| johnnyloadproductions |
03-29-2016 11:53 PM |
Craigslist python scraper script inside
If anyone wants it.
print statements commented out for debugging.
Craigslist structure is the same so just change or add to append ["web","m4m","w4w"....]
Get the personals or whatever other category you want.
If you only have 1 ip don't hit them more than 10-15k times in one day or they'll ban the ip. There are ips services that can be purchased but that's up to you and it's a little harder for craigslist. Craigslist is one of the hardest sites to crawl with 3rd party ips.
You'll need PyMySQL, SQLAlchemy, BeautifulSoup4, and Python 2.7 for this to work. Mine runs every 2 days on a Raspberry Pi. :)
Code:
from sqlalchemy.ext.automap import automap_base
from sqlalchemy import create_engine
from sqlalchemy import MetaData
from sqlalchemy import Table, Column, Integer, Numeric, String, ForeignKey, text
from sqlalchemy import insert
Base = automap_base()
from bs4 import BeautifulSoup
import urllib2
import os
import datetime
import time
from random import randint
# Time the script starter
start = time.time()
def get_city_listings(url, city, dir):
# Start timing the speed of this script
"""
Main page filtering for the selected craigslist page, extract links that will then be followed.
"""
# - - - - - - - - - - Main Page Links
directory = os.path.dirname(os.path.realpath(__file__)) + "/" + city + "/"
if not os.path.exists(directory):
os.makedirs(directory)
# proxy_support = urllib2.ProxyHandler({"http":"159.203.245.77:80"})
# opener = urllib2.build_opener(proxy_support)
# urllib2.install_opener(opener)
time.sleep(randint(1,4) + randint(5,8) + randint(8,12))
array = os.listdir(directory)
print "This is the url: " + url
# This was just using a text file from the directory
# print directory + array[5]
response = urllib2.urlopen(url)
# file = open(directory + "text.txt", "r")
# print file
soup = BeautifulSoup(response.read(), "lxml")
# print soup
# tag = soup.find_all(string=re.compile('<div class='))
tag = soup.find_all("div", attrs={"class": "content"})
final_filter = BeautifulSoup(str(tag), "lxml")
# This further filters out all the links that lead to posts
links = final_filter.find_all("a", attrs={"class": "hdrlnk"})
# - - - - - - - - - - Postings scraper
#This is the link text description for each post
job_title = []
# The link to the posting
job_listings = []
for i in links:
print i.get_text()
value = i['href']
if value[0] == '/' and value[1] == '/':
pass
else:
job_title.append(i.get_text())
job_listings.append(i['href'])
# Gets actual listing (whatever it may be) from the extract URL
# def get_listing(post_url, city):
for title, listing in zip(job_title, job_listings):
# Build the url
url = "http://" + city + ".craigslist.com" + listing
print "This is the url to be read: " + url
time.sleep(randint(1,4))
engine = create_engine('mysql+pymysql://"username_goes_here":"passwordgoeshere'
'@104.131.175.187/craigslist')
duplicate_link = False
listed_url = url
check_duplicate = engine.execute("SELECT post_url FROM %s" % dir)
for i in check_duplicate:
if url == i['post_url']:
# print "This has a duplicate url"
# print i['post_url']
duplicate_link = True
break
else:
duplicate_link = False
if not duplicate_link:
print check_duplicate
response = urllib2.urlopen(url)
# print "this is the url we are using: " + post_url
# print "directory: " + os.path.dirname(os.path.realpath(__file__))
page = BeautifulSoup(response.read(), "lxml")
# info on what the job or post is about
post_body = page.find_all("section", attrs={"id": "postingbody"})
# pulls out the post text from the content post
extracted_text = BeautifulSoup(str(post_body), "lxml")
# print extracted_text.get_text()
string = extracted_text.get_text()
string = string.strip('[\\n')
string = string.strip('\\n]')
# strips the post entry and is entirely text
print "This is the url: " + url
print "This is the title: " + title
print "This is the descriptino: " + string
# - - - - - - - - - - All database logic
# session = Session(engine)
# Base.prepare(engine, reflect=True)
# mapped classes are now created with names by default
# matching that of the table name.
# s = select(['posts'])
metadata = MetaData()
craigslist = Table(dir, metadata,
Column('post_id', Integer(), primary_key=True),
Column('post_date', String(20)),
Column('post_name', String(255), index=True),
Column('post_url', String(255)),
Column('post_content', String(20000))
)
metadata.create_all(engine)
# Make sure to add items here that were parsed
ins = insert(craigslist).values(
# post_id is auto inserted
post_date=datetime.date.today(),
post_name=title,
post_url=url,
post_content=string,
)
# insert into database the parsed logic
engine.execute(ins)
# - - - - - - - - - - -
# prepend = ["ames"]
prepend = ["auburn", "birmingham", "dothan", "shoals", "gadsden", "mobile", "montgomery", "tuscaloosa", "huntsville", "fairbanks", "anchorage", "flagstaff", "phoenix", "prescott",
"tucson", "yuma", "fayetteville", "jonesboro", "texarkana", "bakersfield", "chico", "merced", "modesto", "redding", "sacramento", "stockton", "susanville", "boulder", "denver", "pueblo",
"hartford", "delaware", "washington", "losangeles", "monterey", "orangecounty", "palmsprings", "sandiego", "sfbay", "slo", "santabarbara", "santamaria", "siskiyou", "ventura", "visalia", "yubasutter",
"humboldt", "imperial", "inlandempire", "rockies", "westslope", "newlondon", "newhaven", "showlow", "sierravista", "fortsmith", "littlerock", "goldcountry", "hanford", "nwct", "mohave", "juneau",
"cosprings", "eastco", "fortcollins", "kenai", "gainesville", "jacksonville", "lakeland", "ocala", "orlando", "pensacola", "tallahassee", "panamacity", "sarasota", "miami", "spacecoast", "staugustine", "tampa", "treasure",
"miami", "fortmyers", "cfl", "lakecity", "okaloosa", "keys", "daytona", "albany", "athens", "atlanta", "augusta", "brunswick", "columbus", "statesboro", "valdosta", "macon", "nwga", "savannah",
"hawaii", "boise", "lasalle", "mattoon", "peoria", "rockford", "carbondale", "decatur", "springfieldil", "quincy", "chicago", "bn", "twinfalls", "lewiston", "eastidaho", "bloomington", "evansville", "indianapolis", "kokomo", "muncie",
"tippecanoe", "richmondin", "southbend", "terrehaute", "fortwayne", "ames", "dubuque", "iowacity", "masoncity", "quadcities", "siouxcity", "ottumwa", "waterloo", "fortdodge", "desmoines", "cedarrapids", "lawrence",
"salina", "seks", "swks", "topeka", "wichita", "lexington", "houma", "lafayette", "monroe", "shreveport", "maine", "annapolis", "baltimore", "frederick", "louisville", "owensboro", "westky", "batonrouge", "cenla", "lakecharles",
"neworleans", "easternshore", "smd", "westmd", "eastky", "albany", "binghamton", "buffalo", "catskills", "chautauqua", "newyork", "oneonta", "plattsburgh", "longisland", "potsdam", "rochester", "syracuse", "twintiers", "utica", "watertown",
"elmira", "fingerlakes", "glensfalls", "hudsonvalley", "ithaca", "jxn", "kalamazoo", "lansing", "monroemi", "muskegon", "nmi", "porthuron", "saginaw", "swmi", "thumb", "battlecreek", "centralmich", "detroit", "flint", "grandrapids",
"holland", "up", "annarbor", "laredo", "mcallen", "odessa", "sanangelo", "sanantonio", "charlottesville", "danville", "fredericksburg", "harrisonburg", "richmond", "roanoke", "winchester", "charleston", "morgantown", "swva",
"martinsburg", "huntington", "wheeling", "parkersburg", "swv", "wv", "blacksburg", "lynchburg", "norfolk", "sanmarcos", "corpuschristi",
"dallas", "nacogdoches", "delrio", "elpaso", "galveston", "houston", "killeen", "bigbend", "texoma", "easttexas", "amarillo", "austin", "beaumont", "brownsville", "collegestation", "victoriatx", "wichitafalls", "abilene"]
append = ['web',]
for dir in append:
for city in prepend:
print city
time.sleep(5.5)
# ensure_dir(city)
url = "http://{}.craigslist.com/search/{}".format(city, dir)
print "This is the directory: " + dir
get_city_listings(url, city, dir)
end = time.time()
print(end - start)
|
