[SPARK-19872] [PYTHON] Use the correct deserializer for RDD construction for coalesce/repartition

## What changes were proposed in this pull request?

This PR proposes to use the correct deserializer, `BatchedSerializer` for RDD construction for coalesce/repartition when the shuffle is enabled. Currently, it is passing `UTF8Deserializer` as is not `BatchedSerializer` from the copied one.

with the file, `text.txt` below:

```
a
b

d
e
f
g
h
i
j
k
l

```

- Before

```python
>>> sc.textFile('text.txt').repartition(1).collect()
```

```
UTF8Deserializer(True)
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File ".../spark/python/pyspark/rdd.py", line 811, in collect
    return list(_load_from_socket(port, self._jrdd_deserializer))
  File ".../spark/python/pyspark/serializers.py", line 549, in load_stream
    yield self.loads(stream)
  File ".../spark/python/pyspark/serializers.py", line 544, in loads
    return s.decode("utf-8") if self.use_unicode else s
  File "/System/Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7/encodings/utf_8.py", line 16, in decode
    return codecs.utf_8_decode(input, errors, True)
UnicodeDecodeError: 'utf8' codec can't decode byte 0x80 in position 0: invalid start byte
```

- After

```python
>>> sc.textFile('text.txt').repartition(1).collect()
```

```
[u'a', u'b', u'', u'd', u'e', u'f', u'g', u'h', u'i', u'j', u'k', u'l', u'']
```

## How was this patch tested?

Unit test in `python/pyspark/tests.py`.

Author: hyukjinkwon <gurwls223@gmail.com>

Closes #17282 from HyukjinKwon/SPARK-19872.

(cherry picked from commit 7387126f)
Signed-off-by: Davies Liu <davies.liu@gmail.com>