コード例 #1
0
class SearchData():
    def __init__(self, ip='127.0.0.1', solr_core=SOLR_CORE_NAME):
        self.solr_url = 'http://' + ip + ':8999/solr'
        self.solr_core = solr_core
        self.solr = SOLR(self.solr_url)

    def search_answer(self, select='*:*', scene_topic=[]):
        try:
            fields = ['answers', 'emotion_url', 'media', 'timeout']
            select_parts = []
            for scene in scene_topic:
                s = 'scene_str:' + scene
                if scene_topic[scene] != []:
                    s = '('+ s + ' AND (topic_str:'+ \
                            ' OR topic_str:'.join(scene_topic[scene]) +'))'
                select_parts.append(s)
            select = '(' + ' OR '.join(select_parts) + ') AND ' + select
            data = [
                x for x in self.solr.query_solr(self.solr_core, select, fields,
                                                1).docs
            ]
            data = data[0]
            return {
                'answer': random.sample(data['answers'], 1)[0],
                'emotion': data['emotion_url'][0],
                'media': data['media'][0],
                'timeout': data['timeout'][0]
            }
        except:
            traceback.print_exc()
            return {
                'answer': None,
                'emotion': None,
                'media': None,
                'timeout': None
            }

    def search_questions(self,
                         select='*:*',
                         scene_topic=[],
                         fields=['question'],
                         max_num=10):
        try:

            def pro_data(data):
                for key in data.keys():
                    data[key] = data[key][0]
                return data

            select_parts = []
            for scene in scene_topic:
                s = 'scene_str:' + scene
                if scene_topic[scene] != []:
                    s = '('+ s + ' AND (topic_str:'+ \
                            ' OR topic_str:'.join(scene_topic[scene]) +'))'
                select_parts.append(s)
            select = '(' + ' OR '.join(select_parts) + ') AND ' + select
            data = [
                pro_data(x) for x in self.solr.query_solr(
                    self.solr_core, select, fields, max_num).docs
            ]
            return data
        except:
            traceback.print_exc()
            return None

    def sale_id2description(self, _id, scene):
        try:
            select = 'scene_str:' + scene + ' AND _id_str:' + _id
            fields = ['description']
            data = [
                x for x in self.solr.query_solr(self.solr_core, select, fields,
                                                1).docs
            ]
            data = data[0]
            return data['description']
        except:
            traceback.print_exc()
            return None

    def sale_type2answers(self, scene, t=''):
        try:

            def pro_data(data):
                for key in data.keys():
                    data[key] = data[key][0]
                return data

            if t:
                select = 'scene_str:' + scene + ' AND type_str:' + t
            else:
                select = 'scene_str' + scene + ' AND type_str:*'
            fields = [
                '_id', 'answers', 'type', 'emotion_url', 'media', 'timeout'
            ]
            max_num = 20
            data = [
                pro_data(x) for x in self.solr.query_solr(
                    self.solr_core, select, fields, max_num).docs
            ]
            return data
        except:
            traceback.print_exc()
            return None
コード例 #2
0
ファイル: load_data.py プロジェクト: zhengxin2016/corpus
class SearchSolr():
    def __init__(self, ip='127.0.0.1', solr_core=SOLR_CORE_NAME):
        self.solr_url = 'http://' + ip + ':8999/solr'
        self.solr_core = solr_core
        self.solr = SOLR(self.solr_url)

    def load_data(self, select='*:*', fields=[], max_num=10, flag=False):
        try:

            def pro_x(x):
                y = {}
                y['store_id'] = x['store_id'][0]
                y['category'] = x['category'][0]
                y['instruction'] = x['instruction'][0]
                if 'entities' in x:
                    y['entities'] = x['entities']
                else:
                    y['entities'] = ['']
                y['answers'] = x['answer']
                y['emotion_name'] = 'null'
                y['emotion_url'] = 'null'
                if 'media' in x:
                    y['media'] = x['media'][0]
                    y['timeout'] = '15'
                else:
                    y['media'] = 'null'
                    y['timeout'] = '0'
                return y

            Data = {}

            def pro_y(x):
                y = {}
                y['store_id'] = x['store_id'][0]
                y['category'] = x['category'][0]
                y['intent'] = x['intent']
                y['questions'] = x['question']
                if 'entities' in x:
                    y['entities'] = x['entities']
                else:
                    y['entities'] = ''
                if y['intent'] + '|' + y['entities'] in Data:
                    Data[y['intent'] + '|' +
                         y['entities']]['questions'].append(x['question'][0])
                else:
                    Data[y['intent'] + '|' + y['entities']] = y
                return y

            if flag == True:
                data = [
                    pro_x(x) for x in self.solr.query_solr(
                        self.solr_core, select, fields, max_num).docs
                ]
            else:
                data = [
                    pro_y(x) for x in self.solr.query_solr(
                        self.solr_core, select, fields, max_num).docs
                ]
                data = []
                for key in Data.keys():
                    data.append(Data[key])
            return data
        except:
            traceback.print_exc()
            return None