def __init__(self, task, pov_replace_unk):
import nltk
from nltk.translate.bleu_score import SmoothingFunction
self._smooth_fn = SmoothingFunction().method4
nltk.download('wordnet', quiet=True)
self.pov_replace_unk = pov_replace_unk
dataset_name = PurePosixPath(task.args.data.split(':')[0]).name
if dataset_name.startswith('java'):
import javalang
self._javalang = javalang
self._parse_method = self._parse_java
elif dataset_name.startswith('python'):
self._javalang = None
self._parse_method = self._parse_python
# To fix the bug of original dataset (a temporarily solution).
self._fix_python_orig_bug = dataset_name.endswith('orig')
else:
raise RuntimeError(
'code generation scorer requires only support Java and Python now, '
'dataset name must be started with "java" or "python"')
self.reset()
def parseBed2Json(bedfile):
# test if the file exitst
try:
# try resolve path and get the absolute path
my_abs_path_file = Path(bedfile).resolve()
###### Open bed file and parse to json #####
jsonData = []
# open bed file
infile = open(bedfile, 'r').readlines()
# read header
header = infile.pop(0).strip('\n').split("\t")
for line in infile:
# get row and convert to array
data = line.strip('\n').split("\t")
# test if the row is complete and have the same size as the header
if len(data) == len(header):
datadict = {}
# iterate over my bedfile row
for i, element in enumerate(data):
datadict[header[i]] = element
# push into array the dictionary structure
jsonData.append(datadict)
######## write into file the json structure
# bed filename (without path)
filename = PurePosixPath(my_abs_path_file).name
# json file name
jsonFile = filename + ".json"
# test if file ends with .bed
if filename.endswith('.bed'):
jsonFile = filename.split(".bed")[0] + ".json"
with open(jsonFile, 'w') as outfile:
json.dump(jsonData, outfile, indent=4)
outfile.close()
except FileNotFoundError:
print("your file does not exist")
