()
| 7 | |
| 8 | |
| 9 | def main(): |
| 10 | |
| 11 | output_dir = sys.argv[1] |
| 12 | for fold in ["train", "dev", "test"]: |
| 13 | g_convert = open(path.join(output_dir, fold + "_convert.json"), "w") |
| 14 | with open(path.join(output_dir, fold + ".json"), "r") as g: |
| 15 | print('convert %s to %s' % ( |
| 16 | path.join(output_dir, fold + ".json"), |
| 17 | path.join(output_dir, fold + "_convert.json") |
| 18 | )) |
| 19 | for line in g: |
| 20 | line = json.loads(line) |
| 21 | sentences = line["sentences"] |
| 22 | ner = line["ner"] |
| 23 | relations = line["relations"] |
| 24 | events = line["events"] |
| 25 | sentence_start = line["_sentence_start"] |
| 26 | doc_key = line["doc_key"] |
| 27 | |
| 28 | assert len(sentence_start) == len(ner) == len( |
| 29 | relations) == len(events) == len(sentence_start) |
| 30 | |
| 31 | for sentence, ner, relation, event, s_start in zip(sentences, ner, relations, events, sentence_start): |
| 32 | sentence_annotated = collections.OrderedDict() |
| 33 | sentence_annotated["sentence"] = sentence |
| 34 | sentence_annotated["s_start"] = s_start |
| 35 | sentence_annotated["ner"] = ner |
| 36 | sentence_annotated["relation"] = relation |
| 37 | sentence_annotated["event"] = event |
| 38 | |
| 39 | g_convert.write(json.dumps( |
| 40 | sentence_annotated, default=int) + "\n") |
| 41 | |
| 42 | |
| 43 | if __name__ == "__main__": |
no outgoing calls
no test coverage detected