为公司开发搜索功能的时候使用到了阿里的开放搜索,是使用的python,官方没有sdk,只能自己实现api
这里把一些成熟的代码放上来,就是些简单的功能,包括:
1、文档增删改(dealDocPost)
2、post查询(sendQuery)
3、下拉提示(suggest)
4、生成签名(getSignature)
5、get查询(dealDocGet)
URL = 'http://opensearch-cn-hangzhou.aliyuncs.com/index/doc/' SEARCH_URL = 'http://opensearch-cn-hangzhou.aliyuncs.com/search' ACCESSKEY_ID = 自己的ID ACCESSKEY_SECRET = 自己的密码 table = helloworldapp = helloworld suggest_name = 'helloworld_rule' # 下拉查询规则 suggest_hit = 5 query_hit = 20 action = 'push'def dealDocPost(items): """post请求,增加,删除,修改,都可以从这里,唯一需要变化的是action""" if not items: return json.dumps({'status':'OK'}) curtime = int(time.time()) SignatureNonce = curtime + random.randrange(1000, 9999) timeStamp = datetime.datetime.utcnow().strftime("%Y-%m-%dT%H:%M:%SZ") params = [ ('AccessKeyId', ACCESSKEY_ID), ('SignatureMethod', 'HMAC-SHA1'), ('SignatureNonce', SignatureNonce), ('SignatureVersion', '1.0'), ('Version', 'v2'), ('Timestamp', timeStamp), ('action', action), ('table_name', table), ('items', json.dumps(items))] signature = getSignature(params= sorted(params, key=lambda x: x[0]) , method= 'POST') result_url = URL+ app + '?' + urllib.urlencode(params[:-3]) +'&Signature=' + signature req = urllib2.Request(result_url) req.add_data('&'.join([urllib.quote(str(k)) + '=' + urllib.quote(str(v)) for k, v in params[-3:]])) result_josnstr = urllib2.urlopen(req).read() print result_josnstr return result_josnstr
def sendQuery(request_data):
# 这里是查询:
query_data = {} start = int(request_data.get('start', 0)) hit = int(request_data.get('hit', query_hit)) query = request_data.get('query') key_words = request_data.get('search_text') config = u"start:%d,hit:%d,format:json"%(start, hit) query = query or (u"default:" + u'&'.join([u"'%s'"% kk for kk in key_words.split()])) condition = u"config=%s&&query=%s"% (config, query) query_data['query'] = condition if request_data.get('fields'): query_data['fields'] = request_data.get('fields') query_data['index_name'] = app # summary = u'summary_snipped:3,summary_field:content,summary_len:100;summary_snipped:1,summary_field:title,summary_len:50' result = doc_search(**query_data) return resultdef doc_search(query, fields = None, summary = None, formula_name = None, first_formula_name = None, index_name = None): """ 搜索查询 :param query : 搜索主体,包含所有的查询条件。子句分别包含config子句、query子句、sort子句、filter子句、aggregate子句、distinct子句等。 :param fields : 如果指定了此参数,则服务返回的数据集结构只包含fetch_fields里边指定的field,指定多个field可以使用“;”分隔。 :param summary : 动态摘要的配置,可以指定某些字段的飘红、截断等操作。 :param formula_name : 用户指定的精排表达式名称,搜索会获取您在网站中定义的此名称的表达式,并根据此来搜索。 :param first_formula_name : 用户指定的粗排表达式名称,搜索会获取您在网站中定义的此名称的表达式,并根据此来搜索。 :param index_name : 用户要查询的应用名称,需为UTF-8编码的urlencode结果,多应用查询可以使用“;”分隔。 :return : """ curtime = int(time.time()) SignatureNonce = curtime + random.randrange(1000, 9999) timeStamp = datetime.datetime.utcnow().strftime("%Y-%m-%dT%H:%M:%SZ") params = [ ('Version', 'v2'), ('AccessKeyId', ACCESSKEY_ID), ('SignatureMethod', 'HMAC-SHA1'), ('SignatureVersion', '1.0'), ('SignatureNonce', SignatureNonce), ('Timestamp', timeStamp), ('query', query.encode('utf8')), ('index_name', index_name), ('format', 'json')] # 其他搜索条件 if fields : params.append(('fetch_fields', fields)) if summary : params.append(('summary', summary)) if formula_name : params.append(('formula_name', formula_name)) if first_formula_name : params.append(('first_formula_name', first_formula_name)) signature = getSignature(params = sorted(params, key=lambda x : x[0]) , method='GET') req = urllib2.Request(SEARCH_URL + '?' + urllib.urlencode(params) + '&Signature=' + signature) try: return json.loads(urllib2.urlopen(req).read()) except Exception, e: return Nonedef suggest(query): ''' 用户输入数据获取下列列表提示 :return: ''' query = query.encode('utf8') curtime = int(time.time() * 1000000) SignatureNonce = curtime + random.randrange(1, 9999) timeStamp = datetime.datetime.utcnow().strftime("%Y-%m-%dT%H:%M:%SZ") params = [ ('AccessKeyId', ACCESSKEY_ID),('SignatureMethod', 'HMAC-SHA1'), ('SignatureNonce', SignatureNonce),('SignatureVersion', '1.0'), ('Version', 'v2'),('Timestamp', timeStamp), ('index_name', app),('suggest_name', suggest_name), ('hit', suggest_hit),('query', query)] signature = getSignature(params= sorted(params, key=lambda x: x[0]) , method= 'GET') result_url = 'http://opensearch-cn-hangzhou.aliyuncs.com/suggest?'+urllib.urlencode(params)+'&Signature=' + signature req = urllib2.Request(result_url) res_data = urllib2.urlopen(req) res = eval(res_data.read()) return res
def getSignature(params, method='GET'): """生成签名 . 参考文档:http://help.opensearch.aliyun.com/index.php?title=%E6%8E%88%E6%9D%83%E6%9C%BA%E5%88%B6#.E7.AD.BE.E5.90.8D.E7.94.9F.E6.88.90.E6.96.B9.E6.B3.95 """ str_params = '&'.join([urllib.quote(str(k)) + '=' + urllib.quote(str(v)) for k, v in params]) str_params = str_params.replace('%7E', '~') str_params = method + '&%2F&' + urllib.quote(str_params) StringToSign = str_params.replace('%7E', '~') # 生成签名 signature = base64.b64encode(hmac.new('%s&' % ACCESSKEY_SECRET, StringToSign, digestmod = hashlib.sha1).digest()) return urllib.quote(signature)
def formatItem(dict_list, cmd='add'): insert_items = [] for dict in dict_list: _dict = {} 略 insert_items.append({ 'cmd' : cmd, 'timestamp' : int(time.time()*1000), 'fields' : _dict }) return insert_items def dealDocGet(items): """ get请求 (预留方法) :param items: :return: """ if not items: return json.dumps({'status':'OK'}) curtime = int(time.time()) SignatureNonce = curtime + random.randrange(1000, 9999) timeStamp = datetime.datetime.utcnow().strftime("%Y-%m-%dT%H:%M:%SZ") params = [ ('AccessKeyId', ACCESSKEY_ID), ('SignatureMethod', 'HMAC-SHA1'), ('SignatureNonce', SignatureNonce), ('SignatureVersion', '1.0'), ('Version', 'v2'), ('Timestamp', timeStamp), ('action', action), ('table_name', table), ('items', json.dumps(items))] signature = getSignature(params= sorted(params, key=lambda x: x[0]) , method= 'POST') result_url = URL+ app + '?' + urllib.urlencode(params[:-3]) +'&Signature=' + signature req = urllib2.Request(result_url) req.add_data('&'.join([urllib.quote(str(k)) + '=' + urllib.quote(str(v)) for k, v in params[-3:]])) print urllib2.urlopen(req).read() if __name__ == '__main__': pass
阿里开放搜索 python的API简单封装
最新推荐文章于 2024-06-24 16:20:36 发布