def send_documents(self, docs):
'''Open a connection to the Solr endpoint, and send the documents
for indexing. Multiple requests are sent if a large number of documents
are being sent (see chunk_xml())
Raises SearchHTTPError if the endpoint indicates a failure
'''
core = getattr(g, 'solr_core', 'collection1')
responses = []
connection = httplib.HTTPConnection(self.solr_host, self.solr_port)
chunker = chunk_xml(docs)
headers = {}
headers['Content-Type'] = 'application/xml'
try:
for data in chunker:
# HTTPLib calculates Content-Length header automatically
if getattr(g, 'solr_version', '1').startswith('4'):
connection.request('POST', "/solr/%s/update/" % core, data,
headers)
else:
connection.request('POST', "/solr/update/", data, headers)
response = connection.getresponse()
if 200 <= response.status < 300:
responses.append(response.read())
else:
raise SearchHTTPError(response.status, response.reason,
response.read())
finally:
connection.close()
return responses
def send_documents(self, docs):
'''Open a connection to the cloudsearch endpoint, and send the documents
for indexing. Multiple requests are sent if a large number of documents
are being sent (see chunk_xml())
Raises SearchHTTPError if the endpoint indicates a failure
'''
responses = []
connection = httplib.HTTPConnection(self.doc_api,
port=80,
timeout=_TIMEOUT)
chunker = chunk_xml(docs)
try:
for data in chunker:
headers = {}
headers['Content-Type'] = 'application/xml'
# HTTPLib calculates Content-Length header automatically
connection.request('POST', "/2011-02-01/documents/batch", data,
headers)
response = connection.getresponse()
if 200 <= response.status < 300:
responses.append(response.read())
else:
raise SearchHTTPError(response.status, response.reason,
response.read())
finally:
connection.close()
return responses
def basic_query(query=None,
bq=None,
faceting=None,
size=1000,
start=0,
rank=None,
rank_expressions=None,
return_fields=None,
record_stats=False,
search_api=None):
if search_api is None:
search_api = g.CLOUDSEARCH_SEARCH_API
if faceting is None:
faceting = DEFAULT_FACETS
path = _encode_query(query, bq, faceting, size, start, rank,
rank_expressions, return_fields)
timer = None
if record_stats:
timer = g.stats.get_timer("providers.cloudsearch")
timer.start()
connection = httplib.HTTPConnection(search_api, port=80, timeout=_TIMEOUT)
try:
connection.request('GET', path)
resp = connection.getresponse()
response = resp.read()
if record_stats:
g.stats.action_count("event.search_query", resp.status)
if resp.status >= 300:
try:
reasons = json.loads(response)
except ValueError:
pass
else:
messages = reasons.get("messages", [])
for message in messages:
if message['code'] in INVALID_QUERY_CODES:
raise InvalidQuery(resp.status, resp.reason, message,
search_api, path, reasons)
raise SearchHTTPError(resp.status, resp.reason, search_api, path,
response)
except socket.timeout as e:
g.stats.simple_event('cloudsearch.error.timeout')
raise SearchError(e, search_api, path)
except socket.error as e:
g.stats.simple_event('cloudsearch.error.socket')
raise SearchError(e, search_api, path)
finally:
connection.close()
if timer is not None:
timer.stop()
return json.loads(response)
def basic_query(query=None,
bq=None,
faceting=None,
size=1000,
start=0,
rank="",
return_fields=None,
record_stats=False,
search_api=None):
if search_api is None:
search_api = g.solr_search_host
if faceting is None:
faceting = DEFAULT_FACETS
path = _encode_query(query, faceting, size, start, rank, return_fields)
timer = None
if record_stats:
timer = g.stats.get_timer("solrsearch_timer")
timer.start()
connection = httplib.HTTPConnection(search_api, g.solr_port)
try:
connection.request('GET', path)
resp = connection.getresponse()
response = resp.read()
if record_stats:
g.stats.action_count("event.search_query", resp.status)
if resp.status >= 300:
try:
response_json = json.loads(response)
except ValueError:
pass
else:
if 'error' in response_json:
message = response_json['error'].get(
'msg', 'Unknown error')
raise InvalidQuery(resp.status, resp.reason, message,
search_api, path, response_json)
raise SearchHTTPError(resp.status, resp.reason, search_api, path,
response)
except socket.error as e:
raise SearchError(e, search_api, path)
finally:
connection.close()
if timer is not None:
timer.stop()
return json.loads(response)