Created
August 30, 2021 22:37
-
-
Save haniefhan/8eb52ca0ea4c05db4a17958432e877c6 to your computer and use it in GitHub Desktop.
Parse Aceh CSV test 1
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
import csv | |
import json | |
province = [] # provinsi | |
regency_or_city = [] # kabupaten or kota | |
district = [] # kecamatan | |
administrative_village_or_village = [] # kelurahan or desa | |
# check if string has a number | |
def hasANumber(inputString): | |
return any(char.isdigit() for char in inputString) | |
# get region parent | |
# example: 11.01.01.2001 | |
# 11 is for province code | |
# 11.01 is for regency or city code | |
# 11.01.01 is for district code | |
# 11.01.01.2001 is for administrative village or village code | |
def getRegionParent(ret, code): | |
prev = "" | |
for index, cd in enumerate(code.split(".")): | |
if index == 0: | |
ret["prov_code"] = cd | |
prev = cd | |
if index == 1: | |
ret["kab_code"] = prev + "." + cd | |
prev = prev + "." + cd | |
if index == 2: | |
ret["kec_code"] = prev + "." + cd | |
prev = prev + "." + cd | |
if index == 3: | |
ret["kel_code"] = prev + "." + cd | |
prev = prev + "." + cd | |
return ret | |
# parse region name | |
# because sometimes region name have a order number we don't need | |
def parseName(name): | |
return ''.join([i for i in name if not i.isdigit()]) | |
with open("csv/11-aceh.csv") as csv_file: | |
# read csv file | |
csv_reader = csv.reader(csv_file, delimiter=",", dialect=csv.excel) | |
for index, row in enumerate(csv_reader): | |
code = row[0].strip() | |
if hasANumber(code): | |
if row[1] != '': | |
# must be a province or a regency or city | |
name = row[1].strip() | |
if not hasANumber(name): | |
# province | |
province.append({ | |
"prov_code": code, | |
"prov_name": name | |
}) | |
else: | |
# regency or city | |
name = parseName(name) | |
kab_type = "Kabupaten" | |
if "KAB. " not in name: | |
kab_type = "Kota" | |
kab = { | |
"kab_code": code, | |
"kab_name": name, | |
"kab_type": kab_type, | |
} | |
# we need to know which region parent it from | |
kab = getRegionParent(kab, code) | |
regency_or_city.append(kab) | |
elif row[4] != '' and hasANumber(row[4]): | |
# must be a district | |
name = parseName(row[4].strip()) | |
district.append({ | |
'kec_code': code, | |
'kec_name': name | |
}) | |
elif row[5] != '' and hasANumber(row[5]): | |
name = parseName(row[6].strip()) | |
administrative_village_or_village.append({ | |
'kel_code': code, | |
'kel_name': name, | |
'kel_type': "Kelurahan" | |
}) | |
elif row[6] != '' and hasANumber(row[6]): | |
name = parseName(row[6].strip()) | |
administrative_village_or_village.append({ | |
'kel_code': code, | |
'kel_name': name, | |
'kel_type': "Desa" | |
}) | |
data = { | |
'province': province, | |
'regency_or_city': regency_or_city, | |
'district': district, | |
'administrative_village_or_village': administrative_village_or_village, | |
} | |
with open('json_result/11-aceh.json', 'w') as outfile: | |
json.dump(data, outfile) |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment