def _parseRestaurantPage(self, pool, region_name, area_name, restaurant_name, href):
utils.log("[%s] parsing restaurant '%s.%s.%s' (%s)" % (self, region_name, area_name, restaurant_name, href))
try:
soup = utils.getSoup(href)
except:
utils.log("[%s] error downloading page %s" % (self, href))
return
# parse the address for the current restaurant
addr = soup.find('span', {'class' : 'adr'})
street = addr.find('span', {'class' : 'street-address'}).getText().strip()
locality = addr.find('span', {'class' : 'locality'}).getText().strip()
region = addr.find('span', {'class' : 'region'}).getText().strip()
zipcode = addr.find('a', {'class' : re.compile('postal-code')}).getText().strip()
address = "%s, %s, %s %s" % (street, locality, region, zipcode)
# add the current restaurant to the output for this crawler
entity = Entity()
entity.subcategory = "restaurant"
entity.title = restaurant_name
entity.address = address
entity.sources.urbanspoon = {
'uurl' : href,
}
self._output.put(entity)
def _parseEntity(self, sheet, index, numEntities):
if numEntities > 100 and ((index - 1) % (numEntities / 100)) == 0:
utils.log("[%s] done parsing %s" % \
(self.NAME, utils.getStatusStr(index - 1 - Globals.options.offset, numEntities)))
time.sleep(0.1)
row = sheet.row_values(index)
entity = Entity()
entity.subcategory = "restaurant"
entity.title = row[1]
entity.address = row[3] + ', ' + \
row[4] + ', ' + \
row[5] + ' ' + \
row[6]
entity.openTable = {
'rid' : int(row[8]),
'reserveURL' : row[9],
'countryID' : row[10],
'metroName' : row[0],
'neighborhoodName' : row[2],
}
# don't make external calls to opentable in test mode
if not Globals.options.test:
result = OpenTableParser.parseEntity(entity)
if result is None:
return
if entity is not None:
#print entity.title
#from pprint import pprint
#pprint(entity.getDataAsDict())
self._output.put(entity)
def _parseEntity(self, result):
entity = Entity()
entity.subcategory = 'other'
if 'titleNoFormatting' in result:
entity.title = result['titleNoFormatting']
if 'addressLines' in result:
entity.address = string.joinfields(result['addressLines'], ', ')
entity.subtitle = entity.address
if 'lat' in result and 'lng' in result:
entity.lat = float(result['lat'])
entity.lng = float(result['lng'])
if 'region' in result:
entity.vicinity = result['region']
if 'phoneNumbers' in result:
phoneNumbers = result['phoneNumbers']
if len(phoneNumbers) > 0:
entity.phone = phoneNumbers[0]['number']
entity.googleLocal = {}
entity.titlel = entity.title.lower()
return entity
def _parseResultsPage(self, pool, href):
try:
soup = utils.getSoup(href)
except urllib2.HTTPError:
utils.log("[%s] error parsing page %s" % (self, href))
return
results = soup.find('div', { 'name' : 'LocationDirectory' }).findAll('h3')
for result in results:
try:
name = result.find('a').getText().strip()
except Exception:
continue
try:
raw_address = result.findNext('span', { 'class' : 'address'}).getText()
street = raw_address[0:raw_address.find('(')].strip()
locale = raw_address[raw_address.find(')')+1:raw_address.find('CA')+2].strip()
addr = '{0}, {1}'.format(street, locale)
except Exception:
addr = ''
utils.log("[%s] error parsing %s (%s)" % (self, addr, href))
continue
if addr == '':
continue
if name == '':
continue
if (name, addr) in self._seen:
continue
self._seen.add((name, addr))
entity = Entity()
entity.subcategory = "bar"
entity.title = name
entity.address = addr
entity.sources.sfweekly = { }
self._output.put(entity)
# try the next page
try:
pagination = soup.find('span', { 'class' : 'Pagination' }).getText()
if 'Next' in pagination:
pagination = soup.find('span', { 'class' : 'Pagination' })
href_get = pagination.find('span', { 'class' : 'PaginationSelected' }).findNext('a').get('href')
next_page = '{0}{1}'.format('http://www.sfweekly.com', href_get)
else:
next_page = ''
except Exception:
next_page = ''
if next_page != '':
pool.spawn(self._parseResultsPage, pool, next_page)
def _parseResultsPage(self, pool, href):
try:
soup = utils.getSoup(href)
except urllib2.HTTPError:
utils.log("[%s] error parsing page %s" % (self, href))
return
results = soup.find('div', { 'class' : 'split-right-column' }).findAll('div', { 'class' : 'clear' })
for result in results:
try:
name = result.findNext('div').find('h2').find('a').getText().strip()
except Exception:
continue
try:
street = result.findNext('div').find('address').getText()
locale = '{0}, {1}'.format('Los Angeles', 'CA')
addr = '{0}, {1}'.format(street, locale)
except Exception:
addr = ''
continue
if addr == '':
continue
if name == '':
continue
if (name, addr) in self._seen:
continue
self._seen.add((name, addr))
entity = Entity()
if 'Bars' in result.findNext('span').getText():
entity.subcategory = "bar"
else:
entity.subcategory = "restaurant"
entity.title = name
entity.address = addr
entity.sources.timeout_la = { }
self._output.put(entity)
# try the next page
try:
href_get = soup.find('div', { 'class' : 'next' }).find('a').get('href')
next_page = '{0}{1}'.format('http://www.timeout.com', href_get)
except Exception:
next_page = ''
if next_page != '':
pool.spawn(self._parseResultsPage, pool, next_page)
def _parseResultsPage(self, pool, href):
try:
soup = utils.getSoup(href)
except urllib2.HTTPError:
utils.log("[%s] error parsing page %s" % (self, href))
return
results = soup.findAll('h3')
for result in results:
try:
name = result.find('span', { 'style' : 'cursor:pointer;' }).getText().strip()
except AttributeError:
utils.log("[%s] error parsing %s (%s)" % (self, name, href))
return
try:
address1 = result.findNext('span', { 'class' : 'addresslinecaps' }).getText().strip()
if '(' in address1:
# sf mag does not provide any city, state or zip information,
# so inserting basic universal info manually.
addr = '{0}, {1}'.format(address1.split('(')[0].strip(), 'San Francisco, CA')
else:
addr = '{0}, {1}'.format(address1, 'San Francisco, CA')
except AttributeError:
utils.log("[%s] error parsing %s (%s)" % (self, addr, href))
return
entity = Entity()
entity.subcategory = "restaurant"
entity.title = name
entity.address = addr
entity.sources.sfmag = { }
self._output.put(entity)
#locate total pages and compare against current page num to determine if we should iterate again
try:
total_pages = soup.find('span', { 'class' : 'last' }).findPrevious('span').getText().strip()
except AttributeError:
# crawling of pages is done
return
index = href.find('&page=')
end = href.find('&keyword')
page = href[index+6:end]
if int(page) <= int(total_pages)-1:
next_page = href.replace('&page=' + str(page), '&page=' + str(int(page)+1))
pool.spawn(self._parseResultsPage, pool, next_page)
else:
return
time.sleep(0.01)
def _parseResultsPage(self, pool, href):
try:
soup = utils.getSoup(href)
except urllib2.HTTPError:
utils.log("[%s] error parsing page %s" % (self, href))
return
results = soup.find('div', { 'class' : 'search_results' }).findAll('div', { 'class' : 'restaurant'})
for result in results:
try:
name = result.find('h3').find('a').getText().strip()
except Exception:
continue
try:
street = result.find('br').previousSibling.strip()
locale = '{0}, {1}'.format(result.find('br').nextSibling.strip(), 'CA')
addr = '{0}, {1}'.format(street, locale)
except Exception:
addr = ''
utils.log("[%s] error parsing %s (%s)" % (self, addr, href))
continue
if addr == '':
continue
if name == '':
continue
if (name, addr) in self._seen:
continue
self._seen.add((name, addr))
entity = Entity()
entity.subcategory = "restaurant"
entity.title = name
entity.address = addr
entity.sources.sfgate = { }
self._output.put(entity)
# try the next page
try:
href_get = soup.find('li', { 'class' : 'next' }).find('a').get('href')
next_page = '{0}{1}'.format('http://www.sfgate.com', href_get)
except Exception:
next_page = ''
if next_page != '':
pool.spawn(self._parseResultsPage, pool, next_page)
def _parseRestaurantPage(self, pool, region_name, city_name, restaurant_name, href):
utils.log("[%s] parsing restaurant '%s.%s.%s' (%s)" % (self, region_name, city_name, restaurant_name, href))
try:
soup = utils.getSoup(href)
except:
utils.printException()
utils.log("[%s] error downloading page %s" % (self, href))
return
# parse the address for the current restaurant
addr = soup.find('div', {'class' : 'address'})
street = addr.find('span', {'class' : 'street'}).getText().strip()
geo = addr.find('span', {'class' : 'geo'}).getText().strip()
address = "%s, %s" % (street, geo)
# add the current restaurant to the output for this crawler
entity = Entity()
entity.subcategory = "restaurant"
entity.title = restaurant_name
entity.address = address
entity.sources.zagat = {
'zurl' : self.base + href,
}
#self._globals['soup'] = soup
# parse cuisine
header = soup.find('div', {'id' : "block-zagat_restaurants-14"})
if header is not None:
header = header.find('ul').find('li', {'class' : 'first'})
if header is not None:
entity.cuisine = header.getText()
# parse website
site = soup.find('span', {'class' : 'website'})
if site is not None:
site = site.find('a')
if site is not None:
entity.site = site.get('href')
# parse preview image
img = soup.find('div', {'id' : 'content'}).find('div', {'class' : 'photo'})
if img is not None:
img = img.find('img')
if img is not None:
entity.image = img.get('src')
self._output.put(entity)
def _parseRestaurantPage(self, pool, queue, url, name, base=False):
utils.log('[%s] parsing restaurant page %s (%s)' % (self, name, url))
try:
soup = utils.getSoup(url)
except:
#utils.printException()
utils.log("[%s] error downloading page %s (%s)" % (self, name, url))
return
content = soup.find('div', { 'id' : 'content'})
if content is None:
return
entity = Entity()
entity.title = content.find('h1').getText()
entity.subcategory = "restaurant"
entity.seattletimes = {}
details = content.find('div', {'id' : 'edbtext'})
desc = details.find('p').getText()
if desc is not None:
entity.desc = desc
details = details.findAll('p', {'class' : 'list'})
address = details[0].renderContents().strip().replace('<br />', '')
address = re.sub('[ \n\t]+', ' ', address)
entity.address = address
if len(details) > 1:
site = details[1].get('href')
if site is not None:
entity.site = site
if len(details) > 2:
hoursOfOperation = details[2].getText()
if hoursOfOperation is not None:
entity.hoursOfOperation = hoursOfOperation
key = (entity.title, entity.address)
if key in self.seen or '(closed)' in entity.title.lower():
return
self.seen.add(key)
self._output.put(entity)
def _parseDetailPage(self, name, href, subcategory):
try:
soup = utils.getSoup(href)
except urllib2.HTTPError:
utils.log("[%s] error parsing %s (%s)" % (self, name, href))
return
summ = soup.find('div', {'class' : 'summary-address'})
try:
addrp = summ.find('p', {'class' : 'adr'})
street_addr = addrp.find('span', {'class' : 'street-address'}).getText().strip()
locality = addrp.find('span', {'class' : 'locality'}).getText().strip()
region = addrp.find('span', {'class' : 'region'}).getText().strip()
try:
postal_code = addrp.find('span', {'class' : 'postal-code'}).getText().strip()
except AttributeError:
postal_code = ""
addr = "%s, %s, %s %s" % (street_addr, locality, region, postal_code)
except AttributeError:
try:
p = summ.find('p').getText()
r = re.compile('(.*)nr\. ', re.DOTALL)
m = r.match(p)
if m is None:
r = re.compile('(.*)at[. ]', re.DOTALL)
m = r.match(p)
addr = m.groups()[0].replace('\n', ' ').strip()
except AttributeError:
utils.log("[%s] error parsing %s (%s)" % (self, name, href))
return
entity = Entity()
entity.subcategory = subcategory
entity.title = name
entity.address = addr
entity.nymag = { }
self._output.put(entity)
def _parseEntity(self, row, count):
# utils.log("[%s] parsing entity %d" % (self, count))
name = row["name"].lower().strip()
collapsed = False
if name in to_collapse:
if to_collapse[name]:
self.numCollapsed += 1
return
to_collapse[name] = True
collapsed = True
# record how many times we've encountered each restaurant
# if not hasattr(self, 'seen'):
# self.seen = {}
# if name in self.seen:
# self.seen[name] += 1
# else:
# self.seen[name] = 1
entity = Entity()
entity.subcategory = "restaurant"
entity.factual = {"table": "US_Restaurants_V2.csv"}
if not collapsed:
address = FactualUtils.parseAddress(row)
if address is not None:
entity.address = address
for srcKey, destKey in self._map.iteritems():
if srcKey in row and row[srcKey]:
entity[destKey] = row[srcKey]
self._output.put(entity)
def _parseResultsPage(self, pool, href):
try:
soup = utils.getSoup(href)
except urllib2.HTTPError:
utils.log("[%s] error parsing page %s" % (self, href))
return
results = soup.find('div', { 'id' : 'searchResults' }).findAll('td', { 'class' : 'start' })
for result in results:
try:
name = result.find('a').getText().strip()
except AttributeError:
utils.log("[%s] error parsing %s (%s)" % (self, name, href))
return
x = 0
for r in result.findAll('br'):
x+=1
if x == 3:
try:
addr = '{0}, {1}'.format(result.find('a').nextSibling.strip(), result.find('br').nextSibling.strip())
except Exception:
utils.log("[%s] error parsing %s (%s)" % (self, addr, href))
return
elif x == 4:
try:
addr = '{0}, {1}'.format(result.contents[3].strip(), result.contents[5].strip())
except Exception:
utils.log("[%s] error parsing %s (%s)" % (self, addr, href))
return
else:
addr = ''
if addr == '':
continue
if 'CLOSED' in name:
continue
if (name, addr) in self._seen:
continue
self._seen.add((name, addr))
entity = Entity()
entity.subcategory = "restaurant"
entity.title = name
entity.address = addr
entity.sources.bostonmag = { }
self._output.put(entity)
# try the next page
next_page_ending = soup.find('div', { 'class' : 'right_align' }).findAll('a')
next_page = ''
for n in next_page_ending:
if 'Next' in str(n):
next_page = href.replace(href[href.find('?'):], n.get('href'))
else:
pass
if next_page != '':
pool.spawn(self._parseResultsPage, pool, next_page)
def _parseResultsPage(self, pool, href):
try:
soup = utils.getSoup(href)
except urllib2.HTTPError:
utils.log("[%s] error parsing page %s" % (self, href))
return
results = soup.find('ul', { 'id' : 'search_pagination' }).findAll('div', { 'class' : 'listing_item' })
for result in results:
try:
name = result.find('h2').getText().strip()
except AttributeError:
utils.log("[%s] error parsing %s (%s)" % (self, name, href))
return
try:
addr = result.find('span', { 'class' : 'address' }).getText().strip()
except AttributeError:
utils.log("[%s] error parsing %s (%s)" % (self, addr, href))
return
if addr == '':
continue
if 'CLOSED' in name:
continue
if addr in self._seen:
continue
self._seen.add(addr)
if name in self._count:
if self._count[name] < 3:
self._count[name] = self._count[name] + 1
else:
continue
else:
self._count[name] = 1
entity = Entity()
entity.subcategory = "restaurant"
entity.title = name
entity.address = addr
entity.sources.latimes = { }
self._output.put(entity)
#try the next page
try:
next_page = soup.find('a', {'class': 'next_page'}).get("href")
if next_page != '':
next_page_url = "{0}{1}".format('http://findlocal.latimes.com', next_page)
pool.spawn(self._parseResultsPage, pool, next_page_url)
except AttributeError:
# crawling of pages is done
#utils.log("Done crawling: %s" % href)
pass
def _parseResultsPage(self, pool, url, offset=0, base=False):
utils.log('[%s] parsing page %s' % (self, url))
max_offset = 8
if offset < max_offset:
# optimistically process the next results page before processing this one
if 'start=' in url:
start = self.start_re.match(url).groups()[0]
nexti = int(start) + self.results_per_page
url2 = url.replace('start=%s' % start, 'start=%d' % nexti)
else:
url2 = "%s&start=%d" % (url, self.results_per_page)
pool.spawn(self._parseResultsPage, pool, url2, offset + 1)
try:
soup = utils.getSoup(url)
except:
utils.printException()
utils.log("[%s] error downloading page %s" % (self, url))
return
if offset >= max_offset:
next_pagel = soup.find('a', {'id' : 'pager_page_next'})
if next_pagel is not None:
href = self.base + next_pagel.get('href')
pool.spawn(self._parseResultsPage, pool, href, 0)
time.sleep(0.01)
if base:
categories = soup.findAll('a', {'id' : self.category_re})
if categories is not None:
for category in categories:
href = self.base + category.get('href')
pool.spawn(self._parseResultsPage, pool, href, 0)
# yield so other threads have a chance to start working
time.sleep(0.01)
separator = '___'
results = soup.findAll('div', {'class' : re.compile('businessresult')})
if results is None:
return
for result in results:
entity = Entity()
entity.subcategory = 'restaurant'
entity.sources.yelp = { }
titlel = result.find('a')
title = titlel.getText()
entity.title = self.title_re.match(title).groups()[0]
entity.yurl = self.base + titlel.get('href')
addr = result.find('address').getText(separator)
match = self.address_re.match(addr).groups()
entity.address = "%s, %s" % (match[0], match[1])
entity.phone = match[2]
rating = result.find('img')
if rating is not None:
entity.yrating = float(self.rating_reviews_re.match(rating.get('title')).groups()[0])
reviews = result.find('span', {'class' : 'reviews'})
if reviews is not None:
entity.yreviews = int(self.rating_reviews_re.match(reviews.getText()).groups()[0])
key = (entity.title, entity.address)
if key not in self.seen:
self.seen.add(key)
self._output.put(entity)
def _parseResultsPage(self, pool, href):
try:
soup = utils.getSoup(href)
except urllib2.HTTPError:
utils.log("[%s] error parsing page %s" % (self, href))
return
results = soup.find('div', { 'class' : 'searchresults' }).findAll('div', { 'class' : 'fs1-sans' })
for result in results:
if 'Price' in result.getText():
continue
if 'Kid' in result.getText():
continue
if 'Other' in result.getText():
continue
if 'Wheelchair' in result.getText():
continue
if 'Cuisines' in result.getText():
continue
if 'Rating' in result.getText():
continue
if 'Latest' in result.getText():
continue
try:
name = result.find('strong').getText().strip()
except Exception:
continue
try:
addr = '{0} {1}, {2}, {3}'.format(result.find('span').getText(),
result.find('span').findNext('span').getText(),
result.find('span').findNext('span').findNext('span').getText(),
result.find('span').findNext('span').findNext('span').findNext('span').getText())
except Exception:
addr = ''
utils.log("[%s] error parsing %s (%s)" % (self, addr, href))
continue
if addr == '':
continue
if name == '':
continue
if (name, addr) in self._seen:
continue
self._seen.add((name, addr))
entity = Entity()
entity.subcategory = "restaurant"
entity.title = name
entity.address = addr
entity.sources.washmag = { }
self._output.put(entity)
return
def _parseResultsPage(self, pool, href):
try:
soup = utils.getSoup(href)
except urllib2.HTTPError:
utils.log("[%s] error parsing page %s" % (self, href))
return
results = soup.find('td', { 'id' : 'search-results' }).findAll('tr')
for result in results:
try:
name = result.find('td', { 'class' : 'business-name' }).find('a').getText().strip()
except Exception:
continue
try:
result.find('td', { 'class' : 'contact' }).find('br').previousSibling.strip()
result.find('td', { 'class' : 'contact' }).find('br').nextSibling.strip()
addr = '{0}, {1}'.format(result.find('td', { 'class' : 'contact' }).find('br').previousSibling.strip(), result.find('td', { 'class' : 'contact' }).find('br').nextSibling.strip())
except Exception:
addr = ''
utils.log("[%s] error parsing %s (%s)" % (self, addr, href))
continue
if 'OPENING SOON' in result.find('td', { 'class' : 'categories' }).getText():
continue
if addr == '':
continue
if name == '':
continue
if 'CLOSED' in name:
continue
if (name, addr) in self._seen:
continue
self._seen.add((name, addr))
entity = Entity()
entity.subcategory = "restaurant"
entity.title = name
entity.address = addr
entity.sources.chicagomag = { }
self._output.put(entity)
# try the next page
next_page_all= soup.find('div', { 'id' : 'pager' }).findAll('a')
next_page = ''
for n in next_page_all:
if 'Next' in n.getText():
next_page = n.get('href')
else:
pass
if next_page != '':
pool.spawn(self._parseResultsPage, pool, next_page)
