#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Thu Mar 31 16:36:28 2022
@author: cuxac
"""
from geopy.geocoders import Nominatim
#import time
import functools
import unidecode
import re
#import pycountry
from iso3166 import countries
import json
import sys
osm = Nominatim(timeout=60,user_agent='inistTDM')
locator = functools.lru_cache(maxsize=128)(functools.partial(osm.geocode, timeout=60,language='en'))
locator_rev = functools.lru_cache(maxsize=128)(functools.partial(osm.reverse, timeout=60,language='en'))
#fin=json.loads(json.dumps([{
# "id":1,
# "value":"université sciences et technologies bordeaux 1 institut national de physique nucléaire et de physique des particules du cnrs in2p3 UMR5797"},
# {"id":2,
# "value":"uar76 / ups76 centre national de la recherche scientifique cnrs institut de l'information scientifique et technique inist"},
# {"id":3,
# "value":"centre de recherches sur la géologie des matières premières minérales et énergétiques cregu université de lorraine ul umr7359 centre national de la recherche scientifique"},
# {"id":4,
# "value":"auf der morgenstelle 8, 72076 tuebingen"
# },
# {"id":5,
# "value":"z.i. de kermelin,16, rue ampère, 56017 Vannes"
# },
# {"id":6,
# "value":"campus de santa apolónia, 5300-253 bragança"},
# {"id":7,
# "value":"campus romanus de maron"} ,
# {"id":8,
# "value":"Inist-CNRS, vandoeuvre les Nancy, France"}
# ]))
def placeaddress(address):
place, (lat, lng)=locator(address)
res=locator_rev(str(lat)+","+str(lng))
try :
country=res.raw['address']['country']
except KeyError:
country=res.raw['address']['country_code']
try:
village=res.raw['address']['village']
except KeyError:
village=''
return country,lat,lng,village
#for line in fin:
for line in sys.stdin:
country=""
data=json.loads(line)
aff=data['value']
lines=aff.replace('"','')
adr=lines.strip()
adr=adr.replace('&',';')
ad=adr.split(';')
ads=[i.strip() for i in ad]
for a in set(ads) :
a=a.lower()
a=a.replace('cedex','')
a=re.sub(' U[0-9]+','',a,flags=re.IGNORECASE)
a=re.sub("both in ",'',a,flags=re.IGNORECASE)
a=re.sub(' UMR[0-9]+','',a,flags=re.IGNORECASE)
a=a.replace('...',' ')
a=a.replace('_',' ')
a=re.sub('\([a-zA-Z0-9 ,:\–]+\)','',a)
##-1 on traite avec libpostal et on regarde s'il y a un pays
#nl=parse_address(a.strip().strip('.'))
#d=dict(map(reversed,nl))
ars=re.split(',|:',a)
if re.match(r'u[amsrp]+ ?[0-9]+',a):
country='France'
else:
try:
#country=d['country']
country=countries.get(ars[-1].strip()).name
# if len(country)==3:
# cd=country.upper()
# #x=pycountry.countries.get(alpha_3=cd)
# country=countries.get(cd).name
#
# #country=x.name
country=placeaddress(country)[0]
except (KeyError,TypeError) as error:
#- 2 s'il n'y a pas de pays avec libpostal on essaye avec nominatim de trouver un pays
try:
country=placeaddress(a)[0]
if placeaddress(a)[3].lower()!=a.lower():
country=country
#-3 si ça ne fonctionne pas on essaye de spliter l'adresse et de traiter les éléments séparéément
except (KeyError,TypeError) as error:
a=a.lower()
a2=''
if "university" in unidecode.unidecode(a).lower():#.split()[:-1]:
a1=re.findall("faculty of [a-zA-Z]+",unidecode.unidecode(a), flags=re.IGNORECASE)
a11=re.findall("college of [a-zA-Z]+",unidecode.unidecode(a), flags=re.IGNORECASE)
if len(a1)>0:
a=a.replace(a1[0],'')
if len(a11)>0:
a=a.replace(a11[0],'')
if "university" in a.lower().split()[:-1]:
a=a.replace(' hospital ',' ')
a2=re.findall("university [a-zA-Z&]+",unidecode.unidecode(a), flags=re.IGNORECASE)
if len(a2)>1:
a3=a2[0].replace("university of ","university ")
if len(a2)<1:
a2=unidecode.unidecode(a).lower().replace('university','$ university').split('$ ')
a2.insert(0,a3)
try:
a2.remove('university of science')
except ValueError:
uos=0
elif ("university" ==a.split()[-1].lower()) or ("university," in a.lower()):
a2=re.findall("[ a-zA-Z]+ university",unidecode.unidecode(a), flags=re.IGNORECASE)
a2.insert(0,'')
elif "universite " in unidecode.unidecode(a).lower():
a2=re.findall("universite (?:[A-Za-z]+\s){1}[A-Za-z]+",unidecode.unidecode(a), flags=re.IGNORECASE)
a3=re.findall("(?:[A-Za-z]+\s){1}[A-Za-z]+ universite",unidecode.unidecode(a), flags=re.IGNORECASE)
if len(a3)==0:
a3=[' ']
a2.insert(0,a3[0])
if len(a2)<2:
a2=unidecode.unidecode(a).lower().replace('universite','$ universite').split('$ ')
elif "universidad " in unidecode.unidecode(a).lower():
a2=re.findall("universidad [ a-zA-Z]+",unidecode.unidecode(a), flags=re.IGNORECASE)
a2.insert(0,'')
if len(a2)<2:
a2=unidecode.unidecode(a).lower().replace('universidad','$ universidad').split('$ ')
elif "universitat " in unidecode.unidecode(a).lower():
a2=re.findall("universitat [ a-zA-Z]+",unidecode.unidecode(a), flags=re.IGNORECASE)
a2.insert(0,'')
if len(a2)<2:
a2=unidecode.unidecode(a).lower().replace('universitat','$ universitat').split('$ ')
elif "hopital " in unidecode.unidecode(a).lower():
a2=re.findall("hopital [a-zA-Z']+",unidecode.unidecode(a), flags=re.IGNORECASE)
a2.insert(0,'')
elif ("chu" and "chru") in unidecode.unidecode(a).lower():
if ("chu de " and "chru de ") in unidecode.unidecode(a).lower():
a2=re.findall("ch[r]?u de [a-zA-Z']+",unidecode.unidecode(a), flags=re.IGNORECASE)
a2.insert(0,'')
else:
a2=re.findall("ch[r]?u [a-zA-Z']+",unidecode.unidecode(a), flags=re.IGNORECASE)
a3=a2[0].replace("chru","chu")
a2.insert(0,a3)
elif ',' in a:
a2=a.split(',')
elif ':' in a:
a2=a.split(':')
elif " et " in a.split()[:-2]:
a2=a.split(' et ')
elif (' — ' and ' - ') in a:
if ' — ' in a:
a2=a.split(' — ')
else:
a2=a.split(' - ')
elif ' an der ' in a:
a2=a.split(' an der ')
elif '-' in a:
a2=re.findall(r'\w+(?:-\w+)+',a)
a2.insert(0,'')
if len(a2)>1:
lon=1
for aa in reversed(a2):
aa=aa.strip()
#time.sleep(1)
try:
country=placeaddress(aa)[0]
lat=placeaddress(aa)[1]
lng=placeaddress(aa)[2]
break
except TypeError:
to=1
if lon==len(a2):
country="unknown"
lon+=1
#
# elif re.match(r'u[amsrp]+ ?[0-9]+',a):
# country='France'
else:
country="Unknown"
try:
data["value"]=[country, countries.get(country).alpha3]
except KeyError:
data["value"]=[country, 'N/A']
sys.stdout.write(json.dumps(data))
sys.stdout.write('\n')
