taos.new_bind_params的timestamp绑定有很多隐式操作不容易理解造成bug

[lonelyleaf]

1. 参数中如果传入的是List[float]，会自动根据precision再做乘法，但int又不会，我批量写入的时候debug很久都没找到timestamp out of range的问题就是隐式做了乘法，既然我都指定精度了，为什么sdk内部还要乘法？

如果设计如此可不可以完善文档说明这些问题？

比如这样一个用例

timestamp_list  = (series.index.astype('int64') / 10 ** 6).tolist()
params[0].timestamp(timestamp_list)

会报错 Timestamp data out of range

stmt = c_void_p(2809582648032)

    def taos_stmt_execute(stmt):
        # type: (ctypes.c_void_p) -> None
        """Execute a statement query
        @stmt: TAOS_STMT*
        """
        res = _libtaos.taos_stmt_execute(stmt)
        if res != 0:
>           raise StatementError(msg=taos_stmt_errstr(stmt), errno=res)
E           taos.error.StatementError: [0x060b]: Timestamp data out of range

2. 所有的bind都不支持直接传入numpy的array，这点不太方便

3. 如果我传入一个List[datetime]但带时区的，绑定timestamp也会报错

from datetime import timedelta, datetime

import pandas as pd
import numpy as np
import taos
from iso8601 import iso8601
from sqlalchemy import create_engine, text
from taos import TaosMultiBind

conn = taos.TaosConnection = (
    taos.connect(host="127.0.0.1",
                 user="root",
                 port=6030,
                 password="taosdata",
                 timezone="Asia/Shanghai")
)

start_time = iso8601.parse_date('2024-04-28T00:00:00+08:00')
end_time = start_time + timedelta(hours=24)
time_range = pd.date_range(start_time, end_time, freq='min')
random_values = np.random.uniform(0, 45, len(time_range))
random_series = pd.Series(random_values, index=time_range)

stmt = conn.statement("INSERT INTO ? USING abc.metric TAGS(?) VALUES(?, ?, ?)")
tags = taos.new_bind_params(1)
tags[0].binary(['test_tag'])

params: TaosMultiBind = taos.new_bind_params(3)

time_list = random_series.index.to_pydatetime().tolist()
value_list = random_series.values.tolist()
gateway_id_list = np.full(random_series.size, 'test_id').tolist()
# datetime with timezone will throw exception
# TypeError: can't subtract offset-naive and offset-aware datetimes
params[0].timestamp(time_list)
params[1].float(value_list)
params[2].nchar(gateway_id_list)

conn.close()

# print index
print('batch write success')

error:

Traceback (most recent call last):
  File "E:\projects\python\py-wi-water-model-edge\venv\lib\site-packages\taos\bind.py", line 428, in timestamp
    buffer = cast(values, c_void_p)
  File "C:\Python310\lib\ctypes\__init__.py", line 510, in cast
    return _cast(obj, obj, typ)
ctypes.ArgumentError: argument 1: <class 'TypeError'>: wrong type

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "E:\projects\python\py-wi-water-model-edge\py_wi_water_model_edge\ts\tdengine_write_test.py", line 34, in <module>
    params[0].timestamp(time_list)
  File "E:\projects\python\py-wi-water-model-edge\venv\lib\site-packages\taos\bind.py", line 431, in timestamp
    buffer = buffer_type(*[_datetime_to_timestamp(value, precision) for value in values])
  File "E:\projects\python\py-wi-water-model-edge\venv\lib\site-packages\taos\bind.py", line 431, in <listcomp>
    buffer = buffer_type(*[_datetime_to_timestamp(value, precision) for value in values])
  File "E:\projects\python\py-wi-water-model-edge\venv\lib\site-packages\taos\bind.py", line 215, in _datetime_to_timestamp
    return int(round((value - _datetime_epoch).total_seconds() * 1000))
TypeError: can't subtract offset-naive and offset-aware datetimes
Exception ignored in: <function TaosConnection.__del__ at 0x000002C9AEB44280>
Traceback (most recent call last):
  File "E:\projects\python\py-wi-water-model-edge\venv\lib\site-packages\taos\connection.py", line 341, in __del__
  File "E:\projects\python\py-wi-water-model-edge\venv\lib\site-packages\taos\connection.py", line 62, in close
TypeError: 'NoneType' object is not callable