Re: [Tarantool-patches] [PATCH] test: make strings compatible with Python 3

[Sergey Bronnikov <sergeyb@tarantool.org>]

Hi,


thanks for review


On 23.12.2020 12:59, Leonid Vasiliev wrote:
> Hi! Thank you for the patch.
> Generally I'm comfortable with the patch. But I have a few questions:
>
> On 22.12.2020 11:15, sergeyb@tarantool.org wrote:
>> From: Sergey Bronnikov <sergeyb@tarantool.org>
>>
>> The largest change in Python 3 is the handling of strings.
>> In Python 2, the str type was used for two different
>> kinds of values – text and bytes, whereas in Python 3,
>> these are separate and incompatible types.
>> Patch converts strings to byte strings where it is required
>> to make tests compatible with Python 3.
>>
>> Part of #5538
>> ---
>>   test/box-py/bad_trigger.test.py      |  2 +-
>>   test/box-py/iproto.test.py           | 42 +++++++++++++++++-----------
>>   test/replication-py/conflict.test.py |  5 ++--
>>   3 files changed, 29 insertions(+), 20 deletions(-)
>>
>> diff --git a/test/box-py/bad_trigger.test.py 
>> b/test/box-py/bad_trigger.test.py
>> index 9dd6e17c9..03b7efcda 100644
>> --- a/test/box-py/bad_trigger.test.py
>> +++ b/test/box-py/bad_trigger.test.py
>> @@ -41,7 +41,7 @@ unpacker.feed(packet)
>>   header = unpacker.unpack()
>>   body = unpacker.unpack()
>>   print("error code {}".format((header[IPROTO_CODE] & 
>> (REQUEST_TYPE_ERROR - 1))))
>> -print("error message:  {}".format(body[IPROTO_ERROR]))
>> +print("error message: {}".format(body[IPROTO_ERROR].decode("utf-8")))
>>   print("eof: {}".format(len(s.recv(1024)) == 0))
>>   s.close()
>>   diff --git a/test/box-py/iproto.test.py b/test/box-py/iproto.test.py
>> index 6723e03f5..ff0469f10 100644
>> --- a/test/box-py/iproto.test.py
>> +++ b/test/box-py/iproto.test.py
>> @@ -1,5 +1,6 @@
>>   from __future__ import print_function
>>   +import binascii
>
> Where is the module used?

Left after another implementation, removed and force-pushed.


>
>>   import os
>>   import sys
>>   import struct
>> @@ -163,7 +164,7 @@ class RawInsert(Request):
>>         def __init__(self, conn, space_no, blob):
>>           super(RawInsert, self).__init__(conn)
>> -        request_body = "\x82" + msgpack.dumps(IPROTO_SPACE_ID) + \
>> +        request_body = b'\x82' + msgpack.dumps(IPROTO_SPACE_ID) + \
>>               msgpack.dumps(space_id) + msgpack.dumps(IPROTO_TUPLE) + 
>> blob
>>           self._body = request_body
>>   @@ -172,7 +173,7 @@ class RawSelect(Request):
>>         def __init__(self, conn, space_no, blob):
>>           super(RawSelect, self).__init__(conn)
>> -        request_body = "\x83" + msgpack.dumps(IPROTO_SPACE_ID) + \
>> +        request_body = b'\x83' + msgpack.dumps(IPROTO_SPACE_ID) + \
>>               msgpack.dumps(space_id) + msgpack.dumps(IPROTO_KEY) + 
>> blob + \
>>               msgpack.dumps(IPROTO_LIMIT) + msgpack.dumps(100);
>>           self._body = request_body
>> @@ -182,13 +183,13 @@ space = c.space("test")
>>   space_id = space.space_no
>>     TESTS = [
>> -    (1,     "\xa1", "\xd9\x01", "\xda\x00\x01", 
>> "\xdb\x00\x00\x00\x01"),
>> -    (31,    "\xbf", "\xd9\x1f", "\xda\x00\x1f", 
>> "\xdb\x00\x00\x00\x1f"),
>> -    (32,    "\xd9\x20", "\xda\x00\x20", "\xdb\x00\x00\x00\x20"),
>> -    (255,   "\xd9\xff", "\xda\x00\xff", "\xdb\x00\x00\x00\xff"),
>> -    (256,   "\xda\x01\x00", "\xdb\x00\x00\x01\x00"),
>> -    (65535, "\xda\xff\xff", "\xdb\x00\x00\xff\xff"),
>> -    (65536, "\xdb\x00\x01\x00\x00"),
>> +    (1,     b'\xa1', b'\xd9\x01', b'\xda\x00\x01', 
>> b'\xdb\x00\x00\x00\x01'),
>> +    (31,    b'\xbf', b'\xd9\x1f', b'\xda\x00\x1f', 
>> b'\xdb\x00\x00\x00\x1f'),
>> +    (32,    b'\xd9\x20', b'\xda\x00\x20', b'\xdb\x00\x00\x00\x20'),
>> +    (255,   b'\xd9\xff', b'\xda\x00\xff', b'\xdb\x00\x00\x00\xff'),
>> +    (256,   b'\xda\x01\x00', b'\xdb\x00\x00\x01\x00'),
>> +    (65535, b'\xda\xff\xff', b'\xdb\x00\x00\xff\xff'),
>> +    (65536, b'\xdb\x00\x01\x00\x00'),
>
> Why do you choose single quotes for byte strings and double quotes for
> strings? Maybe we will use double quotes for all strings? (applies to
> the whole patch)

I use double quotes for strings and single quotes for bytes literals, 
see [1].

1. https://www.python.org/dev/peps/pep-3112/

>
>>   ]
>>     for test in TESTS:
>> @@ -197,16 +198,19 @@ for test in TESTS:
>>       print("STR", size)
>>       print("--")
>>       for fmt in it:
>> -        print("0x" + fmt.encode("hex"), "=>", end=" ")
>> +        if sys.version[0] == '2':
>> +            print("0x" + fmt.encode("hex"), "=>", end=" ")
>> +        else:
>> +            print("0x" + fmt.hex(), "=>", end=" ")
>>           field = "*" * size
>> -        c._send_request(RawInsert(c, space_id, "\x91" + fmt + field))
>> +        c._send_request(RawInsert(c, space_id, b'\x91' + fmt + 
>> field.encode("utf-8")))
>>           tuple = space.select(field)[0]
>>           print(len(tuple[0])== size and "ok" or "fail", end=" ")
>>           it2 = iter(test)
>>           next(it2)
>>           for fmt2 in it2:
>>               tuple = c._send_request(RawSelect(c, space_id,
>> -                "\x91" + fmt2 + field))[0]
>> +                b'\x91' + fmt2 + field.encode("utf-8")))[0]
>>               print(len(tuple[0]) == size and "ok" or "fail", end=" ")
>>           tuple = space.delete(field)[0]
>>           print(len(tuple[0]) == size and "ok" or "fail", end="")
>> @@ -357,15 +361,18 @@ s = c._socket
>>   header = { "hello": "world"}
>>   body = { "bug": 272 }
>>   resp = test_request(header, body)
>> -print("sync={}, {}".format(resp["header"][IPROTO_SYNC], 
>> resp["body"].get(IPROTO_ERROR)))
>> +print("sync={}, {}".format(resp["header"][IPROTO_SYNC],
>> +        resp["body"].get(IPROTO_ERROR).decode("utf-8")))
>>   header = { IPROTO_CODE : REQUEST_TYPE_SELECT }
>>   header[IPROTO_SYNC] = 1234
>>   resp = test_request(header, body)
>> -print("sync={}, {}".format(resp["header"][IPROTO_SYNC], 
>> resp["body"].get(IPROTO_ERROR)))
>> +print("sync={}, {}".format(resp["header"][IPROTO_SYNC],
>> +        resp["body"].get(IPROTO_ERROR).decode("utf-8")))
>>   header[IPROTO_SYNC] = 5678
>>   body = { IPROTO_SPACE_ID: 304, IPROTO_KEY: [], IPROTO_LIMIT: 1 }
>>   resp = test_request(header, body)
>> -print("sync={}, {}".format(resp["header"][IPROTO_SYNC], 
>> resp["body"].get(IPROTO_ERROR)))
>> +print("sync={}, {}".format(resp["header"][IPROTO_SYNC],
>> +        resp["body"].get(IPROTO_ERROR).decode("utf-8")))
>>   c.close()
>>     @@ -424,7 +431,10 @@ header = { IPROTO_CODE: REQUEST_TYPE_CALL, 
>> IPROTO_SYNC: 100 }
>>   body = { IPROTO_FUNCTION_NAME: "kek" }
>>   resp = test_request(header, body)
>>   print("Sync: ", resp["header"][IPROTO_SYNC])
>> -print("Retcode: ", resp["body"][IPROTO_DATA])
>> +body = resp["body"][IPROTO_DATA]
>> +if sys.version[0] == '3':
>> +    body = [body[0].decode("utf-8")]
>> +print("Retcode: ", body)
>>     c.close()
>>   diff --git a/test/replication-py/conflict.test.py 
>> b/test/replication-py/conflict.test.py
>> index 5e19d0c40..925ae046b 100644
>> --- a/test/replication-py/conflict.test.py
>> +++ b/test/replication-py/conflict.test.py
>> @@ -20,9 +20,8 @@ replica.deploy()
>>   def parallel_run(cmd1, cmd2, compare):
>>       print("parallel send: {}".format(cmd1))
>>       print("parallel send: {}".format(cmd2))
>> -    master.admin.socket.sendall("{}\n".format(cmd1))
>> -    replica.admin.socket.sendall("{}\n".format(cmd2))
>> -
>> +    master.admin.socket.sendall(cmd1.encode("utf-8") + b'\n')
>> +    replica.admin.socket.sendall(cmd2.encode("utf-8") + b'\n')
>>       master.admin.socket.recv(2048)
>>       replica.admin.socket.recv(2048)
>>