ex:
value = {
'message': message.value.decode('utf-8'),
'error': '{}'.format(ex)
}
result = self._index.index_into(value, key)
if result:
for assignment in self._consumer.assignment():
pos = self._consumer.position(assignment)
if pos != self._consumer.committed(assignment):
self._consumer.commit(
{assignment: OffsetAndMetadata(pos, "")})
# self._time_logger.info("Consumed and indexed one message.")
return result
class TestElasticProducer(object):
def setup_class(self):
self.admin = KafkaAdminClient(bootstrap_servers='localhost:9092')
self.index = ElasticIndex('test-elastic-producer', 'doc')
self.index.index_into({'test': 1}, 0)
self.index.index_into({'test': 2}, 1)
self.index.index_into({'test': 3}, 2)
self.index.index_into({'test': 4}, 3)
self.index.index_into({'test': 5}, 4)
self.producer = ElasticProducer('configs/elastic/test_elastic_producer_producer.yml')
self.consumer = SimpleConsumer('configs/elastic/test_elastic_producer_consumer.yml')
def teardown_class(self):
self.consumer.close()
self.admin.delete_topics(['test-elastic-producer'])
self.admin.close()
self.index.delete()
#@pytest.mark.skip()
def test_produce(self):
self.producer.process()
key, message = self.consumer.consume()
assert key == '0'
assert message == '{"test": 1}'
class TestElasticIndex(object):
def setup_class(self):
self.index = ElasticIndex('test')
def teardown_class(self):
self.index.delete_index()
def test_scroll(self):
self.index.index_into({'test': True}, 1)
self.index.index_into({'test': False}, 2)
self.index.index_into({'test': True}, 3)
self.index.index_into({'test': False}, 4)
for i in self.index.scroll():
assert isinstance(i, list)
def test_index_into(self):
result = self.index.index_into({'test': True, 'object': "This is a string"}, 5)
assert result
result = self.index.index_into({'test': True, 'object': {'sub-object': "another string"}}, 6)
assert not result
result = self.index.index_into({'test': False}, 'HAN000827182')
assert result
def test_search(self):
data = list()
data.append({'id': '1234', 'test': True})
self.index.bulk(data=data, identifier_key='id')
result = self.index.search()
assert len(result) == 7
def test_search_not_unpack(self):
result = self.index.search(unpack=False)
assert len(result) == 7
def test_alias(self):
self.index.add_to_alias('test1')
assert self.index.instance.indices.get_alias('test1')
self.index.remove_from_alias('test1')
with pytest.raises(NotFoundError):
self.index.instance.indices.get_alias('test1')
def test_count(self):
result = self.index.count()
assert result == 7
index = ElasticIndex('kafka-dsv05-*', 'record')
for key in collect_stems:
if key == '0001':
# TODO: A combined manuscript.
continue
item = dict()
cn = transform_call_number(key)
query = {
'_source': ['call_number', 'identifiers.*'],
'query': {
'term': {
'call_number.keyword': cn
}
}
}
results = index.scan_index(query=query)
if len(results) == 1:
result = results[0]
item['hits'] = collect_stems[key]
if 'doi' not in result['identifiers']:
if key in dois:
item['doi'] = dois[key]
target.index_into(item, result['identifiers']['dsv05'])
else:
print(key, cn, results)
def enrich_user_data(config):
for index in config['indexes']:
instance = ElasticIndex(**index['index'])
query = {'query': {'match_all': {}}}
for results in instance.scroll(query=query):
for item in results:
identifier = item['identifier']
database = item['database']
sys_number = item['identifiers'][database]
if 'error_tags' in item:
item['error_tags'] = set(item['error_tags'])
total = 0
# swissbib
hits, error_tags = swissbib.enrich(identifier)
item['hits']['swissbib'] = hits
total += hits['total']
for tag in error_tags:
item['error_tags'].add(tag)
# opac
hits, error_tags = opac.enrich(opac_index, sys_number)
item['hits']['opac-access'] = hits
total += hits['total']
for tag in error_tags:
item['error_tags'].add(tag)
# aleph
hits, error_tags = aleph.enrich(aleph_index, sys_number,
database)
item['hits']['aleph'] = hits
total += hits['loans']['total']
for tag in error_tags:
item['error_tags'].add(tag)
if database == 'dsv05':
# e-rara
hits, error_tags = e_rara.enrich(e_rara_index, sys_number)
item['hits']['e-rara'] = hits
total += hits['bau']['total']
for tag in error_tags:
item['error_tags'].add(tag)
# e-manuscripta
hits, error_tags = e_manuscripta.enrich(
e_manuscripta_index, sys_number)
item['hits']['e-manuscripta'] = hits
total += hits['bau']['total']
total += hits['swa']['total']
for tag in error_tags:
item['error_tags'].add(tag)
# e-codices
hits, doi, error_tags = e_codices.enrich(
e_codices_index, sys_number)
item['hits']['e-codices'] = hits
total += hits['total']
for tag in error_tags:
item['error_tags'].add(tag)
if doi is not None:
if 'doi' in item['identifiers']:
if isinstance(item['identifiers']['doi'], list):
item['identifiers']['doi'].append(doi)
else:
item['identifiers']['doi'] = [
item['identifiers']['doi'], doi
]
# e-mails dsv05
# TODO
item['error_tags'] = list(item['error_tags'])
item['hits']['total'] = total
instance.index_into(item, item['identifier'])
from simple_elastic import ElasticIndex
from datetime import date
dsv01_full_export = ElasticIndex('dsv01-sys-numbers-before-1900', 'record')
with open('data/dsv01_system_numbers_vor_1900_arc_export_20180802.csv',
'r',
encoding='utf-16') as file:
for line in file:
year, sys_number = line.split(',')
doc = dict()
while len(sys_number) != 10:
sys_number = '0' + sys_number
doc['system_number'] = sys_number.strip()
doc['publication_date'] = year
doc['index_date'] = date.today().isoformat()
dsv01_full_export.index_into(doc, doc['system_number'])
# TODO: Rework how this works and think about how I need the data?
wb = load_workbook('data/VERZ_DSV01-Ausleihen-Vormerkungen_20180802_bmt.xlsx')
ws = wb['vor-1900-vormerkungen']
index = ElasticIndex('reservations', 'record')
for row in ws.iter_rows(min_row=2, min_col=2, max_col=4):
doc = dict()
system_number = str(row[0].value)
while len(system_number) != 9:
system_number = '0' + system_number
doc['system_number'] = system_number
doc['reservations'] = row[1].value
doc['year'] = str(row[2].value)
index.index_into(doc, doc['system_number'] + doc['year'])
ws = wb['vor-1900-ausleihen']
for row in ws.iter_rows(min_row=2, min_col=2, max_col=4):
doc = dict()
system_number = str(row[0].value)
while len(system_number) != 9:
system_number = '0' + system_number
doc['system_number'] = system_number
doc['loans'] = row[1].value
doc['year'] = str(row[2].value)
record = index.get(doc['system_number'] + doc['year'])
if record is not None:
doc['reservations'] = record['reservations']
index.index_into(doc, doc['system_number'] + doc['year'])
