def consume_next(self, pred_pack: MultiPack, _):
query_pack: DataPack = pred_pack.get_pack(self.configs.pack_name)
query = list(query_pack.get(Query))[0]
rank = 1
for pid, _ in query.results.items():
doc_id: Optional[str] = query_pack.pack_name
if doc_id is None:
raise ProcessExecutionException(
'Doc ID of the query pack is not set, '
'please double check the reader.')
self.predicted_results.append((doc_id, pid, str(rank)))
rank += 1
def get_pack_index(self, pack_id: int) -> int:
"""
Get the pack index from the global pack id.
Args:
pack_id: The global pack id to find.
Returns:
"""
try:
return self._inverse_pack_ref[pack_id]
except KeyError as e:
raise ProcessExecutionException(
f"Pack {pack_id} is not in this multi-pack.") from e
def _process(self, input_pack: DataPack):
doc = input_pack.text
# Do all process.
if self.nlp is None:
raise ProcessExecutionException(
"The SpaCy pipeline is not initialized, maybe you "
"haven't called the initialization function.")
result = self.nlp(doc)
# Record NER results.
self._process_ner(result, input_pack)
# Process sentence parses.
self._process_parser(result.sents, input_pack)
def consume_next(self, pred_pack: MultiPack, _):
#print(self.configs.pack_name)
query_pack: DataPack = pred_pack.get_pack(self.configs.pack_name)
query = list(query_pack.get(Query))[0]
query_text = query_pack.text
#print(pred_pack.get_pack('passage_6').text)
sorted_query_results = sorted(list(query.results.items()),
key=lambda x: x[1],
reverse=True)
rank = 1
for pid, _ in sorted_query_results:
doc_id: Optional[str] = query_pack.pack_name
if doc_id is None:
raise ProcessExecutionException(
'Doc ID of the query pack is not set, '
'please double check the reader.')
self.predicted_results.append((doc_id, pid, str(rank)))
rank += 1
def __del__(self):
if len(self._pending_entries) > 0:
raise ProcessExecutionException(
f"There are {len(self._pending_entries)} "
f"entries not added to the index correctly.")