scientific-programming-exer.../exam/ex01/receive.py

81 lines
2.2 KiB
Python
Raw Normal View History

2019-02-02 10:18:57 +00:00
import logging
from url import construct_url
2019-02-02 15:06:57 +00:00
from proxy import get_data_with_proxy, NoMoreProxiesException
2019-02-02 10:18:57 +00:00
logger = logging.getLogger(__name__)
2019-02-15 11:46:32 +00:00
def ignore_title(title):
ignore_starters = ["Help:"
, "Wikipedia:"
, "Template:"
, "Template_talk:"
, "Category:"
]
for ignore in ignore_starters:
if(title.startswith(ignore)):
return True
return False
2019-02-02 10:18:57 +00:00
def _receive_links(title, connection):
url = construct_url(title)
result = get_data_with_proxy(url, connection)
# This is basically because we don't know the page ID.
for k, page_data in result["query"]["pages"].items():
cursor = connection.cursor()
if(not "links" in page_data):
destination_title = page_data["title"].replace(" ", "_")
# avoid 1-loops
if(destination_title == title):
2019-02-15 11:46:32 +00:00
continue
if(ignore_title(title)):
continue
cursor.execute("INSERT INTO links(source, destination) VALUES(%s, %s)", (title, destination_title))
2019-02-02 10:18:57 +00:00
yield destination_title
else:
for destination in page_data["links"]:
2019-02-15 11:46:32 +00:00
if(ignore_title(title)):
continue
2019-02-02 10:18:57 +00:00
destination_title = destination["title"].replace(" ", "_")
2019-02-15 11:46:32 +00:00
try:
cursor.execute("INSERT INTO links(source, destination) VALUES(%s, %s)", (title, destination_title))
except Exception as e:
print(destination_title)
raise e
2019-02-02 10:18:57 +00:00
yield destination_title
connection.commit()
def receive_links(title, connection):
return list(_receive_links(title, connection))
def receive_link_graph(title, connection, depth):
if(depth < 0):
# end of recursion
return
cursor = connection.cursor()
2019-02-15 11:46:32 +00:00
print(repr(title))
cursor.execute("SELECT COUNT(source) FROM links WHERE source=%s", (title,))
2019-02-02 10:18:57 +00:00
if(cursor.fetchone()[0] != 0):
# we fetched that title already
return
logger.info("fetching links for {}".format(title))
for link in _receive_links(title, connection):
2019-02-02 15:06:57 +00:00
try:
receive_link_graph(link, connection, depth - 1)
except NoMoreProxiesException as e:
logger.exception("All proxies are blocked")
# Wikipedia blocked all our proxies.
# Retry later, so we have to store our list that is still to fetch.
cursor = connection.cursor()
2019-02-15 11:46:32 +00:00
cursor.execute("INSERT INTO failed_to_fetch(title, depth) VALUES(%s, %d)", (link, depth - 1))
2019-02-02 15:06:57 +00:00
connection.commit()
2019-02-02 10:18:57 +00:00