阿里開放搜索 python的API簡單封裝

爲公司開發搜索功能的時候使用到了阿里的開放搜索,是使用的python,官方沒有sdk,只能自己實現api
這裏把一些成熟的代碼放上來,就是些簡單的功能,包括:

1、文檔增刪改(dealDocPost)
2、post查詢(sendQuery)
3、下拉提示(suggest)
4、生成簽名(getSignature)
5、get查詢(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 = 自己的密碼
table            =  helloworld
app              =  helloworld
suggest_name     = 'helloworld_rule' # 下拉查詢規則
suggest_hit      = 5
query_hit        = 20
action           = 'push'
def dealDocPost(items): """post請求,增加,刪除,修改,都可以從這裏,唯一需要變化的是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):
    # 這裏是查詢:
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 result
def doc_search(query, fields = None, summary = None, formula_name = None, first_formula_name = None, index_name = None):
    """
    搜索查詢

    :param query                :   搜索主體,包含所有的查詢條件。子句分別包含config子句、query子句、sort子句、filter子句、aggregate子句、distinct子句等。
    :param fields               :   如果指定了此參數,則服務返回的數據集結構只包含fetch_fields裏邊指定的field,指定多個field可以使用“;”分隔。
    :param summary              :   動態摘要的配置,可以指定某些字段的飄紅、截斷等操作。
    :param formula_name         :   用戶指定的精排表達式名稱,搜索會獲取您在網站中定義的此名稱的表達式,並根據此來搜索。
    :param first_formula_name   :   用戶指定的粗排表達式名稱,搜索會獲取您在網站中定義的此名稱的表達式,並根據此來搜索。
    :param index_name           :   用戶要查詢的應用名稱,需爲UTF-8編碼的urlencode結果,多應用查詢可以使用“;”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')]

    # 其他搜索條件
    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 None

def suggest(query): ''' 用戶輸入數據獲取下列列表提示 :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'):
    """生成簽名
    . 參考文檔: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', '~')
    # 生成簽名
    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請求 (預留方法)
    :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
發表評論
所有評論
還沒有人評論,想成為第一個評論的人麼? 請在上方評論欄輸入並且點擊發布.
相關文章