def __init__(self, catalog, xquery):
self._catalog = catalog
self._xquery = xquery
# Enquire
enquire = Enquire(catalog._db)
enquire.set_query(xquery)
self._enquire = enquire
# Max
max = enquire.get_mset(0,0).get_matches_upper_bound()
self._max = enquire.get_mset(0, max).size()
def morelike(self, model_clz, query, select=["*"], limit=10, offset=0, order=None,
ascending=True):
""" Find documents in the database most relevant to the given terms.
'select' - The set of keys from the documents picked data
dictionary to return or use to construct a record. '*' means
all available keys.
'limit' - The number of records to return.
'offset' - How many records to skip before returning.
'order' - The number or name (if order_names was provided to
the constructor) of the column to sort the results by.
'ascending' - Whether to sort the results in ascending or
descending order.
"""
self.database.reopen()
suggested_terms = self.suggest(model_clz, query, limit, offset, order, ascending)
terms = [term[0] for term in suggested_terms]
enq = Enquire(self.database)
enq.set_collapse_key(DOC_ITEM_ID)
if order is not None:
if issubclass(type(order), basestring):
order = self.sort_names.index(order)
if order == -1:
raise TypeError("There is no sort name %s" % order)
enq.set_sort_by_value(order, ascending)
else:
enq.set_sort_by_relevance()
query = Query(Query.OP_ELITE_SET, terms, limit)
query = self._add_model_query(query,model_clz)
enq.set_query(query)
mset = enq.get_mset(offset, limit)
return self._generate_records(mset, select)
def select(self, model_clz, query, select=set(["*"]), limit=10, offset=0, order=None,
ascending=True, partial=False):
"""Select documents from the database matching 'query'.
'select' - The set of keys from the documents picked data
dictionary to return or use to construct a record. '*' means
all available keys.
'limit' - The number of records to return.
'offset' - How many records to skip before returning.
'order' - The number or name (if order_names was provided to
the constructor) of the column to sort the results by.
'ascending' - Whether to sort the results in ascending or
descending order.
'partial' - Wether to support wildcard partial queries like "foo*".
"""
self.database.reopen()
enq = Enquire(self.database)
enq.set_collapse_key(DOC_ITEM_ID)
if order is not None:
if issubclass(type(order), basestring):
order = self.sort_names.index(order)
if order == -1:
raise TypeError("There is no sort name %s" % order)
enq.set_sort_by_value(order, ascending)
else:
enq.set_sort_by_relevance()
query, query_parser = self._parse_query(query, partial=partial)
query = self._add_model_query(query,model_clz)
log.debug(query)
enq.set_query(query)
mset = enq.get_mset(offset, limit)
return self._generate_records(mset, select)
def __init__(self, catalog, xquery):
self._catalog = catalog
self._xquery = xquery
# Enquire
enquire = Enquire(catalog._db)
enquire.set_query(xquery)
self._enquire = enquire
# Max
max = enquire.get_mset(0, 0).get_matches_upper_bound()
self._max = enquire.get_mset(0, max).size()
def estimate(self, model_clz, query, limit=10, partial=False):
"""
Estimate the number of documents that will be yielded with
the given query. limit tells the estimator the minimum number
of documents to consider.
"""
self.database.reopen()
enq = Enquire(self.database)
enq.set_collapse_key(DOC_ITEM_ID)
query, query_parser = self._parse_query(query, partial=partial)
query = self._add_model_query(query,model_clz)
log.debug(query)
enq.set_query(query)
return enq.get_mset(0, 0, limit).get_matches_estimated()
def get_docs(db):
enquire = Enquire(db)
enquire.set_query(Query(''))
docs_max = enquire.get_mset(0, 0).get_matches_upper_bound()
return [doc.get_document() for doc in enquire.get_mset(0, docs_max)]
def find(self, query):
offset = query.pop('offset', 0)
limit = query.pop('limit', MAX_QUERY_LIMIT)
order_by = query.pop('order_by', [])
query_string = query.pop('query', None)
query_parser = QueryParser()
query_parser.set_database(self._database)
enquire = Enquire(self._database)
enquire.set_query(query_parser.parse_query(query, query_string))
# This will assure that the results count is exact.
check_at_least = offset + limit + 1
if not order_by:
order_by = '+timestamp'
else:
order_by = order_by[0]
if order_by == '+timestamp':
enquire.set_sort_by_value(_VALUE_TIMESTAMP, True)
elif order_by == '-timestamp':
enquire.set_sort_by_value(_VALUE_TIMESTAMP, False)
elif order_by == '+title':
enquire.set_sort_by_value(_VALUE_TITLE, True)
elif order_by == '-title':
enquire.set_sort_by_value(_VALUE_TITLE, False)
elif order_by == '+filesize':
enquire.set_sort_by_value(_VALUE_FILESIZE, True)
elif order_by == '-filesize':
enquire.set_sort_by_value(_VALUE_FILESIZE, False)
elif order_by == '+creation_time':
enquire.set_sort_by_value(_VALUE_CREATION_TIME, True)
elif order_by == '-creation_time':
enquire.set_sort_by_value(_VALUE_CREATION_TIME, False)
else:
logging.warning('Unsupported property for sorting: %s', order_by)
query_result = enquire.get_mset(offset, limit, check_at_least)
total_count = query_result.get_matches_estimated()
uids = []
for hit in query_result:
uids.append(hit.document.get_value(_VALUE_UID))
return (uids, total_count)
def find(self, query):
offset = query.pop('offset', 0)
limit = query.pop('limit', MAX_QUERY_LIMIT)
order_by = query.pop('order_by', [])
query_string = query.pop('query', None)
query_parser = QueryParser()
query_parser.set_database(self._database)
enquire = Enquire(self._database)
enquire.set_query(query_parser.parse_query(query, query_string))
# This will assure that the results count is exact.
check_at_least = offset + limit + 1
if not order_by:
order_by = '+timestamp'
else:
order_by = order_by[0]
if order_by == '+timestamp':
enquire.set_sort_by_value(_VALUE_TIMESTAMP, True)
elif order_by == '-timestamp':
enquire.set_sort_by_value(_VALUE_TIMESTAMP, False)
elif order_by == '+title':
enquire.set_sort_by_value(_VALUE_TITLE, True)
elif order_by == '-title':
enquire.set_sort_by_value(_VALUE_TITLE, False)
elif order_by == '+filesize':
enquire.set_sort_by_value(_VALUE_FILESIZE, True)
elif order_by == '-filesize':
enquire.set_sort_by_value(_VALUE_FILESIZE, False)
elif order_by == '+creation_time':
enquire.set_sort_by_value(_VALUE_CREATION_TIME, True)
elif order_by == '-creation_time':
enquire.set_sort_by_value(_VALUE_CREATION_TIME, False)
else:
logging.warning('Unsupported property for sorting: %s', order_by)
query_result = enquire.get_mset(offset, limit, check_at_least)
total_count = query_result.get_matches_estimated()
uids = []
for hit in query_result:
uids.append(hit.document.get_value(_VALUE_UID))
return (uids, total_count)
def _enquire(self):
enquire = Enquire(self._catalog._db)
enquire.set_query(self._xquery)
return enquire
def _enquire(self):
enquire = Enquire(self._database.catalog._db)
enquire.set_query(self._xquery)
return enquire
def get_docs(db):
enquire = Enquire(db)
enquire.set_query(Query(''))
docs_max = enquire.get_mset(0,0).get_matches_upper_bound()
return [doc.document for doc in enquire.get_mset(0, docs_max)]
def suggest(self, model_clz, query, limit=10, offset=0, order=None, ascending=True):
"""
Suggest terms that would possibly yield more relevant results
for the given query.
"""
self.database.reopen()
enq = Enquire(self.database)
enq.set_collapse_key(DOC_ITEM_ID)
if order is not None:
if issubclass(type(order), basestring):
order = self.sort_names.index(order)
if order == -1:
raise TypeError("There is no sort name %s" % order)
enq.set_sort_by_value(order, ascending)
else:
enq.set_sort_by_relevance()
query_parser = self._query_parser()
query = query_parser.parse_query(query)
query = self._add_model_query(query,model_clz)
log.debug(query)
enq.set_query(query)
mset = enq.get_mset(offset, limit)
rset = RSet()
for m in mset:
rset.add_document(m[MSET_DID])
eset = enq.get_eset(limit, rset)
for item in eset.items:
yield (item[0].decode('utf8'), item[1])