[BUG] ArrowArray failed with error

Describe the bug

ArrowArray failed with error.

2020-12-05 22:16:43,592 mars.scheduler.operands.common 57 ERROR    Attempt 4: Unexpected error TypeError occurred in executing operand 8d11327f36b6ad13ad323d4719ac9104 in 33.13.158.99:13525
Traceback (most recent call last):
  File "/home/admin/work/public-mars-0.5.2.zip/mars/promise.py", line 100, in _wrapped
    result = func(*args, **kwargs)
  File "/home/admin/work/public-mars-0.5.2.zip/mars/worker/calc.py", line 301, in <lambda>
    .then(lambda context_dict: _start_calc(context_dict)) \
  File "/home/admin/work/public-mars-0.5.2.zip/mars/worker/calc.py", line 276, in _start_calc
    return self._calc_results(session_id, graph_key, graph, context_dict, chunk_targets)
  File "/home/admin/work/public-mars-0.5.2.zip/mars/utils.py", line 377, in _wrapped
    return func(*args, **kwargs)
  File "/home/admin/work/public-mars-0.5.2.zip/mars/worker/calc.py", line 200, in _calc_results
    chunk_targets, retval=False).result()
  File "src/gevent/event.py", line 383, in gevent._gevent_cevent.AsyncResult.result
  File "src/gevent/event.py", line 305, in gevent._gevent_cevent.AsyncResult.get
  File "src/gevent/event.py", line 335, in gevent._gevent_cevent.AsyncResult.get
  File "src/gevent/event.py", line 323, in gevent._gevent_cevent.AsyncResult.get
  File "src/gevent/event.py", line 303, in gevent._gevent_cevent.AsyncResult._raise_exception
  File "/opt/conda/lib/python3.7/site-packages/gevent/_compat.py", line 65, in reraise
    raise value.with_traceback(tb)
  File "/opt/conda/lib/python3.7/site-packages/gevent/threadpool.py", line 142, in __run_task
    thread_result.set(func(*args, **kwargs))
  File "mars/actors/pool/gevent_pool.pyx", line 127, in mars.actors.pool.gevent_pool.GeventThreadPool._wrap_watch.inner
    result = fn(*args, **kwargs)
  File "/home/admin/work/public-mars-0.5.2.zip/mars/executor.py", line 690, in execute_graph
    res = graph_execution.execute(retval)
  File "/home/admin/work/public-mars-0.5.2.zip/mars/executor.py", line 571, in execute
    future.result()
  File "/opt/conda/lib/python3.7/concurrent/futures/_base.py", line 435, in result
    return self.__get_result()
  File "/opt/conda/lib/python3.7/concurrent/futures/_base.py", line 384, in __get_result
    raise self._exception
  File "/opt/conda/lib/python3.7/concurrent/futures/thread.py", line 57, in run
    result = self.fn(*self.args, **self.kwargs)
  File "/home/admin/work/public-mars-0.5.2.zip/mars/utils.py", line 451, in _inner
    return func(*args, **kwargs)
  File "/home/admin/work/public-mars-0.5.2.zip/mars/executor.py", line 443, in _execute_operand
    Executor.handle(first_op, results, self._mock)
  File "/home/admin/work/public-mars-0.5.2.zip/mars/executor.py", line 641, in handle
    return runner(results, op)
  File "/home/admin/work/public-pyodps-0.10.1.zip/odps/mars_extension/dataframe/datastore.py", line 243, in execute
    odps_schema = pd_to_df_schema(to_store_data, unknown_as_string=op.unknown_as_string)
  File "/home/admin/work/public-pyodps-0.10.1.zip/odps/df/backends/pd/types.py", line 92, in pd_to_df_schema
    name=names[i]))
  File "/home/admin/work/public-pyodps-0.10.1.zip/odps/df/backends/pd/types.py", line 54, in np_type_to_df_type
    for it in arr:
  File "/opt/conda/lib/python3.7/site-packages/pandas/core/arrays/base.py", line 352, in __iter__
    yield self[i]
  File "/home/admin/work/public-mars-0.5.2.zip/mars/dataframe/arrays.py", line 386, in __getitem__
    return cls(array[item], dtype=self._dtype)
  File "/home/admin/work/public-mars-0.5.2.zip/mars/dataframe/arrays.py", line 538, in __init__
    ArrowArray.__init__(self, values, ArrowStringDtype(), copy=copy)
  File "/home/admin/work/public-mars-0.5.2.zip/mars/dataframe/arrays.py", line 204, in __init__
    self._init_by_arrow(values, dtype=dtype, copy=copy)
  File "/home/admin/work/public-mars-0.5.2.zip/mars/dataframe/arrays.py", line 234, in _init_by_arrow
    arrow_array = pa.chunked_array([pa.array(values, type=dtype.arrow_type)])
  File "pyarrow/array.pxi", line 211, in pyarrow.lib.array
    return _sequence_to_array(obj, mask, size, type, pool, c_from_pandas)
  File "pyarrow/array.pxi", line 36, in pyarrow.lib._sequence_to_array
    check_status(ConvertPySequence(sequence, mask, options, &out))
TypeError: 'NoneType' object is not iterable

To Reproduce
To help us reproducing this bug, please provide information below:

1. Your Python version
2. The version of Mars you use
3. Versions of crucial packages, such as numpy, scipy and protobuf
4. Full stack of the error.
5. Minimized code to reproduce the error.

Start with mars/dataframe/arrays.py, especially ArrowArray._init_by_arrow and the ArrowStringArray constructor shown in the traceback. Trace the None value reaching pyarrow.array, then create a minimal reproduction from the reported stack and verify that the failing ArrowArray path no longer raises the TypeError.