BUG: read_parquet unexpected output with pyarrow 0.16.0 and nullable usigned int dtype

[cat-Yu]

data = [np.nan, 1, 2, 3, 4]
df = pd.DataFrame()
df['col'] = pd.Series(data, dtype = "UInt32")
df.to_parquet('example.parquet')
res = pd.read_parquet('example.parquet')
print(res)

Problem description

The above cope prints:

    col
0  <NA>
1     0
2     1
3     0
4     2

This happened after I upgraded to pyarrow 0.16.0, and UInt32 is the only dtype that I had trouble with.
The problem should be with read_parquet, I tried reading parquet with another package, and got the expected output.

Expected Output

    col
0  <NA>
1     1
2     2
3     3
4     4

Output of pd.show_versions()

INSTALLED VERSIONS

commit : None
python : 3.7.3.final.0
python-bits : 64
OS : Darwin
OS-release : 18.7.0
machine : x86_64
processor : i386
byteorder : little
LC_ALL : None
LANG : en_US.UTF-8
LOCALE : en_US.UTF-8

pandas : 1.0.1
numpy : 1.17.3
pytz : 2019.3
dateutil : 2.8.1
pip : 20.0.2
setuptools : 45.2.0.post20200210
Cython : None
pytest : 5.3.2
hypothesis : None
sphinx : 2.3.0
blosc : None
feather : None
xlsxwriter : None
lxml.etree : 4.4.2
html5lib : None
pymysql : None
psycopg2 : None
jinja2 : 2.11.1
IPython : 7.11.1
pandas_datareader: None
bs4 : None
bottleneck : None
fastparquet : None
gcsfs : None
lxml.etree : 4.4.2
matplotlib : None
numexpr : None
odfpy : None
openpyxl : None
pandas_gbq : None
pyarrow : 0.16.0
pytables : None
pytest : 5.3.2
pyxlsb : None
s3fs : 0.4.0
scipy : 1.4.0
sqlalchemy : None
tables : None
tabulate : None
xarray : None
xlrd : None
xlwt : None
xlsxwriter : None
numba : 0.48.0

[jorisvandenbossche]

@cat-Yu Thanks for the report!

So it seems that "uint32" data cannot be stored in parquet, and they are rather stored as int64. When reading back in, pandas still assumes that the data is uint32 (without checking that is actually the case) and thus incorrectly interprets the int64 data as uint32.