What is this?

bareun-apis is the generated python classes of GRPC API for bareun.ai.

The bareun.ai provides several service for deep learning NLP features. This api has all of main features, which provides tokenizing, POS tagging for Korean. It has also customized dictionary service.

How to install

pip3 install bareun-apis

How to use

• You can create your own baikal language service client.
• It is used for bareunlpy, the official bareun package for python.

from google.protobuf.json_format import MessageToDict

import bareun.language_service_pb2 as pb
import bareun.language_service_pb2_grpc as ls

MAX_MESSAGE_LENGTH = 100*1024*1024

class BareunLanguageServiceClient:

    def __init__(self, remote):
        channel = grpc.insecure_channel(
            remote,
            options=[
                ('grpc.max_send_message_length', MAX_MESSAGE_LENGTH),
                ('grpc.max_receive_message_length', MAX_MESSAGE_LENGTH),
            ])

        self.stub = ls.LanguageServiceStub(channel)

    def analyze_syntax(self, document, auto_split=False):
        req = pb.AnalyzeSyntaxRequest()
        req.document.content = document
        req.document.language = "ko_KR"
        req.encoding_type = pb.EncodingType.UTF32
        req.auto_split_sentence = auto_split

        res = self.stub.AnalyzeSyntax(req)
        # print_syntax_as_json(res)
        return res

    def tokenize(self, document, auto_split=False):
        req = pb.TokenizeRequest()
        req.document.content = document
        req.document.language = "ko_KR"
        req.encoding_type = pb.EncodingType.UTF32
        req.auto_split_sentence = auto_split

        res = self.stub.Tokenize(req)
        # print_tokens_as_json(res)
        return res

def print_syntax_as_json(res: pb.AnalyzeSyntaxResponse, logf=sys.stdout):
    d = MessageToDict(res)
    import json
    json_str = json.dumps(d, ensure_ascii=False, indent=2)
    logf.write(json_str)
    logf.write('\n')

def print_tokens_as_json(res: pb.TokenizeResponse, logf=sys.stdout):
    d = MessageToDict(res)
    import json
    json_str = json.dumps(d, ensure_ascii=False, indent=2)
    logf.write(json_str)
    logf.write('\n')