Last active
August 2, 2019 02:22
-
-
Save ohoroyoi/3d616c2f94e39e32455db682dd50229e to your computer and use it in GitHub Desktop.
get passenger airlines from https://en.wikipedia.org/wiki/List_of_passenger_airlines
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 requests | |
from lxml import etree | |
import wikipedia | |
import csv | |
import wikipediaapi | |
wiki_en = wikipediaapi.Wikipedia('en') | |
wiki_en = wikipediaapi.Wikipedia( | |
language='en', | |
extract_format=wikipediaapi.ExtractFormat.WIKI | |
) | |
with open("/Users/n18016/Documents/workspace/tour-geolocation-data/get_these_airlines.txt", "r") as f: | |
airline_list = list(f) | |
with open('/Users/n18016/Documents/workspace/tour-geolocation-data/works/airlines/airline_detail_6.csv', 'w', newline='') as outcsv: | |
fields = ["airline", "IATA", "ICAO", "Callsign", "Destination count", "url"] | |
writer = csv.DictWriter(outcsv, fieldnames=fields) | |
writer.writeheader() | |
for line in airline_list: | |
if line is not '\n': | |
print("**************************************************") | |
print(line) | |
if any(x in line for x in wikipedia.search(line)): | |
print("이거 ") | |
print(line) | |
try: | |
each_airline = wikipedia.page(line) | |
print("존재함: ") | |
print("each_airline.url : ") | |
print(each_airline.url) | |
except wikipedia.exceptions.DisambiguationError as e: | |
print(line, file=open('/Users/n18016/Documents/workspace/tour-geolocation-data/works/airlines/airline_error_list.txt', 'a')) | |
except wikipedia.exceptions.PageError as e: | |
print(line, file=open('/Users/n18016/Documents/workspace/tour-geolocation-data/works/airlines/airline_error_list.txt', 'a')) | |
pass | |
print("**************************************************") | |
if each_airline.url is not '': | |
req = requests.get(each_airline.url) | |
store = etree.fromstring(req.text) | |
airline_codes = store.xpath('//td[@class="nickname"]') | |
destination_count = store.xpath('//table[@class="infobox vcard"]//tr[th/text()="Destinations"]/td') | |
one_row = { "airline" : line.rstrip('\n') } | |
if len(destination_count) > 0: | |
one_row["Destination count"] = destination_count[0].text | |
else: | |
one_row["Destination count"] = "-" | |
# one_row = { "airline" : line.rstrip('\n') } | |
div_cnt = 0 | |
if len(airline_codes) == 3: | |
for div in airline_codes: | |
if div.text is not None: | |
one_row[div_cnt] = (div.text).rstrip('\n') | |
elif div.text is '': | |
one_row[div_cnt] = "-" | |
else: | |
one_row[div_cnt] = "-" | |
div_cnt = div_cnt + 1 | |
one_row["IATA"] = one_row.pop(0) | |
one_row["ICAO"] = one_row.pop(1) | |
one_row["Callsign"] = one_row.pop(2) | |
elif len(airline_codes) > 0 & len(airline_codes) != 3: | |
print(line, file=open('/Users/n18016/Documents/workspace/tour-geolocation-data/works/airlines/airline_error_list.txt', 'a')) | |
pass | |
else: | |
one_row["IATA"] = "-" | |
one_row["ICAO"] = "-" | |
one_row["Callsign"] = "-" | |
one_row["url"] = each_airline.url | |
print("one_row: ") | |
print(one_row) | |
writer.writerow(one_row) | |
else: | |
print("없음 : ") | |
print(line, file=open('/Users/n18016/Documents/workspace/tour-geolocation-data/works/airlines/airline_error_list.txt', 'a')) | |
pass |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment