def _safe_extract_all(self, zipfile, target_dir):
"""Safer version of ZipFile.extractall -- does not allow absolute or upwards-relative paths"""
for zipinfo in zipfile.infolist():
# skip absolute or upwards-relative files
if zipinfo.filename.startswith(('/', '..')):
warnings.warn('Skipping potentially unsafe file: ' + zipinfo.filename, RuntimeWarning)
continue
# target_dir is base directory; extract will create subpaths as necessary
zipfile.extract(zipinfo, target_dir)
def safe_extract_from_zip(zipfile, filename, out_dir):
try:
return zipfile.extract(filename, out_dir)
except (KeyError) as e:
pass
return ''
def uncompress(srcfile, destdir):
import gzip
import tarfile, zipfile
file = os.path.basename(srcfile)
if os.path.isfile(file):
shortname, fmt = os.path.splitext(file)
fmt = fmt[1:]
if fmt in ('tgz', 'tar'):
try:
tar = tarfile.open(srcfile)
names = tar.getnames()
for name in names:
tar.extract(name, destdir)
tar.close()
except Exception as e:
print("Can't uncompress {} for {}".format(file, e))
elif fmt == 'zip':
try:
zipfile = zipfile.ZipFile(srcfile)
for names in zipfile.namelist():
zipfile.extract(names, destdir)
zipfile.close()
except Exception as e:
print("Can't uncompress {} for {}".format(file, e))
elif fmt == 'gz':
try:
fname = os.path.join(destdir, os.path.basename(srcfile))
gfile = gzip.GzipFile(srcfile)
open(fname, "w+").write(gfile.read())
# gzip对象用read()打开后，写入open()建立的文件中。
gfile.close()
# 关闭gzip对象
except Exception as e:
return False, e, fmt
'''
elif fmt == 'rar':
try:
rar = rarfile.RarFile(srcfile)
os.chdir(destdir)
rar.extractall()
rar.close()
except Exception as e :
return (False, e, filefmt)
'''
else:
print('文件格式不支持或者不是压缩文件')
return None
def unzip(zip_file_path):
try:
zip_file = zipfile.ZipFile(zip_file_path, 'r')
content_files = zip_file.namelist()
for file_name in content_files:
if os.path.splitext(file_name)[1] == '.dblog':
zipfile.extract(zip_dir, file_name)
return zip_dir + '\\' + file_name
return '' # no debug log files found in the zip file
except:
if (os.path.getsize(zip_file_path) == 0):
print('empty file')
else:
print("FAILED to extract: " + zip_dir + '\\' + zip_file_name)
return ''
def arff_to_df(URL=URL_child,
arff_file='Autism-Child-Data.arff',
force_download=False):
if force_download or not os.path.exists(arff_file):
zipped = urlretrieve(URL, 'autism.zip')
zipfile = ZipFile(zipped[0], 'r')
arff_file = zipfile.extract(arff_file)
##extracting the data dictionary and column names (description)
data, description = arff.loadarff(arff_file)
columns = [i for i in description]
new_columns = [
'A1_Score', 'A2_Score', 'A3_Score', 'A4_Score', 'A5_Score', 'A6_Score',
'A7_Score', 'A8_Score', 'A9_Score', 'A10_Score', 'age', 'gender',
'ethnicity', 'jaundice', 'autism', 'country_of_res', 'used_app_before',
'result', 'age_range', 'relation', 'Class/ASD'
]
df = pd.DataFrame(data, columns=columns)
df.columns = new_columns
#changing utf-8 coding to categorical variables
for column in list(df.select_dtypes(include=['object']).columns):
df[column] = df[column].str.decode('utf-8').astype('category')
#Fixing ethnicity discrepancies
df.ethnicity = df.ethnicity.str.replace('?', "Unknown").str.replace(
' ',
'_').str.replace('\'Middle_Eastern_\'', "Middle_Eastern").str.replace(
'\'South_Asian\'',
'South_Asian').str.replace('others', 'Others').astype('category')
#Fill in missing age with median age
df.age = df.age.fillna(value=df.age.median())
#Fixing relation discrepancies
relation_mapper = {
'Parent': 'Family Member',
'Relative': 'Family Member',
'\'Health care professional\'': 'Health care professional',
'?': 'Unknown',
'Self': 'Self',
'self': 'Self'
}
df.relation = df.relation.map(relation_mapper).astype('category')
df.relation = df.relation.fillna("Unknown")
#Fixing various binary inputs
jaun_mapper = {'yes': 1, 'no': 0}
aut_mapper = {'yes': 1, 'no': 0}
class_mapper = {'YES': 1, 'NO': 0}
app_mapper = {'yes': 1, 'no': 0}
mapper_list = list([jaun_mapper, aut_mapper, class_mapper, app_mapper])
for x, y in zip(['jaundice', 'autism', 'Class/ASD', 'used_app_before'],
mapper_list):
df[x] = df[x].map(y)
df = df.drop('country_of_res', axis=1)
return df
