Initial test of encoding efficiency in various formats under Python (Json,bson, bz2, Lzma,msgpack)

The # -*- coding:utf-8 -*-' test proves that Bson and JSON require a similar number of bytes, or more. The effect of the compression algorithm depends on its degree of redundancy. "' From __future__ import divisionimport collections  # from python  standard library.import datetimetest = {"name": "Lemo",  "Age":12,          "Address":{              "City": "Suzhou",             "Country": "China",             "code":215000} ,             # "Timestamp":d atetime.datetime.now (),      "scores": [{"Name": "中文版", "Grade": 3.0, "code":215000},               {"name": "Chinese", "Grade": 2.0, "code":215000},                {"name": "Kerea", "Grade": 3.0, "code":232000},               {"Name": "France", "Grade": 4.0, "code":235000},               {"name": "Japanese", "Grade": 4.0, "code":235000},               {"name": "Kerea", "Grade": 4.0, " Code ": 235000},              {" name ":" Kerea "," Grade ": 4.0," code ": 235300}    ]} #import  bson#from bson.codec_options  import codecoptions#data = bson. Bson.encode (test) #print   ("BSON:",  len (data)  ) #decoded_doc  = bson. Bson.decode (data) import jsonjson_str = json.dumps ( test ) print  ("JSON:", Len (JSON _STR)  ) ### import bz2compressor = bz2. Bz2compressor () compressor.compress (JSON_str.encode (' Utf-8 '))  bz2_str = compressor.flush ()  print  ("bz2:", Len (BZ2_STR),   len (BZ2_STR)/len (json_str)  ) Import sysif sys.version_info.major*1000+ sys.version_ Info.minor>=3003:    import lzma    lzc = lzma. Lzmacompressor ()     out1 = lzc.compress (Json_str.encode (' Utf-8 ')  )      out2 = lzc.flush ()     # concatenate all the  partial results:    result = b "". Join ([Out1, out2])      print  ("Lzma:", Len (Result),   len (Result)/len (JSON_STR)  )      lzd = lzma. Lzmadecompressor ()     dec_str = lzd.decompress (result )      assert json_str.encode (' Utf-8 ')  == dec_str### msgpackimport datetimeimport  msgpack#useful_dict = {    # "id":  1,    # "created":  Datetime.datetime.now (), #}def decode_datetime (obj):     if b ' __datetime__ '   In obj:        obj = datetime.datetime.strptime (obj["as _str "], "%y%m%dt%h:%m:%s.%f ")     return objdef encode_datetime (obj):     if isinstance (obj, datetime.datetime):         return {' __datetime__ ': true,  ' as_str ':  obj.strftime ("%y%m%dt%h:%m:%s.%f")}     RETURN OBJPACKED_DICT = MSGPACK.PACKB (Test, default=encode_datetime) print  ("Msgpack:", Len (packed_dict),   len (packed_dict)/len (json_str)  ) #compressor  =  bz2. Bz2compressor (compresslevel=1) #compressor. Compress (packed_dict)   #bz2_str  = compressor.flush ()   #print  ("msgpack_bz2:", Len (BZ2_STR),   len (BZ2_STR)/len (json_str)  ) this_dict_again =  MSGPACK.UNPACKB (Packed_dict, object_hook=decode_datetime)

Initial test of encoding efficiency in various formats under Python (Json,bson, bz2, Lzma,msgpack)